COLUMBIA UNIVERSITY 
 EDWARD G. JANEWAY 
 MEMORIAL LIBRARY 
 
GLYCOSURIA 
 
 AND ALLIED CONDITIONS 
 
Digitized by the Internet Archive 
 
 in 2010 with funding from 
 
 Open Knowledge Commons 
 
 http://www.archive.org/details/glycosuriaalliedOOcamm 
 
GLYCOSURIA 
 
 AND ALLIED CONDITIONS 
 
 BY 
 
 P. J. CAMMIDGE, M.D. (Lond.) 
 
 x\EW YORK 
 LONGMANS, GREEN AND CO. 
 
 LONDON: EDWARD ARNOLD 
 1913 
 

 CU-v. 
 
 1^- ^o ^ b^ 
 
PREFACE 
 
 The work on diseases of the pancreas which I commenced some 
 fourteen or fifteen years ago made it necessary for me to famiharise 
 myseK with much that had been written on cUabetes and alhed 
 conditions. Since then I have kept in touch with the htera- 
 ture, and also carried out many researches bearing on these sub- 
 jects, while an increasing number of cases of pancreatic disease and 
 diabetes having been sent to me for observation and treatment, 
 I have had considerable opportunity of extending my chnical know- 
 ledge of the various types of glycosuria. At the request of several 
 friends in the profession I have eventually consented to collect in 
 an accessible form the conclusions I have arrived at, and at the 
 same time summarised the work of others. 
 
 Glycosuria is essentially a chemical problem, and it therefore 
 seemed to me advisable that it should be attacked from a chemical 
 standpoint, leading from this to its pathology, symptomatology, 
 diagnosis, and treatment. I have consequently first dealt with the 
 tests, differentiation, and quantitative estimation of the reducing 
 substances met with in the urine, prefacing this by a brief summary 
 of the chemistry and physiology of the carbohydrates and their 
 derivatives. For those who are interested in the chemical questions 
 involved, I have considered this part of the subject more fully in 
 the Appendix. 
 
 The experimental production of glycosuria in animals has natu- 
 rally received attention before the pathology of human diabetes 
 has been considered. In this connection I have dealt at consider- 
 able length with the relation of the ductless glands, and particularly 
 the pancreas, to glycosuria, but, I hope, not more fully than the 
 importance of the subject warrants. 
 
 Although for convenience' sake the simpler forms of alimentary, 
 transitory, and intermittent glycosuria have been dealt with 
 separately, it is important that it should be reahsed that they pass 
 by a gradual transition, with no well-defined boundary, into per- 
 sistent glycosuria, and this in its turn into typical " diabetes." 
 
 As yet the treatment of glycosuria is primarily dietetic, and it is 
 therefore important that the metabolism of the healthy organism, 
 and the variations that occur in association with the presence of 
 sugar in the urine should be thoroughly understood, for this reason 
 they have been written of in some detaiL Since in my opinion it 
 is not enough that a diabetic should be given a list of foods that he 
 " may take " and " must avoid," I have outhned a system that I 
 have adopted by which the diet is worked out, not oiily as regards 
 
GLYCOSURIA 
 
 AND ALLIED CONDITIONS 
 
 CHAPTER I 
 
 CLASSIFICATION, PROPERTIES, AND PHYSIOLOGY OF THE 
 CARBOHYDRATES AND THEIR DERIVATIVES 
 
 Strictly speaking, the term " glycosuria," or " glucosuria," 
 should be used only to describe the existence of an abnormal 
 amount of glucose or dextrose in the urine, but it has for so long 
 been employed to designate conditions in Avhich an excess of any 
 sugar is met with that it is convenient to retain it as a generic 
 term with that significance, and to employ the names " dextros- 
 uria," " levulosuria," " pentosuria," &c., when speaking of con- 
 ditions in which dextrose, levulose, pentoses, and other reducing 
 substances are alone present. In recent years German medical 
 writers have made use of the term " mellituria " in speaking of the 
 excretion of sugar in the urine generally. Although it has occa- 
 sionally been used by English authors in this sense, it has also 
 sometimes been employed as synonymous with saccharosuria, the 
 condition when cane-sugar is present in the urine, and is therefore 
 not without objection. 
 
 The accumulated experience of over one hundred years, since 
 Dobson of Liverpool first obtained sugar from diabetic urines, in 
 1776, has shown that there are few symptoms that are associated 
 with so many distinct and widely different pathological conditions 
 as glycosuria. The discovery that the reducing substance that 
 occurs in the urine is not always dextrose, and may not even be 
 a sugar, made it necessary to revise some of the earlier work that 
 had been done on glycosuria ; but, although it is now certain that 
 a reduction previously ascribed to glucose is in some instances 
 due to the presence of other bodies, the list of conditions in 
 which dextrosuria may occur is still a lengthy one, and includes 
 the folloAA ing : — 
 
 Diseases of the ]oancreas, liver, thja'oid, pituitary gland, adre- 
 nals, kidneys ; cholelithiasis, intestinal disorders (enteritis, colitis, 
 
2 GLYCOSURIA 
 
 corrosive and food poisoning), chyluria, diseases of the nervous 
 system, j)sychic conditions (worry, shpck, mental strain), pregnancy, 
 tumonrs of the uterus and ovaries, inanition (vagabond's glycos- 
 uria, &c.), asphyxia, cold immersion (attempted drowning), acute 
 fevers (pneumonia, scarlet fever, measles, mumps, variola, malaria, 
 acute rheumatism, phlegmonous diseases), the administration of 
 thjToid extract, adrenalin, and drugs (atropine, morphine, strych- 
 nine, curare, amyl nitrate, copaiba, phosphorus, perchloride of 
 mercury, uranium salts, phloridzin, acetone, chloroform, ether, 
 nitrous ether, &c.), coal-gas, and carbon monoxide, poisoning, and 
 alcoholic excess (especially champagne and beer). 
 
 Some of these are exceedingly vrare, and have only been re- 
 ported in one or two cases, while others are comparatively common ; 
 but the difficulty arises that glycosuria is not a constant symptom 
 of any one of the conditions mentioned. The problem of the 
 essential cause of glycosuria was still further complicated when 
 it was shoAvn by experiment on animals that sugar may be made 
 to appear in the urine as the result of a variety of different pro- 
 cedures. A discussion of glycosuria and allied conditions covers, 
 therefore, a very wide field. 
 
 Before considering in detail the different varieties of glycos- 
 uria, the symptoms with which they are associated and the con- 
 ditions with which they may be confounded, it will be convenient 
 to first deal briefly with the chemistry of the commoner carbo- 
 hydrates, the changes they undergo during digestion and assimila- 
 tion in the body, the means by which the sugars are recognised, 
 differentiated, and estimated in the urine, and the bearing of 
 modern experimental work on the production of glycosuria. 
 
 Classification, Composition, Configuration, and Properties 
 OF THE Carbohydrates 
 
 The carbohydrates constitute an ill-defined group of sub- 
 stances differing widely in their properties and constitution, so 
 that it is difficult to give a satisfactory definition. They may be 
 roughly defined as bodies composed of carbon, hydrogen, and 
 oxygen in which the ratio of hydrogen to oxygen is the same as 
 in water. This definition includes, however, bodies such as inosite, 
 lactic acid, &c., which are not regarded as carbohydrates. The 
 name carbohydrate was originally given to the group because its 
 constituents may be represented as if they were composed of 
 carbon and water in different proportions — e.g. CgfHgO)^,, Q,-^^(KS))^,^, 
 with a general formula C„(H20)n, in which " n " is a variable quan- 
 
THE CARBOHYDRATES 3 
 
 tity, but in reality they are much more complex. The group in- 
 cludes all the principal constituents of plants, except water, and, 
 with fat and albumens, the carbohydrates form the chief substances 
 necessary for animal life. 
 
 The carbohydrates are divisible into three main groups : — 
 
 1. The simple sugars, or monosaccharides, or saccharoses. 
 
 2. The invertible sugars, or clisaccharicles. 
 
 3. The colloidal, non-crystallisable, polysaccharides or polj^oses. 
 
 The Monosaccharides. — The naturally occurring monosacchar- 
 ides are colourless, odourless, crystalline substances which, in a 
 pure state, are not hygroscopic, but are easily soluble in water, 
 feebly soluble in alcohol, and insoluble in ether. They diffuse 
 through animal membrane. Their aqueous solutions are neutral 
 in reaction, and have a sweet taste, varying in intensity with 
 the kind of sugar. On being boiled with dilute acids they are not 
 resolved into simpler sugars. 
 
 The monosaccharides have the general formula C„HonO„, and 
 may be conveniently subdivided, according to the number of 
 carbon atoms they contain, into — 
 
 Trioses (CgHgOg) containing 3 carbon atoms {e.g. glyceric aldehyde) 
 Tetroses (C4H8O4) „ 4 „ „ {e.g. erythrose) 
 
 Pentoses (CgHjfjOj) „ 5 „ ,, {e.g. arabinose) 
 Hexoses (OgHj20(;) „ 6 „ „ {e.g. glucose) 
 Heptoses (7 carbon atoms)^ octoses (8 carbon atoms), &c. 
 
 Many of these sugars have only been prepared artificially, and 
 are merely of theoretical interest. The six, and to a less extent 
 the five, carbon atom sugars are those of chief importance in the 
 animal economy. 
 
 The Hexoses. — The hexoses are represented by the molecular 
 formula V^^Jd^. The most important are dextrose (glucose, or 
 grape-sugar), levulose (fructose, or fruit-sugar), galactose, and 
 mannose. Some of the hexoses, such as dextrose and levulose, are 
 found free in nature, or result as hydrolytic decomposition jDro- 
 ducts from the more complex carbohydrates or related nitrogenous 
 substances, the so-called glucQsides. For long the hexoses were 
 regarded as consisting of a simple chain of six carbon atoms bound 
 to each other by a single valency, the remaining valencies of five 
 being satisfied by hydrogen and hydroxyl groups, while the sixth 
 was joined to an oxygen atom by a double bond >C-0. It was 
 
4 GLYCOSURIA 
 
 believed that the carbonyl group (>C-0) might be situated at the 
 end of the chain, as, for example, in dextrose — 
 
 H HO HO H HO 
 
 I I I i i 
 HO— C— C— 0— C— C— C— H 
 
 H H H OH H O 
 
 or might lie between two other carbon atoms, as in levulose — 
 
 H H H OH H 
 
 I 1 I I I 
 
 HO— C— C— C— C— C— C- OH 
 
 I I I I II I 
 HOHOHH H 
 
 Sugars which were thought to have the former structure are 
 known as aldoses, because they contain the aldehyde group 
 ( - CHO) ; those with the latter as ketoses, since they contain the 
 divalent ketonic or carbcnyl group (=C0). The chemical activities 
 of the sugars are not, however, as marked as might be expected 
 from this simple chain formula, and more recently it has been 
 proposed to represent them by a formula in which four of the 
 carbon atoms, together with one oxygen atom, are included in 
 a ring : — 
 
 H H HO H 
 
 II II 
 
 g^yC/ OHHO NcH.CH(OH)CHa.OH ^^^^h^^^C<^ H HO NcH.CHa.OH 
 
 \o/ \o/ 
 
 Dextrose ' Levulose 
 
 According to this view the aldehydic and ketonic characters 
 exhibited by the sugars are developed on rupture of the ring by 
 hydrolysis, with the formation of an open chain such as the older 
 formulae represented. The ring formula has now been very widely 
 accepted, as it is more in accordance with the chemical properties 
 of the sugars. This formula, which would allow of there being 
 sterio-isomeric forms, also accounts for a peculiar physical property 
 of solutions of dextrose and other sugars having aldehyde pro- 
 perties, known as mutorotation or multirotation. 
 
 The Pentoses are represented by the molecular formula C'^H^oOg. 
 They are widely distributed in the vegetable kingdom as poly- 
 saccharides of high molecular weight, the " pentosans," and are 
 never found as simple sugars. They occur in most fruits, par- 
 ticularly in cherries, apples, pears, and plums, and to some extent 
 
THE CARBOHYDRATES 5 
 
 in corn and other vegetable tissues. In the animal body pentoses 
 are an important constituent of the nucleo-proteins and nucleic 
 acids, being most abundant in the pancreas. According to Griind, 
 the percentage of pentose in the dry weight of the pancreas is 
 nearly five times as great as in any other organ of the body (pan- 
 creas 2-48%, liver 0-56%, thymus 0-56%, kidney 0-49%, muscle 
 0-11%). 
 
 The natural jDentoses (arabinose and xylose) are closely related 
 structurally to the natural hexoses. The arrangement of the 
 groups attached to the first four carbon atoms is the same in 
 arabinose as in galactose, and in xylose as in glucose. In this con- 
 nection it is interesting to note that both xylose and glucose are 
 yielded by some polysaccharides on hydrolysis, and that arabinose 
 and galactose occur together in some gums (Armstrong). 
 
 The chemical characters of the monosaccharides are dependent 
 j)artly upon the hydroxyl groups, and partly upon the carbonyl 
 group, they contain. Through the presence of the hydroxyl 
 groups they are capable of forming esters, or etheral salts, the 
 best known of which is henzoyl-ester, which is sometimes used for 
 their separation and recognition. Owing to the presence of the 
 carbonyl group they are easily oxidised, and so reduce alkaline 
 solutions of the heavy metals. With an ammoniacal silver solution 
 they give a metallic mirror of silver, and reduce alkaline solutions 
 of copper, bismuth, and other metallic salts to oxides and hydroxides 
 of the metal. On this property is based various tests, such as 
 Tiommer's, Fehling's, and Bottger's, that are commonly employed 
 for their detection. 
 
 The monosaccharide sugars are not precipitated by lead acetate, 
 or sub-acetate, but separate out on making the solution alkaline 
 with ammonia. On being heated they char and form a brown sub- 
 stance, soluble in water, known as caramel. When heated ^^ith 
 an alkali, solutions of these sugars turn brown and the sugars 
 are decomposed, forming a variety of substances, including lactic 
 acid, formic acid, and various aldehydes (Moore's test). On being 
 treated with strong acids they break down, yielding furfurol, which 
 can be recognised by the colour reaction it gives with alpha -naphthol, 
 &c. (Molisch's test). The ease with which furfurol is liberated 
 varies with the different sugars, the readiness with which it is 
 evolved from the pentoses forming the basis of several of the more 
 characteristic tests for this group (Phlorogiucin test, Orcin test). 
 With phenylhyclrazin, these sugars form characteristic compounds 
 which serve to demonstrate their presence in very dilute solutions, 
 and, to a certain extent, to differentiate the various monosaccharides 
 
6 GLYCOSURIA 
 
 from each other. Asymmetrical substituted hyclrazins of the type 
 NH0.NR.CgH5, such as methyl-phenylhyclrazin, di-phenylhydra- 
 zine, para-brom-phenylhydrazin, also react wdth the monosac- 
 charides, and, in some cases, form sparingly soluble com]3ounds 
 Avhich are characteristic of a particular sugar, and are therefore of 
 great ser^^ice in their recognition. 
 
 Dextrose and levulose are fermented by yeast, yielding carbon 
 dioxide and alcohol. Galactose is fermented with much greater 
 diflicultj^ and many varieties of yeast do not act upon it at all. 
 The pentoses are unfermentable, but are attacked and slowly 
 broken down by bacteria. 
 
 Like other substances containing an asymmetrical carbon atom, 
 the monosaccharides possess the power of rotating the plane of 
 polarisation of a luminous ray, the exact space relation of the 
 h3^droxyl groups relative to the skeleton chain of carbon atoms 
 determining whether the ray shall be deflected to the right or to 
 the left — that is to say, whether sugars with the same gross struc- 
 ture shall be dextro-rotatory or levo-rotatory. The power of 
 rotating polarised light possessed by a particular sugar is, under 
 certain circumstances, a fixed quantity known as its " specific 
 rotation," and, as this property is also exerted by solutions of the 
 sugars, the angle through which rotation occurs serves for their 
 accurate estimation. 
 
 The most reliable observations of the specific rotatory power 
 of the monosaccharides in solutions containing about 10 per cent, 
 are as follows : — 
 
 Hcxoses — d-dextrose 
 
 . + 52-7° 
 
 d-levulose 
 
 . - 93-8° 
 
 d-galactose 
 
 . + 81-0° 
 
 d-inannose 
 
 . + 14-05 
 
 Pentoses — 1-xylose . 
 
 . + 18-09 
 
 1-arabinose . 
 
 . +105-1" 
 
 The action of a particular sugar on polarised light is indicated 
 by the prefixes d- and 1- ; but, owing to a convention by which 
 this prefix is also attached to its derivatives, it comes about that 
 sugars that are actually levo-rotatory may be designated as 
 d-sugars {e.g. d-levulose), or the reverse (e.g. 1-xylose). A mixture 
 of equal parts of dextro-rotatory and levo-rotatory sugars is 
 optically inactive, and this is sho^\Ti by the prefix i- (inactive) or 
 r- (racemic). 
 
 The Disaccharides are anhydrides, or ether-like derivatives, of 
 the simple monosaccharides. They contain twelve carbon atoms. 
 
THE CARBOHYDRATES 7 
 
 and consist of two simple six-carbon atom residues united through 
 an oxygen atom. They are therefore analogous to the simple 
 glucosides. When acted on by hydrolytic agents, such as dilute 
 acids or enzymes, they break down, with the addition of a molecule 
 of water, into their constituent hexoses, which may be either 
 aldoses or ketoses — 
 
 C12H22O11 + H,0 = C(iHi20o+ CoHioOe- 
 
 Some of the disaccharides occur in nature as such, but others 
 result from the decomposition of still more complex carbohydrates. 
 The more important members of the group are cane-sugar (sac- 
 charose, or sucrose), lactose (or milk-sugar), maltose (or malt- 
 sugar), and isomaltose. 
 
 In their general properties the disaccharides closely resemble 
 the monosaccharides. Like these they have a sweet taste, are 
 crystallisable, are capable of passing through animal membranes, 
 and are optically active. The specific rotation of solutions con- 
 taining about 10 per cent, are as follows : — 
 
 Sucrose . . . . . -f 66"5^ 
 
 Maltose -h 138-0° 
 
 Lactose -t- 52*5° 
 
 The disaccharides as such are not fermentable. A solution of 
 cane-sugar or maltose will, however, undergo alcoholic fermenta- 
 tion when exposed to the action of yeast, but this is due to the 
 existence in the yeast of specific ferments, kno'v^ii as '"invertase" 
 and " maltase " respectively, which have the power of hydrolysing 
 the disaccharides into their constituent monosaccharides, which 
 are then attacked and fermented. Ordinary yeast does not contain 
 the ferment " lactase," which has the power of hydrolysing lactose, 
 hence milk sugar is not fermented, although it may be slowly 
 broken down into lactic acid and butyric acid by contaminating 
 organisms. 
 
 The chemical characters of the disaccharides vary according to 
 the way in which the constituent hexoses are bound together. 
 In all of them the properties of the aldehyde group of one of the 
 hexoses is masked, owing to the second being attached to it in 
 place of an hydrogen atom in the hydroxyl group combined with 
 the carbon atom, which exercises aldehydic functions in the open 
 chain form. The aldehj^dic, or ketonic, group of the second 
 hexose may either remain functional or cUsappear. In the one 
 case the cUsaccharide behaves like a monosaccharide, reducing 
 
8 GLYCOSURIA 
 
 salts of the heavy metals, formmg osazones, &c., and in the other 
 these properties are lost. The disaccharicles may accordingly be 
 divided into two groups : — 
 
 I. Reducing Disaccharicles. 
 
 Maltose = dextrose + dextrose 
 
 Isomaltose = dextrose + dextrose 
 
 Lactose. ..... =dextrose+ galactose 
 
 Isolactose = dextrose + galactose 
 
 II. Non-reducing Disaccliaricles. 
 Saccharose (cane-sugar) = dextrose + levulose (invert sugar) 
 
 The Polysaccharides, like the disaccharides, are to be regarded 
 as condensation products of the monosaccharides. They have a 
 common empirical formula (CgHjoOg),,, in which '^ n " is a variable 
 factor that always exceeds two. In many cases the value of 
 " n " is unknown, but it is probably always large : in starch, for 
 example, it is 108. The polysaccharides are a very numerous 
 class, and, although chiefly met with in the vegetable kingdom, 
 are also found in the animal body. They may be conveniently 
 di\ided into the following groups : — 
 
 1. The starch group (starch, inulin, glycogen). 
 
 2. The dextrins. 
 
 3. The cellulose group (cellulose, hemicellulose, tunicin). 
 
 4. The gum group (plant gums, mucilage, animal gums). 
 
 They are non-volatile and, with very few exceptions, are 
 amorphous. As a class they are insoluble in alcohol, but usually 
 cUssolve in water to form solutions, which are often opalescent and 
 exert a marked rotatory effect on polarised light. As a rule they 
 do not diffuse through animal membrane. Their solutions are 
 not sweet to the taste, are neutral in reaction, and yield the poly- 
 saccharide in the form of a x^recipitate on treatment with certain 
 neutral salts (e.g. ammonium sul]3hate). On hydrolysis mono- 
 saccharides appear among other products. With the exception 
 of the dextrins they do not reduce metallic oxides in alkaline 
 solution, and none of them combine with phenylhydrazin to form 
 osazone. They are not directly fermented by yeast, but, like the 
 cUsaccharides, they may be hydrolised by the action of ferments 
 or acids to monosaccharides, which can be fermented. 
 
 Many of the polysaccharides combine with iodine to form 
 characteristic coloured compounds. Owing to their physical 
 characters and feeble chemical affinities they are often difficult to 
 obtain in a state of purity. 
 
THE CARBOHYDRATES 
 
 Acids and Acid-derivations of the Sugar Series 
 
 Certain acids, and acid derivatives, of the fatty series are related 
 to the carbohydrates, and since some of these are natural products 
 of the chemistry of the body, while others make their appearance 
 in conditions where carbohydrate metabolism is interfered with, 
 it is essential that their structure and relations should be clearly 
 understood before the fate of the sugars under normal and patho- 
 logical conditions is considered. 
 
 Acids having the general formula C,iH2,,02 are known as fatty 
 acids, since the higher members of the series occur in natural fats 
 (e.g. palmitic acid C^gHggOg, stearic acid C^^gHggOg). The two oxygen 
 atoms, one of the carbon, and one of the hydrogen atoms are 
 always in the combination -COOH, which is known as a " carboxyl " 
 radical, and it is to the presence of this that the acid properties of 
 the compound are due. The lowest member of the series is formic 
 acid (H.COOH) ; the next is acetic acid (CH3.COOH) ; then comes 
 propionic acid (CH3.CH2.COOH), then butyric acid (CHg.CH.^. 
 CHg.COOH), and so on up the series, every member counting one 
 more -CH2 group than its predecessor. If, by a process of oxida- 
 tion, an oxygen atom is introduced into the molecule, we have 
 formed an oxy-acid. Thus from acetic acid (CH3.COOH) is derived 
 hydroxy-acetic, or glycollic, acid (CH-j.OH.COOH), which is the 
 lowest member of the series of acids of the sugar group ; and 
 from propionic acid we get oxy-propionic, or lactic, acid 
 (CH3.CH(0H).C00H). When the acid contains several CH groups, 
 each division of the molecule is named according to its relation to 
 the fundamental unchanged carboxyl group, the one nearest the 
 carboxyl radical being said to be in the a-position, the next on the 
 left in the /5-position, and the next to it in the y-position. From 
 any one acid we can therefore have produced a series of oxy-acids 
 differing in the position of the OH grou]3, and known as a-,- /?-, 
 7-oxy-acids respectively. Thus from butyric acid (CHg.CH.^.CHg. 
 COOH) there can be theoretically derived three such acids : — 
 
 a— CH3.CH2.0H(OH)OOOH 
 ;8— 0H3.CH(OH).CH.,.COOH 
 y— CHo.(OH).CH..CHo.COOH 
 
 In the body oxidation appears to take place most readily in 
 the /3-position, and hence in the above series, for instance, it is 
 /J-hydroxybutjTic acid that is of most physiological importance. 
 
 In the laboratory the whole series of oxy-acids may be ob- 
 
10 GLYCOSURIA 
 
 tained by suitable means, and, moreover, more than one group 
 niay undergo the change — e.g. glyceric acid (CH2.(0H)CH(0H) 
 C.OOH), gluconic, mannonic, or galactonic acid (CH2.CH(CH.0H)^. 
 COOH). If /3-oxybutyric acid undergoes further oxidation, and 
 this takes place in the /3-position, as it is supposed to do in the 
 body, aceto-acetic, or di-acetic, acid and water result : — 
 
 CH3.CH(OH).CH2.COOH + O = CH3.CO.CH2.COOH + H2O 
 /5-oxybutyric acid aceto-acetic acid water 
 
 By further oxidation, and the siolitting off of a molecule of carbon 
 dioxide, it is joossible to derive acetone from aceto-acetic acid — 
 
 CH3.CO.CH2.COOH = CH3.CO.CH3+ CO2 
 aceto-acetic acid acetone carbon dioxide 
 
 A further series of acids can be derived from the simple fatty 
 acids by the substitution of a carboxyl group for one of the hj^drogen 
 atoms of the terminal -CH group ; thus carboxyl-acetic, or malonic, 
 acid (COOH.CH2.COOH) may be regarded as derived from acetic 
 acid (CHg.COOH) by such a substitution. Another important 
 member of this series is oxalic acid (COOH.COOH). These acids, 
 since they contain two carboxyl grou^^s, are dibasic, while those 
 pre^dously considered, which contained only one carboxyl group, 
 are monobasic. Oxidation of the dibasic, as of the monobasic 
 acids, gives rise to a series of oxy-acids ; thus tartronic acid 
 [COOH(CH.OH)COOH] is the oxy-acid of malonic acid (COOH. 
 CHG.COOH). Other important cli-basic oxy-acids are tartaric 
 [COOH.(CH.OH),.COOH], and saccharic and mucic acids [COOH 
 (CH.0H)4.C00H]. 
 
 An important oxidation jD^oduct of the sugar series, inter- 
 mediate between the mono- and di-basic acids, and possessing 
 both acid and aldehydic properties, is glucuronic, or glycuronic 
 acid [CH0(CH.0H)4.C00H]. 
 
 (A more detailed account of the properties of the carbohydrates, 
 and the acids, and acid derivatives of the sugar series, is given in 
 the Appendix.) 
 
 The Digestion and Assimilation of Carbohydrates 
 
 Onlj^ a small proportion of the carbohydrates of the food are 
 in a form fitted for immediate absorption, the greater part con- 
 sisting of starches, sugars, &c., which must undergo cleavage and 
 hydrol\i;ic changes before they can be taken up by the walls of 
 the intestinal tract. These changes are brought about by the 
 
THE CARBOHYDRATES 11 
 
 action of ferments found in the saliva, pancreatic secretion, and 
 intestinal juices. 
 
 Two distinct classes of ferments can be recognised — (1) those 
 known as " amylolytic " or " diastatic " ferments, which act upon 
 starches, producing sugars and dextrins ; and (2) those which 
 act upon various saccharoses, giving rise to glucose, called " in- 
 verting " ferments. The two chief amylolytic ferments are the 
 ptyalin of the saliva and the amylopsin of the pancreatic juice, 
 while the inverting ferment is found in the mucous membrane of 
 the small intestine and in the succus entericus. 
 
 The starches contained in the food have usually been broken 
 up and partly converted into dextrins by cooking before they 
 are consumed. Their digestion is commenced in the mouth, 
 soluble starch, certain forms of dextrin, maltose, and traces of 
 dextrose resulting from the action of the ptyalin of the saliva. 
 But digestion is not carried very far here, especially when the 
 food is not well masticated. In man salivary digestion pl^-ys quite 
 a secondary role. When the starchy foods reach the stomach the 
 diastatic fermentation initiated in the mouth is quickly, although 
 not immediately, stopped, the exact stage dei^ending upon the 
 rapidity with which hydrochloric acid in the free state appears 
 in the gastric contents. The hydrochloric acid of the gastric 
 juice may now bring about a certain amount of hydrolysis of the 
 cane-sugar and maltose arising during salivary cUgestion, or already 
 present in the food, but the inversion of saccharoses occurs mainly 
 in the intestine. Absorption of sugars takes place to a shght 
 extent through the stomach wall, especially when there is a con- 
 centrated solution, but it is never very marked. The presence of 
 alcohol increases absorption, even from dilute solution, and may 
 help to account for the glycosuria that occasionally follows the 
 ingestion of alcoholic beverages, and particularly champagne 
 and beer. 
 
 Carbohydrate digestion is essentially effected in the small in- 
 testine. Through the agency of the pancreatic ferment, amylopsin. 
 the insoluble starch is converted into soluble starch, or amylodex- 
 trin, and is then successively decomposed, by gradual hydrolysis., 
 into erythrodextrin, achroodextrin, isomaltose, and maltose. 
 
 Glycogen is similarly decomposed and, like starch, gives rise to 
 isomaltose and maltose. Cellulose is not affected by any of the 
 digestive ferments, but under the influence of the intestinal bac- 
 teria it undergoes a certain amount of fermentative change, par- 
 ticularly in herbivorous animals, so that only a fraction of the 
 ingested cellulose aj)pears as such in the feeces. Maltose or dex- 
 
12 GLYCOSUKIA 
 
 trose have not, however, been found as products of the fermentation 
 of cellulose in the intestinal tract. The intestinal bacteria also 
 appear to have the power of transforming a certain amount of 
 starch into maltose and other products ; for Avhen the pancreatic 
 secretion is prevented from entering the intestine, either by extir- 
 pating the organ or ligaturing the duct, from 47 to 71 per cent, 
 of the ingested starch of the food appears to be utilised. In 
 severe cases of pancreatic disease in man, where the digestive 
 functions of the pancreas are seriously interfered with, it is found 
 that there is not the marked failure of starch digestion that might be 
 expected, and that analysis of the faeces shows to have occurred with 
 the fats and proteids. This is x)robably due to the action of intestinal 
 bacteria ; but it has also been suggested that the epithelial cells of 
 the small intestine are capable of inverting dextrin to maltose, and 
 so replacing to a certain extent the functions of the pancreas. 
 
 Maltose is the chief sugar formed by the action of amylopsin 
 on starch, but before absorption takes place this and other poly- 
 saccharides present in the intestinal contents are converted into 
 monosaccharides. The change is effected chiefly by the inverting 
 ferments, maltase and lactase. The succus entericus possesses only 
 feebly diastatic powers, but by means of these ferments, which are 
 contained in it, it rapidly converts maltose into grape-sugar and 
 Isvulose. These are absorbed by the epithelial lining of the in- 
 testinal mucosa, and, passing thence by way of the mesenteric 
 veins, reach the portal system. Probably less than one per cent, 
 normally goes through the lymphatics and the thoracic duct 
 directly into the venous system. It is probable that a small pro- 
 portion of the dextrin formed in starch hydrolysis, and possibly 
 also some of the saccharoses, are absorbed by the intestinal epi- 
 thelium as such, and are converted by the cells before being passed 
 into the blood stream. 
 
 The rapidity with which carbohydrates are digested and ab- 
 sorbed varies considerably^ Albertini found that when 100 grams 
 of dextrose are given to dogs, 60 grams are absorbed in the first 
 hour ; while of similar doses of maltose and cane-sugar from 70 
 to 80 grams, and of lactose from 20 to 40 grams, disappeared 
 \\ithin the same period. The fact that no soluble carbohydrates 
 can be found in the faeces does not prove that they have all been 
 absorbed and utilised. Even in health there is always some waste 
 from fermentation by bacteria in the intestine, acetic acid, lactic 
 acid, butyric acid, succinic acid, formic acid, alcohol, carbon 
 dioxide, methane, hydrogen, and other bodies being formed. 
 In conditions of disease, where the intestine flora is abnormal, 
 
THE CARBOHYDRATES 13 
 
 such changes may be much exaggerated and give rise to disturb- 
 ances of intestinal digestion and diarrhoea, which are favoured 
 by the ingestion of large amounts of sugar. In such cases the 
 loss of caloric potential may be very considerable and lead to 
 marked inanition. It is not improbable that the liking which 
 many diabetics exhibit for sweet substances may, through the 
 long-continued changes set up in the intestine and pancreas by 
 excessive intestinal fermentation, have been the cause of the 
 imperfect sugar metabolism which they subsequently develop. 
 Starches do not appear to have the same effect as sugar, probably 
 because inversion and absorption run parallel, so that the intes- 
 tinal bacteria have little chance of setting up fermentative changes 
 and giving rise to the formation, of irritating by-products such as 
 is afforded by an abnormally large amount of sugar. 
 
 The digestive processes undergone by the carbohydrates of 
 the food are definitely known, and it is also certain that they are 
 an important source of energy, but exactly how they are dis- 
 tributed and utilised is not quite so clear. When sugar, whatever 
 its source, is absorbed into the blood, it is fixed by the cells of the 
 body in the form of glycogen. This glycogen bears a similar 
 relation to the living cell that the coal in its tender bears to a 
 locomotive. It forms a store on which the tissues can draw in 
 the course of their metabolism, converting the potential energy 
 of carbohydrates into work, and ultimately breaking it dowii into 
 carbon dioxide and water. It was at one time thought that the 
 carbohydrates of the food were the only source of energy for the 
 body, and that the proteicls were used only to repair tissue waste. 
 Such, however, is not the case, and the difficulties of the problem 
 of carbohydrate metabolism have been increased by the discover}^ 
 that most proteins contain a carbohydrate radicle, and the possi- 
 bility that this may be split off and utilised by the organism. The 
 results of experimental investigation suggest that this does occur, 
 and that glycogen may be formed from such carbohydrate groups. 
 But proteicls, such as casein, which contain no jDreformed carbo- 
 hydrate complex, increase the sugar output in the diabetic organism, 
 and it is therefore probable that sugar may in addition be derived 
 from other decomposition products of the protein molecule. It 
 also appears certain that glycogen may be formed by an animal 
 from the proteins of its OAvn tissues. Which are the degradation 
 products of the jDrotein molecule that can be converted into carbo- 
 hydrate are not knoAvn for certain, but the amino-acids, containing 
 six or three carbon atoms, suggest themselves as the most likely 
 source. 
 
14 GLYCOSURIA 
 
 The neutral fats, consisting as they do of gtycerine combined 
 with fatty acids, are another possible source of sugar in the organism. 
 The transformation of glycerine into sugar is not a difficult chemical 
 operation, and experiments mth phloridzinised animals, and dogs 
 after the removal of the pancreas, suggest that such a transforma- 
 tion may occur A^ithin the body. The glycerine is probably first 
 converted into glycerose, and this in its turn is converted into 
 levulose. The fatty acid portion of the neutral fats cannot 
 probably be converted into sugar, although some authors (e.g. 
 V. Noorden) mamtain that even this transformation can be effected. 
 In the j)resent state of our knowledge it may be concluded that 
 fats may be one source of sugar in the organs, but that there is 
 an absence of conclusive proof to this effect as yet. 
 
 It is evident, therefore, that the subject of carbohydrate meta- 
 bolism is much more complicated than at first sight it might appear 
 to be, and that in considering the question we have not only to 
 take into account the carbohydrates of the food, but also the 
 food and tissue proteins, and jarobably also the fats. 
 
 The most generaUy accepted view dates from the discovery 
 by Claude Bernard, in 1857, that the liver contains very little 
 free sugar but a considerable amount of glycogen. It is based 
 upon the theory that the liver glycogen represents the carbo- 
 hydrate absorbed in excess of the immediate requirements of the 
 body, and that this is converted into sugar by the action of a 
 ferment, and is transported to the tissues as they require it by the 
 systemic blood stream. According to this, which may be termed 
 the classical view, the sugar taken up by the blood from the in- 
 testine is conveyed by the portal vein to the liver, where it is 
 converted into glycogen and stored for the future needs of the 
 body. In support of this is adduced the experimental evidence 
 that during absorption the blood in the portal vein contains a 
 much higher percentage of sugar (0*2-0'4%) than the systemic 
 blood (0-05-0-2%), but during fasting the percentage is the same. 
 The amount of glycogen in the liver depends largely on the intake 
 of food, but never exceeds about 150 grams (about 5 ounces), and 
 as this only disappears after several weeks' starvation, it cannot 
 account for the whole of the sugar which is absorbed in a short 
 space of time when a meal rich in starch and sugar is taken. It 
 is therefore assumed that the excess passes through the liver and 
 is laid down in the muscles and other tissues, also in the form of 
 glycogen, thus accounting for another 150 grams. Even if the 
 whole of the glycogen in the liver, muscles, and other tissues of 
 the body, and the sugar in the circulating blood, which is about 
 
THE CARBOHYDRATES 15 
 
 10 grams (J ounce) or less, are allowed for, they do not still 
 represent the whole of the carbohydrate that may be absorbed. 
 It is consequently supposed that the balance enters into the con- 
 stitution of the proteins, nucleo -proteins, and albuminoids, from 
 which a carbohydrate material has been obtained on treatment 
 with acids ; it is probable also that a certain proportion may be 
 turned into fat. The glycogen in the liver, and to a less extent 
 in the muscles, is believed to form a readily available store from 
 which the wants of the body can be quickly supplied, the glycogen 
 being converted into dextrose by the action of special ferments 
 as the need arises. The sugar thus formed in the liver is convej^ed 
 to the tissues, which seize upon it and utilise it in their metaboUsm, 
 splitting it up and oxidising it. According to this theory provi- 
 sion is made in the body for a certain percentage of sugar to be 
 constantly present in the blood. If from any cause the amount 
 circulating exceeds more than from 0-1 to 0-2 per cent., the excess 
 is excreted by the kidneys. This is normally prevented by the 
 fact already referred to, that the liver and muscles at once store 
 up any excess above the normal, resulting from too rapid absorp- 
 tion, as glycogen. If, on the other hand, the percentage in the 
 blood sinks below the normal, owing to the consumption being 
 increased from work or heat production on the part of the tissues, 
 the liver, and later the muscles, at once give back a portion of 
 their glycogen to the blood in the form of sugar. If the stored 
 glycogen is insufficient for this purpose, fat and albuminoids are 
 made use of, the percentage in the blood remaining constant, 
 even after long-continued starvation. 
 
 The theory that the carbohydrates of the food are destined to 
 pass through the circulation to the tissues in the form of sugar 
 has been strenuously opposed by Pavy. He considers that the 
 assumption as to the impermeability of the kidneys to sugar 
 involved in this theory is a fiction, and that in reality the urine 
 stands in very sensitive relationship to the blood with respect 
 to sugar. According to his view, all the food which has been 
 broken down in the intestine, and placed in a fit state for absorj^tion, 
 is at once dealt with at the seat of absorption, being rebuilt, before 
 reaching the circulation, into molecules of sufficient size to prevent 
 their flowing off with the urine in its passage through the kidneys. 
 The building-up process, he believes, is effected by the lymphoc^'tes, 
 thus accounting for the lymphocytosis which accompanies diges- 
 tion. The lymphocytes, carrying the elaborated food material, j)ass 
 from the villi into the absorbent vessels, and thence through the 
 thoracic duct into the vascular system. There thej^ break do^^^l 
 
16 GLYCOSURIA 
 
 and are transformed into the protein constituents of the chyle 
 and blood plasma, thus bringing the elaborated food into direct 
 relation Avith the tissues. The carbohydrates, fat, and nitrogen con- 
 taining material enters into the protoplasmic complex of the cells 
 and, interacting with the oxygen, also brought by the blood, give 
 rise to energy and the phenomena of life. Any oxidisable material 
 taken on in excess of the consumption is cleaved off and stored 
 for future use in the shape of glycogen and fat. Sugar which is 
 nob dis]oosed of at the seat of absorption in the manner described, 
 and particularly when a large amount of carbohydrate food is 
 ingested, is supposed to pass to the liver, and there be checked 
 from further progress by being taken into the liver cells and con- 
 verted into glycogen, and possibly also into fat. The liver thus 
 forms a second line of defence against the passage of absorbed 
 sugar into the systemic circulation, and prevents the onset of the 
 glycosuria that would occur if it flowed on instead of being re- 
 tained. The glycogen in the liver and muscles is, according to 
 this idea, to be regarded simply as a reserve of carbohydrate 
 material ready to be drawn uj)on and utilised as it becomes wanted, 
 its special accumulation in the liver being accounted for by the 
 position which that organ occupies in relation to the food supply. 
 In the muscles the amount stored depends chiefly upon the extent 
 to which they are used, diminishing with exercise and accumulating 
 at rest. As in the case of starch in the vegetable kingdom, the 
 glycogen is probably broken down into sugar before being absorbed 
 into the protoplasm of the tissue, and it can be inferred that this 
 change is brought about by the action of an enzyme. The glycogen 
 in the liver when required is similarly broken down by enzyme 
 action into sugar, but this, instead of passing directly into the 
 circulation, is assumed to be loosely linked on as a side-chain to a 
 protein nucleus, and to be conveyed in this locked-up condition to 
 the tissues, where the carbohydrate radicle is taken off and utilised 
 as required, the protein molecule thus set free being available for 
 the attachment of a fresh sugar side-chain. The essential point 
 in this theory is that the carbohydrate is transported from the 
 seat of accumulation in the liver to the seat of utilisation in the 
 tissues as part of a large molecule which can pass through the blood 
 without running off with the urine. Pavy's views have not been 
 generally accepted, in spite of the brilhancy and perseverance 
 with which he defended them, but they have undoubtedly had a 
 considerable effect in the way of modifying the theories based 
 originally on Bernard's experiments. 
 
 Whether the carbohydrates of the food inevitably go through 
 
THE CARBOHYDRATES 17 
 
 the glycogen stage or not, there can be no doubt that thej^ ulti- 
 mately reach the tissues and are there broken down, eventually 
 forming carbon dioxide and water. In spite of the large amount 
 of research which has been devoted to the elucidation of the pro- 
 blem of the metabolism of sugar in the animal organism, the exact 
 details of the intermediate steps are as yet not understood. It 
 is probable that the decomposition does not occur, at any rate in 
 its entirety, as a direct oxidation, but that an intermediate series 
 of oxidation and fission products are formed. 
 
 Some interesting observations, made from a chemical stand- 
 point by Adolf Jolles of Vienna, seem calculated to throw con- 
 siderable light on this subject. He has investigated the action of 
 various oxidising agents upon a number of different sugars, in- 
 cluding arabinose, rhamnose, dextrose, levulose, invert sugar, 
 mannose, galactose, cane-sugar, maltose, and lactose, in weak 
 alkaline solutions at 37° C. In general the strength of the solutions 
 was 1 per cent, of sugar made N/100 alkaline with sodium hydrate. 
 In every case, Avith the exception of cane-sugar, a diminution in 
 the rotating power of the solution occurred, together with the 
 formation of acids. He found that neutralisation of the alkali 
 produced marked slowing in the formation of the acids, while 
 with glucose in N/100 acid solution acid formation does not occur 
 and the sugar remained unaltered. By the addition to the solu- 
 tions of hydrogen peroxide the oxidation processes were accelerated, 
 occurring more quickly than when the oxygen in the air was used 
 as the oxidising agent. Levulose was found to produce more acid 
 than dextrose, and therefore to be more easily oxidisable. The 
 oxidation products obtained comprised ethyl alcohol, acet-aldehyd, 
 acetone, formic, acetic, butyric, lactic, glycolic, oxalic, succinic, 
 aceto-acetic, and glucuronic acids. With most sugars the chief 
 product was formic acid, a very small quantity of acet-aldehyd, and 
 an acid which gave Tollen's naphtho-resorcin reaction for glucuronic 
 acid also being formed. Lactic acid was only obtained in alkaline 
 solutions of dextrose without the addition of hydrogen peroxide. 
 The use of oxide of silver as an oxidising reagent gave similar 
 results to those obtained with hydrogen peroxide. Ammonia and 
 sodium carbonate did not influence the decomposition of sugar 
 as strongly as sodium hydrate. 
 
 Jolles is of opinion that the conditions of his experiment ap- 
 proximated to those in the body. The blood with its definite 
 alkalinity permeates the tissues, and the peroxidases, catalases, 
 and other oxidising ferments can be regarded as exerting a similar 
 action to the hydrogen peroxide in his chemical experiments. He 
 
 B 
 
18 GLYCOSURIA 
 
 suggests that the sugar in the tissues is oxidised to acids of low 
 molecular weight, such as formic acid, which are further oxidised 
 in the blood to carbon dioxide and water. He has also shown 
 how his observations can be adapted to explain glycolysis in muscle, 
 the formation of sarcolactic acid, of glucuronic acid, and even some 
 of the features of diabetes and pentosuria. 
 
 A series of experiments carried out by Nef have suggested 
 that glycerin aldehyde is an important intermediate product of 
 the breakdown of the hexoses. Observations conducted by 
 Woodyatt tend to confirm this, and show that in the course of the 
 utilisation of sugar in the body a cleavage of glucose into two 
 molecules of triose is an important event. According to Nef , lactic 
 acid, glycerinic acid, and other oxidation products of glycerin 
 aldehyde are formed by intramolecular rearrangement in this body 
 when there is an insufficient supplj^ of oxygen. 
 
 As the result of a series of researches carried out by Stoklasa 
 and others, it has been assumed that the tissues contained a glyco- 
 lytic enzyme capable of causing true alcoholic fermentation of 
 sugars, and that it is by the action of this ferment that the degra- 
 dation of sugars is brought about in the body. In support of this 
 view there has been cited the well-known phenomenon of the 
 formation of lactic acid in the tissues after death, which has been 
 interj)reted as an intermediate stage in the process of alcoholic 
 fermentation. Harden and Maclean have shown, however, that 
 this theory is probably based upon altogether erroneous observa- 
 tions. Experiments of this kind involve the examination and 
 manipulation of various animal organs and tissues, and it is only 
 with great difficulty that they can be kept free from contamination 
 with bacteria. In most cases hitherto no attempt has been made 
 to do so. Harden and Maclean point out the difficulty of per- 
 forming such experiments under absolutely sterile conditions, but 
 with careful precautions they succeeded on a few occasions. In 
 these cases no trace of alcoholic fermentation could be detected. 
 The presence of an efficient antiseptic leads to a similar negative 
 result. The fermentation occurring under natural conditions must, 
 therefore, be due to the action of bacteria, a large number of which 
 are knoMii to cause rapid fermentation of various sugars. The 
 unavoidable conclusion, therefore, is that there is no satisfactory 
 evidence that alcoholic fermentation occurs in animal tissues after 
 removal from the body, apart from the presence of sugar-fermenting 
 bacteria. 
 
 Carbohydrates in Normal Blood. — Normal blood always con- 
 
THE CARBOHYDRATES 19 
 
 tains traces of sugar, which may be temporarily increased by a 
 diet rich in carbohydrates, and be diminished by muscular exercise 
 and hunger. 
 
 The sugar content of the systemic circulation averages about 
 0*8 grams per 1000 when it is estimated by the ordinary reduction 
 methods and is calculated as dextrose. Limbeck found in the 
 blood of two healthy subjects, five hours after eating, 0*075 per cent, 
 and 0-089 per cent. With the polarimeter, however, a much lower 
 reading is obtained, so that the sugar of the blood must either be 
 a variety differing from dextrose, or be composed of a mixture 
 of sugars with opposite optical characters. 
 
 Many physiologists consider that part at least of the sugar in 
 the blood exists in loose combination with some other substance. 
 Some maintain that this is lecithin, forming the so-called jecorin, 
 first found by Dreschel in the liver, while others believe that the 
 albuminates are the sugar-carriers. Most are agreed that part 
 exists in a free state, but the work of Rona and Michaelis tends 
 to prove that all the sugar in the blood is in a simple state of solu- 
 tion, some in the corpuscular elements, the remainder in the plasma. 
 They have shown that when diluted blood is shaken with certain 
 colloids, such as ferric hydroxide or kaolin, the proteins form a 
 colloidal combination and are absorbed. They can then be quanti- 
 tatively precipitated by the addition of a trace of electrolyte, but 
 that no trace of sugar is removed from the solution by this treat- 
 ment. If the sugar were in any way united with the proteins it 
 would be carried down with them, and as the reagents employed 
 cannot have any disruptive effect, it is not possible that the sugar 
 can exist in combination with the proteins. Another piece of 
 evidence in support of the free state of dextrose in the blood is 
 furnished by the observation that, whereas charcoal absorbs both 
 sugar and protein when shaken with a solution containing these 
 two substances, yet it absorbs the protein, but not the dextrose, 
 when acetone is present ; the acetone being more absorbable 
 than the dextrose, prevents the latter being taken up by the 
 charcoal. Further evidence is also furnished by the results of 
 dialysis experiments. 
 
 Levulose. — Lepine and Boulud have obtained from the blood in 
 certain cases a reducing sugar having the characters of le^Tilose, 
 and explain its presence on the assumption that it has been de- 
 rived from dextrose in the alkaline medium furnished by the blood. 
 
 Maltose has been demonstrated in the blood of healthy 
 rabbits and dogs, and is supposed to be derived from the intestinal 
 
20 GLYCOSUEIA 
 
 contents, or to depend upon imperfect hydrolysis of glycogen in 
 the liver. 
 
 Traces of pentose, and in certain instances of a sugar resembling 
 saccharose, have been found in the blood. The former appears 
 to be a constant constituent, and the latter is supj)osed to be 
 derived, either from the intestinal contents, or be produced in the 
 animal economy by a combination of dextrose with levulose. 
 
 Glucuronic acid has been described as present in the blood of 
 both man and cattle by P. Mayer, and this observation has been 
 extended by Lepine and Boulud to the dog. Since the conjugate 
 glucuronates are levo-rotatory, their presence would help to explain 
 the difference between the readings obtained with an extract of 
 the blood by the polariscope and on reduction. The small quantity 
 in normal blood is, however, against this being the sole explanation. 
 
 Animal Gum. — Freund has obtained from blood a carbo-hydrate- 
 like substance resembling the animal gum of Landwehr. Ox blood 
 was found to contain about 0-02 per cent. 
 
 Glycogen is said to be present in traces in the blood, but it is 
 not improbable that the glycogen found free in the plasma is 
 derived from the leucoc^i^es. which are known to contain it. 
 
 BIBLIOGRAPHY 
 
 Allen's Commercial Organic Analysis, vol. i., 1909. 
 
 Armstrong, The Simple Carbohydrates, 1910. 
 
 Fenton, Journ. Chem. 8oc., 1907. 
 
 Fischer, Untersuch. u. Kolenhyd. u. Fermente, 1909. 
 
 Freund, Centralb. f. Physiol, 1892. 
 
 Griind, Zeit. f. phys. Chem., xxxv., p. 111. 
 
 Harden and Maclean, Journ. of Physiol., 1911. 
 
 Jolles, Weiner med. Woch., 1911. 
 
 Lepine, Le diabete sucre, 1909. 
 
 Lepine and Boulud, C.R. de VAcad. d. Sci., 1901-2. 
 
 Limbeck, Prag. med. Woch., 1893. 
 
 MacLeod, Recent Advances in Physiology, 1906. 
 
 Mayer, Zeit. f. physiol. Chem., 1901. 
 
 Nef, Ann. d. Chem., Liebig's, ccclvii. 
 
 Pavy, Lancet, 1908. 
 
 Rohmann, Biochemie, 1908. 
 
 Kona and Michaelis, Biochem. Zeit., xiv. 
 
 Schryver, Proc. Roy. Soc, 1910. 
 
 Tollens, Eurze Handbuch d. Kolenhyd, 1898. 
 
 Woodyatt, Journ. Amer. Med. Ass., 1910, 
 
CHAPTER II 
 
 the detection and differentiation of sugars and other 
 reducing substances in the urine 
 
 Normal Urine 
 
 The question as to whether the urine of healthy individuals con- 
 tains sugar was for many years a subject of keen controversy. In 
 1848 Lespiau stated that normal urines have reducing powers. 
 Ten years later Briicke confirmed this observation, and declared 
 that normal urines contain sugar. His statements were supported 
 by Bence Jones, Tuchen, Abeles, Meissner and Babo, Udranszky, 
 Wedenski, Molisch, Quinquand, Bruel, Luther, Roos, Moritz, 
 Binet, Allen, Baisch, and Pavy, who maintained that all urines 
 contain small quantities of reducing carbohydrates. Some ob- 
 servers, including Seegen, Friedlander, Malay, Leuken, Kiilz, 
 G. and S. G. Johnson, and others, while they allowed that normal 
 urines possess slight reducing powers, came to the conclusion 
 that this can be entirely explained by the presence of other sub- 
 stances than sugar, and particularly creatinin and uric acid. There 
 can be no doubt that these bodies do partly account for the slight 
 reduction caused by many urines when they are boiled with alkaline 
 solutions of copper, &c., but, although a number of the recorded 
 observations bearing on the question are open to serious criticism, 
 and, as Johnson pointed out, it involves a definition of what is a 
 " normal " urine, the balance of available evidence is in favour 
 of the view that the urines of average healthy individuals j)robably 
 contain minute quantities of glucose, and traces of other reducing 
 carbohydrates. The constant presence of the former has not, 
 however, been absolutely proved, and it is probable that diet, 
 exercise, and mode of life have some bearing on the C[uestion, 
 and also on the excretion of glucuronic acid, the presence of which 
 also contributes to the reduction. Worms examined the urines of 
 507 persons of the labouring class, and found that in every instance 
 they were free from sugar ; but out of 100 samples from persons 
 engaged in sedentary occupations, involving mental activity, he 
 found sugar in ten. My own observations have given somewhat 
 
22 GLYCOSURIA 
 
 similar results, but in my experience the reduction given by the 
 urines of persons engaged in sedentary work appears to be chiefly 
 dependent upon the presence of glucuronic acid. Haas, in 1876, 
 pointed out that normal urines are faintly levo-rotatory, and in 
 1885 Fluciger explained this by the presence of glucuronic acid, 
 for he found that on heating the urine with dilute acid its reducing 
 power is much increased and its optical activities are altered. His 
 opinion has . since been confirmed by Mayer and Neuberg, by 
 Porcher and Nicolas, and by others, who have shown that glucu- 
 ronic acid is a very constant constituent of normal urine. Accord- 
 ing to Mayer and Neuberg, it is usually present in quantities of 
 about 0-004 per cent., mostly in combination with phenol, and to a 
 less extent with indol and skatol. Moritz states that the uric acid 
 and creatinin of the urine account for about 50 per cent, of its reduc- 
 ing power under normal conditions ; but more recently Levesson 
 has given a lower figure, 25 to 33 per cent, of the total. 
 
 The total amount of reducing carbohydrate in normal urine 
 has been variously estimated by different observers, but, as we 
 have seen, some of these variations may be partly explained, in 
 all probability, by the diet and environment of the persons whose 
 urine was investigated, while the different methods of estimation 
 employed offer another partial explanation. According to Sal- 
 kowski, the reducing substance of normal urine varies from 0-254 to 
 0-596 per cent. Rosen and Alfthan obtained from 1-5 to 3-0 grams 
 of precipitate from the twenty-four hours' urine of a healthy man 
 by the benzoyl chloride process, while Baisch states that normal 
 urines contain about 0-12 to 0-32 grams of reducing carbohydrate, 
 of which 0-08 to 0-18 grams is grape-sugar. Baisch and Lemaire 
 have also isolated a sugar which they considered to be isomaltose, 
 besides a dextrin-like substance having the characters of Land- 
 wehr's animal gum and a nitrogen-containing body yielding fur- 
 furol, probably derived from mucin or chondroitin sulphuric acid, 
 from normal urine. Lohenstein places the amount of sugar as low 
 as 0-001 per cent., while Pavy regards 0-05 per cent, as the average 
 amount. Kellas and Wethered state that an average of 0-08 per 
 cent, of substances reacting like grape-sugar may be normally 
 present in the urine. The most recent observations made by 
 Schondorff place the quantity of sugar at about 0-01 per cent. 
 
 Clinically the presence of grape-sugar in normal urines is of 
 minor importance, for in any case the quantity is so exceechngly 
 small that it is unrecognisable by the ordinary methods of testing, 
 and such amounts are associated with no clinical symptoms. It 
 is important, however, to remember that urines from apparently 
 
QUALITATIVE TESTS 23 
 
 healthy individuals contain substances which have reducing powers, 
 and that under certain circumstances these may be sufficient to give 
 puzzling results with some methods of examination. 
 
 Abnormal Urines 
 
 In abnormal conditions the reducing power of the urine may 
 be increased so that a more or less marked reaction is obtained 
 with the ordinary clinical tests. By far the most common and 
 important cause is the presence of an appreciable quantity of 
 dextrose, but a reaction may also be due to the presence of levulose, 
 lactose, galactose, maltose, isomaltose, pentoses, homogentisic 
 acid, or compound glucuronates, and a doubtful result is some- 
 times dependent upon an increased excretion of uric acid and 
 creatinin. The presence of some of these substances indicates an 
 undoubted perversion of the metabolism of the body, which may, 
 or may not, be of a permanent and serious character. Others are 
 of doubtful significance. A few are of no known pathological 
 importance. It is obvious that these groups must be clearly 
 differentiated, and the more important members be definitely 
 recognised, if an analysis of the urine is to be of any use in treat- 
 ment and prognosis. In the succeeding pages of this chapter the 
 means by which these objects may be attained will be considered, 
 and reference will also be made to other carbohydrates and sub- 
 stances of a similar composition, which are occasionally met with 
 in the urine. 
 
 Collection of the Urine. — When selecting a single specimen of 
 urine for examination for sugar, it is best to take one that has 
 been passed during the day, preferably in the evening, for if only 
 a small amount is present the morning urine mil probably con- 
 tain less than the evening, and may even give no reaction at all. 
 It may also happen that a specimen taken out of the collected 
 urine for twenty-four hours will give a doubtful or negative result, 
 whereas one that has been passed three or four hours after a meal, 
 and particularly a meal rich in carbohydrates, will give a decided 
 reaction for sugar. Conversely, a strictly protein diet may cause 
 sugar which has previously been present to disappear. For diag- 
 nostic purposes it is therefore advisable that an examination of a 
 twenty-four hours' sample should first be made, and, if this is 
 negative, another specimen taken three or four hours after a meal 
 containing an average amount of carbohydrate should be investi- 
 gated. All specimens must be examined in as fresh a state as 
 
24 GLYCOSURIA 
 
 possible, since traces of sugar may be destroyed, and escape de- 
 tection, if the urine has been allowed to ferment and decompose. 
 
 The Physical Characters of the Urine often afford some 
 indication as to the presence of sugar. Urines containing much 
 glucose usually present a pale, greenish-yellow colour, combined 
 with a high specific gravity, 1-025 or over. It is not uncommon, 
 however, for urines with a normal or even a low specific gravity to 
 contain sugar, and v. Jaksch has reported examples where the 
 specific gravity has been as low as 1-003. In pentosuria, lactosuria, 
 levulosuria, and similar conditions the specific gravity does not, 
 as a rule, show any marked variation from the normal. The 
 amount of urine excreted in most cases of persistent dextrosuria 
 is excessive, 3 or 4 litres (5 to 7 pints) a day in many instances. Its 
 reaction is generally distinctly acid, and on being shaken it readily 
 forms a persistent froth. Saccharine urines ferment spontaneously, 
 especially in warm weather, forming bubbles of carbonic acid gas, 
 and showing a sediment of yeast microscopically, except when the 
 Tinfermentable sugars lactose, pentoses, &c., are alone present. 
 
 The Chemical Reactions of Saccharine Urines.— The oldest 
 and simplest test for sugar in the urine is afforded by its sweet 
 taste. Celsus and Galen in describing diabetes make constant 
 reference to the " sweet and honey urine." In China and the East, 
 sugar is detected by allowing the urine to evaporate on the ground 
 in the sun, and then watching for the concourse of ants and other 
 insects that are attracted by the sweet residue. A somewhat rough- 
 and-ready test, but one which is much more delicate than might 
 be supposed, and that can be carried out at the bedside mth no 
 more apparatus than can be obtained in any household, is afforded 
 by evaporating a few drops of the suspected urine to dryness in 
 a spoon over the flame of a candle or lamp. The residue is gently 
 heated, and as the temperature rises it will be seen to form a pure 
 yellowish-brown viscid mass, which is sticky to the touch, and 
 gives forth an odour of caramel before it is reduced to ash, at 
 about 200° C, if much sugar is present. Urines free from sugar 
 treated in this way show a dirty grey-brown residue, and give 
 no odour of caramel on being further heated. 
 
 The laboratory tests for sugar may be conveniently divided 
 into : — 
 
 1. General tests, with which a reaction is given by all sugars. 
 
 2. Classifying tests, which separate the sugars into groups 
 
 characteiisecl by the possession of one or more common 
 characters. 
 
QUALITATIVE TESTS 25 
 
 3. Special or confirmatory tests, Avhich serve to more or less 
 _ completely differentiate particular sugars. Many of these 
 can only be satisfactorily aj^pliecl to the separated and 
 purified sugar, however. 
 
 1. General Tests 
 
 In testing urines for sugar the minute traces that may be 
 normally met with are disregarded, and sugar is only considered 
 to be present when a characteristic reaction is obtained by methods 
 which have been j)i"oved by clinical experience to show" a patho- 
 logical amount. The most usually employed tests are based upon 
 the reducing powers of the sugars, but since other reducing sub- 
 stances are also met with in the urine, it is advisable, and in all 
 doubtful cases necessary, to confirm a positive reaction by other 
 methods before it is concluded that any reduction that has taken 
 place is due to sugar. 
 
 1. Moore-Heller Test. — This is one of the earHest described 
 tests for sugar in the urine, but it is now rarely made use of, as it only 
 gives a characteristic reaction with a relatively high percentage. 
 
 A few cubic centimetres of sodium, or potassium^ hydroxide are added 
 to about three or four times their volume of the urine, in a test-tube, 
 and boiled for 2-3 minutes. If a considerable amount of sugar is present 
 the fluid begins to turn brown at about 60° C, and gradually darkens as 
 the heating is continued. The reaction is only characteristic of sugar if 
 the colour is a dark yellow to a dark brown, or with diluted urines an 
 intense yellow. It is a wise precaution to compare the result with that 
 given by a normal urine under similar circumstances. If the mixture is 
 allowed to cool and is then cautiously acidulated with sulphuric acid an 
 odour of burnt sugar should be produced. 
 
 With pure sugar solutions the test is very sensitive, but only 
 urines containing at least 0-5 to 1 per cent, of sugar give a charac- 
 teristic brown colouration. Sugar-free urines may give a dark 
 yellow coloration when boiled with a caustic alkali, especially if 
 they are high-colourecl to start with. All urines containing mucus 
 darken somewhat. Any albumen that is present must be removed 
 by acidifying, boiling, and filtering before appljdng the test. A 
 flocculent precipitate of earthy phosphates is generally produced 
 when the alkali is added to the urine, and this collects into large 
 flocculi when heat is applied. It is, however, quite a normal 
 23henomenon. 
 
 2. Trommer's Test. — If a few drops of a dilute solution of 
 
26 GLYCOSURIA 
 
 copper sulphate are added to a solution of caustic potash or soda, 
 a blue precipitate of hydrated cupric acid is formed — 
 
 CUS04+ 2K0H = K^S04 + Cu(0H)2 . 
 
 On heating the liquid the precipitate blackens, owing to the 
 formation of cupric oxide (CuO). If, however, glycerine, tartrates, 
 and various other substances are present in the solution, the cupric 
 hydrate is not precipitated when the copper and alkaline solutions 
 are mixed, but forms a deep blue solution, which on being heated 
 does not blacken. Dextrose acts like glycerine and tartrates in 
 keeping the cupric hydrate in solution, forming with it a compound 
 with the formula CgH-^2C>6-^^^(^-^)2' ^^^ when the temperature of 
 the solution is raised to near the boiling-point reduction occurs, 
 the blue colour of the solution being discharged, and a yellow pre- 
 cipitate of cuprous hydrate (Cu2(OH)2), or a red precipitate of 
 cuprous oxide (CugO) appearing. 
 
 With urine the test is carried out as follows : — 
 
 To a test-tube about half filled with the suspected urine is added from, 
 a quarter to a third of its volume of a 10 per cent, solution of sodium, or 
 potassium, hydrate. A 5 to 10 per cent, solution of copper sulphate is then 
 added drop by drop, shaking after each addition, until a faint trace of copper 
 hydroxide remains undissolved. If the urine is found to take up much 
 copper before a permanent precipitate appears, and it assumes a deep blue 
 colour, the presence of sugar is probable, the amount of copper required 
 being a rough indication of the quantity of sugar. On gently heatings but 
 not actually boiling, the upper part of the mixture, a yellow or greenish tur- 
 bidity will appear in the heated portion and spread downward through the 
 blue fluid, if the urine contains sugar in an appreciable amount. Eventually, 
 as the heating of the mixture is continued, the blue colour is more or less 
 completely discharged, and a yellow or red precipitate settles to the bottom 
 of the test-tube. If the urine contains a high percentage of sugar, metallic 
 copper may separate out on the walls of the tube as a brownish-red coating.. 
 
 A typical reaction is only obtained with urines that contain a 
 distinctly pathological amount of sugar, but no other substances. 
 than the reducing sugars give a quite characteristic reaction. 
 When only traces are present (under 0-5 per cent.) the fluid may 
 turn yellow, but no precipitation of copper oxide occurs. The 
 formation of an intense brilliant yellow colour, while suggestive of 
 the presence of a small amount of sugar, is not conclusive evidence, 
 for other substances may bring about this less typical reduction. 
 
 In carrying out Trommer's test it is important to bear in mind 
 the following points : — 
 
 1. That normal urines contain substances, such as uric acid.. 
 
QUALITATIVE TESTS 27 
 
 creatimn, and salts of ammonia, which are able to dissolve cupric 
 hydrate, hence, as a rule, from three to five drops of copper sulphate 
 can be added to each 10 c.c. of urine before precipitation occurs. 
 The resulting fluid is, however, greenish-blue rather than a distinct 
 blue. 
 
 2. From the presence of uric acid, creatinin, glucuronic acid 
 compounds, traces of carbohydrates, and pyrocatechin, &c., normal 
 urines have some reducing power, the total reduction that takes place 
 being usually equal to about 0*25 to 0*5 per cent, of glucose. On 
 heating a normal urine with an alkaline solution of copper hydrate 
 to boiling, therefore, the colour of the solution may change to a 
 deep yellow by transmitted, and a reddish yellow by reflected, 
 light, owing to a partial reduction of the copper hydroxide, but 
 no actual precipitation occurs, the fluid remaining perfectly clear. 
 The sediment of phosphates at the bottom of the test-tube pro- 
 duced by the addition of the alkali to the urine may, however, be 
 coloured reddish-brown by traces of entangled cuprous oxide and 
 a grey-green turbidity, due to the formation of an amorphous copper 
 compound of xanthin bases, and uric acid may also be given by 
 concentrated (febrile) urines. The copper precipitate formed by 
 sugar is granular and not amorphous. Since the reducing power 
 of a normal urine is greater than its power of holding copper 
 hydroxide in solution, faintly ammoniacal urines, which dissolve 
 more copper, and so allow fuller play to its reducing action, may 
 give a precipitate, even when no abnormal amount of sugar is 
 present, although more often the black cupric oxide is formed. 
 A urine containing both ammonium carbonate (alkahne fermenta- 
 tion) and sugar may chssolve, and reduce, a good deal of cojDper 
 hydrate, but yield no precipitate of suboxide, because the latter 
 is held in solution by the ammonia. 
 
 3. That normal urines contain substances such as uric acid, 
 creatinin, salts of ammonia, &c., which can hold a certain amount 
 of reduced copper suboxide in solution, and, since this power is 
 more marked than are the reducing abilities of such urines, they only 
 give a colour change, and yield no precipitate, on being boiled ^^ith 
 an alkaline solution of copper. When a normal urine is treated A^dth 
 dextrose until a solution of about 0-5 per cent, strength results, and 
 this is tested by Trommer's method, it is found that a considerable 
 amount of copper hydroxide is dissolved, and a strongly marked 
 yellow colour is produced on heating, but no precipitation of copper 
 suboxide occurs while the heat is being applied, or immediately 
 after. This is due to the suboxide, formed both by the reducing 
 substances and by the oxidation of the sugar, being kept in solu- 
 
28 GLYCOSURIA 
 
 tion. It is only A^•hen a still larger quantity of sugar is present, 
 and has given rise to an excess of the suboxide, that it separates 
 out. By diluting the urine its power of holding the suboxide in 
 solution is diminished to a greater degree than the reducing power 
 of any sugar that may be present, so that a granular precipitate 
 may be formed when the undiluted urine gave no definite reaction. 
 Some observers, therefore, advise a dilution of 1 in 2, or 5, in all 
 cases before applying the test. As ^Dolyuria, with a decrease in 
 the proportion of those substances that prevent precipitation, is 
 a natural phenomenon in many cases of persistent dextrosuria, a 
 definite precipitate is often formed when as little as 0-2 per cent, of 
 sugar is present. 
 
 4. Albumen increases the copper hydroxide holding power of 
 the urine, giving, however, a violet rather than a blue colour to 
 the solution. It does not interfere with reduction, but prevents 
 the precipitation of the reduced suboxide that forms, hence a 
 small quantity of sugar may easily escape detection in an albu- 
 minous urine. Before appljdng the test, therefore, the urine 
 should be tested for albumen ; and if more than a trace (about 
 0-5 per cent.) is found to be present, it should be removed by 
 acidulating with a few drops of acetic acid, boiling, and filtering. 
 
 5. Saccharin and salts of ammonia also interfere with the 
 precipitation of the suboxide, and so their presence may j^revent 
 the detection of small quantities of sugar. 
 
 Trommer's test is not very frequently employed in this country, 
 but it has distinct advantages over many others that are more 
 commonly used. The reagents are stable, all the stej)s of the 
 process are under control, the sources of error are evident, and 
 the presence of disturbing influences is readily detected. More- 
 over, the amount of sugar can be roughly estimated. Should the 
 test readily succeed even with insufficient hydroxide saturation, 
 the amount of sugar is large ; but should no precipitate appear, 
 or only become evident on very accurate saturation with copper 
 hydroxide, and more sensitive tests are found to be positive, the 
 sugar-content of the urine is not more than 0*2 to 0-4 per cent. 
 
 To perform Trommer's test satisfactorily it is necessary that 
 the proportions of the reagents emj)loyed should be fairlj^ accurately 
 adjusted to the quantity of sugar present, especially in doubtful 
 cases, and that the mixture should not be raised to too high a 
 temperature. It is particularly important that an excess of copper 
 sulphate should be avoided, and more should never be added than 
 is necessary to give rise to a few flakes of undissolved hydroxide 
 on gently shaking, since the black oxide of CDj)per that forms on 
 
QUALITATIVE TESTS 29 
 
 heating an excess may disguise the precipitate of cuprous hj^lrate. 
 One part of sugar can reduce about five parts of copper hydrate, 
 and the test should be arranged so that this proportion is, as nearly 
 as possible, present. The cloud of earthy phosphates precipitated 
 by the addition of the alkali to the urine often makes it difficult 
 to decide when the necessary slight excess of copjDer hydroxide 
 remains undissolved, however. 
 
 The colour of the precipitate is generally said to depend upon 
 the alkalinity of the fluid, the brick-red cuprous oxide appearing 
 in strongly alkaline solutions, and the yellow cuprous hydrate 
 in solutions that are relatively only feebly alkaline. Neumayer 
 states, however, that it is to the presence of creatinin in the urine 
 that the formation of the amorphous yellow precipitate is due. 
 and that the crystalline red precipitate, like that given by pure 
 solutions of dextrose, appears when this substance is present in 
 relatively small amounts. 
 
 Beside the reducing sugars and the normal urinary constituents 
 already referred to, a number of substances appearing in the urine 
 under pathological conditions, or as the result of the administration 
 of drugs, &c., may give a more or less marked reduction with 
 Trommer's test. Briickner gives the following list : — 
 
 1. Normal urinary constituents which, according to their pre- 
 
 sence in larger or smaller amounts, may produce varying 
 degrees of reduction — 
 
 Traces of carbohydrates, such as dextrose, isomaltose, 
 pentose, animal gum, glucuronic acid compounds. 
 
 Pyrocatechin, bile pigment, urinary pigment, uric 
 acid, indican, creatinin, urobilin, urobilinogen. 
 
 Concentrated urines are particularly liable to effect 
 reduction, and the same is true of urines containino- a 
 moderate or large quantitj'- of formed elements such as 
 leucocytes, erythrocytes, epithelial cells, &c. 
 
 2. Products of abnormal metabolism which effect reduction : — 
 
 Hexoses (dextrose, levulose, isomaltose, lactose), pen- 
 toses, glycogen, increased quantities of glucuronic com- 
 pounds, homogentisic acid. 
 
 3. Reducing substances added to the urine as ^preservatives : — 
 
 Formaldehyde, chloroform. 
 
 4. Drugs or their derivatives : — 
 
 Acetphenetidin, antifebrin, arbutin, benzoic acid, ben- 
 zosol, copaiba balsam, chloral, glucuronic acid compounds 
 of drugs, morphin, jDhenacetin, saccharin, salicylic acid, 
 salol, sulphonal, turj)entine, thallin, urethan. 
 
30 GLYCOSUKIA 
 
 3. Fehling- (Worm-Miiller) Test. — Glucose cannot dissolve 
 nearly as much copper as it can reduce, so that if the proportion 
 of cupric hydrate in solution can be increased, the reduction is 
 likely to be more evident with small quantities of sugar and the 
 test be made more delicate. This is effected in Fehling's test, 
 and the modifications of it which have been introduced, by adding 
 tartrates, gtycerine, &c., which also have the power of dissolving 
 cupric hydrate, to the test solution, so that there may be a maxi- 
 mum amount of copper in solution and an optimum chance of 
 precipitation, without the possibility of the appearance of a black 
 deposit of cupric oxide. 
 
 In Fehling's solution Rochelle salt (potassium-sodium tartrate) 
 is the substance employed to keep the cupric hydrate in solution. 
 
 The test solution is prepared by mixing equal parts of two liquids which 
 may be conveniently referred to as " Fehling A," and " Fehling B." They 
 are best prepared in the following manner (known as Soxhlet's modifica- 
 tion). — (A) 34"64 grams of pure crystallised copper sulphate (free from 
 iron and moisture) are dissolved in distilled water, and the solution diluted 
 to 500 c.c. ; (B) 70 grams of sodium hydroxide of good quality (not less 
 than 97 per cent, of KaOH), and 175 grams of recrystallised potassium- 
 sodium tartrate, are dissolved in about 400 c.c. of distilled water, and the 
 solution made up to 500 c.c. 
 
 Although more delicate than Trommer's test, showing 0-08 
 per cent, of glucose as compared with 0-25 per cent, bj^ the latter, 
 Fehlmg's has all the faults of the other test, and to a somewhat 
 greater degree, owing to its enhanced delicacy. 
 
 Different methods of carrying out the test have been advocated 
 by different observers, with a view to minimising the chance of error. 
 Probably the most usual way is to bring a few cubic centimetres 
 of Fehhng's solution to the boil and then add the urine in small 
 quantities, boiling after each addition, until reduction occurs, or 
 an amount of urine corresponding to half the bulk of the Fehling 
 solution employed has been added. By this method the quantitj'' 
 of sugar present can be roughly estimated, the larger the amount of 
 urine required to effect reduction the lower being the ijercentage 
 of sugar ; but it has the disadvantage that the prolonged boiling 
 required when the urine is sugar-free, or only contains a trace, 
 may bring about reduction from other causes. A second method 
 is to add from ^ to 1 c.c. of the urine to about 10 c.c. of boiling 
 Fehling solution. If sugar is abundant, a yellowish or brick-red 
 opacity and deposit are produced. If no reduction occurs, traces 
 of sugar are tested for by adding 5 c.c. of the urine to a fresh supply 
 
QUALITATIVE TESTS 
 
 31 
 
 of 10 CO. of hot Fehling, heating agam to boihng, and then setting 
 aside to cool. If no turbidity appears within a minute, the urine 
 is considered to be free from sugar, or to contain a quantity so 
 small as to be of no pathological importance. If the liquid loses 
 its transparency, and passes from a clear bluish-green to an opaque 
 lightish green, the precipitation of cuprous oxide only taking 
 place as the mixture cools, it is probable that a small quantity of 
 dextrose, under 0-5 per cent., is present. A third method is to boil 
 equal parts of Fehling's solution and urine in two separate test- 
 tubes, allow them to cool for one minute, and then pour one into 
 the other. Any reduction that takes place within 5 to 10 minutes 
 is then regarded as being due to a pathological amount of sugar. 
 By this method reduction at a temperature not exceeding 60° to 
 70° C. is ensured, and the reducing effect of uric acid and creatinin 
 is excluded. Another method of guarding against this source of 
 error is to dilute the urine to a specific gravity of 1-005 (Zeehuisen), 
 or 1-012 to 1-015 (Kellas and Wethered). The same end may also 
 be attained by varying the cj^uantity of the Fehling solution em- 
 ployed according to the specific gravity of the urine, but always 
 using the same amount of urine (Kellas and Wethered). Thus : — 
 
 Specific Gravity of Urine. 
 
 Urine. 
 
 Fehling's Solution. 
 
 up to 1 -020 
 1-020-1-025 
 1-025-1-030 
 1-030-1-035 
 1-035-1-040 
 1-040-1-045 
 
 Cubic Centimetres. 
 2 
 2 
 2 
 2 
 2 
 2 
 
 Cubic Centimetres. 
 2-0 
 2-5 
 30 
 3-5 
 4-0 
 4-5 
 
 The mixture is boiled for a few seconds. If no precipitate forms 
 within two minutes, it is stated that any sugar present is of no 
 pathological significance. 
 
 At times the phosphate precipitate produced by adding the 
 alkaline Fehling solution to a urine rich in phosphates may cause 
 an ambiguous result. No change is observed in the cold, but on 
 heating a fine greenish-yellow precipitate, giving an opalescent 
 appearance to the fluid, forms. This reaction differs somewhat 
 from the definite reduction of Fehling's solution due to traces of 
 sugar, inasmuch as the phosphate precipitate soon becomes floc- 
 culent and separates out in more or less distinct masses. If the 
 urine is made alkaline vnih sodium carbonate, the phosphates may 
 be removed by filtration, and the filtrate will no longer give a 
 reaction with Fehling's solution. 
 
32 GLYCOSURIA 
 
 To avoid the effects of interfering substances when testing for small 
 quantities of sugar a variety of other methods of removing them, or mini- 
 mising their eftects, have been devised. Among these maj^ be mentioned : — 
 
 - 3 («). Allen's Method. — In this modification of Fehling's test advan- 
 tage is taken of the precipitating power of cupric acetate to remove from 
 the urine the majority of those substances which interfere with the de- 
 tection of sugar, either by themselves reducing the alkaline copper sulphate 
 solution, retaining the cuprous oxide in solution, or producing a fiocculent 
 precipitate which masks the true reaction of the sugar. 
 
 From 7 to 8 c.c. of the urine are heated to boiling, and, without sepa- 
 rating any precipitate of proteins that may form, 5 c.c. of the solution of 
 copper sulphate used for preparing Fehling's solution are added, and the 
 liquid again boiled. This produces a precipitate, principally uric acid, 
 xanthine, hypoxanthine, and phosphates. To render the precipitation 
 complete, however, it is desirable to add to the liquid, when partially 
 cooled, from 1 to 2 c.c. of a saturated solution of sodium acetate, having a 
 feebly acid reaction. The liquid is filtered, and to the filtrate, which will 
 have a bluish-green colour, 5 c.c. of the alkaline tartrate mixture used for 
 Fehling's solution are next added, and the mixture boiled for 15 to 20 
 seconds. In the presence of more than 0'25 per cent, of sugar separation of 
 cuprous oxide occurs before the boiling-point is reached, but with smaller 
 proportions precipitation takes place during the cooling of the solution, 
 which becomes greenish, opaque, and suddenly deposits cuprous oxide as a 
 fine orange-yellow precipitate. When a urine rich in sugar is under exa- 
 mination the volume employed can be advantageously reduced to 2, or 3, 
 c.c, or even less, water being added to make it up to 7 or 8 c.c. It is im- 
 portant that the sodium acetate should not be added until the liquid has 
 partially cooled, in order to avoid any chance of a reaction of the resulting 
 cupric acetate with the glucose, as in Barfoed's test. 
 
 3 (b). Seegen's Method. — This is based upon the fact that animal 
 charcoal absorbs glucose and other reducing substances from the urine, 
 but that those which interfere with the precipitation of cuprous oxide are 
 retained much longer than the sugar on washing the charcoal with water. 
 
 The urine is made into a thin paste with purified animal charcoal, and 
 left to stand for 20 or 30 minutes. The mixture is then poured on to a 
 moist filter and the urine allowed to run through. The residue on the 
 filter paper is now extracted with a quantity of water equal in bulk to the 
 urine employed for the test, and when this has filtered through, the char- 
 coal is again washed twice, with a similar quantity of water. The filtrate 
 from each washing is kept separate and tested for sugar by Fehling's (or 
 Trommer's) test. Seegen claims that a positive reaction with the second 
 or third washing is absolute proof of the presence of sugar, since these 
 washings from a normal urine will no longer reduce. 
 
 3 (c). Carneluttiand Valente recommend that 100 c.c. of the urine should 
 be evaporated to a syrup on a water-bath, 1 c.c. of 25 per cent, solution of 
 zinc chloride, previously mixed with a quarter of its volume of hydrochloric 
 acid, is added, then two volumes of absolute alcohol, and the whole allowed 
 
QUALITATIVE TESTS 33 
 
 to stand for some hours. The hquid is filtered, the residue washed with 
 alcohol, the alcohol evaporated from the solution, and the residual liquid 
 made up to 100 c.c. with distilled water. Excellent results are said to be 
 obtained with Fehling's solution by this method. 
 
 One serious disadvantage of Fehling's test is that the test 
 solution is not very stable. Apart the "A" and "B" solutions 
 keep fairly well, especially if they are protected from light and air ; 
 but when once they have been mixed the mixture soon begins to 
 deteriorate, and a more or less marked reduction will occur on 
 merely boiling the solution with distilled water. It is therefore 
 advisable that a control should be carried out with plain water 
 before using a stock solution, especially if it has been made for 
 some time. To avoid the danger of a misleading result from 
 this cause, various modifications of Fehling's solution, in which 
 the sodium potassium tartrate is replaced by some more stable 
 substance, have been devised. Two only will be mentioned here. 
 
 3 {d). Maine's Test. — In Haine's test glycerine is the substance 
 used to hold the cupric hydrate in solution. The test solution is 
 prepared as follows :— 
 
 Dissolve 30 grains (1-914 grams) of pure copper sulphate, and 4 drachms 
 (15'55 grams) of pure glycerine, in an ounce of water (28'42 c.c.) Mix this 
 solution with 3 drachms (11 '66 grams) of caustic potash dissolved in five 
 ounces (142"1 c.c.) of water. To carry out the test, about one drachm 
 (3'5 c.c.) of the solution is boiled in a test-tube, and from 5 to 8 drops, not 
 more, of the urine are added. The mixture is then again boiled. In the 
 presence of sugar a copious yellow, or yellowish-red, precipitate appears. 
 If no precipitate forms, the urine is free from any appreciable amount of 
 sugar. This solution has the advantage of being quite stable and keeping 
 indefinitely. 
 
 3 (e). Benedict's Test. —In place of the Rochelle salt of Fehling's 
 solution Benedict uses sodium citrate, and he also replaces the 
 sodium hydroxide by sodium carbonate. He points out that the 
 reducing action of glucose is dependent on the formation of a 
 substance arising from the action of the alkali on it, and that this 
 substance is destroyed by strong alkalies, such as caustic soda, 
 but not by sodium carbonate. The delicacy of the test is there- 
 fore much enhanced by the use of the carbonate in place of the 
 caustic alkali, since traces of sugar are not destroyed before they 
 can overcome the inhibiting action of the creatinin, &c., of the 
 urine. The test solution is prepared as follows : — 
 
 With the aid of heat dissolve 173 grams of sodium (or potassium) 
 citrate, and 100 grams of anhydrous (or 200 grams of crystallised) sodium 
 carbonate, in about 700 c.c. of distilled water. Filter if necessary. Dis- 
 
 C 
 
34 GLYCOSUEIA 
 
 solve 17 '3 grams of pure crystallised copper sulphate in about 100 c.c. of 
 distilled water, and pour slowlj^, with constant stirring, into the carbonate- 
 citrate solution. Cool, and make up to 1000 c.c. 
 
 To test for sugar, 5 c.c. of the solution are placed in a test- 
 tube, and 8 to 10 drops, not more, of the urine to be examined 
 are added. The mixture is heated to vigorous boiling, and is 
 kept at this temperature for one or two minutes. It is then 
 allowed to cool spontaneously. In the presence of glucose the 
 entire body of the solution will be filled with a j)recipitate, which 
 may be red, yellow, or greenish in tinge. If the quantity of sugar 
 is low (under 0-3 per cent.) the precipitate forms only on cooling. 
 If no sugar is present, the solution either remains perfectly clear 
 or shows a faint turbidit}^ that is blue in colour, and consists of 
 precipitated urates. The chief points to be borne in mind in using 
 this reagent are : (1) The addition of a small quantity of urine, 
 8 or 10 drops, to 5 c.c. of the reagent ; this being desirable, not 
 because large amounts of normal urine would cause reduction 
 of the reagent, but because more delicate results are obtained 
 by this procedure. (2) Vigorous boiling of the liquid after adding 
 the urine, and then allowing the mixture to cool spontaneously. 
 (3) If sugar is present, the solution, either before or after cool- 
 ing, "wdll be filled from top to bottom with a precipitate, so that 
 the mixture becomes opaque. Since the bulk, not the colour, 
 of the ]3recipitate is made the basis of a positive reaction, 
 the test can be as readily carried out by artificial light as in 
 daylight, even when examining for very small quantities of sugar. 
 The solution is not dark-coloured, like Fehling's solution, so that 
 the precipitate may be readily observed without waiting for it 
 to settle. 
 
 According to Benedict this solution is about ten times as 
 sensitive to sugar in urine as Fehling's or Haine's solution, and, 
 unlike these, is not appreciably reduced by creatinin, uric acid, 
 chloroform, or the simple aldehydes, but it reacts with lactose, 
 greatly increased amounts of glucuronic acid, homogentisic acid, 
 &c. The solution keeps indefinitely in uncoloured glass, or cork, 
 stoppered bottles. I have been using Benedict's solution as a 
 routine test for sugar in my laboratory for over two years now 
 with most satisfactory results. At first the findings were checked 
 hj other reduction methods, but these have been discarded for 
 some time, and I now rely on it alone as a preliminary test in all 
 cases. 
 
 4. Almen-Nylandep's (modified Bottgrer's) Bismuth Test.— 
 
QUALITATIVE TESTS 35 
 
 In the original bismuth test, as described by Bottger, the urine 
 was heated to boihng with sodium hydrate and a small xoinch of 
 basic bismuth nitrate. As it was found that alterations in the 
 alkalinity of the fluid controlled the results to a certain extent, 
 Almen and Nylander worked out a test solution which would 
 . give more constant and reliable findings. 
 
 This is prepared by dissolving 4 grams of potassium-sodium tartrate in 
 100 CO. of a 10 per cent, solution of sodium hydroxide (sp. gr. at 19° C. 
 1*115) with gentle heat, and then saturating with bismuth subnitrate 
 (about 2 grains are necessary). After cooling the solution is filtered 
 through glass-wool, and kept in a dark-coloured bottle awaj' from the 
 light. Preserved in this waj' the reagent is permanent for j-ears. 
 
 To carry out the test, about one-tenth of its volume of the 
 reagent is added to a specimen of urine, and the mixture boiled for 
 from two to five minutes. The fluid may be prevented from boil- 
 ing over by introducing a coil of platinum ^Wre, or it may be heated 
 in a boiling water-bath. If sugar is present the solution turns 
 black, and a black precipitate of bismuth settles out. Where 
 there is over 0-2 per cent, of sugar the yellow colour of Moore's test 
 is first seen. If there is no sugar, a white precipitate of phosphates 
 only \\dll appear. A very small trace of sugar will merely turn 
 the fluid brown, although it may appear black by transmitted 
 light, and will tinge the phosphate deposit a slight grey — a change 
 which is more marked on the upper surface than in the depths of 
 the deposit after it has settled. If no change is observed after two 
 minutes' boiling, the full five minutes should be allowed, as a sudden 
 darkening may take place in a urine that has previously shown 
 no change. It is most important that the reagent should be accu- 
 rately one-tenth of the volume of the urine if only traces of sugar 
 are to be detected. If the solution onh^ turns dark on cooling, 
 the test is not positive. 
 
 This test is useful as a confirmation of Trommer's, or Fehling's. 
 It is not affected by most of the more important disturbing sub- 
 stances which interfere with the reliability of those tests, and 
 gives no reaction with normal urines, yet it is dehcate enough to 
 show 0-025 per cent, of sugar. Concentrated urines may, however, 
 give a positive reaction. A considerable amount of combined gluc- 
 uronic acid will give a reaction, and the reduction that occurs after 
 the ingestion of senna, rhubarb, eucalyptus, kairin, quinine in 
 large doses, and oil of turpentine, probably depends upon the 
 presence of this substance. With rhubarb and senna the fluid is 
 brownish-red from the action of the alkali. Uroer-\i,hrin and 
 
36 GLYCOSURIA 
 
 hsematoporphyrin may give deceptive results, tinging the phosphate 
 deposit a dark brown. If the urine is ammoniacal the action of 
 the test may be interfered with, as part of the sodium hydrate is 
 consumed in replacing the ammonia, leaving the solution insuflfi- 
 ciently alkaline. One of the most important disturbing substances 
 is albumen. This produces a precipitate of sulphide of bismuth 
 which, with small quantities (0-6 j:er cent.), may be distinguished 
 by its reddish-brown colour; but large amounts of protein (1-2 per 
 cent.) yield a brownish-black precipitate which is easily confounded 
 with that due to sugar. All albumen should therefore be removed 
 before appljdng the test. The sulphur-containing compound 
 present in the urine after eating asparagus also yields a similar 
 precipitate, and is a fruitful source of error. 
 
 According to Biiickner the disturbing influences with this test 
 may be classified as follows : — 
 
 (1) Normal urinary constituents which, according to their 
 
 presence in larger or smaller amounts, may produce 
 varying degrees of reduction, or change, in colour of the 
 earthy phosphates : uroerythrin and urinary pigments 
 (urobilin particularly) when present in greatly increased 
 quantities cause a brownish discoloration of the phosphate 
 deposit, indican. 
 
 (2) Products of abnormal metabolism which effect reduction : 
 
 hexoses (dextrose, levulose, isomaltose, lactose), pen- 
 toses, glycogen, increased quantities of glucuronic acid 
 compounds, blood pigments, increased quantities of 
 hsematoporphyrin, homogentisic acid (only in concen- 
 trated solution). 
 
 (3) Drugs, or derivatives from them as the result of metabohc 
 
 changes : antipyrin, arbutin, benzoic acid, benzosol, large 
 doses of quinine, chloral, eucalyptol, glucuronic acid com- 
 pounds of drugs, indican, kairin, rheum, frangula, cascara 
 sagrada, salol, senna, sulphonal, turpentine, trional. 
 
 (4) Substances Avhich influence the sugar reaction : ammonium 
 
 carbonate, albumen in considerable amounts. 
 Modifications of the bismuth test have been devised by Briicke 
 and Maschke, but, as they possess no striking advantage, it is not 
 necessary to refer to them further. 
 
 5. Mercury Tests. — Alkaline solutions of mercury salts are 
 reduced by the sugars, giving a grey deposit of mercury, but as 
 they are almost exclusively used for quantitative work they wiE 
 be considered under that heading. (See Sachsse's and Knapp's- 
 
QUALITATIVE TESTS 37 
 
 methods.) They are open to the same fallacies as copper solu- 
 tions, and have no compensating advantages. 
 
 6. Picric Acid (Braun). — Picric acid (Trinitro-phenol) was at 
 one time much used as a test for sugar, and its advantages were 
 strongl}' insisted upon by Sir G. Johnson. 
 
 The test is carried out by mixing equal volumes of the sus- 
 pected urine and a saturated solution of picric acid, and then 
 adding one-fourth the volume of a 6 per cent, solution of potassium 
 hydrate (Liq. potasses). If much sugar is present, a well-marked 
 orange-red colour develojDS. On boiling the mixture the colour 
 deepens, the extent of the change cle]3ending upon the sugar-content 
 of the urine. If there is a considerable amount the liquid becomes 
 an intense bro'WTiish-red, so deep as to be almost opaque ; but 
 with small quantities the colour is a cherry-red, which it is not 
 easy to distinguish from the similar coloration given by many 
 normal urines. It is advisable to compare the result with a control 
 carried out with a urine knowoi to be free from sugar. Allen states 
 that a serviceable and permanent reagent may be prepared by 
 mixing two volumes of a cold saturated solution of picric acid 
 with one volume of a normal caustic soda solution (4 per cent.), 
 disregarding any crystals that separate out. 
 
 Picric acid gives no reaction with uric acid, urates, and mucin ; 
 but with creatinin, creatin, and glucuronic acid it gives a very 
 similar reaction to glucose. Normal urines, therefore, yield a red 
 coloration with an alkaline solution of picric acid, even in the 
 cold, and this is intensified by boiling, so that it is difficult to be 
 sure of the presence of traces of sugar. If the creatinin is removed 
 by treating the urine with 25 per cent, of its volume of a cold 
 saturated solution of mercuric chloride, and 5 per cent, of a cold 
 saturated solution of sodium acetate, boiling for five minutes and 
 filtering hot, acidulating the filtrate with acetic acid, boiling for 
 ten minutes with zinc dust, and again filtering to remove the 
 mercury, the picric acid test will indicate very small quantities of 
 glucose. 
 
 7. Indig-O Test (Hoppe-SeyleP). — An alkahne solutioniof ortho- 
 nitro-phenylpropiolic acid is reduced on heating with a solution 
 of glucose to indigo blue. 
 
 The test solution is prepared by dissolving 5 "76 grams of the acid in 
 100 c.c. of 10 per cent, sodium hydrate solution, and diluting to one litre 
 with water. 
 
 Five cubic centimetres of this solution are heated with ten drops 
 of urine for quarter to half a minute when, if at least 0-5 per cent. 
 
38 GLYCOSURIA 
 
 of sugar is present, a blue coloration is seen. A high, percentage of 
 sugar may, however, cause the indigo-blue to be still further re- 
 duced to indigo- white, but on shaking the fluid with air it will give 
 a blue foam. Traces of indigo may be detected by shaking the 
 solution with chloroform, which will give a blue solution if indigo 
 is present. It is important that not more than ten drops of the 
 urine should be used, as it is found that a larger amount (20 drops) 
 will give a slight reaction with normal urines. Concentrated urines 
 should be diluted before applying the test. 
 
 It is said that no other substances than the sugar commonly 
 met with in the urine give any reaction if the test is correctly and 
 carefully carried out. Over 2 per cent, of albumen causes the solu- 
 tion to assume a dark-red colour. 
 
 A solution of indigotin-disulphonic acid (Mulder) saturated with 
 sodium carbonate, and boiled with the urine, will turn successively 
 green, purple, red, and yellow if sugar is present. On shaking the 
 warm solution the colour changes are reversed. Glucuronic acid, 
 inosite, gallic, tannic, and salicylic acids and their compounds 
 will give a similar reaction. 
 
 8, Aniline Dye Tests. — Various coal-tar dyes are decolorised 
 on being heated with alkaline solutions of dextrose and other 
 reducing agents. Wender, Crismer, and others have utilised this 
 fact for the recognition, and estimation, of sugar in the urine. The 
 former employed methylene blue, and the latter safranin. 
 
 8 (a). Methylene Blue Reaction (Wender).— A solution of 
 methylene blue is prepared by dissolving 1 gram in 3000 c.c. of 
 distilled water. Six cubic centimetres of this solution are mixed 
 with 2 c.c. of a normal solution of caustic potash (5-6 per cent.). 
 The urine is diluted ten times with water, and 2 c.c. of the dilution 
 are added to the alkaline solution of methylene blue. The mixture 
 is then boiled for one or two minutes, avoiding agitation, or contact 
 with the air, as much as possible. If the urine contains 0-5 per 
 cent., or more, of sugar the blue colour \\ill be completely dis- 
 charged. By using different jsroportions of urine, and heating in 
 a series of test-tubes, in a water-bath, an indication of the amount 
 of sugar present may be obtained. 
 
 8|(&). Safranin Test (Crismer).— Two cubic centimetres of 
 solution of safranin, made bj^ dissolving 1 gram of the dye in a Htre 
 of distilled water, are mixed with 2 c.c. of a normal solution of 
 caustic soda (4 per cent), or caustic potash (5-6 per cent.), and 2 c.c. 
 of the urine added. The mixture is then boiled. If glucose is 
 present to the extent of 0-1 per cent, the red is changed to a pale 
 
QUALITATIVE TESTS 39 
 
 yellow colour, and the liquid becomes turbid from the separation 
 of the insoluble leuco-derivatives. If the sugar is not present in 
 considerable excess the red colour returns on agitating the liquid, 
 or exposing it to the air. 
 
 Urines containing a high percentage of sugar should be well 
 diluted before applying the test. Safranin in alkaline solutions 
 is not decolorised when heated by creatinin, creatin, uric acid, 
 urates, chloral, chloroform, hydrogen peroxide, or salts of hydro- 
 xylamine, or by mucin. It is only slowly affected by albumen. 
 
 The great objection to the safranin, and the methylene blue, 
 tests is that they are so sensitive that normal urines, as a rule, 
 give a distinct reaction. Kellas and Wethered found that with 
 the safranin test a reduction corresponding to 0-07 to 0-08 per cent, 
 of glucose was given by normal urines. They consider, however, 
 that, in spite of this drawback, it is the most convenient and reliable 
 test for sugar that can be applied in the present state of our know- 
 ledge, especially if a reduction due to the presence of glucuronic 
 acid is excluded in doubtful cases, and that, used as an auxiliary 
 to Fehling's test, it is sufficient to settle many troublesome cases 
 where small quantities of sugar, and large quantities of creatinin, 
 cause the findings of the latter to be uncertain. 
 
 9. Phenylhydrazin Test (Fischer, v. Jaksch).— The applica- 
 tion of phenylhydrazin as a reagent for the detection of sugar in 
 the urine marked an epoch in the investigation of glucosuria and 
 allied conditions, for not only is it more delicate, sho\^'ing 0-01 per 
 cent, of sugar, than any test previously employed, but it is unaffected 
 by other substances, such as creatin, creatinin, hippuric acid, 
 homogentisic acid, and excess of uric acid and urates as met ^\ith 
 in human urine, that are liable to give rise to difficulties when 
 most other methods are relied upon. So delicate is it that, theo- 
 retically, the small amount of sugar in normal urine should give 
 a reaction, but practically this is not found to be the case, unless 
 a special technique is followed. Zunz states that osazone crystals 
 are not found unless the urine also reduces Fehling's solution to 
 some extent. On the other hand, the test may fail even when the 
 urine is known to contain sugar, if it is not carefully carried out. 
 Success depends chiefly (1) upon the purity of the phenylhydrazin. 
 hence the more stable hydrochloride is generally preferred ; (2) the 
 amounts of the reagents used, theoretically 1 part of sugar, 2 of 
 phenylhydrazin, and 3 of sodium acetate are best ; and (3) the time 
 allowed to cool. Even under the most favourable circumstances 
 all the sugar is not precipitated. From a 5 per cent, solution of 
 
40 GLYCOSURIA 
 
 glucose, Fischer found that the maximum precipitate represented 
 from 85 to 90 ]Der cent, of the sugar in the solution. Since albumen 
 interferes "with the separation of the crystals, it should be removed 
 before applying the test. 
 
 The method of performing the phenylhydrazin test described 
 bj^ von Jaksch has been much modified by subsequent writers, 
 and, since the physical conditions under which it is carried out 
 materially affect the result, it will be necessary to refer to the 
 chief variations proposed. 
 
 9 (a). Von Jaksch originally recommended that 50 c.c. of the 
 urine to be tested should be mixed with 2 grams of sodium acetate, 
 and from 1 to 2 grams of phenylhydrazin hydrochloride, and that 
 the mixture be heated in a water-bath for twenty minutes to 
 half an hour. If glucose is present, the osazone then separates 
 out, on cooling, as an amorphous, or crystalline, deposit of a yellow 
 or reddish colour. If amorphous, the precii^itate can be recovered 
 in a crystalline form by filtering it off, washing with distilled water, 
 dissolving the residue on the filter in hot 50 per cent, alcohol, 
 diluting the solution wdth w^ater, boiling to expel the alcohol, 
 and cooling. 
 
 9 (6). Two drops of a concentrated solution of lead acetate 
 are added to 10 c.c. of the urine, and the precipitate filtered off. 
 One drojo of acetic acid, or enough to acidify the filtrate, a piece 
 of phenylhj'drazin hydrochloride the size of a pea, and sodium 
 acetate the size of a bean, are then added, and the mixture boiled 
 on a water-bath for from one to two hours. It is then filtered 
 hot, returned to the w^ater-bath, and allowed to cool slowdy. 
 
 9 (c). These methods of performing the test take time, and 
 require the use of sj)ecial apparatus Avhich is not always available. 
 To obviate these difficulties, and render the reaction convenient 
 for clinical use, it has been suggested that 0-5 grams of phenyl- 
 hydrazin hj'drochloride, and 1-5 grams of sodium acetate, should 
 be dissolved by gentle heat in a few cubic centimetres of water in a 
 test-tube, and then 5 to 10 c.c. of the urine added. The mixture is 
 brought to the boiling-point, and maintained there for three minutes 
 with strong, and five minutes with weak, solutions of sugar. The 
 test-tube is then set aside to cool, and the deposit examined for 
 osazone crystals in five or ten minutes. In my experience this 
 rapid method of performing the phenylhydrazin test gives a re- 
 action with all sugars when 10 c.c. of the urine is used, and the 
 heating continued for at least five minutes. In the water-bath, 
 however, even an hour is not sufiicient to obtain a satisfactory 
 yield with maltose, lactose, and pentose, an hour and a half, or 
 
QUALITATIVE TESTS 41 
 
 even two hours, being required to demonstrate their presence 
 satisfactorily, especially when the solution is weak. 
 
 9{d). Some authorities, folloMdng E. Fischer, have preferred 
 to use phenylhydrazin and not the hydrochloride, but it has the 
 disadvantage of not keeping well. It should be almost straw- 
 coloured, and is conveniently kei^t in sealed bottles containing 
 ■only a small quantity, which can be quickly used when once opened. 
 A knife-point of sodium acetate is added to 10 c.c. of the urine, then 
 1 to 2 c.c. of 10 per cent, acetic acid, and 5 drops of pure phenyl- 
 hydrazin. The mixture is heated in the water-bath, or over the 
 free flame, in the same way as when the hydrochloride is employed. 
 
 9 (e). A modification of this method, suggested by Kowarski, 
 which gives very satisfactory results, and is more delicate, consists 
 in mixing 5 drops of pure phenylhydrazin in a test-tube with 
 10 drops of acetic acid, gently shaking, and then adding about 
 1 c.c. of a saturated solution of sodium chloride. To the soHd 
 mass that forms is added 3 to 5 c.c. of the urine, and the test-tube 
 is then heated, in the free flame, for two minutes after its contents 
 begin to boil. On cooHng, the osazone crystals separate from urines 
 containing over 0-2 per cent, of sugar in one minute, and from 
 weaker solutions in about five minutes. 
 
 Beside giving a crystalline osazone with the sugars dextrose, 
 levulose, lactose, maltose, isomaltose, and the pentoses (arabinose 
 and xylose) met with in the urine, phenylhydrazin also forms a 
 compound with glucuronic acid, and exceptionally with acetone, 
 aceto-acetic acid, oxalic acid, and in very concentrated urines 
 with uric acid. As human urine is relatively poor in uric acid, 
 the last named does not call for further remark. The acetone 
 compound occurs as needles which melt to oily globules at 16° C, 
 while the oxalic acid salt separates as insoluble, glancing, colour- 
 less plates which melt at 172° to 173° C. 
 
 Alkaline salts of glucuronic acid form a compound with phenyl- 
 hydrazin, which slowly separates on cooling as yellow needles, 
 usually of a somewhat darker colour, of smaller size, and more 
 irregular shape, than the osazones of the sugars. Although the 
 dejDosit is small, it maj^ be easily mistaken for traces of sugar 
 compounds, and so give rise to an incorrect diagnosis. It is this 
 mistake that has to be chiefly guarded against in using the phenyl- 
 hydrazin test for diagnostic purposes. Some compounds of glucu- 
 ronic acid undergo the decomposition that must occur before it 
 eombines with phenylhydrazin more easily than others, so that 
 the readiness with which the reaction takes place varies. A urine 
 •containing a compound such as urochohc acid, which readih' 
 
42 GLYCOSURIA 
 
 splits up, quickly responds to the test ; but the more resistant- 
 phenol and indol compounds require prolonged treatment before 
 they react. When the reaction takes place in a solution containing 
 a free acid, the phenylhydrazin compound is precipitated in the= 
 form of dark-brown granules, and does not assume the crystalline 
 form. It has been stated that prolonged heating tends to cause 
 the glucuronic acid compound to be j)recipitated on cooling as a. 
 brown amorphous mass and not in the crystalline form (Purdy), 
 w^hich is liable to be mistaken for the osazone of a sugar. In the 
 examination of a hundred normal urines I sought to test the truth 
 of this statement, and found that, when heated in the water-bath. 
 for an hour, four of them showed a crystalline deposit, while by 
 boiling in the free flame for five minutes, six specimens gave a^ 
 positive result. Using the same methods, but shortening the period 
 of heating to twenty minutes and two minutes respecfcively, exactly 
 the same results were obtained. So that the mere time, or method, 
 of applying heat cannot be relied upon to differentiate glucuronic 
 acid from the sugars. In practice the phenylhydrazin compounds 
 of glucuronic acid and the sugars can only be differentiated by 
 experience of the different appearance of their crystals, which is, 
 however, a somewhat fallacious guide, and by a consideration of 
 their physical and chemical characters. These A^ill be fully dealt 
 with when the classifying tests are considered. 
 
 9 (/). A modification of the phenylhydrazin test, which is useful 
 when no microscope is at hand to examine the osazone crystals, 
 is carried out as follows : — 
 
 Eiegler's Modification. — 0-1 grams of phenylhydrazin hydrochloride, and' 
 0'5 grams of sodium acetate, are dissolved in a few drops of water, and 
 boiled for several minutes with 1 c.c. of the urine ; 1 c.c. of caustic soda 
 (10 per cent.) is then added. If sugar is present a deep red-violet colour 
 should appear within five minutes. The test is best carried out in a 
 porcelain dish, as the colour changes are better seen, and appear as rings,, 
 or streaks, that are very striking. The test is said to react with a solution 
 containing 0-01 per cent, of glucose. Other sugars, and formaldehyde, also 
 give the reaction. As the solution turns rose-red from oxidation on standing 
 for an hour or so, it is essential that a distinct colour reaction should be ob- 
 served in five, or at most ten, minutes after the addition of the soda solution. 
 
 II. Classifying Tests 
 
 1. Fermentation. — The fermentation test serves to distinguish 
 the fermentable sugars from those reducing substances that are 
 not attacked, or only with difficulty, by ordinary brewer's yeast. 
 It is important, however, that a time limit of six to twelve hours, 
 should be set in which fermentation at a definite temperature 
 
QUALITATIVE TESTS 43 
 
 (34° to 36° C.) should occur, as otherwise gas may appear as a 
 result of other causes and lead to a mistaken diagnosis. The test 
 is performed as follows : — 
 
 A piece of compressed yeast, about the size of a large pea, is rubbed up 
 with 25 to 30 c.c. of the urine and the mixture poured into a test-tube until 
 it is filled to the brim. The tube is then closed with a perforated cork 
 which carries a V-shaped piece of glass tubing, and is placed in a beaker 
 and kept at a temperature of 34° to 36° 0. in the incubator (or a pro- 
 perly constructed fermentation apparatus may be used). If the urine 
 contains a fermentable sugar, gas will accumulate in the upper end of the 
 tube and expel a corresponding amount of urine into the beaker. To 
 prove that the gas is carbon dioxide, the cork should be removed from the 
 end of the tube vinder mercury or water, and a small quantity of caustic 
 soda introduced. It will then be seen that, owing to the absorption of the 
 gas, the liquid will again rise to the top of the tube. The presence of 
 alcohol may be shown by distilling the urine and testing the distillate with 
 iodine and caustic potash for iodoform. Two control tests should always 
 be carried out : (1) One with a normal urine, to which a little dextrose and 
 yeast have been added, to prove the activity of the yeast ; (2) another with 
 normal urine and the yeast alone, to show that there is no gas formation 
 apart from fermentation. The last test is essential, as compressed yeast 
 often develops gas with normal specimens, particularly if the urine is only 
 faintly acid and the time of fermentation is prolonged. This is due to 
 ammoniacal changes brought about by bacteria contaminating the yeast, 
 which develop carbonic acid from the ammonium carbonate derived from, 
 the urea. The ammoniacal fermentation is slower in its development than 
 the alcoholic fermentation of the sugars, and can be suppressed by boiling 
 the urine to sterilise it. Boiling also frees it from air, which is another 
 possible source of fallacy. Some prefer to add 1 per cent, sodium fluoride to 
 the urine, or to render it distinctly acid with tartaric acid, boil for a few 
 minutes, and then cool, before applying the test. Some samples of yeast 
 contain traces of sugar, and give rise to gas formation from the fermenta- 
 tion of this. By washing the yeast the sugar may be removed, and its 
 presence proved by the washings responding to the tests for sugar. As a 
 rule, however, compressed yeast is free from sugar. Certain samples of 
 yeast give rise to gas formation by what is termed " self-fermentation," a 
 process apparently related to the amount of contained glycogen. Both 
 these sources of error are detected by carrying out a control test. 
 
 By this test the sugars and other reducing substances occurring 
 in the urine may be divided into the following groups : — 
 
 (1) Those that are quickly fermented with brewer's yeast 
 
 (6 to 12 hours) : dextrose, levulose, maltose. 
 
 (2) Those that are slowly broken down and fermented (20 to 
 
 30 hours) : cane-sugar, lactose, galactose, isomaltose. 
 
 (3) Those that show no gas formation (in 20 to 30 hours) : pen- 
 
 toses, glucuronic acid, laiose, dextrin, glycogen, homo- 
 gentisic acid, inosite, uric acid, creatin, and creatinin. 
 
44 GLYCOSURIA 
 
 2. The Polapiscope. — The specific rotatory powers of the sugars, 
 and related reducing substances, occurring in the urine divides 
 them into three classes :— 
 
 (1) Those that are dextro-rotatory (dextrose, galactose, lactose, 
 
 maltose, isomaltose, and 1-arabinose). 
 
 (2) Those that are levo-rotatory (levulose, laiose, and most 
 
 compound glucuronates). 
 
 (3) Those that are ojitically inactive (i-arabinose). 
 
 A determination of the specific rotatory power of the urine, 
 especially when this is carried out quantitatively and is compared 
 mth its reducing power, helps in the detection of the particular 
 variety of reducing substances present. In the recognition of 
 traces of sugar care must be exercised, since normal urines are 
 slightly levo-rotatory (about 0-05° to 0-17°), and occasionally 
 urines are dextro-rotatory when sugxr is absent from the presence 
 of glucuronic acid compounds (Borntrager in two morphia habitueB). 
 The presence of albumen interferes with the recognition of sugars 
 by the polariscope, since it is levo-rotatory, and hence may cover 
 a slight dextro-rotation due to that cause. Cystin, oxybutyric 
 acid, and most paired glucuronates are also levo-rotatory and may 
 have a similar effect. All these substances are not fermented 
 by yeast, so that the polarimetric reading due to their presence 
 is the same after as before fermentation ; but if a urine contains 
 a fermentable sugar the reading is lowered by fermentation. 
 When a urine contains both a dextro-rotatory and levo-rotatory 
 fermentable sugar {e.g. dextrose and levulose), and these alone, 
 the polarimetric estimation will give a smaller value than that 
 obtained by titration, and when it is fermented its rotatory power 
 should be nil, or only equal to that of a normal urine. If, how- 
 ever, there is also an unfermentable levo-rotatory substance, such 
 as /3-oxybutjTic acid, present, not only will the polarimetric and 
 titration reachngs not agree, but the urine will still be levo-rotatory 
 after fermentation, but this will not be sufficient to entirely account 
 for the different results obtained with the polariscope and by 
 titration. The presence of paired glucuronic acid may have a 
 similar effect. The pentose usually met with in the urine in chronic 
 pentosuria is optically inactive (i-arabinose), but Luzzato has de- 
 scribed dextro-rotatory 1-arabinose as being present in one case. 
 The latter is also met with in alimentary pentosuria. Since the 
 degree of polarisation induced by maltose is much more intense 
 than that produced by dextrose, a small quantity of the former 
 may give the same reading as a much larger quantitj^ of the latter ; 
 but the reducing power of maltose is increased by hydrol^J'sis with 
 
QUALITATIVE TESTS 45 
 
 a dilute acid, while that of glucose is unchanged, or even diminished, 
 from the formation of human substances. 
 
 3. Bapfoed's Test. — When carried out under certain conditions 
 this test serves to distinguish (1) the monosaccharides (dextrose, 
 levulose, and galactose) from (2) the reducing disaccharides (lactose 
 and maltose). 
 
 From five to fifteen drops of the urine are mixed with 5 c.c. of the 
 test solution (made by dissolving 13-3 grams of crystallised neutral 
 copper acetate in 200 c.c. of 1 per cent, acetic acid), and boiled in a water- 
 bath for from 3 to 5 minutes. The members of the first group will reduce 
 the solution, giving a yellow or red precipitate within the times men- 
 tioned, but the disaccharides cause no change. 
 
 4. PhlOPOg-lucin Test. — This test is chiefly used to detect the 
 presence of pentoses, but it is also given by glucuronic acid, and, 
 as regards the colour change, by lactose and galactose. It is there- 
 fore not specific, but is merely a classifying test, helping to dis- 
 tinguish these substances from the other sugars. 
 
 According to Wheeler and ToUens it is carried out as follows. 
 To a few cubic centimetres of the lu-ine are added an equal quantity 
 of fuming hydrochloric acid (sp. gr. 1-19), and from 25 to 30 milligxams 
 of phloroglucin. The solution is warmed until a red coloiir develops. 
 On examination with the spectroscope the presence of a pentose, or 
 glucuronic acid, is shown by the appearance of a band between D and E 
 (yellow and green). If this is not found the solution is brought to the 
 boil and again examined with the spectroscope. Lactose and galactose 
 give the red coloration, but do not show the band on spectroscopic 
 examination. Normal urines frequently give a doubtful reaction from 
 the presence of glucuronic acid. 
 
 Salkowski recommends the following modification. Five or six cubic 
 centimetres of fuming hydrochloric acid are warmed and saturated with 
 phloroglucin, leaving a little undissolved. This solution is divided into 
 two parts. To one-half is added 0-5 c.c. of the urine to be examined, 
 and to the other the same amoimt of a normal urine. Both are placed in 
 a beaker of boiling water. A positive reaction is shown by the appear- 
 ance of an intense red colour, which begins above and extends down- 
 ward, in the mixture containing the suspected urine, while the control 
 exhibits no marked colour change. Examination with the spectroscope 
 gives the same result as in the preceding method. The colour change 
 is better seen if the urines are decolorised by being warmed with animal 
 charcoal, and filtered, before commencing the test. The specimens 
 should be removed from the water-bath as soon as the colour has 
 developed distinctly, as prolonged heating interferes with the clear- 
 ness of the reaction. If direct examination of the liquid with the 
 spectroscope is negative the solution should be extracted with amyl 
 alcohol, and this extract be examined spectroscoiDically. 
 
46 GLYCOSURIA 
 
 5. Physical and Chemical Characters of the Phenylosa- 
 ZOnes. — The phenylhydrazin compounds of the sugars possess 
 certain physical and chemical characters by which they can he 
 more or less readily differentiated. 
 
 The chief of these are : (1) The rate of osazone formation, 
 (2) the microscopical characters of the crystals ; (3) the solubilities 
 of the osazones in various reagents ; (4) the specific rotatory power 
 of these solutions ; (5) the melting-points of the purified products ; 
 (6) their percentage content of nitrogen. 
 
 1. The rate of osazone formation is a point of considerable 
 importance in determining the variety of sugar present in a par- 
 ticular solution. Under experimental conditions, with solutions 
 of definite strength, exact time limits for the appearance of the 
 osazone can be laid down. Although this is not feasible with 
 such a liquid as urine, where the proportion of sugar present is 
 unknown, valuable information can be obtained by observing the 
 conditions and rate of osazone formation. The compound formed 
 by dextrose and le\ailose separates from the hot solution after a 
 comparatively brief interval, the former in most cases with charac- 
 teristic suddenness. The osazones of maltose, lactose, and the 
 pentoses only form after much more prolonged heating, and 
 although crystals of pentosazone may eventually separate from 
 the hot solution, maltosazone and lactosazone never appear until 
 the fluid cools, no matter how long it may be boiled. 
 
 2. Microscopical examinaiion of the crystalline deposit ob- 
 tained after treatment mth phenylhydrazin shows certain dif- 
 ferences in the characters, size, and arrangement of the crystals 
 which are suggestive. The large yellow needles arranged in sheaves 
 and rosettes yielded by dextrose are well known. Levulosazone 
 resembles glucosazone, but there is less tendency to rosette forma- 
 tion, and the crystals are somewhat longer and more slender. 
 Lactosazone occurs as spherical masses of crystals resembling a 
 shaggy yellow chrysanthemum. At the periphery the separate 
 crystals can be distinguished, and are seen to be slender, flexible, 
 and hair-like, but in the centre they are felted together into a 
 brown semi-opaque mass. In preparations made by the rapid 
 method the individual crystals are usually not as distinct as those 
 formed after prolonged heating in the water-bath, the aj)pearance 
 j)resented being that of a brown central boss surrounded by a 
 light yellow halo showing " fine " radial striations. The crystals 
 of maltosazone are short, stiff, and sword-hke, and are arranged 
 in small rosettes when j^repared by the water-bath method. Pre- 
 parations made by heating in the free flame are less characteristic, 
 
QUALITATIVE TESTS 47 
 
 consisting of narrow crystals grouped in small bushy sheaves. 
 Isomaltose yields masses of aggregated flexible needles of a golden- 
 yellow colour, mostly arranged in spheres. The pentose crystals 
 are silky, tangled, curved needles, generally arranged in rosettes. 
 The shape, size, and arrangement of the crystals is influenced to 
 a certain extent by the manner in which the test is carried out, 
 and by the rate of cooling, so that not only is some experience 
 of typical preparations required in forming an opinion, but the 
 physical conditions under which the sample under examination 
 was prepared must also be taken into account. In some cases the 
 crystalline form can only be determined satisfactorily after the 
 specimen has been purified by recrystalHsation from alcohol, boihng 
 acetone, pyridin, &c. 
 
 3. The osazones can be differentiated to a certain extent by 
 their solubilities. Thus pentosazone is readily soluble in water 
 at 60° C, but dextrosazone is almost insoluble. One part of 
 lactosazone dissolves in eighty to ninety of boiling water, while 
 maltosazone is still more soluble (1 in 70 to 75). Isomaltosazone 
 dissolves one part in four of water at 100° C. 
 
 Dextrosazone is only very slightly soluble in cold methyl 
 alcohol, levulosazone is a little more soluble, but the pentosazones 
 and glucuronic acid compound readily dissolve. All the osazones 
 are soluble in hot 50 per cent, alcohol, and advantage is taken of 
 this fact to prepare them in a pure form. 
 
 I have found that the rate of solution of the osazones in dilute 
 sulphuric acid is of some value in distinguishing the johenylhydrazin 
 compounds of dextrose and levulose from those of other carbo- 
 hydrates. On irrigating a preparation of the latter with a 33 per 
 cent, solution of the acid, the crystals turn brown and dissolve 
 rapidly, while dextrosazone and levulosazone only slowly assume a 
 brown coloration and take several minutes before they disappear. 
 
 4. Specific Rotatory Power. — When a solution of 0-2 gram of 
 the purified osazone is dissolved in 4 grams of pjTidine and 6 grams 
 of absolute alcohol and examined with the polariscope in a 100 
 mm. tube, the nature and degree of rotation is found to vary 
 with the sugar. 
 
 Three are dextro-rotatory . . 1-arabinose ( + 1'1°), galactose ( + 0'48°), 
 
 maltose (-1-1 -3°). 
 
 Four are levo-rotatory . . . 1-xylose ( -0"5°), dextrose ( — 1-3°), levu- 
 lose ( — 1"3°), and mannose (— TS'). 
 
 One is inactive .... lactose (±0"00°). 
 
 Solutions of dextrosazone, maltosazone, &c., in glacial acetic 
 
48 GLYCOSURIA 
 
 acid are levo-rotatory, but a solution of galactosazone is optically 
 inactive when it contains less than 4 per cent. ; over that amount 
 it is faintly levo-rotatory. 
 
 Examination of the crystals mounted in the mother liquor, 
 with polarised light, under the microscope, also helps to distinguish 
 the osazones of the reducing sugars from glucuronic acid crystals, 
 for the former stand out bright, and aj)pear green, red, &c., while 
 the latter are dark and uncoloured. 
 
 5. The melting-point of the crystals obtained from urine by 
 the phenylhydrazin reaction is one of their most useful and charac- 
 teristic properties. The product employed for the purpose must, 
 however, be pure, or doubtful and misleading results will follow. 
 Thus the melting-point of pure dextrosazone is 204° to 205° C, but 
 the impure crystals obtained direct from the urine generally melt 
 somewhere between 173° and 194° C. Levulosazone melts at the 
 same temperature as dextrosazone. The melting-point of the 
 phenylhydrazin compound of lactose is about 210° C, and of 
 maltose 206° to 207° C. Pure galactosazone melts at 194° to 197° C, 
 but when separated from the urine, at 171° to 174° C. Isomaltose 
 begins to form drops at 140° to 145° C, melts at 150° to 153° C, 
 and blackens at 200° C. The osazone of 1-arabinose, in a pure form, 
 melts at 160° C, and r-arabinose at 166° to 168° C, but as obtained 
 from the urine the melting-point lies between 156° and 158° C. 
 As the osazones undergo decomposition on prolonged heating,, 
 it is necessary that the temperature should be rapidly raised at 
 first, and then gradually increased as the point of fusion is ap- 
 proached, or charring of the specimen may obscure the change of 
 state. 
 
 6. The formula of the osazones formed by the monosaccharides- 
 dextrose andlevulose is C^gHooN^O^, while that of the dissaccharides 
 lactose, maltose, and isomaltose is Cg^HgoN^Og. Hence the former 
 may be expected to yield 15-64 per cent., and the latter 10-76percent. 
 of nitrogen. In practice slightly lower readings are found to be the 
 rule, about 15-58 per cent, being obtained for dextrose and levulose, 
 and 10-67 per cent., or thereabout, for the disaccharides. The 
 percentage of nitrogen contained in the pentosazones is 17-07, as the 
 formula is Cj^-HooN^Og. By determining the percentage of nitrogen 
 in an osazone it is therefore possible to decide to which of these 
 three classes of carbohj^drate the sugar belongs. A satisfactory 
 determination is, however, only possible when a sufficient amount 
 of the pure product and the necessary apparatus are available for 
 a combustion experiment, as Kjeldahl's process is useless for the. 
 purpose, the separation of the nitrogen not being complete. 
 
QUALITATIVE TESTS 49 
 
 III. Confirmatory or Special Tests 
 
 In the preceding pages the means by which sugars, and other 
 reducing substances, occurring in the urine can be detected, and 
 the tests by which these can be classified, have been described. 
 A careful consideration of the results of the classifying tests in 
 any particular instance will have indicated which particular sub- 
 stance, giving the tests of the first group, is probably present. 
 The special or confirmatory tests by which a definite diagnosis 
 can be made now remain to be dealt with. 
 
 Most of the confirmatory reactions, although more or less 
 specific, are not absolute, and much depends on the way that 
 they are carried out. In some instances, too, a satisfactory result 
 is only obtained when the sugar, or other reducing substance, has 
 been isolated from the urine and the test is applied to a pure solu- 
 tion. We shall therefore first briefly describe the methods usually 
 employed for isolating sugars from the urine. 
 
 A. When the urine contains from 5 to 10 per cent, of dextrose mere 
 evaporation on a water-bath, to the consistency of a syrup, will often cause 
 it to separate out in tabular crystals, or irregular w^arty masses, on cool- 
 ing and standing for some days. More frequently it is deposited in 
 crystals consisting of a compound of dextrose with soditun chloride, 
 which is more soluble in water, but less soluble in alcohol than glucose 
 itself. Sometimes, however, the sugar will not separate out, even when 
 the syrup is left at rest for many days. In siich eases treatment with 
 ether, about half the volume of the syrup, which is subsequently allowed 
 to evaporate spontaneously, will induce the separation of the crystals. 
 The syrujD, and any glucose crystals that it may contain, are filtered 
 off, and pvirified from urea and extractive matter by treatment with a 
 small amount of cold absolute alcohol, which leaves most of the sugar 
 undissolved. The residue is then boiled with absolute alcohol, which 
 dissolves the sugar and leaves an insoluble residue of sulphates, phos- 
 phates, and urates. The hot alcohol extract is filtered, and evaporated 
 to a small bulk. On cooling crystals of dextrose separate out. These 
 may be purified by recrystallisation from methyl-alcohol. Levulose 
 can be separated from dextrose by treating the syrup with lime, with 
 which the former forms an insoluble compovmd. This can be separated 
 from the soluble dextrose compound by filtration, washed, and decom- 
 posed with oxalic acid. 
 
 B. Precipitation by metallic salts. 
 
 (a) Lead. — Carbohydrates, and most of the oxidation, and reduction 
 products of carbohydrates, form compounds with lead by which they 
 may be isolated, and by a process of fractional precipitation it is pos- 
 sible to separate them to a certain extent. The separation, however, 
 is not quite sharp, particularly^ in mixtvires and impiure solutions, 
 
50 GLYCOSURIA 
 
 such as iirine. For the isolation of small quantities of sugar from the 
 tirine the following procedure may be adopted (Briicke, Pavy). A 
 large quantity of the nrine is treated with half its voluine of a 10 per 
 cent, solution of neutral lead acetate. The precipitate that forms (1) is 
 filtered off, and the filtrate treated with basic lead acetate, any further 
 precipitate being also filtered off (2). The filtrate from this is treated 
 with ammonia and a further supply of basic lead acetate, unless a 
 distinct excess of the latter has been already used. The precipitate 
 that forms (3) is filtered off, and the filtrate again treated with basic 
 lead acetate and ammonia. Any precipitate that forms is separated 
 by filtration (4). The precipitate (1) contains the urates, uric acid, 
 xanthin, sulphates, phosphates, and colouring matter, beside glycogen 
 and part of any levtilose that may be present. The precipitate produced 
 by basic lead acetate in the acid lu-ine (2) contains glucuronic acid, 
 laiose, and glycogen. The third (3) and fourth (4) precipitates con- 
 tain the sugars (dextrose, levulose, maltose), beside some glucuronic 
 acid and laiose which were not completely precipitated in the acid 
 solution. Galactose and lactose are only incompletely precipitated, 
 even by basic lead acetate in the presence of ammonia. To separate 
 the sugars, &c., from their combination with lead, the mixed third and 
 fourth precipitates are washed with distilled water until the wash 
 water is neutral, or only faintly alkaline. The residue is then suspended 
 in water and decomposed with a stream of sulphuretted hydrogen 
 (sulphiu"ic, or oxalic acid, may also be used, but the product is more 
 highly coloured). The liqioid is now cautiously treated with sodium 
 carbonate until just neutral, when the coloiiring matter separates out 
 and may be filtered off. Further decolorisation may be effected by 
 slightly acidifying the liquid with acetic acid, and shaking with a little 
 animal charcoal (previously freed from phosphates, &c., by boiling 
 with hydrochloric acid and washing until the washings are no longer 
 acid to litmus). The sugars may be recovered from this liquid by 
 evaporating and crystallising, or the solution itself may be used for the 
 necessary tests. 
 
 To separate lactose and galactose from the lorine, it is saturated with 
 lead acetate and filtered, the filtrate treated with ammonia, and the 
 precipitate that forms washed with distilled water. The filtrate, and 
 wash water, are again treated with lead acetate and ammonia, and the 
 process repeated until a filtrate is obtained that is no longer dextro- 
 rotatory. Any resulting precipitates are added to that first obtained. 
 The combined precipitates are suspended in water and treated with a 
 stream of sulphuretted hydrogen. The lead sulphide is filtered off, 
 the optionally active filtrate is shaken with silver oxide, filtered, and 
 the dissolved silver removed with sulphuretted hydrogen. The filtrate 
 from this is evaporated, in the presence of barium carbonate, to a small 
 volume, and filtered. It is then treated with 90 per cent, alcohol and the 
 fiocculent precipitate that forms filtered off. The sugar is crystallised 
 out from the filtrate over sulphuric acid, and the crystals purified by 
 dissolving them in water, decolorising with animal charcoal, and re- 
 
QUALITATIVE TESTS 51 
 
 crystallising. A fresh crop of crystals may be obtained from the 
 mother liquor by a further addition of alcohol. 
 
 (6) Copper. — According to Salkowski the sugars, and particularly 
 glucose, can be precipitated by copper sulphate and an alkali, forming 
 an insoluble blue or green double compound. The formation of in- 
 soluble copper compounds is most coramonly employed for the separa- 
 tion of the carbohydrates from protein substances, and for this purpose 
 copper acetate, or chloride, is generally employed. The neixtral solu- 
 tion to be examined is mixed with a large amount of a concentrated 
 solution of copper chloride and any precipitate that forms is filtered 
 off. To the filtrate is then added sodium hydrate in sufficient quantity 
 to combine with all the hydrochloric acid of the copper chloride used, 
 and also give two molecules for each molecule of copper chloride. The 
 precipitate is filtered off and well washed with hot water. As some 
 copper compound of the carbohydrate still remains in solution, the 
 filtrate is mixed with a large excess of alcohol. The precipitate that 
 forms is filtered off, and washed with alcohol containing sodium 
 hydrate in solution, till the filtrate no longer gives the biiiret reaction. 
 The combined precipitates are dissolved in dilute hydrochloric acid, 
 and re -precipitated by the calculated amount of alkali. The purified 
 precipitate is suspended in water, the copper precipitated by a stream of 
 sulphuretted hydrogen and the filtrate evaporated down in vacuo. 
 The sugar is then crystallised out or precipitated with alcohol or methyl - 
 alcohol. 
 
 C. Alkaline Earths. — The hydroxyl groups of the carbohydrates 
 combine with oxides of the alkaline earths to form more or less in- 
 soluble compounds. Levulose, for instance, forms a characteristic 
 calcium compound. Glucose, in methyl-alcohol solution, forms a 
 barium compound. These earthy salts are most readily precipitated 
 out by alcohol. The sugars can be recovered from them by treatment 
 with sulphuric, oxalic, or carbonic acids. The non-reducing di- and 
 polysaccharides also form insoluble compounds by which they can be 
 separated {e.g. cane-sugar and strontium). 
 
 D. Benzoyl Chloride. — Sodium hydrate is added to the urine to 
 precipitate out the phosphates. To each litre of the clear filtrate are 
 then added about 40 c.c. of benzoyl chloride and 400 c.c. of a 10 per 
 cent, solution of sodium hydrate. The mixture is placed in a large 
 stoppered-bottle, and well shaken matil the smell of benzoyl chloride has 
 disappeared. At the end of the reaction the solution must still be alkaline. 
 The mixture is left to stand overnight on ice, and the precipitated ester 
 filtered off, well -washed with water, and dried. It is piirified by being re- 
 crystallised, fractionally, from warm absolute alcohol. The carbohydrate 
 is recovered by adding to each 10 grams of the ester, 7-5 grams of 
 metallic sodium dissolved in 300 c.c. of absolute alcohol, the sodiLim 
 ethylate solution being cooled to 5° C. and the finely divided ester added 
 to it slowly, shaking well after each addition. After about 20 to 40 
 minutes the decomposition is complete, and a sample on being taken 
 out and mixed with an equal quantity of water is no longer found to 
 
52 GLYCOSURIA 
 
 give a turbidity. Siifficient sulphviric acid is now weighed out to con- 
 vert the sodiuni into an acid sulphate, and, after mixing it with a& 
 much water as alcohol was used in the first operation, it is added to 
 the solution. The freed benzoic acid is removed by extracting the 
 solution with an equal voluixie of ether, three times. The separated 
 ether is extracted with water, to recover traces of sugar that have 
 dissolved in the ether, and the extract added to the alcohol sugar solu- 
 tion. This is then neutralised with sodiuni hydrate until it is only 
 faintly acid, and finally completely neutralised with sodium carbonate. 
 The reaction must not be alkaline. The solution is now mixed with 
 three vohuxies of alcohol, and left overnight for the sodium sulphate tO' 
 crystallise out. The filtrate of the feebly acid solution is evaporated in 
 vacuo. The concentrated brown solution can be decolorised with 
 lead acetate and basic lead acetate, the excess of lead being removed 
 with sulphuretted hydrogen and purified from the latter with carbonic- 
 acid gas. This method which was at one time very extensively exn- 
 ployed is now not often used, for it is very laborious, and benzoyl 
 esters of other substances, which are not easily separated from the 
 sugar compound, are also formed and carried down in the precipitate. 
 Many researches on the reducing sugar-content of normal urines were 
 conducted with benzoyl chloride. 
 
 E. Hydrazones and Osazones.— The compounds formed by many of 
 the sugars with phenylhydrazin, and the substituted hydrazines, are 
 more or less insoluble in water and other solvents, and so serve for 
 their separation. From these the sugar can be recovered by appro- 
 priate treatment. (See Appendix.) 
 
 The confirmatory tests for each variety of sugar, &c., will now 
 be considered. 
 
 Dextrose (g'lucose) is by far the most common sugar met with 
 in the urine. Any urine that gives a marked reduction, readily 
 ferments mth yeast, and is dextro-rotatory is almost certain to 
 contain it. 
 
 1. Buhner's test is a modification of the Moore-Heller test. 
 When carried out under the foUomng conchtions a positive result 
 is characteristic of dextrose. 
 
 Ten cubic centimetres of a concentrated solution of neutral lead 
 acetate (one part of lead acetate to ten of distilled water) are mixed with 
 10 c.c. of the urine. The mixture is filtered, and ammonia carefully added 
 to the filtrate, drop by drop, until a caseous precipitate just remains on 
 shaking. It is then heated in a water-bath at 80° C. (not liigher). 
 If glucose is present the sokition turns a beautiful red and the precipi- 
 tate becomes rose or salmon-pink. If the urine is concentrated it is 
 advisable to dilute it, so that the sjDecific gravity does not exceed 1"010. 
 An excess of ammonia must be avoided as it ruins the test, and the 
 temperature must not be raised above 80° C, since lactose and maltose 
 give a similar reaction when the solution is boiled. When the test is care- 
 
QUALITATIVE TESTS 53 
 
 fiilly carried out it is very reliable and delicate. Under the conditions 
 described lactose gives a yellowish-pink or brown coloration, but no 
 red precipitate ; maltose a slight yellow colour ; and le\ailose no 
 colour at all. 
 
 2. Di-phenylhydrazin. — With this substance dextrose forms a 
 hydrazone by which it can be distinguished from levulose, A\hich 
 does not form a similar insoluble compound. Galactose and the 
 pentoses also react with di-phenylhydrazin, but their hydrazones 
 can be distinguished from the dextrose compound by their melting- 
 points (galactose 157° C, r-arabinose 204° to 205° C, 1-arabinose 
 216° to 218° C, xylose 107° to 108° C). Dextrose di-phenylhydra- 
 zone has a melting-point of 161° to 162° C. 
 
 Owing to the feeble solubility of di-phenylhydrazin in ^^ater an 
 alcoholic solution of the requisite amount of the reagent must be used 
 in carrying out the test. The mixture is left at the temperatiore of the 
 room for two or three days, or may be heated in a water-bath for two 
 hours. The hydrazone is precipitated ovit by the cautious addition 
 of ether. It separates as small colourless prisms that are easily soluble 
 in water and hot alcohol, but are insoluble in ether, chloroform, or 
 benzol. 
 
 3. Methyl-j^henylhydrazin gives a hydrazone with a melting- 
 point of 130° C, which character distinguishes it from the similar 
 hydrazones yielded by levulose (M.P. 158° to 160° C.) and galactose (M.P. 
 180° C). It is separated out by concentrating the solution in which 
 it forms, treating the syrup with alcohol, and recrystallising from 
 alcohol. 
 
 4. Benzyl-phenylhydrazin gives the same hydrazone with both 
 dextrose and levulose, but its melting-point (165° C.) differentiates it 
 from the compound formed with galactose (154° C). It appears as 
 light yellow needles that are slightly soluble in ethyl and methyl alcohol, 
 soluble in pyridin, and insoluble in water. The rotatory powers of its 
 solutions are, methyl-alcohol —33°, glacial acetic acid — 20"2°, pyridin 
 -45-33°. 
 
 5. Beta-najJthyl-hydrazin gives two hydrazones with dextrose, one 
 that melts at 95° C. and the other at 179° C. The hydrazone of levu- 
 lose melts at 152° C. They separate out as broAvn crystals slightly 
 soluble in water and 95 per cent, alcohol, easily soluble in pure methyl- 
 alcohol ( + 402°). The acetic acid solution is optically inactive. 
 
 6. Para-hrom-'phenylhydrazm forms a crystalline osazone with 
 all the monosaccharides and some disaccharides. The dextrose 
 and levulose compounds melt at 222° C, the 1-arabinosazone at 
 196° to 208° C, the r-arabinosazone at 200° to 202° C, the 1-xylose 
 compound at 208° C, the maltosazone at 198° C. Glucuronic acid 
 yields a crystalline hydrazone that melts at 236° C. The test 
 may be carried out for clinical purposes as follows : — 
 
54 GLYCOSURIA 
 
 About one-third of an inch in depth of para-broni-phenylhydrazin is 
 introduced into the bottom of a test-tube, an ecjual bulk of sodium 
 acetate is added, and the test-tube filled to one-third of its capacity 
 with the urine. The mixture is then boiled for two minutes. Crystals, 
 rather longer and paler than those given with phenylhydrazin, are 
 obtained if sugar is present. Performed in this manner the test only 
 responds to an amoiint of glucose that is beyond the physiological 
 limit. 
 
 Levulose (Fructose). — When levnlose occurs in a urine alone 
 a positive fermentation test, and levo-rotation on polariscoj)ic exa- 
 mination, \\ill indicate its presence. If glucose and levulose are 
 present together, as is generally the case, the percentage of sugar 
 as determined by titration will be in excess of that indicated bj" 
 the polariscope, and after fermentation the reducing and optical 
 characters will be lost, provided that proteins, glucuronic acid, 
 beta-oxybutjTic acid, and other optically active substances are 
 absent. Should the urine be levo-rotatory after complete fer- 
 mentation the presence of one, or more, of these substances is 
 indicated. 
 
 Levulose jdelds the same osazones with phenylhydrazin and 
 para-broni-j)henylhydrazin as le^iilose, but it does not form an 
 insoluble crystalline hydrazone with di-phenylhydrazin. Its pre- 
 sence can be confirmed by the aid of the following tests : — 
 
 1. Seliwanoffs Reaction. — This test distinguishes the ketoses 
 from th.e aldoses ; but since levulose (and cane-sugar) is the only 
 ketose met with, in the urine, it may be used for the detection of 
 that sugar. The aldoses also give a reaction if the hydrochloric 
 acid is too strong, or the heating is too prolonged. The details of 
 the test must therefore be strictly adhered to. It is carried out 
 as follows : — 
 
 To the urine is added an equal volume of a solution consisting of 
 0*5 grams of resorcin, 30 c.c. of water, and 30 c.c. of concentrated 
 hydrochloric acid, and the mixture heated in a water-bath. If levulose 
 is present a beautiful Biu-gundy-red colour develops, and a red pre- 
 cipitate settles out on standing. Care must be taken, however, not to 
 confuse the rose tint given on boiling many urines with hydrochloric 
 acid, with the tjrpical colour reaction due to levulose. The former 
 is only a light shade of red, and fades entirely on standing, or cooling, 
 while the latter is a dark magenta-red, clouding the entire specimen 
 and deepening on standing, or on rapid cooling. The colour due to 
 le^n.^lose persists for days, and there is deposited on the bottom of the 
 test-tube a dark red precipitate. Moreover, the red coloxir due to 
 levulose appears at once, and not after prolonged heating. According 
 to Gulart and Grimbert, the appearance of a precipitate on cooling is 
 
QUALITATIVE TESTS 55 
 
 more characteristic than the red coloi.ir of the solution. If the acid 
 be neutralised with sodium carbonate, and the solution is extracted 
 with amyl alcohol, the alcohol takes on the red coloixr, and on examining 
 it with the spectroscope a band between E and B, in the green, is seen 
 with dilute solutions, and a second band at F, in the blue, with con- 
 centrated solutions. On shaking the amyl alcohol extract repeatedly 
 with water the colour is extracted and the alcohol appears yellow. 
 
 Borchardat describes another method of carrying out the test, which 
 he claims gives more reliable results. A few cubic centimetres of the 
 urine are mixed with an equal volume of 25 per cent, hydrochloric acid, 
 a few granules of resorcin are added, and the mixture is quickly brought 
 to the boil. A red colour should appear at once if levulose is present. 
 The solution is then cooled, made alkaline with caustic soda solution, 
 and extracted with acetic ether. In the presence of levulose the acetic 
 ether extract is coloured yellow. Nitrites and indican also give the 
 reaction if present in more than traces, and must therefore be excluded, 
 but this modification is said not to react with urobilin and bile pigments. 
 
 2. Pinojf's Test. — According to Pinoff, levulose can be recognised 
 in a mixture with other sugars, by heating 10 c.c. of the solution, with 
 10 c.c. of a 4 per cent, solution of ammonium molybdate, and 0*2 c.c. of 
 glacial acetic acid, in a water-bath, at 95° to 98° C. for three minutes. 
 Levulose gives a bright blue coloration, whereas other sugars give no 
 colour within the time limit, but may yield a dark green coloiu" after 
 half an hour. Any free acid must be carefully neutralised before carry- 
 ing out the test, since in the presence of free acid other sugars give a 
 blue coloT-U". Schoorl and Kalmthout found that with solutions con- 
 taining 0"05 gxams of dextrose a faint blue colour develops in ten 
 minutes, with cane-sugar after ten minutes a green, and after twenty 
 minutes a blue colour, and it was only with milk-sugar that twenty 
 minutes elapsed before the green coloration was seen, 
 
 3. Methyl-phenylhydrazin. — This is a most important reagent 
 for the recognition and differentiation of levulose, with which it 
 forms an osazone consisting of bright yellow needles that melt at 
 158° to 160° C. A solution of the osazone in pyridin -alcohol 
 (0*2 grams in 4-0 grams pyridin, and 6-0 grams of absolute alcohol) 
 is dextro-rotatory (+1-40°). Dextrose and galactose do not yield 
 osazones, but form hydrazones which melt at 130° C. and 180° C. 
 respectively. 
 
 To obtain the osazone from the separated sugar an alcoholic solution 
 of the reagent acidified with acetic acid (4 c.c. of a 50 per cent, solution to 
 10 c.c.) is mixed with a solution of the sugar and heated for five or ten 
 minutes on the water-bath. The bright yellow osazone crystals appear 
 in a few hoius, or on the following day. They are separated off and 
 recrystallised from 10 per cent, alcohol. 
 
 To separate dextrose and levulose when present in a mixture : — 
 A neutral alcoholic, and not too strongly saline, solution of the sugars 
 
56 GLYCOSURIA 
 
 is heated with methyl -jDhenylhydrazin in a water -bath for some time. 
 When a few crystals of glucose methyl -phenylhydrazone appear in the 
 syrup it is set aside to crystallise for several days. It is then mixed 
 with absolute alcohol and the glucose derivative separated off. The 
 filtrate is acidified with acetic acid, heated on a water -bath for a short 
 time, and set aside to crystallise. The resulting leviolose-methyl-phenyl- 
 osazone is piu-ified by being recrystallised from 10 per cent, alcohol. 
 
 Methyl -phenyl -levulosazone can, according to Neuberg and Strauss, 
 be prepared directly from the urine by the following procedure. The 
 urine is acidified with a few drops of acetic acid, boiled, and filtered, 
 to remove any albumen. It is then evaporated in vacuo, at 40° C, to a 
 syrup, taking care that the reaction remains faintly acid. The syrup 
 is mixed with half its bulk of 98 per cent, alcohol, heated on a water-bath 
 for five minutes, cooled, and filtered. If the residue on the filter is found 
 to have any reducing power it is mixed with a little water and extracted 
 with alcohol once or twice more. The mixed alcoholic solutions are 
 now, if necessary, filtered off from any fiocculent precipitate that may 
 have formed, and decolorised with animal charcoal. The sugar- 
 content of the solution is then estimated by titration, and for each 
 molecule of sugar found to be present, tliree molecules of methyl- 
 phenylhydrazin are allowed. The sugar solution is evaporated to a small 
 bulk (30 c.c), cooled, and left to stand for one hour. If any precipitate 
 forms it is filtered off. The filtrate, or original solution, is now acidi- 
 fied by being ixiixed with a weight of 50 per cent, acetic acid equal to 
 that of the methyl-phenylhydrazin employed, and as much alcohol added 
 as is necessary to produce a clear solution. The mixture is placed in 
 a boiling water-bath for five minutes, or may be left at 40° C. for 
 twenty -four hours. The osazone separates out on cooling, and adding 
 a few drops of water, in a crystalline form if much sugar is present, 
 but as an oil if there is only a small amount. In the latter case it is 
 separated by strongly cooling the solution. The product is recrystal- 
 lised from alcohol by cooling, or by treating it with hot water and 
 enough pyridin to dissolve it, decolorising with animal charcoal, and 
 filtering. 
 
 4. Beta-napthyl-hydrazin can also be employed to separate 
 levulose from dextrose. With levulose it forms a hydrazone with 
 a melting-point of 162° C, with dextrose two hydrazones that 
 melt at 95° C. and 179° C. respectively. 
 
 The mixture of sugars is dissolved in two parts of water, and to it is 
 added two parts of beta-napthyl-hydrazin dissolved in absolute alcohol. 
 The mixture is left to stand for two days, shaking it at frequent inter- 
 vals. The feebly soluble hydrazones of dextrose separates out first, 
 and are filtered off. The filtrate is evaporated to dryness in vacuo over 
 siUphuric acid and the residue dissolved in chloroform. It yields on 
 recrystallisation the pure beta-napthyl-phenylhydrazone of levulose. 
 
 5. Benzyl - phenylhydrazin forms hydrazones with levulose, 
 
QUALITATIVE TESTS 57 
 
 dextrose, and other sugars. The compounds formed with levulose 
 and dextrose have the same melting-point (165° to 170° C), but 
 the latter is decomposed into its constituents by boiling water, 
 whereas the former is not affected. L-arabinose yields a hydrazone 
 with a melting-point of 170° to 174° C, but it is insoluble in alcohol. 
 The hydrazone of r-arabinose melts at 185° C. Galactose forms a 
 hydrazone with a melting-point of 154° to 158° C. that is only 
 feebly soluble in alcohol, but appears much later than the arabinose 
 compounds. 
 
 To prepare the hydrazones a solution of the sugar in 96 per cent, 
 alcohol is mixed with the calculated amoiint of benzyl-phenylhydrazin, 
 and heated in a water-bath for five or six hours. The solution is eva- 
 porated down, and the product recrystallised out of alcohol. The 
 hydrazones of the aldoses and ketoses may then be distinguished by 
 their behaviour with boiling water. 
 
 Lactose (Milk-SUg'ar). — Lactose reduces alkaline solutions of 
 copper and bismuth, although more slowly than dextrose, but 
 does not reduce Barfoed's solution. Boiling with dilute mineral 
 acids increases the reducing power of its solutions, but boiling 
 with citric acid produces no change. Lactose does not ferment 
 with ordinary brewer's yeast within twenty-four hours, but it 
 may be slowly broken down by contaminating bacteria. This 
 spurious fermentation may be prevented by the addition of socUum, 
 or ammonium, fluoride. A urine which still reduces, and is dextro- 
 rotatory, after being fermented with yeast probably contains lac- 
 tose, or possibly a pentose. The confirmatory tests are as follows : — 
 
 1. Rubner^s Test. — On carrying out this test as described for 
 dextrose, a yellowish-pink or brown colour is obtained, but the 
 precijaitate is white. If the solution is boiled it turns yellow, then 
 intense brick-red. On standing the fluid becomes colourless, and a 
 copper-red precipitate settles out. Maltose gives a similar reaction. 
 
 2. Wohlk {Malfatti) Test. — Lactose may be detected in the 
 urine by mixing it with half its volume of concentrated ammonia, 
 and heating the mixture in a water-bath that is not quite boiling 
 for from five to fifteen minutes. The mixture turns red if lactose 
 is present. 
 
 In Malfatti's modification of this test 5 c.c. of the urine are mixed 
 with 2 to 5 c.c. of strong ammonia, and five drops of caustic potash 
 solution added. The mixture is heated, but not quite to boiling, in a 
 water-bath. In the presence of milk-sugar a red coloration develops in 
 about five minutes. By this test it is stated that 0-1 per cent, of lactose 
 ■can be detected in an otherwise sugar-free urine. Maltose gives the 
 same reaction, but with glucose a yellow or brown colour is obtained. 
 
58 GLYCOSURIA 
 
 3. Mucic Acid Test. — ^Mucic acid is a characteristic derivative of 
 galactose. It is formed, along with saccharic acid, when lactose 
 is hydroHsed and oxidised, and owing to its insolubility, the acid 
 may be readily prepared and separated from solutions containing 
 lactose. Although the reaction is satisfactory for pure solutions, 
 only some 50 to 60 per cent, of the theoretical yield is often obtained 
 from the urine, so that small amounts may be easily missed. 
 
 According to Langstein and Steinitz, the test can be carried out as 
 follows. The Tirine is treated with lead acetate and ammonia, and 
 the resulting precipitate filtered off. This is washed with water, and 
 decomposed with sulphuretted hydrogen. The lead sulphide is re- 
 moved by filtration and the excess of sulphioretted hydrogen expelled 
 from the filtrate by warming. It is then evaporated down three times 
 with ammonia (sp. gr. 1-2). The presence of mucic acid is shown by 
 dissolving in ammonia, evaporating, and subjecting the residue to dry 
 distillation, when pyrrol is formed. This is recognised by the red 
 violet coloration it gives with a pinewood splinter moistened with 
 hydrochloric acid. 
 
 Bauer's method is as follows. 100 c.c. of the urine are mixed with 
 20 c.c. of prue concentrated nitric acid (sp. gr. 1*4), in a small beaker, 
 and placed in a boiling water-bath. At first the solution is dark- 
 coloured, but later becomes clear yellow. When this stage is reached, 
 and it is seen to contain a fine white precipitate, generally when it has 
 evaporated down to about 20 c.c, the beaker is removed from the bath 
 and its contents poured into a smaller, into which any precipitate is 
 washed with two small portions of water. It is then left to cool over 
 night. After being diluted with water the precipitate is filtered off, 
 repeatedly washed with cold distilled water, and dried. Mucic acid has 
 a melting-point of 213° to 215° C, and after recrystallising from boiling 
 water 217° to 225° C. The precipitate may also be dissolved in ammonia 
 and tested for pyrrol, as in the preceding method. 
 
 4. Phenylhydrazin.— The phenylosazone of lactose differs in 
 its appearance, melting-point (about 210° to 212° C), and optical 
 activities (pyridin-alcohol solution ± 0-00), from those of other 
 sugars, but owing to its being relatively soluble in water, and the 
 small amount of lactose generally present (under 1 per cent.), the 
 osazone cannot usually be prepared directly from the urine. A 
 negative phenylhydrazin test with a reducing urine is therefore sug- 
 gestive of the presence of lactose. To prepare the osazone the 
 sugar must be first isolated from the urine, or the urine may be 
 treated with Patein-Dufau's reagent, which precipitates out uric 
 acid, creatinin, albumen, &c., and jdelds a colourless filtrate con- 
 taining the sugar. 
 
 Patein-Dufau Reagent. — 220 grams of red oxide of mercury are 
 
QUALITATIVE TESTS 59 
 
 mixed with 160 c.c. of nitric acid (sp. gr. 1'39) in a porcelain basin. 
 After standing for five or six ininutes the mixture is diluted with 160 c.c. 
 of water, and heated until the oxide is completely dissolved. On cool- 
 ing, 40 c.c. of a 10 per cent, solution of sodimn hydroxide are gradually 
 added, with constant stirring, and the mixture diluted to 1000 c.c. 
 The solution is filtered, and preserved in dark glass bottles. One part 
 of the reagent is used to precipitate four parts of urine, and the filtrate 
 used for the phenylhydrazin test, after removing the excess of mercury 
 with sulphuretted hydrogen. 
 
 If a urine suspected to contain lactose is boiled with 5 per cent, 
 sulphuric acid for a short time, and the excess of acid neutralised 
 with ammonia, the phenylhydrazin test should show crystals of 
 dextrosazone, and, with proper precautions, galactosazone also. 
 
 5. Beta-benzyl-phenylhydrazin. — With this reagent lactose forms a 
 hydrazone that melts at 128° C. It appears as light yellow needles, 
 slightly soluble in alcohol, and soluble in inethyl -alcohol. Its methyl- 
 alcohol solution is levo-rotatory ( — 25-7°). 
 
 Pentoses. — A urine containing a pentose gives a reaction with 
 an alkaline solution of copper. When dextrose is not also present 
 the reduction is stated not to take place at once, but only after 
 heating for some time, and then occur suddenly throughout the 
 whole bulk of the fluid. Nylander's solution is only slightly re- 
 duced, a grey precipitate being formed. A pentose-containing 
 urine does not ferment with yeast. It may be either dextro-rotatory, 
 or inactive, according to the variety that is present. Such a 
 urine also gives the phloroglucin test. From the results of the re- 
 duction and fermentation tests a pentose may be easily mistaken 
 for lactose, but a pentose-containing urine should show a charac- 
 teristic spectrum on examining an amyl alcohol extract of the 
 phloroglucin test with the spectroscope. 
 
 As glucuronic acid gives similar results, confirmatory tests 
 must, however, be applied. These are as follows : — 
 
 1. Orcin Tests. — The most easily applied of the confirmatory 
 tests for the presence of a pentose is the orcin reaction. When 
 a solution of a pentose is heated with strong hydrochloric acid 
 and orcin the fluid develops a violet-blue colour, or a green if 
 iron is present. To be reliable, however, the test must be very 
 carefully carried out in every detail, and the reagents employed 
 must be of the exact strength and kind described. ^ Several moch- 
 fications of the test have been described, but Bial's is the one most 
 commonly employed clinically. 
 
 1 The hydrochloric acid must be pure, and sp. gr. 1-195, the commercial acid 
 will not do : orcin is not the same thing as the dye orcein, with which I have 
 more than once seen the test attempted. 
 
60 GLYCOSURIA 
 
 (a) SalkowskV s Method. — A few cubic centimetres of the urine are 
 mixed with an equal cjuantity of strong hydrochloric acid (sp. gr. 1-195), 
 and a few graniiles of orcin, in a test-tube. The mixture is then heated 
 in the flame for twenty to thirty seconds. If a pentose is present the 
 solution turns a reddish-blue, or, if the acid contains traces of iron, a 
 dark green colour, and a dark blue, or green, precipitate forms. If the 
 solution is cooled until it is just warm, and extracted with amyl alcohol, 
 it gives a beautiful dark blue, or green, extract, which on examination 
 with the spectroscope shows a characteristic absorption band between 
 C and D (red and yellow). Glucuronic acid gives the same colour 
 reactions and spectrum as the pentoses, but as a rule it is not split off 
 from its combinations by such brief heating. 
 
 (h) BiaVs Modification.— Foxa or five cubic centimetres of a reagent 
 made by mixing 500 c.c. of fuming hydrochloric acid (sp. gr. 1'195), 
 1 gram of orcin, and twenty-five drops of a 10 per cent, solution of 
 perchloride of iron, are heated to boiling, and then removed from the 
 flame. The urine is immediately added drop by drop, agitating the 
 liquid, and observing its colour between each addition, until either the 
 characteristic result is obtained, or 1 c.c. of the urine has been added, in 
 all. If a pentose is present a green coloizr should appear at once, or almost 
 immediately, and an amyl alcohol extract of the cooled fltxid should 
 yield a green fluid, which shows the same spectrum as is obtained with 
 Salkowski's modification. It is claimed by Bial that his test is much 
 more sensitive than the original method, and yet does not give a 
 reaction with glucuronic acid when properly performed. 
 
 (c) Jones' Test. — Jolles does not consider that Bial's reaction can 
 be relied upon to differentiate the pentoses from glucuronic acid, and 
 recoimnends the following procedure : — 10 to 20 c.c. of the urine are 
 mixed with 1 gram of phenylhydrazin hydrochloride, and 2 grams of 
 sodium acetate. The mixtiore is shaken, and heated for about an hour 
 in a boiling water-bath. It is then cooled for a couple of hours in 
 water. The resulting osazone is filtered off on to an asbestos filter, 
 washed with 3 or 4 c.c. of cold water, and, with the asbestos, introduced 
 into a distillation flask, containing 20 c.c. of water and 5 c.c. of con- 
 centrated hydrochloric acid. Five c.c. are distilled over into 5 c.c. of 
 cold water, and 1 c.c. of the mixture tested with Bial's reagent. If a 
 pentose is present an intense green colour develops, and the character- 
 istic spectriim is seen. Jolles states that by this method 0*05 per cent, 
 of arabinose can be detected, but that glucose and glucuronic acid give 
 no reaction. 
 
 (d) Neumann'' s Test. — Ten drops of the suspected urine are mixed 
 in a test-tube with 5 c.c. of glacial acetic acid (99 per cent.) and a drop 
 of a 5 per cent, alcoholic solution of orcin. The mixture is shaken, and 
 raised quite to the boiling-point. The test-tube is then held in a test- 
 tube holder, and concentrated sulphuric acid dropped in, with constant 
 shaking, until a faint violet -blue colour appears. As a rule not more 
 than fifty drops of sulphuric acid are necessary, and an excess obscures 
 the tint. 
 
QUALITATIVE TESTS 61 
 
 2. Phenylhydrazin. — ^According to Salkowski, this test is 
 best performed as follows : — 
 
 200 c.c. of the iirine are placed in a beaker, and mixed with 5 grams 
 of phenylhydrazin and the same quantity of 50 per cent, acetic acid. The 
 mixture is well shaken, gently heated on a wire gauze, and then in a 
 water-bath, but not to boiling. The fluid is filtered while hot, and 
 cooled by placing the beaker in cold water. The resulting pentosazone 
 crystals are filtered off, and purified by repeated recrystallisation from 
 hot water. They differ from glucosazone crystals in their appearance 
 microscopically, their greater solubility in hot water, and their much 
 lower melting-point. The last varies from 155° to 168° C. according 
 to the way in which the heat is applied, and the variety of pentose 
 present, the inactive forms giving an osazone with a higher melting- 
 point than the active varieties. 
 
 Should the urine contain both pentose and hexoses, Kltlz and 
 Vogel suggest that they can be separated by the following proce- 
 dure : — 
 
 From I'G to .3"2 litres of the urines are taken, and, for each 100 grams 
 of dextrose, 200 grams of phenylhydrazin and 100 grams of glacial 
 acetic acid are added. The mixture is heated on a water-bath for an 
 hour and a half, cooled, and filtered. The filtrate is again heated on 
 the water-bath for an hour and a half, and filtered. The combined 
 precipitates are well washed with cold water, and the pentosazone ex- 
 tracted by digesting with water at 60° C, one litre of water for each 
 100 grams of sugar being used, and the digestion being continued for 
 twelve hovirs. This is repeated fifteen times. The hot extracts are 
 filtered, and allowed to cool. The pentosazone will then separate out. 
 It is purified by recrystallisation from water at 60° C, or from acetone,, 
 till the melting-point is constant. 
 
 The variety of pentose present can be determined from a con- 
 sideration of the following characters : — 
 
 1-arabinOSe. — l. The urine, or solution of the sugar, is dextro- 
 rotatory (+104-4°). 
 
 2. The phenylosazone forms a voluminous precipitate, con- 
 sisting of yellow crystals, which are insoluble in cold water, ether, 
 benzol, and ligroin, and are soluble in hot water, alcohol, acetone, 
 and pyridin. A 4 per cent, alcoholic solution, when freshly prepared, 
 is dextro-rotatory ( -f 18-9°), but on standing becomes optically in- 
 active. The pyridin-alcohol solution is dextro-rotatory (+1-1°). 
 The crystals washed with water and recrystallised from hot water 
 and acetone melt at 160° C. on rapid heating, 
 
 3. Di-phenylhydrazin. — The hydrazone that 1-arabinose forms 
 with di-phenylhydrazin is one of its most insoluble compounds, and 
 is therefore of great use in separating and identifying it. It appears 
 
62 GLYCOSURIA 
 
 as wliite needles, which on being raiDidly heated melt at 216° to 
 218° C. Its pyridin-alcohol solution is shghtly dextro-rotatory 
 ( + 0-42°). It can be prejoared directly from the urine in the fol- 
 lowing manner (JSTeuberg and Wohlgemuth) : — 
 
 100 c.c. of the urine are feebly acidified with acetic acid, evaporated 
 to 40 c.c, and mixed with an equal volmne of alcohol. The precipitate 
 that forms after standing for two hours is filtered off, and washed with 
 50 per cent, alcohol. The filtrate is mixed with 1-4 grams of di-phenylhy- 
 drazin, and heated on a water-bath for half an hour, the loss of alcohol 
 from evaporation being made up as required. It is then left to cool, 
 when the crystalline hydrazone will separate out. On treating each 
 gram of the hydrazone with 4 c.c. of formalin and a httle water it is 
 broken up into foiraaldehyde-diphenylhydrazin and the pentose. The 
 former can be separated by shaking out with ether, lea\dng the sugar 
 in solution, from wliich it can be crystalHsed. 
 
 4. Para-hrom-phenylhydrazin, — ^With para-brom-phenylhydrazin 
 1-arabinose forms a very characteristic insoluble hydrazone, by 
 which it can be distinguished from xylose and the hexoses. Its 
 melting-point of 160° to 162° C. also serves to distinguish this 
 sugar from glucuronic acid, the para-brom-phenylhydrazin com- 
 pound of which melts at 236° C. 
 
 The hydrazone is prepared by mixing each part of sugar with a freshly 
 prepared solution, consisting of one part of para-brom-phenylhydraziu, 
 three and a half parts of 50 per cent, acetic acid, and twelve parts of 
 water, and standing for some time, when it separates out as fine crystals. 
 
 1-arabinose also forms an osazone T\ith para-brom-phenylhy- 
 clrazin, which is easily soluble in hot water, alcohol, acetone, benzol, 
 ether, and pyridin, but is feebly soluble in cold water, and insoluble 
 in ligroin. Its pyridin-alcohol solution is feebly dextro-rotatory 
 ( -1- 0-28°). From alcohol it crystalhses as yellow needles, and from 
 pyridin as six-sided plates. It softens at 185° C, and melts at 
 196° to 200° C. 
 
 5. M ethyl- 'phenylhydrazin forms a hydrazone that is easily 
 soluble in alcohol and pjrridin, slightly soluble in water, and in- 
 soluble in ether. Its alcoholic solution is dextro-rotatory ( -f 4-3°). 
 A solution in acetic acid is levo-rotatory ( - 21-8°). A pyridin solu- 
 tion is opticaUy inactive. It forms yellow crystals that melt at 
 161° to 164° C. 
 
 6. Benzyl-phenylhydrazin forms an osazone consisting of white 
 crystals that are soluble iu methyl alcohol (-12-1°), and glacial acetic 
 acid ( - 14-6°). It has a melting-point of 170° to 174° C. 
 
QUALITATIVE TESTS 63 
 
 i-arabinose. — l. The urine is optically inactive. 
 
 2. PhenylJiydrazin forms an osazone, consisting of yellow needles 
 or prisms, which when pure melt at 166° to 168° C. The osazone 
 prepared directly from the urine usually melts at a much lower 
 temperature, generally at about 156° C. 
 
 3. Di-phenylhydrazin. — On warming an alcoholic solution of 
 the sugar with an equivalent mass of di-phenylhydrazin, the hydra- 
 zone separates out as long white needles, that are insoluble in cold 
 water and alcohol, slightly soluble in chloroform, hot water, and 
 alcohol, and readily soluble in acetic acid and pyridin. The pure 
 product melts at 204° to 205° C. 
 
 4. Para-hrom-phenylhydrazin gives a hydrazone that is easily 
 .soluble in pyridin, but less soluble in water, alcohol, and ether. 
 It has a melting-point of 160° C. Para-brom-phenylhydrazin also 
 forms an osazone, consisting of long yellow needles that melt at 
 200° to 202° C. 
 
 6. Methyl-phenylhydrazin forms a hydrazone that is easily soluble 
 in. water, pyridin, and hot alcohol, less soluble in cold alcohol, acetone, 
 and chloroform, and is insoluble in benzol. Crystallised out from alcohol 
 it melts at 173° C. 
 
 6. Benzyl-phenylhydrazin forms a hydrazone, consisting of light 
 yellow needles, that are soluble in hot water, alcohol, and chloroform, 
 are less soluble in ether, benzol, and ligroin, and are easily soluble in 
 pyridin. It melts at 185° C. 
 
 1-xylose. — 1. Its solution is dextro-rotatory (4-18-10°). 
 
 2. PhenylJiydrazin gives an osazone that crystallises out in 
 light yellow shining needles, or plates. It is easily soluble in ether 
 and acetone, feebly soluble in water, and easily soluble in alcohol, 
 but less so in acetone. A solution in alcohol is strongly levo- 
 rotatory ( - 43-4°). The melting-points given by different authors 
 vary between 152° C. and 170° C. According to Wheeler and 
 ToUens, the pure product melts at 159° to 160° C. 
 
 3. Di-phenylhydrazin yields a hydrazone with a melting-point 
 ■of 107° to 108° C. It is, however, much more soluble than the 
 ■corresponding arabinose compound, and so serves to separate that 
 .sugar from xylose. 
 
 4. Para-hrom-phenylhydrazin gives only a soluble hydrazone 
 with xylose, and so can be differentiated from arabinose. It 
 iorms an insoluble osazone, consisting of yellow needles, solutions 
 of which have the same rotatory powers as those of the arabinose 
 ^compound. Its melting-point is also very similar (208° C). 
 
64 GLYCOSURIA 
 
 5. Methyl-phenylhydrazin gives a soluble hydrazone, consisting of 
 yellow crystals that dissolve in water, alcohol, acetone, acetic, ether, 
 chloroform, and pyridin. It has a melting-point of 108° to 110° C. 
 
 6. Benzyl-'phenylhydrazin. — The hydrazone of xylose forms 
 needles with a melting-point of 93° C. It is only feebly soluble in 
 water, is more easily soluble in ether, and is very soluble in alcohol- 
 Its alcoholic solution is strongly levo-rotatory ( - 33°). 
 
 7. Brucin. — ^With brucin xylose forms a crystalline salt, on 
 being warmed with a faintly alkaline solution. It separates as 
 rhombic tables, and has a melting-point of 172° to 174° C. It is. 
 almost insoluble in cold water and alcohol. 
 
 8. Buhner's Test. — On applying Rubner's test and boiling, xylose' 
 gives a deep orange precipitate. 
 
 9. Xylonic Acid Test. — The most important and characteristic 
 evidence of the presence of xylose in a solution is obtained by 
 converting it into xylonic acid. This is effected by oxidising it 
 with bromine, and separating out the acid as the insoluble double 
 cadmium-bromine salt. The test is carried out as follows : — 
 
 0*2 gram of xylose, or double the volume of the solution, 1 cm. 
 of water, 0-25 gram (seven to eight drops) of bromine, and 0*5 gram of 
 cadmium carbonate are mixed in a test-tube, shaken, and gently 
 warmed. The loosely corked test-tube is then set aside, for twelve to 
 twenty-four hours. The contents are now evaporated to dryness in a 
 porcelain basin, and the residue dissolved in 4 to 5 c.c. of water. The 
 solution is filtered, evaporated to dryness, and mixed with 1 c.c. of 
 alcohol. If pure xylose was present a crystalline precipitate separates. 
 out, and on microscopical examination this is seen to consist of needle- 
 like, or whetstone -like, crystals. 
 
 Glucuronic Acid. — As a rule the glucuronic acid compounds 
 met with in the urine only reduce alkaline solutions of copper after 
 prolonged boiling, but if the urine has been previously heated with. 
 1 per cent, sulphuric acid for from one to five minutes, an immediate 
 reduction occurs. Urocholic acid and paramidophenyl-glucuronic 
 acid reduce Fehling's solution as readily as dextrose without any 
 preliminary treatment, and phenol-glucuronic acid reduces after 
 being boiled with an alkaline solution of copper for a short time. 
 Like the pentoses, glucuronic acid and compound glucuronates are 
 not fermented by yeast. Although paired glucuronic acid does 
 not give the phloroglucin and orcin tests until the compounds have 
 been decomposed and the glucuronic acid set free by prolonged 
 heating, or boiling with dilute mineral acids, some are more readily 
 decomposed than others, and may give rise to difficulties with, 
 these tests unless they are carefully carried out. 
 
QUALITATIVE TESTS 65 
 
 1. A urine containing compound glucuronates is levo-rotatory, 
 but, since the free acid, and its alkaline salts, are dextro-rotatory, 
 on boiling with dilute acid the optical activity will be changed. 
 If the urine contains dextrose, its dextro-rotatory power will be 
 raised. The levo-rotation of normal urines (about 0-05 per cent.) 
 may be increased to 0-25 per cent, by the presence of indoxyl 
 and phenol-glucuronic acids, but if it is over 0-15 per cent, it is 
 probable that they are present in excess. The presence of albu- 
 men, and other levo-rotatory substances, must first be excluded, 
 although albumen up to 0*5 per cent, may be neglected, as this 
 amount does not appreciably affect the optical activity of the urine. 
 If a reaction for acetone is given, the levo-rotation may be due 
 to the presence of beta-oxybutyfic acid. This should be removed 
 by extracting the urine three times with ether before taking the read- 
 ing. If the urine is dark-coloured, and it is necessary to clear it for 
 examination with the polariscope, it should be remembered that some 
 glucuronates {e.g. uroculoric acid, phenol-, menthol-, and napthol- 
 glucuronic acid) are precipitated by lead acetate, while others, such 
 as the camphor compound, are not. The urine must be acid in 
 reaction, since the levo-rotation is less in alkaline solutions. An 
 optically inactive, or even a dextro-rotatory, urine may contain 
 glucuronates, for glucuronic acid is set free spontaneously from 
 some, such as the menthol compound. 
 
 2. Phenylhydrazin. — Most paired glucuronates do not form a 
 crystalline compound with phenylhydrazin when the test is applied 
 directly to the urine, but after boiling with dilute sulphuric or 
 hydrochloric acid, or even in some instances on simply heating 
 for some time with water, the glucuronic acid is set free, and yields 
 crystals that may be easily mistaken for the osazone of a sugar. 
 The phenylhydrazin compound of glucuronic acid is readily soluble 
 in hot alcohol, but generally separates from this solution, after 
 diluting with water and boiling, in an amorphous form. The 
 crystals are feebly soluble in water and hot benzol, are easily 
 soluble in acetone, and very easily dissolve in pyridin, yielding a 
 levo-rotatory solution. In methyl-alcohol they dissolve with ease, 
 being thus distinguished from dextrosazone, which is only very 
 slightly soluble. The phenylhydrazin compound of glucuronic 
 acid dissolves in about one to two minutes when irrigated under 
 the microscope with 33 per cent, sulphuric acid. 
 
 Examined under the microscope with polarised light they are 
 invisible, unlike the osazones of the sugar, which appear bright, 
 and of a green and red colour. The melting-point of the crystalline 
 variety is 114° to 115° C, but the amorphous form sho\A's no change 
 
 E 
 
66 GLYCOSURIA 
 
 until the temperature has been raised to 150° C. or so. With 
 sjsecimens isolated from the urine the melting-point may be 
 anywhere between 114° and 217° C. 
 
 Since the formula of the phenylhydrazin compound of glycu- 
 ronic is CjoH^gNj(,OjQ, it is calculated to yield 16-4 per cent, of nitrogen 
 on combustion. It is not at all an easy matter, however, to obtain 
 a sufficiently pure specimen from the urine to allow of an accurate 
 determination of either the melting-point or the nitrogen content. 
 
 As a rule the glucuronates cannot be satisfactorily chfierentiated 
 from traces of sugar by the phenylhydrazin test, since they do 
 not yield a pure 2)roduct in an amount sufficient for a complete 
 examination. 
 
 3. Para-hrom-'phenylhydrazin gives the most characteristic crys- 
 talhne compound by which glucuronic acid can be recognised. 
 Unlike the corresponding compounds yielded by the sugars, it is 
 insoluble in absolute alcohol. The raw product melts at 200° to 
 206° C, but after being recrystallised from hot 60 per cent, alcohol 
 it has a melting-point of 236° C. Its solution in pjT:'idin-alcohol is 
 .strongly levo-rotatory ( - 7"25°). The test is carried out as follows : — 
 
 A boiling solution of 5 grams of para-brom-phenylhydrazin, and 6 
 grams of sodiimi acetate, is added to the urine, in whichthe glucuronic acid 
 has been previously set free, or a solution containing the separated acid, 
 and heated on the water-bath to 60° C. At first the ixdxtiore is clear, 
 but in from five to ten minutes a yellow precipitate separates out. The 
 solution is now allowed to cool, the crystals are filtered off, and the 
 filtrate is heated afresh. A second crop of crystals separates out. 
 These are filtered off, and the filtrate is again heated on the water- 
 bath, the process being repeated so long as a precipitate forms. The 
 crystals on the filter are carefully washed with a little warm water, then 
 with absolute alcohol, dried, and the melting-point is taken. 
 
 4. Napfifio-resorcinol Test (Tollens). — On being heated with 
 naphtho-resorcinol and hydrochloric acid, glucuronic acid forms 
 a blue substance, soluble in ether. The pentoses do not give this 
 reaction, so that glucuronic acid can be detected by means of it 
 in their presence. The test is carried out as follows : — 
 
 Five or six cubic centimetres of the m-ine, or a piece of the solid 
 glucm-onic the size of a pea dissolved in 5 to 6 c.c. of water, are ixiixed 
 ■R-ith O'o to 1*0 c.c. of a 1 per cent, alcoholic solution of naphtho-resorcinol, 
 and 5 to 7 c.c. of hydrochloric acid (1'19), and gently boiled in a wide 
 test-tube for one minute. After standing for foiu" minutes the liquid is 
 cooled, mixed with an equal volume of ether, and well shaken. If glucu- 
 ronic acid is present, the ethereal solution has a blue or red colovir, and 
 exhibits a blue fluorescence. Examined with the sjDectroseope, it shows 
 a band slightly to the right of the D-line. As a reaction is obtained 
 
QUALITATIVE TESTS 67 
 
 with 0' 1 per cent . , or less, a positive result is given by many normal urines. 
 The presence of indoxyl may vitiate the test, and it should therefore 
 be previously removed by treating the urine with mercvmc acetate. 
 
 5. Quinine forms with glucuronic acid an insoluble salt, con- 
 sisting of microscopic needles, with a melting-point of 204° C, 
 which are strongly dextro-rotatory (-^138-6°). The solution of 
 glucuronic acid is heated to boiling, and C|uinine added until it no 
 longer dissolves. On cooling the quinine salt separates out. 
 
 6. Brucine also forms an insoluble salt, with a melting-point 
 of 200° C. 
 
 7. Benzoyl chloride. — On shaking a solution of free glucuronic acid 
 with benzoyl chloride and sodium hydrate (in 10 per cent, solution), 
 it is precipitated out as dibenzoyi-glucuronic acid. The precipitate 
 is insoluble in w^ater, but is easily soluble in alcohol, particularly 
 in warm alcohol. It reduces Fehling's solution, and melts at 
 107° C. If too much soda is used the precipitation is interfered 
 with, so that for each molecule of glucuronic acid as nearly as 
 possible 9 molecules of benzoyl chloride, and 12 of sodium hydrate, 
 should be employed. 
 
 Separation (a) by Lead. — Paired glucuronic acid may be separated 
 from the urine by concentrating, treating with lead acetate, then -with 
 tribasic lead acetate, and eventually with ammonia and tribasic lead 
 acetate. The lead precipitate is washed and suspended in water, 
 treated with sulphuretted hydrogen, the lead sulphate removed by 
 filtration, and the filtrate heated at 100° C. with 1 per cent, sulphuric 
 acid, in a flask pro\'ided -with reflux condenser. The fliiid is now 
 neutralised with sodium carbonate, and treated with para-brom- 
 phenylhydrazin acetate. After heating for about ten minutes the para- 
 brom-phenylhydrazin separates out. 
 
 {b) Barium. — Glucuronic acid may also be separated as the insol- 
 uble barium salt. The virine is decolorised with animal charcoal, and 
 evaporated to a syrup. It is then digested with a large quantity of 
 damp barium hydrate, at a gentle heat, on a water-bath. The mixtiu-e 
 is extracted with absolute alcohol, and the residue mixed with water, 
 and filtered. More baryta is added to the filtrate, and it is again 
 filtered, and the filtrate evaporated down on a water-bath. An amor- 
 phous bariiun compound of glucuronic acid separates out. This is washed 
 with water, decomposed with sulphuric acid, the barium sulphate 
 filtered off, and the filtrate evaporated down, and dried in vacuo. 
 ■Crystals of the anhydride of glucvironic acid can thus be obtained. 
 
 Maltose. — Maltose reduces alkaline solutions of copper and 
 bismuth, but not Barfoed's reagent until after the mixture has been 
 heated for some time. It is fermented by yeast as easily as dex- 
 trose, and without previous inversion by acids. Its solutions are 
 
68 GLYCOSURIA 
 
 strongly dextro-rotatory, deflecting the plane of polarised light 
 more than tAvice as much to the right as a solution of dextrose of 
 equal strength. Since the amount of cuprous oxide precipitated by 
 maltose from Fehling's solution is only 62 per cent, of that produced 
 by an equal weight of dextrose, the readings obtained with the 
 polariscope, and by reduction, differ very widely when a urine 
 containing maltose is examined by these two methods. After 
 hydrolysis with dilute acid the urine becomes less dextro-rotatory, 
 but reduces Fehling's solution to a greater extent than before. 
 
 1. Phenylhydrazin. — Maltose is most surely recognised and 
 differentiated by the osazone that it forms with phenylhydrazin. 
 This is prepared by prolonged heating (1| hours on the water-bath), 
 and does not separate out from the hot solution, but only on cool- 
 ing. It appears as fine yellow needles, in marked contrast to the 
 coarse crystals of dextrosazone and levulosazone. On taking the 
 melting-point of the purified product it is found to soften at 190*^ 
 to 193° C, and to melt, on rapid heating, at 202° to 208° C. Its 
 solution in pyridin-alcohol is dextro-rotatory ( +1-30°), in contrast 
 to dextrosazone and levulosazone, which are levo-rotatory ( — 1-30°). 
 Maltosazone is much more easily soluble in hot water than dextro- 
 sazone, and so can be separated by fractional crystallisation. It 
 is also more easily soluble in acetone, and can be further purified 
 by extraction with 50 per cent, acetone. 
 
 2. Para-brom-'phenylhydrazin. — Maltose does not give an in- 
 soluble crystalline hydrazone with para-brom-phenylhydrazin, but 
 it forms an osazone with a melting-point of 198° C. The osazone 
 is prepared by standing an alcoholic solution of the sugar with 
 para-brom-phenylhydrazin for several days at 40° C. It appears 
 as needles that are soluble in hot alcohol and acetone, less soluble 
 in acetic, ether, benzol, and chloroform, and are insoluble in ether 
 and ligroin. 
 
 Isomaltose. — Isomaltose is dextro-rotatory, and gives much the 
 same reactions as maltose. It reduces Nylander's and Fehling's 
 solutions to four-ninths the extent of dextrose, but only ferments 
 with yeast after prolonged treatment, and gives a different osazone 
 with phenylhydrazin. It is by the characters of its osazone that 
 it has generally been recognised in the urine. It has also been 
 separated by fermenting the carbohydrates precipitated out with 
 benzoyl chloride. 
 
 Phenylhydrazin. — On heating a 20 per cent, solution of isomaltose with 
 phenylhydrazin acetate, and adding two volumes of cold water, the 
 osazone separates as a flocculent yellow precipitate. On microscopical 
 
QUALITATIVE TESTS 69 
 
 exainination this is found to consist of spherical aggregates of bent 
 yellow needles, that are more readily soluble in hot water, and hot 
 alcohol, than maltosazone, but are insoluble in ether, acetone, and 
 water-free acetic acid. On drying they tvirn orange -yellow, and at 
 100° C. dark yellow. At 142° C. they soften, and melt at 14.5° (Ost) 
 or 153° (Fischer). They can be purified by recrystallisation from 
 warm acetic acid. Their acetic ether solution is levo-rotatory ( — 20°). 
 
 From a mixture of osazones prepared from the urine isomalto- 
 sazone can be separated out, along with maltosazone, by its solu- 
 bility in hot water. From maltosazone it can be differentiated by 
 its comparative insolubility in acetone. Mayer points out, however, 
 that it is most unsatisfactory to depend solely on the melting- 
 point of an osazone for its recognition, as several observers have 
 done in the case of isomaltose {e.g. Pavy and Siau) ; and he sug- 
 gests that, in some instances at least, the sugar regarded as isomal- 
 tose was probably glucuronic acid. 
 
 Galactose. — Galactose gives the ordinary reduction tests, like 
 other monosaccharides. It is not fermented hj brewer's yeast, 
 but is slowly broken down by bacteria. A urine which contains 
 only galactose shows no gas formation in six hours. It is more 
 strongly dextro-rotatory than either dextrose or lactose ( [a]j, for 
 dextrose -h52-5°, for lactose -f52-5°, for galactose +81°). It may 
 be distinguished from other sugars by the following tests : — 
 
 1. Phenylhydrazin. — With phenylhydrazin galactose forms an 
 osazone which is distinguished from dextrosazone by its melting- 
 point, and from lactosazone by its being less soluble in both cold 
 and hot water. The osazone separates out in stout yellow needles, 
 that are only slightly soluble in cold water, more soluble in hot 
 water and alcohol, and easily soluble in hot 60 per cent, alcohol. 
 In pyridin-alcohol the rotation is +0-48° (Neuberg). The melting- 
 point varies very much with the purity of the product, the un- 
 X3urified osazone melting at 171° to 174° C, the purified crystals 
 at 194° to 195° C. The osazone can be prepared direct in the usual 
 way from urines rich in galactose ; but when only small quantities 
 are present the urine must be previously treated with the Patein- 
 Dufau reagent, or the sugar must be isolated. Any admixed lacto- 
 sazone can be removed by washing the crystals with hot water, 
 and the galactosazone be purified by recrystallising from cUlute 
 alcohol and washing with ether, in which it is insoluble. 
 
 2. Methyl-yhenylhydrazin. — The methyl-phenylhydrazone is the 
 most characteristic compound by which galactose can be recog- 
 nised and separated, in the presence of other sugars. It forms 
 colourless needles, with a melting-point of 180° to 188° C, that are 
 
70 GLYCOSUEIA 
 
 only slightly soluble in water and alcohol, but are easily soluble in 
 methyl-alcohol. The hydrazone is prepared by treating a hot- 
 concentrated solution of the sugar with the calculated amount 
 of methyl-phenylhydrazin. 
 
 3. Di-johenylhydrazin. — With di-phenylhydi-azin galactose forms 
 an hydrazone with a melting-point of 157° C, which cannot, however, 
 be distinguished in practice from the hydrazone formed with dextrose 
 (M.P. 161° C). 
 
 4. Benzyl -phenylhydrazin. — Galactose forms a hydrazone consisting 
 of hght yellow needles, slightly soluble in water and alcohol, that melts 
 at 154° to 158° C. Its solution in pyridin is levo-rotatory (— 14"63°), and 
 the methyl-alcohol solution is also levo-rotatory ( — 17*2°). 
 
 5. Mucic Acid. — On oxidising galactose with nitric acid a 
 feebly soluble, sandy, crystalline powder, consisting of mucic acid, 
 is formed. The same procedure may be followed as for the pre- 
 paration of mucic acid from lactose, or the sugar may be isolated 
 and treated in the following manner : — 
 
 The sugar is mixed with about twelve times its bulk of nitric acid 
 (sp. gr. 1*15), and heated for some time on a water-bath. The excess 
 of nitric acid is then evaporated off, and the residue, mixed with a little 
 water, is left to crystalhse out until the following day. The precipitate 
 that forms is washed with water, and the crystalline mucic acid removed 
 by filtration. This is purified by further washing. On microscopical 
 examination it is found to consist of short prisms, which are insoluble 
 in alcohol and ether, and have a melting-point of 225° C. on being 
 quickly heated. On dissolving the mucic acid in ammonia, evaporat- 
 ing, and subjecting the product to dry distillation, CO^, HjO, NHj, and 
 pyrrol are formed. The latter can be recognised by the red-violet 
 colour given by a pine-splinter moistened with hydrochloric acid. 
 Mucic acid also gives a characteristic yellow to reddish -yellow colora- 
 tion "with a reagent consisting of two drops of perchloride of iron, two 
 drops of strong hydrochloric acid, and 100 c.c. of water. 
 
 6. Galactose pentabenzoate crystallises in microscopic needles which 
 melt at 165° C. It is, however, mixed with yellowish drops of an 
 amorphous modification which melts at 82° C. 
 
 Laiose. — This substance has not been obtained in a crystalline 
 form, so that the reactions of the pure product are not definitely 
 known. Its solutions are levo-rotatory ( - 26-07°). It is not fer- 
 mented by yeast, but reduces Fehling's solution, although to a 
 less extent than dextrose or levulose, and only after prolonged 
 boiling. It gives a slight reaction with Moore's test, and forms 
 with phenylhydrazin an oily compound. It has been variously 
 regarded as a hexose (levulose), a pentose (d- xylose), and a heptose. 
 
QUALITATIVE TESTS 71. 
 
 To separate laiose from the iirine it is treated with lead acetate, 
 and the resulting precipitate filtered off. Ammonia is then added to 
 the filtrate. This second precipitate, which contains the laiose and 
 any other sugars, is suspended in water, and decomposed with a stream 
 of sulphuretted hydrogen. The filtrate is evaporated in vacuo, over 
 sulphuric acid, to a syrup, and the syrup treated with methyl-alcohol. 
 The sugar is then precipitated out with a methyl -alcohol solution of 
 baryta, and quickly filtered off. The filtrate is left to stand over 
 sulphuric acid, treated with carbonic acid, and the filtrate from this 
 concentrated in vacuo, to remove the methyl-alcohol. The residue is 
 dissolved in water, and the baryta still in solution is jorecipitated with 
 sulphiiric acid, and the chlorides removed as a silver salt. 
 
 Cane-SUgfar. — Cane-sugar is introduced into the urine by malin- 
 gerers, or may accidentally find its way there. Pure cane-sugar 
 has no reducing action on cupric oxide, but, since the commercial 
 variety contains other sugars as impurities, it may give a positive, 
 although not quite typical, reaction with Trommer's test or Fehl- 
 ing's solution. For the same reason phenylhydrazin may also 
 give a few^ osazone crystals. A urine containing cane-sugar fer- 
 ments only very slowly, is often of a high specific gravity, and 
 is dextro-rotatory. On boiling with dilute hydrochloric acid for 
 twenty to forty minutes, and neutralising with sodium carbonate, 
 it will be found to be levo-rotatory, from inversion of the cane- 
 sugar. It will then also give the typical tests for dextrose and 
 levulose. 
 
 Other Reducing" Substances. — In addition to those already 
 described, other reducing substances have been reported as present 
 in the urine by several observers. 
 
 Salkowski and Blumenthal separated from the urine of several 
 cases, of pneumonia a fermentable, dextro-rotatory body, yielding 
 with phenylhydrazin an osazone having a melting-point of 195° C. 
 and a nitrogen content of 16-06 per cent. 
 
 Jacoby described a reducing substance, recovered from the 
 urine of a case of Addison's disease, that yielded an osazone with 
 a melting-point of 175° to 180° C. 
 
 Rosenberg isolated a sugar, which he regarded as a hejJtose, 
 from the urine of a case of diabetes. It reduced alkaline solutions 
 of copper, and formed with phenylhydrazin an osazone that melted 
 at 195° C. The osazone was soluble in pyridin, and this solution 
 was optically inactive. 
 
 Geelmuyden gave the name " paidose " to a sugar that he 
 isolated from the urine of diabetic children. It was optically 
 inactive, slowly reduced Fehling's solution, and gave an osazone 
 
72 GLYCOSURIA 
 
 with a melting-point of 175° to 190° C. It did not give the phloro- 
 glucin and ore in reactions. 
 
 Glucosamine. — An amino sugar, glucosamine (CgH^jOg.NHg), 
 prepared from chitinin, has been found in the urine after it has 
 been given by the mouth, or subcutaneously. It reduces alkaline 
 solutions of copper, but not as strongly as dextrose, and can be 
 differentiated from glucose by converting it into the tetrabenzoate. 
 According to Kueny, the melting-point of this compound is 197° 
 to 198° C, according to Pum, 203° C. On decomposing the ben- 
 zoate the glucosamine can be recovered and identified. 
 
 Animal Gum (Landwehr) is probably not one, but a group of 
 bodies, precipitated from the urine by alcohol. It is said to be 
 present in traces in all urines, and to be increased in some patho- 
 logical conditions. It is slightly dextro-rotatory, and is not fer- 
 mented by yeast. With the copper tests it gives a precipitate 
 which does not blacken on boiling, but, after prolonged heating 
 with dilute sulphuric acid, it yields a reducing substance. Unlike 
 glycogen, it does not give a colour reaction with iodine. 
 
 Glycog'en (or Erythrodextrin). — Urines containing this sub- 
 stance are dextro-rotatory. They do not reduce alkaline solutions 
 of copper at once, but on prolonged heating the fluid becomes 
 green, then yellow, and sometimes dark brown. 
 
 To separate glycogen from the urine, it is evaporated to a syrup, 
 and potassium hydrate and absokite alcohol added until a cloud, due 
 to the separation of the potassium salts, is obtained. The fluid is 
 decanted, and the precipitate washed several times with absolvite 
 alcohol. It is then dissolved in acetic acid, and reprecipitated with 
 absolute alcohol. The purified precipitate is warmed with alcohol 
 and dried. A white powder, soluble in water, giving a brown colour 
 with iodine, and slowly reducing Fehling's solution is obtained. 
 
 Alkaptonuria. — In this condition the urine, when fresh, is acid 
 in reaction, and of a normal colour. On standing it rapidly 
 darkens, commencing at the surface, and passes through various 
 shades of brown to absolute blackness, owing to absorption of 
 oxygen from the air. The change of colour takes place more 
 quickly if the urine is made alkaline. Linen and woollen fabrics 
 moistened with the urine are stained brown or black, and it is by 
 this staining of the linen that attention is often drawn to the 
 condition. On heating the urine with Fehling's solution a deep- 
 brown colour develops, and a copious reduction occurs, but the 
 browning of the liquid, in which the orange precipitate is suc- 
 pended, gives to the test a peculiar appearance which distinguishes 
 
QUALITATIVE TESTS 73 
 
 it from the ordinary reduction by the sugars. An ammoniacal 
 sokition of silver nitrate is rapidly reduced even in the cold. On 
 heating the urine with Nylander's solution, it is at once darkened 
 by the alkali of the reagent, but no reduction of the bismuth 
 occurs. The urine is optically inactive, does not ferment with 
 yeast, and does not yield a crystalline osazone with phenylhydrazin. 
 The most striking reaction is joroduced by adding a dilute solution 
 ■of ferric chloride to the urine, drop by drop. The addition of 
 each drop produces a deep blue colour, which lasts for only a 
 moment, but is repeated until oxidation is complete. 
 
 The characteristic reactions of the urine are due to the j^resence 
 of homogentisic acid (para-di-oxy-benzene-acetic acid or hydro- 
 quinone-acetic acid, CgH3.(OH)2.CH2.COOH). 
 
 This may be isolated by heating the urine to boiling, and adding 
 5 grams of solid neutral lead acetate for each 100 c.c. The dense 
 precipitate that forms is filtered off while the urine is still hot, and the 
 clear yellow filtrate is put aside in a cool place to stand for twenty-fovir 
 hours. The crystalline lead compound of homogentisic acid that 
 separates out is filtered off, washed, and dried. The free acid may be 
 recovered by dissolving the powdered lead homogentisate in ether, and 
 decomposing it with a stream of sulphuretted hydrogen. The filtrate 
 from this is allowed to evaporate, and the colourless crystals of homo- 
 gentisic acid, with a melting-point of 146° to 147° C, are left. 
 
 In the routine examination of urines for sugar different obser- 
 vers employ different preliminarj^ tests. In this country Fehling's, 
 and on the Continent Trommer's, is the one most commonly used ; 
 but some authors recommend Nylander's reagent, as they contend 
 that it keeps well, and does not give a reaction with many of the 
 disturbing substance j that reduce alkaline solutions of copper. 
 Others strongly advocate Crismer's safranin test, for they point 
 out that, while it is not affected by creatinin, uric acid, and other 
 reducing substances occurring in normal urines, it is very delicate. 
 Although its extreme delicacy is a drawback, since many normal 
 urines give a slight reaction, it has the compensating advantage 
 that, should the test be negative, the presence of even a trace of 
 sugar is excluded, and there is no need to proceed further. If it is 
 positive, the result must always be confirmed by other methods. 
 In my own work, as I have already mentioned, I have for some 
 time been regularly using Benedict's test, with very satisfactory 
 results. The solution keeps well, it is sufficiently delicate to 
 reveal any pathological excess of sugar, but does not react with 
 normal urines, and is not reduced by most of the substances giving 
 rise to difficulties when Fehling's solution is employed. Whichever 
 
s 
 
 o ^^ 
 
 -=-£- 
 
 
 
 o 
 
 -25S_ 
 
 
 
 C8 
 1 
 
 
 
 
 Ph 
 
 
 
 
 
 -2 
 
 
 ^ 
 
 
 
 
 
 
 h 
 
 
 d 
 
 
 t-^ 
 
 
 
 
 U 
 
 
 o 
 
 
 O 
 
 
 
 
 
 
 ;;^ 
 
 m 
 
 
 
 3 
 
 (uS 
 
 
 
 O 
 
 
 
 - 
 
 
 o ^ 
 
 rt 
 
 1 
 
 tH 
 
 ^ 
 
 CD g 
 
 cs 
 
 
 
 
 
 O 
 
 S 
 
 
 ^ 02 
 
 
 
 
 
 
 ."S QJ 
 
 
 
 ^.« 
 
 
 
 Ol P< 
 
 03 
 
 
 ^ to 
 
 O 
 
 
 '5 -i^ — 
 
 
 
 
 
 
 C5 " 
 
 
 
 
 
 
 
 
 
 
 
 
 !>vr! ' 
 
 
 
 
 12 
 
 
 -" 2-fi 
 
 '3 
 
 
 ^.-CJ-^i 
 
 te 
 
 
 2 S> 
 
 o 
 
 
 ^ __ ^ 
 
 
 
 o 3 5 5 
 
 60 -S s 
 <H 'o3 3 
 
 !2; 1-^ 
 
 fe ^ 
 
 
 o-E 
 
QUALITATIVE TESTS 75 
 
 test is selected, it is important that its failings should be carefully 
 borne in mind, and that it should be constantly practised, for it 
 is only by practice that doubtful results can be satisfactorily 
 explained, and mistakes in diagnosis be guarded against. 
 
 A well-marked reduction with any of the preliminary tests 
 associated with the symptoms of diabetes leaves little doubt that 
 dextrose, and possibly also levulose, is present. If only a slight, 
 or ambiguous, reaction is obtained the result must be confirmed, 
 and for this the fermentation and phenylhydrazin tests are the 
 best. By then proceeding in the manner indicated in the preceding 
 pages, and outlined in the table (p. 74), it is possible to determine 
 the particular substance to which the reduction is due. A com- 
 plete investigation is advisable in the more marked cases also ; 
 for, although this may not be necessary for diagnostic purposes, 
 it is probable that much-needed light would be thrown on the 
 urinary changes in diabetes, and allied conditions, if more ex- 
 haustive routine analyses of the urine Avere made. 
 
 BIBLIOGRAPHY 
 
 Abeles, Centralb. f. d. med. Wissensch., 1879. 
 
 Allen, Lancet, 1894 ; Chemistry of the Urine, 1895 ; Commerc. Org. 
 Analysis, 1909, i. 
 
 Allen and ToUens, Liebig's Ann., 1890. 
 
 Baisch, Zeit. f. phys. Chem., xix., xx. 
 
 Barfoed, Journ. f. prakt. Chem., 1872; Zeit. f. analytic. Chem., xii. 
 
 Bauer, Zeit. f. phys. Chetn., 1907. 
 
 Benedict, Journ. Biolog. Chem., 1909 ; Journ. Amer. Med. As- 
 soc, 1911. 
 
 Bial, Deut. med. Woch., 1902, 1903. 
 
 Binet, Rev. d. I. Suisse romande, 1892. 
 
 Borchardat, Zeit. /. phys. Chem., 1908. 
 
 Bottger, Zeit. f. prakt. Chem., Ixx. ; Chem. Centralb., 1857. 
 
 Briicke, Wiener med. Woch., 1858 ; Sitzungsber. d. k. Akad. d. 
 Wiss. z. Wien, 1875. 
 
 Briickner, Aertzliche Rundschau, 1899. 
 
 Bruel, Arch. f. exp. Path., 1898. 
 
 Cammidge, Tr. Med. Chi. Soc, Ixxxviii. 
 
 Crismer, Pharm. Zeit., xxxiii. ; Pharm. Journ., xix. 
 
 Fischer, Berich. d. Deut. Chem. Gesellsch., 1884. 
 
 Fluciger, Zeit. f. physiolog. Chem., 1885. 
 
 Friedlander, Arch. d. Heilkunde, vi. 
 
 Geelmuyden, Jahresberich. f. Tierch., 1903. 
 
 Haas, Centralb. f. d. med. Wissensch., 1876. 
 
 Heller, Dessen. Archiv., 1844. 
 
76 GLYCOSURIA 
 
 Hoppe-Seyler, Zeit. f. phys. Chem., 1892. 
 
 Jacoby, CharM Annalen, 1898. 
 
 V. Jaksch, Zeit. /. hlin. Med., xi. 
 
 Johnson, Lancet, 1894 ; Tr. Pharmaceut. Journ., 1895 ; Proc. 
 
 Roy. Soc, xliii. ; Tr. Roy. Med. Chi. Soc, Ixxvi. 
 Jolles, Centralb.f. inn. Med., 1903. 
 Jones, Bence, Quarterly Journ. Chem. Soc, xiv. 
 Kellas and Wethered, Lancet, 1906. 
 Kowarski, Berl. Klin. Woch., 1899, 1900. 
 Kiilz, Pfliiger's Arch., xiii. 
 Langstein and Steinitz, Biochem. Zeit., 1906. 
 Levesson, Biochem. Zeit., 1907. 
 Leuken, Apoth. Zeit., i. 
 Lowenstein, Allg. med. Centralzeit., 1900. 
 
 Lohnstein, Pflilgefs Arch., 1896 ; Chem. Zentralb., 1896, 1899. 
 Luther, Chem. Centralb., 1891. 
 Maclean, Brit. Med. Journ., 1907. 
 Malay, Wiener Akad. Sitzungsber., Ixiii. 
 Malfatti, Centralb. f. Harn. u. SexvMorg., 1905. 
 Maschke, Zeit. analytic. Chem., 1877. 
 Mayer, Zeit. f. phys. Chem., 1901. 
 Mayer and Neuberg, Zeit. f. physiol. Chem., 1900. 
 Meissner and Babo, Zeit.f. Ration. Med., ii. 
 Molisch, Monatschr. d. Chem., vh. 
 Moore, Lancet, 1844. 
 Moritz, Deut. Arch. f. klin. Med., 1890. 
 
 Mnlder, Arch. f. d. holldnd. Beitrage, 1861-2. 
 
 Neubauer and Strauss, Zeit.f. phys. Chem., xxxv., 1902. 
 
 Neuberg and Wohlgemuth, Zeit.f. phys. Chem., xxxv., 1902. 
 
 Neumann, Berl. klin. Woch., 1904. 
 
 Neumayer, Deut. Arch. f. klin. Med., 1900. 
 
 Nylander, Zeit. f. physiol. Chem., 1883-4. 
 
 Pavy, Guy's Hosp. Rep., xxi. ; Physiol, of Carbohydrates, 1894 ; 
 Carbohydrate Metabolism and Diabetes, 1906. 
 
 Pinoff, Virchow's Arch., cvii. 
 
 Porcher and Nicholas, Journ. d. Physiol., 1901. 
 
 Purdy, Pract. Uranalysis, 1894. 
 
 Qumquand, C. R. d. Soc. d. Biol., 1889. 
 
 Riegler, Deut. med. Woch., 1901, 1903. 
 
 Roos, Zeit.f. phys. Chem., xv. 
 
 Rosenberger, Zeit. f. j)hys. Chem., 1906. 
 
 Rubner, Zeitschr. f. Biol., 1884. 
 
 Salkowski, Zentralb. f. med. Wissensch., 1892 ; Zeit. f. phys. Chem., 
 1879, 1899 ; Berl. klin. Woch., 1905. 
 
 Salkowski and Blmnenthal, Charite Annalen, 1898. 
 
 Schondroff, Pflilger's Arch., 1908. 
 
 Schoorl and Kalmthout, Chem. Berich., 1906. 
 Seegeli, Pflilger''s Arch., Ixiv. 
 
QUALITATIVE TESTS 77 
 
 Soxhlet, J own. /. prakt. Chem., xxi. 
 
 Tollens, Chem. Berich., 1896 ; Zeit. v. cleut. Ziickerind., 1908. 
 
 Trommer, Ann. d. Chem. u. Pharm., 1841. 
 
 Tuchen, Virchow's Arch., xxvii. ; Zeit. f. physiol. Chem., 1888. 
 
 Wedenski, Zeit. J. physiol. Chem., xiii. 
 
 Wender, Pharm,. Post, xxvi. ; Analyst, xviii. 
 
 Wheeler and Tollens, Chem. Ber., 1889 ; Lieb. Ann., 1889. 
 
 Wohle, Zeit. f. anal. Chem., 1904. 
 
 Worm-Muller, Pflilger's Arch., xvi., xvii., xxii., xxvii. 
 
 Worms, Bull. d. VAcad. d. Med. d. Paris, 1895. 
 
 Zunz, Journ. Med. de Bruxelles, 1902. 
 
CHAPTEK III 
 
 QUANTITATIVE AN'ALYSIS OF SUGARS, ACETONE BODIES, ETC., 
 IN THE URINE 
 
 Ha^tng discovered that a patient is passing sugar in the urine 
 the first point to settle is the gravity of the case. This is deter- 
 mined partly by the kind of sugar present. If lactose alone is 
 found, the glycosuria is only a temporary one, and A\ill cease when 
 the condition gi^^^g rise to it is removed ; a pentose may be of 
 alimentary origin, or be e^ddence of an inherent defect of meta- 
 boHsm which, although likely to be of a permanent nature, is not 
 of serious import ; the presence of dextrose, with or without some 
 other sugar, indicates a metabolic defect which is always of serious 
 significance, and shows that the patient is either actually, or poten- 
 tially, in a condition that is likely to result in a fatal train of 
 symptoms. The prognosis in any particular case of dextrosuria 
 depends upon (1) the intensity of the glycosuria, (2) the presence and 
 degree of secondary abnormalities of metabolism, and (3) the extent 
 to which nitrogenous equilibrium is interfered with. To determine 
 these points a quantitative analysis of the urine for (a) sugar, (b) 
 acetone bodies, (c) total nitrogen is necessary. In this chapter 
 we shall consider only the methods employed, leaving for subse- 
 quent consideration the interpretation of the results obtained. 
 
 A quantitative analysis should never be made with an odd 
 samj^le of urine, as the results are apt to be most misleacUng. The 
 whole excretion for twenty-four hours should be collected, and 
 carefully measured, starting, for examj)le, with that passed after 
 8 A.M., and continuing to collect until, but not including, all up to 
 8 A.M. next da\^ A sample from the mixed excretion having 
 been analysed, the results may then be worked out for the total 
 twenty-four hours' specimen. 
 
 (a) The reducing sugars in the urine are commonly estimated by 
 titration with an alkaline solution of copper, but an alkaline solu- 
 tion of a mercury salt, estimations with the polariscope, and by 
 fermentation may also be employed. In sj)ecial cases {e.g. the 
 pentoses, &c.) other methods are also made use of. 
 
 When sugar is kno\Mi to be present in considerable quantity, 
 
daily 
 
 excretion of 2 
 
 litres 
 
 and 
 
 a 
 
 sp. 
 
 
 3 
 
 6- 
 
 10 „ 
 
 
 
 
 QUANTITATIVE METHODS 79 
 
 a rough estimate of the amount may be arrived at by Naunyn's 
 table : — 
 
 gr of l-028-l-030 = about 2-3 % 
 1-028-1 -032= „ 3-5 % 
 l-030-l-035= „ 5-7 % 
 1-030-1 -042= „ 6-20 % 
 
 The percentage of sugar in the urine can also be approximately 
 determined from the quantity and specific gravity, by deducting 
 from the specific gravity as much as would represent the specific 
 gravity of a normal urine diluted to the same amount as that 
 passed by the patient, and multiplying the difference by 230. 
 Thus if a diabetic passes 3 litres of urine with a specific gravity 
 of 1-030, and it is assumed that a normal person passes 2 litres 
 with a specific gravity of 1-015 in the twenty-four hours, the latter 
 on being diluted to 3 litres would have a specific gravity of 1-010. 
 
 2x1-015+ 1-000 ^ j.QjQ 
 3 
 
 Deducting this from the sjsecific gravity of the diabetic urine 
 (1-030-1-010) =0-020, and (0-020 x 230) =4-6, so the urine would 
 contain about 4-6 per cent, of sugar. Such a calculation is not, 
 however, quite justifiable, for it assumes that the change in specific 
 gravity is entirely due to dextrose ; but this is not the case, since 
 alterations in the excretion of urea and inorganic salts may, and 
 generally do, occur, and so influence the result. 
 
 I. Titration with Alkaline Solutions of Copper 
 
 1. Fehling'-Soxhlet Method. — This may be used either volu- 
 metrically or gravimetrically. The latter is more accurate, but 
 the volumetric method is generally employed for clinical purposes, 
 as it is simpler and is more rapid. The duration of the operation, 
 partial re-oxidation of the precipitate, &c., are apt to introduce 
 errors, but with care and experience these disadvantages can be 
 minimised, and any unavoidable errors are probably more than 
 counterbalanced hj errors in collecting the urine and working 
 ■out the daily output. 
 
 (i) Volumetric. — The estimation is based upon an exact de- 
 coloration of a known volume of Fehling's solution by a measured 
 volume of the urine, which should be diluted to contain approxi- 
 mately 1-0 per cent, of reducing sugar. Dilution of the urine 
 tends to eliminate the effect of the normal reducing substances, 
 and is particularly necessary when a high percentage of sugar is 
 
80 GLYCOSURIA 
 
 ]3resent, since the theoretical reduction only takes place when 
 the solution contains between 0-5 and 1 per cent. ; above the latter 
 point there may be an error of 2 per cent. The greater the amount 
 of sugar, therefore the higher must be the dilution. The necessary 
 dilution may be roughly guessed at from the specific gra\dty of the 
 specimen. Thus Avith a specific gravitj^ of 1-025, or under, the 
 urine should be diluted five times, and with a specific gravity of 
 1-030, or over, ten times. Urines containing less than 0-5 per cent, 
 of sugar cannot be satisfactorily titrated with Fehling's solution 
 unless, as Hammerstein suggests, a weighed quantity of dextrose 
 sufiicient to allow of a ten times dilution is added, and alloAved for 
 in the subsequent calculations. If there is less than 0-5 per cent, 
 of sugar the fermented and unfermented urine may both be 
 titrated, and the difference in the result taken to represent the 
 fermentable sugar. 
 
 Any albumen that may be present should be removed by acidi- 
 fying, boiling, and filtering before the titration is commenced, as 
 it interferes with the settling of the precipitate. Allowance must 
 of course be made for any alteration in volume caused by the 
 process. 
 
 In carrying out the titration 10 c.c. of Fehling's solution (p. 30) 
 are generally used. This amount is completely reduced by 0-05 
 gram of dextrose in 1 per cent, solution. Other sugars give slightly 
 different figures, 
 
 ^0-0500 gram of dextrose 
 0-0543 gram of levulose 
 0-0511 gram of galactose 
 0-0678 gram of lactose 
 0-0807 gram of maltose 
 
 ^0'0430 gram of pentose (arabinose) 
 
 Knowing the dilution of the urine {d), and the number of cubic 
 centimetres (x) required to effect complete reduction of the 10 c.c. 
 of Fehling, the weight per litre of sugar present in the original 
 urine is calculated as follows : — 
 
 0-05 (&c.) X 1000 X fZ ,.^ 
 ^ = grams per litre 
 
 10 c.c. of Fehling being reduced by 
 
 From this the total daily output is arrived at by multiplying 
 the result by the measured volume for the twenty-four hours, 
 expressed in litres and decimals of a litre. If it is preferred to 
 obtain the result in grains per fluid ounce, it is only necessary to 
 remember that 1 gram = 15-5 grains approximately, and that 
 a fluid ounce = 28-4 c.c. 
 
QUANTITATIVE METHODS 81 
 
 For accurate work the Fehling solution must be standardised 
 by being titrated against a solution of dextrose containing exactly 5 
 grams per litre, or, since pure cane-sugar is more easily obtained than 
 pure dextrose, with a solution of invert sugar of known strength. 
 
 To prepare a standard solution of invert sugar dissolve 4-75 grams 
 of pure crystallised cane-sugar, that has been dried at 100° C, in 75 c.c. 
 of distilled water, and add 5 c.c. of pure hydrochloric acid (sp. gr. 
 1*188). Place the mixture in a water-bath at 70° C, and keep at that 
 temperature for ten minutes. Cool rapidly to 20° C, exactly neutralise 
 the acid with sodimn hydrate, and dilute to 1000 c.c. with distilled 
 water. Every 10 c.c. of this solution contains 0-05 gram of dextrose, 
 and should reduce 10 c.c. of Fehling solution completely. If the 
 Fehling solution is too strong the necessary dilution can be determined 
 from the following equation : — 
 
 Nxd _ „ 
 n 
 
 (Q =the number of c.c. of water to be added to the remaining 
 
 solution. 
 N =the number of c.c. of solution remaining after the titration. 
 n =the number of c.c. consumed in one titration. 
 d =the difference between the niunber of c.c. theoretically 
 
 required, and actually consumed, in the titration. 
 
 Where^ 
 
 To estimate the amount of sugar in a given specimen of urine 
 the procedure is as follows : — 
 
 By means of a graduated pipette, 5 c.c. of Fehling's solution " A," 
 and 5 c.c. of solution " B," are carefully measiired, and mixed, in a 
 small flask. The mixtiu-e is then diluted with 40 c.c. of distilled water, 
 a few fragments of clay-pipe stem are introduced to prevent bumping, 
 and it is boiled on a wire-gauze, or sand-bath. The urine, which has 
 been previously diluted if necessary, is vvm. in from a burette, adding 
 1 or 2 c.c. at a time at first, and boiling between each addition. 
 Later, as the end-point is neared, the m"ine is added drop by drop, gently 
 boiling meanwhile. The end of the reaction is reached when, on re- 
 moving the flask and allowing the cuprous oxide precipitate to settle, 
 the supernatent fluid is no longer blue when viewed against a white 
 surface. The result obtained with this first estimation will, however, 
 be only approximate, as a portion of the suboxide dissolves in the 
 ammonia liberated by the prolonged boiling of the urine and becomes 
 re-oxidised. It is therefore desirable to carry out a second titration, 
 pouring the approximate amount of urine required to effect reduction 
 into the Fehling solution to start with, boiling for two minutes, re- 
 moving the flame, and examining the clear layer of fluid just below the 
 meniscus. This is most satisfactorily accomplished by lookmg through 
 the flask with the eye on a level with the meniscus against a piece of 
 white paper held in front of a window. If the solution is foimd to be 
 
 F 
 
82 GLYCOSUEIA 
 
 still slightly blue, or the 23recipitate is slow in settling and a clear line 
 does not quickly form, an insufficient quantity of urine has been added, 
 or it has not been sufficiently diluted, and a tliird titration must be 
 performed, using a little more urine, or a more dilute specimen. If, on 
 the other hand, the fluid has a yellow tinge, too much sugar is present, the 
 excess having given Moore's test, and another estimation with a smaller 
 quantity of virine must be carried out. By a series of trials in this 
 way the amoTont of urine required to at once decolorise 10 c.c. of 
 Fehling's solution may be determined. (Soxhlet's modification.) 
 
 Various other methods of using Fehling's solution for tlie 
 estimation of sugar in the urine have been proposed, some of them 
 depending upon the number of drops of urine required to de- 
 colorise the solution, but none of them are as accurate as that 
 just described, and many give most unreliable results. 
 
 The chief practical difficulty in carrying out volumetric estima- 
 tions of sugar A\'ith Fehling's solution is the determination of the 
 end-point. The colour of the mass of the fluid cannot be relied 
 upon, as it is full of the red precipitate of copper oxide, and to 
 attempt to remove this by filtration vitiates the result, since it is 
 partially dissolved by the ammonia present, and is re-oxidised on 
 contact with the air to the blue cupric salt. The fluid cannot 
 be allowed to stand too long for the precipitate to settle, as the 
 yellow h3"droxide deposited on the glass gives to it the comple- 
 mentary colour, blue, and, moreover, re-oxidation from the air 
 is very rapid. For the latter reason it is also necessary to carry 
 out the titration as quickly as possible, and it is advisable to con- 
 duct it in a flask, in an atmosphere of steam, so that the influence 
 of the atmospheric oxygen may be avoided as much as possible. 
 It has been suggested that the end-point should be determined 
 by touching a piece of filter-paper soaked in a dilute solution of 
 potassium ferrocj^anide with a drop of the solution, or filtering 
 a few drops of the liquid through a small filter into a mixture of 
 acetic acid and dilute potassium ferrocyanide contained in a porce- 
 lain crucible or placed on a white plate. If copper is still present 
 a more or less marked brown coloration will be given at once. 
 At times, however, it is difficult to separate the hydroxide j)re- 
 cipitate effectually by filtration, and a doubtful result is obtained 
 in consequence. Ling, Rendle, and Jones have suggested ferrous 
 thiocyanate as the indicator, and do not filter the solution. When 
 a drop of this reagent is brought into contact with a drop of the 
 solution, removed "svith a glass rod, it gives the characteristic red 
 colour of ferric thiocyanate if any unreduced cupric salt is still 
 present. 
 
QUANTITATIVE METHODS 83 
 
 The ferrous thiocyanate indicator is prepared by dissolving 1 gram of 
 ferrous anunonium sulphate, and 1'5 grams of ammonium thiocyanate, 
 in 10 c.c. of water at 120° C, and immediately cooling, 2-5 c.c. of con- 
 centrated hydrochloric acid are then added. The resulting brown 
 solution is decolorised by adding a trace of zinc dust. On keeping, 
 the brown colour retiorns, but may be removed with a little more zinc 
 dust. After being decolorised several times the delicacy of the reagent 
 is reduced, but as it is at first too delicate it is best to prepare it the 
 day before it is required, and use it after the second decolorisation. 
 
 It has been proposed to facilitate the separation of the -pve- 
 cipitate by the addition of pure calcium carbonate, or calcium 
 chloride (10 per cent.), to the Fehhng solution, but most observers 
 find that they offer no great advantage. 
 
 Various modifications of Fehling's method have been intro- 
 duced with a view to rendering the end reaction more definite. 
 Among these may be mentioned — 
 
 (a) Gerrard's Cyanide Process. — Depends upon the formation 
 of a colourless double cyanide of potassium and copper, when a 
 solution of potassium cyanide is added to Fehling's solution. This 
 cyanide has the power of holding in solution any cuprous oxide 
 formed by the reduction of an excess of Fehling's solution by any 
 sugar present, so that the end of the reduction is indicated by de- 
 colorisation of the fluid, but without the formation of a precij)itate. 
 
 Ten c.c. of freshly prepared Fehling's solution are diluted to 50 c.c. 
 with water, and boiled. To the boiling liquid is then added a 5 per cent, 
 solution of potassium cyanide, with constant stirring, until the blue 
 coloiu- is just discharged, carefully avoiding an excess. Another 10 c.c. 
 of Fehling are then introduced, and the urine is rapidly run in from a 
 bi-irette, boiling, and stirring the liquid meanwhile, until the blue colour 
 is just discharged. Only the second 10 c.c. of Fehling undergoes re- 
 duction and is reckoned in the calculation. This process is strongly 
 recominended by Allen, but it is not easy to add the exact amount of 
 potassium cyanide required, and, moreover, objectionable cyanide 
 fumes escape during the heating process. 
 
 (6) Wilhamson and others avoid the difficult}^ of the end- 
 reaction by adding potassium ferrocyanide. As the ferrocyanide 
 itself exerts some reducing action on the copper solution, very 
 accurate results are not possible. 
 
 Williamson's method is carried out as follows : 10 c.c. of Fehling's 
 solution are diluted with 20 c.c. of distilled water, and 4 c.c. of a ^L per 
 cent, solution of potassiimi ferrocyanide added. The diluted iirine is 
 then rvm, drop by drop, into the boiling mixture imtil the blue colour 
 has entirely disappeared. 
 
 Repiton uses potassium ferrocyanide with Pastem-'s copper solution 
 
84 GLYCOSURIA 
 
 (136 grams of sodium hydroxide, 80 grams of potassium hydroxide, 
 and 105 grams of tartaric acid, in 500 c.c. of water ; and 45 to 48 grams 
 of crystallised copper sulphate in 500 c.c. of water). To 5 c.c. of this 
 solution is added 0*2 gram of potassium ferrocyanide, and the mixtixre 
 is heated to boiling. The urine is then added, drop by drop, until 
 the blue disappears, and is replaced by a golden-yellow coloration. 
 He states that the ratio of dextrose required to reduce this solution 
 as compared with the ordinary copper solution is 82 : 100. 
 
 (c) DrescheVs gaunin method (Klimmer) is based upon the fact 
 that, in the presence of gaunin, the red suboxide forms a less 
 readily oxidisable compound of a white colour. It gives a sharp 
 end-reaction, but is only satisfactory with pure gaunin, which is 
 very expensive. 
 
 A solution of gaunin, containing 0-00755 gram per c.c. is prepared 
 by dissolving 9 "375 gxams of gaunin hydrochloride in 1000 c.c. of a 1 per 
 cent, solution of sodiiom hydrate: 15 c.c. of this solution are added to 
 10 c.c. of Fehling's solution, and the ixiixture diluted with 25 c.c. of 
 distilled water. This is then titrated with the diluted urine in the 
 usual way. If the reduction produces a red colour the titration must 
 be repeated with the addition of more gaunin to the Fehling solution. 
 
 (d) Bancfs method is now very largely used for clinical work, as 
 it has a definite end-point ; is rapid and reliable. To obtain good 
 results, however, it is essential that the directions should be fol- 
 lowed in every detail, particularly as regards the j)urity and con- 
 centration of the components of the standard solutions, the tem- 
 peratures at Avhich they are dissolved and mixed, the time and 
 method of boiling, &c. Bang's method is based u^son the fact 
 that cuprous oxide, in the presence of potassium thiocyanate, forms 
 cuprous thiocyanate, if the solution contains only alkaline car- 
 bonates, and no alkali hydroxide. The excess of copper not re- 
 duced is determined by converting it into cuprous thiocyanate 
 with hydroxylamine. 
 
 Two solutions are prepared — (A) 500 grains of potassium carbonate, 
 100 grams of potassium hydrogen carbonate, and 400 grams of potas- 
 sium thiocyanate are dissolved in about 1200 c.c. of distilled water at 
 50° to 60° C, and cooled to 30° C. Exactly 25 graixis of pure copper 
 sulphate (CuS04-r5Br20), previously dissolved in 150 c.c. of hot dis- 
 tilled water, and cooled, are then added. After standing for twenty- 
 four hours, the mixture is made up to 2000 c.c. and filtered. (B) 200 
 grams of potassium thiocyanate, dissolved in about 1500 c.c. of dis- 
 tilled water, and 6*55 grams of hydroxylamine sulphate, dissolved in 
 aboiit 300 c.c. of water, are mixed together. The mixture is then made \vp 
 to 2000 c.c. This solution must be carefully standardised, and be titrated 
 against the " A " solution. It must be kept in a dark-coloiu-ed bottle 
 with a few grains of thymol added to prevent the growth of fvuigi. 
 
QUANTITATIVE METHODS 
 
 85 
 
 To carry oiit the estimation 50 c.c. of the copper solution (A) are 
 placed in a 200 c.c. Jena-glass flask, and 10 c.c. of the urine, or less 
 made up to 10 c.c. with water, are accurately measured out with a 
 pipette and added. The ixiixture is then heated to boiling on a wire- 
 gauze {not a sand-bath, or asbestos board), and boiled for exactly three 
 minutes. The solution must remain of a distinct blue tint ; if it is 
 completely decolorised the experiment must be commenced afresh, 
 using a smaller quantity of urine. The flask is now rapidly cooled to 
 the room temperatvu-e in a stream of cold water. The hydroxylamine 
 solution (B) is then run in until the solution is decolorised and the 
 last trace of green disappears, leaving only a yellow tint when the flask 
 is held over a white slab or paper. Since 50 c.c. of the copper solution 
 (A) correspond to about 60 mg. of sugar (50 ixig. of dextrose = 0-1 376 
 grams of copper), and the amount of unreduced copper is shown 
 by the quantity of hydroxylamine solution (B) that is used, the qtian- 
 tity of sugar present in the urine taken may be calculated, but as 
 experiment has shown that the relation is not quite constant for 
 different strengths it is better to employ an empirical table. (See p. 86). 
 
 If the urine contains 0-6 per cent., or over, of sugar the 50 c.c. of 
 copper solution will be completely decolorised before the mixture has 
 been boiled for the full three minutes, so that it should be previously 
 diluted according to the following table : — 
 With a sugar-content of 0-0-0"5 % use 10 c.c. 
 
 ,, 5 c.c. 
 
 0-5-1 -0 
 
 1 •0-2-5 
 
 2-5-6-0 
 
 over 6-0 
 
 2 c.c. 
 1 c.c. 
 
 of urine 
 
 ,, and 5-0 c.c. water 
 
 ,, and 8-0 c.c. ,, 
 
 ,, and 9-0 c.c. ,, 
 
 0-5 c.c. 
 
 and 9-5 c.c. 
 
 The degxee of dilution necessary can usually be estimated fairly 
 accTorately from a consideration of the specific gravity, and the rate at 
 which the copper is reduced in the qualitative tests. Albumen does not 
 interfere with the reaction, and need not therefore be renaoved. 
 
 (e) Benedict has modified his qualitative test solution for use in 
 quantitative work by adding potassium sulphocyanate and potas- 
 sium ferrocyanide, with the result that the reduced copper is pre- 
 cipitated as cuprous sulphocyanate, a snow-white compound, which 
 is rather an aid than a hindrance to accurate observation of the 
 disappearance of the last trace of blue colour. 
 
 The solution for cjuantitative work has the following composition : — 
 
 grams or c.c. 
 
 Copper sulphate (pure crystallised) . . . . . 18-0 
 
 Sodium carbonate (crystallised) '...... 200-0 
 
 Sodium, or potassium, citrate ...... 200-0 
 
 Potassium sulphocyanate ....... 125-0 
 
 Five per cent, potassimn ferrocyanide solution . . . 5-0 
 
 Distilled water to make a total volmne of ... . 1,000-0 
 
 ^ One-half the weight of the anhydrous salt may be used. 
 
GLYCOSURIA 
 
 Bang's Method 
 
 C.c. of 
 
 Mg. of 
 
 c.c. of 
 
 Mg. of 
 
 c.c. of 
 
 Mg. of 
 
 Bang, "B." 
 
 Sugar. 
 
 Bang, "B." 
 
 Sugar. 
 
 Bang, "B." 
 
 Sugar. 
 
 0-75 
 
 60-0 
 
 19-00 
 
 31-4 
 
 37-50 
 
 10-9 
 
 1-00 
 
 59-4 
 
 19-50 
 
 30-8 
 
 38-00 
 
 10-4 
 
 1-50 
 
 58-4 
 
 20-00 
 
 30-2 ! 
 
 38-50 
 
 9-9 
 
 2-00 
 
 57-3 
 
 20-50 
 
 29-6 
 
 39 00 
 
 9-4 
 
 2-50 
 
 56-2 
 
 21-00 
 
 29-0 
 
 39-50 
 
 9-0 
 
 3-00 
 
 55-0 
 
 21-50 
 
 28-3 
 
 40-00 
 
 8-5 
 
 3-50 
 
 54-3 
 
 22-00 
 
 27-7 
 
 40-50 
 
 8-1 
 
 4-00 
 
 53-4 
 
 22-50 
 
 27-1 
 
 41-00 
 
 7-6 
 
 ■ 4-50 
 
 52-6 
 
 23-00 
 
 26-5 
 
 41-50 
 
 7-2 
 
 5-00 
 
 51-6 
 
 23-50 
 
 25-8 
 
 42-00 
 
 6-7 
 
 5-50 
 
 50-7 
 
 24-00 
 
 25-2 
 
 42-50 
 
 6-3 
 
 600 
 
 49-8 
 
 24-50 
 
 24-6 
 
 43-00 
 
 5-8 
 
 6-50 
 
 48-9 
 
 25-00 
 
 24-1 
 
 43-50 
 
 5-4 
 
 7'00 
 
 48-0 
 
 25-50 
 
 23-3 
 
 44-00 
 
 4-9 
 
 7-50 
 
 47-2 
 
 26-00 
 
 22-9 
 
 44-50 
 
 4-5 
 
 8-00 
 
 46-3 
 
 26-50 
 
 22-3 
 
 45-00 
 
 41 
 
 8-50 
 
 45-5 
 
 27-00 
 
 21-8 
 
 45-50 
 
 3-7 
 
 9-00 
 
 44-7 
 
 27-50 
 
 21-2 
 
 46-00 
 
 3-3 
 
 9-50 
 
 44-0 
 
 28-00 
 
 20-7 
 
 46-50 
 
 2-9 
 
 10-00 
 
 43-3 
 
 28-50 
 
 20-1 
 
 47-00 
 
 2-5 
 
 10-50 
 
 42-5 
 
 29-00 
 
 19-6 
 
 47-50 
 
 2-1 
 
 11-00 
 
 41-8 
 
 29-50 
 
 19-1 
 
 48-00 
 
 1-7 
 
 11-50 
 
 41-1 
 
 30-00 
 
 18-6 
 
 48-50 
 
 1-3 
 
 12-00 
 
 40-4 
 
 30-50 
 
 18-0 
 
 49-00 
 
 0-9 
 
 12-50 
 
 39-7 
 
 3100 
 
 17-5 
 
 49-50 
 
 0-5 
 
 13-00 
 
 39-0 
 
 31-50 
 
 17-0 
 
 50-00 
 
 0-0 
 
 13-50 
 
 38-3 
 
 32-00 
 
 16-5 
 
 
 
 14-00 
 
 37-7 
 
 32-50 
 
 15-9 
 
 .. 
 
 
 
 14-50 
 
 371 
 
 33-00 
 
 15-4 
 
 
 
 
 15-00 
 
 36-4 
 
 33-50 
 
 14-9 
 
 
 
 
 15-50 
 
 35-8 
 
 34-00 
 
 14-4 
 
 .. 
 
 
 
 16-00 
 
 35-1 
 
 34-50 
 
 13-9 
 
 ■• 
 
 
 
 16-50 
 
 34-5 
 
 35-00 
 
 13-4 
 
 .. 
 
 
 
 17-00 
 
 33-9 
 
 35-50 
 
 12-9 
 
 
 
 
 17-50 
 
 33-3 
 
 36-00 
 
 12-4 
 
 
 
 
 18-00 
 
 32-6 
 
 36-50 
 
 11-9 
 
 
 
 
 18-50 
 
 32-0 
 
 37-00 
 
 11-4 
 
 1 
 
 
 
 With the aid of heat dissolve the carbonate, citrate, and suJpho- 
 cyanate in enough water to make about 800 c.c. of the mixture, and 
 filter if necessary. Dissolve the copper sulphate separately in about 
 100 c.c. of water, and pour the solution slowly into the other liquid, 
 with constant stirring. Add the ferrocyanide solution, cool, and 
 dilute to exactly one litre. Of the various constituents, the copper 
 salt only need be weighed with exactness. Twenty-five c.c. of the 
 reagent are reduced by 50 mg. of glucose. 
 
 Sugar estimations are conducted as follows : The urine, 10 c.c. of 
 
QUANTITATIVE METHODS 87 
 
 which should be dikited with water to 100 c.c. (unless the sugar-content 
 is believed to be low), is povired into a 50 c.c. burette up to the zero 
 mark. Twenty-five c.c. of the reagent are measured with a pipette into 
 a porcelain evaporation dish (25 to 30 cm. in diameter), 10 to 20 grams 
 of crystallised sodium carbonate (or one-half the weight of the anhy- 
 drous salt) are added, together with a small quantity of powdered 
 pumice stone, or talcum, and the mixture heated to boiling over a free 
 iiame until the carbonate has entirely dissolved. The diluted urine 
 is now run in from the burette, rather rapidly, until a chalk- white 
 precipitate forms, and the blue colour of the mixtvire begins to lessen 
 perceptibly, after which the solution from the biu-ette must be run in a 
 few drops at a time, until the disappearance of the last trace of blue 
 colour, which marks the end-point. The solution must be kept vigor- 
 ously boiling throughout the entire titration. If the mixture becomes 
 too concentrated during the process, water may be added from time 
 to time to replace the volume lost by evaporation. The calculation of 
 the percentage of sugar in the original sample of urine is very simple. 
 The 25 c.c. of copper solution are reduced by exactly 50 mg. of glucose. 
 Therefore the volume run out of the burette to effect the reduction 
 contained 50 mg. of the sugar. When the urine is diluted 1 : 10, as in 
 the usual titration of diabetic lu-ines, the formula for calculating the 
 percentage of sugar is the following — 
 
 X 1000= per cent, in original sample, 
 
 where x is the number of c.c. of the diluted urine required to reduce 
 25 c.c. of the copper solution. 
 
 In this method chloroform must not be present dxwing the titration. 
 If used as a preservative in the lu-ine it may be removed by boiling a 
 sample for a few minutes, and then diluting to its original volume. 
 
 This reagent, like that employed for qualitative work, will keep 
 indefinitely. Benedict states that repeated determinations, and com- 
 parisons of the results with those obtained by the polariscope, and by 
 AUihn's gravimetric process, have shown the method to be probably 
 more exact than any other titration method available for sugar 
 work. 
 
 (/') Pavy's Ammoniacal Copper Method. — This modification of 
 Fehling's solution is based upon the fact that, in the presence of 
 a sufficient excess of ammonium hydroxide, the cuprous oxide is 
 not precipitated, but forms a colourless solution, so that the end 
 of the reaction is indicated by decolorisation of the blue fluid. 
 
 As the ammoniacal cuprous solution is very readily oxidised, it 
 is necessary to prevent access of air during the titration. This may be 
 done by attaching the nozzle of the burette containing the urine to a 
 tube passing through an indiarubber stopper closing the flask that 
 contains the copper solution. A second tube conveys the steam and 
 ammonia gas generated into a flask of cold water. To prevent the 
 
88 GLYCOSURIA 
 
 tendency to " suck back," the delivery end of this tube should dip 
 below the siorface of a little mercury at the bottom of the flask. A 
 better arrangement is to have a third tube in the stopper by which a 
 slow current of hydrogen, or coal-gas, is passed through the flask con- 
 taining the boiling copper solution. When coal-gas is used a red film 
 of cuprous acetylide is apt to form on the siu-face, but can be disre- 
 garded. 
 
 To prepare the ammoniacal Pavy's solution 120 c.c. of Fehling's 
 solution are mixed with 300 c.c. of strong ammonia (sp. gr. 0'880), 
 and 400 c.c. of a 12 per cent, solution of sodium hydroxide (sp. gr. 1"14). 
 The mixture is then made up to 1000 c.c. One hundred c.c. of this 
 solution corresponds to 10 c.c. of Fehling's solution — that is to say, is 
 reduced by 0-05 gram of dextrose. 
 
 To carry out the titration 100 c.c. of the solution are placed in a 
 flask, a few fragments of pumice, or pipe-stem, added, the biirette and 
 tubes adjusted, and the liquid brought to boiling. The diluted urine 
 is then run in from the burette until the blue colovir of the liquid has 
 entirely disappeared, the boiling being meanwhile continued regularly. 
 The end-reaction is very sharply marked, and may be confirmed by 
 bubbling air through the liquid directly decolorisation is complete. 
 The blue colour should reappear in a few seconds unless too much 
 urine has been used, when a longer time will elapse. Reduction occurs 
 much more slowly than with Fehling's solution, and Pavy's solution 
 has a different oxidising power for maltose, lactose, and invert sugar. 
 The practical disadvantages of this method are the inconvenience of 
 working with the ammoniacal solution, the great dilution of the solution 
 as compared with Fehling's, and the necessity for using a special 
 apparatus. 
 
 The following method of carrying out the test, which is simpler and 
 more convenient, is sometimes used in clinical work, it is, however, 
 less accxarate : 10 c.c. of Pavy's sokition are accurately measured into a 
 wide test-tube, a few fragments of clay -pipe are dropped in, and 8 or 
 10 drops of petroleimi, or paraffin-oil, are poured on the stu-face to 
 exclude the air. The test-tube is then suspended in the neck of a wide- 
 mouthed flask containing hot water, which is heated until the contents 
 of the tube boil. The urine to be tested is treated with an eqiial volume 
 of ammonia, and filtered from the precipitated phosphates. A known 
 volume of the filtrate is then further diluted with a definite quantity 
 of water, according to the proportion of sugar believed to be present. 
 The diluted urine is added, drop by drop, to the boiling Pavy's solution 
 from a graduated pij)ette, or small burette, until the disappearance of 
 the colour indicates the termination of the reaction. The 10 c.c. of 
 Pavy's solution used corresponds to 0-005 gram of sugar. 
 
 Healthy urine may exert a reducing action on Pavy's solution equal 
 to a liquid containing 0-1 to 0-3 per cent, of dextrose. Of this one 
 quarter is ascribed to uric acid (removable by lead acetate), and the 
 remainder to creatinin (removable by mercuric chloride). 
 
jCopper siilphate cryst. 4-158 grs, 
 \Distilled water to . . 500 c.c. 
 
 QUANTITATIVE METHODS 89 
 
 Sahli advises a more concentrated solution, which is prepared in 
 t wo parts that are mixed in equal volumes as required :— 
 
 'Sodium-potassivim tartrate 20*4 
 Potassium hydroxide . 20*4 
 
 Ammonia (sp. gr. 0-88) . 300-0 
 ^Distilled water to . . . 500-0 
 
 Five c.c. of each of these solutions are placed in a flask, holding 
 about 100 c.c, and 30 c.c. of water added. The contents are heated 
 over an asbestos board, with a micro-burner that can be regulated, so 
 that they are gently, but not vigorously boiled. The urine, diluted 
 so that a fairly large amount, but not a sufficient quantity to appreci- 
 ably decrease the concentration of the alkali in the solution, is rim 
 in from a burette, taking care that the boiling is never interrupted, 
 until the fluid is decolorised. 
 
 {g) Purdy's Method. — A solution prepared by mixing 4-752 grams of 
 copper sulphate, dissolved in water, with 23-5 grams of potassium 
 hydrate, previously dissolved in water, 350 c.c. of ammonia (sp. gr. 0-9), 
 and 38 c.c. of pure glycerine, is made up to 1000 c.c. with distilled 
 water. To 35 c.c. of this fluid, placed in a 150 to 200 c.c. flask, is added 
 enough water to half fill the flask. The flask is then closed with a 
 perforated cork. Through one hole the tip of the burette containing 
 the urine is passed, and through another a tube bent at right angles 
 to carry the steam and fmnes away from the face. The fluid is brought 
 to the boil, and the urine added drop by drop, until the blue coloixr 
 begins to fade ; then more slowly, three to five seconds elapsing between 
 each drop, until the blue colovir completely disappears and leaves the 
 test solution perfectly transparent and coloiirless. If it is boiled too 
 long, so that too much ammonia escapes, the copper hydroxide will 
 precipitate. If there is over 5 per cent, of sugar the urine should be 
 diluted with three vohunes of water {i.e. four times). Each 35 c.c. 
 of Purdy's fluid is reduced by 0-02 gram of dextrose. 
 
 (h) The iodometric method (Lehmann) consists in heating the urine 
 with Fehling's solution, filtering, treating with sulphuric acid and 
 potassium iodide, and titrating the Hberated iodine with sodium 
 thiosulphate and starch. It is based on the principle that sugar 
 precipitates a proportionate amount of the copper in the Fehhng 
 solution, and that the precipitated cuprous oxide remains behind 
 on the filter. By the action of the sulphuric acid and potassium 
 iodide on the filtrate, one atom of iodine is liberated for each molecule 
 of the copper sulphate. The tedious decolorisation of the Fehhng 
 test is avoided, and the iodine reaction in its place is much plainer. 
 Only the copper still in solution — not that already reduced — 
 induces the liberation of the iodine, and the cuprous oxide becomes 
 quantitatively transformed into cuprovts iodide. According to 
 Citron, the best technique is as follows : — 
 
90 GLYCOSURIA 
 
 Ten c.c. of the potassimn-sodium tartrate solution (see below) are 
 placed in a porcelain dish. Ten c.c. of copper solution are measured 
 out, and nearly all is poured in. The mixture is now thoroughly 
 stirred. Then exactly 1 c.c. of urine is introduced, and, when all is 
 dissolved, the mixture is heated over a small flame for a minute and a 
 half, to boiling, and then cooled by placing the dish in a vessel of cold 
 water. (While it is cooling the starch solution is prepared.) On add- 
 ing 4 c.c. of a 25 per cent, solution of sulphiiric acid a thick, viscid pre- 
 cipitate is at once thrown down. The solution of sodium thiosulphate 
 is then run in from a bm-ette, graduated for percentages, a line at a 
 time, and a few c.c. of the starch solution are added. The fluid then 
 turns deep blue, or possibly a little more of the sulphuric acid solution 
 and starch solution may be required. More of the thiosulphate solu- 
 tion is then run in, a line at a time, until the fluid turns milk white, 
 and remains so even when it is stirred. The rest of the copper solution 
 is then poLired in, and also the rinsing water. If the fluid still remains 
 white, the percentage line on the brunette is the index of the proportion 
 of sugar in the urine. If a blue tint rettu-ns, more of the sodium thio- 
 sulphate is added, drop by drop, until a permanent milky tint results. 
 Citron uses a special isodometric burette graduated for m-ine percen- 
 tages, and a special pijDctte graduated for 1 c.c. of lu-ine, used with a 
 syringe with a capacity of 2 c.c. 
 
 The solutions used are (a) 34'639 grams of copper sulphate dissolved 
 in 500 grams distilled water, to which 1 c.c. of sulphuric acid is added 
 to prevent the formation of copper carbonate ; (b) 173 grams potassium- 
 sodium tartrate, 50 grams sodivmi hydroxide, and 100 grams potassium 
 iodide dissolved in 500 c.c. water ; (c) a one-fifth normal sokition of sodium 
 thiosulphate (49*6 grams per litre), containing also 0-1 per cent, sodium 
 arsenate to make it keep better ; (d) a 25 per cent, solution of svilphuric 
 acid ; and (e) the starch solution, which he makes by dissolving in half 
 a reagent glass of water a small piece, the size of a hazel nut, of a paste 
 preparation used by jDhotographers. The starch solution is heated to 
 boiling and then cooled. 
 
 If an ordinary burette is used the percentage of sugar may be 
 calculated from the following data — 1 c.c. of the thiosulphate solution 
 = 0-0127 gram of copper, and 10 c.c. of the copper solution=13"9 c.c. 
 of the thiosulphate. If " T " be the number of cubic centimetres of 
 thiocyanate used (13'9 — T x 0-0127) = the amount of reduced copper, 
 and from this the quantity of dextrose present can be calculated by 
 referring to the table given on page 91. 
 
 (ii) Gravimbtiiic Estimation of Reducing Sugars with 
 Fehling's Solution. — This method gives very accurate results, 
 provided the details of the manipulation are closely attended to. 
 It is too complicated for routine clinical use, but is necessary for 
 careful scientific investigation. 
 
 Pfluger-Allihn Method. — 34"630 grams of crystallised copper sul- 
 phate are dissolved in distilled water, and diluted to 500 c.c. A second 
 solution containing|173 grams of sodimn-potassiimi tartrate, and 125 
 
QUANTITATIVE METHODS 
 
 91 
 
 grams of potassimn hydroxide, dissolved in water, and made up to 
 500 c.c. is also prepared. These solutions are preserved separately, 
 and are mixed in equal volumes for use. To carry out the estimation 
 60 c.c. of the mixtiu-e (30 c.c. of A, and 30 c.c. of B) are placed in a 
 beaker holding about 300 c.c, diluted with 60 c.c. of water, and heated 
 on a sand-bath to boiling. Twenty-five c.c. of the urine, diluted so as 
 to contain not more than I per cent, of sugar, are then added to the 
 actively boiling fluid, and the mixttire boiled for exactly two minutes. 
 The precipitated red oxide is immediately filtered off, and the amoimt 
 of copper determined by one of the following methods : — 
 
 (a) Reduction in Hydrogen. — The cuprous oxide is collected on an 
 asbestos film supported by a perforated disc, or cone, of platinum in a 
 hard glass tube, which has been previously weighed, using suction. 
 After all the precipitate has been transferred to the filter, it is well 
 washed with hot water, then with alcohol, and afterwards with ether, 
 being careful that no precipitate is left near the top of the tube, where 
 it would be removed by the cork used during the process of reduction. 
 After drying, the tube is connected with an apparatus supplying a 
 current of dry hydrogen, and is gently heated until the cuprous oxide 
 is completely reduced to the metallic state. It is then cooled in a current 
 of hydrogen, and weighed. 
 
 (6) Direct Weighing of Cuprous Oxide. — Prepare a gooch crucible 
 with an asbestos felt. First thoroughly wash the asbestos with water, 
 to remove small particles, then follow successively with 10 c.c. of 
 alcohol and 10 c.c. of ether, and dry the crucible and its contents for 
 thirty minutes in a water-oven at the temperatxire of boiling water. 
 Cool, and weigh. Collect the precipitate of cuprous oxide on the 
 asbestos felt, thoroughly wash with hot water, then with alcohol, and 
 finally with ether. Dry the precipitate for thirty minvites in a hot 
 water-oven at the temperature of boiling water. Cool, and weigh. 
 The weight of cuprous oxide multiplied by 0-8883 gives the weight of 
 metallic copper. 
 
 (c) Electrolysis. — The metallic copper in the precipitate may also 
 be deposited electrolytically from a nitric, or sulphiu-ic, acid solution, 
 on to platinimi and then weighed, or it may be estimated volumetri- 
 cally by the perraanganate method. (See Allen's Commercial Organic 
 Analysis, vol. i. p. 323, 1909.) 
 
 The weight of sugar corresponding to the amoiint of copper, or 
 copper oxide, obtained by any one of these methods is calculated by 
 the use of the following factors : — • 
 
 
 Dextrose. 
 
 Lactose. Maltose. 
 
 Cane-su!,'ar after 
 Inversion. 
 
 Oil . . . 
 Cu,0 . . 
 CuO . . . 
 
 0-5634 
 0-5042 
 0-4535 
 
 •7707 0-9089 
 0-6843 0-8132 
 0-6153 0-7314 
 
 0-5395 
 0-4790 
 0-4308 
 
 Or from Allihn's tables. 
 
92 GLYCOSURIA 
 
 II. Estimation with Alkaline Solutions of Mercury 
 
 Several metliods of estimating sugars by their reducing action 
 on mercuric solutions have been described, but the one that is 
 generally used in clinical work is Knapp's method. 
 
 (a) Knapp's Method. — This depends upon the fact that an 
 alkaline solution of mercuric cyanide is reduced to metallic mercury 
 in the presence of sugar. The titration can be satisfactorily carried 
 out in artificial light, and is apxalicable when only a small amount 
 of sugar is present. The solution keeps well. It is reduced by 
 creatin, creatinin, &c., and all albumen must be carefulty removed 
 before carrying out the titration. The urine should not contain 
 more than 0-5 to TO per cent, of sugar, and must therefore be 
 diluted if necessary. 
 
 Knapp's solution is prepared by dissolving 10 grams of pure, dry, 
 mercuric cyanide in water, adding 100 c.c. of sodium hydroxide solution 
 (sp. gr. 1-145), and diluting to 1000 c.c. One hundred c.c. of this solu- 
 tion are equivalent to 0"202 grams of dextrose, in 0*5 per cent, solution, 
 and 0*201 gram, in 1 per cent, solution. For practical purposes 20 c.c. 
 of the solution may be taken as equivalent to O'Oo gram of dextrose. 
 
 Twenty c.c. of the solution are diluted with 80 c.c. of water, heated to 
 boiling, and the urine, diluted so as to contain about 0*5 per cent, of 
 sugar, is run in as quickly as possible, until the whole of the mercury 
 is reduced, gently boiling all the time. As the end-point is reached the 
 solution begins to clear, and the mercury, together with the phosphates, 
 settles to the bottom. The final point is determined by moistening a 
 strip of white filter -paper with the clear fluid and exposing it to the 
 fumes of concentrated hydrochloric acid and sulphuretted hydrogen 
 successively. If the cyanide of mercury has not been completely re- 
 duced a yellow spot will appear. As soon as the mercury has been 
 completely reduced the reading is taken, and the number of grams per 
 litre of sugar worked out according to the formiila — 
 
 0-05 X 1000 x(^ . 1 . 
 = grams oi dextrose per litre 
 
 where d = the dilution of the lu'ine, and x = the number of c.c. con- 
 sumed in the titration. 
 
 (6) Sachsse's Method. — The mercuric solution is prepared by dis- 
 solving 18 grams of piore, dry, mercuric iodide in a solution of 25 grams 
 of potassium iodide, adding 80 grams of potassium hydroxide, and 
 making up to 1000 c.c. Twenty c.c. of this solution are boiled in a 
 porcelain basin, and the diluted urine is gradually added. The com- 
 pletion of the reduction is indicated when a drop of the supernatent 
 liquid ceases to give a brown colour with a drop of a very alkaline 
 solution of stannous chloride. 
 
 Soxhlet has shown that less dextrose is required to effect reduction 
 
QUANTITATIVE METHODS 93 
 
 the more slowly it is added, and that the concentration of the fluid is 
 of considerable importance. It is therefore necessary that the urine 
 should be diluted to contain as nearly as possible 1 per cent, of sugar, 
 and that the titration should be performed rapidly. Twenty c.c. of 
 Sachsse's fluid are reduced by 0*66 gram of dextrose in 1 per cent, 
 solution, and by 0*65 per cent, gram in 5 per cent, solution. 
 
 Oerum recommends that, for clinical work, the reduced mercury 
 should be collected on a filter, washed with warm 1 per cent, hydro- 
 chloric acid, then thoroughly washed with water, dissolved in boiling 
 nitric acid, and titrated with decinormal amixionimii thiocyanate, 
 using iron alum as an indicator, as in Volhard's method, the solution 
 having been previously standardised with a solution of dextrose, or 
 invert sugar, of known strength. 
 
 III. Titration with Safranine 
 
 Safranine and other dyes have been used for the quantitative 
 estimation of sugar. It is claimed that the reagents are simple 
 and do not deteriorate on keeping, and that the method is more 
 particularly applicable when only small quantities of sugar are 
 present, since the reduction due to other substances than the 
 sugars is only about one-fourth of that found with Fehling's, 
 Bang's, and similar solutions. As proteins do not interfere with 
 the reaction, their preliminary removal is not necessary. 
 
 To carry out the titration equal parts of a solution of safranine 
 (1 in 10,000) and potassium hydroxide (1 in 100) are mixed. To 2 c.c. 
 of this mixtLire, or more if the quantity of sugar in the urine is large, 
 are added a known number of drops of the urine, and the niixtvire is 
 warmed in a water-bath. The amount of sugar is estimated by ascer- 
 taining the UTunber of drops which are required to just decolorise the 
 alkaline mixture, which can be standardised against a sugar solution 
 of known strength. To prevent the oxidising effect of contact with the 
 air the surface of the alkaline mixtiire may be covered with a layer of 
 petroleum oil, or xylol, and the urine be introduced by a fine pointed, 
 narrow pipette, or biu-ette, dipping below the surface. (Hasselbalch and 
 Lindhard.) 
 
 IV. Quantitative Fermentation Tests 
 
 1. Differential Density Method (Robert's).— This method 
 is useful in clinical work, and, in inexperienced hands, is more to 
 be relied upon than titration with Fehling's solution, but the urine 
 must contain at least 0-5 per cent, of sugar. When carefully carried 
 out it is accurate up to about 0*1 per cent., although inattention 
 to detail may give rise to an error as high as 5 per cent, of the 
 total amount. It has the advantage that it is not necessary to 
 
94 GLYCOSURIA 
 
 remove any protein that may be present, and, moreover, the results 
 are not affected by the reducing substances of normal urine. It 
 is based upon the fact that fermentation reduces the high specific 
 gravity of sugar-containing urines. By comparing the specific 
 gravity before and after fermentation, and making use of an em- 
 pirical factor, which has been found by dividing the amount of 
 sugar ascertained by titration or polarisation by the cUfference in 
 density of urines before and after fermentation, the percentage of 
 sugar in a particular specimen may be determined. The emxmical 
 factor depends upon the particular method employed, each modi- 
 fication requiring a different coefficient. 
 
 The specific gravity of the urine is first carefully ascertained with a 
 pyknometer, or urinometer, graduated to the foiu-th decimal place, and 
 provided with a thermometer indicating tenths of a degree. A piece 
 of compressed yeast, the size of a hazel nut (washed free from sugar 
 if necessary), is added to a measured quantity, 100 to 200 c.c, of the 
 urine. The mixture is placed in a flask, gently shaken, and loosely 
 stoppered, or closed with a cork provided with a water, or mercury, 
 valve. After standing at the room temperature for twenty-four hours, 
 if but little sugar is present, or forty-eight hoiirs, if there is much, the 
 loss by evaporation is made up by the addition of distilled water, and 
 the clear fluid is povu*ed off from the yeast, which will have settled to 
 the bottom if fermentation is complete, or it may be filtered. It is first 
 tested with Fehling's, or Trommer's, test, to make sure that all the sugar 
 has been fermented off, then the specific gravity is again taken, with 
 the same precautions as before. The difference between the two 
 readings is multiplied by 0"230, the product giving the percentage of 
 sugar in the lu-ine. If the temperatures at which the two readings are 
 taken are found not to be the same, one-third of a degree of the urinometer 
 scale roust be added to, or subtracted from, the specific gravity deter- 
 mined after fermentation for each degree of temperature, adding for an 
 increase of temperature, and subtracting for a decrease. It is important 
 that the xxrine should be quite clear and free from yeast before the 
 specific gravity is taken, and if simple filtration does not suffice it may 
 be shaken up with magnesia usta, or keiselgur, and again filtered. The 
 process of fermentation may be hastened by adding to each 100 c.c. 
 of urine 2 grams of diacidsodiLxm phosphate, and 2 grams of sodium- 
 potassium tartrate, with 10 grams of yeast, and incubating at 30° to 
 34° C. If but little sugar is present the fermentation will then take 
 about two to three hours. The urine in which the specific gravity is 
 taken before fermentation should be treated in the same way, but 
 jDractically it is sufficient to add 0-022 to the observed specific gravity 
 to make up for the increased density from the added salts. In every 
 case the urine must be perfectly fresh, as fermentation generally 
 begins spontaneously, even after standing a short time. 
 
QUANTITATIVE METHODS 95 
 
 2. Volumetric Determination of Carbon Dioxide Evolved. 
 
 — The volume of carbon dioxide evolved by a saccharine fluid on 
 fermentation with yeast being proportional to the amount of fer- 
 mentable sugar present, it is possible to compute the sugar in a 
 diabetic urine by this means. Theoretically, the quantity of gas 
 given off should be read with all the precautions necessary in gas 
 analysis, taking into account the barometric pressure, the tem- 
 perature, tension of the water vapour, &c., but the introduction 
 of such physical details into the process is an attempt at refinement 
 not warranted by the basic inaccuracy of the method. 
 
 (a) Einhorn's Saccharifneter. — One gram of yeast is carefully 
 shaken with 10 c.c. of the urine, and the mixture poured into the 
 bulb and upright tube of the instrument, completely filling the latter. 
 No bubbles of air must be allowed to remain at the top of the tube, 
 nor should any collect within the first few minutes after the apparatus 
 is filled. It is then placed in an incubator at 30° to 34° C. for twelve 
 hours, when the reading is taken. The scale is graduated empirically 
 in percentages of sugar (dextrose). To prevent the escape of gas a 
 little mercury should be poured into the bent limb of the tube. It is 
 advisable to render the ixrine acid by the addition of a little tartaric 
 acid, and to boil it for several minutes to expel any contained air or 
 gas, subsequently cooling to the temperature of the room, before 
 ■carrying out the test. When the urine is rich in sugar it must be 
 .diluted four to ten, or more, times. The yeast employed must be fresh 
 .and active. Since the volume of gas evolved is dependent upon the 
 ;amount of yeast, its activity, the temperature, and many other factors, 
 and the so-called spontaneous fermentation of the yeast may give rise 
 to some gas, the accuracy of this method has been very seriously ques- 
 tioned, and it has been abandoned by many observers. There are others, 
 however, who maintain that most of the disturbing factors can be con- 
 trolled and that, with care, sufficiently reliable results for clinical work can 
 be obtained, particularly if a blank experiment, using normal urine and 
 the same yeast under similar conditions, is carried out at the same time. 
 
 (6) Lohenstein^ s Sacckarimeter is a modification of Einhorn's in- 
 :strument constructed on manometric principles, which is said to give 
 much more accurate results. It requires a minimum amount of urine, 
 and since the gas evolved is collected over mercury, no loss from 
 absorption occurs as in the latter instrimient. Its chief drawback is 
 that, the readings being taken on an empirical scale, the observer is 
 •dependent upon the maker for their accm-acy. The sugar estimation 
 is carried out as follows. A small piece of yeast is rubbed up with 
 two or three volumes of water, in a suitable vessel, to a thin paste. By 
 means of a syringe, provided with the instrument, 0-5 c.c. of the 
 previously boiled and cooled urine is introduced on to the surface of the 
 mercury contained in the bulb. From 0-1 to 0-2 c.c. of the yeast paste 
 .is then added, and the stopper fitted into the neck of the bulb, so 
 
96 GLYCOSURIA 
 
 that the hole in the side comes to he over the corresponding hole in the 
 neck of the bulb. The removable scale is now fitted to the tube, and 
 the surface of the mercury is adjusted to the zero mark by slightly 
 tilting the apparatus. It is fixed in this position by turning the stopper,. 
 so that the holes are no longer in communication. A weight is placed 
 on the stopper, and the apparatus set aside for fermentation to take 
 place, either at the temperature of the room (20° C), or in an incubator 
 at 32° to 38° C. In the latter case fermentation is complete, even with 
 a high percentage of sugar, in from tlxree to four hours. The apparatus 
 may also be placed in a water-bath at a temperature of 32° to 38° C.,. 
 the scale having previously been removed. When the column of 
 mercury ceases to rise, showing that fermentation is complete, the 
 apparatus is allowed to cool to the temperature of the room at which it 
 was filled, and the reading taken, using the appropriate scale for the 
 conditions under which fermentations took place. If greater accuracy 
 is desired the percentage of sugar can be calculated according to the 
 following formula — 
 
 p-T 
 
 P -{35 -t) 
 
 15 
 
 Where P=the reading on the scale at 35° C, p =the reading on the 
 20° scale, and t — the temperatiore previously noted at which the urine 
 was placed in the apparatus, and to which after completion of the- 
 fermentation it has again been cooled. As the instrument only gives 
 satisfactory resists if the urine contains 1-5 per cent., or less, of sugar- 
 it is necessary to dilute with water if more is present. 
 
 V. Estimation with the Polaeiscope 
 
 The property possessed by dextrose and other sugars of rotating- 
 the plane of polarised light proportionally to the amount present in. 
 solution is utilised to determine the percentage of sugar in a given 
 urine. The instrument employed for the purpose is known as a 
 polariscope, or polarimeter. Reduced to its essentials it consists 
 of a Nicol's prism, furnishing a beam of polarised light, and an 
 analyser of the same structure for examining it. A tube with clear- 
 glass ends containing the urine is placed between the polariser and 
 analyser in the path of the polarised beam. The length of the 
 tube is usually 100 mm., or multiples of this, and the reading is in 
 degrees. The angle of rotation must therefore be divided by 527 to 
 give the percentage of sugar, since the specific rotation of glucose is 
 Md= +52-74°, or generally for any sugar according to the formula — 
 
 _ 100 a 
 
 where j) = per cent., a = the observed rotation, [a] a = the specific 
 rotation of the sugar for sodium light, and I = the length of the 
 
QUANTITATIVE METHODS 97 
 
 tube in decimetreo. For clinical purposes, however, a modified 
 instrument with a tube of 188-6 mm., or better, 189-4 mm., long 
 is generally used, and the readings are taken directly in percentages 
 of dextrose. In some instruments the rotation caused by the 
 sugar is balanced by a compensating quartz wedge, marked with 
 an empirical scale, and ordinary white light, such as that given by 
 a Welsbach burner, is employed. 
 
 A cheap form of apparatus specially designed for the estimation 
 of sugar and albumen in urine is Mitscherlich's half-shadow polari- 
 scope, with Laurent's polariser. Behind the analyser is a small 
 telescope, and behind the polariser a semicircular plate of quartz 
 which covers one-half of the field of vision. The presence of the 
 quartz plate causes an alteration in phase of half a wave length, 
 and allows light to pass even when the polariser and analyser are 
 crossed. If the analyser is rotated so as to cause the quartz plate 
 to become dark, light passes through the uncovered half of the 
 field. In a position intermediate between these two the halves 
 of the field appear equally illuminated ; this is the zero point of 
 the instrument. The slightest deviation from this neutral position 
 causes one-half of the field to appear darker, and the other lighter 
 than before. Attached to the analyser is a pointer, which moves 
 to the right or left over a metal disc divided into angular degrees, by 
 which the readings are taken. By means of a vernier, ten divisions 
 of which correspond to nine divisions of the disc, readings to a 
 tenth, and estimations to a twentieth, of a degree may be made. 
 The instrument is constructed for homogeneous light, and a sodium 
 lamp must therefore be used as the source of illumination. Sodium 
 chloride in a Bunsen flame is generally employed, but a fused 
 mixture of sodium chloride and phosphate is better. With the 
 pointer and vernier at zero, and the telescope focused on the 
 quartz plate, the field of vision should appear as an evenly illu- 
 minate circle, divided into two halves by a vertical line. If a tube 
 containing a saccharine fluid is now put into position between 
 the polariser and analyser, and the apparatus is refocused, one- 
 half of the field will appear brighter than the other, but on rotating 
 the eye-piece until the optical disturbance produced by the intro- 
 duction of the rotating fluid is compensated, equal illumination 
 of the whole field can again be attained. The angle of rotation 
 necessary to effect this indicates the rotatory power of the sugar 
 in solution. Employing a tube 189-4 mm. in length, the per- 
 centage of dextrose present in the urine can at once be read olf , each 
 degree being equal to 1 gram per 100 c.c. For highly coloured 
 urine a tube 94-7 mm. is used. 
 
 G 
 
98 GLYCOSURIA 
 
 Before being exaroined with the polariscope the tirine must be 
 freed from proteins by acidifying 100 c.c. with a few drops of acetic 
 acid, boihng, cooling, filtering, and making up to 100 c.c. again, at the 
 same temperatvu-e as the original measurement was made at. The 
 removal of proteins is essential, for each gram of albumen rotates the 
 plane of polarised light equally, but in an opposite direction, to one gram 
 of dextrose. 
 
 If the urine is too highly coloured to permit of a satisfactory estima- 
 tion with either the long or the short tube it may be clarified, but it is 
 necessary to allow for any alteration in volume if this is done. Bone- 
 black has been recommended for the purpose, but as it has been shown 
 to remove a considerable amomit of sugar it should not be employed. 
 Kieselgulir is useful to remove turbidity, but it is not always satis- 
 factory, and may also remove some sugar. Keutral acetate of lead is 
 perhaps, on the whole, the most generally useful clearing reagent. It 
 may be used as crystals, or as a 10 per cent, solution, but an excess 
 alters the physical properties of the urine and removes some sugar 
 from solution. 
 
 Before the tube is filled with the virine it is carefully cleaned and 
 dried. The cover-glasses should be free from scratches, and also be 
 thoroughly clean and dry. Unnecessary warming by the heat of the 
 hand shovild be avoided. To fill the tube it is first closed at one end 
 with the cover-glass and screw-cap, then grasped at the other with the 
 thixmb and finger, and the urine poured in untU the meniscus projects 
 slightly above the opening. Air-bubbles are allowed time to rise, 
 and the second cover-glass pushed horizontally over the end of the 
 tube in such a way that the excess of liquid is carried over the side, 
 leaving the cover-glass exactly closing the tube, without air-bubbles 
 beneath it, and without any liquid on its upper surface. AVhen the 
 cover-glass is in position the tube is closed with the screw-cap, care being 
 taken that too great pressure is not exerted, for this ixdght induce 
 double refraction, and give rotatory power to the glass itself and thus 
 give rise to erroneous readings. The rubber washers must therefore 
 be placed in proper position and the caps be screwed on lightly. 
 
 Before taking a reading the zero point of the instrument must be 
 determined, as this changes somewhat with the temperature, &c. 
 The tube is then inserted, the field focused sharply, and the rotation 
 estimated. To find the end -point, with fields of equal ilhmiination, two 
 methods are commonly used. In one the analyser is rotated until a 
 black band seems to cross the division of the fields. This shadow, 
 which is a purely subjective j^henomenon, appears a little too soon, 
 but by taking an average of the readings made from both directions, 
 an accurate determination can be made. In the second method the 
 analyser is slowly turned, always in the same direction, the eye being 
 used for only a few seconds at a time, im.til the end-point seems to be 
 just reached — that is to say, until there seems to be no perceptible 
 difference between the two halves of the field. But since this point is 
 always reached a little too soon, the degree depending upon the 
 
QUANTITATIVE METHODS 99 
 
 sensitiveness of the instrimient, an average of several readings, taken 
 in both directions, must be made to obtain the exact figure. In all 
 cases it should be remembered that the eye must be used for only a 
 few seconds at a time, to prevent fatigue of the retina. The depth of 
 illumination of the whole field, and not of contiguotis portions, should 
 be judged. If when the tube is rotated on its long axis the fields change 
 in relative intensity it means either that the cover-glasses are not at 
 right angles to the long axis of the tube, or that they are subjected to 
 too much pressure from the screw-caps. If the whole of the field is 
 not equally sharp the solution is not homogeneous, or the tube is dirty. 
 Since normal Lirine is slightly levo-rotatory (0-05° to 0-18°) a trace 
 of dextrose may produce no noteworthy rotation. Indican, peptones, 
 cholesterin, and the administration of certain alkaloids cause the 
 iirine to be levo-rotatory, but indican is usually only present in small 
 amounts in diabetic urines, cholesterin occurs but rarely, and the 
 ■concurrence of glucose and peptones has not been observed. The 
 most important levo-rotatory substances are levulose and beta- 
 oxybutyric acid. Oxybutyric acid is practically alw^ays associated with 
 the presence of dextrose, and may be recognised by fermenting the 
 urine, filtering, and again examining with the polariscope, when it will 
 be found to be distinctly levo-rotatory. Levulose also generally occurs 
 in association with dextrose. They may be estimated together in the 
 following manner : — 
 
 Let X represent the dextrose, and y the levulose. If fermentation 
 of the urine shows 2 per cent, of sugar and polarisation 1-2 per cent., 
 then — 
 
 x + y = 2-0 
 
 x-y=l-2 
 
 .-. 2x = 3-2 
 
 and a; = 1 -6, and y = 0'4t. 
 If, after complete fermentation, the polariscope still shows 0'3 per 
 ■cent., owing to the presence of yS-oxybutyric acid, then — 
 x + y = 2-0 
 
 .- r-?/=l-2 + 0-3 =l-5 
 /.'IS = 3-5 
 and x = l*75, and y = 0.25. 
 
 Estimation of Levulose. — The reducing actions of levulose and 
 •dextrose on Fehlinr/s and other alkaline copper solutions are usually 
 assumed to be the same, although Soxhlet states that the reducing 
 power of the former is sensibly less than of the latter, 10 c.c. of 
 Fehling being reduced by 0-0543 gram of levulose as compared 
 with 0-05 gram of dextrose. On account of its ketonic nature 
 levulose reduces alkaline solutions of copper more rapidly, and at 
 .a lower temperature, than other sugars, and this j)roperty has been 
 utilised for its estimation. 
 
 Pieraerts obtained the best results with a cupro-glycocoll solu- 
 
100 GLYCOSURIA 
 
 tion, consisting of 6 grams of cupric hydroxide, 12 grams of glyco- 
 coll, and 50 grams of potassium carbonate, dissolved in water and 
 made up to 1000 c.c. This solution is reduced by levulose at 
 normal temperature in twelve hours, but is totally unaffected by 
 other sugars in twenty-four hours. To carry out the determination, 
 the urine is treated with basic lead acetate, filtered, and the excess 
 of lead removed by adding a saturated solution of sodium sulphate. 
 After half an hour the mixture is again filtered, diluted if necessary 
 so as to contain 5 per cent, of reducing sugar, and mixed with the 
 reagent. 
 
 The reducing action of levulose on Knapp's solution is about 
 the same as that of dextrose, 1 gram of levulose in 1 per cent, 
 solution reducing 508-5 c.c. of the reagent ; but Sachsse's solution 
 is more markedly reduced by levulose than by dextrose, 1 gram 
 of levulose in 1 per cent, solution reducing 449-5 c.c. 
 
 When a urine contains only levulose it is most readily esti- 
 mated by means of the polariscope, the levo-rotation due to the 
 sugar being distinguished from that arising from other causes 
 by the former disappearing after complete fermentation. It is 
 to be noted, however, that the rotatory power of levulose is markedly 
 influenced by the concentration, and temperature, of the solution, 
 and particularly by the latter. At 15° C. [a] has a value of - 93-8°. 
 This value decreases 0-6385° for each rise of 1° C. in the tempera- 
 ture, so that at 87-2° C. it is -52-7°, which is equal, but opposite 
 to that of dextrose at the same temperature. This change in 
 optical activity affords a means of estimating levulose in the 
 presence of other sugars. The solution is examined in a jacketed 
 polarimeter tube provided with a thermometer, and the rotation 
 noted at two temperatures as far apart as possible. Provided that 
 the solution is fairly strong, the difference between the two tem- 
 peratures multiplied by 1-277, and the product divided into 100 
 times the alteration measured in rotation in circular degrees, in a 
 2 decimetre tube by sodium light, gives the number of grams of 
 levulose in 100 c.c. As the rotatory power of other sugars is also 
 affected somewhat by change of temperature, the results are not 
 strictly accurate. 
 
 For clinical purposes in urinary analysis it is generally assumed 
 that the reducing powers of levulose and dextrose are identical, 
 and when the two are present together in a urine, as is generally 
 the case, the rotation is usually measured in terms of dextrose. 
 The percentages of each sugar present is then estimated by com- 
 paring the readings obtained with the polariscope and by titra- 
 tion, or fermentation, as in the example already given (p. 99). 
 
QUANTITATIVE METHODS 101 
 
 Estimation of Maltose. — The reducing power of maltose for 
 hot Fehlincfs solution is only about 62 per cent, of that shown by 
 dextrose, 10 c.c. of Fehling being reduced by 0-0807 grams of maltose, 
 as compared with 0-05 grams of dextrose. The amount of re- 
 duction that occurs varies with the concentration of the solutions. 
 Soxhlet states that the cupric reducing power is 61 when the maltose 
 is present in 1 per cent, solution, the Fehling solution is undiluted, 
 and is employed in the exact proportion necessarj^ — 64-1 when the 
 copper solution is previously diluted with 4 volumes of water, 
 and 65-3 per cent, when twice as much of the diluted Fehling 
 solution is used as is required for the reaction. If a solution of 
 maltose is treated with a volume of Fehling' s solution sufficient 
 for its oxidation, the mixture heated and the cuprous oxide filtered 
 off, a solution is obtained which, if acidulated with hydrochloric 
 acid and heated, acquires the property of reducing an additional 
 quantity of Fehling solution. The second reduction is somewhat 
 more than half the first, so that the two together approach the 
 reducing power of dextrose. 
 
 The reducing action of maltose on Pavy's solution is the same 
 as upon the ordinary Fehling's solution, but the addition of more 
 sodium hydroxide, in the presence of ammonia, increases the oxi- 
 dising power of the copper solution to a notable extent. 
 
 One gram of maltose in 1 per cent, solution corresponds to 317'5 
 c.c. of Knapp's, and 197-6 c.c. of Sachsse's, solution. 
 
 Maltose has a value for [a]^= + 138°. 
 
 Maltose is not directly fermented by ordinarj^ brewer's yeast, 
 but is first hydrolised to dextrose by the enzyme maltase present. 
 Some species of yeast — ^S*. marxianus, S. exiguus, S. Ludwigii, 
 W . anomala, W. Saturnus — do not contain " maltases," and are 
 therefore incapable of fermenting maltose, so that by means of them 
 dextrose and levulose may be detected, and estimated, in the presence 
 of maltose, and by comparing the results obtained by titration and 
 fermentation the percentage of maltose present may be estimated. 
 
 Estimation of Lactose. — The reducing power of milk-sugar 
 for Fehling's solution is considerably less than that of dextrose, 
 10 c.c. of Fehling requiring 0-0678 gram of lactose to effect complete 
 reduction, as compared with 0-05 gram of dextrose. 
 
 If the reduced copjoer oxide is filtered off, the filtrate acidulated 
 with hydrochloric acid and heated, hydrolysis of the disaccharide 
 occurs, and a further quantity of Fehling's solution will then be 
 reduced on heating, as in the case of maltose, the second reduction 
 being about half the first. 
 
102 GLYCOSURIA 
 
 Lactose is dextro-rotatory, the value of [a]^ = +52-7°. 
 
 Estimation of Pentoses. — Pentoses occurring in the urine- 
 cannot be accurately estimated by titration with Fehlincfs solution, 
 as they do not give a sharp end-point. Bial has made use of 
 Knapp's solution, and Salkowski estimated the amount of pentose 
 in his case from the quantity of pentosazone obtained from a 
 measured volume of the urine. The following method for deter- 
 mining pentoses in urine has been suggested by Jolles as being 
 more accurate. 
 
 1. Jolles^ Method.- — The furfurol formed by distilling the jsentose 
 containing urine with hydrochloric acid is estimated by combining 
 it with sodium hj^drogen sulphite. 
 
 The urine is acidified with a few drops of acetic acid, boiled, and 
 concentrated if necessary, to free it from interfering volatile substances.. 
 One hundred cubic centimetres are then mixed with 150 c.c. of hydro- 
 chloric acid (sp. gr. 1-06, i.e. about 12 per cent. HCl), placed in a flask, 
 and submitted to steam distillation, taking care that the contents of 
 the flask do not fall below about 100 c.c. When the distillate amounts 
 to about 1000 c.c, and a sample no longer gives Bial's orcin reaction, 
 100 c.c. are neutralised with 20 per cent, sodium hydroxide, using^ 
 methyl orange as the indicator. A drop or two of half normal hydro- 
 chloric acid is then added, so that the red coloration is just restored. 
 A measured vokmie of standard sodium hydrogen sulphite, sufficient 
 to more than combine with all the furfurol present, is then added and 
 the mixture allowed to stand for two hours. The excess of sulphito 
 is now titrated with Yo^h normal iodine solution, using starch paste as 
 the indicator. Each cubic centimetre of normal sodium hydrogen 
 sulphite lost corresponds to 0-07505 gram of pentose. 
 
 2. Phloroglucin Method. — The furfurol formed on distilhng a 
 pentose-containing urine with 12 -pev cent, hj^drochloric acid by 
 direct heat may be estimated as a phloroglucinol compound. 
 
 A quantity of p\are phloroglucinol (see Allen's Commercial Analysis, 
 vol. i., 401, 1909), about double the amount of furfurol expected, is 
 dissolved in 12 per cent, hydrochloric acid, and added to the distillate. 
 The mixture is well stirred, and left to stand overnight. The amorphous 
 black precipitate that forms is filtered off tlirough a weighed gooch 
 crucible, provided with an asbestos felt, and carefully washed with 
 water, then dried at 100° C. for four hovirs, cooled, and weighed. To' 
 calculate the furfuroldehyde or pentose present the following formulae 
 (Krober) are used, where " a " represents the weights of phloroglucin. 
 foimd. 
 
 For weights of phloroglucide under 0-03 grams — 
 rurfiarol = (a + 0-0052)x 0-5190 
 Pentose =(a-f-0-0052) x 1-0170 
 
QUANTITATIVE METHODS 103 
 
 For weights of phloroglucide from 0'03 grams to 0"3 grams. 
 Fvu-furol = (a + 0-0052) x 0-5 185 
 Pentose =(a + 0-0052) x 1-0075 
 
 For weights of phloroglucide over 0-300 grams — 
 FLirftirol = (a + 0-0052) x 0-51 80 
 Pentose =(a + 0-0052) x 1-0026. 
 
 Griind gives the following formulse for arabinose and xylose — 
 Arabinose = phloroglucide x 1-148 + 0-0025 
 Xylose =phloroglucidex 1-045 + 0-00305 
 
 Estimation of Glucuronic Acid. — Like the pentoses, glucu- 
 ronic acid cannot be satisfactorily estimated by the usual titration 
 methods, but it yields furfurol on being distilled with hydrochloric 
 acid, and by combining this with phloroglucin the quantity may 
 be estimated. One part of furfurol-phloroglucide corresponds to 
 three parts of glucuronic anhydride. Other furfurol-jaelding sub- 
 stances, and particularly the pentoses, must of course be first 
 excluded. 
 
 On being heated with hydrochloric acid, glucuronic acid yields 
 carbon dioxide in addition to furfurol. Lefevre and Tollens have 
 suggested that by weighing the carbon dioxide evolved when its 
 solution is boiled for three or four hours with hydrochloric acid 
 (sp. gr, 106), and allowing one part of carbon dioxide for four 
 parts of glucuronic anhydride, the amount of glucuronic acid pre- 
 sent in a solution may be determined.^ The value of this method 
 as applied directly to urine has not yet been proved, but it is 
 probable that the result will be too high, owing to the presence of 
 other substances yielding carbon dioxide. 
 
 Since the pentoses do not yield carbon dioxide on being heated 
 with hydrochloric acid, it is possible to estimate the glucuronic 
 acid and the pentose in a sample, when they are present together, 
 the former by the weight of carbon dioxide evolved, and the latter 
 from the quantity of furfurol formed. 
 
 Neuberg has employed the para-brom-phenylhydrazin compound 
 of glucuronic acid for its estimation. 
 
 Estimation of Sug^ar in the Blood. — A weighed or mea- 
 sured quantity of blood is freed from albumen by boiling with 
 sodium sulphate, and filtering. The precipitate is well washed, 
 and the quantity of sugar in the filtrate is estimated. A certain 
 amount of sugar is carried down by the precipitate, and unless the 
 blood is quite fresh the action of the glycolytic ferment may cause 
 serious error. 
 
 ^ See Abderhalden's Handbuch dc Biochcm. Arhcitsmcth. (1909), ii. p. 139. 
 
104 GLYCOSURIA 
 
 For clinical purposes the sugar in the peripheral blood stream 
 may be estimated bj^ collecting 5 c.c. and at once mixing it with 
 100 c.c. of absolute alcohol, filtering off the precipitate, washing 
 Avith hot water several times, evaporating the combined filtrates 
 on a water-bath to about 0-5 c.c, and estimating the sugar in this. 
 
 The estimation may be carried out by titration with alkaline 
 safranin, or by Lohenstein's saccharimeter. If a large amount of 
 blood is available, titration with Fehling's solution, &c., or polari- 
 metry, may be employed. 
 
 Wacker has described a colorimetric method by which it is 
 claimed that sugar can be accurately estimated in such small 
 quantities of fluid as 0-3 to 0-4 c.c, with an error of but 0-01 per 
 cent. The j)rotein of the blood is removed with iron alum and 
 sodium carbonate. To the filtrate phenylhydrazine-p-sulphonic 
 acid and sodium hydroxide are added, and the red colour is com- 
 pared with a colour scale made with the same reagent and a stan- 
 dard solution of dextrose. The figures given for normal blood by 
 this method are higher than those found by previous observers. 
 
 (b) Acetone Bodies 
 
 The presence of the secondary disturbances of metabolism that 
 are liable to occur in cases of persistent dextrosuria are shown by 
 the appearance of the acetone bodies (acetone, aceto-acetic acid, 
 and beta-oxybutyric acid) in the urine, and their extent is indi- 
 cated by the amount of these substances excreted daily. It is 
 therefore most important that their presence should be recognised 
 and that the daily output should be estimated as accurately as 
 possible, both from the j)oint of view of prognosis and treatment. 
 
 Chemically the acetone bodies are, as we have seen, intimately 
 related, the oxidation of beta-oxybutyric acid giving rise to aceto- 
 acetic acid and water, and acetone being easily decomposed into 
 acetone and carbon dioxide. It is believed that this series of 
 changes normally takes place in the body. Beta-oxybutyric acid 
 is not met ^^ith in health urines, although Gerhardt produced an 
 excretion of about 9 grams in the twenty-four hours by feeding 
 a normal individual on a purely protein diet for some days, and its 
 presence may be regarded as distinct evidence of a perversion of 
 metabolism. Since it is the mother substance of acetone and aceto- 
 acetic acid, the latter are always present in the urine when beta- 
 oxybutyric acid is being excreted. Only a trace, if any, aceto- 
 acetic acid occurs in normal urines, and probably none on a mixed 
 diet. If it is found, it may be taken for granted that acetone will 
 also be there, and generally, but not necessarily, oxybutyric acid 
 
QUANTITATIVE METHODS 105 
 
 also. Acetone, unlike the two other acetone bodies, appear to be 
 a constant constituent of the urine, although in very small amounts, 
 0-01 to 0-03 gram in the twenty-four hours, a quantity which is not 
 revealed by the ordinary clinical tests. The excretion of acetone 
 bodies is usually most marked when the output of sugar is high, 
 but there is no constant relation between the two. 
 
 (A) These substances may be recognised by the following 
 qualitative tests : — 
 
 Acetone. — The urine must be as fresh as possible. If only a 
 small amount of acetone is present it may be ad\'isable to distil, 
 and apply the tests to the distillate, but this is rarely necessary if 
 Lange's or Rothera's modifications of the nitro-prusside test is 
 employed. If these are negative, it may be concluded any acetone 
 that may be present is not of pathological significance. 
 
 1. Nitro-Prusside Test. — A fresh solution of sodium nitro- 
 prusside when added to an alkaline solution of acetone gives a 
 ruby-red colour, which rapidly changes to yellow. If, while the 
 solution is still red, an excess of acetic acid is added, the colour 
 changes to a purple, and later to \dolet. Creatinin gives with 
 nitro-prusside and an alkali a similar colour reaction, but it more 
 quickly changes to yellow, an alteration ^^•hich is instantly brought 
 about by the addition of acetic acid, so that the reaction due to 
 acetone can be distinguished from that caused by creatinin by this 
 means. Paracresol gives a reddish-yellow solution, acetic acid a 
 clear rose colour. Aceto-acetic acid gives the same reaction as 
 acetone. 
 
 On these facts are based the following tests : — 
 
 LegaVs. — About 5 c.c. of the urine (or the distillate) are mixed with 
 5 drops of a freshly prepared, concentrated (10 per cent.), solution of 
 sodium nitro-prusside, and about 1 c.c. of a solution of sodimn, or 
 potassium, hydrate (15 per cent.) is added. An excess of glacial 
 acetic acid is then introduced. In the presence of much acetone a 
 purple red colour appears with a large excess of acetone a deep 
 violet, and with traces a rose-violet. According to Bohrisch this 
 test, when directly applied to the urine gives a satisfactory reaction 
 with O'l per cent, of acetone. 
 
 Le Nobel. — To exclude aldehyde, Le Xobel and Lee advise the use 
 ■of ammonia, or ammonimn carbonate, in place of sodiiun or potassiiun 
 hydrate, but the colour change does not take place so rapidly unless a 
 drop of acetic acid is added. 
 
 Lange. — As in the preceding method, ammonia is the alkali used, 
 but the delicacy of the reaction is increased by carrying it out as a 
 " ring-test." To 15 c.c. of the urine, contained in a test-tube, are added 
 0*5 to 1*0 c.c. of glacial acetic acid, and a few drops of a freshly prepared 
 
106 GLYCOSURIA 
 
 concentrated solution of sodium nitro-prusside. These are mixed, 
 and then a few cubic centimetres of strong ammonia are run on to the 
 top. In the presence of acetone a violet ring forms at the junction, 
 its depth and intensity varjang with the amount of acetone. This 
 test is said to show 0-0025 per cent, of acetone. 
 
 Rothera has suggested the follo-ndng modification of nitro-prusside 
 test : 5 to 10 c.c. of the tu-ine are satiu-ated with solid ammonium 
 sulphate (or according to Garrod, mixed with an equal volimie of a 
 saturated solution) by vigorous shaking. Two or tlxree drops of a 
 fresh 5 per cent, solution of sodium nitro-prusside are then introduced, 
 and 1 to 2 c.c. of concentrated aixmionia added. The presence of 
 acetone is shown by the formation of a very characteristic " perman- 
 ganate " colour, which is well seen in the foam on the top of the mixture. 
 
 2. Salicylaldehyde Test {Frommer). — This irj based upon the 
 fact that acetone reacts with salicylaldehyde to form dioxj^diben- 
 zoylacetone. The test may be applied directly to the urine, and 
 gives no reaction with aceto-acetic acid if the temperature is not 
 raised too high. It is said to indicate the presence of 0-000001 
 gram of acetone in 8 c.c. of water, and is not affected by the 
 presence of protein. 
 
 Ten c.c. of the urine are rendered strongly alkaline ^^•ith potassium 
 hydroxide, and 10 to 12 drops of a 10 per cent, solution of salicylalde- 
 hyde in absolute alcohol are added. The mixture is then warmed to 
 70° C. In the presence of acetone the fluid becomes yellow, then red, 
 and later purplish-red, tiu-ning to dark red on iDrolonged standing. If 
 acetone is absent the colour of the urine is practically unchanged. 
 It may also be carried out as a ring-test, adding 1 gram of solid sodimn 
 hydroxide to about 10 c.c. of the urine, and, without waiting for it to 
 dissolve, running in 10 or 12 drops of the salicylaldehyde solution and 
 warming to 70° C. 
 
 Other methods of detecting acetone in the urine have been 
 described by Reynolds (fresh mercuric oxide), Penzoldt (indigO' 
 test), Frohner (hydro xy lamine) , Voumasos (isonitrile test), and 
 others, but it is not necessary to describe them in detail here, 
 and the reader is referred to the original papers, to which reference 
 is made in the bibliography. 
 
 3. Iodoform Test. — Like other substances containing an easily 
 oxidised - CHg group, acetone forms with iodine, in alkaline solu- 
 tions, iodoform, and by this its j)resence can be detected. The 
 test cannot be applied directly to the urine, but only to the 
 distillate. 
 
 Distillation. — 200 to 250 c.c. of the urine are acidified with acetic, 
 or better with phosphoric acid, and distilled with an efficient, well- 
 cooled condenser. The latter precaution is especially necessary for 
 
QUANTITATIVE METHODS 107 
 
 quantitative work. If the urine is distilled with steam, as ^^ Jaksch 
 advises, it is not necessary to acidify it. Most of the acetone passes- 
 over in the first 10 to 30 c.c, which are used for the tests. If it is 
 desired to exclude aceto-acetic acid, which is seldom necessary, the 
 urine should be previously made alkaline and thoroughly extracted with 
 alcohol free ether. The ether extract may be shaken out with water, 
 and the latter used for the aceto-acetic acid tests. Instead of being 
 distilled the acid virine may, according to Bolxrisch, be extracted with 
 ether, 20 c.c. for each 30 to 50 c.c. of the urine, the ether extract be 
 shaken out with water, the residual ether be driven off from the watery 
 extract by warming at 40° C. for about five minutes, and the residue, 
 which contains the greater jDart of the acetone, can be used for the 
 tests. 
 
 Lieheris Iodoform Beaction. — One c.c. of the distillate is mixed with 
 a drop of strong sodimn hydrate solution, and sufficient iodo-potassiimi 
 iodide solution (iodine 1 : pot. iod. 2 : water 100) to give a feeble 
 yellow colour. If no tvu-bidity appears in a few minutes the test-tube 
 is placed in a water-bath at 70° C. for a few minutes, and then allowed 
 to cool. With only O'Ol mg. of acetone a precipitate of iodoform, 
 recognised by its smell and appearance microscopically as yellow six- 
 sided plates or stars, appears in 1 to 3 minutes, and with 0-0001 ixig. 
 in 24 hours. The iodoform crystals must be carefully distingxiished 
 from the stellate triple phosphate crystals that are also often found. 
 If necessary, the precipitate may be recrystallised from ether. Urines 
 preserved with thymol give a pink, or red, coloiir due to the formation 
 of di-iodothjanol, but no other substance than acetone occurring in 
 the urine yields iodoform. 
 
 Gunning's Modification. — To 5 c.c. of the distillate add a few drops 
 (5 to 10) of 10 per cent, ammonia, and about o drops of alcoholic iodine 
 solution (tinct. iodi). A deep black precipitate of nitrogen iodide 
 forms. On standing this gradually disappears and leaves a yellow 
 deposit of iodoform, if acetone is present. 
 
 The preceding tests that can be applied directly to the urine may 
 also be used for the detection of acetone in the distillate, also Bela's 
 dinitro -benzol test, Rosenthaler's vanillin test, and Reynolds' mercuric 
 oxide test, which only shows over 1 per cent, in the urine direct. 
 
 Aceto-acetic (Di-acetic) Acid.— The urine used for testing 
 for this substance must be fresh, for it quickly breaks down, and 
 may not be detected after standing for twenty-four to forty-eight 
 hours. 
 
 1. Aceto-acetic acid, and its salts, give with ferric chloride a 
 violet-red colour, which is red-brown with an excess of the reagent. 
 The colour fades on standing for twenty-four hours in the cold, 
 and more rapidly on warming, or adding a mineral acid. 
 
 Gerhardfs Ferric Chloride Reactio7i. — To 10 or 20 c.c. of the urine add 
 a solution of ferric chloride, which must not be too acid, drop by drop. 
 
108 GLYCOSURIA 
 
 so long as any precipitate forms, and then filter. To the filtrate is 
 now added a few more drops of ferric chloride, when, if aceto -acetic 
 acid is present, the urine will assume a Bordeaux -red colour, which is 
 ■cherry-red by transmitted, and purple-red by reflected light. On 
 heating or standing the intensity of the colour diminishes. 
 
 A similar colour reaction is also given by other substances that may 
 occur in the urine, such as cyanates, sodium acetate, salicyclic acid, 
 phenol, skatoxyl compounds, and substances excreted after the ad- 
 ministration of aspirin, diuretin, salol, antipyrin, thallin, chinisol, &c. 
 In the case of antipyrin, thallin, and chinisol urines the colour does 
 not appear until after the lapse of two or three minutes. With 
 salicyluric acid the colour is violet, with antipyrin and thallin piurple- 
 red, turning in the latter to red-brown in three or four hours. The 
 colour with all these substances, in contrast to that given by aceto - 
 acetic acid, is stable, and on heating only the colours due to aceto - 
 acetic acid, ferric acetate, and thallin are destroyed. If the test is 
 repeated with a specimen of the urine which has been feebly acidified 
 with sulphuric acid, boiled, and cooled, the red colour should be less 
 marked, as the aceto-acetic acid will have been partly broken down 
 into acetone, but boiling for half an hour will not decompose it entirely. 
 
 The aceto-acetic acid may be separated and tested for by acidifying 
 the urine with sulphuric acid, extracting with ether, shaking the ether 
 extract with water, and adding ferric chloride, when a violet-red colour 
 will appear in the watery layer. Salicyluric acid may be removed by 
 shaking the ixrine with benzol, or chloroform, in which aceto-acetic 
 acid is insoluble. 
 
 2. Arnold- Lipliaivshi Test. — This test is said to be more delicate 
 than Gerhardt's, and does not give a confusing reaction with other 
 substances. 
 
 Two solutions are required — (1) para-amidoacteophenone 1 gram, 
 concentrated hydrochloric acid 2 c.c, distilled water 100 c.c. ; (2) 
 potassium, or sodium, nitrite 1 gram, distilled water 100 c.c. Six c.c. 
 of the first solution, and three of the second, are mixed together. An 
 equal volume of the i-U-ine is then added, a drop of ammonia introduced, 
 and the mixture shaken. A bright red colour results. From a 
 J to 2 c.c. of the liquid, the amount depending upon the quantity of 
 aceto-acetic acid present, is then mixed with 10 to 20 times its bulk 
 of concentrated hydrochloric acid (1*19) to which 2 to 4 drops of ferric 
 chloride have been added. From 2 to 3 c.c. of chloroform are then 
 added, and mixed, withovit shaking. In the presence of aceto-acetic 
 acid the chloroform is coloured violet in | to 1 minute. Otherwise 
 it only turns yellow. 
 
 3. Morner's Test. — When a urine containing aceto-acetic acid 
 is heated with a httle iodide of potassium, and an excess of ferric 
 chloride, it gives rise to the smell of acetone iodide. The reaction 
 
QUANTITATIVE METHODS 109 
 
 is not specific, however, as urines containing much acetone, but no 
 aceto-acetic acid, also give it (v. Jaksch). 
 
 Bondi and Schwartz" s Modification. — To 5 c.c. of the Lirine is added 
 a solution of iodo-potassiixm iodine, drop by drop, \antil the flmd is 
 orange-red. It is then decolorised by being gently warmed. A 
 further addition of the iodine solution is made i.intil the urine remains 
 a faint red, even on warming. If the mixture be now gently boiled, 
 the piercing smell of acetone iodide will be noticed if aceto-acetic acid 
 is present. This reaction, while more delicate than Gerhardt's, is 
 said not to be given by acetone, or oxybutyric acid. It is only ob- 
 tained with neutral, or feebly acid, urines, so that alkaline urines must 
 be previously made neutral with acetic acid. 
 
 Riegler's test has been much criticised and its negative nature is a 
 distinct drawback. 
 
 One to two c.c. of normal vu-ine are mixed with. 2 c.c. of 10 per cent, 
 solution of iodic acid, and 3 c.c. of chloroform, and shaken. The 
 chloroform is thereby coloured violet. If 10 c.c. of the suspected lu-ine 
 are now added, and the mixtiu-e shaken, the presence of aceto-acetic 
 acid is shown by the chloroform being decolorised, owing to the acid 
 combining with the free iodine. 
 
 Beta-oxybutyric Acid.— This is the mother substance of aceto- 
 acetic acicl, and consequently need be only looked for when the 
 urine gives the ferric chloride, or a similar reaction, for aceto- 
 acetic acid. It may not, however, be found even when aceto- 
 acetic acid is present. If some simple test for oxybutyric acid 
 could be devised, it would be of great value in the investigation of 
 cases of glycosuria, but, unfortunately, such does not exist, and 
 the presence of beta-oxybutyric acid can only be demonstrated 
 by investigations too complicated for routine clinical work, or 
 be inferred from indirect evidence. 
 
 1, Beta-oxybutyric acid may be suspected if the amount of 
 sugar in a urine, as determined by titration or fermentation, is 
 distinctly larger than with the polariscope, since the acid is levo- 
 rotatory, [aj^, = -24-12°. Other levo-rotatory bodies, such as 
 albumen, paired glucuronic acid, and levulose, may, however, 
 bring about a similar result, and must first be excluded. 
 
 2. It is probably present if the fermented urine of a diabetic 
 is distinctly levo-rotatory after other levo-rotatory substances have 
 been removed by treating it with basic lead acetate and ammonia, 
 and filtering off the precipitate. The levo-rotation may still, how- 
 ever, be due to compound glucuronates, but these may be excluded 
 by strongly acidulating the fermented urine with phosphoric acid, 
 extracting with ether, and examining the ethereal extract with 
 
110 GLYCOSURIA 
 
 the polariscope. If this is also levo-rotatory, the presence of 
 oxybutyric acid may be assumed with some confidence. 
 
 3. Conversion into Crotonic Acid. — On being heated with 
 sulphuric acid, oxybutyric acid is converted into crotonic acid, 
 which separates out as white crystals with a melting-point of 
 71° to 72° C. 
 
 The urine is ixiixed with an equal quantity of strong sulphuric acid, 
 placed in a flask, provided with a drop -funnel, and distilled. From 
 time to time additions of water are made through the funnel to keep 
 the mixture at a constant bulk. The first part of the distillate is 
 collected in a test-tube and well cooled, when, if considerable amount 
 of crotonic acid is present, it will crystallise out. The crystals are 
 separated, dried between blotting-paper, and their melting-point taken. 
 This shoiild lie between 70° to 72° C. If no crystals appear the dis- 
 tillate is shaken with ether, the ether extract separated, and allowed 
 to evaporate spontaneously. Any crystals that appear are washed 
 with water, dried, and examined for their melting-point. The crystals 
 may be further purified by redissolving them in ether, evaporating, 
 and precipitating the residual fluid with petroleum ether, which 
 separates the traces of fatty acids that also distil over. (Embden and 
 Schmitz.) 
 
 4. Conversion into Aceto-acetic Acid. — On oxidising beta- 
 oxybutyric acid with hydrogen peroxide and ferrous sulphate, it 
 is converted into aceto-acetic acid, which can be recognised by 
 the ferric chloride reaction. 
 
 Black's Method. — 5 to 20 c.c. of the m-ine are evaporated, by gently 
 heating on a water-bath, to ^ to J of their volume, thereby expelling 
 the preformed aceto-acetic acid. The residue is acidified with a few 
 .drops of concentrated hydrochloric acid, and mixed with recently 
 burned gypsmn to a thick paste until it begins to solidify. The mixtiore 
 is then thoroughly ground and extracted twice with ether, stirring 
 well. The ether is evaporated off, the residue dissolved in water, and 
 neutrahsed with barimn carbonate. Two or three drops of hydrogen 
 peroxide are added to the neutral fluid, and then a few drops of 5 per 
 cent, ferric chloride, that contains a trace of ferrous salt. If the urine 
 contained oxybutyric acid a rose -red colour appears in a few seconds. 
 On standing the coloiu- gradually intensifies for a time, but subse- 
 quently slowly fades. The test solution must be cold and neutral in 
 reaction. An excess of hydrogen peroxide and iron must be avoided. 
 According to Black, this test shows 1 : 10000 of oxybutyric acid. 
 
 Embden and Schmitz have modified the preceding method by 
 •directly testing the acid ether extract of the virine, that has been freed 
 from aceto-acetic acid by heat. To the ethereal extract they add a 
 few drops of hydrogen peroxide, and then cautiously, drop by drop, 
 a 5 per cent, solution of ferrous sulphate. If beta-oxybutyric acid is 
 
QUANTITATIVE METHODS 111 
 
 present the typical \'iolet colour of the ferric chloride reaction for 
 aceto- acetic acid is produced. 
 
 5. Conversion in Acetone. — By more prolonged heating the 
 aceto-acetic acid may be converted into acetone, which can be 
 tested for by the nitro-prusside test. 
 
 Harfs Method. — Twenty c.c. of the urine are diluted with an equal 
 amount of water, acidified with a few drops of acetic acid, and evaporated 
 down to 10 c.c. to drive off the acetone and aceto-acetic acid originally 
 present. The residue is made up to 20 c.c. again. One c.c. of hydrogen 
 peroxide is added, and the mixture warmed, but not boiled, and then 
 cooled. It is placed in a test-tube, 1 c.c. of glacial acetic acid, and a 
 few drops of a freshly prepared concentrated solution of sodium nitro- 
 prusside added, and the mixtvire overlaid with strong ammonia. If 
 the urine contained oxybutyric acid a pLU"ple-red ring will appear at 
 the junction in 4 to 5 hours, and on shaking the whole of the fluid will 
 become coloured. This test is said to show 0*3 per cent, of beta- 
 oxybutyric acid. 
 
 (B) The acetone bodies in the urine may be estimated separately 
 or together by direct methods, but, as the processes involved are 
 lengthy and laborious, it is more usual for clinical purposes to infer 
 the extent of the metabolic disturbance which their presence 
 indicates by an indirect method — namely, by a determination of 
 the amount of nitrogen excreted in the twenty-four hours in the 
 form of ammonia. 
 
 Ammonia nitrogen. — A small fraction, about one-twentieth as 
 a rule, of the total nitrogen eliminated in the urine is in the form 
 of ammonium compounds, and is usually referred to analytically 
 as ammonia nitrogen. The normal amount of ammonia nitrogen 
 contained in adult urine varies between 0-3 and 1-2 grams in the 
 twenty-four hours, with an average of 0-7 grams (Neubauer). In 
 persistent glycosuria the excretion is often very much increased, 
 occasionally reaching as much as 8 grams in the twenty-four hours, 
 the excess being accounted for by the ammonia combined with 
 aceto-acetic and oxybutyric acids. The amount of ammonia 
 nitrogen eliminated may therefore be used as an index of the 
 quantity of these two substances that is present. Although suffi- 
 ciently exact for routine clinical purposes, estimation of the am- 
 monia nitrogen only approximately measures the pathological acids, 
 for, as we shall see when we come to consider diabetic coma and 
 " acidosis," ammonia is only used to neutralise these acids when 
 fixed alkalies are no longer available. According to Magnus-Levy, 
 about 1 to 2 grams of the ammonia excreted by a diabetic are 
 used to neutralise the fixed excess of acid resulting from the protein 
 
112 GLYCOSURIA 
 
 diet, and it is only the remainder that is the measure of the patho- 
 logical excretion of acid. A diabetic with 4 grams of ammonia 
 in his urine is consequentlj^ not excreting twice, but nearly three 
 times, as much pathological acid as one with 2 grams, since 1 gram 
 or more of the 2, or 4, respective^ is combined mth the acids nor- 
 mally excreted. According to this view, when 2 grams of ammonia 
 are excreted, about 1 gram is combined with oxybutyric and 
 aceto-acetic acids. An output of 5 grams leaves 3 to 4 grams for 
 this purpose, and an excretion of 8 grams gives from 6 to 7 grams 
 of ammonia as the measure of the pathological acids. These 
 figures correspond to 5 grams, 20 grams, and 36 to 40 grams of 
 beta-oxybutyric acid respectively. Magnus-Levy points out that 
 an estimation of the pathological acids from the ammonia excretion 
 may vary considerably as the result of a temporary retention, or 
 excretion, of fixed alkalies, and that an isolated observation may 
 lead to erroneous conclusions. The estimations must therefore 
 be made with the total twenty-four hours' urine, and on several 
 successive days. He considers that 6 to 7 grams of ammonia, 
 corresponding to 36 to 40 grams of beta-oxybutyric acid, is the 
 maximum amount the organism can provide for the neutrahsation 
 of pathological acids, although much larger quantities of acid are 
 sometimes excreted. 
 
 Estimation of Ammonia Nitrogen. — The urine must be as 
 fresh as possible to prevent the inclusion in the estimation of 
 ammonia formed by the bacterial decomposition of urea. L^rines 
 which are alkaline from changes undergone in the bladder, or on 
 standing, cannot be used. 
 
 1. Malfatti-Jager. — This method depends upon the fact that 
 ammonium salts react ^^'ith formaldehyde, in neutral solutions, to 
 form hexamethylenetetramine, setting free the acids, which can 
 then be titrated with an alkali. The addition of neutral potassium 
 oxalate has been shown by Jager to give a sharjDer end-point, when 
 phenolphthyalin is used as the indicator. 
 
 Ten c.c. of the urine are diluted with about 50 c.c. of distilled water. 
 Three or four drops of a 1 per cent, alcoholic solution of phenolphthyalin,, 
 and a few crystals, or better a small quantity of powdered, neutral 
 potassium oxalates are added. Decinornial sodium hydrate sokition is 
 then run in from a burette, and the reading taken when a permanent 
 faint pink colour appears. Three c.c. of a neutral solution of formal- 
 dehyde (prepared by diluting commercial " formalin " with an equal 
 volume of distilled water, and adding just sufficient decinormal soda 
 to neutralise the mixtiire, using phenolphthyalin as the indicator) 
 are then added, when it will be found that the pink colour will disappear.. 
 
QUANTITATIVE METHODS 113 
 
 A fiirther addition of decinormal soda is then made until the pink colour 
 is just restored, and the reading is again taken. The difference between 
 the first and second readings gives the amount of acid that was com- 
 bined with ammonia. This mtiltiplied by 0-0014 gives the quantity 
 of ammonia nitrogen in 10 c.c. of the urine in grams. The estimation 
 includes the amino acids and substituted ammonias present in the 
 lu-ine, but as a rule these may be neglected. 
 
 2. Folin. — Twenty-five c.c. of the lu-ine are placed in an areometer 
 cylinder, about 45 cm. high and 5 cm. in diameter. About 1 gram of dry 
 sodium carbonate, 8 to 10 grams of sodium chloride, and a little petroleum 
 or toluol are added, and the neck is closed with a rubber cork provided 
 with a long tube that passes below the surface of the liquid, and a 
 short one that ends in the space above. The shorter tube is connected 
 with a calcium chloride tube filled with cotton-wool, which in its turn 
 is connected with a gas absorption bottle, containing 20 c.c. of decinormal 
 sulphuric acid, 100 c.c. of distilled water, and a few drops of alizarin 
 red as an indicator. The tube leading from the calcium chloride tube 
 terminates below the surface of the contents of the absorption bottle 
 in a special perforated extremity which brings about good contact of 
 the air that passes through with the acid, and so ensures complete 
 absorption of the contained ammonia. A second tube leading from the 
 absorption bottle is connected with a filter pxmap, and air is drawn 
 through the apparatus for an hour and a half, or two and a half hours 
 if much ammonia is present, at the rate of 600 to 700 litres per hour. 
 The acid in the absorption bottle is then titrated with decinormal 
 sodiiim hydrate until a red, not a violet, coloiu" appears, and the number 
 of cubic centimetres of acid neutralised by the ammonia is thus deter- 
 mined. Several estimations may be made at the same time by con- 
 necting the sets of apparatus in series. This is probably the most 
 satisfactory method for determining the ammonia content of the 
 urine, but it needs a well -fitted laboratory. 
 
 Estimation of Beta-oxy butyric Acid.— 1. Beta-oxybutyric 
 acid is oxidised by potassium bichromate and sulphuric acid to 
 acetone, the yield being practically a quantitative one when a 
 solution containing 5 per cent, of sulphuric acid is distilled, ^^ith 
 the addition of the necessary quantity of a 0-1 to 0-5 per cent, 
 solution of potassium bichromate. 
 
 Shaffer's Method. — 25 to 250 c.c. of the vu-ine, according to tie 
 amount of oxybutyric acid expected to be present, are mixed, in a 
 500 c.c. flask, with an excess of basic lead acetate, and 10 c.c. of con- 
 centrated ammonia. The contents of the flask are diluted to 500 c.c. 
 with water, well shaken, and filtered tlirough a dry filter -paper. Two 
 hvmdred c.c. of the filtrate are mixed with 200 c.c. of water, 15 c.c. of 
 concentrated sulphuric acid, and a little talciun powder in a distillation 
 flask provided with a drop -funnel. The mixtiu-e is then distilled until 
 200 to 250 c.c. have passed over. This distillate, " A," contains the 
 
 H 
 
114 GLYCOSUEIA 
 
 preformed acetone, and the acetone derived from the aceto-acetic 
 acid, and may be used for their determination. The residue in the 
 flask is diluted with 400 to 600 of a 0-1 to 0*5 per cent, solution of 
 potassium biclxromate, through the funnel, and again distilled. During 
 the distillation the volume of the liquid is kept constant by adding 
 water tlxrough the drop -funnel to replace that which distils over. If 
 the liquid in the flask assumes a green colour, showing that the bi- 
 clxromate has been used up, a further supply must be immediately 
 introduced tlirough the fvinnel. After about 500 c.c. of the distillate, 
 " B," have been collected it is mixed with 20 c.c. of hydrogen peroxide, 
 3 per cent., and a few cubic centimetres of sodimn hydrate sohition, 
 and redistilled. Three hundred c.c. of this distillate, " Bg," are 
 collected and mixed with 25, or 50, c.c. of decinormal iodine solution, 
 and 10 c.c. of a 40 per cent, solution of potassium hydroxide. The 
 mixtixre is well shaken and allowed to stand for five minutes, 10 c.c. 
 of concentrated hydrochloric acid are then added, and the mixture 
 titrated with decinormal sodium thiosulphate solution, using starch 
 paste as the indicator. The amoiint of iodine absorbed by the 
 acetone present represents the oxy butyric acid that the urine con- 
 tained. The latter may be calculated by allowing 1-794 mg. of 
 beta-oxybutyric acid for each cubic centimetre of the iodine solution 
 absorbed. 
 
 2. Polariscope. — Since beta-oxybutyric acid is levo-rotatory, 
 [o.]^ = - 24-1°, it may be estimated by means of the polariscope. 
 
 The fermented urine is freed from albumen, and decolorised with 
 lead acetate and ammonia, or mercuric nitrate, and its rotatory power 
 determined. According to Minkowski, a rotation of —1° corresponds to 
 
 ^— — per cent., or about 5 per cent., of beta-oxybutyric acid, when the 
 
 100 mm. tube is used. As the rotatory power of the uric acid, creatinin, 
 and other normal levo-rotatory constituents of the urine is very slight 
 they may be neglected, especially as in most cases of diabetes the 
 amount present is reduced, owing to the polyuria. The compound 
 glucvironates, if present in excess, will affect the result, and it is advis- 
 able to exclude them by testing the fermented urine before proceeding 
 to the examination. 
 
 Estimation of Acetone (Folin's Method). — The same appa- 
 ratus is used as in Folin's method for the estimation of ammonia. 
 
 Twenty-five c.c. of the urine, 10 drops of phosphoric acid (10 per 
 cent.), 10 grams of sodium chloride, and a little petrolemii or toluol, 
 are placed in the areometer cylinder. Into the absorption bottle are 
 introduced 150 c.c. of water, 10 c.c. of potassium hydroxide solution 
 (40 per cent.), and a measured voltmie of decinormal iodine solution. 
 The apparatus is then connected up, and a current of air drawn through 
 for twenty to twenty-five minutes. The acetone is carried over into 
 
QUANTITATIVE METHODS 115 
 
 the alkaline hypo-iodite solution and converted into iodoform. Ten c.c. 
 of concentrated hydrochloric acid are now added to the contents of 
 the absorption flask, and, if the iodine present has been in excess of 
 the amount required to combine with the acetone, the original brown 
 colour will be restored. The excess of iodine is then titrated with 
 decinormal sodium thiosulphate sokition, using starch as the indicator. 
 Each cubic centimetre of decinormal iodine that has been absorbed 
 corresponds to 0*96 mg. of acetone. 
 
 Estimation of AcetO-Acetic Acid. — 1. The combined acetone 
 and aceto-acetic acid may be determined from the "A " distillate 
 obtained in Shaffer's method for oxybutyric acid, by making it 
 faintly alkaline, redistilling, and estimating the total acetone by 
 the preceding method. The difference between the preformed 
 acetone, as obtained by Folin's method, and this result will repre- 
 sent the acetone derived from aceto-acetic acid. 
 
 2. Harfs Modification of Folin's Method. — The same apparatus, 
 and reagents, are used as in Folin's method for estimating acetone. 
 
 After the preformed acetone has been removed, the xirine is heated 
 nearly to boiling by immersing the areometer cylinder in a water-bath, 
 another absorption flask, charged with a fresh supply of decinormal 
 iodine, being attached. A current of air is drawn through for twenty- 
 five minutes, during which time all the aceto-acetic acid is converted 
 into acetone and carried over into the alkaline hypo-iodite solution. 
 The same procedure is then followed as for the estimation of acetone, 
 the quantity found representing the aceto-a(.cetic acid of the urine. 
 
 Numerous other methods for the estimation of oxybutyric acid, 
 aceto-acetic acid, and acetone in the urine have been described, 
 but they are all very complicated and give no more reliable results 
 than the preceding, which are comparatively simple, and are carried 
 out Avith similar apparatus and reagents. 
 
 (c) Total Nitrogen 
 
 A healthy man, on an average mixed diet, passes in his urine 
 in twenty-four hours from 10 to 16 grams of nitrogen, which, with 
 a daily total of 1-5 litres, works out at 0-67 to 1-07 per cent. The 
 nitrogen-containing constituents of the urine are degradation pro- 
 ducts of albumen, and their amount is therefore dependent upon 
 the body weight of the individual, the rate of protein metabolism, 
 and the amount of albuminous food material absorbed from the 
 intestine. As we shall see later, an estimation of the total nitrogen 
 of the urine is of importance in persistent glycosuria, both in making 
 a prognosis and estimating the progress of the case, since it indicates 
 the state of protein metabolism within the organism, and also 
 
116 GLYCOSURIA 
 
 gives data from which, the exact degree of the difficulty the patient 
 has in dealing with carbohydrates can be calculated. 
 
 Estimations of total nitrogen in the urine are now almost 
 exclusively made b^T-Kjeldahl's process, or some modification of it, 
 for no other gives such consistently reliable results. The principle 
 of the method is, that the organic constituents of the urine are 
 destroyed by oxidation when it is heated with concentrated sul- 
 phuric acid, the nitrogen of those substances which do not contain 
 it combined "wdth oxygen, forming ammonium sulphate. The urea 
 is converted into ammonia and carbon dioxide. On treating the 
 acid solution ■\\'ith sodium or potassium hydroxide, and distilHng, 
 the ammonia passes over, and may be estimated by collecting it 
 in a measured volume of an acid solution of known strength. From 
 this the nitrogen may be easily calculated. 
 
 1. Gunning's Modification of KjeldahVs Process. — Five c.c. of the 
 clear filtered urine are accurately measured, with a pipette, into a 
 Kjeldahl's digestion flask of Jena glass of about 250 c.c. capacity. 
 Ten c.c. of concentrated sxilphuric acid (sp. gr. 1*84), and 0-2 to 0-5 
 gram of crystallised copper sulphate are added. The flask is loosely 
 closed with a glass ball, and placed in a slightly inclined position on a 
 wire gauze in a fume-chamber. It is then gently heated, with a low 
 flame, until frothing ceases and the black material begins to wash down 
 the sides of the flask. The contents of the flask are allowed to cool 
 somewhat, and 5 to 10 grams of potassium sulphate are introduced. 
 The temperatiu-e is now gradually raised until the acid boils briskly, 
 and the boiling is continued until the solution becomes clear and has 
 only a pale blue, or green, colour. It is then allowed to cool. When 
 it is quite cold, about 100 c.c. of ammonia-free distilled water are 
 cautiously added, and the mixture transferred, by means of a small 
 funnel to a round-bottomed distillation flask of about a litre capacity. 
 The glass stopper, the digestion flask, and the funnel are rinsed several 
 times with distilled water, and the washings added to the contents of 
 the distillation flask, which are made up to 350 to 400 c.c. About 
 15 grams of talc, or a few pieces of granulated zinc, and a few drops of 
 an alcoholic solution (1 per cent.) of phenolphthalein are added. The 
 neck of the flask is now closed with a stopper provided with a tapped 
 fimnel, and a Reitmair or Hopkins biilb-tube, which is attached to an 
 efficient condenser. The delivery tube from the condenser is con- 
 nected with a glass tube, provided with a bulb about half-way along its 
 length, which dips to the bottom of a carefully measured quantitv 
 (50 c.c.) of decinormal hydrochloric, or sulphuric, acid contained in 
 an Erlenmeyer flask of about 250 c.c. capacity. The contents of the 
 distillation flask are now made distinctly alkaline by adding about 
 80 c.c. of sodium hydrate solution (sp. gr. 1-230, or 250 grams per litre) 
 tliroiigh the funnel, and mixed by a gentle rotatory motion. They are 
 
QUANTITATIVE METHODS 117 
 
 then distilled. When 100 to 150 c.c. have distilled over, and the steam 
 no longer gives fumes of ammonium chloride when a glass rod moistened 
 with hydrochloric acid is held in front of the open end of the delivery 
 tube, the contents of the receiver are titrated with decinormal sodium 
 hydrate solution, using cochineal, methyl orange, or congo-red as the 
 indicator. The number of cubic centimetres of decinormal sodium 
 hydroxide used is deducted from the number of cubic centimetres of 
 decinormal acid placed in the receiver, and the difference multiplied 
 by 1-404. The result gives the amount of nitrogen contained in the 
 quantity of urine taken, in milligrams. 
 
 2. At best the distillation process is a lengthy one, so to save 
 time, and also to avoid the error introduced by the action of the 
 hot alkali on the glass, the ammonia formed can be directly 
 estimated in the neutralised oxidation product by the Malfatti 
 formalin process, as described for ammonia nitrogen. 
 
 Rona and Ottenberg give the following directions : 5 c.c. of the urine 
 are mixed with 10 c.c. of concentrated sulphuric acid, and 5 to 8 drops 
 of a 1 per cent, solution of platinum chloride, and heated in a Kjeldahl 
 flask. The digested product is transferred to a .350 c.c. Erlenmeyer flask 
 and made up to 100 c.c. with distilled water, 6 or 7 drops of neutral 
 litmus solution (Kiibel-Tiemann), and 20 c.c. of a 30 per cent, solution 
 of sodium hydrate, are then added, and the flask is well cooled in running- 
 water. When it is quite cold, 30 per cent, sodiiim hydrate is gradually 
 added, at first a cubic centimetre at a time, and later drop by drop, 
 until the fluid is blue, taking care to keep it as cool as possible all the 
 time. It is next feebly acidified with fifth normal acid, and then very 
 carefully neutralised with fifth normal sodimn hydrate and fifth normal 
 acid, comparing the colour with that of a solution consisting of 150 c.c. 
 of distilled water, 1 c.c. of fifth normal sodimn hydrate, and 1 drop of 
 litmus. To the neutral solution is now added 30 c.c. of neutral formal- 
 dehyde, and 1 c.c. of a J per cent, solution of phenolphthalein. The 
 mixture is titrated with fifth normal caustic soda until a violet coloior, 
 due to the combined effect of the litmus with phenolphthalein, appears. 
 The number of cubic centimetres of soda solution used gives the 
 ammonia nitrogen content. After a little practice the titration process 
 takes about ten minutes to carry out. 
 
 BIBLIOGRAPHY 
 
 iVrnold, Zentralb. f. inn. Med., 1900; Weiner klin. Woch., 1899. 
 
 Bang, Biochem. Zeit., 1906. 
 
 Bela, Ann. d. Chem., 1892. 
 
 Benedict, Journ. Anier. Med. Assoc, 1911. 
 
 Black, Journ. of Biolog. Chem., 1908-9. 
 
 Bohrisch, Pharm. Zentralhalle, 1907. 
 
118 GLYCOSURIA 
 
 Bondi and Schwarz, Wiener klin. Woch., 1906. 
 
 Byrac, Bull. d. I. Soc. Chim., 1894. 
 
 Citron, Deut. med. Woch., 1909. 
 
 Einhorn, New York Med. Record, 1887. 
 
 Embden and Schmitz, Abderhalden's Handb. d. Biochem. 
 
 Arbeitmeth., ii., 926-7, 1910. 
 Fehling, Arch. f. physiol. Heilk., 1848 ; Ann. d. Chem. u. Pharm.,. 
 
 1849. 
 Folin, Zeit.f. physiol Chem., 1902-3. 
 Frohner, Deut. med. Woch., 1901. 
 Frommer, Berl. klin. Woch., 1905. 
 Gerhardt, Wiener med. Presse, 1865. 
 
 Gerhardt and Schleisinger, Arch. f. exp. Path. u. Pharm., 1898. 
 Gerrard, Chem. Zentralb., 1896 ; Allen's Chem. of the Urine, 1895. 
 Gunnung, Journ. d. pharm. et d. Chim., 1881 ; Zeit. anal. Chem., 
 
 xxviii. 
 Hart, Amer. Journ. of Med. Sci., 1909. 
 Hasselbalch and Lindhard, Biochem. Zeit., 1910. 
 Henninger, Compt. rend. d. I. Soc. d. Biol., 1884. 
 Hopkins, Journ. Amer. Chem. Soc, 1896. 
 De Jager, Zeit. f. phys. Chem., 1910. 
 Jolles, Zeit. anal. Chem., 1907. 
 Kjeldahl, Zeit. anal. Chem., xxii. 
 Klimmer, Zeit.f. Tiermed. N.F., 1898. 
 Knapp, Ann. d. Chem. u. Pharm., cliv. 
 Lange, Milnch. med. Woch., 1906. 
 Lefevre and ToUens, Ber. d. deut. Chem. Ges., 1907. 
 Legal, Breslauer arz. Zeitschr., 1883. 
 Lehmann, Arch. f. Hygiene, 1898. 
 Lieben, Ann. d. Chem. u. Pharm. Suppl., 1870. 
 Ling, Rendle, and Jones, Analyst, 1905 and 1908. 
 Liphawsky, Deut. tned. Woch., 1901. 
 
 Lohnstein, Milnch. med. Woch., 1899 ; AUg. m,ed. Zentralzeit., 1906, 
 Malfatti, Zeit.f. analytic. Chem., 1908. 
 Morner, Skand. Archiv., 1895. 
 Neubaner, Journ. f. prakt. Chem. , Ixiv. 
 Nenberg, Ber. d. deut. Chem. Gesch., 1899. 
 Le Nobel, Arch. d. exp. Path., 1884. 
 Oerum, Zeit. anal. Chem., 1904. 
 Pavy, Journ. Chem. Soc, 1880 ; Lancet, 1884 ; Chem. Zentralb., 
 
 1879. 
 Penzoldt, Arch. f. klin. Med., 1883. 
 Pfliiger and Allihn, Pflilger's Arch., 1898. 
 Pieraerts, Bull. Assoc Chim. Sucr. et Dist., 1908. 
 Reitmair, Zeit.f. analyt. Chem., 1886. 
 Repiton, Chem. Zentralb., 1907. 
 Reynolds, Proc Roy. Soc, 1871. 
 Riegler, Milnch. med. Woch., 1906. ' 
 
QUANTITATIVE METHODS 119 
 
 Roberts, Edin. Med. Journ., 1861 ; Lancet, 1862. 
 Rona and Ottenberg, Biochem. Zeit., 1910. 
 Rosenthaler, Zentralb. f. analyt. Chem., 1905. 
 Rothera, Journ. of Physiol., 1908. 
 Sachsse, Chem. Zentralb., 1877. 
 Sahl, Deut. med. Woch., 1905. 
 Shaffer, Journ. of Biol. Chem., 1908. 
 Soxlilet, Journ. prakt. Chem. N. F., 1880. 
 Udransky, Zeit. f. phys. Chem., 1892. 
 Vournasos, Bull. Soc. Chim. d. Paris, 1904. 
 Wacker, Zeit. f. physiol. Chem., 1910. 
 
CHAPTER IV 
 
 EXPERIMENTAL GLYCOSURIA 
 
 I. Puncture Diabetes 
 
 In the year 1849 Claude Bernard discovered that if the floor of the 
 fourth ventricle of the brain, between the roots of origin of the 
 eighth and tenth pairs of nerves, is punctured, sugar will shortly 
 afterwards appear in the urine. The glycosuria begins about 
 an hour after the puncture has been made, and lasts only as long 
 as glycogen remains in the liver. If the animal be killed, and 
 the liver examined after the sugar has disappeared from the urine, 
 no glycogen, or at most a trace, is found. Puncture of the ventricle 
 in an animal that has previously been starved for several days 
 gives rise to no glycosuria, and tying the hepatic vessels also 
 prevents the excretion of sugar in the urine. 
 
 During the time that the glycosuria exists the proportion of 
 sugar in the blood is considerably increased, and it is from the 
 escape of this excess through the kidneys that the sugar in the 
 urine arises. According to Bernard's theory the hyperglycaemia, 
 and consequently the glycosuria, is due to hyperglycogenesis, but 
 if Pavy's view is accejDted it must be assumed that they arise from 
 conversion of the liver glycogen into sugar, instead of into fat and 
 j)roteid. According to many physiologists the balance of available 
 evidence suggests that normally there is a constant tendency for 
 the liver to convert glycogen into sugar within its cells, through 
 the action of a diastatic, or glycogenolytic, ferment, but that this 
 tendency is held in check by an inhibitory mechanism controlled 
 from a centre in the floor of the fourth ventricle. 
 
 This diabetic, or glycogenic, centre appears to exist in all animals, 
 and in man glycosuria has been observed to be associated with 
 the presence of tumours, and other lesions, involving this part of 
 the brain. That puncture of the medulla acts by producing irrita- 
 tion of the centre, and not through its destruction, is j)roved by 
 the fact that puncture in an anaesthetised animal does not cause 
 sugar to appear in the urine, by the short duration of the glyco- 
 suria, and more particularly by the exactly similar effects induced 
 by afferent stimuli reaching the centre through the vagus and other 
 
 120 
 
EXPERIMENTAL GLYCOSURIA 121 
 
 nerves. Cutting the vagus in the neck causes a transitory glycos- 
 uria, and subsequent stimulation of the cut central end causes 
 sugar to again appear in the urine. Stimulation of the central 
 end of the cut sciatic nerve, or irritation of the cardiac depressor 
 nerve, acts in a similar way. Tumours pressing on the vagus nerve, 
 and severe neuritis of the sciatic, or facial, nerves, in man have 
 been observed to be associated with glycosuria. 
 
 It is generally stated that the centre exercises its control over 
 the liver through efferent impulses travelling along the spinal cord 
 as far as the upper thoracic region, thence by the upper thoracic 
 spinal roots and rami communicantes into the lower cervical and 
 upper thoracic sympathetic ganglia, from which they pass by the 
 splanchnic nerves to the liver. This conclusion has been arrived 
 at by a study of the effects produced by dividing various parts of 
 these nerve-tracts in animals with a diabetic puncture, but, as 
 MacLeod points out, the procedure involved in all the experiments 
 in which it was found that glycosuria did not follow puncture, 
 coincidentally establish a condition of extreme splanchnic vaso- 
 dilatation, and a consequent fall of blood pressure, which alone is 
 sufficient to cause the glycosuria produced by vagus stimulation 
 to disappear, or become very much less marked. 
 
 How the nervous impulses cause the discharge of sugar is not 
 known with certainty. It is possible that the splanchnic nerves 
 contain secretory fibres which control ferment production by the 
 hepatic cells, as appears to be the case with nerves su]oplying other 
 glands ; but against this is the fact that atropin, which paralyses 
 all true secretory nerve-endings, does not prevent jDuncture diabetes. 
 Bernard thought that the increased sugar production was due 
 merely to vasodilatation, but stimulation of the central end of the 
 cut vagus, which is accompanied by a rise of blood pressure and 
 irritation of the cardiac depressor nerve, associated with a fall in 
 the abdominal blood pressure, both give rise to glycosuria. It 
 has also been suggested that the primary effect of the stimulus 
 is exerted upon the pancreas, and that the changes that ensue in 
 the liver are secondary to this. The most recently advanced view 
 is that the excessive output of sugar by the liver in diabetic punc- 
 ture is dependent upon an increased flow of the secretion of the 
 supra -renals, which are caused to function more actively as a 
 result of the nerve stimulation. By the ingenious method of 
 stimulating the splanchnic nerves after aU the abdominal viscera, 
 except the adrenals, had been removed, Asher has shown that these 
 nerves control the secretory functions of the glands. If, therefore, 
 the splanchnic system is abnormally stimulated, a rise of blood 
 
122 GLYCOSURIA 
 
 pressure and other symptoms, including glycosuria, will result. 
 It has also been found that if the left supra-renal is cut off from the 
 left sympathetic nerve, no sugar appears in the urine after the 
 medulla has been punctured, and it is therefore thought that the 
 stimulus is transmitted by the left sympathetic to the left supra- 
 renal, whence it passes to the right by the connecting nerves, and 
 as a result the supra-renals are aroused to abnormal functional 
 activity. 
 
 In man glycosuria may result from injuries to the head, shock 
 to the central nervous system, and even strong psychical excite- 
 ment, presumably, at any rate in some instances, through irrita- 
 tion, or stimulation, of the glycogenic centre. In many nervous 
 diseases more or less transient glycosuria may occur, and an existing 
 glycosuria may be increased by intercurrent nervous excitement 
 or irritation. 
 
 II. Drug Glycosuria 
 
 [a) Phloridzin Glycosuria. — This is the best known, and 
 most thoroughly studied, form of glycosuria produced by the action 
 of drugs, and offers many points of particular interest. It was first 
 described in a series of papers published by von Mering between the 
 years 1886 and 1889. Phloridzin is a glucoside obtained from the 
 bark of apple, pear, and cherry trees, which, by boiling with dilute 
 acids, is split into dextrose (38-1 per cent.) and phloretin. When 
 given in doses of about 1 gram per kilo by the mouth, or subcu- 
 taneously in doses of about 0*3 to 0*5 grams per kilo, to dogs or other 
 animals, it causes a marked but transient glycosuria, which passes 
 off in a few hours, unless the drug is readministered. In well-fed 
 animals as much as 19*1 per cent, of sugar may be present in the 
 urine. During starvation the sugar excretion is less (0-3 to 2-5 per 
 cent.), but is more constant. The amount of sugar excreted does 
 not depend, within wide limits, upon the dose of phloridzin used, if 
 enough is given to produce the maximum effect, so that the sugar 
 in the phloridzin itself cannot be an important factor in the sugar 
 excretion. It is to the action of the phloretin that the gtycosuric 
 effect must be ascribed. A single dose of phloridzin does not ap- 
 preciably affect the amount of glycogen in the liver and muscles, 
 but after repeated doses the glycogen may disappear to a large 
 extent from the liver, and also from the muscles, although the 
 reduction is not as complete as in some other forms of experimental 
 glycosuria. The most remarkable difference between phloridzin 
 and other glycosurias is, however, the absence of a hyperglycsemia, 
 such as might be expected by analogy with the effects produced 
 
EXPERIMENTAL GLYCOSURIA 123 
 
 by a diabetic puncture, and other conditions, causing the appearance 
 of sugar in the urine. It has been stated that a decreased amount 
 of sugar in the blood (hypo-glycsemia) is, in fact, a characteristic 
 feature of phloridzin diabetes, but the experiments of Pavy, and 
 later of Coolen and Kolisch, have shown that the sugar content of 
 the blood is if anything slightly increased, but never sufficiently 
 to cause marked hyperglycsemia. 
 
 Apparently, therefore, phloridzin does not act upon the glyco- 
 genic functions of the liver and other organs, but simply causes a 
 draining away of sugar into the urine through the blood, to replace 
 which the stored glycogen is converted into sugar. In confirma- 
 tion of this view is the fact that, while in other forms of experi- 
 mental glycosuria ligature of the ureters, or excision of the kidneys, 
 causes a marked rise in the percentage of sugar in the blood, a similar 
 rise does not occur under the same conditions in phloridzin diabetes. 
 The absence of hyperglycsemia, and the relatively slight disappear- 
 ance of glycogen, in phloridzin glycosuria led von Mering to suggest 
 that the drug has a direct action on the kidneys and makes them 
 more permeable to dextrose, so that the normal sugar of the blood 
 is drained off. Later, von Mering suggested that the phloridzin is 
 taken up by the renal cells and decomposed by them into dextrose 
 and phloretin, the dextrose being excreted in the urine and the 
 phloretin passing back into the blood. There it combines with 
 dextrose to form phloridzin again, and this, in its turn, is decom- 
 posed by the renal cells into sugar and phloretin. Owing to a 
 gradual escape of phloretin into the urine the action of the drug 
 slowly diminishes, and a renewed dose has to be given to keep up 
 the glycosuria. The renal character of phloridzin diabetes was 
 first demonstrated by Zuntz, who placed cannulae in the upper 
 portions of the two ureters, and injected phloridzin into the 
 renal artery of one side. On the injected side sugar appeared in 
 the urme in two minutes, but it was three minutes later before 
 sugar could be detected in the urine coming from the other 
 side, the delay being due to the lapse of time necessary for the 
 transportation of the phloridzin from the injected to the other 
 kidney. 
 
 It would seem, however, that the excretion of sugar does not 
 depend upon a condition of the kidneys which simply allows it 
 to leak away, but rather upon actual secretory activity, for disease 
 or injur}'- of the renal tissue diminishes or prevents the glycosuria. 
 Biel and Kolisch have also shown that when phloridzin is passed 
 through the vessels of an isolated kidney, sugar is found in the 
 urine. Phloridzin itself causes necrotic changes in the renal 
 
124 GLYCOSURIA 
 
 epithelium after prolonged administration, and diuretics, although 
 they increase the volume of urine, do not raise the excretion of 
 sugar as they do in diabetes mth hyperglycaemia. Paw, Brodie, 
 and Siau have suggested that the sugar that appears in the urme 
 in phloridzin glycosuria is formed in the kidney out of a precursor 
 contained in the blood, and is not the blood sugar. According to 
 their theory the jorecursor is the sugar that is loosely combined 
 with the serum proteid. This is split off in the kidneys and ex- 
 creted, the proteid returning into the circulation and there com- 
 bining -v^ith more sugar. When, however, the animal is starved 
 so that no fresh sugar is available, the renal cells attack the proteid 
 molecule itself, setting free the sugar that can be derived from it, 
 and also the nitrogen-containing moiety. Anah^ses of the urine 
 of animals to which phloridzin has been repeatedly given, and that 
 are then starved, have shown that the elimination of sugar does 
 not cease, but continues at a lower level, A\hile at the same time 
 the elimination of nitrogen is increased until a constant maximum 
 ratio of dextrose to nitrogen (D : IM) of 2-8:1, that is strikingly 
 similar in different animals, is established. Stiles and Lusk, 
 however, obtained a higher ratio of 3-65 : 1 in dogs with normal 
 kidneys after subcutaneous injections of phloridzin, which corre- 
 sponds to the ratio found by Mandel and Lusk in human diabetes 
 v/hen the patient is given a diet of meat and fat. Reilley. Xolan, 
 and Lusk have shown that in fasting phloridzinised dogs the 
 D : IS!" ratio does not vary after the ingestion of sufficient meat to 
 double the quantity of nitrogen in the urine, for the sugar is at the 
 same time doubled. The sugar production is therefore propor- 
 tional to the j)roteid metabolism, and apparently must be derived 
 from the proteid. That an excessive breaking do^^^n of proteids 
 occurs in advanced poisoning by phloridzin during starvation is 
 also shown by the presence of ^-oxybutjTric acid, aceto-acetic acid, 
 &c., in the urine. 
 
 Phloridzin glycosuria is readily produced in man as well as in 
 animals, and the use of phloridzin has been suggested as a test 
 of the functional acti\dty of the kidne3's. Von Mering quotes a 
 case in which he gave 1 gram of jobloridzin night and morning 
 for a month, \\ith the result that the patient jDassed from 2| to 3 
 litres of urine a day containing 2-7 to 3-7 per cent, of sugar, with 
 a total excretion of 2 kilos 728 grams of sugar for the thirty days. 
 Achard and Delamare, after injecting 15 mg. of phloridzin 
 subcutaneous^, found that sugar appeared in the urine in half 
 an hour, and that the glj^cosuria j)ersistecl for three and a half 
 hours. Fifty mg. injected into the same j)atient produced a 
 
EXPERIMENTAL GLYCOSURIA 125 
 
 glycosuria lasting six hours, and the quantity of sugar excreted 
 was 14 grams. 
 
 (b) Glycosuria due to Drug's and other Toxic Influences. 
 
 — Besides phloridzin, a number of other substances, when taken 
 by the mouth, or injected subcutaneously, &c., may cause a more 
 or less marked glycosuria. Some of these probably do so by stimu- 
 lating the glycogenic centre in the floor of the fourth ventricle ,•; 
 others appear to act by stimulating the splanchnic sympathetics, 
 A\'hile others again seem to exert a direct influence upon the cells 
 of the liver and pancreas, or upon the renal epithelium. 
 
 Bock and Hoffman in 1871 made the interesting observation 
 that the intravenous injection of a slightly hyper-tonic solution 
 of sodium chloride into rabbits gives rise to a glycosuria, which 
 can be prevented by the addition of a soluble salt of calcium to 
 the solution. That the glycosuria in this instance is of central 
 origin, is suggested by the fact that it does not occur if thesjalanchnic 
 nerves are cut before the injection is made. If, however, the salt 
 solution is injected into the central end of the axillary artery, so 
 that the salt may reach the medulla by way of the vertebral, glycos- 
 uria is produced, and is then not prevented by the simultaneous 
 injection of calcium chloride (Fischer). 
 
 It has been repeatedly noticed that the administration of 
 morphia may cause hyperglycsemia, and glycosuria. The weU- 
 known affinity of this drug for the nervous system suggests that it 
 acts uxoon the glycogenic centre, but it is possible that, like ether 
 and other asphyxiating substances, it may stimulate the splanchnic 
 sympathetics. It has also been suggested that it exerts a direct 
 effect upon the liver cells (Lepine). In this connection the question 
 of asphyxial glycosuria may be considered, for it has been shown 
 by Underbill that the hyperglycaemia and glycosuria produced by- 
 certain drugs, such as morphine, nicotine, p3rridin, anaesthetics, &c., 
 can be prevented by the free administration of oxygen. 
 
 Asphyxial glycosuria is most easily induced experimentally by 
 clamping the trachea, or injecting curare. After section of all the 
 hepatic branches of the coeliac plexus, glycosuria is no longer 
 produced by clamping the trachea, suggesting that the condition 
 is due to asphyxial stimulation of the nerve centres. It has been 
 found, however, that hyperglycsemia does occur after such isola- 
 tion of the liver from the nervous system, when the asphyxia is 
 effected by injecting curare, indicating that the profound venosity 
 of the blood, beside acting on the nervous centres, may also chrectly 
 influence the liver cells. Further experiment has suggested that 
 
126 GLYCOSUKIA 
 
 the latter action is probably due to the carbon dioxide contained 
 in the intensely venous blood (Macleod), and it is probable that 
 the free administration of oxygen in certain drug gtycosurias pre- 
 vents the hj^perglycsemia and appearance of sugar in the urine 
 by removing the excess of carbon dioxide {e.g. curare, morphine, 
 amyl nitrite, general anaesthetics). 
 
 Glycosuria may follow the administration of caffein, theo- 
 bromine, and diuretin ; and as the chief effect of these drugs 
 is to produce polyuria, it has been suggested that the glycosuria 
 is a result of their action on the kidneys. The experiments of 
 Nishi have shown, however, that, with diuretin at least, the ex- 
 cretion of sugar is dependent upon impulses transmitted through 
 the sympathetic nerves to the supra-renals. Uranium salts 
 (Chittenden and Lambert), chromic acid and chromates (Pal), 
 and cantharides (Richter) are suiDposed to j)roduce glycosuria by 
 their action on the renal epithelium. 
 
 Other substances, among which maj^ be mentioned alcohol and 
 the toxines of acute infectious diseases, appear to induce glycos- 
 uria, chiefly through their action upon the pancreas. 
 
 The administration of strychnine, phosphorus, arsenic, mercury, 
 hydrocyanic acid and cyanides, chloral, nitro-benzol, chloroform, 
 acetone, ether, amyl nitrite, and organic or mineral acids, the 
 inhalation of carbon monoxide, extensive hsemorrhages, immersion 
 in water, &c., are said to produce glycosuria. With some there 
 can be no doubt that sugar does appear in the urine, and the ex- 
 jjlanation is one of those given above ; but mth others the mecha- 
 nism of its production is uncertain, and with others, again, it is 
 probable that glucuronic acid, and not dextrose, is the substance 
 that gives the urine its reducing power. In any case the condition 
 is a shght and transitory one not likely to cause any practical 
 difficulty. 
 
 III. Glandular Glycosuria 
 Since the pioneer work of Baumann on the thyroid gland, in 
 1895, which resulted in the isolation of iodothyrin, physiologists 
 have learnt to appreciate the far-reaching influence of the exchange 
 of matter that is continually going on between the living cells of 
 the organism, and to realise that every tissue of the body is 
 continually forming intermediate and end-products that serve as 
 stimulants for other tissues, or in some way exercise an influence 
 upon the metaboHc processes of other organs. It is now believed 
 that chemical correlation, and chemical control, exists between 
 practically every gland and tissue of the body. In some instances 
 
EXPERIMENTAL GLYCOSURIA 127 
 
 it has been possible to isolate the specific chemical messenger, or 
 " hormone," through which one tissue activates another. Thus 
 Bayliss and Starling have shown that the hydrochloric acid secreted 
 by the stomach acts upon the mucous membrane of the upper 
 part of the intestine, and gives rise to a substance to which they 
 have given the name " secretin " ; this being absorbed into the 
 blood travels to the pancreas, arousing it to functional activity 
 and causing a flow of its secretion. Again, we find the intestine 
 secreting a specific substance, " entero-kinase," which, reacting 
 with the trypsinogen of the pancreatic juice, converts it into the 
 active proteolytic ferment trypsin. These and similar observations 
 tend to emphasise the part that specific chemical products play in 
 regulating the metabolism of the body. 
 
 The ductless glands are believed to be particularly imijortant 
 in this connection, for they appear to be engaged in the elaboration 
 of internal secretions, the active substances of which are of special 
 importance in the metabolism of the organism. Their histological 
 characters, the absence of a duct system, and their intimate con- 
 nection with the circulation and lymphatics, suggest that, on the 
 one hand, they receive material from the blood and lymph, and, 
 on the other, return specific chemical products to the circulation. 
 Moreover, it has been shown, as the result of their experimental 
 ablation in animals, that their internal secretions are necessary for 
 the normal activity of the nervous system, for the circulation, and 
 for the healthy metabolism and growth of the organism. If these 
 glands cease to function, the loss of their specific products leads 
 to disease, and possibly to death, while an increase in their activities 
 may give rise to equally disastrous results. The many-sided rela- 
 tions, and inter-relations, existing between them and the tissues 
 generally render it difficult to clearly demonstrate the particular 
 part that any one gland plays in the economy, so that jDrogress 
 has been slow, and the explanations given by different authors of 
 the phenomena observed on removing a gland are often at variance. 
 Still, sufficient evidence has now been accumulated to render 
 possible a pro\dsional theory of the results, and to provide material 
 for a working hypothesis as to their action upon each other, and the 
 ■organism generally. 
 
 Among their other functions the ductless glands exert a pro- 
 found influence upon the carbohydrate metabolism of the bodj^, 
 and although for many years interest has been chiefly centred in 
 the pancreas, there can be no doubt that the su23ra-renals, the 
 pituitary gland, and the thyroid also play a part which is probably 
 no less important. We shall first consider the experimental facts 
 
128 GLYCOSURIA 
 
 connecting each of these glands with carbohydrate metabolism, 
 and subsequently deal with the theories that have been advanced 
 with a view to unifying the results. 
 
 {a) The Pancreas, — The anatomical similarity of the pancreas 
 to the salivary glands led the early observers to consider that their 
 functions were also of the same nature, and it was not until C. 
 Bernard discovered, in 1849, that the pancreatic juice was con- 
 cerned in the digestion of fats, and, in 1856, that it was also capable- 
 of acting on proteid materials, that the vastly greater importance- 
 of the pancreas, even as a digestive organ, came to be recognised. 
 The part that it j)lays in the internal metabolism of carbohydrates 
 does not appear to have been suspected until 1875, when Bouchardat 
 suggested that lesions of the j)ancreas were capable of causing 
 diabetes. The subsequent failure of experimental attempts to 
 produce glycosuria in animals through damming back the pan- 
 creatic secretion, by ligature of the ducts, or by conveying the 
 secretion outside the body by a fistula, led such an authority as 
 Cohnheim to deny the connection, and to regard all pancreatic 
 changes met "nith in diabetes as secondary,- or accidental compli- 
 cations. In 1889 von Mering and Minkowski placed the pancreatic 
 theory of diabetes on a secure footing, by showing that total 
 extirpation of the pancreas in dogs is followed by severe diabetes, 
 which joersists until the death of the animal. In a series of papers 
 published between 1890 and 1893 these observers, and JVIinkowski 
 alone, gave further details of the operation, and recorded their 
 experience of its effect on a large number, and great varietj^, of 
 animals. 
 
 After complete extirpation of the pancreas in dogs sugar usually 
 appears in the urine the day after the operation, but occasionally 
 in from three to five hours, and gradually increases in amount 
 until it reaches a maximum of 8 to 10 per cent, about the third 
 day. On a diet of bread and meat, a dog of 8 kilos ^\dll then be found 
 to be passing from 70 to 80 grams of sugar in the twenty-four 
 hours. If no food is given, a gradual fall occurs after the third 
 day, and continues until a constant level is attained, but even. 
 after seven days' starvation glycosuria is still present. There is a 
 marked increase in the quantity of urine passed, a dog of 7 kilos 
 voiding from 1000 to 1200 c.c. in the twenty-four hours. Although, 
 a depancreatised animal eats and drinks well, it rapidly wastes 
 and loses strength, so that death takes place from inanition in 
 about four weeks, even when lung disease, or troubles arising from 
 the invariable disinclination of the operation wound to heal, do not 
 bring about a fatal issue at an earlier date. When the animal is- 
 
EXPERIMENTAL GLYCOSURIA 129 
 
 too weak to move about, the excretion of sugar begins to diminish, 
 although food is being taken, and a few days before death it may 
 disappear altogether, especially when there is suppurative peri- 
 tonitis. Coincidentally with the fall in sugar excretion acetone, 
 aceto-acetic acid, and /3-oxybutyric acid make their appearance 
 in the urine. 
 
 Examination of the blood of animals with pancreatic glycosuria 
 shows that it contains a proportion of sugar much in excess of 
 the normal 0-1 per cent., sometimes as much as 0-4 per cent, being 
 found, and when the ureters are tied, or the kidneys are removed, 
 the proportion is still further increased, thus pointing to an accu- 
 mulation of sugar in the blood as the immediate cause of the glycos- 
 uria. Post-mortem the subcutaneous and tissue fat are found 
 to have disappeared, and the muscles are diminished in weight. 
 The most striking phenomenon is a relative hypertrophy of the 
 liver, due apparently to fatty infiltration. In one instance, quoted 
 by Carnot, the liver was found to constitute 8-37 per cent, of the 
 total weight of the body, and to yield 47"5 per cent, of fat. If the 
 animal is killed a few days after the pancreas has been removed, 
 the glycogen normally present in the liver and muscles is sesn 
 to have disappeared, or only to be present in traces, no matter 
 whether the animal has been starved or liberally fed. It is usually 
 found, however, in excessive quantities in some situations, such 
 as the epithelium of Henle's tubes, the heart muscle, and the leuco- 
 cytes, where it does not normally occur in abundance. 
 
 Minkowski observed that more sugar is excreted at the height 
 of the glycosuria by a dog that has been well fed previous to the 
 operation, than by one that has had little food. It would therefore 
 seem that the increased proportion of sugar in the blood, and 
 that excreted in the urine in the first instance, comes from the 
 glycogen stored in the liver and muscles from the food previously 
 taken, and that excision of the pancreas has in some way interfered 
 with the power of these tissues to replenish their stock. The 
 source of the sugar when all the available glycogen has been used, 
 and the sugar excretion has fallen to a constant level, appears to 
 be the proteins of the tissues. That this is so is suggested by the 
 results of observations on the dextrose to nitrogen ratio (D : N) 
 in the urine. If this is determined in a fasting depancreatised 
 dog, it is found to have a constant value, which, according to Min- 
 kowski, who investigated the urines of seven animals on twenty- 
 two different days, averages about 2-8 : L This relation of 2-8 
 gi-ams of dextrose to 1 gram of nitrogen remains practically un- 
 changed, no matter whether the animal is fasting, or is fed on 
 
 I 
 
130 GLYCOSURIA 
 
 flesh alone, or on other forms of protein food {e.g. plasmon, nutrose, 
 casein, &c.), indicating that the sugar and the nitrogen are derived 
 from a common source, the proteins of the tissues and food. The 
 D : N ratio 2-8:1 is, however, not what ought theoretically to 
 be obtained if all the carbon of the proteins were converted into 
 dextrose, for we should then have a ratio of about 7:1. It is 
 therefore apparent that only some 45 per cent, of the possible 
 sugar is excreted, which must mean, either that all the carbon of 
 the protein is not converted into sugar, or, that some of the sugar 
 formed is destroyed. The former appears to be the more probable 
 explanation, although many observers maintain that depancreatised 
 dogs still retain slight sugar-destroying powers. 
 
 The administration of a moderate amount of dextrose to a 
 depancreatised dog results in the whole, or nearly the whole, re- 
 appearing in the urine. At the same time, however, the nitrogen 
 elimination usually falls below the level of the fasting state, sug- 
 gesting that the sugar either undergoes partial oxidation and so 
 spares the jaroteids, or its presence interferes with diffusion from 
 the tissue cells, and so diminishes proteid destruction by raising 
 the proportion of sugar in the blood. The effects of small quantities 
 of sugar cannot be accurately estimated, owing to the natural 
 variations in sugar excretion, and the influence of large amounts 
 on the sugar excretion is complicated by the diarrhoea and intes- 
 tinal disturbances that they produce. 
 
 Experiments have also been made with other forms of carbo- 
 hydrate. The most interesting results are those that have been 
 obtained with levulose, and substances containing it. When a large 
 amount of levulose is given to a depancreatised dog, with a constant 
 D : N ratio, a rise in the dextrose excretion occurs, and a small 
 quantity of Isvulose appears in the urine ; but if only small quantities 
 of levulose are given the increase in dextrose excretion is compara- 
 tively slight, and no levulose is found. Examination of the glyco- 
 gen content of the liver and muscles of such a dog shows a con- 
 siderable amount to be present (8-14 per cent, in the liver, and 0-81 
 per cent, in the muscle in one case), so that, although the power to 
 form glycogen from the dextrose has been lost, the power to convert 
 levulose into glycogen is retained to a considerable extent. It is 
 apparently only when more levulose is given than can be quickly 
 converted into glycogen that the excess is excreted in the urine, 
 partly as unchanged levulose, and partly as dextrose. That the 
 glycogen in the liver and muscles comes from the levulose itself, is 
 shown by the rise in the D : N ratio that follows its administration, 
 readings of 11 to 13-5 being met with in some instances. With inulin, 
 
EXPERIMENTAL GLYCOSURIA 131 
 
 which only yields levulose on hydrolysis, similar results have been 
 •obtained. The administration of cane-sugar causes an increase 
 in the dextrose excretion corresponding to a little more than half 
 the sugar administered, shoA^dng that the whole of the dextrose, 
 but only part of the contained levulose, is excreted in the urine. 
 •Since maltose on hydrolysis only yields dextrose, the sugar ex- 
 ■cretion is increased as though dextrose alone had been given. 
 Lactose very readily undergoes fermentation in the intestine, so that 
 •exact observations are not possible, but it causes a considerable 
 increase in the dextrose excretion, and it is probable that this is 
 not only due to the contained dextrose, but that a certain amount 
 is derived from the galactose as well. Owing to the absence of 
 the secretions of pancreas from the intestine, starches are very 
 imperfectly digested, and appear to a large extent unchanged 
 in the foeces. A certain proportion is also broken down by the in- 
 testinal bacteria. The small part that is digested appears to be 
 absorbed in the usual way as dextrose, and consequently increases 
 the output of this sugar in the urine. 
 
 The effects produced by removing the pancreas have been 
 most thoroughly and completely investigated in dogs, but ana- 
 logous results have also been obtained with other members of the 
 vertebrate series, including cats and pigs (Minkowski and Harley), 
 ■carnivorous birds (Weintraud, Kausel, and Langendorff), frogs and 
 turtles (Aldehoff and Markuse), and eels (Capparelli). The pro- 
 portion of sugar in the blood has been shown by Kausel to be in- 
 creased in herbivorous birds by removing the pancreas, but glycos- 
 uria was found to only occasionally occur, j^robably because the 
 kidneys of these animals are not readily pervious to sugar. Ex- 
 periments on rabbits have usually been unsuccessful, owing to 
 the technical difficulties encountered in total extirpation of the 
 :gland, which is spread out in a diffuse manner between the layers 
 ■of the mesentery ; but Hedon, and later Sauerbeck, succeeded in 
 producing atrophy, and transient glycosuria, by injecting oil into 
 the duct of Wirsung. The glycosuria appeared at the earliest 
 ■on the twentieth day, and was at its height from the thirtieth to 
 the thirty-eighth day after the injection. 
 
 Partial extirpation of the pancreas may, or may not, give rise 
 to glycosuria, according to the amount left behind, and its con- 
 dition. If about a fourth, or fifth, of the gland is left, glycosuria 
 ■only occurs if carbohydrates are present in the food {" alimentary 
 glycosuria "). A larger proportion usually prevents the condi- 
 tion. Less gives rise to frank diabetes. Even when sugar does 
 .not appear in the urine after partial extirpation, it will do so if the 
 
132 GLYCOSURIA 
 
 remnant is subsequently removed, and may gradually develop 
 as the fragment atrophies. Sandmeyer found that the first trace 
 of sugar appeared in the urine of a dog, part of whose pancreas 
 had been removed, seven weeks after the operation, but it was 
 not until after the lapse of thirteen and a half months that the 
 diabetes became permanent. Death occurred eight months later. 
 At the post-mortem a remnant of pancreatic tissue, weighing 
 0-36 gram, and showing no trace of glandular structure, was found 
 adherent to the posterior wall of the stomach, while attached to 
 the lower part of the duodenum was a piece of slightly changed 
 gland- tissue about the size of a pea. It has been remarked that 
 partial removal of the pancreas usually gives rise to more serious 
 polydipsia, polyphagia, and polyuria than total extirpation of the 
 gland. 
 
 The first exj)lanation of the results of these experiments that 
 suggests itself is that removal of the pancreas leads to impaired 
 digestion from absence of the j)ancreatic juice from the intestine, 
 and that this is in some way responsible for the glycosuria. But 
 the fact that diabetic symptoms do not supervene unless almost 
 ' the entire gland has been removed is against such a theory. More- 
 over, if the secretion of the gland is diverted, and intestinal diges- 
 tion is thus prevented, diabetes does not follow, although marked 
 wasting may occur. Ligature of the pancreatic duct like^\ise fails 
 to give rise to glycosuria as a rule. 
 
 Disease of the solar plexus has been regarded by some as the 
 cause of the diabetes, and, as the solar j)lexus is almost unavoidably 
 injured in the removal of the pancreas, this might possibly be 
 the explanation of the symptoms caused by the depancreatisation 
 of animals. But it was shown by Minkowski that, if the descending 
 portion of the gland in a dog is transplanted into the subcutaneous 
 tissue of the abdominal wall and allowed to become engrafted 
 there, the intra-abdominal portion can be removed, after the graft 
 has been severed from its nervous connections, ^vithout producing 
 glycosuria, and that diabetes only develops if the graft is sub- 
 sequently removed, or atrophies. It is not improbable, however, 
 that the transitory glycosuria that follows immediately after partial 
 extirpation of the gland, and is sometimes seen as a consequence of 
 manipulation, or irritation, of the pancreas, or jDcri-pancreatic region, 
 may be of reflex nervous origin (MinkoAvski). 
 
 The cause of the hypergiycaemia, and the consequent glycosuria, 
 appears, therefore, to be the loss of some influence which the 
 pancreas exerts by way of the blood, or lymph, stream. It is 
 conceivable that this influence may be exerted in four difi^erent 
 
EXPERIMENTAL GLYCOSURIA 133 
 
 ways : (1) The pancreas may destroy sugar coining to it in the 
 Iblood ; (2) it may secrete an enzyme that destroys sugar in the 
 blood ; (3) the cells of the pancreas may normally destroy, or 
 modify, some toxic substance produced in other parts of the body 
 which interferes with the utilisation of sugar by the tissues ; (4) 
 the pancreas may produce an internal secretion which is neces- 
 sary for the splitting up and use of sugars by the other cells of 
 the organism. 
 
 1. As to the first possibility, that the pancreas normally 
 destroys sugar coming to it in the blood, there is no evidence to 
 support it. Numerous exjDeriments have failed to show that the 
 pancreas possesses greater glycolytic powers than other organs, 
 and since only a small fraction of the gland, which may be merely 
 engrafted in the subcutaneous tissue, is sufficient to prevent glycos- 
 uria, it is hardly conceivable that it is necessary for the bulk 
 of the blood to actually transfuse the joancreas to be subjected 
 to some glycolytic action. 
 
 2. The second hypothesis, that the chief function of the pan- 
 creas in carbohydrate metabolism is to furnish a glycolytic enzyme 
 to the blood, is particularly associated with the name of Lepine, 
 who has stoutly maintained the presence of such an enzyme in 
 healthy blood, and states that it is diminished in depancreatised 
 dogs. Crofton, who has supported Lepine's views, claims to have 
 isolated the glycolytic enzyme, and to have identified it with 
 trypsin. Other observers have found that, when precautions are 
 taken against contamination, normal blood possesses no glycolytic 
 power, and consider that contrary results have been due to post- 
 mortem changes, and to the action of micro-organisms. Blumen- 
 thal and others have asserted that the cells of the pancreas, liver, 
 spleen, muscle, &c., have strong glycolytic powers, which are much 
 increased if pancreatic extract is mixed with the cell- juice of other 
 organs ; but, according to Umber, M'hen careful precautions are 
 taken against contamination, the tissues outside the body exhibit 
 only very slight glycolytic powers. The same observer also showed 
 that the sugar-splitting power of the blood in the pancreatic vein 
 is not greater than in the general arterial or venous system, as it 
 should be if the pancreas secretes a sugar-destroying ferment. 
 
 3. The anti-intoxication theory is that which was at first 
 favoured by Minkowski, but it was later abandoned by him in 
 favour of the fourth hypothesis. He, and v. Mering, showed that 
 if the blood of a depancreatised dog is transfused into a healthy 
 animal glycosuria does not ensue, as it might be expected to do 
 if the sugar excretion were dependent upon an accumulation of 
 
134 GLYCOSURIA 
 
 toxic substances. Lombroso has proved that the glycosuria is- 
 not the result of an absorption of toxic substances from the in- 
 testine, formed in consequence of the defective pancreatic diges- 
 tion, as had been suggested, by injecting fluid from a pancreatic- 
 fistula in one dog into the duodenum of another, depancreatised, 
 animal. He' found that, although the digestion of the second animal 
 was much imjDro ved, the glycosuria was in no way affected. Tuckett, 
 Bosanquet, and others have from time to time revived the anti- 
 intoxication theory in a modified form. Tuckett has suggested 
 that the pancreas normally forms an internal secretion which 
 enters the circulation by way of the thoracic duct, and there neutra- 
 lises a toxine absorbed by the Ij^nphatics from the intestine during 
 digestion. In support of this he states that if the thoracic lymph 
 from a fasting dog is injected into the portal circulation of a cat, 
 no hyperglycsemia, or glycosuria, results ; but that if the lymph 
 from a dog during digestion is similarly injected, a hyperglyceemia, 
 varjdng from 0-3 to 0-9 per cent., and a glycosuria, varying from 
 1*0 to 9-0 per cent., are produced. Confirmation of his results are, 
 however, lacking. Bosanquet, in his Goulstonian lectures, favoured 
 the view that diabetes is due to an increased internal dissociation 
 of tissue (possibly fat) into sugar, caused by a toxic substance that 
 is produced in the course of normal metabolism, and Avhich is 
 neutralised by the pancreas. 
 
 4. The remaining hypothesis, that the pancreas produces an 
 internal secretion which, j^assing into the blood, brings about the 
 destruction of dextrose elsewhere in the organism, is the one now 
 most generally held, for it most readily and completely explains 
 the phenomena resulting from the removal of the pancreas in 
 animals. 
 
 Numerous attempts have been made to prove the existence of 
 this internal secretion by administering preparations of the pan- 
 creas, or the fresh gland, by the mouth, by the injection of various 
 extracts, and even of serum from the pancreatic vein, subcutaneously, 
 or into the circulation, but they have all proved to be %^ithout 
 effect on the glycosuria, so that it seemed that either the internal 
 secretion of the pancreas did not exist, or that, as Pfliiger sug- 
 gested, the Hving organism supplies the circulating blood ^^ith its 
 secretion so rapidly that there is never more than a trace in the 
 gland at one time. 
 
 Evidence in favour of the latter suggestion was furnished by 
 the experiments of Forsbach. This observer showed that if two 
 dogs are united by skin, muscle, and peritoneum, so that an ex- 
 change of blood and lymph can take place through their com- 
 
EXPERIMENTAL GLYCOSURIA 135 
 
 municating peritoneal cavities, and the pancreas of one is extir- 
 pated, the glycosuria that should occur is checked, but, on sepa- 
 rating the depancreatised dog from the normal animal, sugar 
 appears in the urine of the former to the same extent as usual. 
 Hedon also found that when cross circulation is established between 
 a normal and a diabetic dog, the sugar in the urine of the latter 
 diminishes. When the experiment was repeated with two diabetic 
 dogs, however, it was found that there was also a diminution in 
 the glycosuria, a result which he interprets as being due to an 
 alteration in the permeability of the kidneys for sugar. 
 
 Hedon subsequently carried out a further series of experiments, 
 inserting a portion of the pancreas of a healthy dog into the circu- 
 lation of a diabetic animal. He found that no effect was produced 
 on the glycosuria when the carotid and jugular were used, but 
 that when the pancreas was inserted into the circulation of the 
 spleen the sugar almost disappeared after some hours, to retuni 
 again when the connection between the two animals was severed, 
 thus suggesting that the normal pancreas only checks glycosuria 
 when it is so placed that its internal secretion enters the portal 
 circulation directly. He also found that if blood from the pan- 
 creatic vein of a normal dog were injected into a mesenteric vein 
 of a diabetic animal, the excretion of sugar fell, for a time, almost 
 to the normal. Hedon considers that the first effect of the pan- 
 creatic secretion was to render the kidneys less j^ermeable to sugar, 
 for in all his experiments the sugar in the urine diminished more 
 markedly than the sugar in the blood. These observations, although 
 they need to be confirmed, tend to support the view that the 
 pancreas influences carbohydrate metabolism by the production 
 of an internal secretion, and also suggest that its action is to inhibit 
 the production of sugar by the liver, rather than to facilitate its 
 consumption by the tissues. 
 
 On the other hand, the recently published experiments of 
 Knowlton and Starling, while they confirm the hypothesis that 
 the pancreas controls sugar metabolism through an internal secre- 
 tion, suggest that its presence is necessary for the utilisation of 
 sugar by the tissues. Working with isolated heart-lung prepara- 
 tions, these observers found that the normal heart, fed with normal 
 blood, under approximately physiological conditions, consumes 
 about 4 mg. of sugar per hour per gram of heart muscle, but that 
 the sugar consumption by the hearts of depancreatised dogs is 
 practically nil, or at the least very much less. On feeding the heart 
 from a diabetic dog with blood from a normal animal, it was found 
 that, even in the first hour, the consumption of sugar was con- 
 
136 GLYCOSURIA 
 
 siderably above that of a diabetic heart fed with diabetic blood, 
 and steadily increased during the next two hours. In the reverse 
 experiment, where a normal heart was fed with diabetic blood, 
 the consummation of sugar during the first hour was only slightly 
 below normal, but steadily diminished during the succeeding two 
 hours. On adding a boiled extract of pancreas to the blood circu- 
 lating through the heart of a diabetic animal, the sugar consump- 
 tion was raised to a point apjDroximating to that obtained with 
 normal hearts. These experiments, therefore, suggest that the 
 tissues and blood normally contain a substance derived from the 
 pancreas, which is essential for the direct utilisation of sugar by 
 the tissues. This substance, it would seem, is gradually used up 
 in the process, and has to be continually replaced from the blood 
 if the utilisation of sugar is to continue. 
 
 Cohnheim has stated that expressed muscle juice is inactive 
 to sugar until it has been mixed with expressed tissue juice from 
 the pancreas, which itself has only slight glycolytic properties. 
 The jjroducts of the glycolysis were found to be carbon dioxide and 
 water, if an abundance of oxygen was present ; but in the absence 
 of oxygen, alcohol, then lactic acid, and finally oxybutyric acid are 
 formed. He maintains that enough sugar destruction takes place 
 in such experiments to fully account for the amount that is daily 
 destroyed in the body. Sehrt confirms these results, and states 
 that the combined pancreas-muscle, or pancreas-liver, extract does 
 not destroy levulose as it does dextrose, thus accounting for the 
 known ability of depancreatised animals to utilise levulose. Cohn- 
 heim explains the results he obtained on the lines of Ehrlich's side- 
 chain theory. He argues that the muscles jjroduce a ferment, 
 which is itself incapable of decomposing sugar, but that, when 
 acted on by an " activator substance " derived from the pancreas, 
 it gains this power, in much the same way as complement is acti- 
 vated by amboceptor in haemolysis and similar processes. The 
 presence of blood in the muscle was found to bring about glycolysis 
 without the addition of pancreatic extract, suggesting that the 
 activator substance from the pancreas is j)resent in the blood. 
 Cohnheim states that a very small quantity of the activator sub- 
 stance is required, and that an excess interferes with its effects 
 (c/. Ehrlich's " deviation of complement "). It is soluble in water 
 and alcohol, but is insoluble in ether, so that an ether jorecipitate 
 of the pancreatic extract may be used in the experiments. Since 
 it is not destroyed by boiling he concludes that it is not a ferment, 
 but is analogous in composition to adrenalin, iodothyrin, secretin, 
 and other known products of internal secretion. Cohnheim's 
 
EXPERIMENTAL GLYCOSURIA 137 
 
 conclusions have been supported b}^ Arnheim and Rosenbaum. 
 Hirsch, De Witt, and Hall, but have been criticised by Claus and 
 Enibden, and by Simpson, who doubt their value, and attribute 
 some at least of the observed effects to contamination. 
 
 Even if Cohnheim's hypothesis were proved to be correct, and 
 the decreased glycolysis met with as the result of pancreatectomy 
 could be thus explained, it would not account for all the patho- 
 logical changes that follow removal of the pancreas in animals, 
 and particularly the failure of the liver and muscles to store glycogen 
 and the increased amount of glycogen met wdth in abnormal situa- 
 tions. Von Noorden considers that the pancreas furnishes a fer- 
 ment which favours the polymerisation of sugar into glycogen, or 
 ■else an anti-ferment which prevents the too rapid destruction of 
 glycogen. On account of the power which clepancreatised dogs 
 possess of forming glycogen from le\T.ilose, he inclines to the view 
 that it is rather faulty formation of glycogen from grape-sugar, than 
 its too rapid destruction that explains the defective glycogenesis. 
 Wells suggests that either the glycolytic ferments or the glycogen 
 are normally so combined in the liver cells that they cannot freely 
 act, or be acted on, to form sugar, but that in the absence of the 
 internal secretion of the pancreas this combination ceases. 
 
 Pavy advanced the view that the pancreas supplies a co-ferment 
 or amboceptor which, by effecting the attachment of sugar to the 
 bioplasmic molecule, places it in a position to be disposed of in 
 accordance with the requirements of the existing environment. 
 Normally it undergoes oxidation, and disappears with the libera- 
 tion of energy, or it may be transmuted into glycogen, or trans- 
 formed into fat. If, however, from absence, or disease, of the 
 pancreas, the co-ferment is not produced the sugar molecule passes 
 into the circulation in a free state, and glycosuria, proportional to 
 the carbohj^drate intake, results. The carbohydrate which has 
 already been put into combination may also, under the influence 
 of the abnormal environment, be split off, a disruption, similar to 
 that which occurs in phloridzin glycosuria, taking place, so that 
 the built-up molecule undergoes a series of changes the reverse of 
 those that are naturally met with, and as a result a further amount 
 of sugar appears in the urine. 
 
 An attempt to reconcile the discoveries of pancreatic glycosuria 
 with the hepato-neurogenic theory of diabetes of C. Bernard was 
 made by Chauveaux and Kaufmann. The source of the sugar 
 according to their view is the liver. Normally the production of 
 sugar by the liver is under the regulating influence of the nervous 
 system and the pancreas, the nerves passing to the hver carr3'ing 
 
138 GLYCOSURIA 
 
 stimulating'influences, while the pancreas has an inhibitory actiom 
 on the formation of sugar. Originally they held the view that th& 
 influence of the pancreas was also exerted through the nervous 
 system, but later altered this and suggested that an internal 
 secretion was formed by the pancreas, and carried by the blood 
 to the liver. When the pancreas is extirpated its inhibitory action, 
 in either case ceases, with the result that there is an over-produc- 
 tion of sugar by the liver, and a consequent hyperglycaemia and 
 glycosuria. 
 
 Boruttau has suggested that the adrenals produce a hormone 
 which sets in motion glycogeny in the liver, while the pancreas- 
 furnishes another hormone in its internal secretion which an- 
 tagonises the sugar-mobilising power of the adrenals. 
 
 Practically all observers of weight are now of opinion that 
 the pancreas influences carbohydrate metabolism by the formation 
 of an internal secretion, and all the available evidence points 
 strongly in that direction, but, as we have just seen, there is con- 
 siderable difference of opinion as to how it exerts its influence. 
 
 Another question which has not yet been settled, and on whick 
 there has been much controversy, is the exact seat of origin of the 
 internal secretion, some hold that it is formed by the islands, or 
 areas, of Langerhans, while others maintain that it is a product 
 of the Avhole gland. 
 
 The suggestion that the islands of Langerhans are concerned 
 in the production of the internal secretion of the pancreas wa& 
 oiigir.ally made by Laguesse in 1893. This view was subsequently 
 adopted by Schafer, Diamare, and others. The theory that such 
 a relationship exists is based partly upon histological grounds, and 
 partly on the results of experimental work, but the most important 
 evidence in support of it has been furnished by pathological ob- 
 servations which suggest that pancreatic diabetes in man is due ta 
 a disturbance of the functions of the cell-islets. 
 
 Langerhans, in his description of the pancreas in 1869, was 
 the first to draw attention to these characteristic structures now 
 known as " intertubular cell-clumps," " interacinar islands," or 
 " islands, or areas, of Langerhans." They consist of collections of 
 small spherical or polygonal cells, which in man are apparently 
 scattered irregularly through the gland substance, but in some 
 animals occupy a definite position in the centre of the lobules. It 
 is generally held that in adult life no connection exists between 
 the cell islands and the duct system of the gland, but they are in- 
 timately related to the blood-vessels. The structure and relations 
 of the cell islands has been the subject of numerous researches on 
 
EXPERIMENTAL GLYCOSURIA 139- 
 
 the part of a large number of investigators, who, while agreeing 
 on some points, differ in their description in many important 
 particulars. All those who have devoted attention to the subject 
 agree that very similar structures are found in all vertebrates, but, 
 while some regard them as permanent bodies endowed with special 
 functions, others look upon them as being of a temporary character, 
 and consider that they are in reality resting acini. The cells 
 composing the islands of Langerhans are always smaller than the 
 gland cells, each possesses a centrally placed round, or oval, vesicular 
 nucleus, which differs from those of the secreting cells in being 
 larger relative to the amount of cell proto]Dlasm. The protoplasm 
 itself is very finely granular, does not stain at all with basic nuclear 
 dyes, and not well with acid stains, such as eosin. Lane has de- 
 scribed two types of cell, containing granules of different char- 
 acters, both of which differ in chemical nature from the zymogen 
 granules of the acinar cells. The islands are highly vascular, the 
 veins and capillaries forming a rich plexus of thin-walled vessels 
 (sinusoids) in intimate relation with the epithelial cells. The out- 
 line of the cell-islands is irregular, as De Witt has shown in her 
 reconstruction models. The digestion experiments of Flint have 
 demonstrated a well-defined capsule, connected on the one hand 
 with the alevolar framework, and on the other with septa which 
 stretch across the space mthin the island, dividing it into lacunae, 
 and acting as a support for the cells of which it is composed. Ac- 
 cording to Flint, the connective tissue forming the framework has 
 a characteristic arrangement, in sharp contrast to that of the 
 remainder of the lobule. The size and distribution of the islands 
 is not uniform. Laguesse has distinguished five different types 
 varying from less than 100 fi in diameter to a giant form of over 
 400 /A in diameter. Opie found that the cell-islets were more 
 numerous in the tail, or splenic end, than elsewhere in the human 
 pancreas, and he agrees with Kasahara that the pancreatic tissue 
 in the foetus, and very young children, shows a larger number of 
 islands than that in the adult. This might be explained by the 
 islets being formed during embryological life and persisting un- 
 changed, while the secreting tissue increases in bulk. 
 
 The distribution of the islands in certain bony fishes is of con- 
 siderable interest and importance as bearing on the question of 
 their origin and significance. Rennie states that very large islands 
 are found in the areas of pancreatic tissue along the abdominal 
 vessels in all of the twenty-five species he investigated, and the 
 pancreas itself possesses no islets. In the pancreas of the guinea- 
 pig De Witt has described large, relatively isolated, cell-islets lying 
 
140 GLYCOSURIA 
 
 in the connective tissue around the large ducts, especially about 
 the junction of the splenic and middle thirds of the gland, and states 
 that cell-islets, which appear to be free from pancreatic tissue, are 
 also met with in the mesenteric fat near the periphery of the gland. 
 The splenic portion of the jDancreas of the cat, according to Opie, 
 has constantly a cell-islet near the centre of each lobule. By 
 means of what are claimed to be specific staining reactions, Bensley 
 appears to have proved that the islands of Langerhans are special 
 structures and not merely changed acinous tissue. He states that 
 the total number of islands in the entire pancreas of the guinea-pig 
 may vary between 15,000 and 45,000, the average number per 
 milligram of pancreas working out in different animals at 9-5 to 
 189, and that an equally great variation in the number of islets 
 exists between corresjjonding areas in different glands. 
 
 Light on the vexed question of the significance of these re- 
 markable structures has been sought by a study of their develop- 
 ment, but here again there is a difference of opinion. Hansemann 
 believes that they arise from the interstitial tissue, and have no 
 connection with, the pancreatic acini. Laguesse described a double 
 origin, the " primary islands," arising from the primitive pan- 
 creatic tubules, bemg permanent, and " secondary islands," de- 
 veloping from the acini, being transitory structures. Kiister states 
 that the cell-islets are derived from the ducts, and Helly believes 
 that they are developed from the pancreatic tubules, early in 
 embryological life, as solid outgroAA'ths which subsequently become 
 vascularised. and later develop a reticulum. 
 
 ExjDerimental investigations have been conducted by a number 
 of observers with a view to deciding the question of the individuahty 
 of the islands of Langerhans. Lewaschew maintained that stimu- 
 lation of the pancreas by overfeeding, or by the administration of 
 joilocarpine. causes the transformation of secreting acini into cell- 
 islets, but Statkewitsch states that the changes observed are merely 
 the results of intense activation in the gland cells, and are not 
 stages in a transition to cell-islets. Jarotsky also came to the 
 conclusion that the altered acini met with as the result of changed 
 dietetic conditions are not connected with the islands of Langerhans. 
 He attributes the results obtained by Lewaschew to imperfect 
 fixation. Opie could detect no increase in the cell-islets after the 
 administration of pilocarpine, and points out that in Lewaschew's 
 experiments the normal variation in the number of cell-islets in 
 different parts of the gland was not taken into account. De Witt, 
 after studying the effects of starvation and various diets on the 
 pancreas, came to the conclusion that, while some quahtative 
 
EXPERIMENTAL GLYCOSURIA 141 
 
 changes are produced in the cell-islets, there are none that can 
 be regarded as constant, or show a progressive increase with the 
 duration of the experiment. Dale, and subsequently Laguesse, 
 claim to have demonstrated that starvation increases the number 
 of cell-islets, and Dale also states that the injection of secretin 
 exhausts the acini, thereby converting many of them into islands 
 of Langerhans. Vincent and Thompson, who have rejDeated 
 Dale's experiments, confirm his conclusions. Schulze, Sauerbeck, 
 Ssobolew, Zuntz and Mayer, De Witt, MacCallum, and Lombroso 
 found that after ligature of the pancreas, or its ducts, the 
 glandular acini degenerate, but that the islands of Langerhans 
 are left intact. The experiments carried out by Schulze were re- 
 peated by Mankowski, who states that the pancreatic tissue dis- 
 appears from between, and behind, two ligatures placed round the 
 splenic end of the pancreas, while the cirrhosed portion in front 
 shows as many, if not more, cell-islets than secreting acini, so that 
 although the last-named result might suggest that the cell-islets 
 are independent and more resistent structures than the acini, the 
 same cannot be said of the remaining parts. It has been sug- 
 gested by Hess that failure to permanently occlude all the ducts 
 is the explanation of the persistence of groups of cells resembling 
 cell -islets described by previous experimenters. Bearing this in 
 mind, Pratt, Lamson, and Marks carried out a series of experiments 
 with cats and dogs, and state that degeneration and destruction 
 of both the acini and cell-islets, in constant and striking contrast 
 to the results obtained by other investigators, occurs when the 
 pancreatic secretion is completely excluded from the intestine by 
 ligature of the ducts. 
 
 With so many conflicting results, and opinions, it is ob\dous 
 that the question of the independence of the islands of Langer- 
 hans, and still more of their performing a special function, is as 
 yet in an undecided condition, but from a review of the recent 
 literature the impression is obtained that the case in favour of the 
 cell-islets being separate tissues, with an independent function, is 
 steadily growing stronger. In favour of this hypothesis are the 
 constant presence of these structures at all ages, and in so many 
 different animals, their early appearance in embryonic life, their 
 different staining reactions, their peculiar arrangement, and their 
 apparent independence of the duct system of the gland ; while 
 their rich blood supply may be taken to indicate that they are 
 possibly vascular glands engaged in the elaboration of an internal 
 secretion, which is poured into the blood stream. It was their 
 close resemblance to the para-thyroids and other ductless glands 
 
142 GLYCOSURIA 
 
 that suggested to Laguesse that they were bodies concerned solely 
 A^dth the elaboration of an internal secretion. Those who beheve 
 that the islands of Langerhans are temporarily changed secreting 
 acini, hold that the islets are closely related to the acini, from 
 which, they contend, the}^ are not separated by any definite cap- 
 sule ; both structures have a common blood supply ; the islets 
 •can be shown to open into the jDancreatic ducts ; in some sections 
 various transition stages between typical acini and typical islets 
 ■can be seen ; the number of islets is increased during activity of 
 the gland, and diminished during rest ; by jDrolonged stimulation 
 of the gland, either by overfeeding, by the administration of pilo- 
 carpine, or by the injection of secretin, it is possible to transform 
 secreting acini into islands of Langerhans. Observers who favour 
 this view maintain that the internal secretion is furnished by the 
 acini, and that j)ancreatic diabetes is due to disease of those 
 structures. 
 
 The most strenuous opponent of the theory that the glycos- 
 uria produced bs^ extirpation of the pancreas is dependent 
 upon the lack of an internal secretion formed by the gland has 
 been Pfliiger, who asserts that the pancreas is under the control 
 of the nervous system, and that in the wall of the duodenum there 
 exists an " anti-diabetic " centre, rich in ganglion cells, which, by 
 nerves passing to the gland, controls its anti-diabetic powers, so 
 that the pancreatic diabetes of von Mering and Minkowski is in 
 jeality a " duodenal " diabetes, due to destruction of these nerves. 
 Ex23erimenting with frogs, Pfliiger found that extirpation of that 
 portion of the duodenum which is in contact Avith the pancreas was 
 followed by severe diabetes, and that even cutting the nerves 
 which pass from the duodenum to the pancreas produced the 
 same effect. His experiments with dogs were not quite so satis- 
 factory, owing, he states, to the difficulty of the operation and 
 the speedy death of the animal ; but even with these Pfliiger 
 found that excision of the duodenum produced glycosuria. His 
 conclusions with regard to dogs and other mammals have been dis- 
 proved by Ehrman, LauAvens, Rosenberg, Minkowski, and Pratt, 
 who showed that total removal of the duodenum, or section of 
 the nerves passing from the duodenum to the pancreas, does not 
 give rise to glycosuria. Rosenberg and Lowitt have also demon- 
 strated that " duodenal " diabetes does not exist in the frog, and 
 that the glycosuria observed by Pfliiger was x^robably produced 
 by keeping his animals on ice. 
 
 According to Ott, the injection of fresh j)ancreas into the jugular 
 vein of animals produces slight glycosuria, about a quarter per 
 
EXPERIMENTAL GLYCOSURIA 143 
 
 cent, of sugar usually appearing in the urine. Leschke states that 
 pancreatic extracts derived from frogs or guinea-pigs, whether 
 fresh, inactivated by heating to 70° C. for a quarter of an hour, or 
 after being heated to 100° C. for ten minutes, when injected sub- 
 cutaneously, or intravenouslj^, into normal guinea-pigs or frogs, 
 give rise to glycosuria. When injected into animals with pan- 
 creatic diabetes they increase the glycosuria, at the same time 
 occasioning a severe disturbance of the general condition, which 
 shortly ends in death. 
 
 In spite of the conflicting results of many of the observations 
 on the relation of the pancreas to glycosuria, and the existence 
 of different theories as to the way in which the gland controls 
 the internal carbohydrate metabolism of the body, certain con- 
 clusions stand out as being generally accepted. These may be 
 summarised as follows : — 
 
 (1) Total extirpation of the pancreas always gives rise to fatal 
 
 diabetes. 
 
 (2) The presence of at least a fourth or fifth of the gland in 
 
 a healthy condition prevents glycosuria. 
 
 (3) Transplantation of a portion of the pancreas prevents the 
 
 development of diabetes when the rest of the gland is 
 removed. 
 
 (4) The conclusion is justified that the pancreas is concerned 
 
 with the internal metabolism of sugar. 
 Exactly how it exerts its metabolic functions, the relation of 
 the islands of Langerhans to the acini, and the relative importance 
 of the islands and the acini in carbohydrate metabolism, are ques- 
 tions that are as yet unsettled, and await more conclusive evidence 
 for their solution. 
 
 {b) The Supra-renals. — In 1901 Blum announced that the 
 subcutaneous injection of an aqueous extract of the supra -renal 
 capsules produces transitory glycosuria in dogs. Subsequently 
 Bierry and Gatin-Gruzewska, Patta, Lepine, and others confirmed 
 his results, and showed that intravenous, or intra-peritoneal, injec- 
 tions of extracts of the glands, or of adrenalin chloride, give a 
 more marked result, and also more quickly. Massage of the ex- 
 posed supra-renal capsules was found by Herter to produce glycos- 
 uria, while most minute quantities of adrenalin appHed cUrecth^ 
 to the surface of the pancreas cause the appearance of sugar in 
 the urine. The dose of epinephrin given in all exjoerimental work 
 must, however, be enormous compared with the amount normally 
 poured into the blood, and Ritzmann states that it is only when 
 
144 GLYCOSURIA 
 
 the quantity of epinephrin administered is sufficient to produce a, 
 rise in blood pressure that glj^cosuria results. When a sufficiently 
 dilute solution (1,000,000 to 2,000,000) is used, it can be allowed 
 to flow into a vein at the rate of two cubic centimetres a minute 
 without sugar appearing in the urine ; but if the solution is more 
 concentrated, or the dilute solution is run in faster, glj^cosuria 
 follows. 
 
 The injection of adrenalin greatlj' intensifies the glycosuria in 
 depancreatJsed dogs, and it has been shown by Fronin and Maj'er 
 that, on the other hand, the extirpation of the adrenals prevents 
 the glycosuria that would otherwise result from removal of the 
 pancreas. According to Mayer, extirpation of the adrenals also 
 prevents the appearance of sugar after diabetic i^uncture. Porges 
 has shown that, after the extirpation of both supra-renals in dogs, 
 only a small amount of glycogen is present in the liver and muscles, 
 and that the quantit}^ of sugar in the blood is lowered. Falta 
 states that high section of the spinal cord, or section of the nervous 
 connections of the adrenals with the diabetic centre, leads to a 
 very marked reduction in the amount of sugar in the blood. Pollak 
 found that the administration of adrenalin causes glycosuria when 
 both splanchnics have been cut. 
 
 The blood of animals with glycosuria produced by adrenalin 
 injections has been examined by Zuelzer and Metger, who found 
 that it contains an excess of sugar, so that the immediate cause 
 of the appearance of sugar in the urine, as in pancreatic and punc- 
 ture diabetes, is a hyperglycsemia. Since the most striking lohysio- 
 logical action of adrenalin is its vaso-constricting power, it would 
 seem not unlikely that the hyperglycsemia and glycosuria might 
 result from its effect on the circulation, the pancreas, &c., being 
 rendered anaemic, and thus functionally inactive. Marked vaso- 
 constriction may, however, be induced in guinea-pigs by the ad- 
 ministration of supra-renal extract, without glycosuria resulting, 
 and, as we have seen, the injection of adrenahn intensifies the 
 glycosuria of depancreatised animals. The very marked effect 
 of the direct application of adrenahn to the pancreas has also 
 suggested that it may act directly upon the gland and interfere 
 with its control of sugar metabolism ; but here again the results 
 of injections into animals dex^rived of their pancreas seemed to- 
 negative such an idea, at least in its entirety. It was shown by 
 Iwanoff that adrenalin increases the rate of discharge of sugar 
 from the glycogen-rich liver, through which salt solution is being 
 transfused, suggesting that the internal secretion of the supra-renals 
 acts directly upon the glycogen splitting enzyme of the hver cells ;. 
 
EXPERIMENTAL GLYCOSURIA 145 
 
 but starved rabbits, from whose tissues most of the glycogen has 
 been removed by repeated doses of strychnine, have been sho'ttTi 
 by Pollak to react to adrenahn, so that an increased production 
 of dextrose from glycogen cannot be the sole explanation of the 
 hyiDerglycsemia and glycosuria. Underbill found that piperidine 
 produces similar results to adrenalin, but the glycosuria induced 
 by piperidine and certain other chemicals, such as potassium 
 cyanide, ether, chloroform, morphine, strychnine, curare, &c., 
 differs from that produced by adrenalin, inasmuch as the former 
 is prevented by the administration of oxygen, while the latter is 
 not. Elliott has pointed out that a characteristic of epinephrin 
 is its tendency to stimulate plain muscle and gland cells that are, 
 or have been, in functional union with sympathetic nerve fibres. 
 In harmony with this fact Underhill and Clossen have suggested 
 that the internal secretion of the supra-renals acts upon the liver 
 and other storehouses of carbohydrates through the sjmipathetic 
 nervous system, causing them to throw out the carbohydrate they 
 contain in the form of dextrose, while at the same time there is 
 diminished storage from stimulation of the hepatic cells, which 
 reject the dextrose brought to them by the blood. From the 
 similarity of the effects produced by adrenalin injections and 
 diabetic puncture, and the increase in the blood of adrenalin that 
 the latter is said to produce, Schur and Wiesel have inferred that 
 the sympathetic fibres actuate the formation of sugar in the liver 
 through the adrenals. 
 
 Zuelzer, Dohran, and Marxer have described extensive experi- 
 mental researches which seem to show that adrenalin plays an 
 active part in the mobilisation of sugar both in physiological con- 
 ditions and in pancreatic diabetes. They conclude that adre- 
 nalin, and the hypothetical internal secretion of the X3ancreas, have 
 an antagonistic action, and that it is by their combined effect on 
 the liver that sugar metabolism is regulated. In pancreatic diabetes 
 the lack of the pancreatic secretion, and the consequent predomi- 
 nance of adrenalin, explains the increased output of sugar. When 
 both the pancreas and the supra-renals are incapable of functioning, 
 diabetes does not occur. Zuelzer states that he has extirpated the 
 pancreas from more than a hundred animals, and that pronounced 
 glycosuria invariably followed, excepting in those cases where 
 adrenalin was excluded from the circulation by ligaturing the 
 veins from the supra-renals. Underhill and Fine have found that 
 hydrazine prevents the glycosuria occurring after extirpation of the 
 pancreas, probably because this drug diminishes adrenal activity. 
 
 All these researches tend to show that the adrenals exert an 
 
 K 
 
146 GLYCOSURIA 
 
 important influence on carbohydrate metabolism through, their 
 internal secretion. Physiologically their task appears to be to 
 mobilise the sugar from the liver, and probably also from the 
 other tissues, possibly through the intermediation of the sympa- 
 thetic nervous system. That a body of known composition, such 
 as adrenalin is, should j^erform such an important function is 
 most interesting, and furnishes a noteworthy example of the 
 specific action of an optically inactive substance containing an 
 asymmetrical carbon atom. It \^dll be remembered, however, that 
 adrenalin is the secretion of the medullary, or chromaffin tissue, of 
 the supra-renal capsules only, and that chromaffin tissue is also 
 found in other situations, although in very much smaller amounts. 
 
 (c) Pituitary Gland. — It has long been known that glycosuria 
 is one of the most frequent symptoms of acromegaly, and that in 
 such cases overgrowth of the pituitary gland is found after death. 
 In the group of symptoms known as " Frohlich's syndrome " or 
 " dystrophia adiposo-genitalis," which is believed to result from 
 hypo-pituitarism, on the other hand, glycosuria is almost unknown, 
 although thirst and polyuria are common. 
 
 Experiments on animals have pointed to the hypophysis as 
 being concerned in the metabolism of sugar. After destroying the 
 posterior, without touching the anterior, lobe in a dog, Caselli 
 found that sugar appeared in the urine. Polyuria and sometimes 
 glycosuria occur in animals after partial removal of the anterior, 
 and total removal of the posterior, lobe according to Gushing. 
 It has been shoAvn by Borchardt that the hjrpodermic injection of 
 a boiled, filtered infusion of the gland into rabbits and dogs causes 
 a small amount of sugar to appear in the urine, varying from a 
 trace to 4-2 23er cent. The sugar usually began to appear in the 
 urine about three hours after the injection, but the total amount 
 excreted was very small, generally not more than a centigramme, and 
 was extraordinarily variable. Glycosuria was found to be more 
 readily produced in rabbits than in dogs, but in both the amount 
 of sugar excreted appeared to be independent of the quantity of 
 gland extract injected. Ott and vScott injected 1 c.c. of a 20 per 
 cent, extract of the pituitary (" Infundibulin ") into the muscles 
 of rabbits, and found, in all cases, at the end of two and a half 
 hours about ^ per cent, of sugar in the urine. Intramuscular, and 
 intraperitoneal, injections into cats produced similar results. Sec- 
 tion of the splanchnin nerves was found to arrest the glycosuria, 
 which they suggest indicates that infundibulin acts on the diabetic 
 centre in the medulla by way of the splanchnics. 
 
EXPERIMENTAL GLYCOSURIA 147 
 
 The relation of the pituitary gland to carbohydrate meta- 
 bolism has been very thoroughly investigated in a series of ex- 
 periments carried out by Goetsch, Gushing, and Jacobson. These 
 authors, like their predecessors, come to the conclusion that it 
 is the posterior lobe {'pars nervosa et intermedia) that is particu- 
 larly concerned with this function. They state the secretion of the 
 posterior lobe is discharged into the cavity of the third ventricle, 
 and becomes dissolved in the cerebro-spinal fluid, a medium 
 which passes from the ventricles to the subarachnoid spaces, and 
 thence, in all probability, enters the blood-stream by way of the 
 dural sinuses. Under various forms of operative manipulation of 
 the infundibulum and hypophyseal stalk-structures, which appear 
 to hold the reserve deposit of posterior lobe secretion, a transient 
 hyperglycsemia is produced, presumably due to the setting free 
 of an excess of this secretion, M'hich in turn causes the discharge 
 of stored glycogen. For the succeeding few days the assimilation 
 limit for ingested carbohydrates is considerably diminished, ali- 
 mentary glycosuria being produced by a smaller amount of sugar 
 than was previously required. If the operation has been so con- 
 ducted as to create a subsequent and jjermanent insufficiency of 
 posterior lobe secretion (either owing to the removal of a con- 
 siderable portion of this lobe with its epithelial investment, or 
 thorough interference with its secretory discharge, either by placing 
 a " clip " on the stalk, or by so damaging it that an infundibular 
 cicatrix forms), the temporary lowering of the assimilation limit is 
 succeeded by an abnormal, and enduring, augmentation in the 
 tolerance for sugars. The assimilation limit for carbohydrates, 
 greatly increased under these circumstances, can be promptly 
 lowered by the coincident intravenous or subcutaneous injection 
 of posterior lobe extract. This extract, furthermore, has a pro- 
 nounced effect in lowering the sugar tolerance of the normal animal, 
 in whom it may even cause glycosuria when given in sufficiently 
 large doses. 
 
 id) The Thyroid.— Experimental and clinical evidence both 
 suggest that the thyroid gland is an essential factor in the regula- 
 tion of the metabolism of the body, and that in some way it is 
 connected with the utilisation of carbohydrates. 
 
 McCurdy, Falta, and others have found that extirj^ation of the 
 thyroid raises the tolerance of animals for sugar, so that alimentary 
 glycosuria is produced with difficulty, ]3rovided that the para- 
 thyroids are i^reserved. The administration of adrenalin is said 
 by Falta not to produce glycosuria in thyroid-ectomised dogs, 
 
148 GLYCOSURIA 
 
 although Underhill found that, if a sufficiently large dose is given, 
 sugar may appear in the urine. MacCallum showed that the 
 glycosuria following jpancreatectomy is diminished by extirpation 
 of the thjToid. After extirpating the thyroid in etherised cats, 
 Iea%'ing two or more parathyroids, Ott and Scott found sugar in 
 the urine on the following day in a few cases. On injecting in- 
 fundibuUn mto the jugular vein the amount of sugar was diminished, 
 as compared with animals whose thyroid was intact, the injection 
 causing the appearance of 3 to 4 per cent, of sugar in the urine 
 before removal of the thyroid, and 1 to 2 per cent, after. King 
 states that the addition of thjTToid juice to the Cohnheim pancreas- 
 muscle-dextrose mixture markedly diminishes glycolysis. 
 
 A microscopical examination of the thyroids of three dogs in 
 which chronic pancreatic insufficiency had existed for from five 
 to thirty-four months, by Pratt, showed a partial or complete 
 replacement of the colloid material by desquamating epithehal 
 cells, so that the structure of the gland was scarcely recognisable, 
 suggesting a compensatory activity of that organ. 
 
 The influence of the thjnroid on carbohydrate metabolism has 
 also been noted in clinical medicine. In exophthalmic goitre 
 (hyperthyroidism) glycosuria is often present, and even if it is 
 not, it is readily induced by feeding with large doses of sugar. 
 The continued use of thyroid preparations has also occasionally 
 been followed by diabetes. In myx oedema (hypothyroidism), on 
 the other hand, glycosuria is practically never met with, and even 
 large doses of sugar can be taken without any being excreted in 
 the urine. 
 
 How the thyroid influences carbohydrate metabolism is not 
 yet decided. Opie inchnes to the view that the glycosuria of 
 Grave's disease is due to an associated pancreatic lesion, and such 
 lesions have been found in fatal cases (Cecil). The fact that 
 thyroid administration may produce glycosuria can, however, be 
 only reconciled with such an explanation by supposing that hyper- 
 thjToidism gives rise to changes in the pancreas. Falta has sug- 
 gested that an excess of thjnroid secretion exercises a check upon 
 the functions of the pancreas, either directly, or by stimulating the 
 activity of the chromaffin system — that is to say, of tissues such 
 as the medulla of the supra-renals, &c., which contain cells that 
 readily stain with chromic acid and its salts. In favour of the 
 latter suggestion are the observations of Kraus and Friedenthal, 
 and of Frankel, who detected an increase of adrenalin in the blood 
 of patients with Grave's disease. Falta has also extended the 
 observations of Schmidt and Salomon on the faeces in exoph- 
 
EXPERIMENTAL GLYCOSURIA 149 
 
 thalmic goitre, and points out that the occurrence of fatty stools 
 suggests a coincident restriction of the external secretion of the pan- 
 creas. If hyperthyroidism tends to induce glycosuria by exercising 
 a restraining influence upon the functions of the pancreas, rather 
 than by causing lesions of the gland, we should expect to find that 
 a withdrawal of the normal supply would have the opposite effect, 
 and that the pancreas, relieved of its wonted restraint, w^ould be 
 unusually active as regards carbohydrate metabolism. That is 
 to say, hypothyroidism would increase the tolerance for carbo- 
 hydrates, and this, as we have seen, is the case. Moreover, it has 
 been found that the administration of thyroid extract lowers the 
 tolerance again. There is thus a very striking resemblance between 
 the effects of alterations in the functions of the thyroid and the 
 posterior lobe of the pituitary body, and it seems probable that 
 the mechanism is in both cases the same, or similar, especially as 
 it is known that removal of the thyroid leads to hypertroi^hy of 
 the pituitary. 
 
 (e) The Parathyroids. — The parathyroids appear to have an 
 opposite influence to the thyroid on carbohydrate metabohsm 
 (Eppinger, Falta, and Ruclinger). Underbill and Hilditch found 
 that, after the excision of three of the four parathyroids of a dog, 
 the tolerance of the animal for sugar Avas greatly lowered, and 
 that complete thyreo-parathyroidectomy, besides causing tetany, 
 &c., gave rise to glycosuria. Ott states that after the injection 
 of the nucleo-protein of the parathjrroids into a vein, sugar appears 
 in the urine, and that large doses of parathyroid subcutaneously 
 produce glycosuria in rabbits. 
 
 Theory of the Correlation of the Ductless Glands.— It 
 
 has been pointed out by Eppinger, Falta, and Ru dinger, who have 
 themselves made numerous experiments and added much to our 
 knowledge of the functions of the ductless glands, that the removal 
 of a gland with an internal secretion acts in two ways. There is, 
 first, the direct result of the absence of its secretion, and, secondly, 
 the indirect effect due to the disturbance of metabolism conse- 
 quent on its relations to other glands. They consider that the 
 ductless glands may be arranged in two groups, according to the 
 disturbance in the metabolic functions of the body to which altera- 
 tions in their secretory activities may give rise. Broadly, they 
 may be divided into (1) an accelerator group and (2) an inhibitory 
 group. To the former class belong the thyroid, the pituitary body, 
 and the supra-renals, since a study of the reactions of the body 
 to injections of preparations from these organs, and observations 
 
150 
 
 GLYCOSURIA 
 
 on the results of experimental interference with their functions, 
 suggest that all three increase protein exchange, the adrenals 
 cause mobilisation of carbohydrates, and the thyroid causes in- 
 creased fat destruction. Into the second group fall the pancreas 
 and the j^arathyroids, since both retard protein destruction, the 
 pancreas being the more active, and both restrain the mobilisation 
 of carbohydrates ; here again the pancreas is more active. The 
 pancreas also causes a decrease in fat consumption. These ob- 
 servers hold that the internal secretions of the pancreas and thyroid 
 mutually inhibit each other's activities, and that there is a similar 
 mutual inhibitory action between the pancreas and the chromaffin 
 system of the body, while between the thyroid and the chromaffin 
 system there is reciprocal promotion of action. 
 
 This inter-relation may be schematically represented by a 
 quadrilateral figure, with the pancreas, th3n:oid, parathjTcids, and 
 chromaffin tissue at the four angles : — 
 
 THYROID. 
 
 PANCREAS 
 
 PARATHYROIDS. 
 
 CHROMAFFIN TISSUE. 
 
 It is believed that all the ductless glands act through the 
 discharge of specific hormones into the circulation, and that in the 
 case of epinephrin it has a special affinity for the sympathetic 
 nerves. The mobilisation of sugar in the liver is assumed to be 
 under the influence of the sympathetics, by mediation of the supra- 
 renals, while the pancreatic internal secretion has an opposite effect 
 to epinephrin, inhibiting the mobilisation of sugar in the liver 
 and other tissues. The action of the two secretions is thus anta- 
 gonistic. The secretion of the thyroid is believed to inhibit the 
 activity of the pancreas and to promote that of the chromaffin 
 system in carbohydrate metabolism, the inhibition being, however, 
 stronger than the promotion. 
 
 According to this view the glycosuria seen after extirpation 
 of the j)ancreas may be looked upon as a negative pancreatic 
 diabetes and a positive adrenal diabetes, for the normal inhibitory 
 action of the internal secretion of the pancreas is removed, while 
 
EXPERIMENTAL GLYCOSURIA 151 
 
 the mobilising power of the adrenal secretion is increased by 
 hyperfunction of the chromaffin tissue. The latter results partly 
 from the removal of the inhibitory action that the pancreas nor- 
 mally exerts on the adrenals, &c., and partly as an indirect effect 
 of the absence of the similar inhibitory action that the pancreas 
 has on the thyroid, which is thus at liberty to exert its stimulating 
 action on the chromaffin tissue unchecked. At the same time 
 there is excessive metabolism of proteids, fats, and carbohydrates 
 as a consequence of the hyperfunction of the thjToicl. This 
 explanation would account for the greater intensity of pancreatic 
 as compared with other forms of experimental diabetes. 
 
 The glycosuria foUomng the injection of adrenalin can be 
 regarded as a chrect result of the rapid and excessive mobilisation 
 of carbohydrates arising from hyperfunction of the chromafiin 
 tissue, and indirectly as a result of an increased inhibition of the 
 functions of the pancreas. O^vdng to the simultaneous promo- 
 tion of the action of the thyroid that occurs, there is also an in- 
 creased metabolism of proteids and fats. If the pancreas has been 
 previously removed this is more marked, and the dextrose to 
 nitrogen ratio in the urine is raised. 
 
 Removal of the th}Toid leads to hyperfunction of the pancreas, 
 owing to the removal of its inhibitory action on that gland, while 
 at the same time it diminishes adrenal action, from the absence 
 of the promoting effect of the thjToid on the chromaffin tissue. 
 Hyperthyroidism produces the opposite effects, causing relative 
 pancreatic insufficiency, and increased adrenal activity, with a 
 tendency to glycosuria. 
 
 Extirpation of both the thyroid and pancreas removes at the 
 same time the inhibitory effect of the pancreas and the promoting 
 action of the thyroid on the chromaffin tissue, but as, according 
 to Falta, the former is stronger than the latter, the mobihsation 
 of fats is increased ; omng to the absence of the thyroid, however, 
 their destruction is decreased. Under these circumstances sugar 
 is formed from fat, and there is a raised dextrose to nitrogen ratio 
 in the urine. 
 
 The glycosuria following puncture of the floor of the fourth 
 ventricle is brought into line by the explanation, to which reference 
 has been made, that the increased output of sugar by the hver 
 on which it depends, arises from an excessive flow of epinephrin, 
 brought about by impulses transmitted to the supra-renal capsules 
 through the left sympathetic nerve. 
 
 It is possible that a number of toxic influences also cause the 
 appearance of sugar in the urine by their action on the supra- 
 
152 GLYCOSURIA 
 
 renals, either directly, or through the medium of the sympathetic 
 nervous system or the diabetic centre in the medulla. 
 
 The views just outHned are to some extent theoretical, and 
 this is particularly so as regards the part the chromaffin system 
 plays, but they correlate, in a much more satisfactory manner 
 than has previously been possible, the data that have accumulated 
 on experimental glycosuria, with the exception of phloridzin glycos- 
 uria, which appears for many reasons to stand in a class apart. 
 Whether subsequent observation confirms, or in part disproves, 
 these theories there can be no doubt that they open up a much 
 wider conception of the pathology of diabetes, and indicate that 
 glycosuria is not a disease, but a symjDtom common to many 
 pathological conditions. 
 
 BIBLIOGRAPHY 
 
 Achard, Zeit. f. klin. Med., xvi. 
 
 Achard and Delamare, These de Paris, 1899. 
 
 Aldehoff, Zeit. f. Biol, 1892. 
 
 Arnheim and Rosenbaum, Zeit. f. phys. Chem., 1904. 
 
 Asher, Zeit. f. Biol, 1912. 
 
 Bayliss and Starling, Proc. Roy. Soc, cxix., Ixxiii. 
 
 Bensley, Amer. Journ. of Anat., 1911. 
 
 Bernard, Compt. Rend. d. I. Acad. d. Sci., 1849. 
 
 Biel and Kolisch, quoted Pasteau Jr. Med. Francais, 1911. 
 
 Bierry and Gatin-Gruzewski, Compt. Rend. d. I. Soc. d. Biol., 1905. 
 
 Blum, Deut. Arch. f. klin. Med., 1901. 
 
 Blumenthal, Zeit. f. diat. u. physik. Therap., 1898. 
 
 Bock and Hoffman, Reichert u. du Bois Raytnond's Arch., 1871. 
 
 Borchardat, Zeit. f. klin. Med., 1908. 
 
 Boruttau, quoted Ott, Internal Secretions, 1910. 
 
 Bosanquet, Lancet, 1905, 1908. 
 
 Bouchardat, De la Olycos. et Diab. Sucre, 1875. 
 
 Capparelli, Arch. ital. d. Biol., 1894. 
 
 Carnot, Malad. d. Gland. Saliv. et d. Pancreas, 1908. 
 
 Casein, quoted Ott, Internal Secretions, 1910. 
 
 Cecil, Journ. Exp. Med., 1909. 
 
 Chauveaux and Kaufmann, Compt. Rend. d. I. Soc. d. Biol., 1893-4. 
 
 Chittenden and Lambert, Zeit. f. Biol., 1889. 
 
 Claus and Embden, Beitr. z. chem. Physiol, u. Path., 1905. 
 
 Cohnheim, Zeit. f. physiol. Chem., 1903-4-6. 
 
 Crofton, Amer. Journ. Med. Sci., 1902. 
 
 Dale, Proc. Roy. Soc, 1904. 
 
 Diamare, Internat. Monat. f. Anat. u. Path., 1899. 
 
 Ehrmann, Arch. f. d. ges. Physiol., 1906 ; Zeit. f. exp. Path., 1908. 
 
 Elliott, Journ. Physiol., 1905. 
 
 Eppinger, Falta, and Rudinger, Zeit. klin. Med., 1908-9. 
 
EXPERIMENTAL GLYCOSURIA 153 
 
 Falta, Wein. klin. Woch., 1907 ; Cong. /. inn. Med., 1907 ; Zeit. f. 
 
 klin. Med., 1910. 
 Fischer, P finger's Arch., 1905. 
 Flint, Arch. f. Anat. u. Entwickl., 1903. 
 Forsbach, Arch. f. exp. Path. u. Pharm., 1909. 
 Fronin and Mayer, Cong. f. inn. Med., 1908. 
 
 Goetsch, Gushing, and Jacobson, Johns Hopkins Hosp. Bidl., 1911. 
 Hall, Amer. Journ. Physiol., xviii. 
 Hansemann, Zeit. f. klin. Med., 1894 ; Verhand. d. deut. 2^(^th. 
 
 Gesellsch., 1902. 
 Harley, Journ. Anat. and Physiol., 1891-2 ; Brit. Med. Journ., 1892. 
 Hedon, Compt. Bend. d. I. Soc. d. Biol., 1911. 
 Helly, Arch. f. mik. Anat., 1905. 
 Herter, Med. News, 1902. 
 
 Hess, Mi'mch. med. Woch., 1902 ; Arch. J. d. ges. Physiol., 1907. 
 Hirsch, Beitr. z. chetn. Physiol, u. Path., 1904. 
 Iwanoff, Centr. f. Physiol., 1905. 
 Jarolzky, Virchow's Arch., 1899. 
 Kasahara, Virchow's Arch., 1896. 
 King, quoted Ott Internal Secretions, 1910. 
 Knowlton and Starling, Lancet, 1912. 
 Kiister, Arch. f. mik. Anat., 1904. 
 Lag-uesse, Cotnpt. Rend. d. I. Soc. d. Biol., 1893, 1910; Journ. VAnat. 
 
 et Physiol., 1894. 
 Lane, Amer. Journ. Anat., 1907. 
 Langerhans, Inaug. Dissert., Berhn, 1869. 
 Lauwens, Arch. f. d. ges. Physiol., 1907. 
 Lepine, Diabtte Sucre, 1909. 
 
 Leschke, Arch. f. Anat. u. Physiol., 1910; Chem. Zentralb., 1911. 
 Lewaschew, Arch. f. mik. Anat., 1886. 
 Lonibroso, ref. Biochein. Centralb., 1903 ; Journ. d. Physiol, et 
 
 Path, gen., 1905. 
 Lowitt, Arch. f. exp. Path. u. Pharm., 1910. 
 Lusk, Science of Nutrition, 1906. 
 MacGalluni, Johns Hopk. Hosp. Bull., 1909 ; Journ. Amer. Med. 
 
 Ass., 1910. 
 MacLeod, Recent Advances in Physiol., 1906 ; Journ. Afner. Med. 
 
 Ass., 1910. 
 Mankowski, Arch. f. mik. Anat., 1901. 
 Mayer, Compt. Rend. d. I. Soc. d. Biol., 1906. 
 Mendel and Lusk, Deut. Arch. f. klin. Med., 1904. 
 Von Mering, Cong. f. inn. Med. 1886 ; Zeit. f. klin. Med., 1888-9. 
 Von Mering and Minkowski, Centralb. f. klin. Med., 1889 ; Arch. f. 
 
 exp. Path., 1890 ; Berl. klin. Woch., 1892. 
 Metzger, Micnch. med. Woch., 1902. 
 Minkowski, Berl. klin. Woch., 1890, 1892 ; Centralb. /. klin. Med., 
 
 1890 ; Arch.f. exp. Path. u. Pharm., 1893, 1908. 
 Von Noorden, Diabetes Mellitus, 1906. 
 Opie, Journ. exp. Med. 1901 ; Diseases of the Pancreas, 1903. 
 
154 GLYCOSURIA 
 
 Ott, Internal Sec7-etwns, 1910. 
 Ott and Scott, Internal Secretions, 1910. 
 Pal, Wien. mecl. Presse, 1898 ; Wiener klin. Woch., 1902. 
 Patta, Arch. ital. d. Biol., 1906. 
 Pavy, Lancet, 1908. 
 Pa^^, Brodie, and Siau, Lancet, 1903. 
 Pearce, Amer. Joiirn. Anat., 1903. 
 Pfliiger, Arch. f. d. ges. Physiol., 1907-8. 
 Pollak, Arch.f. cxj). Path. u. Pharm., 1909. 
 Porges, Berl. klin. Woch., xxv. 
 Pratt, Journ. Amer. Med. Assoc, 1910, 1912. 
 Pratt, Lamson, and INIarks, Tr. Assoc. Amer. Phys., 1909. 
 Eeilley, Xolan, and Lusk, Amer. Journ. Physiol., 1898. 
 Rennie, Quart. Journ. Micros. Sci. 1904; Zeit. f. Physiol, 1905. 
 Richter, Dent. med. Woch., 1899. 
 Ritzmann, Arch.f. exp. Path. u. Pharm., 1909. 
 . Rosenberg, Arch.f. d. ges. Physiol, 1908 ; Biochem. Zentralb., 1909. 
 Sandmeyer, Zeit. f. Biol, 1892-4. 
 
 Sauerbeck, Virchow's Arch., 1904; Centralb. f. path. Anal, 1904. 
 Schafer, Lancet, 189.5 ; Brit. Med. Journ., 1895. 
 Schulze, Arch.f. mik. Anat., 1900. 
 
 Schiu- and Wiesel, quoted Ott Internal Secretions, 1910. 
 Sehrt, Zeit.f. klin. Med., 1905. 
 Simpson, Biochem. Journ., 1910. 
 Ssobolew, Virchow's Arch., 1902. 
 Statke\^-itsch, Arch.f. exp. Path. u. Pharm., 1902. 
 Stiles and Lusk, Amer. Journ. Physiol, 1903. 
 Tuckett, Journ. of Physiol, 1899. 
 Umber, Zeit.f. klin. Med., 1900. 
 Underbill, Journ. Biolog. Chem., 1905-6. 
 Underbill and Clossen, Cent. f. Physiol, 1905. 
 Underbill and Fine, Journ. Biolog. Chem., 1911. 
 Underbill and Hilditcb, Amer. Journ. of Physiol, 1897. 
 Vincent and Tbompson, Internat. Monat. f. Anat. u. Physiol, 1907. 
 Visentini, Arch.f. Anat. u. Physiol, 1908. 
 Weintraud, Arch.f. exp. Path., 1894. 
 Wells, Chem. Pathology, 1907. 
 De Witt, Journ. of Exp. Med., 1906. 
 Zuelzer, Berl klin. Woch., 1901. 
 
 Zuelzer, Dobrn, and Marxer, Deut. med. Woch., xxxiv. 
 Zuntz, Arch. f. Physiol, 1895. 
 Zug and Maj^er, Keen. d. I. Acad. roy. d. 7ned. Belgique, 1906. 
 
CHAPTER V 
 
 ALIMENTARY, TRANSITORY, AND INTERMITTENT GLYCOSURIA 
 
 It is well to remember that glycosuria is in reality but a relative 
 term, and that the presence of sugar in the urine does not neces- 
 sarily denote a pathological condition. The urine of a healthy 
 individual, taking an average amount of exercise and food, con- 
 tains at most a mere trace of sugar, that can only be detected by 
 special methods ; but when a large quantity of sugar, and par- 
 ticularly sugar in solution (for example, sweet wine, &c.), is con- 
 sumed, an amount of dextrose that is demonstrable by ordinary 
 clinical tests may appear in the urine. This may be termed 
 physiological glycosuria, for it is transitory, is not associated with 
 any signs of disease, and is but an exaggeration of the normal 
 condition. 
 
 The limit up to which a healthy person can completely utilise 
 sugar varies with the individual, the concUtions under which it 
 is taken, and the nature of the sugar. Moritz found 0-2 to 0-3 per 
 cent, of sugar in the urines of four out of six healthy people who 
 had taken a large quantity of sweets and sw^eet champagne, but 
 the urines of the other two were sugar-free. A smaller dose of 
 sugar usually gives rise to glycosuria when taken on an empty 
 stomach than when consumed after a full meal. The assimilation 
 limit also varies with the time of day, being higher in the early 
 morning than at night. Muscular work raises the power of assi- 
 milation, thus Bruel found that an individual who excreted 2-14 
 grams of sugar, in four hours after taking 200 grams, when at rest, 
 only passed 0'09 grams after the same amount when doing muscular 
 work. Hofmeister showed that the assimilation limit is lower for 
 the starving organism than for the well-fed one. Young persons 
 can as a rule dispose of a larger amount of sugar without sensibly 
 affecting the urine than adults. Bouchard quotes the case of a 
 boy of seventeen who consumed daily, for over five years, 600 grams 
 of cane-sugar (about 13 grams per kilo of his body- weight) without 
 producing glycosuria. The appearance of alimentary glycosuria 
 seems to be favoured by alcoholic drinks. As a rule sugar begins 
 to appear in the urine about half to one hour after it has been 
 
156 GLYCOSURIA 
 
 taken, and reaches its maximum in about two to four hours. With 
 a moderate dose the excretion ceases in about eight to ten hours. 
 
 To explain physiological alimentary glycosuria it is generally 
 assumed that assimilation does not keep pace with absorption, 
 owing to the power of the liver to form and store glycogen being 
 limited. Hence, if too large an amount of sugar is taken into 
 the portal system in a short space of time, only a part can be 
 intercepted and removed by the liver, the remainder passing into 
 the systemic circulation. As soon as this happens sugar appears 
 in the urine, for the kidneys excrete any excess over the normal 
 limit of about 0-1 jier cent, that may be present in the blood. 
 Ginsberg has suggested that when a large quantity of sugar is 
 present in the intestine some may be absorbed by way of the 
 lacteals and thoracic duct, thus escaping the influence of the liver, 
 and so increasing the percentage of sugar in the blood. 
 
 Dextrose has the highest assimilation limit, 150 to 200 grams 
 in a single dose ; for levulose it is a little lower, 140 to 160 grams ; 
 cane-sugar has about the same assimilation limit as dextrose, 150 
 to 200 grams ; but for lactose it is lower, 120 grams with some 
 individuals, and as little as 60 grams for others. The pentoses can 
 only be assimilated to a very moderate degree, for when as little 
 as 30 to 50 grams are given by the mouth, almost half reappears 
 in the urine. Starch can be taken in very large quantities by a 
 healthy individual without producing glycosuria, apparently be- 
 cause its hydrolysis and absorption in the intestine take place so 
 gradually that the assimilative powers of the liver, and other organs, 
 are able to keep pace with the sugar as it enters the circulation. 
 Rubner found that a man consuming a daily ration of 508 to 670 
 grams of carbohydrate, in the form of wheaten bread, left un- 
 absorbed only 0-8 to 2-6 j)er cent. ; with an intake of 357 to 588 
 grams, in the shape of peas, 3-6 to 7-0 per cent, was unabsorbed ; 
 of 718 grams of potato 7-6 per cent, was not utilised, but in no 
 case was glycosuria j)roduced. Hence when a meal of starchy food, 
 of whatever amount, causes sugar to appear in the urine, it ma,y 
 be concluded that there is some definite morbid cause at work, 
 and that the individual is, potentially at least, a diabetic. This 
 differentiation of alimentary glycosuria ex amylo, from alimentary 
 glycosuria e saccharo was first suggested by Naunyn, and is now held 
 by most observers, an inability to assimilate a moderate amount 
 of starchy food without glycosuria resulting, being generally taken 
 as a criterion of pathological glycosuria, and furnishing the basis 
 for the modern dietetic treatment of diabetes, as we shall see later. 
 After the ingestion of an excess of dextrose the same sugar is 
 
TRANSITORY GLYCOSURIA 157 
 
 found in the urine, but Strauss and others have in addition met 
 with a levo-rotatory reducing substance, which P. Mayer regards 
 as glucuronic acid. According to v. Noorden, although some levu- 
 lose is excreted, after an excess of that sugar has been taken, it 
 is chiefly dextrose that appears in the urine. With cane-sugar, and 
 lactose, any excess that is may be absorbed unhydrolised from the 
 intestine into the lymph and blood, reappears in the urine as such, 
 for the unhydrolised sugars cannot be utilised by the tissues. Of 
 the disaccharides, maltose alone is said to be split up in the blood, 
 as the latter normally contains the ferment " maltase." It is 
 stated by v. Noorden that the assimilation limit of some otherwise 
 healthy persons is particularly low for this sugar, and that as 
 little as half a litre of beer, which is the only common food material 
 containing much maltose, is sufficient to cause sugar to appear in 
 the urine. The explanation he offers is that in these individuals 
 the maltose-splitting ferment of the blood is deficient. 
 
 Von Noorden has shown that if the assimilation limit for dextrose 
 is gradually exceeded, during successive experiments on the same 
 person, the whole excess is not excreted in its entirety, but that 
 only a certain proportion, varying with the individual, appears in 
 the urine. In two such series of experiments the following results 
 were obtained : — 
 
 A 
 
 After an intake of 100 grams of dextrose O'OO gram in the urine 
 150 „ „ 0-15 „ 
 
 180 „ „ 0-25 „ 
 
 200 „ „ 0-26 
 
 „ „ 250 „ „ 0"52 „ „ 
 
 B 
 
 After an intake of 100 grams of dextrose 0-00 gram in the urine 
 150 „ „ 0-00 „ 
 
 180 „ „ 0-25 
 
 200 „ „ 0-71 
 
 250 „ „ 0-64 „ 
 
 In any case the total sugar excreted is seldom more than 2 per 
 cent., or even at the highest, 3 per cent, of the intake. Raphael, 
 however, observes the paradoxical fact that some persons appear 
 to deal more efficiently with large than with small doses. 
 
 Injected directly into the circulation, dextrose and levulose, even 
 in large amounts, can be assimilated, provided that the injection 
 is conducted slowly. If the injection is too rapid, or too large 
 a dose is given (more than one gram per kilo of the body- weight), 
 
158 GLYCOSURIA 
 
 glycosuria results. Glucose and le^^llose appear in the urine almost 
 exclusively, but a fraction is said to be modified during its stay 
 in the body, and is excreted as glucuronic acid (Lepine). If elimi- 
 nation by the kidneys is prevented, by tjdng the ureters, the excess 
 of sugar undergoes changes ^vhich give rise to the formation of 
 lactic acid, aceto-acetic acid, acetone, and other substances, ^nth 
 resulting con\Tilsions, coma, and death. Starch in solution, cane- 
 sugar, lactose, and maltose are not assimilated when injected into 
 the blood-stream, but are at once excreted in the urine. After 
 the injection of one gram per kilo, Pavy recovered from the urine 
 l^assed in one hour : — 
 
 Saccharose. 
 
 Maltose. 
 
 Lactose. 
 
 Galactose. 
 
 Le\Tilose. 
 
 Glucose. 
 
 81 
 
 56 -o 
 
 48-7 
 
 28-9 
 
 20-9 
 
 15-6 
 
 The subcutaneous injection of as much as 69 grams of dextrose 
 only gives rise to traces of sugar in the urine, but \\'ith 100 grams 
 Voit found that 2-6 grams were excreted in the following seven 
 hours. Levulose appears to give much the same results, although 
 experiments with this sugar are few. Voit found traces in the urine 
 after injecting 31 grams. The injection of 30 grams of galactose 
 was followed by the appearance of traces of a reducing sugar in 
 the urine, but 28 grams of maltose did not give rise to any glycosuria. 
 Similar doses of saccharoses, and lactose, were found to give rise 
 to glycosuria, the sugars reappearing in the urine unchanged. The 
 injection of as little as 10 grams of sorbose was followed by the 
 appearance of 3-7 grams (36 per cent.) in the urine. 
 
 The differences in the assimilation limit sho^Mi by the organism 
 for various sugars is thus not only dependent upon their chemical 
 composition, but also upon their configuration, a character wliich 
 we have seen to be of great importance in determining the attack 
 of yeasts and micro-organisms on the sugars. Fermentation is 
 but a special instance of a general law, that the space disiDosition 
 of the chemical components of the bioplasm of a li\ing cell must 
 bear a certain well-defined relation to that of the food material 
 offered to it, before the latter can be utihsed for the production 
 of energy and for groA^th. It is a law that applies to the highest 
 as Avell as to the lowest organisms, and is of the utmost importance 
 in the physiology and pathology of assimilation. The first examjale 
 of such a relation between configuration and assimilation was 
 demonstrated by Pasteur, who showed that Penicillium glaucum, 
 Aspergillus niger, and other moulds, Avhen grown in a solution of 
 inactive racemic -tartaric acid, destroyed the levo-rotatory acid, and 
 left the dextro-rotatory variety, which could thus be recovered. 
 
TRANSITORY GLYCOSURIA 159 
 
 Some bacteria have been found to attack dextro-rotatory, others 
 levo-rotatory, lactic acid, while sorbose bacteria develop only in 
 alcohols of a definite composition and space configuration, oxidising 
 such substances as glycerine, erji^hrite, 1-arabinose, d-sorbite, 
 d-mannitol, &c., to the corresponding ketones ; but glycerol, 1-xylite, 
 dulcite, &c., are unaffected. The oxidation of the aldehyde sugars 
 by bacteria to the corresponding acids is in like manner found to 
 depend upon their configuration. 
 
 A similar relation between oxidation and configuration exists 
 in the higher organisms, even in warm-blooded animals. Thus it 
 has been shown by Brion that if the sterioisomeric tartaric acids 
 were given to a dog with its food, in quantities of about 0-2 to 
 0-75 gram per kilo of its body- weight, the following proportions 
 were recovered from the urine : — 
 
 Raceniic -tartaric acid . . . 2o-0 to 42-0 per cent. 
 
 Dextro- „ „ . . . 25-0 „ 29-0 
 
 Levo- ,, ,, . . . 2'7 ,, 6'4 ,, 
 
 Meso- „ „ . . . 2-4 ,, 6-7 
 
 From this it would appear that the racemic- and dextro-acicls 
 are broken down with difficulty in the organism, but that the 
 levo- and meso-varieties are more easily assimilated. 
 
 Ex23eriment has shown that a previously sugar-free dog only 
 begins to pass sugar in its urine when it consumes 1-9 to 2-5 grams 
 of grape-sugar per kilo of its body- weight, but that when galactose 
 is substituted, sugar appears in the urine if 0-2 to 0-4 gram per 
 kilo are taken (Hofmeister). Mannose appears to be even less 
 easily assimilated than galactose. Rosenfeld found that, 20 grams 
 of dextrose caused no glycosuria in a dog of 7 kilos, and that the 
 same amount of galactose caused 3-2 grams of sugar to appear in 
 the urine, while after 20 grams of mannose 4-2 grams of sugar were 
 excreted. Further experiment showed that 1-mannose was assimi- 
 lated \\dth greater difficulty than cl-mannose, Neuberg and Wohl- 
 gemuth administered the sterioisomeric arabinoses by the mouth 
 to rabbits, and recovered from the urine a proj)ortion which varied 
 with the conformation of the sugar molecule : — 
 
 1-arabinose .... 14-49 per cent. (1-) 
 d-arabinose .... 39-97 „ (d-) 
 
 521-59 „ (r-) 
 
 ■ I 9-0 „ (d-) 
 
 With subcutaneous and intravenous' injections very similar 
 readings were obtained. It would therefore seem that 1-arabinose 
 is much more easily assimilated than d-arabinose, and that the 
 
 r-arabinose 
 
160 GLYCOSURIA 
 
 raceniic variety is partly decomj^osed, the 1-portion being oxidised, 
 while part of the d-portion is secreted in the urine unchanged. 
 Only one sugar "v^ith more than six carbon atoms, a-glycoheptose, 
 has been investigated, and this Wohlgemuth found was assimi- 
 lated with much greater difficulty than the hexoses. 
 
 The above, and numerous other experiments on similar lines, 
 have demonstrated that the assimilation, hke the fermentation, of 
 sugars is dependent upon their structure and configuration, and 
 that, although there are individual variations, each sugar has an 
 assimilation limit which is fairly constant for healthy persons. 
 
 Pathological Alimentary Glycosuria 
 
 The assimilation hmit for sugar in disease has been the subject 
 of numerous investigations, and it has been found that it is par- 
 ticularly Hable to be lowered by pathological conditions of (1) the 
 nervous system, (2) the liver, (3) the pancreas, thyroid, &c., and 
 (4) as the result of toxic disturbances dependent upon certain 
 infectious diseases, poisons, drugs, &c. For the purpose of such in- 
 vestigations, dextrose, levulose, and cane-sugar have been chiefly 
 employed, but as the absorption of the last named is very liable 
 to be influenced hy the condition of the intestinal wall and contents, 
 the results obtained by its use are not so reliable as those found 
 after the administration of the simple monosaccharides. 
 
 I. Alimentary Glucosuria. — It is customary when testing 
 the dextrose tolerance of an individual to administer a single dose 
 of 100 grams, dissolved in a quarter of a litre of water or tea, in the 
 morning fasting or a couple of hours after breakfast, so that it is 
 taken into an empty stomach. After such a close a healthy person 
 excretes no sugar in his urine, but if, for any reason, his tolerance 
 is lowered, sugar mil be found in about half an hour, and vnW as a 
 rule persist for three or four hours. Sometimes it happens that 
 sugar A^dll continue to be excreted for several days, a phenomenon 
 which suggests the presence of a serious derangement of carbo- 
 hydrate metabolism. If 100 grams of dextrose suffices to induce 
 glycosuria smaller quantities should be tried on succeeding days, 
 until an amount is found which is not followed by the appearance 
 of any sugar in the urine. This represents the limit of tolerance 
 for dextrose of the particular case under investigation. In some 
 instances it mil be found that 50, 25, or even 10 grams ^\i\\ be 
 sufficient to produce glycosuria, but when the limit is lowered 
 to 10 or 15 grams it will usually prove that the limit of tolerance 
 for starch is also reduced, and the patient is a potential diabetic. 
 
TRANSITORY GLYCOSURIA 161 
 
 1. hi Nervous Diseases. — The discovery by C. Bernard that 
 puncture of the floor of the fourth ventricle in animals gave rise 
 to glycosuria, stimulated research into the relation between diseases 
 of the nervous system and diabetes in man. In the course of these 
 investigations the question of alimentary glycosuria in diseases of 
 the nervous system naturally received considerable attention. 
 
 Strasser administered 100 grams of dextrose to thirty-seven 
 cases of disease of the central nervous system, and found glycos- 
 uria in seven, more often wdth affections of the brain than of 
 the spinal cord. It was found by Oordt that alimentary glycos- 
 uria is more common with cerebral tumours than with other 
 organic diseases of the brain, while Haedke obtained a positive result 
 with fifteen out of twenty-five cases of injury to the skull. Klippel, 
 Vigoroux, and Juquelier state that alimentary glycosuria is most 
 easily jorovoked in patients with mental confusion, hallucinations, 
 and delirium, and that the tendency disappears as the mental 
 condition improves. Epilepsy, according to Oordt, does not 
 appear to predispose to alimentary glycosuria, and in simple 
 hysteria and hypochondriasis it is not common. Strauss investi- 
 gated the sugar tolerance of thirty cases of tabes, and only ob- 
 tained a positive result with one. A long series of cases of nervous 
 disease were investigated by Arndt, mth the following results : — 
 
 55 cases of General paralysis 5 positive (10 per cent.) 
 
 31 ,, Hysteria 
 
 7 ,, Hypochondriasis 
 
 21 ,, Melancholia 
 
 7 ,, Stupor 
 
 6 ,, ]Mania 
 
 13 „ Epilepsy 
 
 11 ,, Traumatic neiu-opsychoses 4 
 
 It is evident that in certain conditions of the nervous system 
 the capacity of the body for assimilating glucose is diminished. 
 This appears to be most common in traumatic neuroses, cerebral 
 tumours, acute diseases of the brain and meninges, neurasthenia, 
 and forms of mental debility, particularly mania and paralysis, 
 in which it may be assumed that either the diabetic centre in the 
 medulla is excited, or that the nervous arc, of which it is the centre, 
 is so affected that the liver is unable to store glycogen in any 
 quantity, esj^ecially when a large amount of sugar is suddenly 
 thrown into the portal vein. 
 
 2. In Pathological Conditions of the Liver. — In 1875 Colrat 
 announced that alimentary glycosuria is an important sign of 
 disease of the liver, and particularly of cirrhosis, pylephlebitis, 
 
 L 
 
 2 
 
 
 ( 6 
 
 
 
 
 
 ( , 
 
 
 5 
 
 
 (24 , 
 
 
 1 
 
 
 (14 , 
 
 
 1 
 
 
 (16 , 
 
 
 
 
 
 ( , 
 
 
 4 
 
 
 (36 
 
 
162 GLYCOSURIA 
 
 and other conditions, in which the portal vein is more or less ob- 
 structed. His conclusions were confirmed by Baylac, who stated 
 that a positive result was always obtained in such cases, excepting 
 when absorption was interfered with. CampagnoUe, and Mepraschk, 
 came to the same conclusion, but found that there was no relation 
 between the alimentary glycosuria and the intensity of the liver 
 disease. De Haan, who examined twenty-nine cases, obtained a 
 positive result with eighteen (62 per cent.). 
 
 The value of these, and many other observations, mostly by 
 members of the French school, is discounted, however, by the fact 
 that cane-sugar was generally used for the tests. Strauss, who 
 employed dextrose, only obtained a positive result with one out of 
 twenty cases of liver disease, in marked contrast to the French 
 observers. Valmont failed to obtain alimentary glycosuria with 
 a single one of the seven cases of cirrhosis of the liver that he 
 examined, and Krauss and Ludwig only obtained a positive result 
 with three out of seven cases, and in two of these the dose of sugar 
 exceeded 100 grams. Bloch examined nine cases of liver disease, 
 mostly cirrhotic, with regard to their tolerance for sugar, and 
 found that the urine of only one gave any reduction, although 
 several others were levo-rotatory, probably from the presence of 
 glucuronic acid. Other observers have also reported negative 
 results with cirrhosis of the liver. Three cases of catarrhal jaundice 
 were investigated by Ziilzer with negative results. The same 
 observer found that cholelythiasis and amyloid disease of the 
 liver did not predispose to alimentary glycosuria. After the 
 administration of 100 grams of dextrose v. Jaksch recovered from 
 the urine of a case of acute phosphorus poisoning 20 grams of sugar, 
 and Walko subsequently reported several similar cases. The 
 presence of alimentary dextrosuria has been recorded by v. Noorden 
 in cases of fatty liver. 
 
 It may be concluded that, although alimentary dextrosuria can 
 be frequently produced in cases where there is disease of the liver, 
 by the administration of 100 grams or more of glucose, it has not 
 the diagnostic significance that it was at one time supposed to 
 have. 
 
 3 (a). In Diseases of the Pancreas. — The relation of the pancreas 
 to alimentary dextrosuria in man was investigated by Wille, who 
 gave 70 to 100 grams of grape-sugar to eight hundred patients, 
 suffering from a variety of diseases, in the morning before food 
 had been taken. Their urine was examined before the test, and 
 at intervals of two hours afterwards. Of these eight hundred indi- 
 viduals seventy-seven subsequently died, and were examined post- 
 
TRANSITORY GLYCOSURIA 163 
 
 mortem. Alimentary glycosuria had been found in fifteen. In 
 ten (65 per cent.) of them there were grave lesions of the pancreas, 
 either primary or secondary to growths of the stomach, liver, or 
 gall-bladder. Alimentary dextrosuria has also been reported by 
 Niepraschk in association with cancer of the pancreas, and by 
 Kraus and Ludwig with a cyst of the pancreas. 
 
 Pratt and Spencer found that the rapid atrophy of the pancreas 
 produced in dogs by satisfactory and complete ligature of the 
 pancreatic ducts, quickly lowers their assimilation limit for dex- 
 trose. In one dog it fell from 121 grams to 65 grams within three 
 Aveeks, and in two others it dropped to 35 grams in from two to 
 three months. 
 
 (&) In Diseases of the Thyroid Gland. — Arndt discovered that ali- 
 mentary dextrosuria is favoured by the presence of exophthalmic 
 goitre, and Bettmann showed that a sufficiently prolonged ad- 
 ministration of thyroid preparations by the mouth produces a 
 similar result. On the other hand, Knopf elmacher states that in 
 myxoedema the power of assimilating sugar is remarkably in- 
 creased, and that this is notably diminished after treatment with 
 thyroid preparations. 
 
 (c) Influence of the Kidneys. — Other things being equal, ahmen- 
 tary dextrosuria is favoured by the administration of a diuretic 
 (Gobbi), and is less liable to occur when the permeability of the 
 kidneys is diminished from any cause (Ac hard and Castaigne). 
 
 4. Influence of Toxines — (a) Alcohol. — Small doses of alcohol 
 do not notably influence the assimilative powers for dextrose, 
 and chronic alcoholism does not predispose to alimentary dextros- 
 uria as much as might be expected. Strauss obtained a positive 
 result with 7 per cent, of the cases he examined, and Arndt with 
 three out of twenty- three cases, two of Avhich had also been 
 exposed to the influence of lead. Acute alcoholism, on the con- 
 trary, gives rise to alimentary dextrosuria in a considerable pro- 
 portion of cases. Strauss found that it was present in 70 ]3er 
 cent, of the cases of delirium tremens that he investigated, and 
 Arndt obtained very similar results, thirteen out of twenty-nine 
 .cases (65 per cent.) of acute alcoholism in his experience having 
 alimentary glycosuria. 
 
 (&) Lead. — Acute and chronic lead-poisoning, but more par- 
 ticularly the former, are said to give rise to alimentary dextrosuria 
 (Strauss). A positive result was obtained by Rosenberg in 60 per 
 cent, of cases, mostly with colic. Since, according to Mosse, lead 
 is deposited in the coeliac plexus in experimental lead coHc, and 
 extirpation of the plexus gives rise to glycosuria, it is probable 
 
164 GLYCOSURIA 
 
 that the tendency to glycosuria seen in such cases is dependent 
 upon disease of the coeliac plexus. 
 
 (c) Carbon Monoxide. — We have seen that spontaneous glycos- 
 uria is not uncommon as a result of poisoning A^dth carbon mon- 
 oxide. Even in those cases Avhere this does not occur ahmentary 
 glycosuria is easih^ induced, according to Muenzer and Palma. 
 
 {d) Other toxic substances, such as chloralhydrate, copaiba, 
 nitro-benzol, &c., have been said to give rise to alimentary glycos- 
 uria on the results of the reduction tests ; but the reduction in 
 these cases is more probably due to glucuronic acid, as the presence 
 of sugar cannot be confirmed by the fermentation and phenyl- 
 hydrazin tests. 
 
 (e) Febrile Diseases. — According to Poll, febrile diseases are 
 frequently associated ^^ith a lowered tolerance for dextrose, from 
 0*45 to 8-0 per cent, of the administered sugar reappearing in the 
 urine in the cases of pneumonia, typhoid, scarlet fever, acute arti- 
 cular rheumatism, &c., that he investigated. CampagnoUe con- 
 firms these results, but Blumenthal found that the reduction, in 
 many cases at least, is due to glucuronic acid. 
 
 (/) Syphilis. — Paris and Dobrovici observed alimentary dex- 
 trosuria in four out of ten cases of syphilis, but their results lack 
 confirmation. 
 
 (g) In other diseases alimentary clextrosuria has also been 
 described by a few observers. Nobecourt obtained a positive 
 result with seven out of twelve rickety children. Nagelschmidt 
 examined seventeen cases of psoriasis, and found that the urines 
 of five gave a reduction. Naunpi states that clilorosis predisposes 
 to ahmentary glycosuria, while Mayer and Pick observed the 
 same ^^ith regard to obesitj^, obtaining a positive result with twenty 
 out of fifty cases they examined. 
 
 II. Pathologfical Alimentary Levulosupia.— This condition 
 has been chiefly investigated with regard to the influence of diseases 
 of the liver, and its presence is regarded bj^' most authorities as 
 evidence of hepatic insufficiency. In 1899 Sachs, working under 
 Strauss' direction, found that after extirpation of the liver in 
 frogs there was a lessened tolerance for levulose when it was in- 
 jected into the lymph sac, but not for dextrose, galactose, or 
 arabinose. In a later paper he stated that the liver was able 
 to form glycogen from levulose, but that the muscles had not 
 this property, whence it might be assumed that alteration in the 
 functions of the liver would be sho"sra by a diminished tolerance 
 for levulose. Clinical experience tended to confirm tliis conclu- 
 
TRANSITORY GLYCOSURIA 165 
 
 sion, for he found that, whereas healthy individuals could assi- 
 milate levulose better than dextrose, tolerance for the same amount 
 ■of levulose was much diminished in persons with liver diseases. 
 Strauss investigated the tolerance for levulose of eighty-seven per- 
 sons, and found that, of twenty-nine who were suffering from 
 disease of the liver, twenty-six (90 per cent.) showed alimentary levu- 
 losuria, while of fifty-eight who had no obvious liver trouble, only 
 six (10 per cent.) passed levulose in their urine ; and of these six two 
 were driinkards and suffered from obesity and gout, one had some 
 Tiepatic congestion secondary to mitral stenosis, one was suffering 
 from ansemia due to gastric haemorrhage, and one had pneumonia, 
 in which, according to Rosenberger, levulosuria is frequently met 
 with. Of the cases of hepatic disease that did not show levu- 
 losuria, one had atrophic cirrhosis and severe diarrhoea, causing 
 interference with the absorption of the sugar, one had acute chole- 
 cystitis of two days' duration, and a third had a cyst of the liver. 
 
 Results exactly corresponding to those of Strauss were obtained 
 by Bruining, who examined eleven cases of cirrhosis of the liver, 
 and found that ten (90 per cent.) showed alimentary levulosuria. 
 Baylac and Arnaud reported a positive result with twenty-one out of 
 twenty-three cases (91 per cent.) of liver disease that they examined, 
 but observed alimentary levulosuria in a much higher proportion 
 •of cases presenting no symptoms referable to the liver than Strauss, 
 seven out of twenty (35 per cent.) passing sugar in their urine after 
 100 grams of levulose. With regard to the latter they point out, 
 iowever, that slight derangements of the hepatic functions are 
 not uncommon in hospital patients such as they examined. Lands- 
 berg was unable to confirm the diagnostic value of the test, and 
 •considered that idiosyncrasy is more important in producing 
 alimentary levulosuria than functional disease of the liver. He 
 ■obtained a positive result with four out of seven healthy persons, but 
 with only nine out of twenty-one cases (43 per cent.) of patients with 
 affections of the liver that he examined, including four out of eleven 
 Avith cirrhosis, one out of four with carcinoma, two with hyper- 
 trophic cirrhosis, and two with chronic obstruction of the bile 
 duct, while one with icterus and biliary calculi, and one of con- 
 gested liver, showed no alimentary levulosuria. He suggests that the 
 higher percentage of positive results recorded by Strauss is to be 
 attributed to his taking only advanced cases in which a collateral 
 ■circulation had been established, thus allowing rapid entrance of 
 levulose into the blood and urine. In 1904 Chajes collected all the 
 •cases of alimentary levulosuria reported up to that date, and 
 added twenty-one observations on normal individuals, shoA^ing 
 
166 GLYCOSURIA 
 
 only one positive result, of his own. Of eighty-four cases of clinical 
 liver disease he found that fifty-two (86 per cent.) showed alimentary 
 levulosuria, while of ninety-nine persons whose liver was supposed 
 to be healthy only 15 per cent, x^assed sugar in their urine as a 
 result of the test. 
 
 Von HaUsz considers the test of great service in deciding 
 between cirrhosis and other conditions. He states that normally 
 100 grams of levulose rarely cause levulosuria, and that a positive 
 result points to diffuse and severe disease of the liver, and is espe- 
 cially indicative of cirrhosis. Goodman, who investigated thirty- 
 two cases, came to the conclusion that, while not indicative of any 
 specific organic lesion of the liver, alimentary levulosuria is most 
 frequently observed in cirrhosis, with which it is almost a con- 
 stant phenomenon, and that the early, or late, appearance of levu- 
 lose in the urine may be regarded as a sign of severe, or mild, hepatic 
 disease. He obtained a positive result with all the twenty cases 
 of cirrhosis he tested, a speedy appearance of sugar coinciding 
 with severe disorders, and a tardy appearance with clinically mild 
 affections. Six cases with chronic passive congestion showed no 
 levulosuria, and Goodman points out the usefulness of this test 
 as a means of diagnosing that condition from cirrhosis. A positive 
 reaction was obtained in 12-5 per cent, of non-hepatic diseases. 
 
 Hohlweg administered 100 grams of levulose, dissolved in 
 300 c.c. of water, fasting, to forty-one cases of various liver affec- 
 tions, and states that the tolerance was most reduced with cirr- 
 hosis, catarrhal jaundice, and obstructions of the common duct 
 by gall-stones. A stone in the cystic duct, or tumour of the liver, 
 did not give rise to alimentary levulosuria. The different effects. 
 XDroduced when the bile duct was obstructed by a tumour, and 
 by a gall-stone, were found to be very striking. He came to the 
 conclusion that all affections that injure the parenchyma, reveal 
 the functional disturbance by alimentary levulosuria, but that 
 the assimilation is scarcely interfered with in cases of enlargement 
 of the liver, leuchaemia, and congestion, or tumours. He con- 
 firmed these clinical findings by the results of exiaeriments in 
 which pathological conditions of the liver were induced by in- 
 jections of toluilendiamin, phosphorus, or a mixture of chloroform, 
 oil, and paraffin. Sabatowski applied the test in seventy-eight 
 cases of liver disease. Positive findings were constantly obtained 
 with cirrhosis of the liver, and were also the rule in infectious 
 diseases, and jaundice of infectious and toxic origin. Jaundice 
 from obstruction did not give rise to alimentary levulosuria unless 
 there were anatomical changes in the parenchyma, and nutmeg 
 
TRANSITORY GLYCOSURIA 167 
 
 liver also gave a negative result, excepting when the parenchyma 
 was much damaged. He believes that alimentary levulosuria is 
 independent of stasis of bile, but is a constant accompaniment 
 of every severe liver affection, and may serve to differentiate an 
 infectious process in the liver. Frey found that alimentary levu- 
 losuria occurred in 10 per cent, of persons with healthy livers, 
 notably in association with affections of the pituitary gland, and 
 in 59 per cent, of those in which that organ was diseased. He 
 states that a positive result is especially frequent when there is 
 cirrhosis, but does not constantly accompany that condition. 
 
 The relation of alimentary levulosuria to diseases of the liver 
 in children was investigated by Brun. He gave various doses of 
 levulose to four hundred sick and about one hundred healthy 
 children, ranging in age from one month to twelve years, and 
 found that " it was an easy, delicate, and harmless test for deter- 
 mining the functional capacity of the liver. It reveals the slightest 
 impairment of function, and also the progress toward recovery 
 or the reverse." He found that the normal liver does not permit 
 levulose to pass unmodified. 
 
 III. Patholog-ical Alimentary Galactosuria.— Comparatively 
 
 few researches into the influence of disease upon the excretion of 
 galactose have been carried out. According to Bauer the ad- 
 ministration of 29 grams of this sugar, fasting, which does not cause 
 the appearance of sugar in the urine in health, is followed by 
 galactosuria in patients suffering from organic, or functional, dis- 
 orders of the liver. One hundred grams cause a more marked 
 excretion, but also gives rise to galactosuria and dextrosuria in 
 the healthy. It is claimed by Bauer and others that the assimi- 
 lative capacity of the body for milk-sugar when 40 grams are 
 given in the morning after free purgation, gives more satisfactory 
 indications of the functional capacity of the liver than levulose, 
 and that it is also more readily taken. Normally only to 1 gram 
 of sugar can be detected in the urine subsequently, and this appears, 
 if at all, in the first hour, and all traces having disappeared by the 
 fourth hour. If, however, the liver cells are diseased, 4 to 10 grams 
 of galactose are eliminated by the kidneys, the largest amount 
 appearing in from three to six hours and continuing for ten to 
 fourteen hours. The test was originally based on the observation 
 by Bauer that a man with Hanot's cirrhosis who took large quan- 
 tities of milk passed an unfermentable sugar in his urine, and 
 that the excretion ceased when the milk was stopped. 
 
 IV. Patholog"ical Alimentary Lactosuria. — Lactosuria has 
 
168 GLYCOSURIA 
 
 been observed by v. Halasz to follow the ingestion of a moderate 
 amount of milk in patients with dilated stomach. It has also 
 been stated by Grosz that an excess of milk sometimes causes a 
 reducing substance, having the reactions of lactose, to apjjear in 
 the urines of infants, particularly when they are suffering from 
 gastro-intestinal catarrh. 
 
 Ziilzer investigated alimentary lactosuria during the puer- 
 perium, and found that 60 grams of lactose did not cause sugar to 
 appear in the urine of any of the cases he examined, but that 
 after 100 grams, eleven out of thirteen cases gave a slight reaction. 
 Lactosuria was only exceptionally met ^dth after a similar dose 
 in normal women. 
 
 Von Xoorden gave 159 grams of cane-sugar to lying-in women, 
 after abortion or premature labour, and recovered milk-sugar 
 from their urine. 
 
 V. Pathological Alimentary Maltosuria. — I am not 
 
 acquainted with any observations on the effects of disease in 
 reducing the tolerance for maltose, but, judging from the readiness 
 with, which comparatively small quantities of beer sometimes 
 produce maltosuria, it is not unlikely the tendency may be in- 
 creased by various pathological conditions. 
 
 VI. Pathologrieal Alimentary Saccharosuria.— Many of the 
 
 original observations on sugar tolerance were made AWth cane- 
 sugar, and although glycosuria was found to follow its ingestion 
 in cirrhosis of the liver, and other conditions giving rise to hepatic 
 insufficiency, bj^ some observers, others obtained conflicting results. 
 It was not until pure levulose and dextrose were employed that 
 the more constant findings already referred to were secured. There 
 can be no doubt that the occurrence of glycosuria after the ad- 
 ministration of saccharose is largely controlled by the condition 
 of the alimentary tract, and for this reason the simple mono- 
 saccharides are to be preferred for experimental investigations. 
 Traces of cane-sugar are said to have been found in the urine in 
 children after a considerable quantity has been consumed, but as 
 a rule it apj)ears in the form of dextrose, possibly accompanied 
 by some levulose. 
 
 VII. Pathological Alimentary Pentosuria.— Xothing is 
 
 known of the effects of disease on the production of alimentary 
 jDento&uria. According to v. Jaksch, large doses of arabinose, 
 xylose, and rhamnose give rise to diarrhoea, and the sugars can 
 be recovered from the urine, a quarter to a half of the quantity 
 ingested being sometimes thus excreted. I have found that in 
 
TRANSITORY GLYCOSURIA 169 
 
 a healthy man the ingestion of 0'5 gram of xylose was followed 
 by the appearance of 0-08 gram of that sugar in the urine within 
 twenty-four hours. 
 
 Transitory and Intermittent Glycosuria 
 
 The transitory forms of glycosuria already considered, whether 
 in health or disease, are produced artificially by the administration 
 ■of a relatively large dose of sugar under experimental conditions, 
 but sj^ontaneous or idiopathic transitory glycosuria is also known 
 to occur. This is met with as an epiphenomenon in the course 
 •of various organic, or constitutional, diseases. It appears without 
 the addition of any extraordinary amount of sugar, or other carbo- 
 hydrate, to the diet, and disappears again as the condition ■\\T.th 
 which it is associated improves. The existence of sugar in the 
 "urine in such cases is usually discovered accidentally in the course 
 ■of a routine examination, and its presence is not associated with 
 any characteristic symptoms. 
 
 I. Transitory Glycosuria 
 
 [a) In Nervous Diseases. — Such a transitory glycosuria, appa- 
 rently of central origin, has been noticed in connection mth lesions 
 of both the central and peripheral nervous system, such as tumours 
 and haemorrhages at the base of the brain, lesions of the floor of 
 the fourth ventricle, cerebral and spinal meningitis, concussion 
 of the brain, fracture of the cerviacal vertebrse, tetanus, and 
 sciatica. It has also been met with after epileptic, hystero-epi- 
 leptic, and apoplectic seizures, in traumatic neuroses, such as 
 those following railway accidents, mental shocks, mental strain, 
 worry, fatigue, and great anxiety. Glycosuria follo\\ing mental 
 strain and anxiety is not at all uncommon in business and j)ro- 
 f essional men, but if the cause of the condition is recognised at a 
 sufficiently early stage, and the patient is placed under satisfactory 
 hygienic conditions, both physical and mental, the sugar gener- 
 ally disappears as the general health improves. In several such 
 patients I have been unable to induce even alimentary glycosuria, 
 by the administration of 100 grams of dextrose or levulose, taken 
 lasting, after a sea- voyage, or other form of " rest-cure." One 
 ■case has been under observation for five years, and another for 
 three, without any return of the glycosuria, but it is as yet too 
 early to say that this may not occur. 
 
 Transitory glycosuria was observed by Siegmund in 52 per cent. 
 
170 GLYCOSUEIA 
 
 of cases of general paralysis, in 7-4 per cent, of epileptics, and in 
 3-77 per cent, of dementia cases, but he did not meet with it in 
 other mental diseases. Glycosuria after epileptic and apoplectic 
 seizures does not, however, appear to fee as common as it is fre- 
 quently stated to be. Von Jaksch examined fifty recent cases of 
 hemiplegia and did not find sugar in the urine in a single instance, 
 while Simon had a similar experience with the urines of a large 
 number of epileptics examined within a few hours of the attack. 
 
 Kausch has reported eleven cases of recent injury, nine of 
 fracture and two of contusion, in which glycosuria was present. 
 He states that the sugar found was in minute c[uantities, the 
 maximum being 1 per cent, and the average 0-5 per cent. It 
 appeared almost immediately after the injury, being present in 
 the first specimens of urine examined. The glycosuria disappeared 
 in all cases in the course of a week, usually about the third day, 
 and was never associated with any other diabetic symptoms. 
 With one exception all the patients remained in a normal con- 
 dition Avith regard to their urine. According to Ott, the transitory 
 glycosuria that occurs during pregnancy, in nervous cases, and 
 ends after delivery, is very probably due to hyper-hypophsy, 
 although the promotion of acti\aty of the chromaffin tissue by the 
 th5T:'oid is most likely also contributory. 
 
 (b) Infectious Diseases. — We have seen that alimentary glycos- 
 uria is favoured by the presence of infectious diseases, and it is 
 therefore not surprising to find that spontaneous transitory glycos- 
 uria is occasionally observed in acute febrile disorders, and more 
 particularly during convalescence from them. 
 
 The existence of a reducing substance in the urine of patients 
 suffering from malaria was pointed out by Burdel, who obtained a 
 positive result with 17 per cent, of non-cachectic, and 76 per cent, 
 of cachectic cases. Calmette examined forty-one soldiers suffering 
 from malaria and found glycosuria in five, about the same pro- 
 portion as in Burdel's non-cachectic cases. Seegen reported five 
 cases of malaria with " diabetes," in which both conditions dis- 
 appeared under the administration of quinine. Although the 
 glycosuria in this instance may be referred to the action of toxines, 
 it would seem that the general condition of the patient has more 
 influence on its production than the intensity of the intoxication. 
 It is possibly akin to the vagabond glycosuria, described by Hoppe- 
 Seyler, who found that tramps who had walked long distances 
 on a diet of bread and potatoes frequently passed urine containing 
 0-5 to 0-7 per cent, of glucose on admission to an institution. The 
 sugar quickly disappeared, even on a diet rich in carbohydrates. 
 
TRANSITOKY GLYCOSURIA 171 
 
 and did not return, although large doses of dextrose were given on 
 an empty stomach. 
 
 Glycosuria appears to be not uncommon in diphtheria. Binet 
 obtained a positive result in twenty-nine out of seventy cases. 
 Of these thirty-eight were severe, mth twenty-seven positive re- 
 sults, and three mild infections, with two positive results. Simon 
 met with a transitory glycosuria in four out of thirty-t^vo cases 
 of the same disease of moderate severity. Hibbard and Morrissey 
 noticed the existence of a glycosuria, of variable duration, and 
 frequently associated with albuminuria, in serious cases of diph- 
 theria, and have also pointed out that the injection of anti-diph- 
 theritic serum is sometimes followed by the temporary appearance 
 of sugar in the urine. 
 
 In one case Teschemacher reports that the injection of Koch's 
 tuberculin was followed by temporary glycosuria. 
 
 Transitory glycosuria has also been met with in typhoid fever, 
 scarlet fever, measles, cholera, influenza, whooping cough, pneu- 
 monia, smallpox, and other infectious diseases, including syphilis. 
 Manchot obtained a reduction with the urines of twelve out of 
 359 patients suffering from the last-mentioned disease in the 
 second stage. 
 
 (c) Toxic Conditions, d;c. — The absorption of toxic materials 
 affecting the nervous system, or possibly the pancreas, may account 
 for the transitory glycosuria noticed by Hoppe-Seyler in cachectic 
 individuals with intestinal troubles, and in dyspeptics by Robin. 
 While a similar explanation possibly holds good for the sugar 
 found in the urine of children affected with tape- worms by Judson, 
 and with round worms by Parry. Glycosuria has also been noticed 
 in association with cysticircus tumours by Marie and Guillain and 
 others, but it is not improbable that this was merely a coincidence. 
 
 (d) Surgical Conditions. — Transitory glycosuria has been re- 
 ported in various surgical affections. Leidy met with it in twO' 
 cases of appendicitis, and Cohen has reported a similar case, in 
 which, after deferring operation because of the sugar in the urine, 
 it was found to disappear two days after the appendicular abscess 
 had been opened. Cohen suggests that the glycosuria was due 
 to irritation of the solar plexus. Malcolm has met with glycosuria 
 in association with an ovarian cyst, which disappeared on removing 
 the tumour. Robinson has reported what appeared to be levu- 
 losuria, during an attack of gonorrhoea, that disappeared when the 
 discharge was cured. As long ago as 1861, Hill described glycos- 
 uria in four cases of severe burns, and, according to Vamiini, it 
 is not an uncommon complication. 
 
172 GLYCOSURIA 
 
 The occurrence of transitory glycosuria in connection with the 
 foregoing diseases, while startling at the time, is usually of no 
 diagnostic or prognostic importance. It is advisable that the 
 urine should be watched and examined at intervals for sugar, since 
 occasionally a persistent glycosuria may subsequently make its 
 appearance, but as a rule the condition is a temporary one, and 
 does not indicate any profound or lasting interference ^nth carbo- 
 hydrate metabolism. 
 
 II. Intermittent Glycosuria 
 
 In addition to the preceding type of case, in which, after the 
 urine has contained sugar for a shorter or longer period, tolerance 
 for carbohydrates becomes completely restored, we have a second 
 group, in A^hich the limit of tolerance for carbohydrates is j)er- 
 manently depressed, and glycosuria occurs whenever more than 
 a certain amount of sugar, or starchy food, is taken, or an inter- 
 current complication temporarily lowers the patient's assimilative 
 powers. To this class belong the cases of alimentary glycosuria 
 ex amylo previously referred to. At present we have no means 
 of distinguishing the truly transitory from the intermittent variety 
 of glycosuria, excepting experience and continued observation ; 
 but as the latter is much more common, and infinitely more serious 
 than the former, it is well to regard every case with sugar in the 
 urine as a possible diabetic and treat it accordingly, unless there 
 is only a trace and the glycosuria speedily clears up after the con- 
 dition with which it was associated has disappeared. 
 
 Intermittent glycosuria may probably be caused hy any patho- 
 logical state that can bring about a j^ersistent excretion of sugar 
 in the urine, but practically nothing is known of any others than 
 disease of the pancreas. Both exjDeriments upon animals, and 
 clinical experience, have shown that the sugar tolerance of patients 
 suffering from chronic inflammatory conditions of the pancreas is 
 lowered in many cases, and it is also known that persistent glycos- 
 uria ultimately develops when the cirrhosis of the gland reaches 
 a certain stage. If, therefore, a patient whose pancreas is partly 
 incapacitated as the result of inflammation, exceeds his lowered 
 limit of tolerance, or has it temporarily reduced still further by 
 an attack of subacute pancreatitis, sugar will appear in the urine, 
 to again disappear when the exciting cause is removed. I have 
 had several cases under my observation in which this course of 
 events apparently occurred. 
 
 In September 1908 I was consulted by a gentleman of sixty-two. 
 
TRANSITORY GLYCOSURIA 173 
 
 who had symptoms suggesting the presence of floating gall-stones 
 in the common bile duct. An analysis of his urine and fseces tended 
 to confirm this diagnosis. Operation was advised but refused. In 
 May 1909 I again saw him and found that his urine, which had pre- 
 viously been sugar-free, contained 19 grams (1-2 per cent.) of dextrose.. 
 Under treatment this was reduced to 9 grams (0-8 per cent.), and by 
 the end of June it had quite disappeared. He was placed on a re- 
 stricted diet, and his virine examined at frequent intervals, but it was 
 always found to be free from sugar during the following three months. 
 In July 1911 he retm-ned, and his urine was now found to contain 
 108 grams (6 per cent.) of dextrose, and his faeces showed a much higher 
 proportion of unsaponified fat, besides giving a well-marked pancreatic 
 insufficiency test (Gross). With careful dieting the amount of sugar 
 in the urine was reduced to 70 grams in the twenty-four hours (4 per 
 cent.), but it could not be removed altogether. 
 
 Another patient, who was sent to me in January 1908, had been 
 examined for life insurance, and passed, six months previously. After 
 a long drive in an open carriage, he complained of pain in the back and 
 under the left shoulder, and his medical man on examining his virine 
 found a specific gravity of 1-032 and 2-8 per cent, of sugar. A twenty- 
 four hoiir sample of the urine that I examined showed 48 grams (3 per 
 cent.) of dextrose, and a trace of acetone, but no aceto-acetic acid. 
 The excretion of ammonia nitrogen was not excessive (0-6 grams). 
 His fseces were acid in reaction, and contained 55 per cent, of un- 
 absorbed fat, including 41 per cent, in the unsaponified form. Casein 
 digestion was found to be imperfect. On physical examination 
 tenderness, and some fulness, in the region of the head of the pancreas 
 could be made out. He was dieted, and treated with pancreatic 
 extracts, urotropine, salicylates, &c., with a view to quieting the 
 pancreatic inflammation, and relieving the gland of as much work 
 as possible. All the symptoms disappeared, and the iirine became 
 sugar-free in about six weeks. He remained in good health for two- 
 years, and then complained of being easily tired. An examination 
 of his urine showed 28 grams of sugar in the twenty-four hours, a 
 well-marked reaction for acetone and aceto-acetic acid, and 1"6 grams 
 of ammonia nitrogen. The glycosiu-ia and acidosis were reduced by 
 treatment, but his urine could not be again made cj[uite sugar-free. 
 
 A case of convsiderable interest in this connection has been 
 described by Garrod. 
 
 The patient, a man aged fifty-eight, came iinder observation in 
 October 1909, suffering from a moderate degree of jaundice. His 
 urine contained 2*8 per cent, of dextrose. By diet the sugar was 
 reduced in amoixnt, and on December 31, 1909, had c|uite disappeared. 
 Repeated examinations were made, but the urine remained sugar-free 
 for over two years, although he had discarded all dietetic restrictions, 
 and took some 14 ovinces of bread a day, in addition to other starchy 
 food. Even with 100 grams of dextrose, taken fasting, no glycosuria 
 
174 GLYCOSUKIA 
 
 resulted. Subsequently it was foiind that the patient was again 
 excreting sugar, but in very small amount, for although a specimen 
 ]Dassed at 1.30 p.m. reduced Fehling's solution, and gave an osazone 
 melting at 206° C, specimens taken in the early morning, and at 
 :2.30 P.M., gave no reduction. 
 
 In these cases it would seem probable that the temporary 
 glycosuria, and in Garrod's case the jaundice also, was due to a 
 subacute exacerbation of a chronic pancreatitis, which slowly 
 3)rogressed, and ultimately so interfered with carbohydrate meta- 
 bolism as to induce the constant presence of sugar in the urine. 
 
 Hirschfeld has recently reported cases of diabetes which de- 
 veloped after some infectious disease. The diabetes was accom- 
 panied by gastrointestinal symptoms, but there was no pronounced 
 polyuria, although the glycosuria lasted for several months. An- 
 other special feature of the cases was the accompanying enlarge- 
 ment of the liver, which gradually subsided completely. In a 
 still more recent case the onset of the diabetes occurred six weeks 
 after a severe infectious sore throat ; there was also enlargement 
 of the liver in this case. In another case, a man of forty-six pre- 
 sented symptoms of diabetes such as accompany a transient affec- 
 iiion of the pancreas. He had had previously numerous attacks 
 of influenza. After a concussion, due to jumping from a moving 
 car, the diabetes returned in more severe form and became chronic. 
 These and other experiences have convinced Hirschfeld that in- 
 fection is an important factor in the origin and aggravation of 
 pancreatitis and diabetes, and that chronic pancreatitis, such as 
 is found so frequently in diabetics, generally has developed on the 
 basis of repeated acute or subacute inflammations. Glycosuria 
 lasting only a few weeks or months corresponds to the acute tran- 
 sient pancreatitis. 
 
 Temporary glycosuria following an attack of mumps, probably 
 •due to involvement of the pancreas, has been reported by Barbieri, 
 and I have had a similar case. The occurrence of temporary 
 glycosuria and acidosis in a man of fifty-eight, in association with 
 unilateral parotitis, Avas recently reported by Routh. Harris has 
 described two cases of diabetes in which sugar was first found in 
 the urine shortly after an attack of mumps. 
 
 The association of temporary glycosuria with gall-stones is 
 most probably to be referred to the pancreatitis that so commonly 
 accompanies the presence of calculi in the common bile duct. 
 Hochhaus has published a case in which a woman passed 33 grams 
 of sugar in twenty-four hours during an attack of hepatic colic ; 
 the next, and four following days, 4 to 5 grams were excreted ; 
 
TEANSITORY GLYCOSURIA 175 
 
 but after that the sugar disappeared. It is, however, a compara- 
 tively rare compHcation. Zinn only found sugar twice in the 
 urines of eighty-nine cases of hepatic colic, and Kausch three times 
 in eighty-five cases. In an examination of the urines from 396 cases 
 of biliary calculi, I met with sugar in forty-two ( 1 1 per cent . ) , usually, 
 however, only in traces. Of the forty-two cases in which sugar 
 was found, twenty-seven were operated on, and in eight the urine 
 was found to be sugar-free when it was examined a week or ten 
 •days later, but in nineteen sugar was still present. I have not 
 been able to obtain the after-histories of many of these cases, but 
 one at least in which the sugar disappeared after the ojaeration 
 developed a permanent glycosuria a few months later, and ulti- 
 mately died of diabetic coma. It must not therefore be too readily 
 assumed that the improvement that follows operation, when there 
 is obstruction of the bile and pancreatic ducts, is due to the relief of 
 tension and subsidence of the associated catarrhal inflammation, for 
 therest in bed and comparative starvation are often important factors 
 in bringing about the disappearance of the glycosuria, which may 
 return when the patient resumes his ordinary diet and mode of life. 
 An example of the association of temporary gtycosuria with 
 acute pancreatitis, in which recovery took place but the patient 
 ultimately became diabetic, has been recorded by Gilford Nash. 
 
 The patient was a man of sixty who had suffered for seven years 
 from " bilious attacks," and discomfort at the pit of the stomach. 
 On October 27, 1901, he was seized with sudden pain in the abdomen. 
 There was no jaundice, and the symptoms suggested intestinal ob- 
 struction. The urine was increased in amount, and contained on 
 November 5th 8'75 grains of sugar per ounce. Operation was under- 
 taken on November 17th. The pancreas was found to be enlarged, 
 there was fat necrosis in the neighboLirhood of the gland, and a large 
 ■calculus was found in the gall-bladder. Cholecystotomy was per- 
 formed, and the patient slowly recovered. On December 28th the 
 Tirine contained 4*5 grains of sugar to the ounce, and, on March 1, 1902, 
 4'5 grains also, but on May 17th the glycosixria had disappeared. In 
 November 1902 I examined a specimen of the urine and found that 
 the sugar had returned, and that it gave a well-marked " pancreatic " 
 reaction. A second specimen examined in February 1904 gave similar 
 results. It was then stated that the patient was in very good health, 
 and had had no illness since his operation. In January 1906 I made a 
 further examination and found that the urine contained 0*95 per cent, 
 of sugar, and no acetone or aceto -acetic acid, but it still gave a positive 
 pancreatic reaction. 
 
 Marwedel has recorded a case of abscess of the j)ancreas in 
 which sugar was excreted for a few days only. 
 
176 GLYCOSURIA 
 
 Intermittent glycosuria has been met with in some cases of 
 pancreatic calcuh, owing probably to a temporary inhibition of 
 the functions of the joancreas from transient attacks of pancrea- 
 titis. Thus in a case rex3orted by Lancereaux sugar was found 
 in the urine with each attack of colic, but in the intervals it was 
 sugar-free, and in a case recorded by Holzmann sugar was found 
 in the urine at intervals, although when the patient was under 
 Minnich's care sometime earlier repeated examinations had given 
 a negative result. 
 
 The temporary appearance of sugar in the urine met with in 
 some cases of cancer of the pancreas is explicable on similar Hues, 
 the inflammatory reaction attendant on the spread of the growth 
 causing temj)orary damage, which subsequently quiets down, and 
 leaves sufficient unaltered tissue to carry on the work of carbo- 
 hydrate metabolism. In a case of this description quoted by Oser 
 there was transient glycosuria for seven months, then eleven, 
 months cachexia %vithout glycosuria. After death, which took 
 place twenty- three months after the onset of the disease, a large 
 hard cancerous mass was found occupying the head, and haK the 
 body, of the pancreas, the rest of the gland appeared to be normal. 
 It has to be borne in mind that where a portion of the jaancreas 
 has been destroyed by growth, or inflammatory changes, the con- 
 dition resembles that produced in animals by partial extirpation 
 of the gland, so that, if carbohydrates are excluded from the diet, 
 an alimentary glycosuria which previously existed may disappear. 
 
 BIBLIOGRAPHY 
 
 Achard and Castaigne, Bull. soc. med. d. hopit. d. Paris, 1897 ; 
 
 Semaine med., 1897 ; Arch. d. Medicine, 1898. 
 Arnat, Deut. Zeit. f. Nervenkrank, 1897. 
 Arndt, Berl. klin. Woch., 1898 ; Deut. Zeit. f. Nervenheilk., 1897^, 
 
 1899. 
 Barbieri, Oazz. degli Ospedali, 1909. 
 Bauer, Wiener med. Woch., 1906-1908. 
 Baylac, Arch. f. Verdauungskrank, iv. 
 
 Baylac and Arnold, Com2:>t. Rend. d. Med., Toulouse, 1902. 
 Bettniann, MiXnch. med. Woch., 1896. 
 Binet, qiTOted Lepine, Diabete Sucre, 1909. 
 Bloch, Zeit.f. klin. Med., 1892. 
 Blumenthal, Path. d. Harnes, 1903. 
 Bouchard, Traite d. Physiol. Oen., iii. 
 Brion, Zeit. J. phys. Chem., 1898. 
 Bruel, Arch. f. spec. Path. u. Pharm., 1898. 
 
TRANSITORY GLYCOSURIA 177 
 
 Bruining, Berl. klin. Woch., 1902. 
 
 Brun, Riforma Medica, 1910. 
 
 Bur del, Union Medical, 1872. 
 
 Calamette, Gaz. hebdoma, 1882. 
 
 CampagnoUe, Deut. Arch. /. klin. Med., Ix. 
 
 Chajes, Deut. med. Woch., 1904. 
 
 Cohen, New York Med. Journ., 1894. 
 
 Colrat, Lyon medicale, 1875. 
 
 Frey, Zeit. f. klin. Med., Ixxii. 
 
 Garrod, Lancet, 1912. 
 
 Ginsberg, Arch. f. d. ges. Physiol., 1889. 
 
 Gobbi, II PolicUnico, 1900. 
 
 Goodinan, Journ. Amer. Med. Assoc, 1909. 
 
 Grosz, Jahrb. f. Kinderheilk., 1892. 
 
 De Haan, Arch. f. Verdauungskrank, 1898. 
 
 Haedke, Deut. med. Woch., 1900. 
 
 Von Halasz, Wiener klin. Woch., 1908 ; Deut. med. Woch, 1908. 
 
 Harris, Boston Med. and Surg. Journ., 1899. 
 
 Hibbard and Morrissey, Journ. of Exp. Med., 1899. 
 
 Hill, Arch, of Med., 1861. 
 
 Hirschfeld, Deut. med. Woch., xxxv. 
 
 Hochhaus, Deut. med. Woch., 1907. 
 
 Hofmeister, Arch. f. exp. Path., 1889, 1890. 
 
 Hohlweg, Deut. Arch. f. klin. Med., xcvii. 
 
 Holzmann, Munch, med. Woch., 1894. 
 
 Hoppe-Seyler, MiXnch. med. Woch., 1900. 
 
 Von Jaksch, Prag. med. Woch., 1892, 1895 ; Zeit. f. Heilkunde, xx. 
 
 Judson, Lancet, 1902. 
 
 Kausch, Deut. med. Woch., 1899 ; Zentralb. f. Chir., 1904. 
 
 Klippel, Vigoroux, and Juquelier, Arch. f. Neurolog, 1902. 
 
 Knopf elmacher, Wiener klin. Woch., 1904. 
 
 I^aus and Liidwig, Wiener klin. Woch., 1891. 
 
 Lancereaux, Journ. d. Med. enterne, 1889. 
 
 Landsberg, Deut. med. Woch., 1903. 
 
 Leidy, Medical News, 1894. 
 
 Lepine, Diabete Sucre, 1909. 
 
 Manchot, Monats. f. prakt. Dermatol, 1898. 
 
 Marie and Guillain, Soc. med. d. hopitaux d. Paris, 1901. 
 
 Marwedel, Munch med. Woch., 1901. 
 
 Mayer, Berl. klin. Woch., 1899 ; Deut. med. Woch., 1901 ; Zeit. f. 
 
 klin. Med., 1902. 
 Mayer and Pick, Berl. klin. Woch., 1902. 
 Moritz, Arch. f. klin. Med., 1890. 
 Mosse, Verh. d. Vereins. f. inn. Med., 1902. 
 Muenzer and Palma, Zeit. f. Heilkunde, 1894. 
 Nagelschmidt, Berl. klin. Woch., 1900. 
 Nash, Lancet, 1902. 
 Neuberg and Wohlgemuth, Zeit. f. phys. Chem., 1902. 
 
 M 
 
178 GLYCOSURIA 
 
 Niepraschk, Inaug. Dissert., Berlin, 1898. 
 
 Nobecourt, Compt. Bend. d. I. Soc. de Biol., 1900. 
 
 Von Noorden, Zeit. f. klin. Med., xxxviii. ; Arch. f. Anat. u. Phys., 
 
 1893. 
 Oordt, Munch, med. Woch., 1898. 
 Oser, Nothnagel's Encyclop. of Pract. Med., 1903. 
 Ott, Internal Secretions, 1910. 
 
 Paris and Dobrovici, Presse Medicale, 1892, 1905. 
 Parry, Lancet, 1895. 
 
 Pasteur, Compt. Rend. d. I. Acad. d. Sci., 1858, 1860. 
 Poll, Fortschr. d. Med., 1896. 
 
 Pratt and Spooner, Journ. Anier. Med. Assoc, 1910. 
 Raphael, Zeit. f. klin. Med., 1899. 
 Robin, Bull. d. I. Acad. d. Med., 1901. 
 Robinson, Compt. Rend. d. I. Soc. d. Biol., 1899. 
 Rosenberg, Inaug. Dissert., Berlin, 1897 ; Centralb. f. inn. Med., 
 
 1900 ; Dezit. med. Woch., 1906. 
 Routh, Brit. Med. Journ., 1912. 
 Rubner, Zeit. f. Biol., 1883. 
 Sabatowski, Wiener klin. Woch., 1908. 
 Sachs, Zeit.f. klin. Med., 1899, 1900. 
 Simon, Clinical Diagnosis, 1900. 
 Strasser, Wiener med. Presse, 1894. 
 Strauss, Berl. klin. Woch., 1899 ; Zeit. f. klin. Med., 1900 ; Deut. 
 
 med. Woch., 1897, 1901. 
 Teschemacher, Deut. med. Woch., 1895. 
 Valmont, These de Paris, 1897. 
 Vannini, Rivista critica di Clin. Med., 1902. 
 Voit, Deut. Arch.f. klin. Med., 1897. 
 Walko, Zeit.f. Heilkunde, 1903. 
 Wille, Deut. Arch. f. klin. Med., 1899. 
 Wohlgemuth, Zeit.f. phys. Chem., 1902. 
 Zinn, Centralb. f. inn. Med., 1898. 
 Ziilzer, quoted, Blumenthal Path. d. Harnes, 1903. 
 
CHAPTER VI 
 
 PERSISTENT GLYCOSURIA — URINARY CHANGES, BLOOD 
 AND CLINICAL SYMPTOMS 
 
 The continuous elimination of sugar in the urine is most commonly- 
 one of a complex of symptoms to which the name diabetes mellitus 
 is applied. In clinical medicine it is usual to distinguish transi- 
 tory, and intermittent, from persistent glycosuria, and many again 
 .separate the latter from diabetes, but it is important that it should 
 be clearly recognised that each is but a phase of the same meta- 
 bolic error, and that no hard and fast line can be drawn between 
 them. Transitory glycosuria is, as we have seen, evidence that 
 the metabolic powers of the body are not capable of dealing with 
 the amount of carbohydrate contained in the food. If the glycos- 
 uria results from an excessive intake of carbohydrate it is not 
 necessarily evidence of a pathological state, although the appear- 
 ance of sugar in the urine after a starchy, as distinguished from a 
 .sugary, diet points to there being defective metabolism, and the 
 patient must be regarded as a potential diabetic. )Should the 
 glycosuria occur with a normal intake of carbohydrate, it is un- 
 doubtedly an evidence of disease, which may be permanent, or only 
 temporary. If the metabolic disturbance is of a transient nature, 
 and disappears with the removal of the cause, the patient may 
 be little or none the worse, but if it persists his tolerance for carbo- 
 hydrate is likely to be still further lowered with the lapse of time, 
 so that eventually j)ersistent hyperglycsemia and glycosuria will 
 result. When this condition is established, the natural tendency 
 is for it to progress, so that eventuallj=' the secondary effects of an 
 excess of sugar in the blood, and the abnormal tissue destruction 
 that it brings in its train, come into evidence. 
 
 The sugar excreted in persistent glycosuria is as a rule dextrose, 
 but in some cases levulose and other sugars are found in addition. 
 The quantity of the latter is, however, always relatively small, 
 and the exact cause and significance of their presence is not under- 
 stood. 
 
 The amount of dextrose that appears in the urine varies very 
 much, but rarely exceeds 200 grams a day on an ordinary mixed 
 
180 GLYCOSUEIA 
 
 diet. Occasionally a much larger quantity is met with, as in the 
 case of a diabetic of nineteen reported by Naunyn, who passed 
 1200 grams of sugar in the twenty-four hours, and in Dickenson' & 
 case where 1500 grams were excreted daily. Niedergesass states 
 that a child of twelve under his care passed 587 grams of sugar 
 in the twenty-four hours, a quantity corresponding to more than 
 3-8 per cent, of his body-weight. When carbohydrates are ex- 
 cluded from the diet, even the most severe cases rarely pass more 
 than 100 grams a day. In mild cases the sugar excretion increases 
 after food, usually reaching its maximum two or three hours after 
 a meal, and diminishes during fasting, hence the urine passed in 
 the early morning, before breakfast, may be sugar-free. In severe 
 cases the variation is less marked, and more sugar may be excreted 
 during the night than in the day. Muscular exercise reduces the 
 amount of sugar in the urine in well-nourished individuals, and 
 massage has the same effect, but in the more severe forms, where 
 the patient is wasted, as much, or even more, may be excreted after 
 exercise than before. Psychic influences undoubtedly affect the 
 sugar excretion in some cases, the amount being increased by worry, 
 nervous excitement, shock, &c. The output of sugar is usually 
 greater in hot than in cold weather. 
 
 Intercurrent affections may influence the glycosuria, some 
 increasing, while others diminish, it. The influence of acute febrile 
 diseases in this respect is very variable, the effect produced appear- 
 ing to depend on the diet and temperature on the one hand, on 
 the extent of the toxa?mia, and consequent interference with the 
 nutrition of the tissues, on the other. Thus in a short sharp in- 
 fection, such as the cases of follicular tonsilitis described by 
 V. Noorden and Mohr, the sugar is increased, but in more chronic 
 conditions, such as pneumonia and influenza, where the diet is 
 restricted and the temperature high, the sugar may diminish, or- 
 even disappear, to return again when convalescence is established,. 
 In a similar way the sugar is seen to diminish in some cases when, 
 the patient is attacked by enteric fever. But in this, as in other 
 diseases where there is an affection of the gastro-intestinal tract, 
 the interference with food absorption probably influences the 
 sugar excretion in the urine. Of the chronic diseases complicating, 
 diabetes the most common is pulmonary tuberculosis, and with 
 this the glycosuria often diminishes, and sometimes entirely ceases 
 a short time before death. When granular kidney is associated 
 with chronic glycosuria the sugar may slowly diminish, and even 
 disappear, leaving only the symptoms of the intercurrent affection. 
 It has been suggested that the disappearance of the sugar in such. 
 
PERSISTENT GLYCOSURIA 181 
 
 cases is due to interference with the excretory functions of the 
 kidneys, biTt it was shown by Strauss that the serous effusions 
 of such cases are not particularly rich in sugar, and according to 
 Lepine there is no hyperglycsemia, but rather the reverse, so that 
 it is more likely that the sugar disappears in consequence of the 
 cachexia that results from the renal changes. According to 
 Leoorche, the excretion of sugar in diabetic womsn is temporarily 
 diminished at each menstrual period. A rare but striking cause 
 of an apparent diminution, or disappearance, of the sugar is the 
 occurrence of fermentative changes within the bladder. This de- 
 pends upon infection with yeasts, or fermentative bacteria, which 
 split up the dextrose into alcohol and carbon dioxide, and, in 
 addition, give rise to hydrogen, carburetted hydrogen, and other by- 
 products. Apart from diet and intercurrent affections, the sugar 
 output is liable to undergo spontaneous variations of considerable 
 amount, which can only be explained by fluctuations in tolerance. 
 
 The volume of the urine passed in persistent dextrosuria is nearly 
 always increased, varying as a rule with the severity of the case. 
 Polyuria is generally not marked unless the urine contains 2 to 
 3 per cent, of sugar. Three or four litres a day is a common amount 
 to be passed, over 5 litres is rare ; but Naunyn has reported one 
 case that passed 16 litres, Fiirbringer 17 litres, Harnack 18 litres, 
 and Bence Jones 28 litres. In some instances the amount of urine 
 excreted does not exceed 1500 c.c. in the twenty-four hours, although 
 a considerable amount of sugar, 20 to 30 grams for instance, is 
 present. Such cases have been described by Frank as " diabetes 
 decipiens." Generally the amount of urine passed rises with an 
 increase in the output of sugar, but at a slower rate. Sometimes, 
 however, the volume is augmented without any more sugar being 
 excreted, and occasionally the reverse occurs. A strict carbo- 
 hydrate-free diet will reduce the quantity of urine, and it is also, 
 diminished by intercurrent febrile affections, diarrha3a, &c. It 
 is important to note that several days before the onset of chabetic 
 coma, and just prior to a fatal termination, a marked reduction 
 often takes place. In health more urine is excreted as a rule during 
 the day than during the night, but with most diabetics this dif- 
 ference is not so marked. In some cases, and particularly those 
 of a mild type, the night urine is markedly less than the day urine, 
 but if the condition is comj)licated hy arterio-sclerosis and granular 
 kidney the reverse condition is often present, especially in the 
 later stages. 
 
 A'ppearance. — When a urine containing sugar is passed it is 
 usually bright and clear. It froths more readily than a normal 
 
182 GLYCOSURIA 
 
 urine, and the froth is more persistent. If the quantity is not- 
 increased, or only slightly, the colour is not altered, but when there 
 is polyuria it is generally of a slight straw colour, and often has 
 a characteristic greenish tint when examined against a white back- 
 ground. On standing, diabetic urine speedily becomes turbid , 
 from the growth of yeasts and fungi. It is often noticed, too, that 
 the " mucous " cloud, instead of forming at the bottom of the vessel 
 as in normal urine, is susjoended in the upj)er layers. A peculiar 
 sweet, or aromatic, odour is frequently observed, and this is parti- 
 cularly^ noticeable when coma has supervened, or is threatened. 
 
 Reaction. — ^When freshly passed diabetic urine is nearly always 
 acid in reaction, often markedly so, esjDecially in advanced cases, 
 and immediately before the onset of coma. When allowed to- 
 stand it remains acid for several daj^s, and may in fact even increase 
 in acidit}'' from the conversion of some of the sugar into lactic acid 
 through the agency of micro-organisms. 
 
 Density. — The specific gravity is usually high, ranging between 
 1-030 and 1-040, but it rarel}^ exceeds 1-050. Bouchard and Prout 
 mention a case that passed a urine with a specific gravity of 
 1*074. When the specific gra\TLty of a urine is over 1-025, and 
 it is clear and not high-coloured, it is probable that it contains 
 sugar. As a rule, the more sugar there is present the higher is the 
 specific gravity, and vice versa. As the density does not depend 
 upon the sugar-content alone, but is also influenced by the amount 
 of other solids in solution, the specific gravity cannot, however, 
 be rehed upon as an index of the quantity of sugar present. It 
 must be clearly understood that a normal, or even a low, specific 
 gravity does not exclude the 23resence of sugar, for in some 10 per 
 cent, of cases of persistent glycosuria a sub-normal specific gravity 
 is found, 1-012 to 1-006, or even lower. A low specific gravity 
 is most commonly met with in cases where the absorption of nitro- 
 genous material from the intestine is defective, in the glycosuria 
 follo^\ing injuries of the head (in which, according to Naunyn, a 
 specific gravitj^ of 1-003 and a sugar-content of about 1 per cent. 
 is often met wdth), associated with early chronic interstitial nephritis, 
 and when the patient is very weak. In a few cases of diabetes ex- 
 ceedingly low readings have been recorded ; thus in a case reported 
 by Waterman there was a specific gravity of 1-002, and Herrick met 
 -with a case in which the density of the urine was only 1 -004. 
 
 Total Nitrogen. — As we have seen, a healthy inchvidual on 
 an ordinary mixed diet, taking a fair amount of exercise, passes 
 from 10 to 16 grams of nitrogen, "wdth an average of 15 grams, in 
 the twenty-four hours. This comprises by far the greater part of 
 
PERSISTENT GLYCOSURIA 183 
 
 the nitrogenous loss of the body, less than 1 gram being eliminated 
 through the intestinal secretions and other channels combined. 
 In diabetes 30, 40, or even 50 grams of nitrogen may be excreted 
 in the twenty-four hours. The cause of this increase in the nitrogen 
 excretion, and its relation to the sugar output, will be considered 
 later. Normally the urinary nitrogen is distributed in various 
 compounds as follows : Urea about 86 per cent., ammonia about 3 
 per cent., creatinin about 3 per cent., uric acid and allied xanthin 
 bases about 2 per cent., the remaining 6 per cent, being furnished 
 in various proportions by substances such as hippuric acid, indol, 
 skatol, &c. The variations in these proportions, and in the total 
 quantities of the nitrogen-containing constituents met with in 
 persistent glycosuria may be summarised as follows : — 
 
 Urea. — The urine of diabetics usually contains a sub-normal 
 proportion of urea, but this is due to the polyuria, and, when the 
 total excretion for the whole twenty-four hours is considered, an 
 excess is frequently found, 50 grams or more being often passed 
 in the twenty-four hours. The increase is no doubt due, to a 
 certain extent, to the large quantities of nitrogenous food con- 
 sumed, but the experiments of Pettenkofer and Voit have shown 
 that diabetics usually excrete more urea than normal individuals. 
 Seegen concludes (a) that the urea excretion is increased in almost 
 all cases of diabetes, but generally not markedly ; (b) there is no 
 relation between the excretion of urea and sugar ; (c) the urea 
 excretion is generally chiefly dependent upon the nitrogen of the 
 food, and in only a few cases is it so abundant that it is necessary 
 to consider that it is derived from the body proteins. 
 
 Hirschfeld has pointed out that in some cases of diabetes the 
 resorption of nitrogenous material from the intestine, and with it 
 the elimination of urea, may be very much below normal, and 
 upon these grounds has advocated the recognition of a distinct 
 form of diabetes characterised by a comparatively rapid course, 
 the occurrence of colicky abdominal pains before, or at the onset, 
 of the diabetic symptoms, a moderate degree of poljmria, and the 
 existence of j)ancreatic lesions. 
 
 Ammonia. — The urine of a healthy person contains as a rule 
 less than 1 gram of ammonia in the twenty-four hours, and nor- 
 mally only about 2 to 5 per cent, of the total nitrogen of the urine 
 exists in the form of ammonia. The amount excreted depends, 
 however, upon the diet to some extent. It is much diminished 
 in vegetarians, in whom the ammonia nitrogen represents about 
 2 to 3 per cent, of the total urinary nitrogen, and is increased by 
 a solely meat diet, when it may rise to 1-2 to 1"5 grams a day, 
 
184 GLYCOSURIA 
 
 the ammonia nitrogen then representing about 5 per cent, of the 
 total nitrogen. In diabetics, who are taking a large amount of 
 meat, the ammonia nitrogen is increased above the average from 
 that cause ; but in some cases a quantity in excess of what can 
 be accounted for by the nature of the diet is passed, while in others 
 the output goes up ^^dthout anj^ increase in the nitrogenous food 
 to 5 or 6 grams a day, representing 10 to 25 per cent, of the total 
 nitrogen. In such cases there is a corresponding decrease in the 
 excretion of urea. The reason for this has already been referred 
 to, and will be more fully discussed -when acidosis and diabetic 
 coma are considered, since it is as a sign of these that an increase 
 in the output of ammonia nitrogen is chiefly important. 
 
 Creatinin. — ^An increased excretion of creatininin the urine, gene- 
 rally said to arise partly from the nitrogenous diet and partly from 
 muscular wasting, is seen in many cases of persistent glycosuria, 
 as much as 2 grams in the twenty-four hours being met with. 
 Mendel and Rose state that carbohydrates, in contrast to other 
 foodstuffs, are capable of jDreventing the excretion of creatinin, 
 and are indisj)ensable for creatin-creatinin metabolism. Thej^ 
 found that experimental interference with carbohydrate meta- 
 bolism, as in 23hloridzin diabetes, leads to the excretion of creatin, 
 and increases the output of total creatinin (creatin plus creatinin). 
 This increase is always accompanied by a rise in the total nitrogen 
 excretion. They ascribe the parallelism between the total creatinin 
 and total nitrogen outputs to true tissue, or endogenous, metabolism, 
 while they came to the conclusion that the metabolism of exo- 
 genous, or reserve, protein is not accompanied by the production 
 of creatin or creatinin. 
 
 Uric Acid. — It has been frequently stated that the urine in 
 diabetes contains only a small amount of uric acid, but Naunjnn 
 and Reiss showed that this is an error due to incorrect estimation, 
 and that the daily quantity is usually increased. This increase 
 is, as a rule, dependent upon the exogenous uric acid derived 
 from the meat diet, however, and Burian and Schur and others have 
 found that in some instances the endogenous uric acid is diminished. 
 Occasionally cases are met mth in which there is a marked increase 
 in the excretion of uric acid, amounting to as much as 3 grams 
 in the twenty-four hours, associated with a diminution or dis- 
 appearance of the sugar. To this condition the name " diabetes 
 altemans " has been appUed. 
 
 Xanthin Bases. — According to the observations of Bischof- 
 swerder, and of Jacoby, the xanthin bases (xanthin, guanin, &c.) 
 are increased in most cases of diabetes. 
 
PERSISTENT GLYCOSURIA 185 
 
 Hippuric Acid and Benzoic Acid are both said to have been 
 found in diabetic urines in demonstrable quantities. 
 
 Tyrosine is not met with in normal urines, but has been de- 
 scribed as present in the urine of diabetic ]oatients hy Mies, Nicola, 
 Adberhalden, and Mohr. 
 
 Indican. — According to Moraczewski, the amount of indican 
 in the urine is increased in most diabetics, often very largely so. 
 This is to be explained by the meat diet, and by the gastro-intes- 
 tinal disturbance and hepatic insufficiency that are so common. 
 
 Glycocoll has been met with in the urines of two diabetics by 
 Mohr. 
 
 Lactic Acid. — The quantity of lactic acid in normal urine is 
 very small, under 0-02 grams per day, but after an abundant carbo- 
 hydrate meal, and as a result of the administration of lactates by 
 the mouth, the excretion may rise to five times the normal. In 
 diabetes there may be an increase in the excretion of lactic acid 
 in the urine, 0-05 to 0-06 grams being cometimes found. Both in 
 health and disease it seems probable that the lactic acid that 
 appears in the urine originates from the sugar of the blood through 
 imperfect oxidation. 
 
 Oxalic Acid. — An abundant deposit of calcium oxalate crystals 
 is often seen in both mild and severe cases of diabetes. Bose 
 states they were detected in the sediment from the urine in 26-5 per 
 ■cent, of his cases, and I have found them in over 20 jaer cent, of the 
 cases that have come under my observation. Although an increased 
 excretion of oxalic acid cannot be argued from the presence of 
 calcium oxalate crystals in the urinary sediment, for other factors 
 than the amount of acid determine their separation in cryotalHne 
 form, it has been shown by quantitative estimations that the ex- 
 cretion of oxalic acid is frequently increased in diabetes. Thus 
 Prerichs found* 0-6 grams, as compared with the normal daily 
 excretion of about 0-015 to 0-02 grams, in one case, and in this 
 instance it was noticed that the output of oxalic acid and of uric 
 acid ran curiously parallel. Barth and Autenrieth, Luzzato, and 
 others have, on the other hand, reported cases in which the ex- 
 cretion was distinctly sub-normal. In some cases of diabetes a 
 so-called " vicarious " elimination of calcium oxalate has been 
 noticed, a diminution in the amount of sugar being associated 
 with an increased deposit of oxalate crystals, and an increase in 
 the sugar output being accompanied by a cUminution or disappear- 
 ance of the oxalates. This has been taken to indicate that glycos- 
 uria and oxaluria are two phases of the same condition, and that 
 there is a relationship between carbohydrate metabolism and 
 
186 GLYCOSURIA 
 
 oxalic acid. The experiments of Mayer tend to support this view, 
 for he found oxalic acid in the liver and urine of rabbits after they 
 had been injected with dextrose, or glucuronic acid. The frequent 
 association of oxaluria with cirrhosis of the joancreas, which I 
 pointed out some years ago, also suggests that the appearance of 
 oxalates in the urine may be a result of the incomplete oxidation 
 of carbohydrates. Since gastro-intestinal disturbances are very 
 apt to be associated with oxaluria, probably in consequence of 
 the defective digestion and subsequent imperfect oxidation of 
 carbohydrates, the influence of such intercurrent conditions must 
 be taken into account in explaining the oxaluria of diabetes and 
 pancreatic disease, both of which are frequently accompanied by 
 disturbances of digestion. The similar relations which the excre- 
 tions of oxahc and uric acid bear to the intensity of the dextros- 
 uria may possibly be accounted for by the fact that oxaluric acid, 
 AA^hich is readily decomposed into oxalic acid, can be derived from 
 uric acid. 
 
 Phosphoric Acid. — The quantity of phosphoric acid excreted in 
 the urine is largely dependent upon the amount ingested, increasing 
 ■with an animal, and decreasing \vith a vegetable, diet. It is also 
 influenced by the amount of tissue destruction. The character 
 of the individual phosphatic salts depends upon the alkalinity of 
 the blood, and ultimately on the quantity of acid set free in the 
 tissues, or absorbed during digestion. In many cases of diabetes 
 the excretion of phosphates, chiefly the earthy phosphates, is con- 
 siderably increased. Boecker in one case found that the earthy 
 phosphates were about three times the normal, and Neubauer met 
 with 0'711 grams of phosphate of lime, in twenty-four hours, in the 
 urine of a child of six, an amount more than double the normal 
 for an adult. Teissier drew attention to a curious relation that 
 exists between the elimination of sugar and phosphates in some 
 cases, the quantity of the latter rising and falling in m verse ratio 
 to the former. Occasionally a condition known as " phosphates 
 diabetes " is met ^%-ith. In this, although sugar is absent from 
 the urine, the patient presents various symptoms commonly asso- 
 ciated with diabetes mellitus, and there is an increased excretion 
 of phosphates, amounting to 7 or even 9 grams in the twenty-foiir 
 hours. 
 
 Chlorides. — As the chlorides that are excreted in the urine 
 are derived from the food, the animal diet is apparently the ex- 
 planation of the increased output of chlorides that is generally 
 seen. 
 
 Sulphates. — An increase in the excretion of sulphates in dia- 
 
PERSISTENT GLYCOSURIA 187 
 
 betes is, like an increase in the excretion of indican, to be accounted 
 for by the abnormal putrefactive changes that go on in the intestine 
 as a result of the highly nitrogenous diet and the associated gastro- 
 intestinal disturbances ; for the etheral sulphates, to which the 
 increase is largely due, are one of the means which the body adopts 
 for harmlessly removing the toxic products of intestinal putrefaction. 
 
 Calcium. — A healthy adult excretes about 3 mg. of calcium 
 per kilogram of body-weight daily, but in many diabetics, and 
 more especially in the grave forms, this amount is very consider- 
 ably exceeded, as Boecker, Dickinson, and others have shown. 
 The excess has been attributed by Tenbaum to the nature of the 
 diet, and, although this may be a partial explanation, there can 
 be no doubt that a large part of the excess is derived from the 
 osseous system in an attempt to neutralise the abnormal acids 
 formed as a result of the perverted metabolism. 
 
 Iron. — The excretion of iron is increased in many cases of 
 diabetes. Jolles and Winkler found in four cases that it oscillated 
 between 83 and 136 mg. in the twenty-four hours. Neumann and 
 Mayer state that the excretion of iron is jjroportional to that of 
 sugar, and, although such a result would be of interest as bearing 
 on the anaemia of diabetes, it has not as yet been confirmed. 
 
 Albumen. — Some authors consider that albuminuria is frequently 
 met with in diabetes, while others are of the contrary opinion. 
 This difference appears to depend partly upon the interpretation 
 of the term that has been adopted, and partly upon the class of 
 case from which the statistics have been drawn. If only a decided 
 reaction for albumen, such as is given by 1 gram or more per litre, 
 is taken into account, it is found that only eome 8 or 10 per cent, 
 of cases suffer from albuminuria, according to the extensive experi- 
 ence of Kiilz and Lepine. If, however, a slight opalescence is also 
 accepted as evidence of albuminuria, some 66 per cent, of cases must 
 be included, as in Schmitz's statistics. For clinical purposes it is 
 therefore advisable to divide the cases into two groups : (1) those 
 in which only a small quantity of albumen is present, and there 
 are no other indications of kidney mischief ; (2) those in which 
 there is a characteristic reaction for albumen, and in addition 
 other signs of nephritis. 
 
 The former are by far the most common. They include those 
 cases in which, according to NaunjTi, the albuminuria arises from 
 the effect of the glycosuria on the kidneys, and also those in which 
 a trace of albumen comes from a small amount of purulent material, 
 mixed with the urine, that is derived from the discharge exuded by 
 the external genitals consequent on the irritation of the saccharine 
 
188 GLYCOSURIA 
 
 urine. Slight albuminuria is also often observed Avhen phthisis 
 is a complication of diabetes, and in elderiy diabetics a small 
 quantity of albumen is not infrequeiitlj^ met Avith in the urine, 
 accompanied by an excess of uric acid. It has been pointed out 
 by Lepine that albummuria is pari:iculariy frequent in traumatic 
 diabetes, and he suggests that the appearance of albumen in such 
 cases is the result of nervous influences. 
 
 In members of the second group there are all the indications of 
 parenchpnatous, or interstitial, nephritis, the disease of the kidney 
 being in some cases, however, probably antecedent to the glycos- 
 uria. A large quantity of albumen along with casts in the urine, 
 o?deme. headache, retinitis, and cardio- vascular changes pomts to a 
 parenchymatous inflammation of the kidneys. In such cases it is 
 often found that as the albumen increases in amount, the sugar dimi- 
 nishes, until only the former remains. In other cases the sjonptoms 
 of granular kidney are present, and here again the sugar may 
 disapj)ear as the kidney lesion advances. In either type of case the 
 disapijearance of the sugar is a grave jarognostic sign. According 
 to Williamson, albuminaria is more common in private practice 
 than in hospital cases. Avhere. as a rule, the aj)pearance of albumen 
 in the urine is a late symptom. A very common sign of approach- 
 ing diabetic coma is the appearance of albumen, generally accom- 
 panied by many granular and hyaline casts in the urine. According 
 to Maguire. albuminuria is always present in diabetic coma. 
 
 The Acetone Bodies. — ^With the exception of sugar itself, the 
 acetone bodies are the most important substances met ^rith in 
 the urine in persistent dextrosuria, for their presence signifies that 
 the case has reached the second stage in its doTMiward progress. 
 
 Acetone is the only one of the three that is met with in normal 
 urine, but never in quantities that can be recognised by the ordi- 
 narj^ clinical tests. It is the first to appear in appreciable amounts 
 when acidosis is develoiDing. Some authors contend that acetone is 
 never preformed in the urine, but is always derived from aceto- 
 acetic acid, and although this vicAv is not generally accepted, it 
 cannot be denied that a part of the acetone obtained by distilla- 
 tion processes from diabetic urines comes from the contained 
 aceto-acetic acid. According to Embden and Schliep, at most a 
 quarter, often a sixth, and very frequently a tenth, only, of the 
 acetone is preformed. 
 
 Normal urine contains very little acetone, 0-01 to 0-03 grams 
 in the twenty-four hours, but in chronic glycosuria it may rise to 
 0-5 grams, and in severe cases of diabetes ma 5^ reach 1-0 to 4-5 grams, 
 or, with a purely protein diet, even 5 grams in the twenty-four 
 
PERSISTENT GLYCOSURIA 189 
 
 hours. In the latter case a great part of the acetone is undoubtedly 
 derived from aceto-acetic acid. A marked increase in the excre- 
 tion may follow slight fever. Acetone also leaves the body in the 
 expired air, as much as 150 mg. an hour being sometimes got 
 rid of in this way. The acetone in the breath is usually considered 
 to be the cause of its characteristic sweet smell in severe cases of 
 diabetes, but this is denied by Folin. 
 
 Aceto-acetic acid appears in the urine in relatively small quan- 
 tities, rarely exceeding 10 per cent, of the total organic acids ; 
 but, owing to the ease with which it is converted into acetone, 
 it is not easy to determine the exact proportion in which the two 
 occur. They are therefore frequently estimated together. 
 
 Beta-oxyhutyric acid is met with in very variable quantities, the 
 amount depending on the extent of the secondary disturbances of 
 metabolism that are present. In mild cases of persistent glycosuria 
 it is usually absent, but in severe cases of diabetes large quantities 
 are found as a rule. Preceding the onset of diabetic coma 15 to 
 20 grams a day may be met with, and Kiilz has reported that in 
 three cases a daily output of 67, 100, and 226 grams respectively 
 was observed. When coma has developed it is frequently found 
 that the elimination does not keep pace with the formation of the 
 acid, so that considerable quantities are retained within the body, 
 and the quantity in the urine drops. Every urine that contains 
 oxybutyria acid also contains acetone and aceto-acetic acid, but 
 the converse is not true. In a few exceptional cases a certain 
 parallelism is seen between the excretion of beta-oxybutyric acid 
 on the one hand, and aceto-acetic acid on the other for a few days, 
 but an inverse relationship is more commonly found, as might 
 be expected. As a rule, when once oxj^butjTic acid has appeared 
 in the urine it increases at a much more rapid rate than the aceto- 
 acetic acid and acetone, so that 30 to 80 grams may be present, 
 while the total amount of the other two rarely exceeds 7 to 8 grams. 
 The parenchymatous degeneration of the heart, kidneys, and some- 
 times of the liver, seen in fatal cases of diabetes, is attributed by 
 Busse to the action of beta-oxybutjo^ic acid. 
 
 The Blood in Diabetes. — Many observers have found that 
 in cases of diabetes, of apparently equal severity, the number of 
 erythrocytes, and the haemoglobin content of the blood, may be 
 increased in one, and diminished in another, or may vary con- 
 siderably from time to time in the same patient. Most recent 
 investigations have shown that, while there may be considerable 
 individual variation in the percentage of haemoglobin, it does not 
 
190 GLYCOSURIA 
 
 differ greatly from the normal when a series of eases is considered, 
 and the same may be said of the red blood cells, for, although they 
 are often increased, this is not a constant condition, and in some 
 instances they are found to be diminished. Leichenstern, finding 
 an excess of haemoglobin in an advanced case and a diminution 
 in an early case, was led to consider that the determining factor 
 is the concentration of the blood, Avhich in its turn depends upon 
 the diuresis. This exj)lanation has been generally accepted, and 
 it is concluded that any increase in the number of erythrocytes is 
 also to be referred to this cause. James, who carefully investigated 
 the blood in thirteen cases of diabetes, found that in five of them 
 the red corpuscles were increased to six millions, or more, per cubic 
 millime re, in five they were normal, in two they were four millions 
 per millimetre, and in one three millions per millimetre, but con- 
 cluded that the sjDecific gravity of the blood was not distinctly raised 
 in those ^^ith an excess, as it should have been were the concentra- 
 tion of the blood the only cause of the polycythsemia. E^ving criti- 
 cises this conclusion, and considers that the specific gravities cited 
 are distincth^ above normal if allowance is made for the percentage 
 of haemoglobin found, and states that James' results indicate a 
 relative anhydremia ^^ith marked reduction of haemoglobin, and 
 but slight loss of red cells. Although it may be allowed that the 
 marked changes seen in the blood of many diabetics may be referred 
 to the uncertain balance between the amount of water absorbed 
 and that excreted in the urine, other factors undoubtedly enter 
 into the problem in the later stages, for the general failure of nutri- 
 tion that occurs must affect the blood. Even in extreme cases, 
 however, the anaemia that results is frequently masked by the 
 concentration of the blood consequent on excessive diuresis. 
 
 As a rule there is no definite change in the number, or relative 
 proportion, of the leucocytes in diabetes, although, according to 
 V. Limbeck, digestion leucocytosis is often very well marked. 
 Gabritschewskj^ has dra\^ai attention to the existence of an excess 
 of " glycogen " in the leucocjiies, and plasma, in severe cases. This 
 is demonstrated by mounting cover-glass preparations in a solution 
 of iodine (1 part), and potassium iodide (3 parts), in water (100 
 parts), containing an excess of gum-arabic. The significance of 
 the brownish extra cellular granules, which are stated to be two or 
 three times as numerous as in normal blood, has been questioned, 
 however, and it has been pointed out that myelin, lecithin, &c., 
 also stain bro'VMi A\ith iodine. Their significance is therefore not 
 certain. According to Locke, the leucocytes only contain the 
 granules in pathological conditions, and they are particularly 
 
PERSISTENT GLYCOSURIA 191 
 
 abundant in diabetes when coma, or gangrene, exists ; but as they 
 are also found in other diseases, sometimes in large numbers, and 
 more especially when there is sej)tic absorption going on, they 
 are of little diagnostic value. The granules are found chiefly in 
 the polymorphonuclear neutrophiles, but also to some extent in 
 the large and small mononuclear cells. According to some authors 
 they are seen in the eosinophile cells only in diabetes, but Habershon 
 believes that the eosinophile granules themselves are related to, 
 or are identical with, glycogen, and he states that normally from 
 1 to 16 per cent, of all leucocytes contain glycogen granules. Fut- 
 terer claims to have demonstrated thrombi composed of glycogen 
 in the brain and medulla in diabetes. 
 
 To the naked eye the blood in diabetes usually presents no 
 striking variation from the normal, but in rare instances it may 
 have a j)ink colour, or an appearance like chocolat au lait. In such 
 cases a milky, or cream-like, serum sejjarates out on standing, 
 and analysis may reveal as much as 19 per cent., 20 jjer cent., or 
 even more of fat, instead of the normal of about 0-1 to 0-2 per cent. 
 Chemical analysis shows that as a rule the fat consists largely of 
 olein, with a small amount of free fatty acid, but in some cases 
 a considerable j)roj)ortion of cholesterin has also been found, a 
 quarter of the total in a case quoted by Javal. Microscopically 
 the fat is seen to be in the form of fine granules, which stain feebly 
 with osmic acid ; but as all that stains with osmic acid is not fat, 
 a control specimen, that has been thoroughly extracted Mdth 
 alcohol and ether previous to staining, should be compared with 
 the original sample. Even in cases where the blood appears normal 
 to the naked eye an excess of fat may be often demonstrated b}^ 
 the use of the microscope, or the haematokrit. The occurrence of 
 lipeemia in alcohohsm, pneumonia, anaemia, and phosphorus poison- 
 ing, in all of which there is defective oxidation, as well as in dia- 
 betes, suggests that an imi^airment of the power of the organism 
 to oxidise fats is the cause of the condition ; but, as we are not 
 acquainted with the essential steps that take place in the oxidation 
 of fat, it is not possible to suggest how the failure of fat destruction 
 is brought about. The origin of the fat in lipsemia is Hkewise un- 
 certain. Ebstein considers that it arises jmrtly from the food, 
 and partly from the fatty degeneration of the cells of the blood, 
 the vessel walls, and the viscera. The presence of a considerable 
 excess of cholesterin in some instances tends to favour the view 
 that some portion at least has the latter origin. Neisser and 
 Derlin conclude that it is merely fat from the food, coming from 
 the chyle that accumulates in the blood. Fischer beheves that 
 
192 GLYCOSURIA 
 
 it is largely derived from the fat stores of the body, but that, 
 owing to a loss of Ij^Dolj'tic power on the part of the blood, it 
 cannot be rendered diffusible and so enter the tissues, where it 
 is normally consumed. It has been stated that diabetic coma and 
 death in diabetes may be caused by fat embolism of the cerebral 
 vessels. While this is possible, it cannot be a common occurrence, 
 for a marked degree of lipsemia is rare, and even then the fat 
 droplets are too small to cause occlusion of the vessels unless 
 they combine to form large droplets. Fischer doubts whether this 
 can ever occur, and supports his contention by experiments and 
 clinical records. 
 
 The alkalinity of the blood in normal individuals varies between 
 300 and 400 mg. of sodium hydrate per 100 grams of blood. In 
 diabetics mth little general disturbance it is found to be slightly 
 reduced, but in cases with marked general disturbance of meta- 
 bolism it is lower than in any other known condition, falling to 
 40 mg. or so of sodium hydrate per 100 mg. of blood (v. Noorden). 
 The reaction of the blood is, however, never acid to litmus. 
 Brandenburg has pointed out that a distinction must be drawn 
 between the total alkalescence of the blood, referable to the non- 
 diffusible combinations of alkalies with albumen, shown by direct 
 titration, and the " alkaline tension," due to the diffusible alkalies, 
 indicated by the j)roportion of carbon dioxide obtainable by dis- 
 sociation of the carbonates. This distinction is particularly im- 
 portant in diabetes, for it is chiefly in diabetic coma that a marked 
 reduction in the alkali tension is met with. Thus in one case 
 Minkowski found as little as 3-3 volumes of carbon dioxide per 
 cent., as compared with the normal 33-37 to 45-3 per cent. This 
 relative acidsemia is referable to the presence of various acid pro- 
 ducts of fat and protein metabolism, including beta-oxybutyric 
 and other fatty acids. 
 
 The chief chemical alteration in the blood in diabetes is the 
 increase in the amount of sugar it contains. While the sugar 
 content of normal blood varies from 0-05 to 0-15 per cent., Pavy 
 and Seegen have found as much as 0-6 per cent, in cases of severe 
 diabetes, and Hoppe-Seyler reported 0-9 per cent, in one instance. 
 Frerichs found the sugar in the blood in diabetes to vary between 
 0-38 and 0-44 per cent, when the urine contained from 5-5 to 8-4 
 per cent. Xaunyn met with 0-7 per cent, of sugar in the blood 
 of a fatal case of diabetes in which the urine contained 4 per cent. 
 Henriques and Kohsh have claimed that an excess of preformed 
 sugar exists in the blood only in alimentary glycosuria, and that 
 in diabetes it is but slightly above the normal. They consider 
 
PERSISTENT GLYCOSURIA 193 
 
 that in the latter condition most of the sugar exists in combina- 
 tion, the blood containing a marked excess of jecorin combined 
 with albumen, which is split up in the kidneys during excretion into 
 dextrose and lecithin. 
 
 Several tests have been devised with the object of differentiating 
 diabetic from non-diabetic blood. 
 
 Williamson'' s Test. — Wilhamson showed that diabetic blood, 
 like diabetic urine, decolorises solutions of methylene blue, while 
 normal blood does not, and he suggests that this test may be useful 
 in cases where a specimen of urine is not available. 
 
 The test is performed as follows : — 20 cm. (2 drops) of blood are 
 mixed with 40 cm. of water, and added to 1 c.c. of methylene blue 
 (1 : 6000), and 40 cm. of liquor potassse (sp. gr. I*0o8), in a small 
 narrow test-tube, and well mixed by shaking. A control specimen 
 of normal blood is then prepared in the same way. Both test-tubes 
 are suspended in a beaker of water, the contents of which are brought 
 to the boiling-point, and the boiling continued for four minutes. 
 The specimen containing the diabetic blood is then seen to have 
 changed from a fairly deep blue to a dirty, pale yellow, while the 
 specimen in which the normal blood has been added remains blue, or 
 occasionally assiunes a bluish-green, or violet, tint. It is important 
 that the tubes should not be shaken while they are being heated, or 
 after the experiment is completed, as the colour of the diabetic speci- 
 men may be restored through the action of the atmospheric oxygen. 
 Williamson obtained a positive result with all of fifty specimens from 
 twenty cases of diabetes, but failed to obtain the reaction in over 100 
 cases of other diseases. 
 
 Bremer'' s Test, — Another test, introduced by Bremer, has also 
 been supposed to depend upon the hyperglycsemia, but it is more 
 probably due to the presence of abnormal acids, as Schneider and 
 others have shown that the reaction runs parallel with the quantity 
 of these present in the blood. 
 
 This test is carried out as follows. A smear preparation of the 
 blood is fixed, by heating it at 60° C. in. equal parts of alcohol and 
 ether for four minutes. It is then stained for four minutes in a freshly 
 prepared solution containing 0-025 to 0-05 grams of a powder (made 
 by mixing twenty-four parts of the dried and washed precipitate 
 formed when saturated aqueous solutions of eosin and methylene blue 
 are mixed, in about equal proportions, with six parts of methylene 
 blue and one of eosin) in 10 c.c. of 33 per cent, alcohol. After washing 
 the stained preparation in water, diabetic blood should have a greenish 
 tint, while normal blood is reddish -violet. On microscopical examina- 
 tion the erythrocytes in diabetes shoiild appear greenish, normal 
 erythrocytes red. 
 
 Various modifications of this test have been introduced from time 
 
194 GLYCOStJRIA 
 
 to time. Bremer found that 1 per cent, solutions of Congo red, or of 
 methylene blue, acting for one and a half to two minutes, stain diabetic 
 blood, that has been fixed by heating to 125° C. for six to ten minutes, 
 very slightly, but that a 1 per cent, solution of Biebrich scarlet stains 
 it very intensely. Directly opposite effects are obtained with normal 
 blood. Rather thick smears should be employed, and the preparations 
 must not be heated to over 140° C. The colours should be compared 
 with the naked eye. Bremer's results have been confirmed by other 
 observers, but similar reactions have been obtained with the blood in 
 leukeemia, Hodgkin's disease, exophthalmic goitre, and multiple 
 neuritis, and a partial reaction in cachectic conditions. Yet in most 
 conditions other than diabetes the reaction has been found to be in- 
 constant, and to occur only in a small proportion of the cases. It is 
 to be noted that a slight variation in the technique appears to vitiate 
 the results of both Williamson's and Bremer's tests. 
 
 Clinical Symptoms of Persistent Dextrosuria. — In many 
 cases of persistent dextrosuria the presence of sugar in the urine 
 is the most prominent, and, in some, the only sign of the metabolic 
 disturbance that exists, so that the condition is only discovered 
 accidentally in the course of a routine examination for life in- 
 surance, or for some other purpose. This is particularly the case 
 when the patient is middle-aged or stout. Inquiry may show that 
 there has been some loss of weight, but often there is no sign of 
 it, and with the exception, perhaps, of a feeling of lassitude, or a 
 general loss of mental and physical tone, no special symptoms can 
 be discovered. In others the occurrence of one or other of the 
 numerous complications to which diabetics are liable, may draw 
 attention to the presence of sugar in the urine. 
 
 With severe cases the clinical picture is very different. It is 
 obvious when the patient is first seen that he is suffering from some 
 serious wasting disease ; his form is emaciated, his face is pale 
 and sunken, the naso-labial folds are deep and well-defined, and 
 are frequently prolonged round the angles of the mouth. There is 
 sometimes a dull red flush on the cheeks, and the tip of the nose 
 and lips are slightly cyanosed. The expression is listless or 
 anxious, and the patient looks older than his years. The skin and 
 hair are often dry and harsh. The tongue may be moist and 
 covered with a thin yellowish white fur, or may be unnaturally 
 clean, red, glazed, and beefy-looking. Sometimes it is fissured 
 and cracked. The mouth is dry, and thirst is often a prominent 
 symptom. The gums are frequently red and inflamed, and the 
 teeth loose. A voracious appetite, which is only temporarily 
 assuaged by food, may be present, yet, in spite of the large quan- 
 tities of food and drink consumed, indigestion is not complained 
 
PERSISTENT GLYCOSURIA 195 
 
 of as often as might be expected. On being weighed the patient 
 is found to be considerably below the standard for an individual 
 of his height. Physical examination of the abdomen usually 
 shows nothing abnormal, although rarely a fullness or swelling may 
 be detected in the region of the pancreas.. No physical signs of 
 disease are found in the chest, unless as the result of tuberculosis 
 and other complications, which mil be mentioned shortly. The 
 temperature is generally normal, or slightly subnormal. 
 
 In many instances the family and 'past history of the patient 
 throw little or no light on the condition, but in some 18 to 20 per 
 cent, it will be found that one or more blood relations have suffered 
 from glycosuria, or died of diabetes. Williamson states that the 
 relatives most frequently affected in order of frequency are — brother, 
 father, mother, and sister. In five out of 250 of his cases (2 per cent.) 
 the husband or wife of the diabetic also suffered from glycosuria, 
 but of course such instances cannot be included in a table of here- 
 dity. Williamson records some striking examples of this family 
 tendency. In one family, consisting of two sons and two daughters, 
 both the sons and one daughter became diabetic. Each had lived 
 in a different town, and several years elapsed between the onset 
 of the glycosuria in the three cases. I have had under my care a 
 patient who informed me that his father, grandfather, and two 
 uncles all passed sugar in their urine, but all of them cUed of some 
 other disease than diabetes. 
 
 Loeb considers that cases in which heredity can be traced are 
 distinct as regards etiology, symptoms, course, complications, and 
 ]3rognosis — that is to say, there is an hereditary variety of dia- 
 betes, which can be differentiated by the following characters : — 
 1. Etiology. — (a) Females in hereditary cases are affected as 
 frequently as, or even more frequently, than males. In diabetes 
 generally males are affected about thrice as frequently as females. 
 In Frankfort, where hereditary diabetes is exceedingly common, 
 the diabetes death-rate is about the same for males and females ; 
 in one year more females than males succumbed. In a family 
 tree, constructed by von Noorden, there were eight cases of diabetes 
 in females and only five in males ; and in a diabetic family Loeb 
 found eight cases in females and four in males. The statistics of 
 Weidenbaum, based on one thousand cases of diabetes, j)oint in 
 the same direction. (6) As regards race, the herecUtary form of 
 diabetes is especially frequent in Jews. Wallach showed that in 
 a period of twelve years the proportion of deaths from diabetes 
 was six times greater among Jews than Gentiles, the greater 
 frequency of the hereditary form among Jews probably accounting 
 
196 GLYCOSURIA 
 
 for the whole of the excess, (c) The hereditary form of diabetes 
 seldom occurs in youth, and usually appears between fifty and 
 sixty. Loeb has seen it only twice in young peoj)le — once in a 
 woman aged twenty-three, and once in a boy of ten, whose aunt 
 had died of diabetes at thirty. Von Noorden knew of a family 
 in Avhich a slight case of diabetes occurred ; in the second genera- 
 tion three female members were attacked in middle age by a 
 rapidly fatal form of diabetes, and in the third generation two 
 children fell victims to the disease, (d) As regards constitution, 
 the hereditary form almost always attacks well-nourished pre- 
 viouslj^ healthy subjects. Frequently, and especialty in women, 
 there is a tendency to obesity, though the diabetes of obese subjects 
 is not necessarily hereditary. Many of these patients are ' ' nervous 
 or suffer from paralysis agitans, hysteria, or mental disease. 
 
 2. Course. — In hereditary diabetes the onset is often insidious. 
 Sugar may recur at varying intervals in the urine, usually in small, 
 though sometimes in large, amounts. The general health at this 
 early stage is not affected. The duration of the disease cannot 
 readily be estimated, as it often exists for years before the onset 
 of general symptoms. Usually it runs a chronic and benign course, 
 so that if the patients are docile and escape intercurrent diseases, 
 they may attain a good age. Acetone bodies are seldom excreted 
 in excess until shortly before death. 
 
 3. Symptoms. — The symptoms are those of a mild form of 
 diabetes, and are readily controlled by diet. They recur after 
 errors of diet, or excitement, and may eventually persist. Arterio- 
 sclerosis, as evidenced by thickening of the radial artery, high blood 
 pressure, and hj^pertrojDhy of the left ventricle, or, in more ad- 
 vanced cases, by asthmatic and anginal attacks, cerebral haemorr- 
 hage, and albuminuria, usually occurs early in hereditary diabetes. 
 Pulmonary tuberculosis is practically unknown in the hereditary 
 form. 
 
 4. Complications. — The dangers which threaten patients with 
 the hereditary form of diabetes are chiefly those of intercurrent 
 diseases (influenza, pneumonia, erysipelas, arterio-sclerosis, and 
 its results — gangrene, thrombosis, and cardiac failure). In only 
 two of sixteen fatal cases was death directly attributable to 
 diabetes. 
 
 In some cases of persistent glycosuria there is a family history 
 of gout, diabetes insipidus, phthisis, exophthalmic goitre, epilepsy, 
 or various neuroses. 
 
 The glycosuria will in some instances be found to have made 
 its appearance after a j)eriod of acute or prolonged nervous strain, 
 
PERSISTENT GLYCOSURIA 197 
 
 such as severe anxiety, grief, or business worry. Occasionally 
 an attack of some infectious disease, such as typhoid fever, or 
 syphilis, may be considered to be the starting-point, and ^\ith 
 regard to these it is important to remember that glycosuria may 
 not appear for some years after the attack, for in such cases the 
 l^ancreas is probably the organ at fault, and the degree of in- 
 flammatory change necessarj^ to give rise to glycosuria is only 
 slowly produced. In one case that has come under my care sugar 
 was found in the urine, and typhoid bacilli were isolated from the 
 faeces ten years after an attack of typhoid fever. 
 
 On September 8, 1908, I was consulted by an American gentleman, 
 who came to me with a letter of introduction from Dr. J. B. iNIurphy 
 of Chicago. In his letter Dr. Murphy stated that the patient was 
 suffering from diabetes, and had been sent to me with the object of 
 determining the condition of his pancreas and clearing up the etiology 
 of the disease. He also informed me that there was a distinct history 
 of cholecystitic infection. 
 
 The patient was a well-built, healthy looking man, fifty-one years 
 of age. He had had no serious illness, except an attack of typhoid 
 fever in 1898. There was no history of syphilis, or gastro -intestinal 
 disturbances, nor had he had any symptoms pointing to the presence 
 of gall-stones. Although he smoked a good deal, he took little or no 
 alcohol. He had never suffered from thirst or polyphagia. The 
 quantity of urine excreted had not increased. So far as he knew he 
 had not lost flesh, nor had he noticed any dmiinution of strength. 
 Three per cent, of sugar had been found in his urine in the course 
 of a routine examination at an American health-resort in January 1908, 
 but on an anti-diabetic diet this had been reduced to about half. 
 
 Physical examination revealed nothing abnormal. The patient's 
 tongue was not red or glazed, and his gums and mouth appeared 
 normal. His skin was moist. His heart was not enlarged, the heart 
 sounds were natural, and his arteries were not thickened. His ab- 
 domen and body generally were well covered with fat. No abdominal 
 tumour or swelling could be discovered on palpation. The liver 
 dullness was normal, and the stomach did not appear to be dilated. 
 Analysis of a specimen of the mixed night and morning urine gave 
 the following results : — ■ 
 
 Reaction, acid ; Specific gravity, 1023 ; Albumin about 1 : 5000 ; 
 Sugar, Fehling's solution — reduced at once, phenylhydrazin — crowds 
 of typical phenylglucosazone crystals, insoluble in 33 per cent, sul- 
 phuric acid in five minutes : quantitatively, copper reduction (Bang's 
 method), 2*3 per cent., polariscope, +2-2 per cent. ; fermentation 
 (Lohenstein's saccharimeter), 1-9 per cent.; Acetone, nil; Aceto 
 acetic acid, nil ; Indican, a fairly well-marked reaction ; Bile, nil ; 
 Urobilin, a pathological excess ; Blood, nil ; Urea, 2-42 per cent. ; 
 Chlorides, 0-8 per cent. ; Phosphates, 0-13 per cent. ; Preformed to 
 
198 GLYCOSURIA 
 
 Conj. sulphates, 8:1. " Critical solution point " (phenol), raised 10° C. 
 Microscopically, many small calciiom oxalate crystals, a few squameoiis 
 and transitional epithelial cells, a few leucocytes, no casts. " Pancreatic " 
 reaction, many typical fine crystals, soluble in 33 per cent, sulphuric 
 acid in five to ten seconds ; melting-point, after recrystallisation, 160° C. 
 
 The urine, therefore, contained a fair amount of a dextro-rotatory 
 fermentable svigar, with traces of a non-fermentable variety. The 
 results of the " pancreatic " reaction suggested that the glycosuria 
 was probably associated with disease of the pancreas, and the presence 
 of many small calcium oxalate crystals in the centrifugalised deposit 
 tended to confirm this conclusion. The pathological excess of 
 lu-obilin pointed to there being a catarrhal condition of the biliary 
 passages, which probably extended to the pancreatic ducts. The 
 abnormal reaction for indican, and the disturbed relation of the pre- 
 formed to the conjugated sulphates, suggested that both the pan- 
 creatic disease and cholangitis were connected with a catarrhal con- 
 dition of the upper part of the intestinal tract. The absence of any 
 trace of acetone, or aceto-acetic acid, showed that there were not the 
 profound tissue changes met with in severe diabetes, and rendered 
 it probable that the pancreatic disease was of a slowly advancing type, 
 such as, in my experience, is the common result of crrrhoses of the 
 pancreas secondary to infection of the ducts. Although the specimen 
 contained traces of albumin, the fairly normal critical solution point, 
 the normal percentages of urea and inorganic salts, and the absence 
 of casts or other evidence of renal mischief, rendered it unlikely that 
 there was any serious disease of the kidneys, while the presence of 
 leucocytes and transitional eptheliiuu suggested that it was not 
 improbably of bladder origin. 
 
 The bowels were stated to be opened regularly every day, and 
 there was neither diarrhcea nor constipation. A sample of the faeces 
 submitted for examination gave the following results : — 
 
 Appearance, dark brown, formed, solid mass ; Reaction, ampho- 
 teric ; Stercohilin, a well-marked reaction ; Occult blood, nil. Micro- 
 scopically, a little vegetable tissue, many partly digested muscle 
 fibres, some fatty acid and soap crystals, no fat globules : Quantitative 
 analysis showed : — 
 
 Organic matter 
 
 88-7 
 
 per 
 
 cent. 
 
 of the dry weight 
 
 Total fat 
 
 16-6 
 
 
 
 
 " Unsaponified " fat^ 
 
 8-6 
 
 
 
 
 Saponified fat 
 
 8-0 
 
 
 
 
 Organic matter not fat 
 
 70-1 
 
 
 
 
 Inorganic ash 
 
 13-8 
 
 
 
 
 The only striking variation from the normal was the large amount 
 of partly digested muscle fibre found microscoiDically, but this was 
 probably to be explained by the highly nitrogenous diet which the 
 
 1 i.e. neutral fats and free fatty acids estimated together. See Brit. Med. 
 Journ., Oct. 28, 1905, p. 1102. 
 
PERSISTENT GLYCOSURIA 199 
 
 patient was taking. There was no marked excess of unabsorbed fat, 
 nor was the relation between the saponified and " unsaponified " fats 
 disturbed in the way that one might expect in serious disease of the 
 pancreas. This, however, I have found in a considerable number of 
 cases of glycosuria of undoubted pancreatic origin in which I have had 
 an opportvinity of examining the stools, and I have come to the con- 
 clusion that any marked interference with fat digestion is of serious 
 jDrognostic significance, and indicates that the patient is in a late 
 stage of the disease. The absence of any trace of occult blood was 
 against there being any malignant growth in the course of the gastro- 
 intestinal tract, to which the pancreatic disease might be secondary. 
 
 Bearing in mind the history of typhoid fever, and the well-re- 
 cognised tendency of the typhoid bacillus to linger in the gall-bladder 
 and bile-ducts, in some cases, long after the patient has quite recovered 
 from the disease, it seemed to me possible that I had to do with a case 
 of tyjDhoidal pancreatitis going on to giycosima, and that a bacterio- 
 logical analysis of the faeces might throw further light on the condition. 
 A portion of the faecal material was taken with a sterile knife from the 
 centre of the mass sent for examination, emulsified in sterile normal 
 saline solution, plated on Drigalski and Conradi's medituii, and in- 
 cubated at 37° C. On being examined twenty-foiir hours later one of 
 the six plates that had been inoculated showed two small, blue, finely 
 granular colonies, with raised centres and filmy edges, suggesting 
 bacillus typhosus. Sub-cultures were made from these on to agar- 
 agar slopes, and incubated at 37° C. for twenty-four hours ; they then 
 showed a thin bluish white, transparent film, that had not spread far 
 from the needle track.. Hanging-drop preparations showed actively 
 motile rod-shaped bacilli. The bacilli did not stain by Gram's method, 
 and cover-glass preparations stained by Van Ermengen's method 
 showed that they were richly flagellate. Glucose-gelatine shake 
 cultures showed no gas formation in forty-eight hours. Lactose 
 litmus broth was not rendered acid, and showed no gas formation. 
 Litmus-milk cultxxres were not coagulated, nor was there any acid 
 fermentation in forty-eight hours. Neutral red broth became tm'bid, 
 but there was no film formation or colour change. Peptone salt 
 cultures incubated at 37° C. for forty-eight hours gave no indol re- 
 action. An emulsion of the bacilli in normal saline solution, prepared 
 from a twenty-foiir hotirs agar culture, gave a similar reaction with a 
 typhoid serum to a laboratory ciilture of bacillus typhosus. The 
 faeces, therefore, contained a small number of bacilli having the appear- 
 ance and characters of bacillus typhosus ; a result which tended to 
 confirm my siirmise as to the probable origin of the disease. 
 
 In a few cases the glycosuria may be traced to changes set 
 up in the pancreas by gall-stones. One of the most common 
 exciting causes in my experience is chronic gastro-enteritis. This 
 has often not been serious, but on going carefully into the history 
 of cases of glycosuria it is not infrequently found that there have 
 
200 GLYCOSURIA 
 
 been sjnnptoms of " intestinal indigestion " extending over many 
 years, and that the patient has been abnormally fond of sweets. 
 Funck has also remarked that gastro-intestinal disturbances are 
 unexpectedly prominent in the histories of diabetics when they 
 are carefully gone into, and suggests that gastritis and chronic 
 enteritis are a primary factor in the production of chronic glycosuria 
 in many cases, or are at least responsible for exacerbations much 
 more frequently than is generally supposed. 
 
 Omng to the special nature of my work for a considerable 
 number of years I have seen an unusually large proportion of cases 
 in which glycosuria has been associated with disease of the pancreas 
 and gall-stones. Taking a consecutive series of two hundred cases 
 I find that biliary calculi were present in just under 12 per cent., 
 and it is noteworthj^ that in nearly half of these (47 per cent.) 
 there was no jaundice, and that there was more sugar than in the 
 jaundiced cases, suggesting that the absence of this striking 
 symptom had deferred the diagnosis and led to more serious 
 IDancreatic mischief. Eight of my cases (4 per cent.) had been 
 operated on for gall-stones at periods varying from two to six years 
 prior to the onset of the glycosuria. In one instance the presence 
 of sugar in the urine was associated with a growth of the ampulla 
 of Vater, in one mth a growth of the duodenum invading the 
 pancreas, and in twelve (6 per cent.) mth primary malignant dis- 
 ease of the gland. In two cases there was a cyst of the pancreas, 
 and in two j)ancreatic calculi were found, in one at operation and 
 in the other post-mortem. In one case a transient glycosuria 
 was associated with parotitis and symptoms of pancreatitis. I 
 found interacinar pancreatitis in three and interlobular pancreatitis 
 in two cases not associated A\'ith gall-stones. One of the latter 
 had been diagnosed as mahgnant disease during life, but post- 
 mortem no secondary growths could be found, and histologically 
 there was no evidence of cancer. Interlobular pancreatitis was 
 also found in three cases of gall-stones with glycosuria that I 
 investigated. In one of my cases there was a history of an accident 
 in which the upper part of the abdomen was crushed eight months 
 before the onset of the glycosuria. Arterio-sclerosis was present 
 in 6 per cent, of the cases, a history of gout was obtained in 4 ]3er 
 cent. Two had suffered from syphilis, one had exophthalmic 
 goitre, one was a member of a family that suffered from haemo- 
 philia and was a " bleeder " himself, in one the glycosuria had 
 come on during pregnancy, four, including the one already men- 
 tioned, gave a history of typhoid fever, and two had had repeated 
 attacks of influenza. In six cases (3 per cent.) glycosuria appeared 
 
PERSISTENT GLYCOSURIA 201 
 
 to be a family disease, and had affected one or more members 
 beside the patient. A history of chronic indigestion was given by 
 10 per cent, of the cases, and an analysis of the urine and faeces 
 tended to show that there was a chronic catarrh of the intestine. 
 In four cases there was a definite history of duodenal ulcer. 
 
 Complications. — Persistent glycosuria is not of itself a 
 serious condition, and many patients whose urine contains sugar 
 live comfortable lives for many years ; it is the complications and 
 secondary effects of the metabolic disturbance to which the gly- 
 cosuria is due that constitute the danger. All the complications 
 met with in diabetes do not threaten the existence of the patient ; 
 some are merely annoying or painful, but the appearance of others 
 is of the very gravest import, and unless they are speedily recognised, 
 and treated, a fatal termination is likely to quickly supervene. 
 
 The secondary effects of persistent glycosuria may be con- 
 veniently considered under the various systems. 
 
 The Skin. — The skin in mild and early cases appears and feels 
 normal, but in severe cases of diabetes it is usually harsh and dry, 
 in striking contrast to the velvety feel sometimes met with in 
 diabetes insipidus. Occasionally, however, it may be moist, and 
 perspiration may be excessive, in cases that are otherwise typical. 
 Some observers state that they have detected sugar in the sweat, 
 but others have failed to do so. In one case sugar is said to have 
 been found in the tears. Pruritis, which may be general, but is 
 more frequently confined to the external genitals, and the skin 
 in their neighbourhood, is sometimes troublesome, and occasionally 
 is one of the first symptoms. Local pruritus about the genital 
 organs is most common in women and may draw attention to the 
 existence of the glycosuria, especially in stout females. The 
 irritation of the skin is generally accompanied by congestion and 
 redness, which may lead to balanitis in men, and vulvitis in women, 
 together with eczema of the neighbouring skin. The majority of 
 cases of eczema of the vulva occurring in women about the climac- 
 teric are due to glycosuria. Sometimes eczema and eriv'thema 
 also occur in other situations. Psoriasis, urticaria, and more or 
 less localised patches of oedema are also met with in some cases. 
 Xanthoma is a very rare complication, which, when present, 
 diminishes as the sugar in the urine is reduced, and reaj^pears 
 with a return of the glycosuria. Boils are among the most common 
 skin lesions in chronic glycosuria. They often occur at an early 
 stage, and may be the first obvious symptom of the condition. It 
 is therefore important that the urine of everyone suffering from 
 
202 GLYCOSURIA 
 
 furunculosis should be repeatedly and carefully examined for 
 sugar. The boils may be met with in any situation, but are most 
 common on the neck, the shoulders, and the buttocks. According 
 to Marechal one-third, and v. Noorden one-fourth of all persons 
 having boils suffer from glycosuria. In advanced cases of diabetes 
 they are rare. Carbuncles, which like boils may be amongst the 
 earliest symptoms, also occur later in the disease, when they have 
 a tendency to spread and become gangrenous, or to give rise to 
 cellulitis, so that they are sometimes the cause of death. They 
 are most commonly seen on the neck, but they also occur on the 
 face, or other parts. Carbuncles are a less frequent complication 
 than boils. Gangrene is sometimes met with, and occasionally 
 may be the first symptom of diabetes ; it may come on spon- 
 taneously, or as the result of slight injury, or be secondary to 
 wounds, boils, carbuncles, &c. The lower limbs are most fre- 
 quently affected, the gangrene commencing in the toes. In a 
 large proj^ortion of cases of gangrene of the leg the lumen of the 
 vessels is reduced by atheroma and the blood supply is conse- 
 quently diminished, hence the intermittent claudication, or limping, 
 first noticed by Charcot. The gangrene may be moist or dry. With 
 the former the constitutional symjatoms, such as loss of appetite, 
 drowsiness, and delirium are more marked than with the latter. 
 Perforating ulcers, resembling those met with in locomotor ataxia, 
 are occasionally seen in diabetes, and are probably of nervous 
 origin. They are chiefly seen on the soles of the feet, and are 
 especially common about the big toe. The starting-point is fre- 
 quently a wound caused by cutting a com, &c. All wounds heal 
 badly in diabetic patients, so that operative interference is avoided 
 as far as possible by most surgeons. 
 
 Digestive System. — In severe cases of diabetes the breath has a 
 peculiar sweet smell, like that of decomposing apples. It is 
 generally referred to the presence of acetone, although this is 
 doubted by some authorities (Folin). The mouth is often dry, 
 and the saliva is, as a rule, scanty. Rarely ptyalism, such as is 
 occasionally met with in association with disease of the pancreas, 
 has been noted. The saliva is usually said to be free from sugar, 
 although this is denied by Redier, is acid in reaction, and some- 
 times does not give the sulphocyanide reaction. The gums are 
 often spongy, tender, and retracted. The teeth are frequently 
 carious, without giving rise to much pain. Alveolar periostitis 
 occasionally occurs, and the teeth may become loose and drop out 
 as the disease progresses. In advanced cases of diabetes aphthous 
 stomatitis is sometimes present. The 'pharynx may be intensely 
 
PERSISTENT GLYCOSURIA 203 
 
 congested, especially about the base of the tongue, and this con- 
 gestion may spread to the larynx. The tonsils may be the seat of 
 abscess or gangrene. The appetite is often, but by no means 
 always, increased. This increase is met with chiefly in the severe 
 forms, or in mild cases when carbohydrate food is being taken in 
 large quantities. The excess of food is liable to cause dilatation 
 of the stomach and may set up gastritis, which in its turn may result 
 in atroph}^ and absence of gastric juice. Griibe and others have 
 observed abdominal crises, resembling those seen in locomotor 
 ataxia. The patient is suddenly seized with violent abdominal 
 pain, especially at the pit of the stomach, the abdomen becomes 
 distended, eructations occur, nausea and the vomiting of acid 
 material follow, and this is sometimes succeeded by diarrhoea 
 and cramp in the legs. Such a condition is not infrequently the 
 precursor of diabetic coma. Diabetic patients often suffer from 
 constipation, especially when they are on a highly nitrogenous 
 diet, and when diabetic coma is imminent. Occasionally symptoms 
 so closely resembling those of acute intestinal obstruction that 
 immediate operation appeared necessary have been met with. 
 Downes and O'Brien have reported two such cases, and it was 
 only when the urine was examined and found to contain a large 
 amount of sugar, beside giving well-marked reactions for acetone 
 bodies, that the true nature of the condition was recognised. 
 Similar cases have also been described by Tirarcl and Da vies. 
 Diarrhoea may sometimes result from intercurrent tubercular 
 enteritis, but is more often dependent upon mal-assimilation of the 
 fatty and nitrogenous foods prescribed. In a few cases typical 
 fatty stools are seen, but steatorrhoea is rare, even when the gly- 
 cosuria is associated with pancreatic disease. In a consecutive 
 series of a hundred cases of diabetes, on ordinary diet, in which I 
 have made an analysis of the faeces, an abnormal proportion of 
 unabsorbed fat was only found in sixteen, but in forty-eight cases 
 an excess of " unsaponified " over saponified fat was met with. Un- 
 digested muscle fibres were discovered microscopically in five cases, 
 and a deficiency of pancreatic juice was sho^^Ti by the casein 
 digestion test, in fifty-two. In thirty-four of these it was slight, 
 and in eighteen well marked. On rare occasions an enlarged pancreas 
 can be made out on abdominal examination, especially in thin 
 subjects and under an anaesthetic. This may be due to inflammatory 
 changes in, and around, the gland, when there is likelj^ to be tender- 
 ness in the epigastrium, and a tender spot, just above and to the 
 right of the umbilicus, with possibly pain in the back under the 
 right scapula, or to cirrhosis of the pancreas, a pancreatic cyst, or 
 
204 GLYCOSURIA 
 
 more rarely a growth of the head of the gland, when there will also 
 be progressively deepening, and painless, jaundice. Sometimes the 
 liver is found to be enlarged from hypersemia, fatty infiltration, 
 or cirrhosis, chieflj^ in gouty or obese subjects who suffer from a 
 mild form of glycosuria. 
 
 Respiratory System. — One of the most common complications 
 of diabetes is pulmonary tuberculosis, a third to a quarter of all 
 cases dying of this disease. It is often latent, and tuberculosis 
 of the lungs is frequently found post-mortem when the disease has 
 not been recognised as present during life. It may run a rapid 
 course with early excavation, which after death is nearly always 
 found to be much more extensive than would be expected from the 
 symptoms and physical signs. Other pulmonary complications 
 that sometimes occur are gangrene of the lung, which also may run 
 a rapidly fatal course, broncho-pneumonia, and acute croupous 
 pneumonia, which occasionally gives rise to few subjective signs, 
 but is nearly always fatal. 
 
 Circulatory System. — Diabetics of all ages frequently exhibit 
 arteriosclerotic changes, which are, however, comparatively rare 
 in severe cases, and are most commonly met with in the mild, 
 chronic, forms of glycosuria occurring in later life. In the former 
 it is not unlikely that the condition of the vessels is dependent 
 upon the action of organic acids, and other toxic substances, 
 circulating in the blood, while in the latter it is probable that the 
 arterio-sclerosis is antecedent to the glycosuria, at any rate in 
 many instances, and is rather a cause than a complication of the 
 condition. The glycosuria in such cases is most likely dependent 
 upon nutritive changes in the pancreas, and other organs control- 
 ling carbohydrate metabolism, arising as a result of the defective 
 blood-supply and the toxsemia that is primarily responsible for 
 the degeneration in the vessel walls. The heart is usually not 
 affected in the earlier stages of diabetes, although towards the 
 termination its action may be rapid and feeble, especially Avhen 
 coma is threatening or has developed. In such cases post-mortem 
 examinations show that it is small, and degenerative changes, 
 ascribed by some to the action of circulating acids and toxines, 
 are found in the myo-cardium. The pulse is usually regular, and 
 of normal frequency and tension. In some cases, however, it is 
 of high tension, even when there is no kidney mischief, and the 
 patient is under middle age, suggesting that some factor, or internal 
 secretion, producing an effect like epinephrin, is at work. 
 
 Renal System. — Albuminuria is more esj)ecially found in those 
 cases where there is gout, arterio-sclerosis, or obesity. Sometimes 
 
PERSISTENT GLYCOSURIA 205 
 
 the albumen is, as we have seen, only small in amount and tem- 
 porarily present, at others it is a sign of intercurrent interstitial 
 nephritis. 
 
 Nervous System. — Individuals suffering from persistent glycos- 
 uria are subject to nervous affections of the most varied kind. 
 Loss of sexual power is not infrequently an early symptom, and 
 the i^atient may first seek advice because of impotence. A loss of 
 sexual power does not, however, always occur, and in some cases 
 an increase of sexual desire has been observed. In females similar 
 changes have been met with, and amenorrhoea is sometimes an 
 early symptom. Conception, pregnancy, and parturition may 
 occur in the normal way in women with diabetes, but there is a 
 great tendency to abortion, which, according to Gaudard, occurs 
 in about 30 per cent, of cases. During pregnancy and the puer- 
 peral state the disease usually follows a rapidly downward course. 
 Cramp in the calf muscles, myalgia, neuralgic pains in the distribu- 
 tion of one or more of the spinal nerves, especially the sciatic, 
 are very common in the severer forms of diabetes, and occasionally 
 may be one of the earliest symptoms. The sciatica is often bi- 
 lateral, and the presence of such a condition should always suggest 
 the presence of glycosuria. Next to the sciatic, the trigeminal 
 nerves are most commonly affected. Multiple neuritis is met with 
 in some cases. It is usually of a mild type, and most often affects 
 the lower extremities. The arms and legs are apparently never 
 affected together, unlike alcoholic neuritis. Sensory disturbances 
 are, as a rule, most prominent, so that numbness, pain, commonly 
 of a dull aching, gnawing, or burning character, along the nerves^ 
 with areas of anaesthesia and hypersesthesia varying in position 
 and degree, are chiefly found. Only in the severe forms of diabetes 
 are the motor powers much impaired. The condition of the reflexes 
 in diabetes has been the subject of numerous researches. Bouchard 
 pointed out that the knee-jerks disappear in about 36 to 37 per cent, 
 of cases. Williamson found that they were absent in half his 
 cases, and that they were lost in a much larger projjortion of 
 persons under, than over, thirty years of age. This fact, which 
 at first sight appears astonishing, is no doubt due to the generally 
 greater severity of the disease in young peojDle. An exaggeration 
 of the knee-jerks has been observed by Zaudy and others, 
 particularly just before the onset of coma. A careful investiga- 
 tion of both the cutaneous and deep reflexes was made by Pitres 
 in thirty-six diabetics, and he found that the cutaneous reflexes 
 (abdominal, cremasteric, and plantar) were diminished, or abolished, 
 more often than the knee-jerks, but that the pupillary reflex is 
 
206 GLYCOSURIA 
 
 usually intact, a useful distinction from locomotor ataxia. Some 
 of the troj)liic disturbances, such as atrophy of the skin, herpes, 
 cracking and shedding of the nails, perforating ulcer, &c., have 
 already been referred to. 
 
 Affections of the spinal cord, such as tabes dorsalis and dis- 
 seminated sclerosis, may occasionally precede the appearance of 
 sugar in the urine, but they only very rarely occur as complications 
 of persistent glycosuria. 
 
 Cerebral complications, which may cause haemiplegia or mono- 
 plegia, are sometimes encountered. They may arise from haemorr- 
 hage, softening from atheroma of the arteries, or the action of 
 some toxic agent which produces the symptoms without causing 
 any recognisable lesion. Of the mental change, depression, irri- 
 tabihty, or restlessness, melancholia, with or without suicidal 
 tendencies, and mania are met "with, especiall}^ in severe cases, when 
 coma is imminent or a very strict diet has been enforced. 
 
 Special Senses. — Ocular changes are not very common. Cataract 
 is the most frequently met mth. It is usually double, and soft, 
 and runs a fairly rapid course. It is most commonly seen in severe 
 cases, and is therefore more frequent in young subjects. Retinitis 
 and retinal haemorrhages are occasionally met mth, but hardly 
 ever in young persons. A toxic amblyopia has been described. 
 Defective accommodation is not uncommon, and according to Hirsch- 
 berg is an early symptom of diabetes. Albuminuric retinitis 
 occurs in cases with complicating kidney mischief. Hsemorrhagic 
 glaucoma, iritis, purulent keratitis, and atrophy of the optic nerve 
 liave been seen, although rarely. 
 
 Furunculosis of the external ear, and rapidly cleveloj^ing otitis 
 media are sometimes mit with. 
 
 Osseous System. — Fragihty of the bones, and delayed union 
 after fracture, have been noticed in severe cases of diabetes. The 
 former probably depends upon the removal of lime salts in an 
 attempt on the part of the organism to neutralise the abnormal 
 acidit}^ of the blood that occurs in such cases, while the latter is 
 part of the general loss of reparative power that is so characteristic 
 of chronic glycosuria. 
 
 Anasarca in non-cachectic cases, without albuminuria or any 
 sign of cardiac failure, is occasionally met with. The oedema 
 chiefly affects the legs, and there is pitting on pressure on the skin 
 about the ankles, on the dorsum of the foot, and over the tibia. A 
 slipper or boot will often leave a characteristic imjoression, which 
 attracts the patient's attention. In a few cases the hands, the face, 
 or other parts are involved. The oedema is believed to depend 
 
PERSISTENT GLYCOSURIA 207 
 
 upon chloride retention, consequent upon damage of the kidneys, 
 and one certainly finds as a rule an abnormally low chloride 
 excretion in cases where this condition exists. 
 
 Acidosis. — It is well knoAvn that abstinence from food, or even 
 the sudden withdrawal of carbohydrates from the diet, gives rise 
 to acetonemia, or acetonuria — that is to say, the acetone bodies, 
 including acetone, aceto-acetic acid, and less frequently beta- 
 oxybutjrric acid, are excreted in the urine. It is therefore not 
 surprising that in persistent dextrosuria, where the po^^'er of 
 utilising dextrose is reduced, and later the general oxidative 
 capacity of the body is interfered with, the acetone bodies and 
 other unoxidised products of metabolism appear in the urine. It 
 would seem probable that the acetone bodies are not abnormal 
 products of metabolism, but that they are normally formed and 
 only accumulate, and are excreted unchanged, when they are not 
 •destroyed, owing to some oxidative defect on the part of the tissues. 
 All cases of diabetes, therefore, are not complicated by acetonemia, 
 although it is a constantly present menace, since a deficiency in the 
 oxidative powers of the body is an essential element of the con- 
 dition. It is only when this reaches a certain stage, however, that 
 the acetone bodies appear in the urine as a necessary consequence. 
 
 As we have already seen acetone, aceto-acetic acid, and beta- 
 oxybutyric acid are closely related, and probably have a common 
 .source, for although the experiments of Embden, Salomon, and 
 Schmidt suggest that acetone may be derived from substances, 
 such as tyrosine, which do not give rise to oxybutjrric acid, it may 
 be taken for granted that when acetone can be detected in the 
 urine by the ordinary qualitative tests, the metabolic powers of 
 the organism are impaired. If acetone alone is present the dis- 
 turbance is not, as yet, a serious one, for the organism is still 
 capable of oxidising oxybutyric acid to aceto-acetic acid, and of 
 breaking up the latter. Should aceto-acetic acid also be found in 
 the urine, it points to there being a much graver interference mth 
 metabolism, for the organism has now not only lost the power of 
 breaking up this substance, but also probably the capacity to 
 oxidise beta-oxybutyric acid, since it is generally found that the 
 two occur together. According to v. Noorden when the acetone 
 in the urine reaches 0-4 to 0-5 grams a day, the perchloride of iron 
 reaction for aceto-acetic acid is invariably obtained, when the 
 acetone reached 0*6 to 1-0 grams a day oxybutyric acid is usually 
 present, when 1-5 grams or more of acetone are met with in the 
 urine beta-oxybutyric acid is rarely absent, although such cases 
 are occasionally met with. 
 
208 GLYCOSURIA 
 
 The acetone bodies may also be detected in the internal organs. 
 Geelmuj'den found considerable quantities in cases of diabetes 
 that he examined, but there was less in the liver than in the other 
 viscera, and the blood contained less acetone than the urine of 
 the same patient. 
 
 The source of the acetone bodies was for long a subject of keen 
 controversy, but it now seems to be practically settled that thej' 
 are principally derived from fat, and particularly from fats con- 
 taining the lower fatty acids. The chemical experiments of Blu- 
 menthal and Neuberg, and of Orgler, have shown that they can 
 also originate from proteins. The observations of Embden, 
 Salomon, and Schmidt on animals prove that they can have this 
 source witliin the body, but as the first step in the process appears 
 to be the splitting off of the ammonia group of the contained amido 
 acids and their conversion into fatty acids, the immediate ante- 
 cedent of the acetone bodies formed is in either case the same. It 
 is not now believed, as was at one time thought, that the acetone 
 bodies can be derived from carbohydrates in a similar way to the 
 closely related lactic acid, chiefly because, as Satta showed, the 
 administration of a proper amount of carbohydrate under certain 
 conditions may cause them to disappear from the urine, and that 
 when a patient is put upon a diet that is almost free from carbo- 
 hydrate they may be eliminated in large quantities. 
 
 The place of origin of the acetone bodies is, according to Embden, 
 Salomon, and Schmidt, most probably the liver, although there 
 are some, and notably J. Miiller, who considered that they are of 
 intestinal origin. The latter view is not consistent with the ob- 
 servations of Liithje or v. Noorden, who found that acetonuria 
 is not diminished bj^ the use of laxatives and intestinal antiseptics, 
 but may, on the contrary, increase as a result of the employment 
 of such measures. Many patients suffering from the earlier cerebral 
 symptoms of diabetic coma are, however, peculiarly sensitive to 
 tiae effects of constipation, and, while the lapse of two or three 
 days without an action of the bowels may greatly exaggerate the 
 symptoms, a free purge is promptly followed by remarkable relief. 
 
 It has long been known that diabetic patients are liable to. 
 suffer from nervous symptoms, and attacks of dyspnoea, succeeded 
 by coma, which are usually followed by death. Since the coma 
 appears Avhen the quantity of acetone bodies is highest, and is 
 absent A^hen the amount is small, or they are not present, it would 
 seem probable that the condition is dependent upon an excess of 
 these substances in the blood, or upon some condition that is 
 associated with such an excess. The occurrence of coma in diabetes 
 
PERSISTENT GLYCOSURIA 209 
 
 appears to have been first recorded by v. Dusch. in 1854. Three 
 years later Fetters discovered what he believed was acetone in 
 the urine, and blood, of a diabetic patient who died with anuria 
 and a subnormal temperature. The presence of this substance 
 in the urine of many cases of diabetes was confirmed by Kaulich, 
 who at the same time pointed out that it is not peculiar to diabetes. 
 In 1874 Rupstein conclusively proved, by an exhaustive chemical 
 analysis of the fluid isolated from the urine of a case of diabetes, 
 that it was really acetone. The same year Kussmaul investigated 
 the toxicity of this substance and found that after large doses had 
 been administered to animals the temperature fell, the pulse became 
 more rapid, and the respirations were shallower. It was therefore 
 assumed that diabetic coma was due to acetone poisoning. It was 
 subsequently shown that a close of 8 grams per kilo in dogs, corre- 
 sponding to 500 grams for an adult man, is rec[uired to bring about 
 a fatal result, also that acetone can be excreted for years without 
 any symptoms of coma, that in a few cases of coma acetone is 
 absent, that moderate doses do not cause the symptoms described 
 by Kussmaul, 4 grams per kilo in dogs only giving rise to the same 
 effects as ethyl alcohol, and, finally, that the symptoms produced, 
 even by large doses, are not identical with those of diabetic coma. 
 
 In 1865 Gerhardt discovered the reaction of diabetic urine with 
 perchloride of iron known by his name, and attributed it to the 
 presence of aceto-acetic ether. Buhl, as the result of experiments 
 performed on rabbits, came to the conclusion that this substance was 
 the cause of diabetic coma. His conclusions were not confirmed, 
 however, by Quincke, who found that aceto-acetic ether is only 
 very slightly toxic for dogs. When it was proved by v. Jaksch 
 that Gerhardt's reaction is in reality dependent upon aceto-acetic 
 acid, it was sought to ascribe diabetic coma to the presence of this 
 substance, but the investigations of Brieger and others proved 
 that it, too, is only slightly toxic. 
 
 In 1880 Goetghens compared the acids and the bases in the 
 urine from a diabetic, and found that the latter were in decided 
 excess of what was required to neutralise the knowTL acids present. 
 He consecj[uently suggested that the urine in diabetes contains 
 some unknown acid. Three years later Stadelmann, following 
 the same method, confirmed Goetghens' results, and succeeded in 
 isolating crotonic acid. He suggested that the symptoms of 
 diabetic coma are due to increased acid formation, and that 
 crotonic acid is the immediate cause. The latter conclusion was 
 soon disproved by Minkowski and Kiilz, who showed that crotonic 
 acid does not exist as such in the urine, but is formed from beta- 
 
 o 
 
210 GLYCOSURIA 
 
 oxybutyric acid as the result of the chemical manipulation to which 
 it had been subjected. They looked upon the oxybutyric acid as 
 the cause of the symptoms of diabetic coma. The experiments 
 of Waldvogel, Desgrez, and others showed, however, that beta- 
 oxj^butyric acid, like acetone and aceto-acetic acid, is only sHghtly 
 toxic. The theory of oxybutyric acid intoxication was further 
 weakened by the observations of Walther, who proved that 
 symptoms similar to those of diabetic coma can be produced with 
 hydrochloric , or phosphoric , acid. Eppinger showed what disastrous 
 effects the administration of such acids has upon rabbits. 
 
 If a rabbit is given repeated small doses of an inorganic acid, 
 which cannot be destroyed by oxidation, the respirations soon 
 become more rapid, the pulse rate is increased, its movements 
 become unsteady, convulsions occur, and stupor, followed by coma, 
 and death supervenes. These manifestations are very character- 
 istic, and particularly the respiratory effects. The animal appears 
 as though it were being asphyxiated (the so-called " air hunger "), 
 yet there is no cyanosis, and the blood is bright red. Analysis 
 of the blood shows that it contains much less carbon dioxide than 
 normal, and that the amount of oxygen is unchanged, but that its 
 alkalinity is diminished. If the urine of such an animal is analysed, 
 it is found to contain increased quantities of the chief inorganic 
 bases, and also of ammonia, indicating the withdrawal of alkalies 
 from the body. Normally the blood carries away the carbon 
 dioxide formed by the tissue, by combining it with the inorganic 
 alkalies it contains. This combination, chiefly bicarbonate of 
 sodium, is decomposed in the lungs, the carbon dioxide escaping, 
 and the carbonate returning to the tissues, to be again converted 
 into bicarbonate. If unoxidisable acids are introduced into the 
 blood, as in the above experiments, they combine with the alkalies, 
 and the blood being thus unable to extract the carbon dioxide, it 
 accumulates in the tissues and the animal consequently succumbs 
 to what has been graphically termed " internal suffocation." This 
 explanation of the effects of acid intoxication is substantiated 
 by the remarkable beneficial effects produced by the intravenous 
 injection of alkalies in many cases. Similar experiments with 
 carnivorous animals, such as dogs, have shown that they are re- 
 latively insusceptible to acid intoxication, so that large amounts 
 of acid are required to produce any appreciable effect (Spiro). 
 Analyses of the urine of dogs subjected to this treatment demonstrate 
 that, while the excretion of inorganic alkalies is but little increased, 
 the elimination of ammonia is very much greater than in herbi- 
 vorous animals. It is consequently believed that a great part of 
 
PERSISTENT GLYCOSURIA 211 
 
 the acid is, in carnivorous animals, neutralised by ammonia, and 
 that this ammonia is derived from a jDortion of the nitrogen that 
 normally goes to form urea. As protein metabohsm is greater, 
 and more rapid, in carnivora than in herbivora, the ammonia 
 available for the neutralisation of acids is also larger and more 
 readily obtained, so that in the former the inorganic alkahes of the 
 blood are spared for a longer period, and acid intoxication is more 
 difficult to produce. Eppinger has shown that the resistance of 
 rabbits to acid intoxication can be increased by the administration 
 of amino acids, ap]3arently owing to the ammonia that is formed 
 in the metabolism of these substances. Winterberg and Limbeck 
 found that by gradually increasing the dose of acid they could 
 exceed the usual fatal dose for rabbits, and that there was then 
 a larger elimination of ammonia in the urine. 
 
 From the close resemblance of the sj^mptoms of diabetic coma 
 to those of experimental acid intoxication, it is now generally agreed 
 that the acetone bodies, or rather beta-oxybutyric acid and to a 
 less extent aceto-acetic acid, produce their effects simply as a 
 result of their acid characters, and not in virtue of any specific 
 poisonous properties. Like the inorganic acids they withdraw 
 alkalies from the body, and as a consequence give rise to " internal 
 suffocation," or " acidosis " (Naunyn). In favour of this view is 
 the fact, demonstrated by Orlowski, that titration of the blood in 
 diabetes invariably reveals a reduction in its alkalinity. Minkowski 
 has also shown that the amount of carbon dioxide carried by the 
 venous blood is lowered, from the normal of about 36 per cent., to 
 as little as 3-3 per cent., a reduction comparable only with that 
 seen in extreme cases of experimental acid intoxication, and which 
 Avas found to be associated with the presence of 46-2 grams of 
 oxybutyric acid in the urine. Further, it is well known that the 
 administration of alkalies will often diminish, or abort, the 
 symptoms of diabetic coma, and that this effect is associated with 
 the elimination of an increased amount of organic acids in the urine, 
 indicating their previous retention within the body, owing to the 
 lack of alkali with which they could combine. It has been calculated 
 that the quantity of native alkah, chiefly sodium carbonate and 
 sodium phosphate, in the entire body is only equivalent to 60 grams 
 of sodium hydroxide. This amount is so small that it would 
 speedily be exhausted by a persistent production of even small 
 amounts of acid, but thanks to the carnivorous habits of man he 
 is able to protect himself against the production of large amounts 
 of acid in virtue of the ammonia that can be derived from the protein 
 of his food. According to v. Noorden, an output of ammonia of 
 
212 
 
 GLYCOSURIA 
 
 from 4 to 6 grams a day is not at all uncommon in diabetic acidosis, 
 and Stadelmann once found 12 grams. When it is remembered 
 that the normal amount of ammonia in the urine lies between 0-3 
 and 0-7 grams, according to the diet, and that each gram above 
 that accounted for by the food corresponds to 6-12 grams of oxy- 
 butyric acid, it will be seen what a great protection the ammonia 
 is, and what large quantities of acid it can neutralise. 
 
 The amount of pathological acid in the urine is most accurately 
 determined by the laborious process adopted by Goetghens of 
 estimating the bases and comparing their total alkali value, ex- 
 pressed in terms of sodium, v^-ith. the total acid value of the known 
 acids of the urine. By this means the following results were ob- 
 tained in a healthy person (A), and a case of advanced diabetes 
 with threatening nervous symptoms (B), by Herter : — 
 
 Bases. 
 
 Grams 
 (A). 
 
 11 
 Grams , ... 
 
 i 
 
 Grams Grams 
 (A). (B). 
 
 K.,0 .... 
 
 Na,0 .... 
 CaO .... 
 Mi^O .... 
 
 NlNHg) . . . 
 
 0-9232 
 3-8810 
 0-2154 
 
 0-0806 
 0-4707 
 
 2-5540 
 2-4450 
 0-8035 
 0-1973 
 3-1130 
 
 SO3 (preformed) 
 SO3 (combined) . 
 P263 (bibasic) . 
 P9O5 (monobasic) 
 Uric acid . . . 
 Ci 
 
 0-9733 0-6857 
 0-0356 - 0-1253 
 0-7432 0-8521 
 0-1987 0-1756 
 0-0584 0-0271 
 4-4760 1-5110 
 
 Total bases . . 
 
 5-5709 
 
 9-1128 
 
 Total acids . . 6-4852 3-3768 
 
 In the healthy person the total acids exceeded the total basis 
 in this instance by 0-9143 gram, but in the diabetic there was an 
 excess of bases over acids which equalled 5-736 grams. In the 
 former case the apparent excess of acid was due to the j)resence 
 of some organic base with which the acid was united, while in the 
 latter the excess of base must correspond to some organic acid not 
 allowed for in the estimation. Assuming that the acid was beta- 
 oxybutyric acid, this amount of sodium would correspond to 
 about 26 grams. A rough, and for clinical purposes sufficient, 
 index of the degree of acidosis can, however, be obtained by 
 estimating the dailj^ excretion of ammonia, for, as a rule, ammonia 
 is the base that is chiefly increased in diabetes. The acetone, 
 aceto-acetic, and oxybutjTic acid may be separately estimated by 
 the methods already described, but the processes involved are too 
 lengthy for routine work. 
 
 The explanation of diabetic coma by the theory of an acid 
 
PERSISTENT GLYCOSURIA 213 
 
 intoxication due to the formation of beta-oxy butyric, and aceto- 
 acetic, acid is at present widely held, but why these acids should 
 be formed is as much a puzzle as ever. The statement that the 
 body has lost, its oxidative powers is, after all, only a cloak for 
 our ignorance on the matter. It has been pointed out by Kraus, 
 Rumpf, and others that in a few cases of diabetes coma develops 
 without any increase in the elimination of organic acids in the urine, 
 and that alkali therapy has not alwaj^s the remedial effect that 
 might be expected if the coma were always a result of acidosis. 
 Moreover, cases have been reported by v. Noorden and other ob- 
 servers in which large amounts of acetone (5 to 6 grams) and oxy- 
 butyric acid (30 to 40 grams) have been excreted in the urine, 
 and yet the joatient has lived comfortably for years. It has been 
 suggested by Naunyn that the coma in cases without increased 
 acid formation, may be due to the presence of some unknown 
 toxine which exerts a direct action on the cerebral cells, and 
 especially those of the respiratory centre. Klemperer holds the 
 view that both the coma and the abnormal production of acid 
 in diabetes are due to the presence of some such toxine. Labbe 
 maintains that while acidosis and diabetic coma are closely related 
 they are probably due to different mechanisms, and points out 
 that many features of coma indicate an intoxication with poly- 
 peptides. It has been joointed out by Forges that, since the 
 amount of carbon dioxide in diabetic blood is abnormally low, and 
 it is known that this substance is of importance in relation to the 
 a.ctivity of the heart, it is possible that several of the symptoms 
 of coma result from the lowered carbon dioxide tension. Although 
 the diversity of the symptoms met with in different cases of 
 diabetic coma suggests that more than one cause may be operative, 
 these explanations belong rather to the realm of theory than of 
 fact, at an}'- rate at present. 
 
 Symptoms of Diabetic Coma. — Coma is one of the most frequent 
 and serious symptoms of diabetes, but is more common in persons 
 under, than over, forty. It is not necessarily a late symptom, 
 but may occur apparently early in the course of the disease. The 
 clinical symptoms are, however, always preceded by the chemical 
 signs of acidosis ; for, in the large majority of cases at least, coma is 
 but the culminating point of an acid intoxication. It is therefore 
 of the utmost importance that the urines of all diabetics should 
 be watched for evidence of this condition. This is true not only 
 of the clinically severe types, but also of the milder forms, since 
 fatigue, anxiety, complicating and intercurrent diseases of various 
 kinds, alcoholic intoxication, general anaesthesia, and complete 
 
214 GLYCOSURIA 
 
 abstinence from carbohydrate food are all liable to rapidly develop 
 the acidosis and bring about a condition of severe intoxication, 
 with its attendant danger of coma, in any case where abnormal 
 acid formation has commenced. 
 
 In the majority of cases the onset of the coma is more or less 
 sudden, often coming on without any apparent cause. In some 
 there may be prodromal symptoms extending over several days, 
 or, occasionally, weeks. The most common of these are loss of 
 appetite, and obstinate constipation. When a diabetic complains 
 of an impairment of appetite and gastric disturbances after every 
 meal, it is always a grave sign, especially if the bowels are not 
 being freely opened, and the liver is found to be enlarged. Ac- 
 cording to Lepine acceleration of the pulse is often an early, 
 although, not a characteristic, symptom. Lassitude and apathy, 
 or mental irritability and restlessness, with, maybe, epigastric 
 pain and vomiting, are often the first indications that are noticed. 
 The patient soon becomes drowsy, and the drowsiness gradually 
 verges into coma. At first it is possible to rouse him, but this 
 becomes more and more difficult. Meanwhile the respirations, 
 although regular, are much increased in range, and inspiration 
 is prolonged, so that the breathing is often of a sighing character, 
 a condition usually described as " air-hunger." The breath has 
 a characteristic sweet, fruity, smell, that often pervades the whole 
 room, and, with the character of the respirations, at once suggests 
 to an experienced observer the cause of the coma, even if the 
 patient has not previously been seen. The action of the heart is 
 rapid and weak, the features become drawn. The eyes are half 
 closed, the pupils are sometimes dilated, sometimes contracted, 
 but they generally react well to light until deep coma sets in. The 
 extremities are cold, and are often cyanosed. The temperature, 
 which at the onset may have been temporarily raised, is subnormal, 
 although it may again rise just before death takes place. Twitch- 
 ings and convulsions occasionally occur. The urine is generally 
 much diminished in quantity, and is markedly acid in reaction. 
 Acetone and aceto-acetic acid, although usually present, are not 
 so abundant as before, but the amount of beta-oxybutyric acid is 
 generally much increased. The total output of nitrogen is increased 
 relative to the sugar, but the proportion of nitrogen excreted in 
 the form of urea is diminished, while the ammonia nitrogen is cor- 
 respondingly increased, but this too is in some instances diminished. 
 The urine nearly always contains albumen, and numerous granular 
 and hyaline casts are seen on microscopical examination of the 
 deposit. When coma is fully developed death almost invariably 
 
PERSISTENT GLYCOSURIA 215 
 
 follows, and usually takes place in from twelve to forty-eight hours 
 after the onset of the characteristic symptoms. 
 
 Beside the preceding, which may be termed the " dyspnoeal 
 form of diabetic coma, there is another type in which sudden collapse, 
 probably from failure of the heart, occurs soon after the drowsiness 
 and coma have developed, and yet a third in which a period of 
 acute excitement and ataxia, suggestive of acute alcoholic intoxica- 
 tion, precedes the coma. In a few instances, where death occurs 
 from coma in diabetes, it is clearly due to the ordinary causes, 
 such as ureemic intoxication, &c., but in these the character of the 
 symptoms, and the presence of definite renal, or other, complications 
 usually suffice to differentiate the condition from true diabetic 
 coma. 
 
 BIBLIOGRAPHY 
 
 Abderhalden, Zeit. f. phys. Chem. , xliv. 
 
 Barth and Autenrieth, Hoppe-Seyler's Zeit., 1902. 
 
 Biscofswerder, Inaug. Dissert., Berlin, 1896. 
 
 Blumenthal and Neuberg, Deut. med. Woch., 1901. 
 
 Boecker, Deut. Klinik., 1853. 
 
 Bose, Brit. Med. Journ., 1907. 
 
 Brandenburg, Zeit. f. klin. Med., xlv. 
 
 Bremer, New York Med. Journ., Ixiii., Ixvi. ; Cent. /. inn. Med., 
 
 1897. 
 Buhl, Zeit.j. Biol., 1880. 
 
 Bvixian and Schnr, Pfluger's Arch., Ixxx., xciv. 
 Busse, MiXnch. med. Woch., 1901. 
 Camxnidge, Lancet, 1904, 1909. 
 Davies, Lancet, 1909. 
 
 Desgrez, Compt. Bend. d. I. Soc. d. Biol., 1907. 
 Dickinson, Diseases of the Kidneys, 1875. 
 
 Downes and O'Brien, Intercol. Med, Journ. of Australia, 1909. 
 Von Dusch, Zeit. f. ration. Med., 1854. 
 Ebstein, Virchow's Arch., 1899. 
 
 Embden, Salomon, and Schmidt, Hofmeister's Beitrdge, 1906. 
 Embden and Schliep, Centralb. f. Path. u. Pharm., 1907. 
 Eppinger, Wiener klin. Woch., 1906 ; Zeit. f. exp. Path. u. Pharm., 
 
 1906. 
 Ewing, Clinical Diagnosis of the Blood, 1904. 
 Fischer, Virchow's Arch., 1903. 
 Frerichs, Diabetes Mellitus, 1884. 
 Folin, Harvey Lectures, 1908. 
 Funck, Deut. med. Woch., 1911. 
 Futterer, Verh. d. pi. med. Gesell., 1888. 
 Gabritschewsky, Arch. f. exp. Path., xxviii. 
 
216 GLYCOSURIA 
 
 Geekxiuyden, Zeit. /. phys. Chem., 1904. 
 
 Gerliardt, Wiener med. Presse, 1865. '• 
 
 Goetghens, Zeit. f. phys. Chem., 1885. 
 
 Griibe, Munch, med. Woch., 1895. 
 
 Habershon, Journ. of Path, and Bact., 1908. 
 
 Henriques, Zeit. f. phys. Chem. , xx vi. 
 
 Herrick, Am^er. Journ. of Med. Sci., 1900. 
 
 Herter, Lectures on Chemical Pathology, 1902. 
 
 Hoppe-Seyler, Physiol. Chem., 1880. 
 
 Jacoby, Zeit. f. klin. Med., 1897. 
 
 James, Edin. Med. Journ., 1896. 
 
 Jolles and Winkler, Arch. f. exp. Path. u. Pharm., 1900. 
 
 Kaulich, Prag. Vierteljah., 1860. 
 
 Kolisch, Wiener klin. Woch., 1897. 
 
 Ivraus, Zeit. f. Heilkunde, x. 
 
 Kiilz, Zeit.f. Biol., 1884. 
 
 Kussmaul, Deut. Arch. f. klin. Med., 1874. 
 
 Labbe, Presse Medicate, 1912. 
 
 Lecorche, Le Diabete chez les femmes, 1886. 
 
 Leichenstern, Untersuch. u. d. Hcemoglob., 1878. 
 
 Lepine, Diabete Sucre, 1909. 
 
 Loch, Zentralb. f. inn. Med., 1905. 
 
 Locke, Boston Med. and Surg. Journ., 1902. 
 
 Liithje, quot. Neuberg and Blumenthal, Deut. Arch. f. klin. Med., 
 
 1903. 
 Luzzato, Festsch. z. v. Salkowski, 1904. 
 Maguire, Fowler's Diet, of Med., 1890. 
 Mayer, Zeit.f. klin. Med., 1901. 
 Mendal and Rose, Journ. of Biol. Chem. , 1911. 
 Mies, Munch, med. Woch., 1894. 
 
 Minkowski, Arch. f. exp. Path., 1884 ; Berl. klin. Woch., 1892. 
 Mohr, Zeit. f. klin. Med., 1901 ; Deut. med. Woch., 1905 ; Zeit. f. 
 
 exp. Path. u. Therap., ii. 
 Moraczewski, Zeit. f. klin. Med., 1904. 
 Naunyn, Nothnagel's Spec. Path., vii. 6. 
 Naunyn and Reiss, Reichert. u. Dubois Arch., 1869. 
 Neisser and Berlin, Zeit. f. klin. Med., 1904. 
 Neubauer, Journ. f. prakt. Chem., Ixvii. 
 Neumann and Mayer, Zeit. f. phys. Chem., 1903. 
 Nicola, Gioin d. Roy. Acad. d. m,id., 1904. 
 Von Noorden, Die Zuckerkrank, 1901 ; Twentieth Cent. Practice, ii. 
 
 99 ; Handb. d. Ernahrungstherapie, 1904. 
 Orgler, Hofmeister's Beitr., 1901-2. 
 Orlowski, Centralb. f. Stoffwechsel, 1902. 
 Pavy, Lancet, 1878. 
 Fetters, Prager. Vierteljahr, 1857. 
 Forges, Wien. klin. Woch., 1911. 
 Quincke, Berl. klin. Woch., 1880. 
 
PERSISTENT GLYCOSURIA 217 
 
 Redier, Stomatologie, 1909. 
 
 Rumpf, Berl. klin. Woch., 1895. 
 
 Rupstein, Centralb. F. d. Med. Wissensch., 1874. 
 
 Satta, Hofmeister's Beitr., 1905. 
 
 Schmitz, Berl. klin. Woch., 1891. 
 
 Schneider, Munch, med. Woch., 1899. 
 
 Seegen, Diabetes Mellitus, 1893. 
 
 Stadelmann, Arch. f. exp. Path., 1883. 
 
 Strauss, Die Chron. Nierenentzund, 1902. 
 
 Tessier, These de Paris, 1876. 
 
 Tenbaum, Zeit. f. Biol, 1896. 
 
 Tirard, Lancet, 1909. 
 
 Waldvogel, Die Acetonkorper, 1903. 
 
 Waterman, New York Med. Record, 1882. 
 
 Williamson, Diabetes Mellitus, 1898 ; Med. Chronicle, 1909. 
 
CHAPTER VIT 
 
 PERSISTENT GLYCOSURIA — PATHOLOGY AND DIAGNOSIS 
 
 In some cases persistent dextrosuria is undoubtedly associated 
 with, pathological changes in the pancreas, which can be demon- 
 strated after death by the naked eye or by the microscope. In 
 others no pancreatic lesion can be discovered, and, while it is pos- 
 sible that, as V. Noorden has suggested, severe disturbances of the 
 chemical functions of an organ are not necessarily associated with 
 recognisable changes in its anatomical structure, it is probable 
 that in such cases the primary lesion is in some other organ. The 
 work of Falta, and others, on the influence exerted by the duct- 
 less glands on carbohydrate metabolism suggests, as we have seen 
 when considering experimental glycosuria, that affections of the 
 thyroid, pituitary, and supra-renal glands play a part in the 
 pathology of diabetes, and that it is to them we must look for the 
 primary cause, at least in some instances. The nervous theory of 
 diabetes that so long held the field has proved not to be universally 
 true, but it is quite clear that lesions, or disturbances of function, 
 of the brain, spinal cord, &c., may also give rise to persistent 
 glycosuria. It is therefore necessary that, in the first place, we 
 should consider what morbid changes have been met with in these 
 various structures in diabetes. 
 
 The Pancreas. — The pancreatic origin of certain cases of 
 diabetes appears to have been suspected by pathological anatomists 
 and clinicians long before a definite pancreatic theory of diabetes 
 was propounded. As far back as 1788 Cowley described a case in 
 which glycosuria was associated with disease of the pancreas. 
 The patient was a very stout man of thirty-five with an alcoholic 
 history ; post-mortem the pancreas was found to be atrophied, 
 and numerous calculi were present in the ducts. Chopart pub- 
 lished a similar case in 1821, and Bright, in 1833, gave an account 
 of a diabetic of nineteen with jaundice and fatty stools, whose 
 pancreas was found post-mortem to be atrophied, and to contain 
 a hard nodular tumour in the head, which was firmly adherent to 
 the duodenum. Isolated examples of similar conditions were sub- 
 sequently published by other observers with increasing frequency. 
 
 2J8 
 
PERSISTENT GLYCOSURIA 219 
 
 The first to definitely suggest a causal relationship between disease of 
 the pancreas and diabetes in £ome cases was Bouchardat, in 1846. He 
 based his belief, however, on the view that the glycosuria was depen- 
 dent upon alterations in the digestive functions of the gland. In his 
 Traite du Diabete, published in 1875, he said : " Si j'ai observe, chez 
 quelques glycosuriques, une alteration bien manifeste du jjancreas 
 ou de ses conduits, il est d'autres observateurs (et je suis moi- 
 meme de ce nombre) qui, pour la plus grande majorite de cas, n'ont 
 rien trouve d'anormal dans le pancreas des glycosuriques," a state- 
 ment which holds good to the present day. In 1877 Lancereaux, 
 basing his conclusions on the literature and two cases of his own, 
 confirmed Boucharclat's conclusions, but sought to distinguish a 
 special type, which he termed " cUabete maigre," characterised by 
 profound wasting and a rapid course, as being characteristic of 
 serious alterations in the structure of the pancreas. The fallacy 
 of this distinction was, however, proved by subsequent observers, 
 who showed that the glycosuria accompanying disease of the 
 pancreas is not associated with any particular symptoms, and that 
 " diabete maigre " may occur without there being serious structural 
 changes in the gland. 
 
 The experimental work of von Mehring and Minkowski, published 
 in 1889, established the pancreatic theory of diabetes on a secure 
 footing, for it proved that extirpation of the gland in animals 
 gives rise to symptoms more nearly resembling those met with in 
 severe human diabetes than can be produced by any other means, 
 and also showed that the influence the pancreas exerts on carbo- 
 hydrate metabolism is independent alike of the external secretion 
 of the gland, and of its nervous connections. These observations 
 aroused fresh interest in the condition of the pancreas in diabetes, 
 and a number of observers published statistics bearing on the point. 
 Among the earliest of these were the investigations of Windle, who 
 reported that in 139 cases of cUabetes the pancreas had been found 
 to be diseased in 74, or 53 per cent. Seegen, however, who analysed 
 the records of 92 cases stated that a pancreatic lesion had been 
 noticed in only 17 (19 per cent.), while Frerichs found disease of 
 the pancreas in 16 out of 44 cases of diabetes (36 per cent.). 
 Hansemann again reported a much higher percentage of pancreatic 
 lesions, 40 out of 54 (74 per cent.) in the cases of diabetes examined 
 after death in the Berlin Pathological Institute. Bloch collected 
 22 cases from the records of the Vienna General Hosjaital, and found 
 that in 12 (55 per cent.) the pancreas had been recognised as 
 abnormal. Oser quotes 42 cases with pancreatic lesions in 161 
 diabetics (26 per cent.). Williamson, in his work on Diabetes 
 
220 GLYCOSURIA 
 
 MeUitus, published in 1898, gives an account of 23 cases in which 
 special attention was paid to the condition of the j)ancreas, and 
 states that in 15 of these (79 j)er cent.) there was evidence of disease. 
 These mdely divergent results are no doubt, to a great extent, to 
 be explained by a difference of opinion as to what is to be regarded 
 as normal and what as pathological, and it is apj)arent from a 
 study of the published records that the more carefully and sys- 
 tematicalty disease of the j)ancreas has been searched for in cases 
 of diabetes, the more frequently has it been found. 
 
 The use of the microscojDe by recent observers has considerably 
 increased the proportion of cases in which lesions of the pancreas 
 have been discovered in association with diabetes, and this has 
 been particularly marked since attention was drawn to the possible 
 relationship of the islands of Langerhans to the internal secretion 
 of the gland. OjDie, who made a histological examination of 
 the pancreas in 19 cases, found some abnormality in 15 (79 per 
 cent.), and in 4 of these it was not until they were submitted to 
 microscoj)ical examination that a lesion was discovered. Bosanquet, 
 using the microsco23e, records disease of the pancreas in 17 out 
 of 19 cases (90 per cent.) that he investigated. Hansemann has 
 recently claimed that every case of " true " diabetes is associated 
 ^^ith demonstrable changes in the pancreas, if the gland is examined 
 quite fresh and before any auto-digestion has taken place. 
 
 According to the older observers the most common lesion of 
 the pancreas met with in diabetes is atrophy of the gland. Windle 
 found it in over 59 per cent, of the cases he examined, and Frerichs 
 in 75 per cent. The statistics quoted by Hansemann from the 
 Berlin hospitals in the space of ten years, show 49 cases of diabetes 
 with disease of the pancreas, in 36 (90 per cent.) of which there 
 was simple atrophy, and in 3 (8 per cent.) atrophy and sclerosis. 
 The more recent observations of Williamson and Opie give much 
 lower figures, the former finding simple atrophy in 4 out of 11 cases 
 (27 per cent.), and the latter in 4 out of 15 (26 j)er cent.). Some 
 explanation of this difference is afforded by the more exact methods 
 of observation employed by the modern observers, and there can 
 be no doubt that in the past too great reliance on the naked-eye 
 characters caused many cases to be classified as simple atrophy 
 which in reality were examples of atrophic changes resulting from 
 chronic inflammation of the gland. 
 
 In a few cases fatty degeneration of the pancreas has been found 
 after death as the only discoverable lesion. Bosanquet met mth 
 a recognisable degree of fatty change in 10 out of 100 cases, which 
 in 3 was combined with some fibrosis. Wilhamson in his series 
 
PERSISTENT GLYCOSURIA 221 
 
 found one case of lipomatosis in which there was atrophy and fatty 
 degeneration, and one where, beside atrophy and fatty degenera- 
 tion, there was e\'idence of inflammatory change. 
 
 The earliest recorded case in which disease of the pancreas was 
 found to be associated with diabetes was, as we have seen, one of 
 'pancreatic calculi. Hansemann, however, was only able to find 
 fourteen instances in 72 cases (19 per cent.) collected from the 
 literature, and Oser quotes but twenty-four examples in 188 cases of 
 diabetes (14 percent.), so that the association is not very common, 
 particularly as the lesion is so obvious that it would not be readily 
 overlooked. The mere presence of calculi cannot be regarded as 
 directly responsible for the diabetes, since blocking of the ducts by 
 ligature, or othermse, has been proved not to cause glycosuria. It 
 is to the fibrotic changes accompanying them that we must therefore 
 look for the explanation. That this is so is shoA\ai by the fact that 
 diabetes is only found in cases where there is very marked over- 
 growth of fibrous tissue, whereas in those instances where the 
 concretions are not associated with advanced interstitial changes 
 sugar does not appear in the urine. 
 
 In a similar way although cysts of the pancreas have been met 
 with in from 5 per cent. (Oser) to 7 per cent. (Dieckhoff ) of diabetics 
 showing pancreatic lesions, there are many cases of cysts in which 
 glycosuria does not occur. In some instances sugar may appear 
 in the urine some time after a cyst has been surgically treated, 
 owing probably to the advance of the chronic inflammatory changes 
 to which the formation of the cyst was originally due. A case of 
 this description, under the care of Dr. Churton, was operated on 
 by Mr. Mayo Robson in June 1896. 
 
 At the time of the operation the urine was free from sugar and 
 showed no other abnorniahty, save that it gave a well-marked "" pan- 
 creatic " reaction. In February 1905, I was able to obtain a twenty- 
 four hours' sample of urine, and a specimen of the faeces, from this 
 case. The former measured 62 oz. and had a sp. gr. of 1-030. It 
 reduced Fehling's solution, and gave characteristic glucosazone 
 crystals with phenylhydrazin. Quantitatively 4-5 per cent, of sugar 
 (80 grams in the twenty-four hoiu-s) was foiind. Aceto-acetic acid 
 was absent, but there was a trace of acetone. There was no albvunen, 
 bile-pigment, lu'obilin, or indican. BiaFs pentose reaction \^as nega- 
 tive. The total nitrogen, urea, m"ic acid, chlorides, phosphates, and 
 sulphates were formd to be normal in amount, but oxalates were in 
 excess (0"32 grams in the twenty-four hours). This specimen of urine 
 also gave a " pancreatic " reaction. The faeces were light yellow in 
 colour, and were faintly alkaline in reaction. There was no inarked 
 excess of fat, but the normal relation between the unsaponified and 
 
222 GLYCOSURIA 
 
 saponified fats was disturbed, the former constituting 15 per cent., 
 and the latter only 5 per cent, of the dry weight, thus poLating to there 
 being some interference with the digestive functions of the pancreas. 
 
 The association of cancer of the pancreas with diabetes is re- 
 latively uncommon. Windle found it in 4 per cent, of his cases, 
 Frerichs in 6 per cent., Dieckhoff in 7 ]3er cent., Heiberg in 3 per cent., 
 and Williamson once in a series of twenty-three consecutive cases. 
 Glycosuria has been met with in 6 per cent, of the cases of primary 
 malignant disease of the pancreas that I have examined, and once 
 where the gland was involved in a secondary growth. The latter 
 is of particular interest, for it demonstrated very clearly the im- 
 portance of the pancreas in carbohydrate metabolism in the 
 human subject, and also the value of the "pancreatic" reaction 
 in diagnosis. 
 
 The patient was first seen in December 1905 ; there was then an 
 abdominal tumour which was suspected to be pancreatic, but an 
 examination of the virine gave no " pancreatic " reaction, and there 
 was also at that time no sugar. An exploratory examination was 
 performed by Mr. Mayo Robson, and a growth was found in the first 
 part of the duodenmii, but quite free from the pancreas. On January 
 18 a second specimen of urine was examined, and foiond to be free 
 from sugar, but it gave a well-marked " pancreatic " reaction, sug- 
 gesting that the pancreas was then involved in the disease. At the 
 request of the patient's friends the abdomen was re-opened a few days 
 later, and it was then found that the growth had invaded the pancreas, 
 as had been suspected. In the early part of May 1906, examination 
 of the vu-ine showed 5*25 per cent, of svigar, and a modified " pancreatic " 
 reaction gave many fine crystals soluble in 33 per cent, sulphuric acid 
 in five to ten seconds. A month later the sugar had increased to 
 7 per cent., and a much less marked "pancreatic" reaction was ob- 
 tained. In July the urine contained 7*25 per cent, of sugar, and the 
 " pancreatic " reaction gave only a few crystals. In August, 7*5 per 
 ■cent, of sugar was present, and no crystals were found on carrying out 
 the modified " pancreatic " test. In October the urine contained 
 9*5 per cent, of sugar, and the " pancreatic " reaction was negative. 
 In spite of the high percentage of sugar in the Lirine, the general con- 
 dition of the patient remained fairly satisfactory, and she complained 
 of no other symptoms than thirst and a voracious appetite. Con- 
 siderable quantities of acetone and aceto -acetic acid were found in 
 the urine in May, but with careful treatment they gradually diminished 
 in amount, tuitil in the early part of October only traces could be 
 detected. Towards the end of October the gall-bladder was discovered 
 to be distended, and a few days later jaundice developed. The patient 
 died deeply jaundiced on November 5th. 
 
 In some cases of malignant disease of the pancreas glycosuria 
 
PERSISTENT GLYCOSURIA 223 
 
 has appeared as an early symptom, and has later disappeared, 
 while in others it has only been met with toward the termination 
 of the disease. The temporary glycosuria is probably to be ex- 
 plained by a transitory disturbance in the functions of the gland, 
 caused by an inflammatory reaction consequent on the spread of 
 the growth. It has also to be borne in mind that when a portion 
 of the pancreas has been destroyed, whether by growth, or as the 
 result of chronic inflammatory changes, the condition resembles 
 that produced in animals by partial extirpation of the gland, so 
 that if carbohydrates are excluded from the diet, or are much 
 reduced, an alimentary glycosuria that previously existed may 
 disappear. In most recorded cases where sugar has appeared in 
 the urine as a terminal symptom, either the whole organ has been 
 rejolaced by a mass of growth, or the portions that have remained 
 have undergone sclerotic changes, so that no normal pancreatic 
 tissue has been left to carry on the functions of the gland. 
 
 The absence of permanent diabetes in most cases of cancer of 
 the pancreas is due to the growth being limited, in many instances, 
 to one portion of the gland, generally the head. In about 29 per 
 cent, of cases, however, this explanation will not hold good, for 
 in about that proportion there is a diffuse growth affecting the 
 whole organ. It is supposed that in these cases either the tumour 
 cells possess the same secretory functions as the normal gland 
 tissue, or that the new growth insinuates itself between the pan- 
 creatic cells in such a way as to obliterate the normal structure of 
 the organ without destroying it entirely. That such a process of 
 growth is possible is shown by the presence, in some instances, 
 of unaltered islands of Langerhans in the midst of the cancerous 
 material, while in support of the former hypothesis Hansemann 
 points out that in primary carcinoma of the supra-renals Addison's 
 disease is rare. Lepine and Heiberg have both reported cases in 
 which cancer of the pancreas occurred in diabetics who had passed 
 sugar in their urine for several years previous to the onset of the 
 symptoms of malignant disease. Such cases rather favour the 
 view advanced by some, that carcinoma of the pancreas may 
 ■originate in groups of cells isolated by fibrosis of the gland in much 
 the same way as primary cancer of the liver appears to arise from 
 groups of cells similarly isolated in cirrhosis of that organ. 
 
 Inflammatory lesions of the pancreas and their sequelae are 
 by far the commonest pathological changes affecting the gland, 
 but until recently they have failed to receive the recognition their 
 importance deserves. It is consequently not surprising to find 
 that the association of diabetes with pancreatitis and its results, 
 
224 GLYCOSURIA 
 
 has been, to a great extent, overlooked, or that the condition has 
 been referred to some other cause. CalcuH and cysts, as we have 
 seen, are not of themselves responsible for the glj^cosuria with 
 which they may be associated, but occur in the course of a chronic 
 inflammation which ultimately destroys the structure of the gland ; 
 the special form of atrophy of the pancreas described by Hansemann 
 as common in diabetes is in reality a fibrosis due to chronic in- 
 flammatory changes, while the diabetes associated with some cases 
 of malignant disease is apparently brought about by a secondary 
 inflammation set up by the presence of the growth. Dieckhoff in 
 his analysis of fifty-three cases found acute pancreatitis in 10 per 
 cent., and chronic pancreatitis in 36 per cent. Williamson met 
 with four instances of cirrhosis of the pancreas in twenty-three 
 cases, and Opie with four of chronic inflammation in nineteen cases, 
 so that inflammatory change probably plays a not unimportant part 
 in the production of diabetes, especially if the secondary manifesta- 
 tions to which reference has been made are taken into account. 
 
 Acute pancreatitis is not a common disease, and for this reason 
 alone is not frequently met with as a cause of diabetes. In 188 
 cases collected by Oser there were three in which glycosuria was 
 associated with haemorrhage into the pancreas, three with necrosis of 
 the gland, and six with abscess. In 100 cases of acute inflammation 
 collected by Fitz and by Sietz glycosuria was present in two. The 
 reason for the comparative rarity with which glycosuria occurs in 
 acute pancreatitis appears to be that when the whole organ is 
 destroyed death usually follows very rapidly, and when the pro- 
 gress of the disease is less acute portions of the gland are left 
 unaffected. The experiments of Guleke on dogs have shown that 
 when complete necrosis of the pancreas has been induced, by 
 injecting oil into the ligatured pancreatic duct, glycosuria always 
 occurs, but that when a portion of the pancreas has been left 
 intact no sugar is found in the urine. A fatal case of hsemorrhagic 
 pancreatitis with destruction of the whole gla.nd and associated 
 with the appearance of sugar in the urine was described by Bosan- 
 quet in his Goulstonian lectures. 
 
 The patient, a laundress aged fifty-three, was under the care of 
 Dr. J. M. Bruce in Charing Cross Hospital. A week before admission 
 she was seized with acute pain in the abdomen, which rapidly swelled 
 and became hard to the touch. She had previously had no symptoms 
 of diabetes, but now complained of thirst, and on examining the urine 
 it was found to contain from 10*2 to 11 -25 grains of sugar in the twenty- 
 four hours. Her temperature rose, and she had rigors on two suc- 
 cessive days. Finally she died collapsed, but without any symptoms 
 
PERSISTENT GLYCOSURIA 225 
 
 of coma. At the necropsy a breaking -down mass, with much bloody 
 flv^id, was found in the situation of the pancreas. There was also 
 diffuse fat necrosis and evidence of recent peritonitis. 
 
 Such a case constitutes a natural experiment on the removal 
 of the pancreas in a human being and, as Bosanquet points out, 
 the results exactly corresponded with those obtained in animals. 
 
 Benda and Stadelmann found 3 to 5 per cent, of sugar in the urine 
 of a patient who succumbed to hsemorrhagic pancreatitis with fat 
 necrosis, in whose previous history there was nothing to suggest 
 antecedent diabetes. In a few instances acute or subacute pan- 
 creatitis has been recovered from, but left the patient with per- 
 sistent glycosuria. Gifford Nash's case already referred to is an 
 example of this, for, although the sugar that appeared in the urine 
 during convalescence disappeared for some six or seven months, 
 a permanent glycosuria was ultimately established. Brentano has 
 described a case of necrosis and sloughing of the pancreas in 
 Avhich a sub-diaphragmatic abscess was opened. The patient even- 
 tually recovered, but left the hospital with a pancreatic fistula 
 and persistent glycosuria. 
 
 Chronic interstitial 'pancreatitis can be divided histologically 
 into two types — (a) an interlobular form, in which the overgrowth 
 of fibrous tissues takes place chiefly between the lobules, and (6) 
 an interacinar form, where the newly formed fibrous tissue is 
 diffusely distributed within the lobules and between the individual 
 acini. According to Opie, the former is rarely found to be associated 
 with diabetes, while glycosuria is a very much more common 
 symptom of the latter. To the first type belong the chronic in- 
 flammatory changes that result from obstruction of the pancreatic 
 duct, gall-stones, &c. The second variety is of unknown origin, 
 but its very constant association with arterio-sclerosis suggests 
 that they have a common cause, possibly an intestinal toxine. 
 Pancreatitis of the interlobular type was found by Opie to be 
 associated with glycosuria in only one out of twenty-nine cases, but 
 seven out of nine cases with interacinar pancreatitis had suffered 
 from diabetes. I have had considerable experience of the glycosuria 
 following symptoms of pancreatic disease and gall-stones, and have 
 already summarised the findings in 200 cases of diabetes that have 
 come under my observation. In another series of sixty-five conse- 
 cutive cases where biliary calculi were discovered in the common bile 
 duct at operation, and the pancreas was enlarged and hard, I was able 
 to detect sugar in the urines of only four (16 per cent.). The quantity 
 was very small in all of them, under 0-2 per cent, in three, and 0-4 
 per cent, in the fourth. After operation no sugar could be found 
 
 p 
 
226 GLYCOSURIA 
 
 in the urine of the former, but it was still present in the fourth ease, 
 where it slowly increased in amount. This patient died of diabetic 
 coma, I am informed, nineteen months after the operation. I 
 have met with interacinar pancreatitis in three cases of diabetes 
 which I have had the opportunity^ of examining after death, and 
 have found interlobular pancreatitis to be the lesion present in 
 six gall-stone cases I have investigated histologically. In 220 
 cases of diabetes collected by Windle gall-stones were present in 
 only one (0-45 per cent.). In another series of 142 cases of diabetes 
 collected by Williamson, biliary calculi had been also found in 
 only one of them (0-7 per cent.). Rolleston states that in a con- 
 secutive series of twenty-s3ven cases of diabetes examined at St. 
 George's Hospital gall-stones were present in four, and that in two 
 of these the calculi were in the common bile duct, and were 
 associated with chronic interstitial pancreatitis. 
 
 One great stumbling-block in the way of the readj^ acceptance 
 of the pancreatic theory of diabetes has been the very frequent 
 occurrence of lesions, and often very marked lesions, of the pancreas 
 without glycosuria ; but if we accept the view that the islands of 
 Langerhans are concerned in the elaboration of the hypothetical 
 internal secretion, and that so long as they are intact carbohydrate 
 metabolism will not be interfered \\dth, this difficulty is met, for 
 experiments upon animals, and observations on the human subject, 
 have sho^^^a that fibrosis of the gland, fatty degeneration, &c., 
 may involve the secreting parenchyma to a remarkable degree and 
 yet leave the cell-islets unaffected. 
 
 The very frequent occurrence of diabetes in interacinar pan- 
 creatitis, and its comparative rarity Avith the interlobular form, is 
 considered by Opie to depend upon the relationship of the fibrous 
 tissue overgrowth to the cell-islets ; for, oMdng to the diffuse distribu- 
 tion of the fibrous tissue in the former, the islands are affected 
 at the same time as the other elements of the gland, while in inter- 
 lobular pancreatitis the proliferating fibrous tissues invades the 
 lobules from the periphery, so that the cell-islets only suffer when 
 the process is far advanced and the secreting parenchyma has 
 been largely replaced by masses of scar-tissue. Opie's case of 
 interlobular pancreatitis mentioned above, in which glycosuria 
 was present, showed far-advanced induration Math fibrosis of the 
 islands of Langerhans, and the two cases of interacinar pancreatitis 
 in which diabetes was absent were both found to be in an early 
 stage of the disease, so that the cell-islets were unaffected. 
 
 Schafer in 1895 was the first to suggest that pathological changes 
 in the islands of Langerhans might be the cause of diabetes. In 
 
PERSISTENT GLYCOSURIA 227 
 
 1900 Ssobolew announced that the cell-islets were absent in two 
 <!ases of diabetes that he examined. The same, and the following 
 year, Opie published accounts of marked lesions of the islands of 
 Langerhans in cases of diabetes investigated by him. The latter 
 stated that he had met with hyaline degeneration of the islands of 
 Langerhans in seven out of nineteen cases (35 per cent.) of diabetes, 
 that in three the condition of the cell-islets was such as to render 
 them almost completely functionless, though the parenchyma was 
 relatively well preserved, in three others the lesion of the cell-islets 
 apj)eared to be less widespread ; while in another with severe 
 diabetes the islets were so altered as to be completely unrecognisable, 
 and the secreting parenchyma was in great part destroyed. Opie's 
 striking results aroused fresh interest in the subject, and stimu- 
 lated a more searching inquiry into the condition of the islands 
 ■of Langerhans in fatal cases coming to autopsy. Other observers 
 have since described hyaline degeneration of the islands of Langer- 
 hans in diabetes, but it would not appear to be by any means as 
 ■common as Opie's experience suggests, and the possibiUty of the 
 degeneration being a secondarv change has also not been excluded. 
 Bosanquet has reported a case in which he found it apart from 
 ■diabetes, in association with extensive arterio-sclerosis of the 
 pancreatic vessels, in a woman who died after an operation for 
 gall-stones. According to Weichselbaum and Stengel, who studied 
 the islands of Langerhans in thirty-five cases of diabetes, the lesion 
 most frequently met with is simple atrophy of the cells, with 
 vacuolisation and liquefaction of the protoplasm. Sclerosis of the 
 ■cell-islets was met with in four out of their second series of sevente3n 
 cases. As the result of an investigation of a further series of 183 
 cases, Weichselbaum came to the conclusion that the changes in the 
 islands of Langerhans are constant and peculiar to diabetes, and 
 that they ma}^ be classified under four headings — (1) Dropsical 
 degeneration, in which the epithelial cells become translucent, 
 while some are destroyed ; there may be small-celled infiltration 
 of the capsule with hypertrophy of the connective tissue, and 
 consequent atrophy and chsappearance of the epithelial cells of the 
 islands ; this lesion is seen in subjects under forty, and is never 
 associated with arterio-sclerosis. (2) Sclerosis of the islands with 
 hypertrophy of the surrounding connective tissue and chronic 
 interstitial pancreatitis, often accompanied by fatty change ; it is 
 seen in persons over fifty, and is always associated with arterio- 
 ■sclerosis. (3) Hyaline degeneration of the sheaths of the vessels 
 with consequent compression of the cells of the islands ; it is 
 ;seen in advanced age, and is associated with sclerosis of the 
 
228 GLYCOSURIA 
 
 pancreatic vessels ; this form is often combined with the second, 
 (4) Haemorrhage into the islands, which may occur apart from 
 diabetes ; regeneration or hypertrophy of the islands often ensues, 
 Weichselbaum did not find analogous lesions in any other disease, 
 acute or chronic ; he accounts for the fact that many have failed 
 to notice them on the ground that often dropsical degeneration 
 only is present, which may easily pass unperceived, and that the 
 lesions may be restricted to portions only of the pancreas. Herzog 
 studied three cases of diabetes in which the cell-islets were th& 
 seat of marked sclerotic changes, and Schmidt has met with twa 
 in which there was advanced interacinar pancreatitis, so seriously 
 involving the islands of Langerhans that they were converted inta 
 fibrous tissue balls resembling fibrosed glomeruU. In a case re- 
 ported by Lepine the islands were surrounded, and in places partly 
 destroyed, by a new growth of fibrous tissue. Gentes has also' 
 described a case of diabetes with chronic interstitial pancreatitis, 
 invading the islands of Langerhans. An acute inflammation, 
 limited to the cell-islets, was met with by Schmidt in the case of 
 a child of ten Avhose urine contained 6" 8 per cent, of sugar. Focal 
 necrosis of the pancreas involving the islands of Langerhans has 
 been described by Opie in one case. Absence of the islands, or a- 
 diminution in their number, has been reported by several observers ;. 
 but as Opie has pointed out, it is necessary that the sections of 
 diseased pancreas should be compared with preparations from 
 corresponding parts of a healthy gland before any conclusion is- 
 dra-v^TL as to the absence or diminution of the cell-islets, for their 
 distribution varies very much in different parts, being very numerous 
 in some situations, particularly the tail, and almost absent in 
 others. Opie compared the size and distribution of the cell-islets- 
 in the head, body, and tail of the pancreas in eight cases of diabetes, 
 and found that the figures obtained showed no constant departure 
 from the normal. A striking diminution in the number of islets 
 was, however, seen in two other cases, and in one of these, a child 
 of fourteen, the diabetes was hereditary, suggesting that it might: 
 be due to a congenital defect of the gland. 
 
 Dale and others, as we have seen, deny, on histological grounds, 
 that the islands of Langerhans are independent structures, and that- 
 they therefore have any particular function in carbohydrate 
 metabolism. Some observers have come to a similar conclusion 
 as the result of pathological investigations. The chief of these 
 has been Hansemann, who investigated thirty-four cases of diabetes, 
 and found that cell-islets were present in all. He states that in 
 some, where nearly the whole parenchyma had been destroyed by 
 
PERSISTENT GLYCOSURIA 229 
 
 fat, or interstitial fibrosis, they were diminished in number but 
 were structurally unchanged. In six cases he found that the 
 islands were invaded by what he regarded as hyaline connective 
 tissue, but since they were not all affected, and there was an 
 accompanying fibrosis of the gland, he considered that it was a 
 matter of chance as to whether the islands were involved or not, 
 .although he admits that he has not met with a case in which 
 fibrosis affected the cell-islets without diabetes being present. 
 Herxheimer, studying the cell-islets in the cirrhotic pancreas so 
 often met with in diabetes, found evidences of their new formation 
 from the small ducts, but regarded the whole pancreas as con- 
 trolling sugar metabolism. He thinks, however, that diabetes 
 may be due to functional changes which may, or may not, be 
 accompanied by visible lesions. He considers that, in man, the 
 islets alone are inadequate for the prevention of diabetes, while, 
 in animals, they seem to be sufficient. This conclusion is not, 
 however, supported by convincing evidence, and is directly contro- 
 verted, as regards human diabetes, by a case described by S. G. 
 .Scott, in which structures having the characters of islands of 
 Langerhans were found as the only recognisable remains of the 
 pancreas in a mass of fibrous tissue associated with malignant 
 disease of the head of the gland, yet there was no trace of sugar 
 in the urine. Pathological changes in the cell-islets without 
 diabetes have been described. Chauffard and Ravant met with 
 .swelling and increase in size of the islands without glycosuria in 
 thirteen cases of enteric fever, two of pneumonia, and one of 
 erysipelas . Salisbury Trevor observed similar changes in pneumonia 
 .and infective endocarditis. 
 
 The balance of available evidence, and the weight of opinion, 
 would seem to be in favour of the view that the islands of Langerhans 
 are specific structures having an internal secretion that is concerned 
 in carbohydrate metabolism. Their connection with diabetes may 
 therefore be assumed, but in view of the conflicting results reported 
 by different observers, many of whom have investigated but a 
 few cases, it is extremely difficult to draw satisfactory conclusions 
 as to the exact part they play, and the nature of the lesions that 
 exist. It is evident that what is needed is a careful correlation, 
 by unprejudiced observers, of the conditions present in a number 
 of patients during life, with the state of the pancreas and cell- 
 islets after death. Statistical studies are here of importance. 
 Sauerbeck in 1902 collected from various sources reports of 176 
 cases of diabetes in which the condition of the cell-islets had been 
 noted, and found that in 117 (66 per cent.) some abnormality bad 
 
230 GLYCOSURIA 
 
 been observed. If, however, the purely quantitative changes were 
 excluded as being of too indefinite a character, there remained ninety- 
 eight (62 percent.) in which qualitative changes had been met with. 
 The most extensive collection of cases j^et made by one observer 
 was pubhshed by Cecil in 1909. He investigated the jDathological 
 anatomy of ninety cases of diabetes, and came to the following 
 conclusions : (1) Anatomical lesions of the pancreas occur in more 
 than seven-eighths of all cases of diabetes mellitus. (2) When 
 lesions are found, the islands of Langerhans constantly show 
 pathological changes (sclerosis, hyaline degeneration, infiltration 
 "s^ith leucocj^tes, and hyiDertrophj-). (3) In some cases (12 out of 
 90) the lesions are limited to the islands of Langerhans. (4) In 
 sixteen cases associated ^^ith hyaline degeneration of the islands of 
 Langerhans, the average duration of the disease was three and a 
 half j^ears ; and in forty-six cases with sclerosis of these bodies the 
 average duration A\"as 3ii j'ears. In six cases associated with an 
 infiltration of leucocytes the average duration was eleven months. 
 (5) Destructive lesions of the islands of Langerhans maj^ be associated 
 Anth compensatory hypertrophy of other interacinar islands. (6) 
 Pecuhar adenoma-hke hj'j^ertxophy of the islands of Langerhans 
 occurred in a small proportion of cases (7 out of 90), and vas 
 associated with adenomata of the thyroid gland in two cases, and 
 of the pituitarj" body in one. (7) Fiftj^ per cent, of cases of diabetes 
 melhtus, occurring before the age of thirty j^ears, are associated 
 with lesions of the jDancreas ; 75 per cent, of all cases of diabetes, 
 in ^^hich the pancreas is normal, occur before the age of thirty 
 years ; 97 per cent, of cases, occurring after the age of thirty, are 
 associated A^ith lesions of the pancreas, and 86 per cent, occur in 
 conjunction T\-ith chronic interacinar j)ancreatitis accoinpanying 
 arterio-sclerosis. (8) Interacinar pancreatitis, which occurs in 
 73 per cent, of cases, is almost constantly associated with arterio- 
 sclerosis and gangrene of the extremities, which occurs with one- 
 fourth of cases of interacinar pancreatitis, and is referable to the 
 same cause. (9) Chronic interlobular pancreatitis, when associated 
 A^ith diabetes, is accompanied by sclerosis or hj^aline degeneration 
 of the islands of Langerhans. 
 
 The association of arterio-sclerosis, gout, sj^phihs, and alcoholism, 
 with chronic gtycosuria is probably to be explained by the fibrotic, 
 and degenerative, changes which each is capable of setting up in 
 the j)ancreas. As a rule the diabetes is of a mild type, such as 
 might be expected from a slowly progressing interstitial pancreatitis. 
 The glycosuria consequent on the presence of gall-stones is, in my 
 experience, usually of a similar mild type. The apparently in- 
 
PERSISTENT GLYCOSURIA 231 
 
 fective origin of some cases of diabetes may be due to the effects 
 produced upon the pancreas by micro-organisms entering the ducts 
 from the duodenum, and the history of digestive distiirbances 
 given in many cases suggests that a chronic duodenal catarrh may, 
 in some instances, give rise to conditions favourable to the onset 
 of a pancreatic lesion that eventually causes glycosuria. Hirschf eld 
 has pubhshed cases which suggest that pancreatitis, with secondary 
 glycosuria, may be set up through an infection carried by the blood, 
 and he thus explains the sugar sometimes met with in the urine 
 after influenza, tonsiUtis, and various infective processes. 
 
 A few cases have been reported in which there has been no 
 glycosuria, although the whole of the pancreas appeared to have 
 been destroyed by mahgnant disease, or inflammatory processes ; 
 but none are of recent date, and in most instances the proof 
 of total destruction has rested upon naked eye examination 
 alone. 
 
 Even if the connection of chseases of the pancreas, including 
 more or less marked lesions of the islands of Langerhans, with 
 diabetes is granted, a certain number of cases remain in which 
 no marked morbid change could be discovered in that organ by 
 competent observers. Opie met with four in his nineteen cases, 
 Williamson with eight out of twenty-two, Ssobolew with two out 
 of fifteen, and in twenty-three examined by Schmidt there were 
 no pathological changes in the pancreas in eight, and in eight 
 others the alterations were so shght as to be considered secondary 
 to the diabetic condition. Cecil states that in 12 per cent, of the 
 cases he examined no distinctive morbid condition of the pancreas 
 could be discovered, although in approximately half of these he 
 considers that the size of the gland, or the number of islands 
 of Langerhans, was less than normal. It is therefore probable 
 that only a proportion of cases of diabetes, possibly some 75 per 
 cent, at the outside, can be referred to structural alterations in the 
 pancreas. There are some who suggest a higher figure, and a few 
 who have claimed that all cases of diabetes are of pancreatic 
 origin, but while allowing that there may be fine histological and 
 chemical changes, which we are unable to detect by the methods 
 at present available, most authorities are now agreed that there is 
 both chnical, and experimental, evidence, that the pathology of 
 all cases of diabetes is not the same, and that some are to be referred 
 to other causes. 
 
 Supra-renals. — Although the experimental evidence in favour 
 of the view that the supra-renals control the metaboHsm of carbo- 
 
232 GLYCOSURIA 
 
 hydrates through their internal secretion is now considerable, 
 the cHnical facts pointing to disease of these structures being 
 responsible for diabetes is as yet very meagre, possibly owing to 
 insufficient care having been devoted to their investigation. 
 Burghart described in detail a case of diabetes occurring in a 
 woman of thirty-nine in which the pancreas was normal, and a 
 sarcoma of the left supra-renal capsule was found post-mortem. 
 Three months before death she suddenlj^ developed thirst and 
 polyuria, and her urine was found to contain 8 to 10 per cent, of 
 sugar. Lepine has reported a case of sarcoma of the right supra- 
 renal capsule in a woman of sixty-four, whose urine contained 
 9 grams of sugar per litre, and who died comatose. There was a 
 history of loss of flesh for two years, and great thirst for one and 
 a half years before death, so that there is a possibiHty that the 
 diabetes antedated the growth. Tumours of the supra-renals 
 have also been observed bj^ Grawitz in association with diabetes. 
 Ogle has described a case in which small patches of fibrinous material 
 were found in the centre of each adrenal. This case is referred to 
 by NaunjTi as one of tuberculosis. Rabe has reported an example 
 of what he considered to be co-existing bronzed diabetes and 
 Addison's disease, in which the supra-renals were totally destroyed 
 by tuberculosis. A case of diabetes melhtus in which the right 
 adrenal was replaced by a small fibrous mass, showing caseous 
 tuberculous changes, and the left was larger than normal has been 
 described by Montgomery. In this instance there was also chronic 
 interstitial pancreatitis, but the islands of Langerhans were said 
 to be normal. The thjToid was thought to be atrophied, and 
 there was arterio-sclerosis. The urine contained from 2-5 per cent, 
 to 3*5 per cent, of sugar, 38 to 40 oz. being passed in the twenty- 
 four hours. It had a specific gravity of 1-024 to 1*028. Acetone and 
 aceto-acetic acid were present. Montgomery came to the con- 
 clusion that there was no causal relationship between the primary 
 tuberculosis of the supra-renal and the glycosuria, and that the 
 association of the two conditions was merely a coincidence. 
 
 HjqDeractivity of the supra-renals has been supposed by Lepine 
 to be the cause of the glycosuria observed to follow prolonged 
 hypertension, Neubauer has detected hyperglycsemia in patients 
 with chronic renal disease and a high tension pulse, and other 
 observers have met with hypertrophy of the chromaffin tissues 
 and hj^peradrenalism in such cases. Examples such as these, con- 
 sidered in the fight of the experimental work that has been done, 
 are suggestive, but the evidence of the interdependence of the 
 glycosuria and a pathological condition of the supra-renals, and 
 
PERSISTENT GLYCOSURIA 233 
 
 more especially of growths of these structures, is so slight that an 
 accidental association cannot be excluded. 
 
 Garrod points out that malignant growths originating in the 
 adrenal medulla are not uncommon in children, yet, in his ex- 
 perience, sugar is never found in the urine, nor is there a raised 
 pulse tension to suggest an excess of adrenalin in the blood. We 
 must remember, however, when considering the effects produced 
 by disease of the supra-renals that chromaffin cells, resembling 
 those of the medulla of those glands, are found in other situations 
 in connection with the symjiathetic nervous system, as strands in 
 the larger nerves and as oval masses about the ganglia, the so- 
 called paraganglia, especially in the carotid gland, and in the lower 
 animals, in the abdomen at the point of division of the aorta 
 (Zuckerkandl's organ), and in the coccygeal glands, &c. Swale 
 Vincent found that an extract of the abdominal chromphile body 
 of the dog has the same powerful effect upon blood pressure as an 
 ■extract from the medulla of the adrenals, so that it would apj)ear 
 that the other chromaffin tissue of the body has an internal secre- 
 tion resembling that of the supra-renals. Destruction of the 
 supra-renals by a growth may therefore be compensated by 
 hypertrophy of the remaining chromaffin tissue, while any influence 
 that causes hyperactivity of the chromaffin tissue as a whole will 
 tend to produce glycosuria. The affinity of ejDinephrin for structures 
 that are, or have been, in intimate comiection with sympathetic 
 nerves, and the close relation of the chromaffin tissue to these 
 nerves, offers an explanation of the way in which a very slight 
 increase in the activity of the chromaffin tissue cells may produce 
 results apparently out of all proportion to the intensity of the 
 lesion present. The experiments of WooUey and Newburgh on the 
 results produced by the injection of indol and tyrosin into animals 
 suggest that these, and possibly other lower derivatives of protein 
 decomposition in the intestine, may cause hyper-activity of the 
 adrenals, and hence be a factor in disturbing the pancreas-chromaffin 
 equilibrium , and give rise to glycosuria as a consequence. 
 
 It may be assumed that in Addison's disease a condition of 
 hypo-adrenalism exists, and that, if the supra-renals have the 
 function in metabolism that is ascribed to them, not only will 
 glycosuria be unknown, but that tolerance for sugar A\dll be raised. 
 Porges has shown that the blood in this disease contains an ab- 
 normally low proportion of sugar, and he, as Avell as Eppinger, 
 Falta, and Rudinger, state that patients suffering from Addison's 
 disease exhibit an abnormally high tolerance for sugar. In seeming 
 contradiction to these results, however, is the case recorded by 
 
234 GLYCOSURIA 
 
 West, in which, although the supra-renals were corapletely atrophied! 
 and the skin was pigmented, there was well-marked glycosuria, 
 Unfortunatelj^ no information is furnished as to the condition of 
 the pancreas and other viscera in this case. 
 
 In view of the small amount of evidence available, and its 
 somewhat contradictory character, it is impossible to come tO' 
 a definite conclusion as to the effects of disease of the supra-renal 
 capsules on carbohj^drate metaboUsm, but it seems not unlikely 
 that some cases of persistent dextrosuria are dependent upon 
 h}^Derfunction of these structures, and possibly also of the 
 remaining chromaffin tissue of the body. 
 
 Pituitary Gland. — One of the earliest reported cases in 
 which disease of the pituitary gland was found to be associated 
 Avith glycosuria was that of Lepine. This patient suffered from 
 acromegaly, and during life passed about 7'1 per cent, of sugar in 
 her urine. After death a large tumour of the pituitary was found. 
 Four years later Marinesco published a similar case. In this- 
 instance the acromegaly was known to have antedated the glycos- 
 uria by about three years. Cases in which acromegaly has been 
 associated A^ith the presence of sugar in the urine have been 
 reported by Naunyn, Williamson, Loeb, Launois, and Roy, and 
 others. Hansemann found that diabetes was present in twelve out 
 of ninety-seven cases of acromegaly, and Schleisinger described three 
 cases — one "oith diabetes, one A^dth alimentarj^ glycosuria, and one 
 with transient glj'cosuria. Borchardt, reviemng the literature in 
 1908, states that in 176 cases of acromegaly, glycosuria had been 
 observed in sixty-three (35 per cent.), and that there was a lowered 
 tolerance for sugar in eight others. The latter symptoms had ap- 
 parently been looked for in only ten cases. He points out that in all 
 the cases of pituitary tumour recorded between 1886 and 1908 in 
 which glycosuria was present there were acromegaHc symptoms. 
 
 As a rule the sequence of the sj^nptoms has left httle doubt that 
 the glycosuria was a consequence of the disease of the pituitary 
 giving rise to the acromegaly. Occasionally, however, sugar has 
 been found in the urine before the s5anptoms of acromegaly ap- 
 peared, as in a case quoted by Schleisinger. In this instance sugar 
 was detected in the urine at the age of six, but under treatment 
 it disappeared until the symptoms of acromegaly were manifested, 
 when it appeared anew. Subsequently it was found at irregular 
 intervals, and without Rny apparent relation to the diet. A 
 similar intermittence in the glycosuria has been noted in other 
 cases. Thus Stiiimpell had under observation for several years a 
 
PERSISTENT GLYCOSURIA 235 
 
 case in which thirst and polyuria were prominent symptoms and 
 the urine contained as much as 100 to 120 grams of sugar in the 
 twenty-four hours, yet during two separate periods, with an 
 interval of four years between them, the sugar completely dis- 
 appeared in spite of a diet containing an abundance of carbo- 
 hydrate. In a few instances the glycosuria, after persisting for 
 years, has permanently disappeared. Such cases are referred to 
 by Borchardt and Pineles. 
 
 Goetsch, Gushing, and Jacobson investigated a series of twenty 
 cases of acromegaly, or gigantism, and failed to discover sugar in 
 the urine of any one of them. They found, on the contrary, that 
 there was a marked increase in the tolerance for carbohydrates. 
 The results of the animal experiments carried out by these ob- 
 servers, to which reference has already been made, appear to 
 offer an explanation of the difference between their findings and 
 those which the experience of previous authors would lead one to 
 expect, for they suggest that in the early stages of the disease there 
 is hyperfunction, not only of the anterior lobe, v/hich leads to the 
 skeletal changes, but also of the posterior lobe, leading to increased 
 carbohydrate tolerance. It may also be mentioned here that, 
 according to Goetsch, Gushing, and Jacobson, primary hypo- 
 function of the posterior lobe may result from direct pressure by 
 interpeduncular tumours, by growths originating in the lobe 
 itself, and by the distant effects of a tumour causing an obstructive 
 hydrocephalus, and thus damming back the fluid medium carrying 
 the posterior lobe secretion. In such cases the same symptoms 
 are seen as in the later stages of acromegaly. There is a high 
 tolerance for sugar, marked adiposity, due to a lowered con- 
 sumption of sugar, and a subnormal temperature, perhaps arising 
 from deficient oxidation. The administration of posterior lobe 
 extract lowers the carbohydrate tolerance, and by giving sugar in 
 the amounts usually just sufficient to cause glycosuria in normal 
 individuals (about 2 grams of dextrose or 1-4 grams of levulose 
 per kilogram), and administering posterior lobe in increasing doses 
 at the same time until glycosuria is produced, it is possible to 
 determine the proper therapeutic dose of the gland for each case. 
 Gushing and his associates are also inclined to think that the 
 transient glycosuria seen in cases of fracture of the base of the 
 skull is due to temporary stimulation of the hypophysis, and that 
 over-activity of the gland also explains the polyuria that occurs 
 after such accidents. The earher hyperactivity of the posterior lobe 
 may later give way to a condition of hypo-activity with increased 
 carbohydrate tolerance, as in a case described by these authors. 
 
236 GLYCOSURIA 
 
 The quantity of sugar excreted in acromegaly and gigantism 
 is often very large, but it frequently varies considerably. In 
 Achard and Loeper's case 380 grams were passed in the day, 
 Lancereaux's case passed 6 to 8 htres of urine containing 180 to 240 
 grams of sugar in the twenty-four hours, that of Perwuschin and 
 Foworski 3 to 8 litres, containing 30 grams of sugar per htre, while 
 in Ravaut's case from 15 to 20 litres, containing 500 to 1500 grams 
 of sugar, were excreted daily during several months. As a rule 
 there are no symptoms of any secondary disturbance of metabohsm 
 such as are seen in severe cases of diabetes. Ravaut and Prau 
 with Proescher have met with acetonemia, but there appears to be 
 only one recorded case, that of Stadelmann, which died of diabetic 
 coma. Herxheimer has reported a case in which there was 
 hpsemia. 
 
 In some cases of acromegaly with glycosuria structural lesions 
 of the pancreas have been described, both of the gland as a whole 
 and of the islands of Langerhans. Dallemange, Hansemann, 
 Pineles, Norris, and Cecil have observed sclerosis and other patho- 
 logical changes. The alterations in the pancreas appear, however, 
 to have been comparatively slight, and Borchardt states that in 
 most cases the gland has been found normal. That disease of the 
 pancreas can exist in acromegaly without there being any glycos- 
 uria, is shown by Stadelmann's case, in which partial sclerosis 
 was found after death. On the other hand, in the fatal case quoted 
 by the same author in which there was severe diabetes and death 
 occurred from diabetic coma, the pancreas was stated to be quite 
 healthy. Cecil considers that the glycosuria in acromegaly may 
 be referred to lesions of the islands of Langerhans, chiefly sclerosis 
 and hyaUne degeneration, with adenoma-hke hypertrophy. Most 
 authorities hold, however, that the interference with the internal 
 secretion of the hj^ophysis is the primary cause of the glycosuria, 
 and that any lesions of the pancreas that may be present are 
 probably secondary. 
 
 An increase in the volume of the thyroid has been noticed 
 in some cases, as, for example, those of Ferrand, Lancereaux, 
 and Henrot. In the two latter there were also the symptoms of 
 exophthalmic goitre. 
 
 The Thyroid. — It has been already pointed out that the 
 tolerance for sugar of patients suffering from hypothyroidism 
 (myxoedema) is increased, while with hyperthyroidism (exoph- 
 thalmic goitre) it is lowered, so that ahmentary glycosuria can 
 be readily produced. A number of cases have also been reported 
 
PERSISTENT GLYCOSURIA 237 
 
 in which intermittent, or persistent, glycosuria has been associated 
 with Grave's disease. DumontpeUier in 1867 appears to have been 
 the first to describe such a case. In his patient the urine contained 
 6 per cent, of sugar. A few years later Lauder Brunton reported 
 a similar case, and in the following year Wilks gave an account of 
 a patient who passed 300 grams of sugar a day. In 1878 Hartmann 
 reported two cases of exophthalmic goitre with glycosuria, and 
 O'Neill described a similar case. Since then a considerable number 
 have been recorded by other authors. In many of these there is 
 no evidence to show which condition preceded the other ; but as 
 up to the present there has been only one recorded case, that de- 
 scribed by Griibe, in which glycosuria was the first symptom, and 
 there is no evidence to suggest that diabetes can give rise to exoph- 
 thalmic goitre, it is not unfair to assume that the condition of 
 the thyroid is in some way responsible for the diabetes, especially 
 as this is in accordance with the results of experimental work. It 
 may also be noted that in a few cases exophthalmic goitre and 
 diabetes have occurred in different members of the same family. 
 Such have been reported by Mautry and Schmey, and Garrod 
 states that he has had under his care a young woman suffering 
 from the severest form of cUabetes whose mother had Graves' 
 disease. 
 
 It would thus seem that in exophthalmic goitre all the three 
 degrees of disturbed carbohydrate metabolism may exist. In some 
 glycosuria only results when sugar is taken — that is to say, it 
 is of the alimentary type. Occasionally only very small doses, 
 20 to 30 grams, are needed to cause sugar to appear in the urine, 
 while in others as much as 100 grams are required. In a few 
 instances the test dose may cause a glycosuria that persists for 
 several days, suggesting that the patient is on the verge of per- 
 sistent spontaneous glycosuria. With the second type of case, 
 into which the last mentioned merges, sugar is present in the urine 
 at intervals, sometimes for a day, at others for several days, prob- 
 ably owing to the patient's ordinary diet exceeding his tolerance 
 at these times. In the third class of case glycosuria is persistently 
 present, although it may be controlled by diet as in other forms of 
 diabetes. 
 
 Soon after the introduction of thyroid preparations into clinical 
 medicine it was noticed by Dale James, and Becl^re, that their 
 continued use might result in a reduction of tolerance for carbo- 
 hydrates and the spontaneous appearance of sugar in the urine, 
 also that on the drug being discontinued the glycosuria promptly 
 disappeared. These results are most readily induced in patients 
 
238 GLYCOSURIA 
 
 whose thryoid secretion is normal. Thus Garrod reports a case 
 in which a thjToid preparation was given to reduce obesity with the 
 result that sugar, to the extent of 4 per cent., appeared in the urine 
 at the end of a week, when only 4 grams had been taken. In 
 myxoedema a similar effect may, however, follow the continued 
 use of the drug. Thus Ewald has reported a case in which glycosuria 
 occurred on three successive occasions after the administration 
 of th^^:oid extract in very moderate doses, such as 5 grains a d&j, 
 as much as 5 to 6 per cent, of sugar appearing in the urine. In the 
 intervals the urine was free from sugar. Garrod states that an 
 examination of the urines of eleven myxosdematous patients, who 
 had been under treatment with tMToid extract in St. Bartholomew's 
 Hospital, showed that four contained sugar, and that the glycos- 
 uria quickly disappeared when the treatment was stopped. It 
 has been thought that in some instances the administration of 
 th}Toid has given rise to actual diabetes. Von Noorden suggests 
 that in such cases a pre-existent diabetic tendency has been 
 awakened into activity by the drug. Miiller has described a case 
 in which rapidly fatal diabetes occurred in a patient with exoph- 
 thalmic goitre who was treated with thyroid extract. 
 
 It must be pointed out that in many cases of exophthalmic 
 goitre no diminution of tolerance for sugar can be detected, while, 
 on the other hand, some patients suffering from myxoedema, or at 
 least from thyroid insufficiency, have passed sugar in their urine, 
 apart from any treatment with th}T:oid extract. A few of the latter 
 would seem to be undoubted examples of hypothyroidism, where 
 glycosuria would not be expected, but in others there is room for 
 doubt as to the correctness of the diagnosis. Such seeming ex- 
 ceptions to the generally observed effects of alterations in the 
 functional activity of the thyroid are possibly explained by the 
 interaction of other glands, and only serve to emphasise the diffi- 
 culty that surrounds the studj^ of organs that are so intimately 
 correlated. Some observers, notably Opie and Cecil, hold that the 
 glycosuria of Grave's disease is due to disease of the pancreas, but 
 although lesions of the gland have been found in a few fatal cases, 
 the fact that thjToid medication may cause sugar to appear in the 
 urine can only be reconciled with such a view by supposing that 
 hyperthyroidism produces a lesion of the pancreas. 
 
 The glycosuria of pregnancy is ascribed by Reichenstein, 
 Gushing, and others to the changes that occur in the ductless glands, 
 and more particularly in the thjrroid and pituitary body. Enlarge- 
 ment of the thyroid is a well-recognised event during pregnancy, 
 and changes in the pituitary have been shown to be common by 
 
PERSISTENT GLYCOSURIA 239 
 
 Erdheim and Stiimme. The case recorded by Marek tends to 
 support the view that changes in the ]Dituitary gland may be in 
 part responsible for the glycosuria in such cases. 
 
 The patient, a j^rimipara, aged twenty-six, developed the charac- 
 teristic symptoms of acromegaly dviring the eighth month of pregnancy. 
 Her hands and feet enlarged, her skin, nose, and lips were thickened, 
 and enlargement of the lower jaw prevented her teeth from meeting. 
 She became very somnolent. Sugar was present in her urine. After 
 delivery the glycosuria ceased in about two months, and by that time 
 all signs of acromegaly had also disappeared. It can hardly be doubted 
 that the glycosLiria and attendant symptoius were alike due to hyper- 
 pituitarism developed in association with the pregnancy. 
 
 It will be convenient to refer here to the glycosuria that has 
 been observed in association with diseases of the female generative 
 organs. Imlach described a case in which removal of the uterine 
 appendages, on account of pyosalpinx with ovarian adhesions, 
 was followed within a week by cessation of glycosuria which had 
 been detected five months before the operation, and which was 
 accompanied by thirst and polyuria. In cases recorded by Croom, 
 Beyea, and Henkel, removal of ovarian tumours, or uterine myomata, 
 has in hke manner been followed in the course of a few months, 
 and apart from restriction of diet, by disappearance of sugar from 
 the urine. In Beyea's case the glucose amounted to as much as 
 7 per cent. It will be noticed that in most cases the cessation of 
 the glycosuria has been much more gradual than after childbirth. 
 
 Nervous System. — There can be no doubt that nervous 
 disturbances enter into the pathology of many cases of diabetes, for 
 not only are there numerous well-authenticated cases where severe 
 or fatal diabetes has developed as the result of a severe mental 
 shock or an injury to the head, but the harmful effect of worry, 
 overwork, and shock on an established glycosuria are well known. 
 
 In such cases it is tempting to suppose that while the primary 
 effect of the emotional stress is on the nervous system, secondary 
 changes are produced in the glands with an internal secretion, and 
 that it is to a disturbance of their functions that the excessive 
 output of sugar is directly due. It is, however, difficult to estimate 
 the exact part that functional and emotional affections play, for 
 they give rise to no obvious lesion, but it is otherwise with the 
 persistent glycosuria that is met with as a result of gross morbid 
 changes in the brain, spinal cord, and nerves. 
 
 An analysis of the post-mortem records of the Allgemeine 
 Krankenhaus at Vienna by Seegen showed that in only eleven out of 
 
240 GLYCOSURIA 
 
 122 caces (9 per cent.) of diabetes was there any marked change in 
 the nervous system, and that in many of these it was doubtful 
 whether the lesion could be regarded as the cause of the glycosuria. 
 
 In some instances a disease involving the fourth ventricle has 
 been found. The earhest recorded case of this description appears- 
 to be that reported by Levrat-Perrotton, in 1859, in which a 
 colloid tumour of the choroid plexus was present. Five years 
 later v. ReckUnghausen published a similar case. Subsequently 
 other observers reported cases in which glycosuria had been found 
 in association with disease of the fourth ventricle, including soften- 
 ing (Richardson, Luys), sclerosis (Frerichs, and others), hsemorrhage 
 (Frerichs, &c.), cysticerus (Michael). A cj^stic sarcoma of the right 
 hah of the medulla, with glycosuria, was reported by Dompeling, a 
 tumour of each pyramid close to the pons by Frerichs, and tubercle 
 of the medulla by De Jonge. Lesions of the spinal cord, inclucUng 
 tumours compressing the cord in the cervical region (Smith), 
 haemorrhage and softening in the cervical and upper dorsal region 
 (Silver and Irvine), have also been met with in cases of diabetes. 
 Changes in the cerebral hemispheres, including a tumour of the 
 right temporal lobe and pachymeningitis, have been reported by 
 Frerichs, and lesions of the cerebellum by the same observer, and 
 by Hosier. Tumours at the base of the brain, and arterial changes 
 with softening, have been described by Richardson, Grossman, 
 and others. Bernardt collected reports of 485 cases of brain 
 tumour, and found that sugar was present in the urines of five — - 
 one being a tumour of the medulla, two of the hypophysis, one of 
 the cerebellum, and one of the cerebral hemispheres. Occasionally 
 glycosuria has been met with in association with disseminated 
 sclerosis, locomotor ataxia, chronic anterior poHomyehtis, and 
 other well-defined diseases of the nervous system, such as menin- 
 gococcal meningitis (Mannkopf, Adler, and others), tuberculous 
 meningitis (Loeb, Still, Aldren Turner, Grainger- Stewart, and 
 Garrod), &c. In most of these glycosuria is a rare comphcation, 
 but according to Garrod it is met with in 30 per cent, of aU cases 
 of tuberculous meningitis. 
 
 The microscopical changes met with in the central nervous 
 sj^stem in diabetes consist of sweUing of the myehn, vacuolar 
 degeneration of both the grey and white matter, and an increase in 
 the neuroglia ; but these, hke the similar changes seen in perni- 
 cious anaemia, tubercle, and cancer are more probably the result 
 of the action of a toxine than the cause of the glycosuria. 
 
 In a few cases tumours pressing on the vagus nerve have been 
 met with (Dulen). Lesions of the coehac plexus have been de- 
 
PERSISTENT GLYCOSURIA 241 
 
 scribed by Klebs and Munk. A thickening of the connective tissue 
 in the neighbourhood of the semi-hmar ganglion was met with 
 by Hale White in several cases. Cazzani has described lesions of 
 the sympathetic nerves, particularly in the neighbourhood of the 
 coeliac plexus. 
 
 Traumatism, particularly of the head and spinal cord, has 
 been noticed to be followed by persistent glycosuria in some cases, 
 but more frequently the condition is transitory. In 145 cases 
 investigated by Jodry the head was involved in seventy-two 
 (50 per cent.), the spinal cord in twenty-seven (20 per cent.), the 
 abdomen in twelve (8 per cent.), in 5 per cent, the patient had fallen 
 on his feet, and in 17 per cent, the site of the traumatism was not 
 clearly indicated. According to Lepine, the glycosuria manifests 
 itself the day after the injury in a third of the cases, and in the 
 majority during the first week. In a few cases, about 20 per cent., it 
 may be delayed, however, for two or three weeks, or even longer. 
 
 As a rule the quantity of sugar passed is small, not usually 
 more than 8 to 10 grams per Utre. According to Garrod, the sugar 
 in tuberculous meningitis rarely exceeds 1 per cent., usually being 
 about 0-3 per cent., and appears during the last week of life, 
 generally during the last two days. 
 
 Although in a small proportion of cases of persistent glycosuria 
 a definite lesion of the nervous system is found, it may be con- 
 cluded that in the majority it is either normal, or only presents 
 slight and unimportant alterations. It is very generally assumed 
 that in those instances where there is some gross cerebral change 
 the glycosuria results from implication of the diabetic centre. 
 This may be the explanation in a very few, but it is probable that 
 in most cases other causes are the active agents in its production. 
 The relation of the nervous system to the ductless glands that has 
 now been worked out, suggests that it is through disturbances in 
 the functions of these organs, either directly, as in the case of the 
 pituitary, or indirectly through nervous influences, as in the case 
 of the supra-renals, that the hyperglycsemia and consequently 
 glycosuria result. 
 
 A special form of neurogenic glycosuria is recognised bj" 
 V. Noorden, which he considers is comparatively harmless and is 
 controlled by suitable restriction of the diet. If neglected or 
 mismanaged it may, however, be converted into a typical diabetes. 
 A case of this description has been recently pubhshed by him in 
 detail. 
 
 The jDatient, a man of forty, had an inherited neuropathic taint, 
 but the glycosuria did not develop until the age of thirty-six, following 
 
 Q 
 
242 GLYCOSURIA 
 
 a period of great anxiety connected with the massacres in Russia. 
 Severe insomnia, loss of appetite and weight, and obstinate consti- 
 pation, were accompanied by 1"5 per cent, sugar in the urine, per- 
 sisting even on anti -diabetic diet. Sleepless nights were followed 
 by exacerbation of the glycosuria. Under slight restriction of carbo- 
 hydrates, with one day a week in which no carbohydrates were taken, 
 the patient lost his glycosuria. He kept well on this diet for a year. 
 Then another physician found the urine free from sugar on two 
 analyses, and advised the patient that anti-diabetic restrictions were 
 no longer necessary, so he commenced to eat at will. It was not 
 long before he had pains in the calves of the legs, he soon began 
 to lose weight, and 5*6 per cent, sugar was found in the m-ine. 
 Attempts to reduce the carbohydrates brought on acetonvu'ia, and 
 the man is now a confirmed diabetic. 
 
 Liver. — From the importance of the liver as a storehouse of 
 carbohydrates, it might be expected that striking lesions of that 
 organ would be found in a considerable number of cases of diabetes. 
 Glycosuria is, however, very rare, even when it is apparent that the 
 functions of the gland must be seriously interfered with. Clinical 
 and pathological experience show that extensive destruction of 
 the parenchjona ma}^ occur in cancer, cirrhosis, phosphorus poison- 
 ing, and other diseases without even a trace of sugar appearing in 
 the urine. Frerichs, who investigated the liver in fifty-five cases 
 of diabetes, came to the conclusion that it is generally normal 
 in volume, sometimes small, and very rarely enlarged. Saundby, 
 on the other hand, states that it is generally enlarged, weighing 
 68 to 80 ounces. The latter opinion agrees with the more recent 
 observations of E,6ssle, who considers that a diagnosis of diabetes 
 can be made in the post-mortem room with more certainty from 
 the appearance of the hver than from the condition of the pancreas. 
 He states that, while the normal hver constitutes from 2-3 to 2-75 
 per cent, of the body- weight, in diabetes it represents from 2-9 to 
 4-1 per cent., averaging 3-54 per cent. Beside the enlargement, it 
 is usually of a rosy colour, and the parenchyma is transparent, or 
 homogeneous, in appearance. Microscopically the diabetic hver, 
 according to this observer, shows interstitial changes, including 
 very constantly fatty degeneration of the stellar ceUs of Kupfier, 
 and homogeneous refractile bands along the capillaries, which he 
 beheves to be specific and characteristic of diabetes. It is generally 
 considered, however, that there are no constant macroscopical 
 or microscopical changes to be found in the hver. Beside enlarge- 
 ment, hyperaemia, and fatty degeneration, which are probably of a 
 secondary nature, and are possibly the result of hj^perfunction of 
 the gland, cirrhosis is the most common lesion met with. Claude 
 
PERSISTENT GLYCOSURIA 243 
 
 Bernard in his lectures on diabetes described a case of cirrhosis of 
 the Uver, occurring in an old alcoholic subject, who in the early 
 months of 1873 passed 6 litres of urine a day, containing 29 grams 
 of sugar per Htre. Later, as his condition became more serious, 
 and the cirrhosis advanced, the sugar began to diminish, until 
 only a trace could be found. The glycosuria was attributed by 
 Bernard to excitation of the glycogenic function of the Hver by the 
 disease in its early stages, and the disappearance of the sugar 
 to its extinction in the later phases. Other observers have also 
 reported the association of diabetes with cirrhosis of the liver, 
 and in some instances the glycosuria has been found to diminish 
 or disappear with the onset of cachexia. Among 128 cases of 
 diabetes observed in hospital by Naunyn seven, and of 158 in his 
 private practice twenty-two, were found to have cirrhosis of the liver. 
 The association appears to be rather accidental than causal, for 
 Frerichs has recorded cases of cirrhosis in which post-mortem ex- 
 amination showed almost complete degeneration of the Hver 
 parenchyma, and yet there was no glycosuria during life, and even a 
 large quantity of sugar taken by the mouth did not produce it. An 
 explanation of these seemingly opposing results was offered by this 
 observer, who pointed out that chronic inflammation of the liver 
 and pancreas are very commonly found together, and that the 
 glycosuria seen in cases with cirrhosis of the liver is probably de- 
 pendent upon an associated pancreatic lesion. Chvostek, Hanse- 
 mann, Dieckhoff , Oser, Opie, and many others, have since confirmed 
 the frequency with which chronic inflammatory changes are met 
 with in the pancreas and liver at the same time. 
 
 The researches of Lefas are of particular interest, as he investi- 
 gated the alteration in the pa,ncreas accompanying different varieties 
 of cirrhosis. With atrophic (Laennec's) cirrhosis he found that the 
 weight of the pancreas was often increased, and that there was a 
 uniformly intralobular overgrowth of connective tissue, poor in 
 cells, penetrating the parenchyma and isolating groups of acini. 
 With hypertrophic biliary (Hanot's) cirrhosis the pancreas was 
 not enlarged, but the interlobular tissue was increased in amount 
 and density. The size and consistency of the pancreas did not, 
 however, appear to be related directly to the condition in the hver. 
 As a rule the liver was more seriously affected, but in some cases 
 the disease of the pancreas was in a more advanced stage. In 
 every case, however, the fibrous tissue in the pancreas was fully 
 formed and poor in nuclei, even when the newly formed fibrous 
 tissue in the liver was of a semi-adult type. It is therefore con- 
 cluded that the cirrhosis of the liver and of the pancreas are due 
 
244 GLYCOSURIA 
 
 to the same etiological factors, but that the condition of the pancreas 
 is independent of, and not secondary to, the lesion in the hver. 
 Hirschfeld considers that, as a rule, the liver and the pancreas 
 are simultaneously affected, but the hver affection generally retro- 
 gresses, although traces of the process can be discovered by the 
 pathologists. If the Hver is in any way deteriorated, as from the 
 action of alcohol or malaria, the process in the liver predominates, 
 and although cirrhosis may develop in the pancreas, yet no diabetes 
 results. He thinks that the portal vein is very rarely the route by 
 which the influences responsible for the changes in the Hver and 
 pancreas arrive. Hirschfe'd cites some cases described by 
 Tsuchiya in which cirrhosis in the liver was the direct result 
 of invasion of the gland through the portal vein by helminths, 
 the Schistosoma Japonica, where the pancreas was intact, and 
 accepts this, as confirming his view that when the infectious agent 
 arrives by the portal route the Hver alone is affected, but when it 
 comes by way of the blood both liver and pancreas suffer. 
 
 A striking illustration of the dependence of chronic lesions of 
 the pancreas and liver upon the same etiological factor is furnished 
 by the condition known as hsemochromatosis, described by v. 
 Recklinghausen in 1889. In this a reddish-yellow iron-containing 
 pigment is deposited in the epithelial cells of the liver, pancreas, 
 and other glands, while a yellow iron-free pigment is found in the 
 muscles of the gastro-intestinal tract, blood-vessels, &c. Asso- 
 ciated with the pigmentary deposit there are cirrhosis of the Hver, 
 chronic interstitial pancreatitis, and histological lesions of the 
 stomach, intestine, heart, spleen, kidneys, &c. 
 
 Closely related to hsemochromatosis is the condition described 
 by Hanot and Chauffard in 1882 as " Bronzed Diabetes.'' In 
 this there is a rapidly fatal form of diabetes with cirrhosis of the 
 Hver and pancreas, associated with pigmentation of the internal 
 organs and skin. The bronzing of the skin is usually general and 
 uniform, but is not accompanied by pigmentation of the buccal 
 mucous membrane, as in Addison's disease. Although found in 
 the majority of cases it is not a constant symptom, being absent 
 in some 15 to 16 per cent. The disease is almost exclusively met 
 with in males, and usually between the ages of thirty and sixty. 
 Alcohol is stated by some to play an important part in the etiology 
 of many cases. Hanot and Chauffard beHeved that the cHabetes 
 was the primary factor in the disease, the changes in the Hver and 
 pancreas being the result of alterations in the blood and the accom- 
 panying endarteritis. Marie, Acard, and Dutourier, Jeanselme, 
 and Ancshultz came to the conclusion that the tissue changes are 
 
PERSISTENT GLYCOSURIA 245 
 
 the consequence of the deposit of pigment in them, and that the 
 pigment arises from the dissolution of haemoglobin from some 
 unknown cause. According to this view the diabetes is a pheno- 
 menon secondary to changes in the x^ancreas. Opie is of opinion 
 that hsemochromatosis is a distinct morbid entity associated with 
 ■chronic interstitial inflammation, notably of the pancreas and Uver. 
 He considers that diabetes only ensues when the pancreatitis 
 has reached a certain grade of intensity, and is usually the 
 terminal event. He found that the pancreatic inflammation is 
 of the interacinar type, and that the islands of Langerhans are 
 implicated in the lesion, their destruction being, in his opinion, re- 
 sponsible for the glycosuria. Cecil agrees with Opie's conclusions, 
 and states that the diabetes associated with hsemochromatosis is 
 referable to pigmentation and destruction of the islands of Langer- 
 hans. Margin found that in a fatal case of bronzed diabetes that 
 he examined, some of the islands of Langerhans were preserved, 
 but that their cells were crowded with pigment. Potter and Milne, 
 from what they have found in the hterature, and from a study of a 
 case under their own care, came to the conclusion that cirrhosis of 
 the Uver is the primary condition. They consider that pancreatic 
 involvement with cUabetes is a sequence, or coincident event, to 
 this ; that the haemochromatosis, always present in slight degree in 
 hver cirrhosis, is in some cases very excessive and causes a general 
 pigmentation which eventually also involves the skin ; and finally, 
 that the whole process is not definite symptom-complex, but a 
 chain of circumstances which rarely seems to be compHcated. In 
 a review of the literature up to 1910 Bernoulh found forty-one 
 detailed reports of cases of so-called bronzed diabetes, and states 
 that in seven there was no tendency to glycosuria. 
 
 Gastro-intestinal Tract. — The stomach in diabetes may show 
 evidence of gastritis, and in a few cases atrophy of the mucosa 
 has been described. The most common morbid condition is hyper- 
 plasia of the gastric and intestinal mucous membrane. Hedon 
 states that animals, which have survived for some time after 
 extirpation of the pancreas, show a thickening of the walls of the 
 stomach and intestine, with MqDertrophy of the mucosa, and 
 according to Martinotti and Boccarcli there is an abnormal de- 
 velopment of the glands of Lieberkiihn. It has consequently been 
 suggested that the hypertrophy met with in human cUabetes 
 is secondary to cUsease of the x^ancreas. Others, however, are 
 inclined to regard the changes in the bowel as primary, and to 
 •consider that there is a glycosuria of intestinal origin. 
 
246 GLYCOSURIA 
 
 Reale and Renzi, in 1890, stated that after extirpating the 
 duodenum in dogs glycosuria results, the animal passing 15 grams 
 or so of sugar in its urine in the following twenty-four hours. Sub- 
 sequently this question was taken up by Pfliiger, who, as we have 
 seen, came to the conclusion that there exists in the wall of the 
 duodenum an " anti-diabetic centre " which controls the pancreas, 
 Herhtzka found that glycosuria could be produced in frogs by 
 injecting nicotine into the duodenum, owing, he thought, to a. 
 paralysis of the sympathetic nerves. Pfliiger's conclusions were, 
 however, denied by Ehrmann, Rosenbaum, Minkowski, and others, 
 and Lepine failed to induce glycosuria in dogs by injections of 
 nicotine. Eichler and Silbergleit, in consequence of the description 
 hy Zak of two cases of corrosive j)oisoning, one hy an alkah and 
 the other with a mineral acid in which glycosuria occurred, made 
 experiments upon dogs, and found that severe corrosion or scorch- 
 ing of the duodenum was followed by the appearance of sugar in 
 the urine. The same result followed similar treatment of other 
 portions of the intestine, however, and they came to the conclusion 
 that the glycosuria was due to an emptying of the glycogen reser- 
 voirs, such as occurs after other violent insults to the organism. 
 
 Another explanation of the association of glycosuria with gastro- 
 intestinal diseases is that an . extension of an infective process 
 to the pancreas takes place. My own experience suggests that in 
 some instances diabetes may result from such a cause. The 
 following case seems to be an example. 
 
 In May 1904 I was asked to examine the urine of a girl of twenty- 
 three, who was complaining of indefinite abdominal symptoms, with 
 loss of appetite and general ill-health. I found a positive "" jDancreatie 
 reaction," a considerable excess of indican, and a rather high pro- 
 portion of ethereal sulphates, and consequently diagnosed " catarrhal 
 pancreatitis, probably secondary to a gastro -intestinal catarrh." It 
 was stated, however, that there was no evidence clinically of disease 
 of the pancreas, and as the surgeon in charge of the case considered 
 that the symptoms were due to appendicitis, he operated for that 
 disease and found a subacutely inflamed appendix. In June 1906- 
 T was asked to see the patient, and foiind that she was much wasted 
 and was passing 5 per cent, of sugar in her urine. She still gave a well- 
 marked ■' pancreatic reaction," and I diagnosed '" pancreatic diabetes." 
 The patient was put on a restricted diet and treated for that condition,, 
 but the disease was now too far advanced to hope for any very great 
 improvement and she died a year later. 
 
 In this case the diabetes was apparently due to disease of the 
 pancreas secondary to an intestinal catarrh. The pancreatitis 
 certainly existed for two years, and probably for more, before the 
 
PERSISTENT GLYCOSURIA 247 
 
 destruction of glandular tissue had advanced sufficiently far to 
 interfere seriously with carbohydrate metaboHsm. 
 
 The experiments of Woolley and Newburgh on the effects 
 produced by the injection of the lower derivatives of protein de- 
 composition on the supra-renals, offers yet another possible ex- 
 planation of the glycosuria that has been found in association with 
 intestinal disturbances, and would also account for the improvement 
 that follows measures taken to control intestinal putrefaction and 
 catarrh in some cases. It is tempting to think that this may have 
 been what happened in Richartz's case. 
 
 The patient was a man aged sixty, whose father and brother had 
 died of diabetes. He stated that he himself had excreted sugar eight 
 years previously when suffering from diarrhoea. With a fresh attack 
 of diarrhoea, which lasted for some weeks, 1-4 per cent, of sugar was 
 found in the urine. The motions were of a Hght yellow coloLir and foul 
 smelling. Under treatment with milk diet, peptone, castor-oil, and 
 tanalbin the stools became less frequent, improved in character, and 
 vrltimately became normal. At the same time the sugar diminished, 
 and completely disappeared in about a fortnight. Afterward it was 
 found that starchy foods did not give rise to glycosiiria, but that 
 30 grams of dextrose still caused sugar to appear in the urine. 
 
 Hiirter has pubhshed a remarkable case of glycosuria in a 
 child, associated with gastro-intestinal s}T3iptoms. 
 
 The patient was a girl of ten, whose parents were healthy, but 
 some of whose relations had suffered from diabetes. In July 1908 
 she had an attack of intestinal catarrh, accompanied by urticaria, 
 which had also affected two other members of her family at the same 
 time. There were several recurrences of the symptoms subsequently. 
 Fovu* weeks before the child came under observation she had suffered 
 from vomiting, after a slight indiscretion in diet, and from that day 
 there had been excessive hunger and thirst. She had lost flesh, her 
 urine was often abundant, and sugar was detected in the third week 
 of the illness. Physical examination revealed nothing abnormal in 
 the abdomen or nervous system, but the urine contained 9 per cent, of 
 sugar, and gave a slight reaction with perchloride of iron. As the 
 sudden onset suggested an affection of the pancreas, Schmidt's test 
 diet was given and his " silk-bag " test applied, but no evidence of 
 pancreatic disease was obtained. The patient was strictly dieted, and 
 the output of sugar fell from 210 to 5 grams in the twenty-foiu- hours. 
 In about a week, and after several days of vegetable diet, the glycos- 
 uria disappeared entirely. On a single occasion some ten days later 
 a reaction for sugar was obtained. Seven months later it was found 
 that the patient had gained 6 kilos, and that all the symptoms had 
 disappeared. A test diet containing 420 grams of white bread caused 
 no glycosuria, and 50 grams of glucose was also without any effect. 
 
248 GLYCOSURIA 
 
 Since then the patient has resiimed an ordinary diet, and up to 1910 
 the urine had remained sugar-free. 
 
 This case is particularly interesting because the excretion of 
 sugar was large, yet a perfect recovery apparently took place ; more- 
 over, the patient was a child with whom the prognosis is usualty grave. 
 
 Cases of intestinal disease with glycosuria have also been re- 
 ported by other observers, including Schmidt, Funck, and Garrod. 
 The latter points out that some of the most convincing recoveries 
 from diabetes have been associated with gastro-intestinal disturb- 
 ances, and that this field of inquiry promises to yield fruitful 
 results, and holds out hope of therapeutic advances. 
 
 Kidneys. — Post-mortem examination shows that the kidneys 
 are frequently abnormal in those who have suffered from per- 
 sistent glycosuria during Ufe. Seegen states that at the Vienna 
 Pathological Institute from 1870 to 1892 disease of the kidneys was 
 found in forty out of ninety-two cases. The most frequent change 
 was congestion and enlargement, but small contracted kidneys 
 were not uncommon, occurring chiefly in gouty patients. The 
 most striking and characteristic change in the kidnej^s in diabetes 
 is the glycogenic degeneration of the epithehum of Henle's tubes, 
 described by Ehrhch and Frerichs. Microscopically the cells are 
 seen to be large and clear. On treating fresh sections with iodine 
 solution the protoplasm stains yellow, and is seen to contain deep 
 brown masses of glycogen. Preparations fixed and hardened in 
 the ordinary wa^y for embedding in paraffin, only show clear round 
 spaces, resembling those left by fat, from which the glycogen has 
 been dissolved out, but the margins, for some unknown reason, 
 tend to take on basic stains. A similar deposit of glycogen has 
 been noticed in depancreatised dogs. The presence of glycogen 
 in this abnormal situation is remarkable in view of its relative 
 deficiency in the more usual situations, and it has been suggested 
 that it is due to the absorption of sugar from the urine. The hyahne 
 degeneration described by Armanni is beheved to be a later stage 
 at which the glycogen has disappeared. Diffuse nephritis with 
 fatty degeneration is also met with in some cases. 
 
 The onset of inflammatory changes in the kidneys is often 
 associated with a diminution in the amount of sugar passed, and, 
 according to Stocvis, it may completely disapjoear in cases where 
 there is granular atrophy when the renal changes have reached a 
 certain stage. Even when albuminuria and other evidences of 
 severe kidney mischief are absent, vascular changes sufficient to 
 render the kidneys more impermeable to sugar may be set up by 
 
PERSISTENT GLYCOSURIA 249 
 
 intercurrent infectious diseases, according to Barrenscheen, thus 
 -accounting for the diminished gtycosuria frequently met with in 
 such cases. 
 
 Other Morbid Changes. — Advanced fatty degeneration of 
 the muscles is a characteristic feature of long-stancUng cases of 
 diabetes. The heart is usually affected, and the mj^ocardium is 
 ]Dale and soft. Rarely there is hypertrophy of the heart. The 
 ■spleen is usually small, pale, and soft, but may be enlarged and 
 congested. Croupous pneumonia and broncho-pneumonia, chronic 
 interstitial pneumonia, and tuberculosis, terminating in gangrene, 
 are often found. The lungs may soften (malacia) and mix Avith the 
 gastric secretions after death, forming the so-called " pneumo- 
 malacia acida." 
 
 Theories of Diabetes. — Theoretically glycosuria may be 
 supposed to occur under three conceivable conditions — (1) if the 
 kidneys become abnormally permeable to sugar and allow its 
 escape into the urine ; (2) if the sugar in the blood exists in an 
 abnormally loose combination ; (3) if the blood from any cause is 
 unusually rich in sugar. 
 
 1. It was at one time supposed, and not unnaturally, that 
 •diabetes is referable to some disease of the kidneys, which permits 
 the escape of sugar from the blood into the urine, in much the same 
 way that lesions of these organs allow the escape of albumen, but 
 it is now universally agreed that this explanation is not true, and 
 that the morbid changes met with in fatal cases are secondary 
 ■effects and are not the cause of the glycosuria. 
 
 In 1896 Klemperer suggested that there probably exists a 
 variety of glycosuria, which is akin to that caused by the ad- 
 ministration of phloridzin, and depends upon a failure on the part 
 of the kidneys to retain the sugar normally present in the blood. 
 This view was accepted by Naunyn, Liithje, and a few other authors, 
 but most authorities denied that sugar might be present in the 
 urine as the result of such a renal insufficiency. '"Renal diabetes''' 
 is said to be characterised by a slight, but persistent, glycosuria, 
 on which variations in the diet have Uttle or no effect, and by a 
 diminution in the quantity of sugar in the blood. Cases of this 
 description have been carefully investigated and described by 
 Ronniger, Siebke, Weiland, Tachau, and Garrod, and there appears 
 to be httle doubt that, though rare, they are a distinct pathological 
 and cHnical entity depending upon a functional disorder of the 
 kidneys, possibly arising from the presence of some toxine in the 
 blood which acts in a similar way to phloridzin. 
 
250 GLYCOSURIA 
 
 GaiTod's case may be quoted as an example. 
 
 The patient was a lady between thirty and forty years of age. Her 
 urine was found to contain sugar in 1907 and 1908, the daily output 
 being about 1 gram, but at no time did she exhibit any symptoms 
 of diabetes. She was placed on a restricted diet, but did not bear it 
 well, and the excretion of sugar was, if anything, greater than when 
 she w^as taking ordinary food. In 1909 Garrod made a series of 
 analyses of her urine and found a total excretion of sugar for the twenty- 
 four hours which varied from .2-0 to 5-3 grams. Restrictions of diet 
 on the one hand, and the administration of 10 grams of glucose on 
 the other, did not obviously affect the glycosuria. Dextrose was 
 definitely proved to be present, but although a great discrepancy 
 between the optical activity and reducing power of the lorine was 
 observed, levulose could not be detected. Weiland noticed a similar 
 discrepancy in one of his cases. Unfortunately no estimations of the 
 sugar-content of the blood were made. Analysis of the urine two years 
 and four months later, the patient meanwhile having only avoided 
 sugar and an excess of starchy food, showed no aggravation of the 
 glycosLma. 
 
 In Bonniger's case an estimation of the sugar-content of the 
 blood was made. 
 
 The patient, a man aged thirty-seven, had for years excreted small 
 quantities of sugar, some 0-2 per cent., in urine with a specific gravity 
 of 1-020. The glycosuria was first detected in the course of an exa- 
 mination for life-insurance, and was proved to be due to dextrose. It 
 varied little in amoiint, and was not influenced by diet. The adminis- 
 tration of 100 grams of dextrose in no way affected it. The blood 
 serum was found to contain 0'097 per cent, of sugar, as compared, 
 with the normal of about 0"1 per cent, or rather less. 
 
 In Weiland' s three cases the average quantity of sugar in the 
 twenty-four hours urine was 7 to 10 grams in the first case, 3 grams 
 in the second, and 10 to 20 grams in the third. In none of them 
 was there any hyperglycsemia. The diagnosis in cases such as 
 these must lie between ordinary mild diabetes, and glycosuria the 
 result of abnormal permeability of the kidneys for sugar. But 
 even if the metabolic findings seem to indicate the latter it is well 
 to adopt an expectant attitude, for careful observation of the case 
 through years is the only means of finally deciding the question. 
 
 2. Lowei conceived the idea that the blood sugar is normally 
 in a loose combination with colloid, so that it cannot pass through 
 the glomeruK of the kidneys. If for any reason this combination 
 is not formed, is not broken up, or more sugar is poured into- 
 the blood than can be combined mth the available colloid, sugar 
 passes through the kidneys and appears in the urine. Stiles and 
 
PERSISTENT GLYCOSURIA 251 
 
 Lusk have suggested that the colloid sugar exists in two forms, 
 a-coUoid and /3-colloid dextrose, and that although both can be 
 utilised by the tissues, the latter is more readily attacked, thus 
 accounting for the difference in the dextrose to nitrogen ratio 
 observed in cases of various degrees of severity. These suggestions 
 are, however, highly speculative. 
 
 3. There can be no doubt that in the vast majority of cases, 
 of glycosuria the presence of dextrose in the urine is due to an 
 excess of sugar in the blood. The limit for normal individuals 
 is somewhat variable, but rarely exceeds one, or at most two, 
 parts per thousand ; in diabetes three or four parts per thousand 
 are common, and occasionally cases are met with in which as much 
 as seven or eight parts per thousand exist. Bernard, Pavy, and 
 most observers are agreed that dextrosuria is invariably the result 
 of hyperglycsemia, but Seegen and others, while admitting that 
 this explanation holds good in the majority of cases, believe that 
 in a few of the milder forms of glycosuria the blood sugar scarcely, 
 if at all, exceeds the normal limit. This difference of opinion 
 depends probably in part upon variations in the method employed 
 for determining the sugar, some of which give the glucose content 
 alone, while others also take into account the whole or part of the 
 colloidal s^igar. Difficulties also arise from the normal variations, 
 met with in different persons, and even in the same patient at 
 different times, and from the fact that only a slight increase of the 
 sugar-content of the blood is followed by its appearance in the 
 urine. Seegen states that glycosuria may occur when the quantity 
 of sugar in the blood is less than 0-2 per cent., an amount no greater 
 than is normally met with in many individuals, although it may 
 possibly be an excessive proportion for some. According to Pavy, 
 the intensity of the glycosuria is directly proportional to the hyper- 
 glycsemia ; but this is not the experience of Seegen, Naunjrn, v. 
 Noorden, and others, who have found that the percentage of sugar 
 in the blood may be only slightly above the normal, in spite of 
 there being marked glycosuria in some cases, while in others there 
 is only a small amount of sugar in the iirine, but a high percentage 
 in the blood. The former are probably due to the causes already 
 mentioned, while the latter may possibly be referred to changes in 
 the kidneys, consequent on the persistent glycosuria, which inter- 
 fere with the elimination of sugar. The fact that hj^oerglycsemia 
 is the usual cause of glycosuria may now be taken as firmly estab- 
 lished. 
 
 The question that next arises is, What is the cause of the 
 hj^perglycsemia ? Numerous explanations have been given, varying 
 
-252 GLYCOSURIA 
 
 according to the prevailing view as to the primary pathological 
 lesions present in diabetes, but most of these are now onl}' of 
 historical interest. After Claude Bernard propounded the doctrine 
 that the Hver is the great storehouse of carbohydrate for the 
 body, and that the sugar of the blood is derived from the glycogen 
 that it contains through the action of an amylolytic ferment, it 
 was concluded that the hjqDergiycsemia met with in diabetes was 
 dependent upon an abnormal output of sugar by the hver. When it 
 was discovered that puncture of the floor of the fourth ventricle 
 in the neighbourhood of the vaso-motor centre gives rise to giycos- 
 nria, it was suggested that an increased vascularity of the liver, 
 of vaso-motor origin, was the true explanation, and that all cases 
 of diabetes were of nervous origin. This vaso-hepatic theory was 
 for a time generally accepted, but it was severely criticised by 
 Oohnheim and others, who j)ointed out that there is no chnical, 
 or experimental, evidence of increased vascularity of the hver in 
 diabetes, and that in any case puncture of the floor of the fourth 
 ventricle could not bring about such an increased vascularity and 
 accelerated blood flow as its supporters assumed. Cohnheim was 
 inchned to take the view that diabetes depends upon the absence 
 of a ferment which in the normal condition initiates the further 
 destruction of dextrose. It was subsequently suggested that the 
 hver might be supphed with glyco-secretory nerves which are 
 independent of the vaso-motors. The absence of any satisfactory 
 evidence of cUsturbances of the nervous system in the majoritj^ of 
 cases of diabetes showed, however, that, at the most, only a small 
 proportion could be explained on some such hypothesis. It was 
 then suggested that the primary defect might he in the tissues, 
 which are, for some reason, unable to oxidise in the ordinary" way 
 the sugar brought to them, so that less is consumed than normal. 
 
 As the result of v. Mering and Minkowski's experimental work, 
 and in consequence of the discovery that j)athological changes in 
 the pancreas exist in many cases of diabetes, the theories that had 
 been previously held w^ere abandoned, or modified to fit in with the 
 view that a deficiency, or lack, of the internal section of that gland 
 Avas responsible for the excess of sugar in the blood. This may be 
 .supposed to arise from some influence which the secretion normally 
 exerts {a) upon the oxidation of sugar by the body cells, (b) upon the 
 sphtting of the sugar molecule and its preparation for oxidation 
 by the tissues, (c) upon the storage of glycogen. 
 
 Up to the present it has not been demonstrated that the general 
 oxidative powers of the tissues are diminished in diabetes, at least 
 in the earher stages. It is known that the products of protein 
 
PERSISTENT GLYCOSURIA 253 
 
 metabolism, urea, uric acid, &c., are approximately normal, and 
 that fats can be oxidised to carbon dioxide and water, that lac- 
 tates, inosite, mannite, and many other substances are also oxi- 
 dised, as in normal individuals, and that benzol is oxidised to phenol. 
 Moreover, in many cases of diabetes the levo-rotatory sugar, levulose, 
 is utihsed by the organism. On the other hand, we find that in 
 respiratory diseases in which there is cyanosis, sugar does not 
 appear in the urine as we should expect it to do if diabetes depended 
 upon diminished oxidation alone. In phosphorus poisoning, again,, 
 the oxidative powers of the body are distinctly reduced, yet 
 glycosuria does not occur. As we should expect when a con- 
 siderable amount of carbohydrate escapes oxidation, the quantity 
 of oxygen lost during respiration is lessened in cUabetes in com- 
 parison with a healthy person, but the recent experiments of 
 Forges and Salomon show that the respiratory quotient of a 
 depancreatised dog is not different from that of a normal animal. 
 
 The fact that the administration of camphor, chloral, and 
 other substances that are eliminated in the urine in combination 
 with glucuronic acid, is followed by almost as copious an excretion 
 of glucuronates in diabetes as in normal individuals, has been taken 
 by some observers to indicate that the first step in sugar oxidation 
 is not interfered \vith. This idea is based on the assumption that 
 glucuronic acid is formed in the preUminary stage of the degrada- 
 tion of dextrose by the tissues, as a comparison of their formulee 
 suggests : — 
 
 OH.C- (CH.OH)^- CHo.OH + 00 = OH.C- (CH.OH)^- COOH + HoO 
 dextrose gluci-ironic acid 
 
 It has also been noticed that glucuronic acid may be eUminated 
 in diabetes without the administration of these substances, and 
 that sometimes it may be found in the urine when, as a result 
 of careful dieting, the excretion of sugar has ceased. In confirma- 
 tion of this view are the recent chemical experiments of Jolles, 
 which indicate that the oxidation of dextrose involves the inter- 
 mediate formation of glucuronic acid, but on the other hand the 
 experiments of Mendel and Jackson suggest that glucuronic acid 
 is only produced in the intermediary metabolism of proteins, and 
 not of carbohydrates at all. 
 
 Baumgarten found that the diabetic organism can oxidise sugar 
 after a start has been made, for depancreatised dogs were able to 
 complete the degradation of partly oxidised carbohydrates, such 
 a d-gluconic acid, d-saccharic acid, mucic acid, glucuronic acid, 
 succinic acid, &c. He therefore concludes that the difficulty 
 
254 GLYCOSURIA 
 
 lies in the first attack on the sugar molecule, and suggests that this 
 is due to the absence of some ferment which normally initiates the 
 process. The possibihty that the pancreas supplies some such fer- 
 ment in its internal secretion has been warmly advocated by some 
 authors. Lepine believes that a giycolj^tic, or sugar- spUtting 
 ferment, which is supplied by the pancreas, exists, and that it is 
 ■diminished, or absent, in diabetes. The experimental facts on 
 which this theory is based have, however, been severely criticised, 
 and it is not generally accepted. More recently Cohnheim has 
 advocated the view that the pancreas supplies an amboceptor-like 
 .substance which acts as a link between the sugar in the blood and 
 the tissue cells. His conclusions have been provisionally accepted 
 hy some authors, but his experimental work and the inferences he 
 draws from it have been adversely criticised by others. Von Noorden 
 has suggested that the internal secretion of the pancreas is necessary 
 to enable the tissues to convert sugar into glycogen, which he 
 assumes is an essential step in the absorption of sugar into the 
 molecule of protoplasm. In support of this view is the poverty 
 of the liver and muscles in glycogen, both in experimental and 
 human diabetes. 
 
 Pavy considers that the excess of sugar in the blood may arise 
 from two causes — first, in the milder forms of diabetes, the sugar 
 absorbed from the intestine is not attached and built up as it 
 should be, owing to the absence of an amboceptor supplied by the 
 pancreas; and secondly, in severe, or what he terms " composite," 
 ■cases, there is in addition to this a molecular disruption, attended 
 with the dissociation and setting free of carbohydrate that has 
 previously been put into combination, brought about in a similar 
 way to that which occxu-s when blood impregnated with phloridzin 
 is circulated in the kidneys. This view appears to be in part 
 accepted by Bosanquet, who points out that the glycogenic theory, 
 as usually set forth, is inadequate to explain the most characteristic 
 feature of human diabetes, the stage at which there is a formation 
 of sugar from the cells of the bod3^ He suggests that there may 
 be in diabetes a poison, acting somewhat Hke phloridzin, which 
 has the poAver of splitting off a saccharine radicle from the proto- 
 plasmic molecule. Such a substance he contends will first attack 
 loose combinations of sugar, such as exist in the hepatic cells, 
 which probably hold their glycogen attached by some mechanism 
 analogous to Ehrlich's side chains, and readily allow it to be split off. 
 The poison would equally attack at an early stage a loose combina- 
 tion of sugar and protoplasm such as Pavy suggests (in the lym- 
 23hocytes) , as the vehicle for the carriage of sugar to the tissues. As, 
 
PERSISTENT GLYCOSURIA 255 
 
 however, the disease advances and more of the poison is formed, its 
 activities will not be confined to these loose compounds, but it will 
 attack the other cells of the body, breaking off from them too a 
 saccharine radicle. Now we know that the hepatic cells can give 
 up and resume their glj^cogen without injury to themselves ; such 
 is their function. But to extract a molecule of sugar from other 
 cells is probably impossible without destro^dng them. Thus it 
 comes about that in the later stages of diabetes there is a de- 
 struction of protoplasm, with formation of acetone bodies, which 
 are so characteristic of grave diabetes. 
 
 It has for some time been felt that even if it is conceded that 
 the pancreas possesses an internal secretion which acts in one or 
 other of the ways suggested, and it is allowed that the islands of 
 Langerhans are concerned in its production, the pancreatic, like 
 its predecessor, the nervous theory, has failed to explain all cases 
 of diabetes. Although its more enthusiastib advocates have sug- 
 gested that functional derangement of the gland may give rise to 
 the condition without there being any structural changes, most 
 observers are agreed that such an explanation is unsatisfactory. 
 Many cases of diabetes are undoubtedly of pancreatic origin, some 
 are due to nervous influences, but a certain proportion appear to 
 be dependent upon other causes, among which are probablj^ to be 
 counted disease of the other ductless glands, while the part that 
 toxic influences take in the production of glycosuria have also to 
 be accounted for. 
 
 There is a tendency at the present time to revert to the doctrine 
 that over-production of sugar is the main cause of diabetes, and 
 to abandon the theory that diminished consumption is the essen- 
 tial element. This view is now held by a number of eminent 
 observers, including v. Noorden. They consider that the hyper- 
 glycsemia, and consequent glycosuria, are dependent upon an 
 excessive output of sugar by the liver, and that this may arise from 
 abnormal stimulation, or impaired inhibition of its glycogenetic 
 function. The stimulus to the liver may come from (1) an excess 
 of carbohydrate food, as in alimentary glycosuria ; (2) an increased 
 call by the tissues ; (3) hyperfunction of the supra-renals, &c. ; or 
 (4) from the central nervous system, through the medium of the 
 chromaffin system. Impaired inhibition may arise from (a) inter- 
 ference with, or suppression of, the functions of the pancreas, as in 
 pancreatic diabetes, or (b) from interference with the controlhng 
 action of the thyroid, or hypophj^sis, onthatorgan, as in exophthahnic 
 goitre and acromegaly. They also consider that in some cases there 
 is probably a primary anomal}^ of the hver cells themselves. 
 
256 GLYCOSURIA 
 
 According to this view the glycosuria that results from puncture 
 of the floor of the fourth ventricle, and that occurs in association 
 with chsease of the nervous system, is not due to direct stimulation 
 of the hver through special glyco-secretory nerves, &c., as was 
 formerly taught, but depends upon the transmission of impulses 
 by the left sympathetic to the left supra-renal, whence it is trans- 
 mitted to the right supra-renal by the connecting nerves ; as a, 
 result of the stimulation of these organs they function more actively, 
 and the increased flow of their secretion in its turn brings about 
 an excessive output of sugar by the hver. Cannon and his asso- 
 ciates have sho"WTi that fright, anger, asphyxia, and the strong 
 stimulation of sensory nerves all cause an increased secretion o£ 
 epinephrin, and consider that the glycosuria that results from worry 
 and mental strain may be contributed to by the same cause. 
 Von Noorden suggests that a number of toxic influences may act 
 in a similar way, the glycosuria to which they give rise being partly 
 the result of an action thej^ exert on the diabetic centre in the 
 medulla, and partly an effect of their stimulating action on the 
 supra-renals, or the sjmipathetic nerves controlHng them, thus in 
 any case bringing about h^-perfunction of the chromaffin system,. 
 with a consequent over-production of sugar by the liver. 
 
 This conception of glycosuria correlates in a much more satisfac- 
 tory maimer than has previously been possible the experimental 
 facts, and all that is known of the etiology of diabetes. At the 
 same time it destroj^s the idea of a sharply defined cUsease, and 
 substitutes a system complex, liable to arise from a variety of 
 causes. According to this view the normal metaboHc level is 
 maintained by the balance existing between a number of mutually 
 controlKng forces. Should the balance be disturbed a lowered 
 tolerance for dextrose, a temporary, or intermittent, glycosuria, a 
 persistent glycosuria, or a tj'pical diabetes may result, according to 
 the nature, extent, and permanencj^ of the disturbance. If this be 
 so, it follows that there is no such thing as a non-diabetic glycos- 
 uria, with possibly the exception of the so-called renal diabetes,. 
 and that any difference is merely of degree, and not of kind. 
 
 Magnus Levy in his Cartwright Lectures maintained on the con- 
 trary that the conceiDtion of diabetes as a concrete unity is justi- 
 fiable, since the metabohc disturbance and its intensity dominate 
 the pathological process. He regards a primary disturbance of 
 the sugar- sphtting process as the essential factor in severe cases, 
 and considers that complete acceptance of the view that chabetes 
 arises only from an increased formation and mobihsation of sugar 
 must lead to the conclusion that proteins and fats are normally 
 
PERSISTENT GLYCOSURIA 257 
 
 transformed into sugar. There is, however, no justification for 
 such a conclusion. He grants that increased mobihsation co- 
 operates in the production of cUabetes, and also that, in a certain 
 sense, there is an increased formation of new sugar, but maintains 
 that this increase in sugar formation and mobihsation is only a 
 secondary consequence of disturbed combustion. He points out 
 that in spite of the fact that sugar is wasted in the body, those 
 organs which need it do not cease in their demands to be supplied. 
 The liver and other producers of new sugar strive to meet their 
 requirements, forming it, and sending it to other tissues, but 
 without any advantage to the organism. This disturbance in the 
 utihsation of sugar increases in severe cases with the lapse of time, 
 but it need never be absolute, some of the newly formed sugar being 
 catabolised. According to this theory the muscles, as the principal 
 organs concerned in the combustion of sugar, would come more 
 into the foreground of the picture, the mobilising organs, such as 
 the supra-renal capsules and the thyroid, taking a secondary place. 
 Levy holds that a form of diabetes exists, in which the muscle 
 system alone is involved, and that, even if the attempts that are 
 being made to isolate the hormone of the pancreas, and to prove 
 the co-operation of the pancreas and muscles in carboh3'drate 
 metabolism are successful, the solution of the problem Hes in the 
 direction he indicates. 
 
 Pathology of the Symptoms and Complications of Persistent 
 Dextrosuria. — Whatever may be the explanation of the imperfect 
 utilisation of sugar by the organism in chronic glycosuria, there can 
 be no doubt that most, if not all, the symptoms can be referred 
 to interference with the normal source of energy that the loss of 
 sugar in the urine entails, to the high sugar-content of the blood, 
 or to the presence of products of imperfect metabolism in the blood 
 and tissues. The manner in which the products of imperfect and 
 abnormal metabolism bring about the sjonptoms of acid intoxica- 
 tion have already been considered in connection with acidosis and 
 diabetic coma, it now only remains, therefore, to briefly refer to the 
 way in which it is believed that others of the main S3nQiptoms and 
 complications are produced. 
 
 The carbohydrates of the food are the most imj)ortant source 
 of energy for the body, and it consequently follows that if a patient 
 on an ordinary cUet is unable to make use of a part or the whole 
 of the carbohydrate he consumes, but passes it in his urine in the 
 form of dextrose, there is a corresponding loss of energy-forming 
 material which normally would be used for the production of power 
 and heat. Each gram of dextrose lost to the organism in this way 
 
258 GLYCOSURIA 
 
 represents about 4 Calories, so that if 100 grams are excreted daily 
 there will be a loss of 400 Calories, or sufficient heat to raise 4000 
 grams of water through 100° C.^ A man of average size doing 
 light work requires from 2500 to 3000 Calories a day, so that such a 
 loss is equivalent to nearty one-sixth of the total caloric expenditure 
 of the body, and is inconsistent with the maintenance of a good 
 state of nutrition. So long as the patient is able to digest and 
 absorb enough food to compensate for what he loses in his urine 
 his weight and powers of work may not seriously suffer, but even- 
 tually a time arrives when this is no longer possible, and the tissue 
 fats and proteins are used to supply the energy required, with the 
 result that there is a gradual loss of weight and increasing iveakness. 
 Eventually, if the patient is not meanwhile carried off by some 
 intercurrent disease or complication, he dies from what is practicalh^ 
 acute starvation, for all the available fat and proteins having been 
 exhausted sufficient energy to maintain the vital functions is no 
 longer obtainable. 
 
 The defective power of metaboHsing carbohydrates is no doubt 
 the explanation of the voracious appetite of many diabetics. In 
 some instances enormous quantities of food are consumed, and 
 yet the patient is hungry. In a case quoted by Quintard a diabetic 
 of sixtj^-two took 4| kilograms (10 lb.) of solid food a day, and Kiilz 
 records a case in which between 5 and 6 kilograms (11 to 13 lb.) 
 were consumed daily. Although the food for a time appeases the 
 pangs of hunger in the stomach, the cravings of the tissues remain 
 unsatisfied, since the chief source of energy, the carbohydrate it 
 contains, cannot be made use of. It is only when an amount of 
 protein and fat sufficient to supplj^ the energy required is added 
 to the diet, and the unusable carbohydrates are eliminated, that 
 the polj^ohagia disappears. The excess of food taken by patients 
 whose diet has not been arranged on scientific fines is apt to cause 
 dilatation of the stomach and set up gastritis, which add further 
 complications to the condition. 
 
 Both the polyuria and thirst that are so frequently found as 
 sjrmptoms of diabetes are referable to the excess of sugar in the 
 blood and its excretion in the urine. Experiment has shown that 
 if the sugar-content of an animal's blood be increased there is a 
 prompt rise in the quantity of urine passed, probably because an 
 additional amount of fluid is needed by the epithefial cells of the 
 kidneys to enable them to separate the sugar and efiminate it from 
 the body. The blood thus tends to diminish in volume, but this 
 
 ^ A large Calorie is the amount of heat required to elevate the temperature 
 of 1000 grams of water 1° C. (See p. 284.) 
 
PERSISTENT GLYCOSURIA 259 
 
 tendency is counteracted by a withdrawal of water from the tissues. 
 In some obscure way this loss of water by the tissues affects the 
 nervous system, calhng forth the sensation of thirst. Another 
 factor that probably assists in the production of this sensation is the 
 presence of chronic gastritis from which so many diabetics suffer. 
 
 Some of the complications met with in diabetes depend upon 
 the fact that the saccharine urine forms an excellent medium for 
 the growth of micro-organisms, while others are due to the altered 
 •chemical and physical characters of the blood and tissues. The 
 qnuritis, dermatitis, and eczema that develop in the parts about the 
 urinar}^ and genital tract, especially when strict cleanhness is not 
 observed, arise from the irritation j)roduced by the growth of 
 fermentative organisms in the urine clinging to the parts. Occa- 
 sionally the urine undergoes similar fermentative changes within 
 the bladder, giving rise to pneumaturia, or, if the urine becomes 
 infected with cohform and other organs, to cystitis. 
 
 The excess of sugar in the blood seems to exercise a deleterious 
 influence upon the tissvies of the body, which is especially seen in 
 the failure of ivounds to heal, and in the tendency of sHght in- 
 juries to lead to extensive tissue necrosis and gangrene. That this 
 tendency to tissue disintegration and necrosis is probably de- 
 pendent upon the hyperglycsemia is suggested by the fact that any 
 measures which result in a reduction in the quantity of sugar in 
 the blood favourably influence the tissue changes, but since the 
 amount of tissue destruction does not appear to bear any direct 
 relation to the intensity of the hyperglycaemia, it is possible that 
 .some unknown toxic influence that accompanies the latter may also 
 be involved. The increased osmotic pressure of the blood from the 
 excess of sugar it contains possibly also plays some part in the 
 tissue changes, but with the exception of Pusey's experiments, 
 which suggest that an alteration in osmotic pressure is an important 
 factor in the production of cataract, no observations on this subject 
 appear to have been carried out. The tendency to gangrene is 
 contributed to, in some instances, by a defective blood supply 
 consequent on arterio-sclerotic changes in the blood-vessels. 
 
 The marked susceptibility of diabetic subjects to various in- 
 fections, and particularly infection by pyogenic organisms and 
 tubercle bacilli, is generally attributed to the abnormally high 
 sugar-content of the blood, which might favour the growth of 
 various bacteria, or possibty neutrahse the bactericidal and other 
 -anti-infectious powers of the blood. There is no experimental evi- 
 dence in favour of the view that increasing the amount of sugar 
 in the blood, within reasonable limits, produces either of these 
 
260 GLYCOSURIA 
 
 effects. Thus Handmann found that the addition of glucose tO' 
 the blood to the extent of 0-5 per cent, to 1 per cent, does not render 
 it a better culture medium for staphylococci than normal blood ; 
 further, that the addition of sugar to the blood does not diminish 
 its bactericidal power for staphylococci, nor does it reduce the 
 opsonic power of the serum with reference to this coccus. It may 
 be noted, however, that Sweet in pancreatectomised dogs found 
 that the serum in the later stages loses its bactericidal power, 
 especially ^vith reference to colon, typhoid, and dysentery bacilh, 
 and also to a large extent its haernolytic power. Da Costa and 
 Beardsley observed a distinct diminution in the opsonic power of 
 the serum in diabetics for streptococci, staphylococci, and tubercle 
 bacilh. Thej^ did not, however, determine whether this diminution 
 in opsonic power was due to a diminution in the thermostabile 
 opsonic element or in the thermolabile element, or complement ; 
 neither did they make repeated observations on the same patients 
 over long periods of time so as to determine what relations, if any, 
 exist between the changes in the opsonic power of the serum and 
 the occurrence of complications due to the usual diabetic infec- 
 tions. It has been suggested that the diminished alkalinity of the 
 blood in diabetes reduces its anti-infectious power, but this 
 possibihty has not been investigated with sufficient thoroughness 
 to develop any definite facts. It has also been urged that there- 
 is no change in the general anti-infectious powers of the blood 
 and other fluids in diabetics, but rather a local loss of resistance 
 concerning the exact nature of which no explanation at all has been 
 advanced. The latter view is based largely on the fact that diabetic 
 patients do not seem to be much more hable to general infections 
 and to miliary tuberculosis than other individuals. It would 
 seem, however, that the observation by Sweet that the serum of 
 pancreatectomised dogs loses its bactericidal power, and the ob- 
 servation by Da Costa and Beardsley of the diminution in opsonic 
 power of the serum of diabetics, point distinctly to general diminu- 
 tion in the anti-infectious powers in diabetes. 
 
 The inability of the tissues to satisfactorily metabohse carbo- 
 hj'^drates probablj^ leads to impaired nutrition of the cells, and this, 
 Avith the injury due to the accumulation of deleterious waste pro- 
 ducts, may predispose them to degenerative changes and the 
 attacks of micro-organisms, thus accounting for the tendency to 
 furunculosis and other forms of suppurative infection, notably 
 phlegmons and carbuncles, which may run on to gangrenous pro- 
 cesses. The comphcation of pulmonary tuberculosis with gangrene 
 of the lung is possibly expUcable on similar Hnes. 
 
PERSISTENT GLYCOSURIA 261 
 
 Diagnosis. — In the diagnosis of glycosuria three main errors 
 have to be guarded against — (1) the possibihty of not recognising 
 a small amount of sugar in the urine, owing to the masking effect 
 of creatinin, albumen, ammonia, &c., when reliance is placed on 
 Fehhng's and similar reduction tests alone ; (2) the reverse mistake, 
 wliich appears to be more common, namely, attributing a reduction 
 due to uric acid, &c., to the presence of sugar ; (3) mistaking a 
 reduction due to pentoses, lactose, levulose, and more particularly 
 to glucuronic acid compounds, for that of dextrose. The precau- 
 tions necessary to avoid these errors, and the methods by which 
 the commonly occurring reducing substances in the urine can be 
 differentiated, have already been fully dealt with ; but I think it 
 is necessary to again draw attention to them, as my experience 
 suggests that too much rehance is often placed upon the results of 
 a single test, and more especially on the findings obtained with 
 Fehling's solution. During five consecutive years I had sent to 
 me thirty-seven cases with a diagnosis of '" diabetes " in which 
 dextrose was absent from the urine. In three the sugar was found 
 to be dextro-rotatory 1-arabinose, apparently of ahmentary origin, 
 in one case inactive arabinose appeared to be present, but I could 
 not obtain sufficient material to thoroughly study the case, small 
 quantities of a sugar resembling maltose were met with in two cases, 
 in two women lactose was found, the reduction in eighteen appeared 
 to be due to glucuronic acid compounds, and in six was probably 
 dej)endent upon uric acid or similar reducing substances, for no 
 trace of sugar could be detected with the phenylhydrazin and other 
 tests ; in five cases I failed to obtain any reaction with Fehhng's 
 solution, &c., and could only refer the. results reported to me to 
 faulty technique on the part of the observer. 
 
 In the present state of our knowledge it is rarely possible to 
 determine with certainty the underljdng cause of a persistent 
 glycosuria in any particular case. A minute examination of the 
 family history, the previous health of the patient, his mode of 
 life and surroundings, and his present symptoms and jDhysical 
 signs will sometimes suggest a probable cause, but it is not often 
 that a definite conclusion can be come to. A diagnosis is most 
 readily arrived at when evidence of pancreatic disease is found. 
 The possibility of this may be suggested by the history of the case, 
 but it is only reached with certainty by a thorough qualitative 
 and quantitative examination of the fajces and urine. 
 
 When making an examination of the faeces in a case of suspected 
 pancreatic disease ni}- procedure is to first notice the naked-eye 
 characters, and then make a careful microscopical investigation 
 
262 GLYCOSURIA 
 
 of specimens taken from various parts of the sample. A chemical 
 analysis for trypsin, amylopsin, stercobilin, and occult blood is- 
 then undertaken, and finally the percentages of total, " saponified,"' 
 and " unsaponified " fats, and the proportion of inorganic ash are 
 quantitatively determined. It will here be only necessary to 
 refer to the method I employ for the estimation of fats, the other 
 tests being described in their appropriate jjlaces subsequently. 
 
 Rapid Estimation of the Fat-content of the Fceces {Cammidge). — 
 For this purpose I have adopted a procedure, wliich, while much, 
 more rapid than Soxhlet's process, gives results which are quite satis- 
 factory for clinical piu-poses. The method is an adai^tation of the 
 Schmidt-Stokes process of estimating fat in milk. It may be briefly 
 described as follows. Two clean, dry, Schmidt -Werner tubes, labelled 
 A and B, and provided with a 10 c.c. mark are taken. Into the lower 
 bulb of each is introduced an accurately weighed quantity of the 
 thoroughly dried and finely powdered f^ces ; I usually employ about 
 half a gram. The residue on the watch-glass used for weigliing, and 
 on the sides of the short-necked funnel with which the powder is: 
 introduced into the tube, is washed down with a fine jet from a wash 
 bottle, which for the A tube contains hydrochloric acid (1 in 3), and 
 for the B tube distilled water. The sides of the tube are also washed 
 down until the whole of the sample is collected into the lower bulb,, 
 and the 10 c.c. mark is reached. The A tube is then heated in bohing: 
 water for a quarter of an hoiu", occasionally rotating it so as to well 
 mix the contents. After it has cooled, both tubes are filled to the- 
 60 c.c. mark with ether, securely corked, and inverted thirty or forty 
 times, allowing the whole of the solid material to run tlirough the ether 
 each time. Each tube is then rotated between the hands, fixed ia 
 an upright position, and left undisturbed for half an hour or more,, 
 so that the whole of the solid residue may be brought into the lower 
 bulb. Considerable care is required in this part of the operation, or 
 a perfectly clear supernatent layer of ether, free from solid, may not 
 be secured. With a pipette exactly 20 c.c. of the ethereal extract is 
 drawn off from each tube and delivered into two COg-flasks of knowa 
 weight, the amount of ether remaining in each tube being noted. 
 The ether in the flasks is evaporated off, the residue dried by heating 
 on a water-bath, and the flasks again weighed. From the amount of 
 extract yielded by the 20 c.c. of ether, and the quantity of ether left 
 in the tubes, the total amount obtained from the weight of dry faeces 
 used may be calculated, and from tliis the percentage in the stool 
 detenuined. For convenience of reference I am in the habit of de- 
 scribing the yield from the A tube as " total fat," that from the B tube 
 as " unsaponified fat," and the difference between the two as " saponi- 
 fied fat," and this is the meaning I shall attach to these terms when 
 employing them subsequently. 
 
 The solid residue in the B tube can be used for detecting stercobilin. 
 For this purpose it is filtered off, extracted with acid alcohol, the acid 
 
PERSISTENT GLYCOSURIA 263 
 
 neutralised with ammonia, and an equal volume of 10 per cent, zinc 
 acetate in alcohol added. The precipitate which forms is removed by 
 filtration, and the clear filtrate examined against a black background 
 for the green fluorescence which indicates the presence of stercobOin. 
 The intensity of the colovir varies with the amoiint of pigment present, 
 so that by always using approximately the same quantities of faeces 
 and of the reagents, any marked variation from the normal is readily 
 detected. 
 
 Lancereaux, in 1877, claimed that diabetes associated with, 
 disease of the pancreas was always accompanied by marked 
 wasting (diajete mavjre), and was further characterised by the 
 brusqueness of its onset, the gravity of the symptoms, and the 
 rapid progress of the disease, while that in which there was no 
 marked loss of flesh {diabete gras), and which ran a longer and 
 more benign course, was due to some other cause. Although it is. 
 true that some cases of pancreatic diabetes do run a rapid course, 
 and are exceedingly grave from the first appearance of the 
 symptoms, this, in my experience, is rather the exception than the 
 rule. I have had the opportunity of investigating a number of 
 cases of diabetes of undoubted pancreatic origin in which there has- 
 been no marked loss of flesh, the onset had been insidious, and the 
 general health was good several years after the discovery of the 
 glycosuria. It is now generally acknowledged that, as a rule, the 
 character of the onset and symptoms afford no evidence on which 
 a reliable opinion can be based as to the origin of the disease. 
 
 FsBCes. — There appears to be a general impression that disease 
 of the pancreas is always accompanied by an alteration in the 
 appearance and physical characters of the faeces. This is far from 
 being the case, for apart from those cases where there is associated 
 bihary obstruction, the pale bulky motions, of a white or oily 
 appearance, described in text-books, are very uncommon, being 
 only seen in very advanced cases. When present the}'^ present 
 characters which, to the experienced eye, distinguish them from 
 those met with in other diseases. They are whiter, more ghstening, 
 and sometimes contain masses of yellow, oily, undigested fat. 
 Their smell, like that of rancid bacon, is very typical. Their lack 
 of colour is partly due to the excess of fat they contain, particularly 
 to the free fatty acid crystals, and partly to a reduction of the 
 normal colouring matter to a colourless derivative through the 
 action of anaerobic bacteria. 
 
 The reaction of the normal faeces is amphoteric, faintly alkahne, 
 or very faintly acid, but in pancreatic diseases it is frequently acid^ 
 sometimes very markedly so. With intestinal catarrhs and bihary 
 
264 GLYCOSURIA 
 
 obstruction the stools are usually alkaline, so that when a pan- 
 creatitis is due to intestinal or bihary trouble the acid reaction maj^ 
 be masked by the alkahnity arising from this cause. In some 
 cases of pancreatic disease the acid reaction is so marked that it 
 produces considerable irritation of the bowels, and I have met with 
 cases of pancreatic glycosuria in which this was so great that the 
 patient's hfe was only made bearable by washing out the rectum 
 several times a day with an alkahne solution. Confirmatory 
 evidence of the presence of an intestinal catarrh is given by an 
 excess of inorganic ash in the faeces, the excess being more marked 
 when the colon, as well as the small intestine, is involved. 
 
 Fats: — One of the most important and characteristic functions 
 of the pancreas is to prepare the fats of the food for absorption 
 by splitting them into fatty acids and glycerine ; the former com- 
 bine wdth bases in the intestine to form soaps, and these, in the 
 presence of bile, are absorbed by the epithehal cells of the in- 
 testinal wall. In a healthy individual, taking an average amount of 
 fat, from 20 to 25 j)er cent, of the dvy weight of the faeces consists 
 of unabsorbed fat, and of this about an equal amount — viz. between 
 8 and 12 per cent. — consists of " saponified " and " unsaponified " 
 fat. Any disease of the pancreas interfering with its digestive powers, 
 such as advanced cirrhosis or cancer, leads to an increase in the 
 proportion of unabsorbed fat, and the greater part of this is usually 
 found to be in the unsaponified form. On the other hand, with 
 diseases of the intestine or obstruction of the bile flow, where fat 
 absorption is interfered mth, the saponified fats are found to be in 
 excess. If there is both interference with the digestive functions 
 of the pancreas and bihary obstruction or intestinal disease, the 
 relation between the saponified and unsaponified fats will depend 
 uj)on the relative extent and intensity of the two conditions. In 
 interpreting the results of an analysis of the faeces for fats, one must 
 therefore take into account the other indications given by the 
 ehnical symptoms and by an analysis of the urine and faeces. In 
 the early stages of chronic pancreatitis, when as yet the disease is 
 of the purely catarrhal ty^Q, there is probably an increased flow of 
 pancreatic juice analogous to the sahvation seen in parotitis, hence 
 fat digestion is often more active than usual, so that a low reading 
 of total fat is obtained on analysing the faeces, and an excessive 
 proportion of the fat is found to be in the saponified form. When, 
 however, the inflammatory changes have persisted for some time, 
 and there is well-marked cirrhosis, analysis of the faeces may show 
 as much as 50, and rarely, even 80, per cent, of unabsorbed fat, and 
 
PERSISTENT GLYCOSURIA 265 
 
 as a rule the unsaponified are in excess of the saponified fats. 
 With cancer of the pancreas the total fat content of the faeces is 
 always high, averaging 71 per cent, for the cases I have examined. 
 
 Intestinal diseases interfering with absorption may, however, 
 also show 50 to 60 per cent, of imabsorbed fat in the stools. Other 
 conditions associated with an excess of fat in the faeces which must 
 be taken into account when investigating the functions of the 
 pancreas are disorders of the stomach, which prevent the breaking 
 down of the connective tissue septa binding the fat together, and 
 a diet containing an abnormal amount of fat, ^particularly when 
 this is of a kind that is digested with difficulty. The mere presence 
 or absence of an excess of fat in the faeces therefore, although 
 suggestive, is by no means conclusive evidence of the existence or 
 not of disease of the pancreas. In a considerable number of cases 
 of pancreatic diabetes no alteration in the fat content of the stools 
 can be discovered. The relation between the saponified and un- 
 saponified fats is a much more important diagnostic point, but 
 even here the modifications produced by the causes I have men- 
 tioned must be constant^ borne in mind. Should an excess of 
 fat, and particularly an abnormal proportion of unsaponified fat 
 when the patient is on an average diet, be reduced by the ad- 
 ministration of fresh pancreas, or an active preparation of the 
 gland, by the mouth, the presence of pancreatic disease is rendered 
 more probable. 
 
 The digestion of proteins is another important function of the 
 pancreas, and the appearance of numerous undigested muscle 
 .fibres in the faeces tends to indicate serious pancreatic mischief, 
 but it is not justifiable to conclude that the functions of the pancreas 
 are interfered with from this alone, for, excluding their presence 
 from an excess of meat in the food, undigested muscle may also be 
 found in cases where, owing to increased peristalsis or putrefactive 
 changes, leading to secondary diarrhoea, the food is hurried through 
 the intestine before it has had time to be digested. Defective 
 gastric secretion may also lead to imperfect digestion of muscle, 
 as the connective tissue binding the fibres is not dissolved. 
 
 According to Schmidt cell-nuclei are not attacked in the stomach 
 but are digested by the pancreatic secretion, hence the discovery 
 of well-preserved cell nuclei in the muscle fibres in the faeces in- 
 dicates pancreatic insufficiency. For the purpose of his test small 
 cubes of meat, enclosed in gauze bags, and previously hardened in 
 alcohol, are swallowed by the patient and subsequently recovered 
 from the faeces. The contents are washed, stained, and examined 
 microscopically. 
 
266 GLYCOSURIA 
 
 Miiller has shown that the normal fseces contain a tryptie 
 ferment which can be detected by allowing an extract of the stool 
 to act upon a serum plate at body temperature, and Schlecht has- 
 made use of this test as an indication of the functional activity of 
 the pancreas. Subsequently Gross and Heiberg substituted an 
 alkahne solution of casein for the serum plate, and test the amount 
 of digestion that has taken place by precipitating the casein that 
 remains unaffected with dilute acetic acid. These tests are of con- 
 siderable value in showing serious interference with the digestive 
 functions of the pancreas, such as is met with in advanced cirrhosis- 
 of the gland and in cancer associated with an estabhshed glycos- 
 uria ; but they do not distinguish the early pre- glycosuria stages 
 of chronic pancreatitis, which are most amenable to treatment. 
 Further, the value of the test may be interfered with by the partial 
 digestion of the casein by erepsin. 
 
 In a paper read at the International Congress of Physiologists- 
 held at Brussels, Boldireff stated that the introduction of a large 
 amount of oil into the stomach causes relaxation of the pylorus- 
 and the reflux of bile and pancreatic juice. Taking advantage of 
 this observation, Volhard suggested a chnical method of testing, 
 the functional activity of the pancreas. By means of an oesophageal 
 tube 150 to 200 c.c. of ohve-oil are introduced into the stomach,, 
 and withdrawn again in a quarter to three-quarters of an hour.. 
 The recovered liquid is then tested for the presence of trypsin, as 
 in Gross' test. It has been found that when the stomach contents 
 are unusually acid the findings are positive only when the acidity 
 has been neutrahsed, so that it is advisable to administer half a 
 teaspoonful of burnt magnesia, or some other alkali, before giving 
 the oil. Volhard obtained satisfactory results with normal indi- 
 viduals, and with a few cases of advanced chsease of the pancreas 
 that he examined. Lewinski, who investigated twenty-nine cases, 
 reports that the absence of trypsin is an important sign of pan- 
 creatic insufiiciency or of a mechanical obstruction to the passage 
 of pancreatic juice into the stomach. Koziczkowsky carried out 
 the test with eight 3' patients suffering from a variety of diseases, 
 using 150 c.c. of oil or 250 c.c. of cream, and obtained a more or 
 less marked reaction for trypsin with seventy-two. In eight no 
 trace could be discovered, and these were found to be suffering 
 from inoperable cancer of the stomach, pernicious anaemia, severe 
 liver affections, or diabetes. The smallest jaroportions of trypsin 
 were met with in patients with catarrh of the small intestine, 
 cholehthiasis, and diabetes insipidus. He concludes that a series 
 of tests is advisable, that the constant absence of trypsin points 
 
PERSISTENT GLYCOSURIA 267 
 
 to severe functional disturbances of the pancreas, and that the 
 presence of large amounts exclude advanced pancreatic mischief. 
 
 Ehrmann has recently suggested a modification of this test. 
 It is based on the fact that a neutral fat is not spUt by anything 
 but the fat- splitting ferment of the pancreas, so that by adding 
 a stain which only acts on the fatty acids to the recovered 
 stomach contents their presence is rendered evident. 
 
 The patient is given a test breakfast of 30 grams of rice starch- 
 dissolved and warmed in a glass of water, a trace of salt is added, and 
 75 grams of commercial palmin, liquefied by heat, stirred in. After 
 two or two and a half hours the stomach contents are siphoned out, 
 and a portion mixed with an equal part of a mixture of 90 parts of 
 petroleum, benzin, and benzol to 100 parts. After the mixture has 
 been well shaken, the supernatent ether layer is decanted and mixed 
 with an equal part of a 3 per cent, solution of copper acetate in water. 
 The ethereal layer then assumes a bright green tint in proportion to 
 the content of fatty acid. 
 
 Starch. — Reduction or failure of the pancreatic secretion might 
 be expected to lead to impaired digestion of starchy foods and 
 the appearance of an excess in the fasces. Observations made by 
 various observers have shown, however, that only a small pro- 
 portion, or none at all, of the starch of the food is excreted in the 
 faeces in an unchanged condition in cases where these conchtions 
 exist. The discovery of unaltered starch granules in the stools 
 is more commonly an evidence of a catarrhal condition of the 
 upper part of the intestine, causing the food to be hurried along at 
 an abnormally rapid rate, than of pancreatic disease. 
 
 Attempts have been made from time to time to estimate the 
 functional activity of the pancreas hy determining the amount of 
 diastase in the faeces and urine, but as a rule with no very satis- 
 factory result. For this purpose the modification of Robert and 
 Strassburger's method devised by Goiffon and Tallarico may be 
 followed. 
 
 A 1 per cent, solution of starch is mixed with an equal part of 
 a 10 per cent, solution of the stool, neutralised, and filtered. The 
 mixture is kept at a constant temperature of 38° C, and at regular 
 intervals a drop is brought in contact with a drop of iodine solution. 
 When it ceases to stain blue the digestion of the starch has been 
 complete. The stool shoiold be fresh, and there should not be the 
 slightest admixtiire of urine. It is often enough merely to mix the 
 stool and starch solutions in a test-tube, heat in hot water, and apply 
 the iodin test. If there is abundant amylase present, the starch will 
 be digested in five minutes. 
 
268 GLYCOSURIA 
 
 When greater precision is desired, it is better to use a 5 per cent, 
 sokition of the faeces, and dihite further with an acid solution made 
 with 10 parts normal solution of hydrochloric acid, 5 parts sodium 
 chloride, and 1000 parts water. About 5 c.c. of the stool are ground 
 in a mortar with 50 c.c. of water. The mixture is poured into a test- 
 tube, and all above the 50 c.c. mark represents the volmne of faeces. 
 Water is then added at the rate of 20 c.c. for each cubic centimetre of 
 faeces. In another test-tube are placed 2 c.c. of the 1 per cent, starch 
 solution, 5 c.c. of the above acid solution, and the whole is heated in 
 water toward 40° C. Then, noting the time, 2 c.c. of the solution of 
 faeces are added, and at regular intervals a drop is added to a drop of 
 the official compound solution of iodin until all bluish or reddish dis- 
 coloration ceases. For extreme precision, the index of the findings 
 can be the time required for digestion of the starch, divided by the 
 dry weight per hundred of the stools. 
 
 The discovery of Roger and Simon that saliva, inhibited in its 
 activity by gastric juice, recovers its digestive powers when trans- 
 ferred to an alkaline medium and a little unmodified saliva or 
 pancreatic juice is added, has been applied by Fedeli and 
 Romanelli to the testing of the functions of the pancreas. 
 
 To 1 c.c. of the patient's saliva they add 5 c.c. of gastric juice, or 
 an equal qiTantity of 2*5 per cent. HCl; shake the mixture, and leave 
 it to rest for half an hour; then add 4 c.c. of a 1 per cent, solution 
 of carbonate of soda, so as to render the mixture slightly alkaline. 
 They next add 20 c.c. of a 10 per cent, starch paste, and place the 
 whole in thermostat at 37° C. for two hoiirs, repeatedly shaking. The 
 amount of sugar formed is then estimated. The next stage consists 
 in adding to the above mixture 10 c.c. of an aqueous solution (1 in 4) 
 of faeces, and leaving the whole in the incubator for twelve hours, and 
 then estuTiating the sugar formed. The difference between the two 
 estimations of sugar represents the degree of pancreatic functionality. 
 To show that this was not due to other constituents of the faeces, the 
 authors tested with bile and saccus entericus, and found that they both 
 gave negative results. 
 
 Even when an examination of the faeces points to the existence 
 of pancreatic insufficiency it does not necessarily follow that disease 
 of the pancreas exists and is the cause of the glycosuria, for a 
 healthy gland may fail to function because it has not been stimu- 
 lated into activity in a normal manner. Modern research has 
 shown that the processes of digestion are intimately linked together 
 and form a continuous chain, each step in the series calHng forth 
 that which succeeds it, so that anj^ inefficiency or break in the 
 chain disorganises the whole process. For the normal perform- 
 ance of the digestive functions of the pancreas it is necessary (1) 
 that the stomach should supply a stimulus in the shape of an acid 
 
PERSISTENT GLYCOSURIA 269 
 
 chyme which will act upon the duodenal mucous membrane, giving 
 rise to the " secretin " that will rouse the pancreas into activity ; 
 (2) that the pancreas itself should be able to respond and secrete 
 the necessary ferments ; (3) that bile should reach the intestine to 
 aid in digestion and absorption ; and (4) that the intestine should 
 secrete the " enterokinase " by which the inactive proteolytic 
 ferment, trypsogen, of the pancreatic juice can be converted into 
 active trj^sin. Failure of any one of these may lead to symptoms 
 of pancreatic insufficiency. 
 
 1 . We will first consider pancreatic insufficiency dependent upon 
 gastric troubles. If the secretion of hj^drochloric acid by the 
 stomach is deficient, or absent, the pancreas will act imperfectly, 
 or not at all, for although fats and a few other substances appear 
 to have the power of forming secretin from the intestinal mucous 
 membrane, their activity in this respect is much less than that 
 possessed by hydrochloric acid. In such cases, too, the pylorus 
 opens at an abnormally early stage in digestion, so that the in- 
 testine is faced with the problem of dealing with materials in which 
 there is not only a deficiency or absence of pancreatic ferments, 
 but which have also been imperfectly prepared for the action of 
 any ferment that may be there, with the result that the symp- 
 toms of " intestinal " indigestion ensue, and an examination of 
 the fffices shows an abnormal amount of undigested food material. 
 The absence of hydrochloric acid also tends to favour the growth of 
 an abnormal intestinal flora. If, on the other hand, an excess of 
 acid is poured out by the stomach, the pancreas may for a time 
 be able to cope with it by secreting a corresponding amount of 
 its alkaline juice, but the over-stimulation of the gland leads to 
 degenerative changes, which are indicated by a marked pancreatic 
 reaction in the urine, and eventually these bring about a diminished 
 secretion. If the hyperchlorhydria continues it will cause an ab- 
 normally acid condition of the intestinal contents, and so interfere 
 with the activity of any pancreatic ferments that may be present, 
 for these are quickly destroyed by free mineral acids, so that pan- 
 creatic insufficiency is again brought about but in an altogether 
 different way. In either case the only certain way to diagnose the 
 cause of the condition is by the administration of a test meal. 
 
 2. The part the bile takes as an adjunct in the cUgestion and 
 absorption of fats is well known, but it also appears to exert a very 
 material influence on the digestion of j)roteids. It is therefore 
 very necessary for satisfactory digestion that bile should be present 
 in the intestine, and its absence is often a contributory factor in 
 the production of pancreatic insufficienc3\ 
 
270 GLYCOSURIA 
 
 3. Another imjDortant adjuvant to the digestive action of the 
 pancreatic Juice is enterokinase, a ferment present in the succus 
 •entericus, which has the power of augmenting the activity of the 
 pancreatic ferments, and more particularly the proteolytic, to a 
 .striking degree. This " ferment of ferments " is secreted by the 
 intestinal mucous membrane, chiefly in the duodenum, apparently 
 through the stimulus afforded by the pancreatic juice. In certain 
 diseases of the intestine it is probable that its formation is interfered 
 with, and there may consequently be defective pancreatic digestion, 
 not from true pancreatic insufficiency, but from a lack of the activat- 
 ing ferment. The diagnosis of such a condition is not easy, and 
 its presence can only be inferred when an analj^sis of the faeces 
 reveals imperfect digestion, particularly of proteids, and there is 
 no evidence of pancreatic disease or true pancreatic insufficiency. 
 
 4. The fourth variety of pancreatic insufficiency is the true 
 form in which, owing to lesions of pancreas or obstruction of the 
 ducts, there is imperfect digestion from diminution or absence of 
 the pancreatic ferments. This is seen in advanced cases of cirrhosis 
 •of the pancreas, in some cases of pancreatic calcuh and cysts, in 
 .cancer of the pancreas, particularly of the head of the gland, and 
 in occlusion of the ampulla of Vater by gall-stones, growths, or 
 stricture. 
 
 The presence of lecithin in the faeces is said to indicate disease 
 ■of the pancreas, but as traces are found in health, and the quantity 
 present is liable to be increased from other causes than interference 
 with the functions of the pancreas, it is not a sign of much value. 
 
 Sometimes chronic pancreatitis giving rise to glj^cosuria is the 
 result of invasion of the pancreas by growths or ulcers of the 
 .stomach or intestine, and the discover}^ of occult blood in the faeces 
 is suggestive of one or the other. If the blood is found in every 
 specimen on four or five successive daj^s it j)oints to a malignant 
 growth, whereas its intermittent presence is suggestive of a duo- 
 denal or gastric ulcer which may be invading the pancreas, or simply 
 be associated with a catarrhal condition of the upper part of the 
 intestine to which the pancreatitis is secondary. Occult blood is 
 also found very constantly in cases of cancer of the pancreas and 
 with growths of the common bile-duct or gall bladder, which may 
 invade the pancreas and cause glycosuria. Occasionally it is met 
 with in advanced pancreatitis without actual alteration of the 
 intestinal mucous membrane, probably as a result of the haemorr- 
 Jiagic tendency. 
 
 The Urine. — As complete an analysis as possible of the mrine 
 
PERSISTENT GLYCOSURIA 271 
 
 should of course be made in all cases of glycosuiia, but this is 
 particularly important Avhen it is suspected that the presence of 
 sugar is dependent upon disease of the pancreas. The changes 
 met with in the urine in association with pancreatic disorders 
 depend partly upon the altered conditions existing in the intestinal 
 tract, and partly upon the perverted chemistry of the body. The 
 ohief alterations that have been observed are in the amounts of 
 indican, ethereal sulphates, total nitrogen, uric acid, j)hosphates, 
 calcium oxalate, urobilin, and bile, and according to some observers 
 the occasional presence of maltose and pentoses. Sahh's test and 
 the so-called " pancreatic " reaction may also be of some assistance 
 in arriving at a diagnosis. 
 
 Sulphates and Indican. — In the normal condition about a tenth 
 of the total sulphates of the urine occur in combination with 
 aromatic alcohols (indoxyl, skatoxyl, cresol, phenol, &c.), and are 
 consequently spoken of as ethereal, or conjugate, sulphates. While 
 their amount is subject to great and inexphcable variations, the 
 quantity present in any case may be considered as a fairly accurate 
 index of the extent to which absorption of the products of intestinal 
 decomposition that can pair with sulphuric acid is taking place. 
 It has been stated that when the flow of pancreatic juice into the 
 intestine diminishes, or ceases, the proportion of ethereal sulphates 
 falls, probably because less inclol, skatol, &c., are formed in the 
 intestine. Several observers have, however, found an increase in 
 both the conjugate sulphates, and in the indican excretion in such 
 cases. M}^ own observations during the past ten years have given 
 such varied results that I have come to the conclusion that no reli- 
 ance can be placed upon estimations of the ethereal sulphate ex- 
 cretion or indican output, in the diagnosis of pancreatic disease, 
 and that an excess of both indicates rather an associated enteritis 
 and hepatic insufficiencJ^ 
 
 Azoturia and Endogenous Uric Acid. — An excess of nitrogen is 
 said to occur in the urine in diseases of the pancreas, but as this 
 also exists in all forms of diabetes, it is of no assistance in the 
 differential diagnosis. Rosenberger states that in pancreatic 
 diabetes there is a diminished excretion of uric acid when the 
 patient is placed on a purin-free diet, and suggests that this may 
 be of use in the diagnosis of the condition. 
 
 Phosphates. — According to Dominicus, an increase in the ex- 
 cretion of phosphoric acid is characteristic of pancreatic lesions, 
 but since the chief source of the phosphoric acid in the urine is the 
 food, the nature of this will largely control the output. It is 
 therefore advisable that, if a diagnosis is attempted by this means, 
 
272 GLYCOSURIA 
 
 the patient should be placed upon a milk diet and the excretion 
 of phosphates be compared before and after the administration of 
 an active pancreatic extract. 
 
 Calcium Oxalate. — From an early stage of my work on diseases 
 of the pancreas I was struck by the frequency with which deposits 
 of calcium oxalate crystals were found in the urine in chronic 
 pancreatitis. Subsequent observation showed that this was due 
 to an increased excretion, and did not depend merely on an altera- 
 tion in the physical characters of the urine that favoured the de- 
 position of the crystals. It has long been known that glycosuria 
 sometimes follows continued oxaluria, and that, in some cases of 
 diabetes, a diitiinution in the output of sugar is associated with an 
 increase in the oxalate deposit. Helen Baldwin found that in dogs 
 the excretion of oxalates is increased by the administration of sugar, 
 and Ssobolew states that overfeeding animals with carbohydrates 
 gives rise to changes in the islands of Langerhans, so that it is not 
 improbable that oxaluria may in some way be due to a disturbance 
 of metabohsm arising from pathological changes in the pancreas. 
 
 Bile and Urobilin. — The presence of bile in the urine shows that 
 there is some obstruction to the free flow of bile into the intestine 
 and, if associated with glycosuria, points to the presence of gall- 
 stones in the pancreatic portion of the common bile duct, malignant 
 disease of the head of the pancreas, or a growth of the common duct, 
 ampulla of Vater, duodenum, &c., involving the pancreas. In my 
 experience urobilinuria is most frequently associated with infection, 
 of the bile ducts, and its presence points to floating gall-stones in. 
 the common bile duct, or to an ascending infection from the in- 
 testine, which may also involve the pancreatic ducts and pancreas.. 
 
 Maltose and 'pentoses will be considered under maltosuria, and 
 pentosuria, respectively, but it may be stated here that their 
 presence in the urine in pancreatic affections is so exceedingly 
 rare that it may be a mere coincidence, and is of no practical 
 diagnostic value. 
 
 A whole series of special tests have been devised with the object 
 of elucidating the state of the functions of the pancreas. 
 
 Sahlis Test. — This test of the functional activity of the 
 pancreas depends upon the fact that if iodoform, enclosed in a. 
 gelatine capsule hardened with formalin, is given by the mouth, 
 it passes through the stomach unchanged, but is dissolved by the 
 pancreatic secretion, so that iodine appears in the urine in from 
 four to eight hours after its administration. The presence of the 
 iodine is detected by adding a drop of nitric acid and shaking with 
 
PERSISTENT GLYCOSURIA 273 
 
 chloroform, when the characteristic violet colour is seen in the 
 chloroform extract. If the digestive functions of the pancreas are 
 impaired, or have ceased, the urinary reaction is delayed, or absent. 
 So many sources of error attach to this method, notably the diffi- 
 culty of properly adjusting the hardness of the capsules, and the 
 dependence of the test on the motor-efficiency of the stomach, that 
 it is not often used. 
 
 Other observers have employed pills containing potassium 
 iodide, sodium salicylate, or methylene blue, coated with keratine, 
 hardened gelatine, or wax. 
 
 The So-called " Pancreatic " Reaction in the Urine. — The " pan- 
 creatic " reaction was first described in my Arris and Gale Lecture 
 at the Royal College of Surgeons in 1904. In 1906 I published 
 an improved method by which some of the defects of the original 
 test were overcome, and the result was, to a certain extent, made 
 independent of the experience of the observer. ^ 
 
 I have now been using this method almost daily for six years, 
 and have examined nearly 3000 urines by means of it, but I 
 have had no reason to alter the opinion that I have repeatedly 
 expressed that, although not ^pathognomonic, the results of the test, 
 when considered in conjunction with the chnical symptoms and 
 an analysis of the faeces, are chnically useful. A positive reaction 
 in my experience is usually associated with some functional dis- 
 turbance of the pancreas ; generally the result of active degenerative 
 changes of an inflammatory character. Both chnical observations 
 and experiments on animals have shown that mahgnant disease 
 and cirrhosis of the pancreas are not, as a rule, accompanied by a 
 positive reaction, for the diagnosis of these conditions rehance 
 must be chiefly placed on an analysis of the faeces. The hterature 
 relating to the " pancreatic " reaction in the urine is now so ex- 
 tensive that to adequately review it would occupy a very consider- 
 able space, and I can here merely state that my observations and 
 conclusions have been borne out by a considerable number of 
 independent observers, and that, if the reactions and controls 
 are carefully carried out, the information obtained is of distinct 
 diagnostic value. 
 
 Taking advantage of the fact that the reaction is due to a 
 substance that is not readily fermented by yeast,^ a modification 
 of the original test was devised for apphcation to urines from cases of 
 diabetes with a view to discovering whether the gtycosuria was asso- 
 ciated with active degenerative changes in the pancreas or not. 
 
 ^ See The Pancreas : Its Surgery and Pathology (W. B. Saunders Co., 1907), p. 243-. 
 2 See Cammidge, Proc, Roy. Soc, 1909. 
 
 S 
 
274 GLYCOSURIA 
 
 In this modified method a specimen of the urine is boiled with 
 hydrochloric acid, neutralised, treated with tri-basic lead acetate, and 
 the excess of lead removed with sulphiu-etted hydrogen in the usual 
 way. It is subsequently warmed to free it from sulphuretted hydrogen, 
 mixed with yeast, and incubated to remove the fermentable sugar ; 
 but, as the " unfermentable " sugars are slowly broken down by the 
 action of organisms contained in the yeast, it is necessary that the 
 incubation should be stopped as soon as the last trace of fermentable 
 sugar has been removed. This point is determined by simultaneously 
 incubating a slightly larger quantity of a " control " specimen of 
 urine, that has been dealt with in exactly the same way, except that 
 it has not been boiled after the addition of the hydrochloric acid, and 
 testing it at frequent intervals for sugar, linmediately no reaction is 
 obtained the phenylhydrazin test is carried out with the specimen 
 that has been hydrolised. The essence of the test therefore consists 
 in fermenting a control only just so long as is necessary to remove 
 the last trace of fermentable sugar, and taking this as a guide to the 
 condition of the sample to be tested. 
 
 I have examined 296 specimens of urine from 168 cases of 
 diabetes by this method, and obtained a positive result, suggesting 
 that the diabetes was probably of pancreatic origin, in 121 (72 per 
 cent.) of the patients. In 47 no reaction was obtained. Of the 
 47 whose urine gave no reaction two were children, and when the 
 pancreas of one of these was examined post-mortem no abnormahty 
 could be discovered, either microscopically or macroscopically ; 
 the pancreas of the other was not examined after death. Two 
 others in this group were also examined post-mortem and the 
 pancreas was said to be normal. In three of the cases that gave 
 a positive reaction the glycosuria followed an attack of acute, or 
 subacute, pancreatitis from which the patient recovered. Twenty- 
 six were beheved to have had chronic pancreatitis, secondary in 
 eight to the presence of gall-stones in the common bile duct, to an 
 attack of typhoid fever in one, to chronic indigestion and duodenal 
 catarrh in eighteen, to duodenal ulcer in three, associated with the 
 presence of pancreatic calcuU in one, and with a cj^st of the pancreas 
 in two. In one there was transient glycosuria associated with 
 mumps, and one patient had had an accident involving the ujoper 
 abdomen. Arterio-sclerosis was found in ten, two were syphihtic, 
 and in seven there was a history of gout. In six there was primary 
 mahgnant disease of the pancreas, and in two extension of a 
 mahgnant growth from a neighbouring organ to the pancreas 
 (cancer of the duodenum in one, and cancer of the common bile 
 duct in the other). Eight of the cases were examined after death ; 
 two showed well-marked interacinar pancreatitis ; in five there was 
 
PERSISTENT GLYCOSURIA 
 
 2/0 
 
 advanced interlobular pancreatitis, and in one well-marked fibrosis 
 of the pancreas associated with calculi in the ducts. 
 
 At the best, however, the fermentation test is not altogether 
 satisfactory. It is verj'- laborious, requires constant attention, and 
 is open to many fallacies. I have for long felt that if the '' pan- 
 creatic " reaction could be made a quantitative one, and could 
 be carried out by a method which would not be influenced by the 
 presence of dextrose, &c., its findings would be of much greater 
 value. After a lengthy series of experiments I have eventually 
 evolved a procedure which appears to fulfil these requirements. 
 
 One hundred c.c. of the filtered urine are mixed with 5 e.c. of 
 hydrochloric acid (sp. gr. 1"16), and boiled on a sand-bath for ten 
 minutes. It is then cooled, made up to 100 c.c. with distilled water, 
 and the excess of acid neutralised with 16 grams of lead carbonate. 
 After standing for a few minutes it is carefully filtered, and the pre- 
 cipitate well washed with cold water, the washings being added to the 
 filtrate. The filtrate is then thoroughly shaken with 12 grams of tri- 
 basic lead acetate, filtered, and the precipitate well washed with water. 
 Several filtrations may be necessary at this stage. To the clear filtrate 
 is now added 1 c.c. of ammonia (sp. gr. 0-880), and the precipitate that 
 forms is collected on an asbestos filter. This precipitate is washed 
 until the washings come through neutral, first with tap -water, and finally 
 with distilled water. The precipitate is now dissolved in 11 c.c. of 
 hydrochloric acid (sp. gr. 1"195), washed with water, and the mixed 
 solution and washings made up to 50 c.c. The mixture is then sub- 
 mitted to steam distillation, and 50 c.c. of the distillate collected. 
 This is neutralised with 20 per cent, caustic soda solution, using methyl 
 
 N 
 orange as the indicator, and then made faintly acid with — hydro- 
 
 N 
 chloric acid. Ten c.c. of an r^ solution of sodium hydrogen sulphite 
 
 solution are now added, and the mixture left to stand overnight. 
 
 N 
 Next day 5 c.c. of -rjr iodine solution are introduced, and the excess 
 
 N . . 
 of sodiijm hydrogen siilphite titrated with -r— ~ iodine solution, using 
 
 two drops of starch paste as the indicator, until the blue colour persists 
 for fifteen seconds. As a control the titration is repeated with 50 c.c. 
 
 N 
 of distilled water to which have been added 10 c.c. of r-K sodimn 
 
 N 
 hydrogen sulphite and 5 c.c. of r^ iodine solution, imtil the same tint 
 
 is obtained. The difference between the quantities of — ^ iodine 
 solution used in the two experiments is termed " the percentage iodine 
 coefficient " of the urine. This, multiplied by the total twenty-four 
 hours' output of tirine expressed in decilitres and decimals of a deci- 
 litre, gives the " total iodine coefficient." 
 
276 GLYCOSURIA 
 
 I have found that the iodine coefficient of healthy urines is 
 nil, and that even in those who are following sedentary occupations 
 or have symptoms pointing to slight digestive or hepatic troubles, 
 the percentage coefficient rarely exceeds 1 or 1-5. In cases of 
 pancreatitis, on the other hand, it generally ranges from 12 to 20 
 per cent., with a total of 100 to 200 for the twenty-four hours. 
 Similar figures have been obtained with several cases of maUgnant 
 disease of the pancreas. A total daily coefficient of 100, or over, 
 has been given by 75 per cent, of the cases of untreated diabetes 
 that I have investigated, a result comparable with that obtained by 
 the fermentation process. In a few cases of severe diabetes readings 
 of between 400 and 500 have been obtained. The coefficient has 
 been found to fall as a result of suitable treatment, ultimately 
 reaching a total of only 5, or even 2-5, in some cases that responded 
 well. In others it has not been possible to reduce it below about 
 50, and these have generally been cases in which no form of treat- 
 ment was followed by permanently satisfactory results. That the 
 iodine coefficient of the urine is independent of its sugar-content 
 is shown by the results of numerous experiments in which various 
 sugars have been added to normal urines. Dextrose, up to as 
 much as 15 per cent., has given a negative result. With levulose 
 it was also negative up to 2 per cent., and after that rose 1-2 for 
 each 5 grams of the added sugar ; as, however, the levulose found in 
 diabetic urines, unlike that of plant origin, is removed by treating 
 the urine with basic lead acetate, it is not likely to introduce any 
 serious error. With lactose and maltose a negative iodine co- 
 efficient was given by quantities up to 2 per cent. The coefficient 
 for arabinose and xylose has been found to be about 30 to 35 for 
 each 0-1 gram of the added pentose. 
 
 Adrenalin Mydriasis. — In 1907, Loewi stated that the instilla- 
 tion of adrenalin into the conjunctival sac has no appreciable effect 
 on normal animals, but that after the pancreas has been removed 
 rapid and pronounced dilatation of the pupil occurs. He suggested 
 that this mydriatic effect might be of use as a sign of pancreatic 
 insufficiency. On applying the test to the human subject, using a 
 few drops of a 1 : 1000 solution of adrenalin or some similar supra- 
 renal preparation, he found that a positive result was obtained with 
 ten out of eighteen cases of diabetes, a marked dilatation of the 
 pupil, that commenced in a few minutes and lasted for some time, 
 taking place, but that only two out of thirty patients suffering from 
 a variety of other diseases gave a reaction. Loewi's findings were 
 confirmed in two cases by Glassner, and also in several cases by 
 Schwarz, who, however, obtained negative results more often in 
 
PERSISTENT GLYCOSURIA 277 
 
 pancreatic diseases, and several times in ordinary diabetes and 
 exophthalmic goitre, " that is in morbid conditions, one of which 
 certainly, and the other possibly, stand in some etiological relation 
 to the pancreas." Quadrio examined twenty-five patients and 
 found that twenty, who presented no evidence of pancreatic disease, 
 gave no reaction, but that pronounced mydriasis resulted in five. 
 One of these was an epileptic, and the other four had tumours 
 of the pancreas or were suffering from advanced diabetes. The 
 explanation of the reaction suggested is that the normal pancreas 
 exerts an inhibitory effect on the sympathetic nervous system 
 and prevents the action of adrenahn on it, so that when the gland 
 is removed, or ceases to function, the sympathetic becomes more 
 excitable and adrenahn exerts its full effect. 
 
 Although it is certain that the methods at present available 
 are not capable of detecting even the majority of cases of glycosuria 
 dependent upon lesions of the pancreas, it is possible by using 
 one or more of the preceding tests, or better by employing several 
 in conjunction, to demonstrate insufficiency of the gland with a 
 considerable degree of certainty in a few instances. In others the 
 results are indefinite or conflicting, so that, at the best, only a 
 ^provisional diagnosis can be made. 
 
 BIBLIOGRAPHY 
 
 Acard, Thesis, Paris, 1895. 
 
 Adler, Zeit. f. Heilkunde, 1904. 
 
 Anschiitz, Deut. Arch. f. klin. Med., 1899. 
 
 Baldwin, quot. Herter, Lect. on Clin. Path, 1898. 
 
 Barker, Med. Record, 1908. 
 
 Barrenscheen, Biochem. Zeit., 1912. 
 
 Bamngarten, Zeit. f. exp. Path. u. Therap, 1905. 
 
 Beclere, Oaz. Med. d. Paris, 1894. 
 
 Benda and Stadelmann, Deut. med. Woch., 1896. 
 
 Bernard, Legons sur le diabete. 
 
 Bernardt, Sympt. u. Diag. d. Hungeschwulste, 1865. 
 
 Bernoulli, Corresp. Blatt. f. Schweizier Aerzte, 1910. 
 
 Beyea, Amer. Journ. of Ohstet., 1900. 
 
 Bloch, quot. Oser, Nothnagel's Ency. of Pract. Med., 1903. 
 
 Bliimenthal, v. Noorden's Clinical Medicine, 1906. 
 
 Bonniger, Deut. med. Woch., 1908. 
 
 Bosanquet, Lancet, 1905, 1909. 
 
 Borchardt, Zeit. f. klin. Med., 1908. 
 
 Bouchardat, Traite du Diabete, 1875. 
 
 Brentano, Arch. f. klin. Chir., 1900. 
 
 Bright, Med. Assoc. Trans., 1833. 
 
278 GLYCOSURIA 
 
 Brunton, St. Barth. Hosp. Rep., 1874. 
 Bvirghart, Verein. /. inn. Med., 1897. 
 CaiTimidge, Lancet, 1904, 1906 ; Med. Chi. Trans., 1906 ; Surg. 
 
 Gynecol, and Obst., 1908 ; Proc. Roy. Soc, 1909 ; Proc. Roy. Soc. 
 
 of Med., 1910 ; The Pancreas, Its Surgery and Pathology, 1907. 
 Cannon, New York Med. Journ., 1912. 
 
 Cavazzani, Zeigler's Centralb., 1893; Centralb. f. inn. Med., 1894. 
 Cecil, Journ. of Exp. Med., 1909. 
 
 Chauffard and Ravant, Arch. d. Med. exp. et d'Anat. path., 1901. 
 Cohnheini, Zeit.f. phys. Chem., 1903. 
 
 Da Costa and Beardsley, Amer. Journ. Med. Sci., 1907, 1908, 
 Cowley, Lond. Med. Journ., 1788. 
 Croom, Brit. Gynaecol. Journ., 1895. 
 Cashing, Johns Hopk. Hosp. Bull., 1911. 
 Dallemange, Arch. d. Med. exp., 1895. 
 Decker, Bull. State Univ., Iowa, 1908. 
 Dieckhoff, Beitrdge z. Path. Anat. Praxis., 1895. 
 Dompeling, Nederl. Arch. v. Genel. Natuurk., 1868. 
 Dule, Dublin Hosp. Gaz., 1857. 
 Dumontpelier, Compt. rend. d. Soc. d. Biol., 1867. 
 Dutourier, Thesis, Paris, 1895. 
 Ehrlich and Frerichs, Zeit.f. klin. Med., vi. 
 
 Ehrmann, Arch. f. ges. Physiol., cxix. ; Berl. klin. Woch., 1912. 
 Eichler and Silbergleit, Berl. klin. Woch., 1908. 
 Eppinger, Falta, and Rudinger, Wein. klin. Woch., 1907 ; Zeit. f. 
 
 klin. Med., 1908-9. 
 Erdheim and Stiimme, Beitr. z. path. Anat. u. z. allg. Path., 1909. 
 Ewald, Berl. klin. Woch., 1895. 
 Falta, Zeit. f. klin. Med., 1910. 
 Fedeh and Romanelli, Riforma Medica, 1909. 
 Ferrand, Rev. Neurolog., 1901. 
 Fitz, Med. Congress, Washington, 1903. 
 
 Frerichs, Ueber Diab. mellitus, 1893 ; Virchow's Arch., 1903. 
 Funck, MUnch. med. Woch., 1909. 
 Garrod, Lancet, 1912. 
 Gentes, Thesis, Bordeaux, 1901. 
 
 Goetsch, dishing, and Jacobson, Bull. Johns Hopk. Hasp., 1911. 
 Goiffon and Tallarico, Arch. d. Malad. d. VAj^p. Digestif., 1912. 
 Goodman, Annals of Surgery, 1909 ; Internal. Clinic, 1909. 
 Grawitz, Fortschr. d. Med., 1897. 
 Grimbert and Bernier, Presse Medicale, 1910. 
 Gross, Deut. med. Woch., 1909. 
 Grossman, Berl. klin. Woch., 1879. 
 Grtibe, Zeit.f. klin. Med., 1895. 
 Griind, Zeit. f. physiol. Chem., xxxv. 
 Guleke, Arch. f. klin. Chir., Ixxxvii. 
 Handmann, Deut. Arch. f. klin. Med., 1911. 
 Hanot and Chauffard, Rev. de Med., 1882. 
 
PERSISTENT GLYCOSURIA 279 
 
 Hanseman, Zeit. f. klin. Med., 1894 ; Berl. klin. Woch., 1887, 1897 ; 
 
 Verhand. d. deut. path. Oesellsch., 1902 ; Berl. klin. Woch., 1912. 
 Hanyn, Der Diabetes. 
 
 Hartmann, Inaug. Dissert., Tubingen, 1878. 
 
 Heiberg, Wien. klin. Woch., 1910 ; Deut. Arch. f. klin. Med., 1911. 
 Henkel, Deut. med. Woch., 1909. 
 Herlitzka, Roy. Accad. d. med. d. Torino, 1908 ; PfliXger's Arch., 
 
 1908. 
 Herxheimer, Virchow's Arch., 1906 ; Verhand. d. deut. path. 
 
 Gesellsch, 1907. 
 Herzog, Virchow's Arch., 1902. 
 Hirschfeld, Berl. klin. Woch., 1905, 1908 ; Deut. med. Woch., 
 
 1910. 
 Hiirter, Mediz. Klinik., 1910. 
 Imlach, Brit. Med. Journ., 1885. 
 James, Brit. Journ. of Dermatol., 1894. 
 Jeanselme, Bull. d. hop. d. Paris, 1897. 
 Jodry, Lehere Diab. Suisse, 1909. 
 Jolles, Wiener 7ned. Woch., 1911. 
 De .Jonge, Arch. f. Psychiat., 1882. 
 Kehr, Munch, med. Woch., Ivi. ; Mitteil a. d. Gfrenzgeb. d. med. u. 
 
 Chir., XX. 
 Kinnicutt, Med. Record, 1909. 
 Klebs and Miink, Hanb. d. path. Anat., iii. 
 Kleniperer, Verhand. d. Vereins /. inn. Med., 1896 ; Berl. klin. 
 
 Woch., 1892. 
 Koziczkowsky, Zeit. f. klin. Med., 1909. 
 Krienitz, Arch. f. Verdauungn. Krank., xv. 
 Lancereaux, Semaine Med., 1895. 
 
 Lannois and Roy, Arch. gen. d. med., 1903 ; Semaine Med., 1903. 
 Lefas, Arch. gen. d. med., 1900. 
 Lepine, Rev. Scientifique, 1891 ; ReiK d. Med., 1896 ; Compt. rend. 
 
 d. Soc. d. Biol., 1903 ; Lyon Medicale, 1903 ; Journ. d. Phys. et d. 
 
 Path, gen., 1905 ; Diabete Sucre, 1909. 
 Levrat and Perrotton, Thesis, Paris, 1859. 
 Levy, Medical Record, 1910. 
 Lewinski, Deut. med. Woch., xxxiv. 
 
 Loeb, Deut. Arch. f. klin. Med., 1884 ; Centralb. f. inn. Med., 1898 
 Loewi, Munch, med. Woch., 1907. 
 Lowei, Arch. f. exp. Path. u. Pharm., 1902. 
 Llithje, Munch, med. Woch., 1901 ; Cong. f. inn. Med., 1907. 
 Liiys, Gaz. Med. d. Paris, 1869. 
 Maass, Mediz. Klinik., 1909. 
 
 Mannkopf, quot. Naionyn, Die Diab. Mellit., 1906. 
 Marck, Zeit. f. Oyncecol., 1911. 
 Margin, Rev. d. Med., 1905. 
 Marie, Semaine Med., 1895. 
 Marinesco, Compt. rend. d. Soc. d. Biol., 1895; Semaine Mid., 1895. 
 
280 GLYCOSURIA 
 
 Mautry, Brit. Med. Journ., 1889. 
 
 Mendel and Jackson, ATner. Journ. of Physiol., 1902. 
 
 Michael, Deut. Arch. f. klin. Med., 1889. 
 
 Minkowski, Arch. f. exp. Path. u. Pharm., 1886, 1893, 1908. 
 
 Mirallie, Gaz. d. hop. d. Paris, 1893. 
 
 Montgomery, Journ. Amer. Med. Assoc, 1912. 
 
 Mosler, Deut. Arch. f. klin. Med., xv. 
 
 Miiller, Verhand. d. xxiii. Cong. f. inn. Med., 1906 ; Arch. f. klin. 
 
 Med., 1908. 
 Nash, Lancet, 1901. 
 Naunyn, Nothnagel's Spec. Path. u. Therap., 1898 ; Der Diab. 
 
 Mellitus, 1906. 
 Neubauer, Biochem. Zeit., 1905. 
 Von Noorden, Diab. Mellitus, 1906 ; Die Zuckerkrank u. ihre Behand., 
 
 1910; Mediz. Klinik., 1911. 
 Norris, Proc. New York Path. Soc, 1907. 
 Ogle, St. George's Hosp. Rep., 1866. 
 Opie, Journ. Expt. Med., 1899, 1901 : Diseases o/ the Pancreas, 
 
 1903. 
 Oser, Nothnagel's Ency. of Pract. Med., 1903. 
 Pavy, Lancet, 1908. 
 
 Perwvischin and Foworski, Neurol. Centralb., 1900. 
 Pfliiger, Arch. f. ges. Physiol, 1907, 1908. 
 Pilcher, Annals of Surgery, 1910. 
 Pinches, Jahrb. d. Wiener. Krank, 1897 ; Bull. d. Sci. Med. d. Scula 
 
 Med. d. Bologna, 1897. 
 Porges, Wien. klin. Woch., 1909 ; Zeit. f. klin. Med., 1910. 
 Porges and Salomon, quot. v. Noorden, Die Zuckerkrank, 1910. 
 Potter and Milne, Amer. Journ. of Med. Sci., 1912. 
 Proescher, Grafe's Arch. /. Ophthal., 1899. 
 Pusey, Arch. f. Ophthal., 1904. 
 Quadrio, II Policlinico, xv. 
 Quintard, Bull. d. Soc. mid. d. Ang., 1894. 
 Rabe, Bull. d. Soc. Anat. d. Paris, 1900. 
 Ravaut, Soc. med. d. hopit. d. Paris, 1900. 
 Reale and Renzi, Verhand. d. X. internat. Med. Cong., 1890; Wien 
 
 med. Woch., 1891 ; Berl. klin. Woch., 1892. 
 Von Recklinghausen, Tagebl. d. Ixii. Versam. deut. Naturf. u. Aertze, 
 
 1889. 
 Reichenstein, Wien. klin. Woch., 1909. 
 Richardson, Med. Times and Gaz., 1866. 
 Richartz, Zentralb. f. inn. Med., 1910. 
 Rolleston, Diseases of the Liver, 1905. 
 Rosenbaiuii, Arch. f. ges. Physiol., cxxi. 
 Rosenberger, Zeit.f. Biol., 1907. 
 Rossle, Vern. deut. path. Gesellsch., 1907. 
 Roth, Zeit.f. klin. Med., 1909. 
 .Sauerbeck, Virchowi's Arch., 1904. 
 
PERSISTENT GLYCOSURIA 281 
 
 Saundby, Led. on Diabetes, 1891. 
 
 Schafer, Lancet, 1895. 
 
 Schleisinger, Wien. klin. Rundschau, 1900; Oesell. /. inn. Med., 
 
 1902. 
 Schmey, Centralb. f. inn. Med., 1897. 
 Schmidt, Prog. u. Therap. d. Zuckerkrank, 1892 ; Milnch. med. 
 
 Woch., 1902 ; Mitteil a. d. Genz. d. Med. u. Chir., 1909. 
 Schroeder, Journ. Amer. Med. Assoc, 1908. 
 Schwarz, Wiener klin. Woch., 1909. 
 Scott, Journ. of Path, and Bact., 1907. 
 Seegen, Der Diab. Mellit., 1893. 
 Siebke, Deut. med. Woch., 1910. 
 Silver and Irvine, Trans. Path. Soc, 1878. 
 Smith, Brit. Med. Journ., 1883. 
 
 Speese and Goodman, Amer. Journ. Med. Sci., 1909. 
 Ssobolew, Centralb. f. allg. path. u. path. Anat., 1900; Arch. /. 
 
 path. Anat., 1902. 
 Stadelmann, Deut. Med. Woch., 1901 ; Zeit. f. klin. Med., 1904. 
 Stiles and Lusk, Amer. Journ. of Physiol., 1903 ; Deut. Arch. / 
 
 klin. Med., 1904. 
 Still, Encyclop. Med., 1901. 
 Striimpell, Deut. Zeit. f. Nervenheilk., 1897. 
 Sweet, Journ. Med. Res., 1903. 
 Tachau, Deut. Arch. f. klin. Med., 1912. 
 Trevor, Practitioner, 1904. 
 
 Turner and Stewart. Text-book of Nervous Dis., 1910. 
 Volhard, Milnch. med. Woch., 1907. 
 Watson, Lancet, 1908 ; Brit. Med. Journ., 1908. 
 Weichselbamn, Wiener, klin. Woch., 1911. 
 Weiland, Deut. Arch. f. klin. Med., 1911. 
 West, Path. Soc. Trans., 1890. 
 White, Path. Soc. Trans., xxxvi. 
 Wilks, Lancet, 1875. 
 Williamson, Diabetes Mellitus, 1898. 
 Windle, Dublin Journ. Med. Sci., 1883. 
 WooUey and Newburgh, Journ. Amer. Med. Assoc, 1911. 
 Zack, Wiener klin. Woch., 1908. 
 
CHAPTER VIII 
 
 PERSISTENT GLYCOSUEIA METABOLISM 
 
 Before proceeding to deal with the metabolism in diabetes, it is 
 advisable that we should briefly consider the metabohsm and food 
 supply of the healthy organism. 
 
 By metabohsm in its widest sense is meant the sum of the 
 chemical changes that go on in the body, but it is also used in a 
 restricted sense to mean the processes by which the cells incorporate 
 food materials of various kinds into their substance. The metaboUc 
 changes that go on in the tissues are of two kinds — (1) constructive, 
 or anabolic, involving change from a lower to a higher state ; 
 (2) destructive, or katabolic, in which the reverse occurs, with the 
 formation of various waste products. The former are chiefly con- 
 cerned in the maintenance of the tissues of the body, while the 
 latter give rise to the phenomena of motion that constitute life, 
 and are the source of the energy of the living organism. In early 
 life the metabolic processes of the body are more intense than later, 
 and anabohsm outstrips katabohsm, with the result that the body 
 grows. In later Hfe the two processes are more nearly balanced, 
 and after growth has ceased they should be approximately equal 
 if the individual is to remain healthy. For this condition to exist 
 it is necessary (a) that the supply of food material reaching the 
 tissues should be sufficient ; (b) that it should be in a form that 
 can be utihsed ; (c) that the physical condition of the environment 
 should be suitable. With the last we are not at present concerned, 
 but the other two are of great importance. 
 
 The questions that arise then are : What is a sufiicient food 
 supply, and how are we to ascertain that it is being made use of by 
 the tissues ? The answer to the latter is the more easily obtained, 
 for by a regular use of the scales we can determine whether weight 
 is being lost or gained — that is to say, whether katabohsm or ana- 
 bohsm preponderates. The information obtained in this way mil 
 partty answer the first question, but not completely, since a de- 
 ficiency may not be in bulk but kind, and a loss of weight may be 
 due to the food not being of a nature, and in the proportions that 
 are most suitable. The direction and nature of the metaboHc 
 processes going on within the boclv can be determined in another 
 
 282" 
 
PERSISTENT GLYCOSURIA 283 
 
 way — viz. b}^ balancing the intake against the output. In the 
 intake are included the food and oxygen, in the output the urine, 
 the faeces, &c. To make a complete quantitative analysis of the 
 ingesta and excreta would involve an enormous expenditure of 
 time and labour, and as a rule sufficient information can be 
 obtained by considering two elements only, the nitrogen and 
 the carbon. Of these the nitrogen is the more important, for the 
 nitrogenous foods are the only elements of the diet able to repair 
 tissue waste, and the nitrogen content of the excreta is a measure 
 of the tissue destruction that is going on within the body. When the 
 nitrogen of the food exactly balances the nitrogen excreted, the 
 body is said to be in " nitrogenous equilibrium," and it may be 
 assvimed that the living material of the tissues is not being increased 
 or diminished in amount. For practical work in dietetics a suffi- 
 ciently reliable estimate of the nitrogen content of the food is 
 obtained by weighing it, working out the amount of protein it 
 contains, and allowing 1 gram of nitrogen for each 6-25 grams of 
 protein, since meat protein contains an average of about 16 per cent, 
 of nitrogen. It is usually not necessary to estimate the nitrogen 
 contained in the faeces, for, as a rule, only a small fraction, about 
 1 or 2 grams a day, escape absorption and are lost in the stool ; it is 
 therefore only needful to determine the nitrogen content of the 
 urine. This is estimated by Kjeldahl's process, or some modifica- 
 tion of it. 
 
 A different level of nitrogenous equilibrium exists in different 
 individuals, and it may also vary in the same person at different 
 times. If the amount of nitrogenous food is diminished the amount 
 of urinary nitrogen will also decrease. Should the amount of food 
 then remain constant the output of nitrogen will likewise remain 
 the same ; but if, on the other hand, more nitrogen is ingested an 
 increased elimination will result, and at the same time a certain 
 proportion will be retained by the body, so that gradually a higher 
 level of equilibrium is established. There is a natural limit, 
 however, to this power of accommodation, and a point is finally 
 reached, varying with different individuals, where a further increase 
 in the amount of ingested nitrogen does not lead to a higher level 
 of equilibrium, and where, consequently^ a further retention of 
 nitrogen does not occur. Over-feeding then results and various 
 digestive disturbances, such as diarrhoea and vomiting, follow. 
 If a person in nitrogenous equihbrium takes an insufficient amount 
 of food it will lead to an increased destruction of the organised 
 albumens, and more nitrogen will be lost in the urine than is ab- 
 sorbed from the food. For a time, the reserve of fats and carbo- 
 hydrates is capable of protecting the body against an unduly rapid 
 
284 GLYCOSURIA 
 
 loss of nitrogen, but finally the protection fails and death occurs 
 from inanition. 
 
 It does not follow that because the body is in nitrogenous 
 equihbrium that the other elements of the food and excreta are 
 also balanced, and we maj^ have an increase in the fats, and to a 
 less extent in the carbohydrates, or the reverse, shown by the 
 carbon, hydrogen, and oxygen of the ingesta being more, or less, 
 than in the excreta. 
 
 Every process of the body is attended with manifestations 
 of energy, shown as mechanical work, electric currents, or heat. 
 This energy is derived from the food, and is an expression of the 
 potential energy that it contains. Now, the researches of Mayer 
 and Joule showed that the amount of power, or energy, that can 
 be obtained from a given weight of matter is connected with, and 
 is proportional to, the heat given out dm-ing its combustion, and 
 as heat is the simplest measure of potential energy that can be 
 obtained, it is convenient to calculate both the potential energy 
 of the food, and the work done by the organism, in terms of heat 
 units. The standard measure of heat, or heat unit, is the calorie. 
 This is the amount of energy required to raise 1 gram of water 
 1° C, and is equivalent to 425-5 units of work, or gram-metres — 
 that is to say, the energy required to raise 1 gram of water 1° C. 
 would lift a weight of 425-5 grams to a height of 1 metre. ^ It is, 
 however, an advantage in practice to use a larger unit than this, 
 viz., the amount of energy required to raise 1000 grams of water 
 1° C, which is termed the Kilo-, or large, Calorie. 
 
 Foods which on oxidation give out the greatest amount of heat 
 should theoretically have the greatest capacitj^ for producing work, 
 but the heat equivalents of organic substances cannot be calculated 
 from their chemical composition ; for part of the heat, varying with 
 different substances, is used in the process of dissociating the 
 molecules, &c., and the heat equivalent has therefore to be deter- 
 mined by direct calorimetric methods. Rubner, as the result of 
 his experiments, came to the conclusion that the heat value of 
 1 gram of protein in an average diet is 4-1 calories, although sKght 
 differences exist between different forms {e.g. casein =4-4, meat 
 protein = 4-233, vegetable protein = 3-96 calories). For fats 
 Stohmann's figures were — ohve-oil = 9-384 cal., animal fat = 9-372, 
 butter fat = 9- 179 cal. Rubner therefore adopted 9-3 as the aver- 
 age heat value of 1 gram of fat in a mixed diet. The following 
 heat values have been found for carbohydrates — dextrose, 3-692 
 to 3-755 cal. ; milk-sugar, 3-877 cal. ; cane-sugar, 3-959 to 4-001 
 cal. ; starch, 4-116 cal. Taking into account the predominating 
 ^ To convert kilogram metres into food pounds x 7-233. 
 
PERSISTENT GLYCOSURIA 285 
 
 importance of starch in the average diet, Rubner gave the carbo- 
 hj^clrate group a heat value of 4-1 calories. Rubner's " standard 
 values " have been widely adopted, and are generally used in. 
 determining the heat value of a mixed diet. They are : — 
 
 1 gram of protein . . . . .4-1 calories 
 
 1 „ „ fat . . . . . 9-3 
 
 1 ,, ,, carbohydrates . . . . 4"1 ,, 
 
 Atwater and Bryant, as the result of over four hundred experi- 
 ments, obtained shghtly different results, which they consider are 
 the heat values absolutely available in computing the average 
 diet : — 
 
 1 gram of protein ..... 4-0 calories 
 
 1 „ „ fat 8-9 
 
 1 ,, ,, carbohydrate .... 4*0 ,, 
 
 The difference between the two standards is probably to be 
 explained by the fact that Rubner used comparatively pure foods, 
 ^vhile the waste in the fseces in Atwater and Bryant's experiments 
 reduced the amount of available nutriment. Since the combustion 
 of 1 gram of fat produces 9-3 calories, and the conversion of 1 gram 
 of protein into urea, carbon dioxide, and water, and of 1 gram of 
 carbohydrate into carbon dioxide, each produces 4*1 calories, the 
 combustion of 100 grams of fat will give rise to an amount of energy 
 equal to that produced by 227 grams of protein, or carbohydrate. 
 This amount of protein, or carbohydrate, is said to be of the same 
 " isodjrnamic value " as 100 grams of fat. 
 
 An average adult man expends daily about 34 to 35 calories for 
 each kilogram of his body-weight when taking moderate exercise, 
 w^hen at rest about 10 to 15 per cent, less, and when taking active 
 exercise about 10 to 15 per cent, more, so that a man weigh- 
 ing 70 kilos (154 lb.) doing hght work would require about 2500 
 calories in the twenty-four hours. This amount of energy would 
 be supplied by — 
 
 118 grams of protein .... 484 calories 
 56 „ „ fat 621 
 
 500 ,, ,, carbohydrates . . . 2050 ,, 
 
 3055 calories (Voit) 
 
 provided that all the food materials were fullj^ utihsed for the 
 purpose of heat production in the body. As a matter of fact, the 
 heat value of foods to the body is a httle less than is shown by the 
 calorimeter, owing to the loss of unoxicUsecl products in the ex- 
 cretions, &c. Atwater, in his more recent experiments, estimates 
 that 4 J per cent, of the nourishment taken is unutihsed by the 
 
286 GLYCOSURIA 
 
 organism, while Rubner gives a slightly higher figure, 5 to 5| per 
 cent. A certain proportion of the food ingested is rendered value- 
 less to the body as the result of putrefactive and fermentative 
 changes in the stomach and intestine, or is lost in the faeces, so that 
 the actual caloric value of the food material to the body is always 
 less than the theoretical value of that taken by the mouth. As a 
 rule, probably somewhere from 10 to 25 per cent, should be deducted 
 from the calculated value to represent this loss, but in cases where 
 abnormal putrefactive and fermentative changes are going on in 
 the gastro-intestinal tract, a much larger deduction must be made. 
 The proportions of protein, fats, and carbohydrates in a mixed 
 ■diet may be varied within limits, but generally it is found that one 
 part of protein food is taken to each four or five parts of non- 
 protein, and that of the latter one part of fat is taken to five or 
 ten of carbohydrate. According to Voit, one-third of the protein 
 should be animal and two-thirds vegetable. Wide variations, 
 however, exist among different races, and among different social 
 grades of the same race. 
 
 Protein Requirement. — In the distribution of the chet men- 
 tioned above, and suggested by Voit for an average labourer working 
 eight to ten hours daily, it will be noticed that 118 grams of protein 
 a day are allowed. This allowance, which represents from 1-4 to 1-7 
 grams per kilogram of body-weight, was arrived at by a statistical 
 method, and showed what the average labourer is in the habit of 
 consuming. Rubner obtained a higher figure, 127 grams, for the 
 same class, and Atwater allows 125 grams. For men doing hard 
 work Voit gives 145 grams, Rubner 165 grams, and Atwater 150 
 grams. 
 
 Experiment has proved that with an ordinary cUet on Voit's 
 standard, muscular exertion increases the metaboHsm from 1-1 per 
 cent, to 18-0 per cent., with an average of 15 per cent. The addition 
 of 10 to 15 grams of protein daily to the physiological minimum, 
 raising it to 44 or 45 grams, might therefore be expected to suffice 
 for the protein requirement when an average amount of work is 
 being performed. Siven found that a man weighing 65 kilos could 
 be maintained in nitrogenous equilibrium for a short period on a 
 diet containing as little as 4 or 5 grams of nitrogen, that is to say, 
 on 25 to 31 grams of protein a day, but Munk and Rosenheim showed 
 that dogs given a quantity of protein only sufficient to maintain 
 nitrogenous equilibrium gradually lost strength, and suffered from 
 digestive disturbances. Numerous investigations undertaken by 
 Voit, Moleschott, Ranke, Foster, Munk, and others have shown that 
 when no food is taken for twenty-four hours the amount of nitrogen 
 
PERSISTENT GLYCOSURIA 287 
 
 excreted varies from 7'5 to 12 grams, the variation probably being 
 due to a difference in the amount of stored and circulating protein 
 and glycogen in the organism. But if the starvation is continued 
 the stored protein and glycogen is all consumed in a few days, and 
 after a period of about five days all the nitrogen in the excretions 
 will be derived from the disintegration of tissue proteins. In 
 fasting people the amount of nitrogen excreted by the kidneys 
 from the fifth day of starvation averages about 4*5 grams per 
 day. If we allow a margin of 1 gram for other modes of excretion, 
 we arrive at the conclusion that the amount of protein absolutely 
 essential to prevent the destruction of tissue proteins is (nitrogen 
 5-5 X 6-25 = ) 34 grams daily, or about 0-5 grams of protein per kilo 
 of body- weight. 
 
 Chittenden has claimed that well-nourished men, following 
 various vocations, can be maintained in good health for a period 
 of several months on a mixed diet containing 40 to 60 grams of 
 protein. He considers that the Voit standard contains about 
 twice the necessary amount, and that the excess, consumed only 
 from habit and seK-indulgence, throws an unnecessary strain on 
 the. liver, kidneys, and other organs concerned in the transforma- 
 tion and ehmination of the end-products of protein metabohsm. 
 Chittenden's results have been much criticised on the ground that 
 the men tested were all of the better classes, and could undergo a 
 greater restriction than the poorer classes, who are not so well 
 nourished, also that the subsequent health of the cases was not 
 reported, so that it remains to be seen whether the quantity of 
 protein in his ration, which is not greater than would be meta- 
 bolised in starvation, is advisable as a permanent standard. It 
 cannot be denied, however, that 50 grams of protein, containing 
 8 grams of nitrogen, are apparently sufficient to maintain the 
 machinery of the body in good repair. Voit himself has stated 
 that a vegetarian can live in nitrogenous equilibrium on a diet 
 containing 48*5 grams of protein, and that an active man, weighing 
 74 kilos, may remain in good condition on less than 118 grams. 
 
 It is beheved that proteins undergo different metabohc changes 
 according to the purpose they are to serve in the body, a part, the 
 so-called " repair-proteins," appear to enter into the hving sub- 
 stance of the tissue cells to make good the waste of their substance 
 that the vital processes involve, while another part, the so-called 
 " energy-protein," is used as a source of work and heat. The repair 
 protein is slowly broken down, giving rise to the uric acid, creatinin, 
 and neutral sulphur compounds of the urine. The energy protein 
 appears to undergo a rapid process of denitrification, the nitrogen- 
 containing fraction which is spht off being oxidised and excreted 
 
288 GLYCOSURIA 
 
 in the form of urea and inorganic sulphates, while the carbonaceous 
 portion is used, Hke the carbohydrates and fat of the food, for the 
 production of energy. The proportion of the total intake of protein 
 used for energy production depends upon circumstances. If the 
 food contains much more protein than is required to maintain 
 nitrogenous equihbrium, the larger part is made use of in this way, 
 but if only a small amount of protein is taken, the greater part, 
 or even the whole, is used for tissue repair. The cleavage of protein 
 is increased by muscular work, but the cleavage is small in pro- 
 portion to the consumption of non-]3rotein material, and it is not 
 cUrectly connected with the work performed. When the body con- 
 tains no stored proteins muscular exertion increases the output of 
 nitrogen 15 per cent. ; but if the body is in good '' condition," that 
 is, containing a store of circulating proteins, the output of nitrogen 
 is larger, because such proteins are consumed as a source of heat 
 and energy ; but the muscle fibres are not broken clown in a greater 
 proportion by the work they perform. In Paton's investigation 
 the excretion of nitrogen was increased both during and after 
 the performance of work, and the metabolism of protein, indicated 
 by the increased nitrogen excretion, accounted for 35 per cent. 
 of the work done. It was evident, therefore, that one-third of the 
 energy expended in work was derived from the metabolism of 
 nitrogenous, and two-thirds from non-nitrogenous matter. 
 
 It is not a matter of indifference, however, to the organism 
 as to whether its energy is derived from a nitrogenous or a non- 
 nitrogenous source, for not only are the waste products from the 
 former more highly organised and more difficult to get rid of than 
 from the latter ; but according to Rubner 28-6 per cent, of the 
 energy of meat protein is never utilised in the Ufe processes of the 
 tissues, but is hberated as free heat during the early cleavage, 
 while with cane-sugar, for instance, only 3-1 per cent, of its energy 
 is dissipated as heat when it is inverted into levulose and dextrose. 
 Rubner found that if the quantity of protein in the chet of a dog is 
 raised above the requirement by 56 per cent., there is an increase 
 in heat production of 19 per cent., with a rise of 90 per cent, heat 
 production is increased 35 per cent., and with a rise of 105 per cent., 
 44 per cent, more heat is produced. The temperature scarcely 
 changes so ^Derfect is the regulatory mechanism, but more rapid 
 respiration indicates the increased oxidation and the efforts of 
 the body to rid itself of the excess of heat. This effect of abundant 
 protein food in raising metabohsm is called by Rubner the " specific 
 dynamic " action of protein. 
 
 In a state of health, the nature and amount of the other con- 
 stituents of the diet exert an influence on the direction of protein. 
 
PERSISTENT GLYCOSURIA 28& 
 
 metabolism. The presence of carbohydrates, fats, and gelatine in 
 the food enables a larger proportion to take part in tissue repair, 
 shielding it from denitrification and from being used for purposes 
 of energy production. Such substances are known as " protein 
 sparers." The exact mechanism of their action is not understood, 
 but it is possibly the effect of " mass influence." If such is the 
 case, it is obviously important that a low protein ration should 
 be spread as evenly as possible over the day, and be thoroughly 
 blended with the non-nitrogenous food materials. The intimate 
 mixture of protein and carbohydrates that exists in vegetable 
 foods possibly explains whj- nitrogenous eqiiiUbrium is attained 
 more readily on a vegetable than on any other form of diet. It 
 would also seem that the kind of protein is not altogether a 
 matter of indifference, but that some varieties form amino acids 
 that more readily yield their carbohydrate moiety than others. 
 
 Fats and Carbohydrates. — The fats and carbohydrates are 
 the chief source of energy, and are much more economical, botk 
 financially and physiologically, than proteins. To obtain equiva- 
 lent amounts of energy, for instance, from protein and fat eleven 
 and a half times more of the former by weight must be de- 
 stroyed than of the latter, since each gram of nitrogen that is lost 
 corresponds to a diminution of body-weight of 33 grams by " flesh," 
 with an energy yield of 0-8 calories per gram, while the oxidation of 
 1 gram of fat simply means the loss of 1 gram of body-weight with 
 an energy yield of 9-3 calories. The ingestion of carbohydrates or 
 fat alone, while supplying power to the organism, does not prevent 
 tissue waste, and, although death is delayed by a fatty or carbo- 
 hydrate diet, it eventually ensues from the gradual weakening of 
 the vital organs that takes place. 
 
 Fat. — Voit showed that the ingestion of 100, 200, and 300 grams 
 of fat by a fasting animal scarcely influences protein metaboHsm at 
 all, tending, in fact, rather to increase than to diminish it if any- 
 thing. It would seem that the ingested fat is simply burned in 
 place of the body fat, the total consumption of protein and fat 
 remaining unchanged. Schulz found that during starvation there 
 is an increase in the quantity of fat in the blood, and Rosenfeld 
 showed that the quantity of fat in the liver rises to 10 per cent, 
 during fasting, and to as much as 25 per cent, if fats alone are taken. 
 
 If carbohydrates are given to a fasting animal the fat-content 
 of the Hver falls to 6 per cent., but if carbohydrates are given along 
 with fats the latter are not retained by the Hver as they are when 
 taken alone, so that there appears to be an antagonism between 
 glycogen and fat deposition in the liver. The effects on metaboHsm 
 
 T 
 
290 GLYCOSURIA 
 
 j)rodiiced by the ingestion of fats and protein together have been 
 investigated by Voit and Korkunoff. They found that much less 
 protein food is required to maintain nitrogenous equiUbrium when 
 the two are taken together than when proteins are consumed alone, 
 and that with increasing quantities of fat there is, for a time, an 
 increased addition of protein to the body. Eventually, however, 
 a higher level of nitrogenous equihbrium is estabhshed and the 
 deposition of protein ceases, any excess being used for energy 
 production, as on a protein diet. 
 
 The continued administration of an excess of fat to the healthy 
 organism leads to its being stored, first in the subcutaneous and 
 intermuscular connective tissue, and then in the abdominal cavity. 
 In the converse condition, when energy is required to meet a chronic 
 deficit in the intake, the abdominal fat is made use of first, then 
 the subcutaneous fat, after that the intermuscular deposits, while 
 the fat in the organs is onty used as a last resort. 
 
 Little is known of the stages through which fat passes in the 
 course of its utiHsation by the body, probably the first step is 
 similar to that which it undergoes in digestion, a cleavage into 
 glycerine and fatty acids. That beta-oxj^butj^ric acid is one of the 
 products of the metaboHsm of fats is suggested by a study of the 
 chemical pathology of diabetes, but Satta and others consider that 
 the available evidence is against this substance being a normal 
 cleavage product. The heat production in the early cleavage pro- 
 cesses of fat appears to be small, and Rubner calculated that the 
 specific dynamic action of fat raises the metabolism 14-4 per cent, 
 at a temj)erature of 31° C. Allowing for the difference in the 
 specific d}Tiamic action of protein and fat, it would appear that 
 these two substances can each rej)lace the other in metabolism in 
 isodjTiamic proportions. 
 
 Carbohydrates. — This class of foodstuffs has been found to 
 protect the tissues from wasting more effectually than fat. Voit 
 gave a fasting dog 500 grams of sugar and noticed a fall in protein 
 metaboHsm from 181 to 170 grams, and Rubner was able to reduce 
 the nitrogen in the urine in a fasting man from 11-9 to 6-3 grams 
 by giving carbohydrates. 
 
 The specific d%Tiamic action of carbohydrates is low, the heat 
 produced in the early cleavage processes being smaller than for fats. 
 It has been calculated by Rubner that cane-sugar, for example, 
 raises the metaboHsm 5-36 \)ev cent, on an average. Hence the 
 ingestion of sugar hy a starving animal raises the general meta- 
 boHsm to Httle above the starvation requirement, the smaU specific 
 dynamic action of the sugar absorbed scarcely exceeding the re- 
 duction due to the cHminution in protein loss, with its much greater 
 
PERSISTENT GLYCOSURIA 291 
 
 specific dynamic action, that ensues. This is well illustrated by an 
 experiment of Rubner's, in which he compared the metabohsm in 
 starvation, on cane-sugar, and on meat : — 
 
 Starvation 2042 Calories 
 
 Sugar (120 per cent, requirement) . . 2087 ,, 
 
 Meat (120 „ „ ) . . 2566 
 
 The administration of an easily absorbable carbohydrate not only 
 reduces protein metabolism, but when given above the energy 
 requirement it is retained within the body, part being stored as 
 glycogen and part as fat. The conversion of sugar into glycogen 
 is universally acknowledged, and definite proof of the conversion 
 of carbohydrate into fat, which was doubted by some, was furnished 
 by the experiments of Meissel and Strohmer, of Lehmann and Voit, 
 of Rubner and others. When a substance rich in oxygen, hke 
 dextrose, is converted into material poor in oxygen, hke fat, the 
 intramolecular oxygen becomes available for purposes of oxidation, 
 with the result that the volume of carbon dioxide expired may 
 increase and be greater than the volume of the inspired oxygen. 
 Johansson, Billstrom, and Heyl found that if 50 to 200 grams of 
 cane-sugar were given to a fasting man the carbon dioxide output 
 increased from 22*6 grams per half-hour to about 30 grams. More 
 carbon dioxide was not found to be eliminated with large than 
 with small doses, showing the regularity with which the process of 
 utihsation proceeds, and indicating that any excess is stored until 
 required by the organism. It has been shown by Sieven that the 
 level of nitrogenous equilibrium can be lowered very considerably 
 by giving carbohydrates along with protein, thus proving their 
 marked action as protein sparers. This action has also been 
 demonstrated in the reverse way by Lusk, who found that the 
 sudden withdrawal of carbohydrates from the diet causes a rise in 
 nitrogen metabohsm. When, however, the carbohydrate in the 
 diet is partly replaced by fat, protein metabohsm is not influenced, 
 or there is only a transitory change. Landergren proved that a 
 diet containing half its calories in carbohydrates, and half in fat, 
 has about the same protein protecting power as one made up of 
 carbohydrate alone. If the carbohydrates of the food are entirely 
 replaced by fat, protein metabohsm rises, but is again reduced on 
 the reintroduction of sugars and starches. The substitution of a 
 carbohydrate-free for a mixed diet also raises the amount of 
 acetone in the urine considerably above the average normal level, 
 but it diminishes, or disappears again, when carbohydrates are 
 taken. The fact that carbohj^drate inanition is the sole cause of 
 acetonuria is explained by Geelmuyden on the hypothesis that 
 carbohydrates, or a derivative, glucuronic acid, unite with acetone 
 
292 GLYCOSURIA 
 
 in intermediary metabolism, and that this process is necessary for 
 the fiurther change of the acetone bodies. If the synthesis is 
 limited, or fails, a collection of the acetone bodies takes place, and 
 they are consequently eliminated in increased quantities in the 
 urine. To summarise the chief points in carbohydrate and fat 
 metabohsm, therefore, it would appear that the former are the most 
 economical foodstuffs from every point of view ; they are the greatest 
 sparers of protein, they may ahnost completelj^ replace fat in the 
 food, they are more completely absorbed from the intestine than 
 fat when given in the form of sugar or cooked starch, although 
 when contained in some vegetables the carbohydrate content 
 of which is small relative to then? bulk {e.g. spinach, lettuce, 
 cabbage, &c.), a considerable proportion may escape absorption. 
 After being absorbed they much more quickly oxidised than other 
 food materials, and are therefore desirable when a quick supply of 
 heat or energy is required. Further, they can be given in a greater 
 variety of form than fat, so that appetite, which, after all, is a most 
 important factor in dietetics, is not blunted by a lack of change. 
 Fats, on the other hand, although they are a much more concen- 
 trated food, are more costly, both to the pocket and to the organism, 
 an excess raising protein metabolism and giving rise to acetonuria 
 when the use of carbohydrates is entirely abandoned. 
 
 Normally, the ingestion of fat has for its object the relief of the 
 intestine from excessive carbohydrate digestion and absorption, 
 but when a large amount is given for a protracted period diges- 
 tive disturbances are apt to ensue. As fat is usually only given 
 in five forms (butter, cream, cheese, and animal, or vegetable, fat), 
 and there is a strong repugnance on the part of many individuals 
 to fatty foods in any form, considerable skill is required to obtain 
 variety, and to make it palatable, when more than the average 
 amount is included in the diet. Experiments by Zuntz and 
 Heineman have shown that there is very Httle difference in the 
 efficiency of the body as a machine whether carbohydrates or fats 
 are taken. Zuntz, for instance, found that in one experiment 
 each kilo-gram-metre of work Avas accompanied by the hberation 
 of 9-39 calories on a fat diet, and 10-37 calories with carbohydrate 
 food. It is a matter of common experience, however, that muscular 
 fatigue is delayed and more work is done when carbohydrates are 
 available. The marked habihty of diabetics to muscular fatigue 
 also points in the same direction. Experimental confirmation of 
 the intimate relation between a deficiency of carbohydrate and 
 fatigue in voluntary muscle is suppHed by the observations of Lee 
 and Harrold, who found that the muscles of a cat, from which the 
 readily combustible sugar had been swept out by treatment with 
 
PERSISTENT GLYCOSURIA 293 
 
 phlorhidzin, contracted only 200 to 400 times a minute on elec- 
 trical stimulation, instead of 800 to 1000, as they should have done. 
 Moreover, the curves of contraction resembled those of normal 
 muscle in the late stages of fatigue. The muscles of control animals, 
 which had been given phlorhidzin for four days and then re- 
 ceived 50 grams of dextrose, were examined after an interval of 
 eight hours, and it was found that their muscles gave 650 contrac- 
 tions a minute, and that the first hundred were quite normal in 
 character, thus showing that the results of the first experiment 
 were not due to any poisonous action of the drug. 
 
 Alcohol. — ^The nutritive value of alcohol has been the subject 
 of much discussion. According to Atwater and Benedict, small 
 quantities can be used in the economy in the place of isoclynamic 
 quantities of carbohydrate or fat, but large doses, according to 
 Miura, increase the waste of tissue protein. One gram of alcohol 
 has a heat value of 7 calories, so that 30 to 60 grams (1 to 2 oz.) 
 in the form of whisky, or Rhine wine, yield 270 to 540 calories, 
 and a litre of German beer, containing 3 to 4 per cent, of alcohol 
 and 5 to 6 per cent, of extractives, jdelds 450 calories, only haK 
 of which comes from the alcohol, however, the remainder being 
 derived from the dextrin and protein-Hke extractives. 
 
 AlcohoHc beverages are generally taken not for their food value, 
 but as stimulants, and for the sake of their flavour. As a stomachic, 
 alcohol is of httle use when gastric digestion is normal ; but it may 
 be of service when the secretory powers of the stomach are de- 
 fective, or to stimulate appetite, especially when considerable 
 quantities of fatty food have to be consumed. 
 
 Metabolic Chang-es in Chronic Glycosuria. — A diabetic 
 patient is in much the same position as a person with an insufficient 
 food supply, for although the food is there his tissues are unable 
 to make use of it in the normal way. 
 
 In mild cases, where the sugar disappears from the urine when 
 carbohydrates are cut out of the diet, and where the patient is 
 still able to make use of protein sugar, the metaboUsm of proteins 
 is not different from that of a person hving on a diet of meat and 
 fat. Lusk has shown that if the conditions under which a diabetic 
 lives be imitated in a healthy man by diminishing the carbohydrates 
 of his food by an amount equivalent to the sugar excreted in the 
 urine of the patient with whom he is being compared, protein 
 destruction in the healthy individual is the same as in the diabetic. 
 
 In severe cases the conditions are more comphcated. While 
 in the healthy organism, and in mild cUabetes, the supply of car- 
 bohydrate derived from protein metaboHsm can be made use of 
 
294 GLYCOSURIA 
 
 to protect the tissues from further destruction, in the severe type 
 this is not the case, for when the j)rotein sugar is withdrawn from 
 the tissue cells there is a large increase in protein metabolism., and 
 consequently in the nitrogen output in the urine. Thus in a case 
 reported by Mendel and Lusk, it was found that the ingestion of 
 broths containing 7*7 grams of nitrogen was followed bj^ an ehmi- 
 nation of 21-7 grams of nitrogen in the urine, or a loss of body 
 nitrogen of approximately 14 grams, and that nitrogen equihbrium 
 could only be maintained by giving 27 grams of protein nitrogen in 
 the food. The abnormal tissue waste that occurs in such cases 
 is commonly, although without definite proof, attributed to the 
 action of toxines, and is consequently termed " toxogenic proteid 
 disintegration." 
 
 Another result of the abnormal tissue destruction is the appear- 
 ance in the urine of an excessive quantity of purin bodies, which are 
 derived from the nuclear substance (endogenous pm-in bocUes). In 
 some cases an increase of 50 per cent., or even 100 per cent., has 
 been met with, 0*25 to 0*30 grams of purin nitrogen being excreted 
 in the urine dailJ^ (Von Noorden.) 
 
 Proteins.— A. large amount of iniormation is now available 
 with regard to the relation between the urinary nitrogen, and the 
 sugar elimination, in the fasting and meat-fed diabetic organism. 
 Since protein contains about 16 per cent, of nitrogen and 50 to 
 55 per cent, of carbon, 100 grams of protein material could theoreti- 
 cally give rise to about 130 grams of sugar and 16 grams of nitrogen, 
 so that if both were completely eliminated in the urine a dextrose 
 to nitrogen ratio (D : N) of about 8 : 1 would be obtained. Accord- 
 ing to the observation of Rubner, however, 1 gram of protein nitrogen 
 supphes 18-6 available calories, and, since 1 gram of dextrose 
 furnishes 3-74 available calories, 4-97 grams of dextrose "svill be 
 required to yield the same amount of energy. On this finding 
 we should expect that at the most each gram of nitrogen in the 
 urine can be accompanied by a Httle less than 5 grams of dextrose 
 when the power to utilise sugar is completely lost. The experi- 
 ments of Minkowski and others have shown that in depancreatised 
 dogs on a protein diet, 2-8 grams of dextrose are excreted in the 
 urine for each gram of nitrogen ehminated, and the same dextrose 
 to nitrogen ratio has been found in the urines of a variety of animals 
 rendered diabetic mth phlorhidzin, or by removal of the pancreas. 
 Assuming that these animals were incapable of using sugar, and 
 that the ratio D : N : : 2-8 : 1 represents the amount of sugar 
 derived from the cleavage of protein within the body, it would 
 appear that rather less than 45 per cent, of the jDrotein molecule is 
 
PERSISTENT GLYCOSURIA 295 
 
 converted into dextrose during metabolism. This follows from 
 the fact that 1 gram of nitrogen in the urine corresponds to a 
 destruction of 6-25 grams of protein in the organism, hence : — 
 1 gram nitrogen := 6-25 grams protein ^2-8 grams dextrose 
 
 ■ •■ ^•^^^QQ = 44-8 per cent. 
 6-25 
 
 It has been shown by Liithje, however, that if depancreatised dogs 
 
 fast complete^ after the operation, the excretion of dextrose in the 
 
 urine ceases, while the percentage of sugar in the blood returns to 
 
 the normal ; hence the bodies of these animals appear to be capable 
 
 of utiHsing a limited amount of sugar. Reilly, Nolan, and Lusk, 
 
 working with dogs made diabetic with phlorhidzin, discovered a 
 
 higher dextrose to nitrogen ratio than was met with by Minkowski, 
 
 D : N : : 3-65 : 1, that would show a yield of slightly over 58 per cent. 
 
 of sugar from protein : — 
 
 3-65x100 _ ^ 
 
 — = oS'i per cent. 
 
 This higher dextrose to nitrogen ratio, 3-65 : 1, was also found in 
 some severe cases of diabetes when the patient was given a diet of 
 meat and fat. They proved that in such cases the relationship 
 between the nitrogen and sugar of the urine is constant, no matter 
 how much protein is given, and is in no way dependent upon varia- 
 tions in the amount of fat in the diet. 
 
 The reason for the existence of the two ratios 2-8 : 1 and 3-65 : 1, 
 each occurring when there is complete intolerance for carbohydrates, 
 is not clear. Mendel and Lusk suggest that the sugar in the blood 
 exists in two forms combined with colloid material. The one, 
 a-colloid dextrose, corresponds to the amount of sugar represented 
 by the lower ratio, or 45 per cent, of the protein ; while the other, 
 ^-colloid dextrose, represents the additional 13-6 per cent, of the 
 protein when the higher ratio is present. The dextrose to nitrogen 
 ratio found would then depend upon whether the /3-dextrose were 
 utilised or not. It is also possible that sugar production varies 
 under different circumstances, and that the organism may be able 
 to form sugar from a certain class of protein decomposition products 
 at times only, and under certain conditions, when the higher ratio 
 will be found. 
 
 Reilly, Nolan, and Lusk attempted to answer the question as to 
 when protein sugar becomes available for use in the organism, by 
 giving a fasting phlorhidzinised dog 500 grams of meat and collect- 
 ing the urine at regular intervals. They found that the fasting 
 relation between the dextrose and nitrogen changed immediately 
 on the ingestion of the meat, more sugar being eliminated than cor- 
 responded to the nitrogen in the urine in the early hours, and less 
 
296 GLYCOSURIA 
 
 in the later j)eriods. It was therefore evident that the sugar 
 elimination took place decidedly before that of the nitrogen, a 
 fact that should be borne in mind when collecting the urine of 
 diabetics for analysis. To allow of the complete ehmination of the 
 nitrogen corresponding to the sugar of a twenty-four hours' sample, 
 it is advisable that the collection should terminate at an early 
 morning hour, before smy food has been taken. 
 
 The energy value of protein to the diabetic organism varies, of 
 course, with the proportion of protein sugar that is excreted vm- 
 utihsed in the lurine. Takuig an extreme case, where a dextrose to 
 nitrogen ratio of 3-65 : 1 exists, we have seen that 52-5 per cent, 
 •of the available energy is lost as dextrose in the urine ; we have 
 also seen that, according to Rubner, 28-5 per cent, of the energy 
 ■of meat protein is Hberated as free heat that cannot be utihsed in 
 the life processes of the tissues, so that there only remains a balance 
 of 19 per cent, that can be made use of for the vital processes of 
 the body. To compensate for this great waste of energy there is 
 a great increase in protein metabolism, hence in severe cases of 
 diabetes 30, 40, or even 50 grams of nitrogen may appear in the 
 urine in the twenty-four hours. The administration of a large 
 amount of meat in severe cases of diabetes is hable not only to 
 increase protein metabolism and the output of sugar, but also to 
 interfere with the fixation of glycogen by the liver and thus tend 
 to promote acidosis, so that a " carbohydrate-free " diet should 
 never be given, at any rate for more than a few days, without very 
 careful consideration. 
 
 Most observers are agreed that vegetable proteins are less 
 harmful in severe cases of diabetes than those of animal origin, 
 and that by substituting them in part, or completely, for meat in 
 the diet the glycosuria is more readily controlled, and the danger 
 of acidosis is more easily averted. Various preparations of vege- 
 table protein are now on the market, and may be used for this 
 purpose. The least expensive and most generally useful vegetable 
 protein food is the soj^ bean {Glycine hispida). This contains 
 38-5 per cent, of protein, and 20 per cent, of fat calculated on a 
 water-free basis, and is almost starch-free, a phenomenon which 
 is said to be due to the presence of a diastatic ferment capable 
 of converting any starch formed into sugar (two-thirds) and dex- 
 trine (one-third). It may be eaten as a vegetable, after being well 
 soaked in water, as a salad, in soups, or the flour may be made 
 into muffins, or cakes, &c., but it must be remembered that it 
 contains from 20 to 40 per cent, of carbohydrate. 
 
 Fats. — It will be recalled that, though fats do not protect the 
 tissues from wasting so effectuall}^ as carbohydrates, the healthj^ 
 
PERSISTENT GLYCOSURIA 297 
 
 organism can be maintained in nitrogenous equilibrimn on a diet 
 of meat and fat alone, and since the power to utilise fats is not 
 apparently affected in the earher stages of persistent glycosuria, 
 fatty foods can be used in place of carbohydrates to meet the energ}^ 
 requirements of the body, and spare the proteins from excessive 
 waste. As the condition progresses, however, fat metaboHsm is 
 more or less impaired, and we see increasing wasting, lipsemia, and 
 acetoneemia. A limitation of the fat in the diet, and its partial 
 replacement by carbohydrate, is then advisable, even though the 
 glycosuria may be thereby increased. Schwarz states that the 
 lower fatty acids increase the amount of acetone bodies more than 
 the higher members of the series; butter, which contains more or 
 less butyric and other lower fatty acids, causes, therefore, more 
 marked acetonuria than animal fats that contain stearin and 
 palmitin. The action of oil depends upon the nature of the con- 
 stituent fatty acids, those containing the oleic groups having the 
 least effect. Joshn considers that Schwarz's results may have been 
 due to lack of absorption, for he found that oleic acid may nearly 
 double the acetonuria in a fasting man, while butyric acid had no 
 effect on the output of acetone bodies. 
 
 According to v. Noorden, there is in most cases of diabetes an 
 impairment of the power of the body to synthesise fats from carbo- 
 hydrate, and it is partly to this that the wasting seen in severe 
 -cases is to be attributed. In some individuals, although sugar 
 utiHsation is defective, the synthesis of fats from carbohydrates 
 is not interfered with, so that the tissues being richly bathed with 
 sugar build up excessive quantities of fat, and obesity results. So 
 long as the power to form fat from sugar is unaffected glycosuria 
 •does not occur, but eventually there is, as a rule, a gradual impair- 
 ment of the synthetic process, and sugar is consequently excreted 
 in the urine. At first it may appear only intermittently and when 
 a large amount of carbohydrate is consumed, but later it is excreted 
 regularly, and we then have the common form of '' diabetes in the 
 obese.'' 
 
 A mixed diet of protein and fat will not ordinarily increase the 
 amount of sugar in the urine in diabetes, but in some cases the 
 quantity of sugar ehminated appears to be greater than can be 
 accounted for by the destruction of the protein of the food and 
 tissues, as measured by the nitrogen excretion : it would therefore 
 seem probable that under these conditions sugar is being formed 
 from fat. Cremer found that the administration of glycerine, one 
 of the cleavage products of fat, will increase the output of sugar in 
 the urine, but there is no direct evidence that the higher fatty acids 
 can be converted into carbohydrates, although v. Noorden main- 
 
298 GLYCOSURIA 
 
 tains that this must occur, since, in some cases of diabetes, he has 
 found a larger amount of sugar in the urine than would be accounted 
 for from other known sources, including glycerine. Lusk states 
 that sugar is not formed from fat in phlorhidzin diabetes in animals, 
 and that if such a formation occurs in diabetes in man, it must be 
 due to a quahtative alteration in the metabohsm in rare and special 
 cases. He and Mendel investigated the metabohsm of dogs with 
 phlorhidzin glycosuria, starving, and after meat ingestion, and 
 found that the latter doubled the protein metabolism and caused 
 a fall in fat metabohsm, as it would do in a normal animal. The 
 dextrose to nitrogen ratio remained unchanged, showing that the 
 amount of sugar was not influenced by the quantity of fat made 
 use of. 
 
 Carbohydrates. — Defective carbohydrate metabohsm is the essen- 
 tial featmre of the diabetic state, and there is always a more 
 or less marked difficulty in dealing with dextrose and substances, 
 such as starch, that give rise to dextrose in the processes of diges- 
 tion. We have seen, however, that the structure and configura- 
 tion of a sugar determine whether it shall, or shall not, be attacked 
 and broken down by living cells, so that it would appear possible 
 that the tissues of the diabetic organism might be able to make use 
 of sugars of different composition, or space arrangement, to dextrose. 
 Minkowski found that although the livers of depancreatised dogs 
 could not form gh^cogen from dextrose, the administration of levu- 
 lose resulted in glycogen being deposited, and caused a reduction 
 in protein metabohsm. also that after 100 to 200 grams (10 to 20 
 grams per kilo) of levulose had been given to a depancreatised dog 
 only half reappeared in the urine, and of this 90 per cent, was 
 excreted as dextrose, 10 per cent, at the most being passed as levu- 
 lose. Observations on diabetic patients bj^ Bouchardat, Kiilz, 
 V. Noorden, and others, have shown that in mild cases of diabetes 
 levulose causes less sugar to appear in the urine than dextrose, and 
 that there is at the same time a rise in the respiratory quotient. 
 The power to utihse levulose ]30ssessecl by such individuals is 
 limited, however, and is soon overtaxed, so that its continuous 
 administration after a time is followed by almost as marked glycos- 
 uria as would be produced hy an equal amount of dextrose. In 
 most severe cases of chabetes a varying proportion of levulose is 
 found, along with the dextrose, in the urine, and in such cases the 
 administration of levulose causes almost as much sugar to appear 
 in the urine as when dextrose or starch is given. Mendel and 
 Lusk investigated the urine of a severe case of diabetes, and found 
 that when 100 grams were taken an increase in the sugar excretion, 
 corresponding to 80 per cent, of the ingested sugar, took place. 
 
PERSISTENT GLYCOSURIA 299 
 
 and that it had no effect on protein metabohsm. The fact that 
 diabetics can frequently utiUse levulose, although the glycogen 
 formed from it must later be converted into dextrose, has been used 
 as an argument in favour of the view that it is not as sugar that 
 the cells make use of carbohydrates, but as glycogen. Neubauer 
 has shown, however, that the power to form glycogen from levu- 
 lose, as opposed to dextrose, is not a characteristic of diabetes, but 
 is also seen in phosphorus poisoning. Arguing from the known 
 relation of alimentary levulosuria to diseases of the liver and 
 Minkowski's experiments on animals, it has been suggested that a 
 failure on the part of a diabetic to make use of levulose points to the 
 condition being dependent upon, or associated with, hepatic disease. 
 
 Pentoses have been suggested as a substitute for dextrose in 
 diabetes, since Cremer and others have shown that a pentose, such 
 as rhamnose, can be utilised by animals and spare an isodynamic 
 equivalent of fat. Lindemann and May found that a healthy man 
 could make use of 90 grams of rhamnose, but when this was given 
 to a diabetic, whose urine had previously been sugar- free, glycos- 
 uria was produced. Von Jaksch found that rhamnose, arabinose, 
 and xylose all increased the glycosuria, tended to raise protein 
 metabolism, and caused diarrhoea when administered to severe 
 cases of diabetes. 
 
 It has long been known that diabetics appear to assimilate 
 some forms of starch much better than others, but that there is a 
 very considerable individual variation in this respect, so that a 
 dietary that will suit one will prove unsatisfactory for another. 
 It has also been found that better results are obtained when one 
 source of carbohydrate is employed alone than when several are 
 mixed together. These differences depend probably in part upon 
 the ease with which they are digested and assimilated, those being 
 tolerated best which take the longest to absorb, and consequently 
 pass only slowly into the blood. Many of the green vegetables, 
 such as cabbage, lettuce, spinach, celery, asparagus, cucumber, &c., 
 and some fruits, such as melon, pineapple, strawberries, rhubarb, &c., 
 contain under 5 per cent, of carbohydrate combined with a con- 
 siderable amount of indigestible cellulose, and as, under ordinary 
 circumstances, a portion even of this starch is not inverted and 
 absorbed, restriction of the diet to such substances is, as Blum 
 j)oints out, almost equivalent to fasting, so far as the tissues are 
 concerned. The way in which these vegetables are prepared for 
 the table exerts some influence on the amount of carbohydrate 
 absorbed, and serves to explain the fluctuations in the amount of 
 sugar excreted after their use. The least harmful source of sugar 
 in each particular case can only be learnt by trial and careful 
 
300 GLYCOSURIA 
 
 analysis of the urine ; at the same time an attempt should be made 
 to discover if the sugar-yielding food can be more safely given at one 
 period of the twent^'^-four hours than another, for such often proves 
 to be the case. 
 
 In his report on the results of monotonous feeding of prisoners 
 Bar says that the continual serving of one kind of food, always in 
 the same way, causes loss of appetite, vomiting, flatulence, and 
 diarrhoea, or obstinate constipation. Such results are not infre- 
 quently met with in diabetics who are restricted to a diet consisting 
 almost entirely of meat, fat, and the so-called diabetic breads, &c. 
 It has also to be remembered that stringent restriction of the diet 
 usually leads to cUsobedience and evasion, which is more harmful 
 than a httle extra licence. Bj^ allowing the patient as much 
 vegetable as possible, and particularly the coarser kinds containing 
 a low percentage of carbohydrate, we not only give a much greater 
 variety of food, and so ensure a more ready acquiescence in our 
 directions, but also furnish a vehicle bj^ which a much larger quan- 
 tity of fatty food can be taken "udth comfort than would other- 
 wise be the case. The alkahne salts contained in vegetables and 
 fruits no doubt assist in the treatment of diabetes, for they aid 
 in the neutraUsation of the acids formed as the result of a highly 
 protein diet and faulty metabolism . 
 
 Diets based upon the different effects of varioiis starchy foods 
 have from time to time been introduced. Mosse has recommended 
 potatoes, as they seemed to diminish the sugar and poljoiria. 
 Although, in some cases, this dietarj^ appears to exert a beneficial 
 effect, partly owing to the causes mentioned, and particularly to 
 its richness in potash salts, the treatment as recommended by 
 Mosse is now rarely used. It may be mentioned here that a small 
 amount of potato, thinly sliced, and cooked by being plunged into 
 boihng fat, is often a useful addition to the diet of a diabetic, for 
 one small potato prepared in this way will fill a tmreen, give a 
 pleasant change, and have a food value of about 2600 calories per 
 pound. The best known and most frequently used of these 
 special chets is the oatmeal " cure " introduced by v. Noorden. 
 This consists in the daily administration of 200 to 250 grams (6 to 8 
 oz.) of oatmeal, 200 to 300 grams (6 to 10 oz.) of fat, in the form 
 of butter, and 100 grams (3| oz.) of protein, in the shape of three 
 or four eggs or 50 to 100 grams of a vegetable protein, prepared 
 in the form of soup or porridge, and given at frequent intervals, 
 with an occasional allowance of black coffee, or tea with lemon- 
 juice, wine, or cognac. The oatmeal is prepared by placing it in 
 three parts of water, sHghtly salted, and thoroughly cooking it for 
 at least six hours ; while stiU hot it is strained through a sieve. 
 
PERSISTENT GLYCOSURIA 301 
 
 The coarse covering of the kernel that remains is rejected, and the 
 butter is then stirred into the hot porridge that has passed through. 
 After three or four days of such a diet there succeed one or two 
 days on which only green vegetables are allowed, then the oatmeal 
 cure is resumed. The treatment is continued in a similar manner 
 for several weeks. As a rule four or five courses are required. 
 According to v. Noorden, it is advisable to precede the oatmeal 
 treatment by a few days of restricted, or vegetable, diet, but 
 Crofton states that in his experience this is unnecessary'-, and is 
 in fact rather detrimental than otherwise. At the commencement 
 of the cure the glycosuria may increase, but in favourable cases it 
 soon diminishes and may disappear altogether while the patient 
 is on the oatmeal diet, and, in some cases, the urine may remain 
 sugar-free even when the patient returns to a restricted diet. The 
 resumption of a general diet, or of a meat-fat- vegetable diet, must 
 however be gradually made, and for a long time, often for months, 
 after the exclusive oatmeal feeding has been stopped, oatmeal 
 should still remain the only carbohydrate taken. A mixture of 
 carbohydrate should be carefully avoided. Animal protein should 
 also be resumed with care ; for even in favourable cases, but more 
 especially in those where the excretion of a sugar and acetone was. 
 only reduced, the addition of animal proteins is promptly followed 
 by increased glycosuria and an increased excretion of acetone bodies. 
 Crofton recommends that, as a precautionary measure against 
 acidosis, from two to four teaspoonfuls of sodium bicarbonate 
 alone, or mixed with equal parts of magnesia usta, should be given 
 daily throughout the oatmeal treatment. Von Noorden found that 
 of three hundred and ten patients who had taken a systematic oat- 
 meal cure only sixty-five were not benefited, but that in thirty- 
 five the condition was aggravated. Blum states that he obtained 
 excellent results with thirty-five cases, the success of the cur& 
 depending on the intensity of the disease. He considers that, 
 while the ordinary dietetic methods are equally effectual in mild 
 cases, even with these the oatmeal treatment is more rapid, is more 
 easily managed, and is accepted better by the patient. Witk 
 severe cases, where there is acidosis, he considers that the oatmeal 
 cure is of great use, but insists that not more than 75 grams of 
 oatmeal should be given for a few days at first, and then a vegetable 
 day should be prescribed. The most brilliant results were obtamed 
 by Crofton in children, particularly when the oatmeal ciure was. 
 started as soon as possible after the glycosuria was chscovered. 
 Cases of long duration did not fare so well. He considers that in 
 adults the treatment is worse than useless when they are of a mild 
 type and can still utihse some carbohydrate, but that when the- 
 
302 GLYCOSURIA 
 
 sugar does not disappear from the urine, even after some time on a 
 carbohydrate-free diet, the effect is often favourable, although the 
 results are not as satisfactory as with many children. As a rule 
 the improvement is only temporary, and is not maintained when 
 the patient returns to an ordinary restricted diet. Most autho- 
 rities are agreed that the oatmeal cure when carefully controlled, 
 and when employed in properly selected cases, is a useful and often 
 indispensable adjuvant in the treatment of diabetes, but that when 
 it is applied promiscuously it is distinctly dangerous. It has been 
 pointed out by Minkowski that the oatmeal diet causes a tendencj- 
 to water retention within the organism, and that oedema and 
 partial retention of urine sometimes result, these symptoms dis- 
 appearing, however, when it is discontinued. 
 
 No very satisfactory explanation of the beneficial effects that 
 the oatmeal cure undoubtedly exerts in some cases of diabetes has 
 yet been advanced. Some consider that it is not due to any specific 
 action of the oatmeal, but depends upon the lowering nature of the 
 diet and the restriction of protein, more particularly the absence 
 of albuminous substances of animal origin, which not only does 
 away with a source of sugar production, but also reduces the meta- 
 bolism as a whole. Blum states that wheatmeal given in a similar 
 wa}^ to oatmeal is just as efficacious, and Griind found that barley- 
 meal gives similar therapeutic results. The experiments of RoUy 
 also suggest that there is no essential difference between these 
 various forms of starch. Magnus-Levy, while he asserts that the 
 advantages of oatmeal are mostly on the negative side, and are 
 largety due to absence of meat, admits that oatmeal starch appears 
 to have some special feature which renders it superior to the starch 
 of other grain foods. This might be due to a difference in con- 
 stitution, or to the presence in the oat grain of some substance 
 that favours the utilisation of its starch. He found that pure 
 oat starch, freed from other substances contained in oats, given 
 "with eggs and sanatogen, was just as efficacious as ordinary oats, 
 so that the superiority of oatmeal in the treatment of diabetes must 
 be ascribed to a pecuharity of the starch itself. It has always been 
 assumed that starch, whatever its origin, is always the same sub- 
 stance, but if the results of these observations be confirmed, it 
 would appear that this is not the case, and that the varying effects 
 of other starchy foods in glycosuria may be partly due to differences 
 in composition as yet unrecognised, Naunyn has suggested that 
 the effects of oatmeal may be due to its not being utiHsed as sugar, 
 and, with Magnus-Levy, beHeves that oat starch undergoes some 
 pecuhar transformation through the action of micro-organisms in 
 the ahmentary tract whereby it is converted into fermentation 
 
PERSISTENT GLYCOSURIA 303 
 
 products rather than into simple sugars, as is ordinarily the case 
 with starch. Falta has expressed the opinion that the action of 
 oatmeal is specific and is not dependent in any way upon a re- 
 striction, or difference in kind, of the protein content of the diet. 
 A series of experiments by Hunt have shown that the resistance 
 of mice to the poisonous effects of aceto-nitrile is very much in- 
 creased by an oatmeal diet, probably owing to a specific effect 
 exerted on the thyroid gland, and it is consequently possible that 
 it may be by some such action that it j^roduces its effect on meta- 
 bolism in diabetes. 
 
 The recent discovery of the effects of the removal of the 
 cuticle of rice in producing beri-beri, and the influence on the 
 disease of giving the pericarp of the cereal, suggest that there 
 may be some part of the potato, of oats, and of rice which has a 
 specific action on metabolism, and which if it can be isolated from 
 the cereals or from the preparations of diastase, may prove to have 
 specific action in the treatment of disease. Von Noorden has 
 suggested that the diminished glycosuria that results from the 
 oatmeal treatment may depend upon an alteration in the permea- 
 bility of the kidneys, and experiments carried out by Barrenscheen 
 on the delay in the excretion of milk-sugar injected intravenously 
 into non-diabetic subjects caused bj' an oatmeal diet tends to 
 support this hypothesis. 
 
 Strauss and others have advocated the use of vegetables rich 
 in inuhn, such as artichokes, dandehon, the roots of black viper's 
 grass, and edible species of the sunflower family, in diabetes. It 
 is claimed that the absorption of inulin from such vegetables takes 
 place very slowly, and that it is a form of carbohydrate that is well 
 borne by many diabetics. The value of inuhn as a source of energy, 
 has, however, been questioned by Lewis, who concluded that the 
 quantities that can be utihsed are insignificant. His experiments 
 showed that inulin taken by the mouth is partly converted by the 
 acid of the gastric juice into levulose, and it is only this that can 
 be made use of, for any that leaves the stomach unchanged under- 
 goes bacterial decomposition in the intestine with the formation of gas, 
 but no sugar, and that the inuhn that escapes the bacterial change 
 appears in the faeces. The extent to which inulin can be utihsed de- 
 pends therefore on the acidity of the gastric juice, and to a certain 
 extent on the diet, and consequently varies very much in different 
 individuals and in the same person under different conditions. 
 
 The Energy Requirement in Diabetes. — In consequence of 
 the serious loss of strength and marked emaciation met with in 
 severe cases of diabetes it might be thought that larger quantities 
 
304 GLYCOSURIA 
 
 of food are necessary than in health, and the fact that there is 
 often a voracious appetite would appear to confirm this. The 
 experiments of most investigators have shown, however, that the 
 energy requirement is httle, if at all, altered, for the diabetic con- 
 dition does not involve a decrease in the quantity of energy pro- 
 duced, but only an alteration in its source. The early experiments 
 of Pettenkofer and Voit showed no change in the metabohsm in 
 diabetes from the normal. Rubner, using phlorhidzimsed dogs, 
 found that the heat production was increased by 7 per cent, when 
 glycosuria was produced, but as the protein metabohsm was raised 
 more than threefold, he attributed this alteration in energy pro- 
 duction to the specific djmamic action of the increased protein 
 metabohsm. Falta could find no evidence that the energy require- 
 ment is increased in diabetes, and Du Bois and Veeder showed 
 that the amount of energy required by diabetics is approximately 
 34 calories per kilogram of body-weight, which is not greater than 
 for a healthy man. Benedict and Joslin, as the result of a series 
 of metabohc experiments on thirteen cases representing various 
 types of glycosuria, state that the heat production may be 15 per 
 cent, above the normal. Some observers have contended, on the 
 other hand, that in severe diabetes there is a lowered energy require- 
 ment. Kohsch, for instance, states that 25 calories -per kilogram may 
 lead to an increase in body-weight, and that in some instances as 
 little as 20 calories are sufficient to maintain the patient's condition. 
 He strongly advises a minimum amount of food, and claims that 
 this offers the best hope of success in the treatment of severe cases. 
 More than a hundred years ago Prout pointed out the great 
 advantage of limiting the quantity of food in diabetes, and it is 
 now generally agreed that high-feeding is injurious, and that in most 
 cases the best results are obtained by arranging the diet so that 
 the normal 34 to 35 calories per kilogram of body- weight are suppHed, 
 due allowance being made for the potential energy lost in the sugar 
 contained in the urine. Some severe cases of diabetes, however, 
 do best when the minimum amount of food that will maintain the 
 body-weight is given. As a rule, not more than 100 to 120 grams 
 of protein should be taken in the day, and this amount should be 
 reduced on the appearance in the urine of signs of acidosis. 
 
 BIBLIOGRAPHY 
 
 Atwater, ATuer. Journ. of Physiol., 1904 ; Proc. Amer. Phys. Soc, 
 Atwater and Benedict, Memoirs of the Nat. Acad, of Sci., 1902. 
 Atwater and Bryant, Report Storr's Agricult. Expt. Station, 1899. 
 Barrenscheen, Biochem. Zeit., 1912. 
 
 XXX. 
 
PERSISTENT GLYCOSURIA 305 
 
 Benedict and Joslin, quot. Lusk, Journ. Amer. Med. Assoc, 1910. 
 
 Blijin, Semaine Medicate, 1911. 
 
 Du Bois and Veeder, Arch. f. int. Med., 1910. 
 
 Chittenden, Physiol. Economy of Nutrition, 1894. 
 
 Cremer, Zeit. f. Biol., 1901 ; MiXnch. med. Woch., 1902. 
 
 Crofton, Journ. Amer. Med. Assoc, 1909. 
 
 Faita, Wien klin. Woch., 1909 ; Germ. Cong. Int. Med., 1911. 
 
 Geelmuyden, Zeit. f. phys. Chem., 1897. 
 
 Grund, Zeit. J. phys. Chem., 1902 ; Germ. Cong. Int. Med., 1911. 
 
 Hunt, Journ. Amer. Med. Assoc, 1911. 
 
 Von Jaksch, Deut. Arch. f. klin. Med., 1899. 
 
 Johansson, Billstrom, and Heyl, Skan. Arch. /. Physiol., 1904. 
 
 JosHn, Journ. of Med. Res., 1904. 
 
 KoHsch, Zeit. f. phys. u. diat. Therap., 1908. 
 
 Landergren, Skan. Arch. f. Physiol., 1903. 
 
 Lee and Harrold, Amer. Journ. of Physiol., 1900. 
 
 Lehmann and Voit, Zeit. f. Biol., 1900. 
 
 Lewis, Journ. Amer. Med. Assoc, 1912. 
 
 Lindemann and May, Deut. Arch. f. klin. Med., 1896. 
 
 Lusk, Zeit. f. Biol., 1890, 1901 ; The Science of Nutrition, 1906. 
 
 Liithje, Miinch. med. Woch., 1902. 
 
 Magnus-Levy, Arch. f. Physiol., 1904 ; Arch. f. exp. Path. u. Pharm., 
 
 1899; Berl. klin. Woch., 1911. 
 Meissel and Strohmer, Sitzungsber. a. k. Acad. d. Wissensch., 1883. 
 Mendel and Lusk, Amer. Journ. Physiol., 1903 ; Deut. Arch. f. klin. 
 
 Med., 1904. 
 Minkowski, Arch. f. exp. Path., 1893 ; Germ. Cong. Int. Med., 1911. 
 Miiira, Zeit. f. klin. Med., 1892. 
 Munk, Arch. f. Physiol, 1891. 
 
 Neubauer, Miinch. med. Woch., 1905 ; Arch. f. exp. Path. u. Pharm., 1909. 
 Neuberg, Ergeb. d. Physiol., 1904. 
 Von Noorden, Herter Lectures on "Diabetes," 1905; Handb. d. Ernah- 
 
 rungstherapie, v. Leyden, 1904 ; Diabetes, 1905. 
 Pettenkofer and Voit, Zeit. f. Biol., 1867. 
 
 Reilly, Nolan, and Lusk, Amer. Journ. of Physiol., 1895, 1898. 
 Roily, Deut. Arch. f. klin. Med., 1912. 
 Rosenfeld, Ergeb. d. Physiol, 1903. 
 Rosenheim, Arch. f. Physiol, 1891. 
 Rubner, Zeit. f. Biol, 1885 ; Gesetze d. Energieverbrauchs, 1902 ; v. 
 
 Leyden's Handbuch, 1903. 
 Satta, Hofmeister's Beitr., 1904-5. 
 Schulz, PfliXger's Arch., 1896. 
 Schwarz, Deut. Arch. f. klin. Med., 1903. 
 Sieven, Skan. Arch. f. Physiol, 1901. 
 Stiles and Lusk, Amer. Journ. Physiol, 1903. 
 Stohmann, Journ. of prakt. Chem., 1885. 
 Voit, Physiol, d. Stoffwechsels u. d. Ernahrung, 1881. 
 Voit and Korkunoff, Zeit. f. Biol, 1895. 
 Zuntz and Heineman, Pjiuger's Arch., 1900. 
 
 U 
 
CHAPTER IX 
 
 PEESISTENT GLYCOSURIA — TREATMENT AND PROGNOSIS 
 
 The Dietetic Treatment. — The aim of the modern dietetic 
 treatment of diabetes is to so balance the diet that as perfect 
 nutrition as possible is maintained, without there being undue 
 strain on the organs of metabolism in any direction. The ad- 
 ministration of carbohydrates in larger amounts than the defective 
 metabohc powers of the diabetic organism can deal with is un- 
 doubtedly injurious, but their permanent exclusion from the diet 
 is even more harmful, and is often quite unnecessary. A temporary 
 exclusion is, in some instances, advisable and beneficial, for not 
 only is the excess of sugar in the blood thereby diminished, so that 
 the cause of many of the troubles to which diabetics are liable is 
 lessened, or abohshed, but at the same time the metabolic powers, 
 like a released spring, often recover their tone to a surprising 
 extent, and the tolerance for carbohj^drates is materially increased. 
 Similar results are obtained in other cases by merely restricting 
 the carbohydrate intake, both in quantity and kind, until the 
 glj^cosuria is controlled. Some diabetic patients have a marked 
 susceptibility to proteins, which as we have seen may yield 45 to 
 60 per cent, of sugar, so that their glycosuria is decreased more by 
 diminishing the protein and allowing a certain amount of carbo- 
 hydrate in the diet, than by excluding the latter altogether. The 
 vegetable proteins are, as a rule, found to be less harmful than 
 those of meat, and consequently periods of vegetable feeding may 
 be advantageously introduced into the treatment. For the same 
 reason prepared vegetable proteins, &c., can often be used in place 
 of a corresponding amount of meat in the ordinary diabetic diet. 
 It must always be borne in mind that some diabetics never become 
 permanently sugar-free, and that an attempt to make them so is 
 Uable to give rise to serious consequences. Even in the most 
 severe cases there is some stored glycogen in the tissues, but if 
 this is used acidosis will quickly develop, and the power to store 
 further glycogen be lost or markedly diminished. It is therefore 
 better to maintain a fairly hberal supply of glycogen in the body 
 and a low sugar-content in the urine than to attempt to estabHsh 
 a sugar-free condition of the urine with the attendant risk of a 
 
 306 
 
PERSISTENT GLYCOSURIA 307 
 
 greatly depleted glycogen store in the tissues. No routine line of 
 treatment, however sound it may be in principle, can be apphed 
 to all cases. The metabolic powers of each patient must be experi- 
 mentally determined and a diet adapted to his requirements be 
 worked out, and not only so, but it must be adjusted from time to 
 time to meet the altering requirements of the case. That is to say, 
 we must treat the patient, not the disease. 
 
 In working out a diet the first point to determine is the gravity 
 of the case. For this purpose we must ascertain, first, the intensity 
 of the diabetes ; secondly, the presence and degree of any secondary 
 abnormahties of metabolism that may exist ; and, thirdly, whether 
 any disturbance of nitrogenous equihbrium is present. To this 
 end the urine passed in twenty-four hours is collected, starting and 
 finishing the collection at an early morning hour before food has 
 been taken. It is then carefully measured, and the total excre- 
 tion of sugar, nitrogen, and ammonia nitrogen is worked out. In 
 order that an accurate estimate may be formed of the power the 
 patient possesses of deahng with sugar, that is to say, the intensity 
 of the diabetes, it is necessary that the intake as well as the output 
 should be known. The latter is given by the urinary analysis. 
 Eor the former the patient must be put upon a diet of known com- 
 position, both as regards kind and quantity, such as that shown in 
 the following table (p. 308), for forty-eight hours before the collection 
 is made. This diet contains 102 grams of carbohydrate, but as, in 
 :severe cases of diabetes, dextrose can also be derived from protein, 
 the possible sugar from this source must also be taken into account 
 in calculating the total sugar value of the diet. Now the 114-2 grams 
 •of protein that it contains are equivalent to 18 -27 grams of nitrogen 
 /114-2\ 
 
 ( i\-9^ )' ^^^ since, at the most, each gram of nitrogen can be accom- 
 panied by 5 grams, or, according to Lusk, 3-65 grams, of dextrose, 
 the possible yield of sugar from this amount of protein would be 
 91-35 grams (18-27 x 5), or 66-68 grams (18-27 x 3-65), according to 
 whichever figure is accepted. We therefore find that the total 
 :sugar value of this particular diet is approximately 169 grams, or 
 possibly 193 grams. 
 
 (1.) The amount of carbohydrate contained in the diet as such 
 will influence the output of dextrose in the urine within a short 
 time after it has been taken, but the protein food does not necessarily 
 affect it at once — moreover, dextrose may be formed from the tissues, 
 so that the quantity of sugar of protein origin appearing in the 
 urine will be governed by the intensity of protein metaboMsm, and 
 does not necessarily bear a direct relation to the protein content 
 -of the food. Since the total nitrogen content of the urine can be 
 
308 
 
 GLYCOSURIA 
 
 Test Diet 
 
 Grams. 
 
 Oz. 
 
 
 Protein. 
 
 Fat. 
 
 Carbohy. 
 
 Calories. 
 
 225 
 
 8 
 
 Coffee (1 large cup) 
 
 
 
 
 
 20 
 
 f 
 
 Cream (1 tablespoonful) 
 
 0-7 
 
 's-i 
 
 0-7 
 
 54 
 
 50 
 
 2 
 
 Egg (1 average) .... 
 
 6-6 
 
 6-0 
 
 
 83 
 
 56 
 
 2 
 
 Bacon (smoked, weighed uncooked) 
 
 5-6 
 
 39-0 
 
 
 378 
 
 15 
 
 \ 
 
 Margarine 
 
 0-2 
 
 11-0 
 
 
 111 
 
 45 
 
 1-1 
 
 White bread (1 thick slice) . 
 
 3-9 
 6-3 
 
 0-5 
 
 22-3 
 
 112 
 
 28 
 
 1 
 
 Sardines (3 average) . 
 
 5-4 
 
 
 77 
 
 28 
 
 1 
 
 Lettuce or endive 
 
 0-3 
 
 0-1 
 
 "l-O 
 
 5 
 
 28 
 
 1 
 
 Tomato 
 
 |'4tablesp. ol. oil 
 
 0-4 
 
 1 ... 
 
 0-1 
 
 1-3 
 
 7 
 
 22 
 
 3 
 
 French dressing! 1 „ vinegar 
 
 16-0 
 
 
 148 
 
 
 
 (2 dessertsp.) j \ teasp. salt, 
 
 f 
 
 
 
 
 
 
 I pepper 
 
 
 
 
 18 
 
 _3. 
 
 4 
 
 Yolk of egg 
 
 2-9 
 
 6-0 
 
 
 68 
 
 96 
 
 3i 
 
 Mutton (roast) .... 
 
 25-2 
 
 21-7 
 
 
 300 
 
 43 
 
 1* 
 
 White bread .... 
 
 3-9 
 
 0-5 
 
 22-3 
 
 112 
 
 15 
 
 4 
 
 Margarine 
 
 0-2 
 
 11-0 
 
 
 111 
 
 225 
 
 56 
 
 8 
 2 
 
 Milk (1 glass) .... 
 
 7-3 
 
 8-8 
 
 11-0 
 
 157 
 
 White fish (cod, hake, sole, 
 
 
 
 
 
 
 
 whiting) ..... 
 
 12-() 
 
 0-2 
 
 012 
 
 50 
 
 28 
 
 1 
 
 Bacon 
 
 2-8 
 
 19-5 
 
 
 189 
 
 84 
 
 3 
 
 Beef (roast) 
 
 18-7 
 
 24-0 
 
 
 300 
 
 150 
 
 H 
 
 Potato (boiled, 1 medium-sized) 
 
 3-7 
 
 01 
 
 31-4 
 
 144 
 
 56 
 
 2 
 
 Spinach or tomato 
 
 1-3 
 
 0-2 
 
 1-5 
 
 9 
 
 15 
 
 4 
 
 Margarine 
 
 0-2 
 
 11-0 
 
 
 111 
 
 14 
 
 i 
 
 White bread .... 
 
 1-3 
 
 02 
 
 7-5 
 
 37 
 
 28 
 
 1 
 
 Cheddar cheese . . . . 
 
 7-7 
 
 10-3 
 
 1-2 
 
 1S3 
 
 14 
 
 4 
 
 Brazil nuts (4 large) . . 
 
 2-3 
 
 9-3 
 
 1-0 
 
 100 
 
 20 
 
 S 
 
 Cream (1 tablespoonful) 
 
 0-7 
 
 5-1 
 
 0-7 
 
 54 
 
 225 
 
 s 
 
 Coffee 
 
 
 
 
 
 
 114-2 
 
 211-1 
 
 102-1 
 
 2850 
 
 Carbohydrate 
 Protein sugar 
 
 114-2 _ 
 
 Total sugar value 
 
 102-1 
 
 grams 
 
 91-4 
 
 ,, 
 
 193-5 
 
 grams 
 
 taken as an index of protein metabolism of the body, the quantity 
 of nitrogen found to be present in any particular twentj^-four hours' 
 sample multiplied by 5 (or 3-65) will give the possible sugar that 
 can be derived from that source, and this, plus the amount of carbo- 
 hydrate known to have been consumed, will represent the sugar 
 that might appear in the urine if the power of metabolising carbo- 
 
PERSISTENT GLYCOSURIA 309 
 
 hydrate were completely lost. The relationship between the sugar 
 that might be excreted, and that which actually appears in the 
 urine, can be expressed in a single term by dividing the latter by 
 the former, but as this would give an inconvenient fraction, or 
 decimal, Falta has suggested that the result should be multi- 
 plied by 100, which will give the percentage of available sugar 
 that is ehminated unutihsed. This he terms the " Coefficient of 
 Excretion." 
 
 Falta's Coefficient of Excretion = 
 
 Total dextrose in urine x 1 00 
 
 (Total urinary nitrogen x 5) + food dextrose 
 If Lusk's factor is used the formula will be : — 
 
 Lusk's Coefficient of Excretion = 
 
 Total dextrose in ur ine x 1 00 
 
 (Total urinary nitrogen x 3-65) + food dextrose 
 
 SUghtly higher figures will of course be obtained when Lusk's 
 formula is used than when Falta's is employed ; but this is of Httle 
 practical importance, as the numbers are not absolute values, and 
 are only useful as an index of the severity of the diabetes by which 
 one case can be compared with another and the progress of the 
 same case can be accurately gauged from time to time, in much the 
 same way as the severity and progress of an ansemia can be esti- 
 mated and watched by blood-counts and haemoglobin estimations, 
 or the course of a fever can be followed with the thermometer. 
 
 (2.) Of the secondary disturbances of metabolism that occur 
 in diabetes the most important is the acidosis that incUcates im- 
 perfect oxidation of fats and proteins and that is the forerunner 
 of diabetic coma. The presence of aceto-acetic acid in the urine 
 shows that this stage in the progress of the case has been reached, 
 but for accurate work it is necessary that an approximate idea 
 at least should be obtained of the extent of the defect. A rough 
 measure of the degree of acidosis is given by the amount of sodium 
 bicarbonate that must be administered before the ferric chloride 
 reaction in the urine disappears, or the quantity in excess of 2 drams 
 required to make the urine alkaline. But the simplest waj^ is to 
 estimate the ammonia nitrogen in the urine. This is not strictly 
 accurate, but for routine clinical work it gives as near an approxi- 
 mation as is necessary. 
 
 (3.) According to Weintraud, acidosis is only dangerous when 
 it is accompanied by a disturbance of nitrogenous equilibrium, 
 that is to say, when more nitrogen is excreted than is absorbed, so 
 that a determination of the total urinary nitrogen is important, 
 
310 GLYCOSUPaA 
 
 not only for the estimation of the possible sugar output, but also 
 as a guide to the significance to be attached to the presence of the 
 acetone bodies andean excess of ammonia nitrogen in the urine. 
 The state of nitrogenous equihbrium is ascertained by balancing 
 the nitrogen- content of the food, estimated from diet tables, and 
 allowing 1 gram of nitrogen for each 6-25 grams of protein, as shown 
 above, against the nitrogen excreted in the urine, plus about 1 gram 
 to represent the unabsorbed nitrogen contained in the faeces. More 
 accurate results coiald of course be obtained by making complete 
 analyses of the food and excreta, but since a difference of 1 or 2 
 grams only is of no practical importance the method suggested is 
 found to be sufficiently reMable. 
 
 Having by these prehminary investigations obtained informa- 
 tion as to the type of case with which we have to deal, the next step 
 is to work out a diet suitable to the conditions found. For this 
 purpose it is necessary to ascertain (1) how much, and what kinds of 
 carbohydrate can be used with safety to supply as much as possible 
 of the energy required, (2) what amount of protein is needed to 
 maintain nitrogenous equilibrium, and (3) how much fat must be 
 added to the diet to make up the balance of the energy requirement. 
 The answers to these questions can only be obtained by metabohc 
 experiments controlled by analyses of the urine and careful regular 
 weighing of the patient. 
 
 If the patient has a low coefficient of excretion, the output of 
 ammonia nitrogen is not high, and nitrogenous equilibrium is not 
 seriously disturbed, he is at once put on a "' carbohydrate-free " diet 
 — that is to say, upon a diet that consists almost entirely of proteins 
 and fats, such as that given on the opposite page. 
 
 Such a diet is not strictly carbohydrate free, for it contains 
 9 grams, chiefly in the form of vegetables. If it is thought advisable 
 these may be omitted ; but they form a useful vehicle for a consider- 
 able amount of fat, and, as part of the raw starch is probably not 
 absorbed, the diet is as free from carbohydrates as it can con- 
 veniently be made and yet remains palatable. 
 
 After the patient has been forty-eight hours on this diet the 
 urine is collected and tested for sugar, acetone, aceto-acetic acid, &c., 
 and the total amount of sugar, ammonia nitrogen, and total nitrogen 
 determined. If the glycosuria has disappeared, or after a few more 
 days on the diet it does so, as it probably will, gradually increased 
 quantities of white bread — starting, say, with half an ounce (14 grams) 
 — are given each day until traces of sugar reappear. The amount of 
 bread which produces this effect incUcates the patient's limit of 
 tolerance for this particular form of starch. He is then again 
 placed on a carbohydrate-free diet for twenty-four hours, and his 
 
PERSISTENT GLYCOSURIA 
 
 " Carhohydrate-free " Diet 
 
 311 
 
 Grams. 
 
 Oz. 
 
 
 Protein. 
 
 Fat. 
 
 Carbohy 
 
 Calories. 
 
 225 
 
 8 
 
 Coffee (1 large cup) 
 
 
 ... 
 
 
 
 18 
 
 I 
 
 Cream ( 1 tablespoonf ul) 
 
 0-7 
 
 5-1 
 
 7 
 
 54 
 
 50 
 
 2 
 
 Egg (1 average) .... 
 
 6-6 
 
 6-0 
 
 
 83 
 
 56 
 
 2 
 
 Bacon (smoked, weigheduncooked) 
 
 5-6 
 
 39-0 
 
 
 378 
 
 14 
 
 1 
 
 2 
 
 Margarine ..... 
 
 0-2 
 
 11-0 
 
 
 103 
 
 28 
 
 1 
 
 Kalaribiscuits(Callard), Tbiscuits 
 
 17-0 
 
 7-5 
 
 
 107 
 
 18 
 
 I 
 
 Yolk of one egg .... 
 
 2-9 
 
 6-0 
 
 
 68 
 
 28 
 
 1 
 
 Lettuce or endive 
 
 0-3 
 
 0-1 
 
 I'-O 
 
 h 
 
 28 
 
 1 
 
 Tomato 
 
 0-4 
 
 0-1 
 
 1-3 
 
 7 
 
 22 
 
 i 
 
 „ . ,„ ('4tablesp. ol. oil 
 
 Dressing (2 h vinegar 
 
 dessertsp.) \^^^^^ ^^^^^^ 
 
 1 
 /■■■ 
 
 16-0 
 
 
 148 
 
 56 
 
 2 
 
 White fish (cod whiting, hake, 
 
 
 
 
 
 
 
 &c.) 
 
 12-0 
 
 0-2 
 
 0-12 
 
 50 
 
 28 
 
 1 
 
 Bacon (smoked, weigheduncooked) 
 
 2-8 
 
 19-5 
 
 
 189 
 
 28 
 
 1 
 
 Cheese (Cheddar .... 
 
 8-0 
 
 11-0 
 
 1-3 
 
 142 
 
 14 
 
 X 
 
 Margarine 
 
 0-2 
 
 11-0 
 
 
 103 
 
 28 
 
 \ 
 
 Kalari biscuits (Callard), 7 biscuits 
 
 17-0 
 
 7-5 
 
 
 107 
 
 18 
 
 3 
 
 4 
 
 Cream (1 tablespoonf ul) 
 
 0-7 
 
 5-1 
 
 0-7 
 
 54 
 
 225 
 
 8 
 
 Coffee (1 large cup) 
 
 
 
 ... 
 
 ... 
 
 28 
 
 1 
 
 Sardines (3 average) . 
 
 6-3 
 
 5-4 
 
 
 77 
 
 28 
 
 1 
 
 Lettuce or endive 
 
 0-3 
 
 0-1 
 
 1-0 
 
 5 
 
 28 
 
 1 
 
 Tomato 
 
 0-4 
 
 0-1 
 
 1-3 
 
 7 
 
 18 
 
 i 
 
 Yolk of one egg .... 
 
 2-9 
 
 60 
 
 
 68 
 
 22 
 
 2 
 4 
 
 Dressing as above (2 dessertspoon- 
 fuls) 
 
 
 16-0 
 
 
 148 
 
 98 
 
 3i 
 
 Roast-beef or mutton . 
 
 22-3 
 
 28-6 
 
 
 357 
 
 18 
 
 f 
 
 Cognac 
 
 
 (8-4) 
 
 0-1 
 
 70 
 
 14 
 
 i 
 
 Brazil nuts (4 large) . 
 
 2-3 
 
 9-3 
 
 1-0 
 
 100 
 
 18 
 
 f 
 
 Cream (1 tablespoonful) 
 
 0-7 
 
 5-1 
 
 0-7 
 
 64 
 
 225 
 
 8 
 
 Coffee (1 large cup) 
 
 
 ... 
 
 
 
 
 109-6 
 
 224-1 
 
 9-3 
 
 2494 
 
 Carbohydrate 
 
 -D . • / 109-6 - 
 
 Frotein sugar/ x 5 
 
 ° \6-25 
 
 Total sugar value 
 
 9-3 grams 
 87-5 ,, 
 9^ grams 
 
 limit of tolerance for other sugars and starches determined by- 
 adding equivalent quantities of dextrose, levulose, oatmeal, potato, 
 milk, &c., a day on a carbohydrate-free diet being interposed be- 
 tween each test. The quantities of the commoner starchy and 
 sugar foods containing the same amount of carbohydrate as one 
 ounce of bread are shown in the following table ; — 
 
312 
 
 GLYCOSURIA 
 
 Grranis 
 
 Oz. 
 
 Carbohydrate Ecjuivalents of 1 oz. Protein. | 
 
 Fats. 
 
 Carbohy. 
 
 Calories. 
 
 
 
 (30 grains) Bread. 
 
 Grrams. 
 
 Grams. 
 
 Grains. 
 
 
 30 
 
 1-0 
 
 Bread, white .... 
 
 2-8 
 
 0-4 
 
 15-9 
 
 80 
 
 26 
 
 0-8 
 
 „ toasted 
 
 
 3-0 
 
 0-4 
 
 15-9 
 
 90 
 
 34 
 
 1-0 
 
 „ brown . 
 
 
 1-8 
 
 0-6 
 
 15-9 
 
 78 
 
 32 
 
 1-0 
 
 ,, wholemeal . 
 
 
 31 
 
 0-3 
 
 15-9 
 
 81 
 
 32 
 
 1-0 
 
 ,, gluten . 
 
 
 3-0 
 
 0-5 
 
 15-9 
 
 82 
 
 72 
 
 2-4 
 
 Farina 
 
 
 1-2 
 
 0-1 
 
 15-9 
 
 41 
 
 21 
 
 0-7 
 
 Force 
 
 
 1-9 
 
 0-3 
 
 15-9 
 
 75 
 
 20 
 
 0-7 
 
 Grape-nuts 
 
 
 2-4 
 
 0-1 
 
 15-9 
 
 77 
 
 90 
 
 3-0 
 
 Hominy (boiled) 
 
 
 20 
 
 0-2 
 
 15-9 
 
 75 
 
 100 
 
 3-3 
 
 Macaroni( „ ) 
 
 
 3-0 
 
 1-5 
 
 15-9 
 
 91 
 
 24 
 
 0-8 
 
 Oatmeal (dry) . 
 
 
 38 
 
 1-7 
 
 15-9 
 
 97 
 
 138 
 
 4-6 
 
 (boiled) 
 
 
 3-9 
 
 0-7 
 
 15-9 
 
 87 
 
 65 
 
 2-2 
 
 Potato (baked) 
 
 
 1-9 
 
 0-1 
 
 15-9 
 
 74 
 
 75 
 
 2-5 
 
 „ (boiled) 
 
 
 1-8 
 
 0-07 
 
 15-9 
 
 73 
 
 34 
 
 1-1 
 
 „ (chips) 
 
 
 2-3 
 
 13-6 
 
 15-9 
 
 202 
 
 65 
 
 2-2 
 
 Rice (boiled) . 
 
 
 1-8 
 
 0-07 
 
 15-9 
 
 73 
 
 21 
 
 0-7 
 
 Shredded wheat biscuits 
 
 
 2-2 
 
 0-3 
 
 15-9 
 
 77 
 
 170 
 
 5-7 
 
 Vermicelli 
 
 
 2-4 
 
 0-4 
 
 15-9 
 
 80 
 
 16 
 
 0-5 
 
 Dextrose . 
 
 
 
 
 16-0 
 
 66 
 
 16 
 
 0-5 
 
 Le^ulose . 
 
 
 
 
 16-0 
 
 66 
 
 317 
 
 10-5 
 
 Milk, whole 
 
 
 10-7 
 
 12" 7 
 
 15-8 
 
 226 
 
 319 
 
 10-6 
 
 „ whey 
 
 
 3-2 
 
 1-0 
 
 15-9 
 
 88 
 
 By comparing the coefficients of excretion, and the total amounts 
 of ammonia nitrogen, one with the other, the effects of each of these 
 substances on sugar excretion and on metabolism can be deter- 
 mined. The results are more readily appreciated if they are j)lotted 
 out on squared " graph " paper, using different coloured inks, or 
 forms of Hne, to represent the intake and output of carbohj^- 
 drate, the total available carbohydrate in the diet, the coefficient 
 of excretion, the ammonia nitrogen, the total nitrogen, quantity of 
 urine, &c. &c. (Chart I.). 
 
 In this chart only the carbohydrate content of the food (F.S.), 
 the sugar excreted in the urine (U.S.), the coefficient of excretion 
 (C.E.), and the ammonia nitrogen (multiphed by ten) (Am. N.) are 
 shown, to avoid confusion. It will be seen that the patient could 
 take 22 grams of carbohydrate, including half an ounce of white 
 bread, mthout producing glycosuria, but that when the carbo- 
 hydrate content of the diet was raised to 30 grams by adding another 
 half ounce of bread, 5-5 per cent, of the available sugar appeared 
 in the urine unused, while 68 grams of carbohydrate, including 
 3 J ounces of bread, gave a coefficient of excretion of 14 per cent. 
 Of the other carbohydrates tried in this case only potato, milk, 
 and levulose appeared to be assimilated with anything like ease. 
 It will be noticed that when the patient was placed on a " carbo- 
 
PERSISTENT GLYCOSURIA 
 
 313 
 
 hydrate-free" (test) diet there was a slight rise in the excretion of 
 ammonia nitrogen, but on adding carbohydrate this fell to below 
 the former level. At no time, however, did it exceed the normal 
 
 LOO 
 
 S 
 
 E- 
 
 pq 
 
 
 PQ 
 
 pq 
 
 ^ 
 
 PM 
 
 s 
 
 iJ 
 
 o 
 
 
 90 
 BO 
 70 
 
 < 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 \ 
 
 
 
 
 
 L 
 
 
 
 
 
 
 
 
 \ 
 
 
 
 
 \ 
 
 \ 
 
 
 
 
 ^^ 
 
 ► F.S. 
 
 60 
 
 
 \ 
 
 
 
 
 / 
 
 \ 
 
 
 
 ) 
 
 '" 
 
 
 
 ( 
 
 
 
 
 / 
 
 \ 
 
 
 
 / 
 
 
 
 50 
 
 
 fr 
 
 
 
 
 + 
 
 -V 
 
 
 
 / 
 
 
 
 40 
 
 
 u 
 
 
 
 
 /- 
 
 
 
 
 r 
 
 
 
 30 
 
 
 w 
 
 
 
 
 / 
 
 
 
 
 
 
 
 
 w 
 
 
 
 / 
 
 
 
 
 
 
 
 
 20 
 
 
 \ 
 
 y 
 
 / 
 
 
 
 
 
 
 
 / 
 
 U.S. 
 
 
 \1 
 
 /\ 
 
 
 
 / 
 
 ^^ 
 
 
 
 
 /' 
 
 ■C.E. 
 
 10 
 
 
 
 
 
 
 /' 
 
 
 P 
 
 V 
 
 
 /' 
 
 
 
 
 
 
 I 
 
 / 
 
 
 
 ^"""^■^ 
 
 u==^ 
 
 / 
 
 
 -0 
 10 
 
 
 
 
 ^ 
 
 -^ 
 
 
 
 
 
 
 
 
 4 
 
 r"" 
 
 V. 
 
 
 
 
 
 
 
 
 
 
 
 X 
 
 > 
 
 y- — 
 
 ■—\ 
 
 , • 
 
 1 , 
 
 y. ( 
 
 1 K 
 
 1 , 
 
 r-— • 
 
 'Am.N. 
 
 n 
 
 
 
 
 
 
 
 
 
 
 
 
 (xlO) 
 
 F. S. = Food Carbohydrate. U. S.= Sugar in Urine. 
 
 C.E. = Coefficient of Excretion. Am. N.=Amnioma Nitrogen x 10. 
 
 Chart I. 
 
 limit, and the patient's urine never contained any aceto-acetic 
 acid, though traces of acetone were present throughout, excepting 
 when potato was being taken. 
 
 It will generally be found that, as in the above case, some forms 
 
314 GLYCOSURIA 
 
 of carbohydrate are better borne than others, and that not only is 
 a smaller proportion excreted in the urine as sugar, but that the 
 tendency to acidosis is reduced. By selection one or more of 
 those which appear to be best suited to the idiosyncrasies of the 
 case, and by giving them in quantities which experiment has 
 shown to be below the toleration point, considerable variation in 
 the diet can often be arranged, a most important consideration 
 from the patient's point of view, and one which generally ensures 
 his more strict adherence to the directions given him. In the case 
 just mentioned, for instance, the patient was allowed half an ounce 
 of white bread, or half a pint of milk, or half an ounce of levulose, 
 with occasional^ an ounce of potato. Only one form of carbo- 
 hydrate should be allowed on one day, and those which are best 
 borne should be most frequently taken. It is essential that the 
 patient should be impressed with the absolute necessity of his not 
 exceeding the quantity of each that is prescribed for him. The 
 time of day at which starchy foods are taken appears to exert some 
 influence upon their effects upon metabolism, as a rule they are 
 best tolerated wdth an early morning meal. 
 
 Having determined the caloric value of the types and quantity 
 of carbohydrate allowable, the diet is completed by adding sufficient 
 protein to produce from 1*5 to 1-6 calories per Idlogram of the body- 
 weight, and then making up the balance of energy to 34 or 35 
 calories per kilogram with fatty foods. Here, again, the diet may 
 be varied by classifying the commoner foodstuffs in groups, accord- 
 ing to whether they are preponderatingly protein or fatty, and sub- 
 dividing each group so that foods of a similar type are collected 
 together. If we then work out the quantity of each which will 
 yield a definite number of calories, this unit for convenience of 
 reference being termed a " ration," we can substitute one for the 
 other as convenience or appetite may direct, and yet be sure that 
 the patient is receiving the required amount of energy. The 
 " ration " I have selected for the unit is an average serving of roast- 
 beef, which weighs 3 ounces, and yields 300 calories. Any of the 
 other rations shown in the tables (pp. 315, 316) may be substituted 
 for this, for they all yield approximately the same amount of energy. 
 
 It wiU be noticed that some of the rations are inconveniently 
 large, and also that in the first table some contain an amount of 
 protein much in excess of that in a ration of roast-beef. To get 
 over this difficulty the quantities actually given are those shown 
 in the column headed "' portion allowed," in which a fraction of the 
 whole ration is indicated. By combining several such fractional 
 portions the food value of a whole ration can be obtained without 
 giving an excess of any one food or too large an amount of protein. 
 
PERSISTENT GLYCOSURIA 
 
 315 
 
 »i 
 
 
 
 
 
 gl 
 
 1— 1 rt^rtlti^H HClH»iH» 
 
 «l»iWn-<»i 
 
 .-.i^-4NH<N-:M«|«-|.= -io 
 
 rt rt 
 
 
 
 
 
 
 r -^ 
 
 
 
 
 
 ^<v 
 
 ooooooo 
 
 o oo 
 
 OOOOOOO 
 
 -* o 
 
 o 
 
 o o o o o o o 
 
 o o o 
 
 OOOOOOO 
 
 O CO 
 
 7i 
 
 CO CO CO fc CO ro M 
 
 CO CO CO 
 
 CO CO CO CO CO CO CO 
 
 M CO 
 
 o 
 
 
 
 
 
 ^m 
 
 
 
 
 
 o^ 
 
 
 
 O . v; IS . -o 
 
 
 j3 g 
 
 
 
 p : v' 7^ : Lo c-i 
 
 
 «■« 
 
 
 
 i ' h- X * -^ ?0 
 
 
 O 
 
 O IM O t^ CO <M O 
 
 
 
 
 C-. M CO 
 
 <7H p >p O O M — 1 
 
 p o 
 
 
 it; r; o^ r* ■* i^'^' ^ 
 
 s^i »b cv 
 
 -^ o M <Jo A( 1^ 30 
 
 •— o 
 
 C-1 (M .-H O'l 1— 1 (M 
 
 Ol r-H 
 
 
 G-] CO 
 
 'oj 
 
 r- Tf< lO C'l o o C5 
 
 CO t- c-i 
 
 CO ip 5-1 .-H lo c: o 
 
 00 o 
 
 P 
 
 00 CT5 05 <ffl o as lis 
 
 CO ^- C5 
 
 i) o in o Tt< ib o 
 
 -i 4ri 
 
 r-H <M M <M Tt< —( O 
 
 O i-l !M 
 
 «3 lO CO CO <M O U-5 
 
 C-l rt 
 
 Ph 
 
 
 
 
 
 
 
 _ 
 
 
 
 
 
 ^ . . . 
 
 oT 
 
 
 
 
 CS 
 
 'o 
 
 
 
 
 Ul 
 
 VJ 
 
 
 .2 
 
 
 cS • • • 
 
 
 , . 
 
 
 
 <u 
 
 
 
 o 
 
 
 -a 
 
 (1) 
 O 
 
 
 O 
 
 
 
 '5 
 
 ^ 
 
 o 
 o 
 
 
 oT ■ * ■ 
 
 "p^ 
 
 
 M 
 
 
 TJ 
 
 
 
 >> 
 
 
 •r; . . . 
 
 « 
 
 , 
 
 2 
 
 52 
 
 
 
 ^ 
 
 ^ 
 
 
 
 & ^ 
 
 r— 1 
 
 
 c 
 
 
 ' -*J • 
 
 
 
 o 
 
 
 «3 -*i ra 
 
 rM ^-~ 
 
 
 >; 
 
 
 o c' "• t^ 
 
 o rx:; 
 
 
 0. 
 
 • ^ ■ • • • • 
 
 O _ _ _ OJ _ 
 
 
 2 
 ■>> 
 
 s 
 o 
 
 'o 
 
 •^. '^^ '^ ■- 
 
 ='■'-' to g 
 
 S -S b/5 ^ 
 
 o 
 
 ^illil^l 
 
 iJ ••0 
 
 bD - 
 bo ' 
 
 
 tC 
 
 <D 
 
 5 
 
 s^laass^ 
 
 0) 
 
 So - _ - J' . , 
 
 o 
 
 2 
 
 J2 
 
 X 
 
 .2 .a = ' ' = s 
 
 ?3 ' ' 
 
 > 
 
 c3 
 
 ^ - - - be - ^ 
 
 bp 
 
 Jh " 
 
 > 
 
 Ph 
 
 fj l« 
 
 
 d 
 
 cS 
 
 
 -<■*) 
 
 HM 
 
 ^l« -iPi 
 
 
 
 ^ CO N r-l F-H ^ CO 
 
 rt S^l rt 
 
 CO M CO .-H O O O 
 i-l C-l 
 
 ••# o 
 
 
 «l'*'^t*H'^ Hwr^H* 
 
 co[-)i -<tn;^ 
 
 ri^ c^'i^ "i'*«;-*«|.ji-'-* 
 
 -k 
 
 O 
 
 CO CO O CO O CO XI 
 
 IQ -^ CO 
 
 CO X> t~ »0 CO 00 C5 
 
 t-CO 
 
 S 
 
 •* ^^ o ^ CO o cq 
 
 za o o 
 
 ■^ X S-1 <M I^ O CO 
 
 oo 
 
 CO O >0 03 -* O CO 
 
 CO CO o 
 
 o -^i o o o o -^ 
 
 o C-. 
 
 2 
 
 rt rH i-H 1-1 <N 
 
 rH 1— ( 1 — 1 
 
 CO oi e-j rt r-( CO c>i 
 
 IM 
 
 C5 
 
 
 
 
 
316 
 
 GLYCOSURIA 
 
 
 
 oocooooooooooooo 
 
 OOOC OOOOoOOOOCOO 
 
 
 
 OOpMMNipooOI;-epMNtp«5p 
 COOOCOCOCOCOfMINSqiMiNiMWiMiMIN 
 
 .'^ICO'-'OO.— lTtHlO-*'<*<CCOM 
 
 
 tj .a — IS s 03 fS 
 
 
 vj aj uj 
 
 a a3 s 
 
 • M to 
 
 P4 
 
 
 
 O OJ 3 (g 3J 
 
 to 4J o -tS 43 
 
 >0(Nl>Oi0«0M'*J0-* 
 
 
 COCOfOOO(30'-HO(M»0»OiOCOOOiO 
 
PERSISTENT GLYCOSURIA 317 
 
 When a patient has had a diet worked out for him he is told 
 how much energy he requires a day, and is then given a specimen 
 diet with a list of various foods that he may substitute for those 
 shown. To avoid the inconvenience of manipulating fractions I 
 am in the habit of taking the " food value " of a unit ration as 100 
 and allotting to the other foods corresponding values, so that the 
 diet chart for proteins and fatty foods given to the patient reads 
 as follows : — 
 
 Meat — Food Value. 
 
 Beef 3 oz., mutton 3^ oz. . . . . , . . 100 
 
 Lamb 2J oz., pork 2J oz., ham If oz,, tongue If oz. . . 50 
 
 Poultry — 
 
 Chicken 3 oz., grouse 3 oz., partridge 3 oz., pheasant 3 oz., 
 
 pigeon 3 oz., duck 2 J oz., goose 2 J oz., turkey 2 oz. . . 50 
 
 Fish — 
 
 Halibut 4 oz., mackerel 3| oz., salmon 2 J oz., sardines 
 
 (tinned in oil) 2 oz. . . . . . . . .50 
 
 Cod 2J oz., haddock 2J oz., hake 2J oz., plaice 2J oz., sole 
 
 2\ oz., whiting 2J oz. ....... 25 
 
 Smelts 2 J oz., trout 2 oz. . . . . ... .20 
 
 Fatty Foods — 
 
 Bacon IJ oz., butter l^^ oz., margarine IJ oz.,lard 1 oz., suet 
 
 1 oz., salad oil 1 oz. ........ 100 
 
 Cream 2 oz., Devonshire cream \\ oz., cheese (cream, 
 
 Cheddar, Stilton) 1 oz., cheese (Cheshire, Roquefort) IJ oz. 50 
 Brazil nuts 1 oz., olives 2 oz. . . . , . .70 
 
 Eggs— 
 
 Average whole, 4, or average yolks, 5 . . . . .100 
 
 Milk— 
 
 (Whole cream) J pint ........ 75 
 
 As a rule it will be found that three or four protein " rations " 
 will contain about 100 to 120 grams of protein, so that this amount, 
 yielding 900 to 1200 calories, should not be exceeded. The balance 
 of the 1800 to 2000 calories required by an average individual must 
 be made up in other ways. If carbohydrate tolerance is low and 
 only a small quantity of starchy food can be given the administra- 
 tion of the large amount of fatty material required may present a 
 serious difficulty, especially with some patients who have a natural 
 repugnance to fats. Bacon, cream, and cheese will supply a con- 
 siderable amount of the energy required, and are usually well 
 taken. A few Brazil nuts, or green olives, and a certain amount 
 of butter, or margarine, may also be given. Frequently, however, 
 a larger quantity of the latter than most patients will tolerate 
 when they are having httle or no bread, and a certain amount of 
 salad (olive) oil, to which many people have a strong objection, 
 
318 
 
 GLYCOSURIA 
 
 have to be introduced into the diet. This difficulty may be over- 
 come to a large extent by giving either oiled butter, or salad dress- 
 ing, with vegetables, especially the coarser kinds. The following 
 are the food values of several such dressings (Locke) : — 
 
 French Dressing — (4 tabsp. olive-oil, 1 tabsp. vinegar, J tsp. salt, 
 
 pepper) — ^one dessertspoonful yields 74 calories. 
 Hollandaise Sauce — {h, cup butter, yolk 2 eggs, 3 teasp. lemon-juice, 
 
 salt, cayenne pepper)- — 2 tablespoonfuls yield 170 calories. 
 Mayonnaise Dressing — (2 eggs, 2 cups olive-oil, 3 tabsp. vinegar, or 
 
 3 tabsp. lemon-juice, salt, pepper, mustard) — one tablespoonful 
 
 yields 187 calories. 
 
 Some vegetables contain only a small proportion of carbo- 
 hydrate, and by selecting those in which the lowest percentage 
 [e.g. under 5 per cent.) is met with, and working out weights of 
 these which contain 1 gram or less of carbohydrate, considerable 
 choice, suitable to the tastes of the patient and the season of the 
 year, can be offered. The commoner vegetables of this class are 
 shown in the following table : — 
 
 The Common Vegetables containiiig under 5 Fer Cent, of Carhohydrate 
 
 Grams. 
 
 Oz. 
 
 
 Protein. 
 
 Fat. 
 
 Carbohy. 
 
 Calories. 
 
 100 
 
 u 
 
 1 av. helping asparagus (cooked) 
 
 1-5 
 
 0-11 
 
 2-8 
 
 18 
 
 100 
 
 3J, 
 
 3 hpcl. tablesp. French beans ,, 
 
 0-8 
 
 1-1 
 
 1-9 
 
 22 
 
 100 
 
 St 
 
 3 ,, cabbage ,, 
 
 0-6 
 
 0-1 
 
 0-4 
 
 5 
 
 100 
 
 3^ 
 
 2 ,, cauliflower ,, 
 
 0-9 
 
 0-1 
 
 0-4 
 
 7 
 
 100 
 
 3+ 
 
 2 larare mushrooms ,, 
 
 3-4 
 
 0-2 
 
 3 
 
 29 
 
 100 
 
 3* 
 
 1 average onion „ 
 
 1-2 
 
 1-8 
 
 4-9 
 
 42 
 
 100 
 
 U 
 
 4 slices parsnip ,, 
 
 0-22 
 
 0-29 
 
 1-46 
 
 10 
 
 100 
 
 U 
 
 2 hpd.talilesp. spinach ,, 
 
 2-7 
 
 0-8 
 
 3-0 
 
 18 
 
 100 
 
 3+ 
 
 2 tablesp. turnip ,, 
 
 0-3 
 
 0-06 
 
 0-6 
 
 4 
 
 100 
 
 3* 
 
 6 small sticks celery (raw) 
 
 1-0 
 
 10 
 
 2-8 
 
 16 
 
 100 
 
 3^ 
 
 16 thin slices cucumber ,, 
 
 0-8 
 
 0-2 
 
 3-1 
 
 18 
 
 100 
 
 3+ 
 
 endive ,, 
 
 1-0 
 
 0-2 
 
 3-0 
 
 16 
 
 100 
 
 U 
 
 lettuce ,, 
 
 0-9 
 
 0-3 
 
 2-9 
 
 17 
 
 100 
 
 U 
 
 radishes ., 
 
 1-2 
 
 0-1 
 
 5-0 
 
 25 
 
 100 
 
 U 
 
 sorrel „ 
 
 2-0 
 
 2-0 
 
 3-0 
 
 25 
 
 100 
 
 3+ 
 
 tomato ., 
 
 1-2 
 
 0-2 
 
 4-0 
 
 23 
 
 100 
 
 3i 
 
 watercress ., 
 
 0-7 
 
 0-0 
 
 3-7 
 
 18 
 
 The weights calculated to contain 1 gram, or less, of carbohydrate, 
 are as follows : — 
 
 Cooked vegetables containing 1 gram or less of carhohydrate — 
 
 Food Value. 
 Cabbage 8 oz., cauliflower 8 oz., tiu-nip 5 oz., parsnip 2 oz., 
 French beans IJ oz., asparagus 1 oz., onions 1 oz., 
 mushroom 1 oz. ......... 5 
 
PERSISTENT GLYCOSURIA 
 
 319 
 
 Raw vegetables containing 1 gram or less of carbohydrate — Food Value. 
 
 Celery 1 oz., lettuce 1 oz., endive 1 oz., sorrel 1 oz., cucumber 
 
 1 oz., watercress 1 oz., tomato 1 oz., radishes 1 oz. . . 2 
 
 By instructing the patient not to take more than one, or two, 
 rations daily of these vegetables we can ensure his not deriving 
 more than one, or two, grams of carbohydrate from this source. 
 Although their energy value is practically negligible, they form a 
 vehicle by which fats may be administered. They also furnish 
 alkalies to the organism, and by their bulk they make the diet 
 more satisfying. 
 
 In cases where there is fair tolerance for carbohydrate the diet 
 may be further varied by allowing certain fruits which contain 
 comparatively small percentages of carbohydrate. The more 
 common, containing 10 per cent., or under, are as follows : — 
 
 The Commoney^ Fruits containing 10 Per Cent., or under, of Garhohydrate 
 
 Grams. 
 
 Oz. 
 
 100 
 
 Hi 
 
 100 
 
 U 
 
 100 
 
 31 
 
 130 
 
 4i 
 
 100 
 
 'd\ 
 
 128 
 
 4* 
 
 100 
 
 ?A 
 
 100 
 
 U 
 
 100 
 
 3* 
 
 300 
 
 \0\ 
 
 3 hpd. tablesp. blackberries 
 
 „ „ bilberries 
 
 ,, ,, cranberries 
 
 1 average lemon 
 61- teasp. ,, juice 
 
 1 average peach 
 Edible pt. 2 slices pineapple 
 
 2 hpd. tablesp. ihubarb . 
 
 4 hpd. ,, strawberries 
 1 large slice water-melon 
 
 Protein. 
 
 1-30 
 1-60 
 1-30 
 0-90 
 
 0-t>4 
 0-40 
 0-40 
 TOO 
 0-60 
 
 Fat. 
 
 1-00 
 0-20 
 010 
 0\S5 
 
 0-13 
 0-30 
 0-60 
 0-60 
 0-30 
 
 Carbohy. Calories, 
 
 10-00 
 7-20 
 9-90 
 7-G7 
 9-80 
 9-86 
 9-70 
 3-. 50 
 7-40 
 8-10 
 
 46 
 H2 
 40 
 44 
 44 
 34 
 40 
 39 
 
 Weights of these containing approximately 5 grams of carbo- 
 hydrate are : — 
 
 Fruits containing ahout 5 (/rai)is of carhohiidrate — 
 
 Food Value. 
 Water-melon 6 oz., rhubarb 4| oz., lemon 2f oz., bilberries 
 2 J oz., peach 2 oz., strawberries 2 oz., blackberries 1^ oz., 
 cranberries 1| oz., lemon- juice 1| oz., pineapple 1^ oz. . 7 
 
 The patient is also supplied with a list of foods that he must 
 not take unless specially ordered, including : — 
 
 Sugar and starchy food in all forms. 
 
 Bread, toast, biscuits, pastry, pies, puddings, rice, sago, tapioca, 
 macaroni, vermicelli, arrowroot, cornflour, oatmeal. 
 
 Potato, carrot, turnips, parsnip, artichokes, beetroot, peas, beans, 
 lentils. 
 
 Fruit, sweets, chocolate, ices, jam, honey. 
 
 Sauces and gravies thickened with flour. Thickened soups and broth. 
 
320 GLYCOSURIA 
 
 Oysters, liver. 
 
 Milk, ale, stout, porter, cider, sweet and sparkling wines, port 
 wine, liqueurs. 
 
 Diabetics are usually allowed meat extracts and unthickened 
 soups, but as Thompson and Wallace have shown that the addition 
 of even small quantities of creatinin to the diet temporarily in- 
 creases the output of sugar by nearly 50 per cent., it would seem 
 that these are best avoided. 
 
 In addition to the substances already mentioned the patient 
 is allowed : — 
 
 Tea or coflfee (without milk or sugar), vinegar, pickles, and saccharin 
 or saxin. 
 
 A word may be said here about the use of saccharin and similar 
 coal-tar sugar substitutes. As a rule diabetic patients are allowed 
 an unlimited quantitj?-, but the experiments of the Referee Board 
 of the United States Department of Agriculture have shown that 
 saccharin in large doses, over 0-3 grams per day, and especially 
 over 1 gram a day, added to the food and taken for considerable 
 periods, are hable to induce digestive disturbances, increase the 
 free hydrochloric acid in the gastric juice, alter the reaction of 
 the fseces, cause a greater formation of putrefactive products in 
 the intestine, and eventually bring about serious distaste for the 
 substance. 
 
 With regard to '" diabetic " breads, and bread substitutes, it is 
 most important that these should be obtained from a thoroughly 
 rehable maker, and an emphatic warning must be entered against 
 the majority of those now in the market. Many are exploited as 
 being " practically starch-free," or as containing starch " only 
 in a form that is readily digested and assimilated by diabetics," 
 but analyses show that the latter have usually been merely sub- 
 jected to heat, and that many of the former contain as much, or 
 nearly as much, starch as ordinary white bread. In one analysis, 
 for instance, the following results were obtained : — 
 
 White bread 100 gTams = 9-2 grams protein, 1-3 grams fat, 53-1 grams 
 
 carbohydrate. 
 Gluten bread 100 gTams = 9-3 grams protein, 1-4 grams fat, 49-8 grams 
 carbohydrate. 
 
 All such bread substitutes are much more expensive, both 
 from a financial and physiological point of view, than ordinary 
 bread, and they are really a serious danger, for by using them the 
 patient may be unconsciously taking an amount of starch far 
 beyond his powers of assimilation. It is better that he should be 
 given a definite amount of a food the danger of which is known, than 
 he should be allowed to live in a fool's paradise by substituting an 
 
PERSISTENT GLYCOSURIA 321 
 
 expensive proprietary preparation of unknown, and often varying, 
 composition. In my ex]3erience very few commercial bread substi- 
 tutes can be relied upon as being starch free, and those that can 
 are usually so unpalatable that most patients soon tire and prefer 
 to do without them. A small quantity of some thoroughly reliable 
 diabetic bread, or biscuit, however, is often useful as a vehicle for 
 the administration of butter, cheese, &c., but its exact composition 
 should be known, and it should be tested for starch from time to time, 
 otherwise it is safer to allow a definite quantity of ordinary bread. 
 
 Some people have an idea that toast is better for a diabetic 
 than fresh bread, but a glance at the table on p. 312 will show 
 that their composition is practically the same, the only difference 
 being the higher proportions of protein, fat, and carbohydrate, 
 due to the loss of water, so that 26 grams of toast are equivalent 
 to 30 grams of fresh bread, so far as the carbohydrate content goes. 
 The superficial layers of starch are partly converted into dextrin 
 in the process of toasting, but this does not alter in its effect on 
 carbohydrate metabolism. 
 
 To merely give a patient a hst of foods that he " may take," 
 and another of those that he " must avoid," is a very perfunctory 
 way of treating a case of persistent glycosuria, and in all but the 
 mildest forms is hkely to eventually lead to disaster, or at least 
 shorten the possible span of hfe. Persistent glycosuria is a disease 
 of the chemistry of the body in much the same way as a defect of 
 vision is an anatomical one, and just as lenses of the proper form 
 and size are worked out for the latter, so must the diet of the former 
 be adapted to the needs of the case, both in quahtj^ and amount. 
 The patient must be taught that instinct is no longer a safe guide, 
 that he must watch every mouthful that he takes, and learn that 
 a certain amount of the foods that he can assimilate must be con- 
 sumed each day. It may be thought that this is a counsel of per- 
 fection, and that most patients will not go to the trouble of weighing 
 their food, but after a very short time it is quite unnecessary to do 
 so, except as an occasional check, for it is surprising how quickly the 
 eye can be trained to estimate the weights of the various foodstuffs 
 that are required. One of the advantages of the method of working 
 out a diet that I have outlined is that it can only be satisfactorily 
 carried out in an institution, or nursing-home, where the food can 
 be carefully prepared and weighed, and as the patient can at the 
 same time be taught the quantities that he is receiving all incon- 
 venience in this respect is avoided when he returns home. Even 
 in mild cases of persistent glycosuria, and in many cases of transitory 
 glycosuria, a quantitative restriction of the diet is advisable. In 
 the more severe cases it is essential. 
 
 X 
 
322 GLYCOSURIA 
 
 As the object of the treatment is to so arrange the intake of food 
 that the metaboUc powers of the patient for carbohj^drates are 
 not overtaxed, but are working below their maximum capacity, so 
 that they may have an opportunity of recovering their tone, con- 
 stant supervision is required for some time. The urine should be 
 examined at intervals, the diet being revised accordingly to the 
 metabohc findings, and adjusted to the progress of the case. 
 
 If faulty carbohydrate metaboHsm were the only factor to 
 be considered in diabetes its treatment would be comparatively 
 simple, but in severe cases, where there is not onty a high coeffi- 
 cient of excretion, but also well-marked secondary disturbances of 
 metabohsm, with a marked excess of acetone bodies, a high total 
 output of ammonia nitrogen, and a serious disturbance of nitro- 
 genous equihbrium, we have to take into account the patient's 
 defective powers of deaUng with proteins, and probably also with 
 fatty acids. In treating such a case we must first endeavour to 
 restore the nitrogen balance as much as possible. This is accom- 
 pHshed by (1) rest in bed, (2) reducing the diuresis, (3) limiting the 
 intake of nitrogenous food, (4) the use of such drugs as codeia. 
 opium, arsenic, &c. 
 
 Since the large amount of urine passed is one of the characteristic 
 features of most severe cases of diabetes, and nitrogenous equilibrium 
 is practically impossible when more than 1200 to 1500 c.c. are 
 excreted in a day, we must, as a preliminary step, endeavour to 
 diminish the diuresis by reducing the glycosuria upon which it 
 depends. The carbohydrate in the test diet is therefore cautiously 
 reduced until the sugar disappears or is diminished, so that the 
 daily excretion of urine falls to about 1500 c.c. As the glycosuria 
 itself is of secondary importance, time should not be spent in the 
 frequently futile, and dangerous, task of attempting to make the 
 urine sugar- free, but attention should be mainly devoted, at least 
 at first, to diminishing the diuresis sufficiently to allow of a more 
 normal nitrogen balance being established, and to controUing the 
 acidosis. Restriction of the carbohydrates of the food always in- 
 creases the acetonuria, but if the restriction can be safely persisted 
 in an ultimate decrease often results. A constant watch must be 
 kept on the excretion of acetone bodies and ammonia nitrogen for 
 indications of serious acidosis, and should there be evidence of this, 
 coupled mth the prodromal symptoms of diabetic coma, more 
 starchy food should be added to the diet, and other means be taken 
 to combat the condition. It is usually safe to diminish the carbo- 
 hydrates in the diet, in spite of mod3iate acidosis, so long as there 
 is a slight nitrogen addition to the body. 
 
 If the measures already taken are not sufficient to restore 
 
PERSISTENT GLYCOSURIA 323 
 
 nitrogenous equilibrium, the coefficient of excretion remains high, 
 and the output of acetone bodies and ammonia nitrogen is ex- 
 cessive, the next step is to reduce the amount of nitrogenous food 
 until 100, 75, or even 50 grams of protein are being taken in the 
 day, and not more than 12 or 15 grams of nitrogen are being ex- 
 creted in the urine. Vegetable proteins and eggs are often found 
 to disturb metaboHsm less than nitrogenous foods of animal origin, 
 and they may be advantageously substituted for an equivalent 
 amount of meat. Such restrictions of the protein of the food 
 frequently exert a very favourable influence on the acidosis, and at 
 the same time reduce the amount of sugar in the urine to a much 
 greater extent than mere limitation of the carbohydrate, even when 
 some starchy food is being taken at the same time. To show the 
 manner in which the diet is worked out in severe diabetes the 
 following example may be quoted. 
 
 The case was a serious one, with a coefficient of excretion of 84 
 per cent., and an ammonia nitrogen excretion of 1-2 grams in the 
 twenty-four hours, on a diet containing 76 grams of carbohydrate, 
 including 3 oz. of bread. On reducing the bread to 2 oz. the co- 
 efficient of excretion fell to 66 per cent., but even with no added 
 bread and a diet containing only 20 grams of carbohydrate, it 
 .stood at 65 per cent., and the ammonia nitrogen rose to 1-8 grams in 
 the twenty-four hours. By cutting down the proteins of the diet 
 the coefficient of excretion was reduced to 62 per cent., and the 
 ammonia nitrogen excretion dropped to 1*5 grams. The tolerance 
 for different forms of carbohydrate was tested bj^ adding dextrose, 
 levulose, milk, oatmeal, potato, and apple successively to the test 
 diet, with the results shown in the chart (Chart II.). 
 
 From this it is apparent thatbothpure dextrose and pure levulose, 
 particularly the latter, were utilised with difficulty, but that the 
 carbohydrate contained in oatmeal and potato were metaboHsed 
 comparatively well. The addition of milk, and potato, was found 
 to be followed by a reduction in the output of ammonia nitrogen, 
 so that they apparently diminished the acidosis. A diet based on 
 these results was worked out and taught to the patient. At the 
 end of his stay in the nursing-home he was put upon a diet contain- 
 ing 24 grains of carbohydrate, and it was found that his coefficient 
 of excretion had fallen to 55 per cent., as compared with 65 per cent, 
 when the treatment was commenced, so that his power of deaHng 
 with carbohj^drate had improved about 10 per cent. 
 
 In cases of this description von Noorden's '" oatmeal ciire " 
 may give very satisfactory results, at least temporarily. This 
 treatment, as we have seen, consists in the administration of 200 
 to 250 grams of oatmeal, 200 to 300 of butter, and 3 or 4 eggs, with 
 
324 
 
 GLYCOSURIA 
 
 nothing else except black coffee or tea in the twenty-four hours. 
 Such a diet will contain only from 50 to 70 grams of egg and vegetable 
 protein, 180 to 300 grams of fat, and 135 to 170 grams of carbo- 
 hydrate, yielding 1900 to 3700 calories, but as an essential part of 
 
 
 n 
 
 
 Eh 
 
 
 Q 
 
 
 S 
 
 o 
 
 PU 
 
 < 
 
 H 
 
 
 120 
 110 
 100 
 90 
 80 
 70 
 60 
 50 
 40 
 30 
 20 
 10 
 
 ( 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 > 
 
 
 
 
 
 
 
 ; 
 
 
 \ 
 
 
 
 
 
 
 
 
 
 
 1 
 
 V \ 
 
 \ \ 
 
 \ \ 
 
 \ 
 
 
 
 
 
 
 
 
 
 
 
 
 I 
 
 U^^ 
 
 :— "' 
 
 
 \ 
 
 \ 
 
 
 
 
 --^ 
 
 U.S. 
 
 
 
 \ 
 
 
 
 
 \ 
 \ 
 
 V ^ 
 
 -- 
 
 / 
 
 {.'"" 
 
 vC.E. 
 
 
 
 \ 
 
 
 / 
 
 
 ^^ 
 
 \ 
 
 \ 
 
 S 
 \ 
 
 --^ 
 
 • 
 
 
 
 
 
 \ 
 
 
 / 
 
 
 s 
 
 X 
 
 "/ 
 
 
 ^^ 
 
 'F.S. 
 
 
 
 \ 
 
 
 / 
 
 
 
 
 
 
 
 
 ( 
 
 
 1 ■* 
 
 "^t 
 
 
 
 1- A 
 
 
 k^ 
 ^^'•i 
 
 1 , 
 
 __,-.-4 
 
 ,Am.N. 
 (xlO) 
 
 
 
 
 
 
 
 
 
 
 
 
 
 F. S.= Food Carbohydrate. U. S.= Sugar in Urine. 
 
 C.E.= Coefficient of Excretion. Am. N.=Animonia NitrogenxiO. 
 
 Chart II. 
 
 the treatment is the interposition of one or two days of pure vege- 
 table diet, the nitrogenous intake is reduced to a very low level. 
 The effect of substituting other starchy foods, such as whole wheat 
 meal, barley meal, potato, &c., for oatmeal, may be tried, as it is 
 sometimes found that they give as good a result, and are better 
 Hked by the patient. One form onlj^ of carbohydrate should.. 
 
PERSISTENT GLYCOSURIA 325 
 
 liowever, be given at one time. Drugs, such as opium, are useful 
 adjuncts in the treatment, when careful dieting fails to reduce 
 the condition. 
 
 In severe cases of diabetes the quantitative arrangement of 
 the diet is even more important than in the milder forms, for not 
 only have we to watch the carbohydrate intake, but the amount of 
 protein consumed must also be carefully regulated. It is, as a 
 rule, practically impossible to keep the patient's urine free from 
 sugar, and it is usually found that he does better when a certain 
 amount of carbohj^drate is allowed than when he is put on a strict 
 carbohydrate-free diet. The kinds of starchy food that are best 
 borne vary with different cases, and should be experimentally 
 determined. Having ascertained the amounts and varieties of 
 ■carbohydrate, and protein, that give the best result, we must add 
 sufficient fat to the diet to make up the 34 or 35 calories per kilogram 
 of body-weight required. Little or no energy for the needs of the 
 patient can be obtained from the carbohydrate, and the proteins 
 ■will probably yield but 10 or 15 per cent, of the total requirement, 
 so that we are faced with the very serious difficulty of suppl3dng 
 fat in an assimilable form to make up the deficiency. This is a 
 problem which will often tax the resources of the physician to the 
 utmost, and requires all the arts of cookery to accomplish. Although 
 it is now known that the acetone bodies are chiefly formed from fats, 
 it would appear that their production in the body is not influenced 
 by the quantity of fat in the diet unless an enormous amount is 
 taken, so that fatty foods can be safely given in diabetes. As the 
 lower members of the fatty acid series are more Hkely to cause 
 acidosis than the higher, fatty foods that contain the latter should 
 be selected. For this reason good margarine is preferable to butter, 
 unless the latter has been washed to remove the butyric acid, &c., 
 that it contains. The coarser vegetables that contain only a smaU 
 proportion of carbohydrate may be used as vehicles by which a 
 considerable quantity of salad-oil and margarine, or butter, can be 
 introduced into the diet. Cream can be given with coffee. Bacon, 
 cream cheese, and sardines in oil are well taken by most patients. 
 Brazil nuts, almonds, and a few green olives may also be used in 
 moderation, while lard and suet can be employed in various ways in 
 the preparation of the other food materials. English patients have 
 usually a constitutional antipathy to fatty foods, and although one 
 may succeed for a time in prevaihng on them to take a sufiicient 
 amount, they quickly tire, or gastro-intestinal disturbances super- 
 vene which cause nausea and distaste. The latter maj^- be pre- 
 vented to a certain extent by giving a small quantit}^ of alcohol 
 with each fattv meal, but the total amount should not exceed an 
 
326 
 
 GLYCOSURIA 
 
 alcohol content of 40 grams a day. The energy value of the alcohol 
 itself- is small, but its use certainly permits of more fat being 
 taken with comfort than would otherwise be the case. The 
 composition of the alcohohc beverages that may be employed in 
 this way is shown in the following table : — 
 
 
 
 
 Alcohol Per 
 
 Extractives 
 
 
 Cc. 
 
 Oz. 
 
 Alcoholic Beverages. 
 
 Cent, by 
 Weight. 
 
 Per Cent, (as 
 Sugar). 
 
 Calories. 
 
 20 
 
 # 
 
 2 dessertsp. brandy (Cognac) . 
 
 55-9 
 
 0-02 
 
 78 
 
 50 
 
 1 + 
 
 3 tablesp. gin 
 
 30'0 
 
 O-50 
 
 116 
 
 50 
 
 U 
 
 ,, rum (Jamaica) . 
 
 69-6 
 
 0-61 
 
 245 
 
 50 
 
 H 
 
 ,, whisky .... 
 
 89-0 
 
 
 137 
 
 135 
 
 4+ 
 
 1 glass champagne (dry) . 
 
 10-4 
 
 2-36 
 
 112 
 
 120 
 
 4 
 
 „ claret ..... 
 
 8-2 
 
 2'42 
 
 81 
 
 120 
 
 4 
 
 ,, French (white) 
 
 9-5 
 
 3-03 
 
 95 
 
 120 
 
 4 
 
 ,, Moselle or Saar (white) 
 
 7-4 
 
 2-31 
 
 73 
 
 120 
 
 4 
 
 „ Khein (white) 
 
 8-1 
 
 •2-91 
 
 83 
 
 30 
 
 1 
 
 ,, sherry .... 
 
 17-5 
 
 3-98 
 
 42 
 
 In some instances the difficulty is best overcome by prescrib- 
 ing fat in an emulsified form. The ordinary commercial cod-hver 
 oils can be used, or an emulsion of ohve, or cod-liver, oil may be 
 prepared : — 
 
 (Contains 33 per cent, of oil) 
 
 Olive, or cod-liver, oil . 
 
 4J 
 
 - oz. 
 
 R Castile soap . 
 
 . i- dr. 
 
 Cherry laurel water 
 
 3 
 
 dr. 
 
 Cherry laurel water 
 
 . 5 „ 
 
 Orange flower water 
 
 1 
 
 fi. oz. 
 
 Orange flower water 
 
 . 10 „ 
 
 Carrageen 
 
 SO 
 
 gr. 
 
 Saccharin 
 
 . 3gr. 
 
 Essence of bitter almonds 
 
 4 
 
 min. 
 
 Essence of peppermint 
 
 . 6 min 
 
 Saccharin 
 
 2 
 
 gr. 
 
 „ „ lemon 
 
 . 6 „ 
 
 Distilled water 
 
 5 
 
 &. oz. 
 
 Olive-oil to 
 
 . 1 pint 
 
 (Contains 90 per cent, of oil) 
 
 A palatable and efficient emulsion may be made by mixing 
 olive-oil, castor-oil, and glycerine, and adding to this calcium lactate 
 or gum-arabic (4 per cent.) with a httle flavouring, such as almond 
 or cinnamon oil, or oil of wintergreen :- 
 
 R OHve-oil 
 Castor-oil 
 Glycerine 
 
 Calcium lactates 
 Almond oil 
 Chloroform water 
 
 3-4 dr. 
 +-1 ,, 
 
 1 ,. 
 
 5gr. 
 
 y min. 
 to 1 oz. 
 
PERSISTENT GLYCOSURIA 327 
 
 This emulsion is particularly useful in overcoming the con- 
 stipation from which most diabetics suffer. 
 
 In cases where there is pancreatic insufficiency and the digestion 
 and absorption of fats are interfered with, solid fats, particularly 
 those with a high melting-point, should be avoided, as they are 
 liable to undergo chemical changes in the intestine, with the forma- 
 tion of irritating by-products, and consequently give rise to dis- 
 comfort. All fats should be given, as far as possible, in an 
 emulsified form, since Abelmann has shown that emulsified are 
 better absorbed than unemulsified fat by dogs w'hose pancreas has 
 been partly removed, and Cavazzani found that, while ordinary fat 
 is rejected, soap is eaten with eagerness by animals after extirpa- 
 tion of the pancreas. Abelmann obtained the best results with a 
 natural emulsion, such as milk, 30 per cent, of the fat of large 
 amounts, and 53 per cent, of small quantities, being absorbed after 
 complete extirpation, and up to 80 per cent, when portions of the 
 pancreas had been left behind. The high sugar and protein content 
 of milk is, as a rule, a drawback to its use in pancreatic insufficiency 
 associated with diabetes. In some cases of persistent glycosuria, 
 however, it is well borne, and its addition to the diet does not cause 
 a marked increase in the glycosuria, or notably affect the excretion 
 of the acetone bodies — in fact, Donkin has stated that by the use 
 of an exclusive skimmed milk diet sugar may often be entirely 
 removed from the urine in a fortnight. This treatment appears to 
 be most useful in fat, gouty, overfed persons. In other cases even 
 a few ounces of milk has a bad effect. Various methods of preparing 
 milk, free from sugar, for the use of diabetics, have been suggested, 
 but most of these are too complicated, or vmcertain, for ordinary use. 
 " Diabetic, sugarless " milks, put up in sterile bottles, can now be 
 obtained from several firms, and may be used in moderation, remem- 
 bering that they contain 10 or 12 per cent, of protein and often | per 
 cent, or more of sugar. As a substitute for milk Williamson has sug- 
 gested a mixture of washed cream and white of egg, suspended in 
 water. Three or four tablespoonfuls of fresh cream are thoroughly 
 mixed with about a pint of water, and allowed to stand for twelve to 
 twenty-four hours, the fat that has meanwhile floated to the surface 
 is skimmed off and mixed with water, the white of an egg is added 
 and the mixture well stirred, on adding a trace of salt and a Httle 
 saccharin "a palatable preparation, closely resembling milk, and 
 practically sugar-free," is obtained. If the functions of the stomach 
 are being carried out satisfactorily a considerable amount of proteid 
 may be digested in spite of a deficiency of pancreatic juice, both 
 in the stomach and upper part of the intestine, where the action 
 of the gastric secretion will continue, owing to the absence of the 
 
328 GLYCOSURIA 
 
 alkaline pancreatic secretion, but even then less than half the 
 albumin of the food is absorbed. Proteids which are cUgested with 
 difficulty, such as pork, white of boiled egg, &c., must be excluded 
 from the diet. The most useful proteid in cases of pancreatic 
 insufficiency is casein. It may be used in all cases, whether the 
 stomach is functioning normally or not, for it alone among the 
 proteids appears to be broken down, without any prehminary pre- 
 paration, by the ferment " erepsin " discovered by Cohnheim in the 
 succus entericus. It may be given in the form of milk, or in larger 
 quantities as one of the artificially prepared powders, biscuits, &c., 
 which can be obtained free from sugar. The digestion of both fats 
 and proteins can be assisted by the administration of some active 
 preparation of pancreas two or three hours after each meal, but it is 
 important to make certain that the preparation selected is active 
 and contains a lipolytic ferment, as so many useless pancreatic 
 extracts are now on the market. 
 
 Pancreatic and other digestive ferments such as taka-cUastase 
 papain, &c., are often useful even when there is no evidence of 
 j)ancreatic insufficiency, since they prevent the stagnation of food 
 in the intestine with consequent excessive putrefactive changes, 
 and the absorption of toxic products. The latter are probably 
 the primary cause of the glycosuria in some instances, and in others 
 no doubt accentuate the concUtion. Regular and satisfactory 
 movements of the bowels, with attention to the digestive tract 
 generally, are nearly always points that repaj^ careful attention. 
 
 Tea and coffee may be taken in moderation by all diabetics, 
 and form a useful vehicle for the administration of cream. They 
 should, of course, be unsweetened, or be sweetened with saccharin. 
 Lemon- juice wdth plain aerated water and sweetened with saccharin, 
 or a lemonade made with citric acid ( 10 grains) , glycerine (4 drachms) , 
 and water (1 pint) may be taken when desired, but the most 
 generally useful drink is probably Vichy (Celestins) water. This 
 contains chiefly sodium bicarbonate, with smaller quantities of 
 sodium chloride, calcium carbonate, potassium bicarbonate, 
 magnesium carbonate, and sodium sulphate, and is therefore 
 useful in counteracting acidosis. Its effect may be increased by 
 adding sodium bicarbonate (15 to 30 grains), sodium benzoate 
 (4 to 8 grains), and lithium benzoate (4 to 8 grains) to each tumbler- 
 ful. A large amount of Hquid should be avoided, as it increases 
 diuresis. Excessive thirst is most effectually controlled by regulating 
 the diet so that the excretion of sugar in the urine is reduced. 
 
 Each case of diabetes is a law unto itself, and can only be 
 satisfactorily treated when all the data obtained by a thorough 
 investigation are available ; but it may be laid down as a general 
 
PERSISTENT GLYCOSURIA 
 
 329 
 
 rule that, since acidosis is the chief danger to be apprehended in 
 severe cases, it is unwise to put the patient on too strict a diet, and 
 that he should not be kept on a carbohydrate-free diet for am^ 
 length of time. Success depends on the establishment of a tolerance 
 for proteins, and until this has been obtained tolerance for carbo- 
 hydrates cannot be regained. 
 
 The plan of arranging the diet in periods, introduced by von 
 Noorden, is very helpful. By this method the patient is first given 
 a diet containing a small quantity of bread, or other starchy food, 
 the amount varying with the type of case ; after two or three days 
 he is put on a low protein diet, consisting chiefly of vegetables, eggs, 
 and fatty foods, such as that shown in the table : — - 
 
 Vegetable Diet 
 
 Grams. 
 
 Oz. 
 
 
 Protein. 
 
 Fats. 
 
 Carboby. 
 
 Calories. 1 
 
 225 
 
 8 
 
 Coffee 
 
 
 
 
 
 40 
 
 U 
 
 Cream (2 table.sp.) 
 
 1-4 
 
 10-0 
 
 1-4 
 
 100 
 
 100 
 
 3+ 
 
 Lettuce ..... 
 
 0-9 
 
 0-3 
 
 2-9 
 
 17 
 
 25 
 
 3 
 4 
 
 Onion [^ large) .... 
 
 03 
 
 0-4 
 
 1-2 
 
 10 
 
 200 
 
 7 
 
 Tomato (1 large) 
 
 2-4 
 
 0-4 
 
 8-0 
 
 46 
 
 22 
 
 3. 
 
 French j'4 tablesp. 01. oil . 
 dressing -[ 1 ,, vinegar 
 (2dessertsp.) \\ teasp. salt, pepper 
 
 \ 
 
 
 
 
 
 
 \ ••• 
 
 16-0 
 
 
 148 
 
 
 
 J 
 
 
 
 
 , 50 
 
 2 
 
 Egg (1 average) .... 
 
 6-6 
 
 6-0 
 
 
 83 
 
 200 
 
 7 
 
 Spinach (4 hpd. tablesp.) 
 
 5-4 
 
 0-6 
 
 60 
 
 36 
 
 15 
 
 i 
 
 Margarine ..... 
 
 0-2 
 
 11-8 
 
 
 111 
 
 249 
 
 8 
 
 Broth (bouillon) .... 
 
 .■•3 
 
 0-5 
 
 0-5 
 
 26 
 
 120 
 
 4i 
 
 Cauliflower (2 hpd. tablesp.) 
 
 1-1 
 
 0-1 
 
 0-5 
 
 8 
 
 15 
 
 4 
 
 Margarine ..... 
 
 0-2 
 
 11-8 
 
 
 111 
 
 56 
 
 2 
 
 Bacon (weighed uncooked) 
 
 5-6 
 
 39-0 
 
 >.. 
 
 378 
 
 60 
 
 2 
 
 Brazil nuts (10 large) . 
 
 10-2 
 
 40-1 
 
 4-2 
 
 432 
 
 30 
 
 1 
 
 Lemon juice (2 tablesp.) 
 
 
 
 3-0 
 
 12 
 
 225 
 
 8 
 
 Tea 
 
 
 
 
 
 40 
 
 U 
 
 Cream (2 tablesp.) 
 
 1-4 
 
 lO-O 
 
 1-4 
 
 100 
 
 56 
 
 2 
 
 Bacon (weighed uncooked) . 
 
 5-9 
 
 39 
 
 
 378 
 
 22 
 
 -| 
 
 French dressing (2 dessertsp.) . 
 
 
 16-0 
 
 
 148 
 
 100 
 
 31. 
 
 Endive or lettuce 
 
 1-0 
 
 0-2 
 
 30 
 
 16 
 
 ICO 
 
 3* 
 
 Asjjaragus (or French beans) 
 
 1-5 
 
 0-1 
 
 2-8 
 
 18 
 
 15 
 
 1 
 
 Margarine 
 
 0-2 
 
 11-8 
 
 ... 
 
 111 
 
 
 49-6 
 
 214-1 
 
 34-9 
 
 2309 
 
 Carbohydate 
 
 X, . • /49-6 ^ 
 rroteni sugar x 5 
 
 Total sugar value 
 
 34 9 
 39-7 
 
 74-6 
 
230 GLYCOSURIA 
 
 This in its turn is replaced in another two or three days by a diet 
 of oatmeal, eggs, and butter or margarine, ^y repeating this pro- 
 cedure for two or three weeks the acidosis is frequently so much 
 lowered that it ceases to be an immediate danger. The subsequent 
 treatment then depends upon the carbohydrate tolerance of the 
 patient. In some instances marked improvement will result from 
 a week or two on a carbohydrate-free diet, in others the best results 
 are obtained hj allowing a certain amount of starchy food, while 
 others again are more benefited by a prolonged period of vegetable 
 or oatmeal feeding. Finally a general diet is worked out on the 
 Unes previously indicated for the milder tj^^De of case, care being: 
 taken, however, to warn the patient that the protein of the food 
 must be strictly hmited to the amount shown on his chart. 
 
 Whatever diet the patient is eventually put upon, the risk of 
 overtaxing his powers of protein metabolism should be avoided by 
 arranging a period of two or three daj^s on vegetable-fat cUet every 
 few weeks, and once or twice a year a longer period of two or three 
 weeks on similar food should be prescribed. 
 
 Org'ano-therapy. — The most serious difficulty in the way of 
 the satisfactory treatment of persistent glycosuria has alwaj's been 
 the obscurit}^ of its etiology. When it was proved that excision of 
 the pancreas in animals gives rise to symptoms closely resembhng 
 those of human diabetes, and that disease of the gland is found post- 
 mortem in some patients who have presented such symptoms during 
 life, it appeared that a solution of the difficulty was at hand. The 
 satisfactory results following the administration of thjTToid extract 
 in myxoeclema and sporadic cretanism naturalty suggested that 
 the use of pancreatic extracts or of the fresh gland might be equally 
 effectual in diabetes, but the results of such treatment have proved 
 most disappointing, and even in those cases of glycosuria where 
 disease of the pancreas was undoubtedly^ present, the oral adminis- 
 tration of fresh or prepared pancreas has, in nearly every instance, 
 failed to produce the hoped-for result. The majority of observers 
 are agreed that, although some improvement of the digestive powers 
 may follow, the glycosuria and other symptoms of the diabetic 
 condition are uninfluenced. In a few instances it has been claimed 
 that some amelioration of the symptoms has been produced. Thus 
 Cowles has reported a case in which an average of over three fresh 
 pancreases were consumed a day, with the result that not only 
 did an improvement in appetite, strength, and weight occur, but 
 the thirst was less, the quantity and specific gravity of the urine 
 fell, and the sugar diminished, although it never c^uite disapj)eared. 
 The diet was not restricted much, and nothing but sugar was for- 
 
PERSISTENT GLYCOSURIA 331 
 
 bidden. Eventually the patient stopped the treatment, the symp- 
 toms returned, a large carbuncle developed on the neck, and he 
 died. Post-mortem, the pancreas was found to be represented 
 by a fibrous cord about one-fourth the size of the normal gland. 
 The same treatment was tried with other patients without benefit, 
 but Cowles states that none of these were able to eat such a large 
 quantity of pancreas or to continue the treatment as long. I have 
 met with one case of diabetes in a child of ten, in which a temporary 
 improvement in the general condition and a reduction of the sugar 
 in the urine by nearly half followed the administration of fresh 
 pancreas. Spooner and Pratt state that the limit of assimilation 
 is raised in animals with experimental atrophy of the pancreas by 
 feeding with the fresh gland. In a dog whose pancreas was entirely 
 separated from the duodenum they determined the assimilation 
 limit at frequent intervals for a year, and found that it was never 
 over 35 grams, but after the administration of three sheep's pan- 
 creases a day for six weeks the limit of assimilation rose to 80 grams, 
 and continued to rise for a short time after the pancreas feeding was 
 discontinued, the maxinumi reached being 100 grams. 
 
 Arguing that pancreatic preparations given by the mouth are 
 destroyed in the stomach, it has been suggested that they should be 
 administered in capsules that will protect them from the action of 
 the gastric secretions. Some observers, and notably Crofton, have 
 reported satisfactory results, even Avhen no alteration was made in 
 the diet, but it has been generally found that this method gives no 
 better results than the preceding. I have sj^stematically given 
 various pancreatic extracts that laboratory experiment has shown 
 to contain active ferments, in various ways by the mouth, and 
 although the digestion and general condition of the patient has 
 frequently shown a marked improvement, especially when there 
 has been evidence of pancreatic insufficiency, I have not found that 
 the glycosuria or secondary disturbances of metabolism were 
 favourably influenced. 
 
 Subcutaneous injection has been resorted to by some observers, 
 with a view to avoiding destruction of the ferments contained in 
 the extract by digestion in the ahmentary tract, but with equally 
 unsatisfactory results. AccorcUng to Leschke, fresh extracts of 
 pancreas are not only useless, but increase the elimination of sugar 
 in diabetic persons and animals, and also induce glj'cosuria in normal 
 animals, with a toxic and often fatal result. Basing his procedure 
 on experiments which suggested that the hormone of the pancreas 
 is antagonistic in its action to the hormone of the supra-renals. and 
 that in pancreatic diabetes the lack of the pancreatic secretion and 
 the predominance of the internal secretion of the supra-renals 
 
332 GLYCOSURIA 
 
 explains the glycosuria, Zuelzer and his associates have attempted 
 to treat diabetes with exjDressed extracts of the pancreas taken at 
 the height of digestion and rendered less toxic by removing the 
 albumens wdth alcohol. Satisfactory results were reported in 
 several cases, the urine being freed from acetone and aceto-acetic 
 acid and the sugar much reduced in amount. Employing the same 
 method, Forschbach was able to confirm Zuelzer's conclusions with 
 depancreatised clogs, but with human diabetics found that although 
 the sugar excretion was temporarily diminished, the acetone bodies 
 were not affected, the temperature was raised, and the patient 
 became acutely ill. He is inclined to attribute the diminished 
 glycosuria to the accompanying fever, and considers that the 
 pancreatic hormone is too dangerous for practical emplojmaent in 
 diabetes. With the idea of restoring the balance between the 
 internal secretions of the pancreas and supra-renals, and working 
 on the principle by which exophthalmic goitre is treated with the 
 serum or milk of thjnroidectomised animals, Briick has suggested 
 that the missing neutrahsing pancreatic secretion might be supphed 
 to diabetics by administering the milk or serum of animals in which 
 adrenalin has been excluded from the circulation, but I am not 
 aware that his suggestion has been carried out in practice. 
 
 When considering experimental diabetes, we saw that Minkowski 
 proved that if a portion of the pancreas were implanted in the 
 subcutaneous tissue of the abdominal wall, it sufficed to prevent 
 glycosuria when the remainder of the gland was removed. In 
 1908 he pubhshed a full report of the effects of grafting pancreatic 
 tissue in dogs, and conclusively j)roved that if the graft secures a 
 sufficient blood supply it grows and functions to such an extent 
 that the animal's own pancreas can be complete^ removed without 
 the occurrence of diabetes. It would seem that this procedure 
 presents great therapeutic possibihties, but I can only find records 
 of two cases in which treatment on these fines has been attempted. 
 One was reported by Watson Wilfiams and the other by J. W. Allan. 
 In the former case a sheep's pancreas was grafted into a patient 
 suffering from severe cfiabetes, but without any result, death taking 
 place a fortnight after the operation from diabetic coma. Post- 
 mortem, the pancreas was found to be " large and apparently quite 
 healthy to the naked eye." In the other case a cat's pancreas was 
 used, but in this too failure resulted, the pancreas dying and 
 sloughing out. It may be pointed out that in both these cases 
 the operation was not attempted until a late stage, and Minkowski 
 has shown that in animals the transplantation must be undertaken 
 before diabetes has been induced, as othenvise the graft does not 
 grow and the wound fails to heal. 
 
PERSISTENT GLYCOSURIA 333 
 
 If the view that the islands of Langerhans are the source of 
 the internal secretion of the pancreas is correct, it would seem 
 probable that an extract prepared from them would have more 
 effect than an extract of the whole gland. With this object Rennie 
 and Eraser treated five diabetics with a preparation of the macro- 
 scopic chief islands found in certain fishes, and state that some 
 improvement followed its use. Their results have not, however, 
 been so far confirmed, and the authors themselves have not pub- 
 lished any additional cases. 
 
 The observation of Hedon that the normal pancreas only checks 
 glycosuria when it is so placed that its internal secretion enters 
 the portal circulation directly, offers a reasonable explanation of 
 the almost uniformly negative results of the various attempts that 
 have been made to treat diabetes with preparations of pancreas, 
 and appears to make the reahsation of a specific treatment along 
 these lines more difficult than ever. It would seem, however, 
 that there is a group of cases in which the administration of pancreas 
 by the mouth materially improves the condition of the patient, 
 but as a rule there is in these some interference "wdth the flow of 
 pancreatic juice into the intestine. 
 
 Basing their treatment on the effects of secretin as a stimulant 
 of the pancreas, Moore, Edie, and Abram employed an acid extract 
 of duodenal mucous membrane in diabetes. They found that 
 when this was given by the mouth, the sugar in the urine gradually 
 diminished in some cases and finally disappeared in a few. In 
 others, although there was an improvement in the digestion, no 
 effect on the sugar output was produced. Bainbridge and Beddard, 
 however, noticed no amehoration of the symptoms in the cases 
 that they treated by this method, and suggested that any improve- 
 ment that takes place is to be attributed to the diet and not to the 
 secretin. Bainbridge also found that the yield of secretin from the 
 duodenal mucous membrane is almost or quite as great in diabetic 
 as in non-diabetic people, and attributes the failure of some other 
 observers to find it to its destruction from rapid post-mortem changes. 
 J. R. Charles, N. B. Foster, and other observers state that in their 
 experience the administration of secretin does not affect the glyco- 
 suria. Even if it could be shown that the glycosuria could be con- 
 trolled by this means in some cases, it is open to question whether 
 the treatment might not in the long run do more harm than good, 
 for the excessive artificial stimulation of the cUseased tissue that 
 may remain in pancreatic diabetes, although it may at first induce 
 increased activity, is hkely to eventually bring about fatigue and 
 cause more rapid degeneration than would have occiu-red if it had 
 been left alone. The intravenous injection of secretin has been 
 
334 GLYCOSURIA 
 
 suggested, but it has been shown by Starhng that it gives rise to 
 acute inflammation of the intestine, and even to gastric ulcers in 
 animals, as the pancreatic juice is not met and neutrahsed by the 
 acid gastric contents which normally cause the flow. This ob- 
 jection does not apply to its administration by the mouth, as the 
 resulting secretion is then gradual, and corresponds to the acidity 
 of the gastric juice reaching the intestine. 
 
 Cohnheim's work wpoii the effects of a mixture of pancreatic 
 .and muscle extracts in glycolysis has suggested the use of such a 
 mixture in diabetes, and it has been employed for this purpose by 
 Crofton. vSewall, and others. 
 
 When considering the alleged favourable results obtained with 
 any form of treatment in diabetes, it is important to bear in mind 
 that the clinical course of the condition is at times subject to con- 
 siderable variation under a great variety of conditions, and that 
 favourable results have been claimed to be produced by a number 
 of remedies that have probably no specific relation to the patho- 
 logical processes present. The effects of alterations in the general 
 hygiene, surroundings, and diet of the jjatient have also to be 
 allowed for. 
 
 Drug's. — The drugs that have been employed at various times 
 in the treatment of diabetes are exceedingly nvunerous, but only 
 very few are generally acknowledged to be of service. Even these 
 are not ciu-ative. and improve the condition of the patient most 
 Avhen the}' are employed in conjunction with careful restriction of 
 the diet. 
 
 Opinm and its derivatives, morphine and codeine, are not infre- 
 quently employed, and have stood the test of time. It is found 
 that as a rule they control the glycosuria and poh^'uria more 
 effectually than any other drug when they are given in sufficiently 
 large doses. In some cases however, opium, like all other remedies, 
 is useless. 
 
 Codeine is probably the opium preparation most frequently 
 prescribed in persistent glycosuria, since it does not tend to make 
 the patient sleepy, nor does it affect the bowels as readily as opium 
 or morphine. At first it should be given in small doses, a quarter 
 to half a grain three times a daj^, and then be gradually increased 
 until the glycosuria is controlled or the patient's Hmit of tolerance 
 for the drug is reached. Often 12 grains or so may be given in thi^ee 
 doses during the day without producing any untoward result, but 
 in msmy instances 6 or 8 grains maj^ cause the patient to lose ground 
 and give rise to diarrhoea. Codeine maj^ be administered in pill 
 form combined "v^dth cascara sagrada : — 
 
PERSISTENT GLYCOSURIA 335 
 
 R Codeinse . . . . . gr. J to | 
 
 Ext. Case. Cas. . . . gr. 2 
 
 Pulv. Glyc. . . . . gr. 2 
 
 Ext. Gent. . . . . q.s. 
 
 or with cascara and nux vomica — 
 
 R Codeinae . . . . gr. J to | 
 
 Ext. Nuc. Vom. . . . gr. |^ 
 
 Ext. Case. Sag. . . . gr. | 
 
 Codeine is, however, considered by some authorities to be inferior 
 in its effects on sugar excretion to opium and morphine, and there 
 can be no doubt that these sometimes succeed when codeine fails ; 
 moreover, as it quickly loses its first sedative effect, it has often to 
 be replaced by other drugs after a short interval. 
 
 Morphine is much less expensive than codeine. According to 
 Mitchell Bruce, acetate of morphia is the most useful salt, and acts 
 better when given by the mouth than when injected subcutaneously. 
 A small dose, one-sixth of a grain or so, three times a day, is suffi- 
 cient to commence with. The dose is gradually increased until one 
 grain, or even more, is being taken three times in the twenty-four 
 hours. 
 
 Opium is believed by many observers to give more uniformly 
 satisfactory results than either of its alkaloids. Starting with half 
 a grain or its equivalent {e.g. Pil Saponis Co., gr. 2 to 4), three 
 times a day, the dose is gradually increased until 12 or 15 grains 
 are being taken daily. Some prefer the watery extract of opium, 
 which can be given in much the same doses as codeia. 
 
 According to Ralfe, opium has the greatest effect in restraining 
 diuresis when taken about one hour after meals, and is also then 
 less likeh^ to cause dyspeptic symptoms or derange the stomach. 
 Although diabetics are very tolerant of opium and its alkaloids, 
 so that large quantities can be taken without producing bad results, 
 they often exhibit individual peculiarities with regard to its different 
 preparations,. Opium for instance, in the form of Pil Saponis Co., 
 can sometimes be taken without discomfort, when morphine or 
 codeine will give rise to headache and giddiness. Occasionally 
 the best effects are obtained by combining crude opium with one or 
 other of its alkaloids. In fixing the dose it must be remembered 
 that this will vary with different cases. As a rule the administra- 
 tion should be pushed until the glycosuria is controlled, or no further 
 reduction in the sugar excretion and volume of urine follows an 
 increase in the dose ; when this point is reached the dose should be 
 maintained at that level. Its effects should always be carefully 
 watched, particularly with regard to the production of constipa- 
 tion and dyspeptic symptoms. When these cannot be otherwise 
 
338 GLYCOSURIA 
 
 controlled, its administration must be discontinued. Since con- 
 stipation is often associated with severe acidosis and threatening 
 coma, opium must be used with great care in such cases. Landergren 
 states, however, that in certain cases opium markedly reduces the 
 amount of acetone and oxybutyric acid in the urine, and claims 
 that in threatening coma opium may often prove of value by 
 diminishing the acidosis. When nephritis is present opium should 
 be avoided, or be given with very great caution. Opium, like its 
 alkaloids, gradually loses its effect, although not so rapidly as a rule. 
 
 The way in which opium and its alkaloids act in diabetes is not 
 definitely understood. According to Lepine it produces an effect 
 on the hver, through the nervous system, preventing the excessive 
 production of sugar. Von Mering and Minkowski suggest that it 
 inhibits the formation of sugar from albumen. Roberts considers 
 that it diminishes the appetite, and so less sugar is excreted in the 
 urine. Others hold that it reduces the level of metabolism generally. 
 In view of the theories that are at present held with regard to the 
 influence of the ductless glands on carbohydrate metaboHsm and 
 their relation to the nervous system, its effect as a nervine sedative, 
 and its known action on vaso-motor system and on glandular 
 secretion, offer the most hkely explanation of the beneficial results 
 that follow its use in most cases. 
 
 Belladonna, either alone or in combination with opium, appears 
 to be useful in some cases, but owing to the dryness of the throat, 
 to which it gives rise, it cannot usually be tolerated for any length 
 of time. Rudisch has strongly recommended the use of atropin 
 methylbromide in diabetes, and states that when given in fairly 
 large doses in conjunction with dietetic treatment it greatly 
 aids in lessening or preventing glycosuria. Forscheimer found 
 that the same results were obtained whether atropine methyl- 
 bromide, atropine sulphate, or belladoima was used, and that not 
 only did the glycosuria diminish in many cases, but that the ex- 
 cretion of acetone bodies also lessened and carbohydrate tolerance 
 was increased. According to this author belladonna is more 
 particularly adapted to mild cases, and does little good in the severe 
 types. In the hands of other observers this method of treatment 
 has not been so satisfactory. Mosenthal, for instance, found that 
 atropine sulphate effects no change in carbohydrate tolerance of 
 sufficient importance to make the drug of chnical value in the 
 treatment of diabetes. Experimenting with depancreatised dogs, 
 Wallace found that the drug had no effect on svigar excretion. 
 
 Pilocarpine injections have been stated to cause a diminution in 
 the amount of urine and sugar excreted in some cases, but in others- 
 they have been found to do no good. 
 
PERSISTENT GLYCOSURIA 337 
 
 Antipyrin was suggested as a remedy in diabetes by Gonner, and, 
 according to Dujardin-Beaumetz, in doses of 30 to 60 grains daily 
 it diminishes the quantity of urine without there being any increase 
 in the percentage of sugar. Its action is, however, only temj)orary, 
 and, as it is liable to give rise to digestive disturbances and 
 albuminuria, it should not be used for long. Pyramidon is said 
 to be less likely to upset the stomach, although it has not such 
 a marked effect on the urine as antipyrin. Many observers have 
 found both drugs useless, but as there appear to be marked idio- 
 syncrasies, and the results vary very much with different patients,, 
 they are worth a trial when other methods of treatment fail. 
 
 Bromides. — In 1866 Begbie employed potassium bromide in 
 the treatment of two cases of diabetes with success. Since then it 
 has been used by other physicians, and has been recommended 
 by V. Noorden, Osier, and Saundby in cases where there is great 
 nervous irritabihty or excitement. It does not appear to have as 
 marked an effect in such cases as might be expected, and, owing 
 to its depressing action in many cases, its use has frequently to 
 be discontinued. Ammonium, or lithium, bromide may be sub- 
 stituted, and do not so readily give rise to depression. Valerian, 
 which was recommended by Trousseau for polyuria and is a nerve 
 stimulant, may be given, in the form of the ammoniated tincture, 
 to counteract the dejDressing effect of the bromides. 
 
 Bromides, antipyrin, and other drugs which depress the nervous 
 system are to be avoided in cases where the patellar reflex is absent 
 or weak. 
 
 Arsenic. — Experiments on animals have shown that the ad- 
 ministration of arsenic in sufficient doses, for an adequate time, 
 causes the glycogen to disappear from the liver, and that puncture 
 of the floor of the fourth ventricle will not then give rise to glycosuria. 
 Toxic effects must, however, be produced, and ordinary medicinal 
 doses are of no use. Varying results have been reported from the 
 use of arsenic in clinical work on diabetes. Some physicians have 
 stated that, after the excretion of sugar has been reduced by diet 
 and opium, arsenic will effect a cure, while others have found that, 
 although it is useful in some mild cases, it has no effect in the 
 severer forms. Forscheimer, on the other hand, states that it is 
 especially indicated in severe cases, but should always be combined 
 with diet. He considers that it is also useful in neurotic, debihtated 
 subjects. He points out that to get the best results mild toxic 
 effects must be produced. Gradually ascending doses are given 
 until this result is obtained, and the dosage is then gradually re- 
 duced. Arsenic used in this manner is said to diminish the glycos- 
 uria and acetonsemia, but does not increase sugar tolerance. Unhke 
 
 Y 
 
338 GLYCOSURIA 
 
 oj^iiim the effect is not lessened by prolonged administration, but 
 when the drug is discontinued a gradual falHng off is observed. A 
 repeated course of the treatment, however, again controls the glycos- 
 uria and excretion of acetone bodies. Arsenic has been usually 
 ^ven in the form of hquor arsenicaHs in diabetes, but arsenite of 
 bromine (Clemens' solution), in doses of 3 to 5 minims once or twice 
 a daj^ after meals, has also been employed, and combinations of 
 lithium and sodium arsenate in pill form, or dissolved in aerated 
 water, have been used, especially in gouty cases. Dujardin- 
 Beaumetz recommends 5 grains of lithium carbonate and 2 minims 
 of hquor arsenicalis in a glass of Vichy, or other alkahne, water. Le 
 Gendre advises five or six pills a day. according to the toleration, 
 of the follo-wdng composition : — 
 
 R Strychnine sulphate 
 Sodium arsenate . 
 Codeine .... 
 
 Quinine valerinate 
 Extract of valerian . . . q.s. 
 
 in cases where there is depression of the nervous system, and the 
 patellar reflex is absent or weak. 
 
 Quinine. — According to Lepine, quinine has an anti-glycogenic 
 action, and is of use in diabetes for this reason. Although some 
 observers have stated that it chminishes the glycosuria, it is gene- 
 rail}^ held that it merely acts as a general tonic, and is therefore only 
 useful in cases where a stimulant Hne of treatment is indicated. 
 It may also be of service when the symptoms of diabetes have 
 followed attacks of malaria, or the patient has resided in a malarious 
 neighbourhood. 
 
 Anti-syphilitic Treatment. — In a small proj)ortion of diabetics 
 the condition appears to be due to pathological changes set up in 
 the pancreas, and possibly also in the nervous system, by syphihs. 
 Peinburg and v. Noorclen have recorded such cases, and found that 
 improvement was frequently brought about by treatment with 
 mercury and potassium iodide. I have had two patients under my 
 care in which persistent glycosuria followed infection with syphilis, 
 and in both a course of anti-s^^ihihtic treatment much improved 
 the general condition and reduced the glycosuria, although the 
 urine was not rendered sugar-free in either. As mercurial stomatitis 
 and intestinal catarrh are very hable to develop in diabetics, the 
 patient should be constantly watched, and the treatment be care- 
 full}^ regulated. Von Noorden states that in several of his cases 
 fatal comphcations, including gangrene of the foot, haemoptysis, and 
 rapidly progressing tuberculosis, occurred during the mercurial 
 course. 
 
PERSISTENT GLYCOSURIA 339 
 
 Iodide of 'potassium has been tried as a remedy in cases of diabetes 
 not of syphilitic origin, but Dickinson, as the result of a series of 
 careful observations, came to the conclusion that any diminution 
 in the excretion of sugar that occurs is due to loss of appetite and 
 general depression of function, and not to real mitigation of the 
 disease. 
 
 Iodoform was found by Moleschott and Frerichs to cause a 
 temporary improvement in some cases, but other observers have 
 stated that no good effect followed its use. 
 
 Uranium nitrate has been recommended by Hughes, and later 
 by West, who state that it caused a marked diminution in the ex- 
 cretion sugar, and also diminished the thirst and amount of urine 
 in the cases where they tried it. By starting with small doses of 
 1 or 2 grains, freely diluted with water, twice a day after meals, the 
 danger of digestive disturbances and albuminuria is avoided. The 
 dose is gradually increased every few days until the desired effect 
 on the urine is produced. In some cases as much as 15 to 20 grains 
 have ultimately been given three times a day. 
 
 Intestinal Antiseptics. — At one time diabetes was thought by 
 some observers, to be a specific infectious disease and attempts 
 were made to reproduce the condition in dogs by feeding them 
 with faecal material from diabetic patients, but without result. 
 According to Herter, glycosuria has been produced in cats and dogs 
 by intravenous injections and by feeding experiments with bacteria 
 isolated from diabetic stools, but the data are so scanty that the 
 evidence cannot be regarded as conclusive. Spontaneous diabetes 
 melhtus has been described in dogs by Friedberger, Fronner, 
 Schnidelka and Eichorn, and others. Some five years ago I 
 isolated a sugar, having the reactions of dextrose and yielding 
 dextrosazone crystals, from the urine of a pet dog whose mistress was 
 suffering from diabetes. The condition has also been met with in 
 horses by Heim, Rueff, and Dieckerhoff , and in monkeys by Leblanc. 
 Cases of apparently infectious origin have been reported by several 
 observers. Teissier, at the first French Congress of Medicine, 
 reported the case of a washerwoman aged sixty-two who became 
 glycosuric after having washed for six months the Hnen of a diabetic, 
 and her little girl also became affected with the disease. There is, 
 too, the case of a gouty patient whose mother had been diabetic for 
 twenty years, and who developed the disease in his turn ; six months 
 later the cook, who had washed the handkerchiefs of her master, 
 became glycosuric ; and finalty, the disease showed itself in a woman 
 of fifty employed by the family for ten years to mend clothes. 
 A coachman became glycosiu-ic shortly after his master, and a 
 restaurant- keeper who took his meals at the same table with a 
 
340 GLYCOSURIA 
 
 diabetic sister-in-law became diabetic at the end of six years. 
 Kuelz has reported a case of conjugal diabetes in which four persons 
 occupying the same house became diabetic, and Naunyn observed 
 five cases of diabetes under the same roof. Senator, at a meeting 
 of the Berlin Medical Society in 1908, quoted the case of a doctor 
 aged fortj^-two who became diabetic after amputating the thigh of 
 a diabetic patient. He discovered that there were four cases of 
 diabetes in the immediate neighbourhood, all in the same street, 
 and all taking their meals at the same restaurant, which was kept 
 by a glycosuric. 
 
 The most striking evidence in favour of an infectious origin in 
 glycosuria is furnished by the occurrence of diabetes in husband and 
 wife (" Conjugal Diabetes "). The possibihty of this occurrence 
 has been supported by Debove, in France, who, out of fifty cases of 
 diabetes, found five in which husband and wife were both affected, 
 and similar cases have been recorded by Dreyfous, Gaucher, Labbe, 
 Rendu, and Schmitz. The last had a very large practice at Neuenahr, 
 in which he had seen an immense number of cases of diabetes, and 
 was certainly inchned to believe in the possibihty of transmission. 
 Senator, in the discussion referred to, stated that he had collected 
 the histories of 516 married pairs, in which either husband or wife was 
 diabetic, and in eighteen cases the second partner had become dia- 
 betic, giving a proportion of 3-5 per cent, cases of conjugal diabetes. 
 Although we have no data by which we can fix accurately the period 
 of incubation of diabetes, supposing it to be an infective cUsease, 
 yet as we know that traumatic diabetes develops in at least six 
 months after the injury, and usually in less time, he therefore accepts 
 that period, and excludes from his hst all those marriages which had 
 lasted for only six months or less ; in this way the percentage of 
 cases rose to 3-7, while by ehminating other seventy-four couples, 
 whose marriages dated from less than a year, the percentage rose to 
 4-1. He admits that it would be right to exclude all cases with dis- 
 tinct hereditary predisposition ; but taking the figures as they are the 
 proportion is too feeble to support the theory were it not that th.ere 
 are other facts in its favour. In the course of the discussion 
 Neumann of Potsdam said that during the previous five years, with 
 the aid of his colleagues and the chemists of the town, he had collected 
 180 cases of cUabetes from a population of 59,881 ; among these 
 he found only three instances in which both married partners were 
 affected, and two of these should be excluded, as in one case the 
 wife was pentosuric only, and in the other the husband became 
 diabetic three j^ears after the death of his ^vife. In the single 
 remaining case the wife was diabetic and suffered from Graves' 
 disease, and the husband developed the disease after an accident. 
 
PERSISTENT GLYCOSURIA 341 
 
 to her. Albu said he had seen no case to justify the view that 
 diabetes was contagious, but he knew of two instances ilhistrating 
 the fallacies surrounding such apparent cases ; in one the wife 
 became diabetic after her husband, and at the time Albu knew of 
 no hereditary tendency on her side, but some years after he treated 
 her nephew for diabetes and obtained unquestionable evidence of 
 the family predisposition. In the other case the glycosuria de- 
 veloped by the second of the married pair turned out to depend upon 
 cancer of the pancreas. Finally, Ewald stated he had met with no 
 case of conjugal diabetes, and, considering that the statistics pre- 
 sented ranged from 1 to 20 per cent., he thought the entire relation 
 was accidental. In all such statistical inquiries the influence of 
 heredity must be borne in mind. In Germany Jews are numerous, 
 and amongst them diabetes is so common that few families escape 
 it altogether, and as they only marry with their own people, among 
 married Jews suffering from diabetes there should be a distinct 
 l^roportion of cases in which both partners show the disease. On 
 the other hand, races not predisposed to diabetes, among whom the 
 annual mortality from it does not exceed 6 or 7 per 100,000 of 
 population, the probability of both members of a married pair 
 developing the disease becomes infinitely small, and it is in accord- 
 ance with this view that so-called conjugal diabetes is very rarely 
 met with in this country. 
 
 Although a specific infectious theory for all cases of diabetes 
 has now been abandoned by nearly all observers, there are many who 
 consider that in some instances, and particularly where disease of 
 the pancreas is the cause of the condition, an abnormal intestinal 
 flora may be the origin of the mischief. The evidence in favour of 
 this view is at present rather cUnical and inferential than experi- 
 mental, but it is not inconsistent with modern views with regard 
 to diabetes, and furnishes a basis for a preventive, if not for a 
 curative treatment. Certain drugs which have been employed in 
 the treatment of diabetes with satisfactory results are intestinal 
 antiseptics, and it is probable that their effect is largely due to 
 this fact. 
 
 Sodium salicylate has been occasionally used for twenty years 
 or more. Some observers, notably Ebstein, have reported a re- 
 duction in the excretion of sugar, and a general improvement in 
 the condition of patients to whom it had been administered. 
 Ebstein considers that it is chiefly useful in recent cases, and advises 
 that it should be given in large doses 75 to 150 grains in the twenty- 
 four hours. Brunton and Ralfe state that it is chiefly useful in 
 gouty glycosuria. Sahcylate of bismuth, in the form of 71 grain 
 powders twice a day, has been employed by Schmitz with good 
 
342 GLYCOSURIA 
 
 results. Other observers have employed phenjd salicylate (salol), 
 or acetyl sahcyhc acid (aspirin). These drugs are all recognised 
 as more or less efficient intestinal antiseptics, and, according to 
 Crow, salicylates are excreted in the pancreatic juice and bile after 
 absorption. Salicylates are also said to stimulate the activity of 
 the thyroid. When giving these drugs, or any other saUcylate 
 compound, it must be remembered that the urine will give a purple 
 coloration with perchloride of iron, which is liable to be mistaken 
 for the similar reaction with aceto-acetic acid, the latter is, however, 
 diminished after the urine has been boiled for some minutes, whereas 
 the sahcylate reaction persists unchanged. As most saUcylate com- 
 pounds have a tendency to produce constipation, the condition of 
 the bowels should be watched during their administration. 
 
 Hexamethylenamine {urotropine) is one of the most generally 
 useful antiseptic drugs for internal administration that we possess. 
 The experimental investigations of Crow have demonstrated that 
 it is absorbed into the circulation, and can be found in the blood for 
 twenty-four hours after its administration by the mouth. It is 
 excreted in the urine, by all the mucous surfaces, in the bile, synovial 
 fluid, pancreatic juice, saHva, plural and cerebro-spinal fluids, &c. &c. 
 A dose of 75 grams a day will prevent bacterial growth in the bile 
 and pancreatic ducts. It is therefore exceedingly iiseful in con- 
 trolling infections of these and other regions, and it is possible 
 that to this its effect in diabetes is due. Forchheimer used it 
 in 5 grain doses in the treatment of diabetes, and found that the 
 glycosuria was, in many instances, reduced, and the carbohydrate 
 tolerance improved. He considers that it is especially indicated in 
 cases where dietetic regulation is impossible, but does not recommend 
 it in severe cases. It is to be noted that after the administration 
 of this drug the urine may reduce copper solutions, but it does not 
 generally reduce bismuth, and never ferments with yeast. 
 
 Yeast is a very old remedy for diabetes. Caessaet and WilHam- 
 son both report good results from its use. The latter gives it 
 in doses of two dessertspoonfuls in water, but points out that it 
 must be fresh and free from hquid, or severe diarrhoea may be 
 produced. Any improvement that follows is, however, only 
 temporary. 
 
 Castor-oil, Carlsbad salts, and other laxatives are frequently 
 required in the treatment of diabetes. Constipation is one of the 
 difficulties that calls for most constant attention, and if the 
 bowels are carefuUy regulated it will be found that most forms of 
 treatment prove more satisfactory. 
 
 The Massive Saline, or Fasting-purgation (Guelpa) Treat- 
 ment. — Taking the view that diabetes is an auto-intoxication 
 
PEKSISTENT GLYCOSURIA 345 
 
 of intestinal origin, Guelpa advises free purgation, combined with, 
 complete abstinence from food for two, three, or even more days. 
 The one without the other, he maintains, is harmful, purgation 
 without complete rest to the digestive tract causing as much con- 
 stitutional disturbance as abstention from food without daily and 
 abundant evacuation of the intestinal contents. The somewhat 
 paradoxical statement is made that the longer starvation is con- 
 tinued by this method, the less is the feeling of hunger ; for the first 
 twenty-four hours there is some discomfort from the interruption 
 of the usual habits, but by the second or third days this is mucli 
 less marked. It is also said that the more abundant the purge, 
 within Umits, the less is the colic produced. If, for example, a single 
 glass of Hunyadi Janos water is drunk several watery stools result, 
 and there is discomfort for some hours, but if a whole bottle is 
 rapidly taken, say, within fifteen minutes, no colic whatever occurs, 
 and a satisfactory and complete action follows within a short time, 
 sometimes in a few minutes. The purge must be dilute, and be 
 taken hot. To apply these principles to the treatment of diabetes 
 the patient is given a whole bottleful of warm Hunyadi Janos, or 
 other dihite sahne solution, each evening for three days, with com- 
 plete abstinence from all nourishment except water, weak tea with. 
 milk, clear strained vegetable soup, or anj^ hot infusion, according 
 to the degree of thirst complained of. He is then put ujjon a diet 
 of green vegetables, beef -tea, or meat extracts (20 oz.), milk (20 oz.), 
 well diluted with water or tea, for another three days. It is said 
 that in many cases the sugar disappears entirely from the urine, 
 acetonsemia is controlled, and a marked improvement in the general 
 symptoms occurs. If the resumption of an ordinary diet causes a 
 return of the sugar, the amount excreted is always much less, and 
 can be again reduced by a further course of treatment — in fact, it is 
 often advisable that three or four periods of fasting and purgation, 
 each not exceeding three days, and extending over six weeks, should 
 be arranged for at the outset. Satisfactory results have been re- 
 ported with the Guelpa treatment by R. W. Phihp, A. A. Warden, 
 Oscar Jennings, and a few others, but I have never been able ta 
 persuade any of my patients to attempt it. 
 
 Vaccines. — If the view that some cases of diabetes are dependent 
 upon bacterial infection and toxaemia is correct, it might be expected 
 that treatment with vaccines would be useful. As it is impossible 
 to make cultures directly from the pancreatic ducts in cases where 
 the pancreas is probably at fault, the only way is to prepare them 
 from the faeces, testing the j)atient's blood against the cultures so 
 made by the opsonic and agglutination tests. I have carried out 
 this method of treatment in three cases of persistent glycosuria 
 
344 GLYCOSURIA 
 
 where there were probabty floating gall-stone in the common bile 
 duct, and which, for various reasons, had not been submitted to 
 operation. In two the results were for a time satisfactory, the 
 g:lycosuria and other symptoms diminishing ; but in the third no 
 •effect was produced, possibly because the causal organism had not 
 been used in the preparation of the vaccine. 
 
 Santonin has been said to reduce glycosviria, but an exhaustive 
 trial by Walter Lofer proved that it had httle or no effect on sugar 
 excretion in the cases in which he used it. 
 
 Jambul. — The seeds of Eugenia Jambolana {Syzygium Jambol- 
 ■anum) are much used in the East in the treatment of diabetes, and 
 although some observers in Europe have reported satisfactory 
 results, they have proved quite useless in the experience of many. 
 The seeds are given in doses of 5 to 30 grains in the form of 
 powders, cachet, or pills, or as the liquid extract, in doses of | to 2 
 drams. According to Colostoni and Martz, the fresh seeds contain a 
 substance that hinders the formation of sugar from starch, and it 
 may be that the varying effects observed, especially in Europe, are 
 •due to changes in this substance, or some other constituent of the 
 seeds, on keeping. 
 
 Taka-diastase. — Beardsley has reported great rehef from the 
 symptoms of diabetes in several cases following the administration 
 of taka-diastase, in 5 grain capsules after each meal and at bedtime. 
 In one instance the sugar disappeared completely, while in others 
 it diminished in amount ; the pohauria was reheved, and the patient 
 gained in weight. Improvement was found to cease when the drug 
 was stopped, but was continued when it was resumed. No unto- 
 ward effect was traceable to its prolonged administration. 
 
 Alkalies. — Large doses of alkalies are said to interfere with 
 glycogenesis, as the glycogenic ferments do not act well in an 
 alkaline medium (Lepine). It is not hkely, however, that the good 
 effects that foUow the usual therapeutic doses given in diabetes 
 are due to this. They undoubted!}^ tend to counteract the acidosis 
 from which all diabetics are hable to suffer, but many cases in which 
 there is no evidence in the urine of excessive acid formation are also 
 benefited by the administration of alkahes. The drug most com- 
 monly selected is bicarbonate of sodium, but the carbonate, or 
 ■acetate, of sodium, the bicarbonate, carbonate, tartrate, citrate, 
 or acetate of potassium, and carbonate of ammonium, or lithium, 
 are also used sometimes, the lithium salt being particularly useful 
 in gouty patients. The drug may be given in simple solution, or 
 the patient may be sent to a Spa, such as Carlsbad, where he gets 
 the carbonate of soda along with a sulphate, or to Vichy, where he 
 gets the carbonate without the sulphate. Poques is recommended 
 
PERSISTENT GLYCOSURIA 345 
 
 in asthenic cases, and Vittel for gouty diabetes. Richardiere advises 
 an alkaline course, of 60 to 150 grains of sodium bicarbonate in 
 twenty-four hours, for two or three weeks, every three or four 
 months, and states that it is most useful in mild cases. AlkaUes 
 are especially indicated when an anti-diabetic regimen has not 
 fully succeeded. Alkahes should be avoided as far as possible 
 in cachectic cases, when pulmonary tuberculosis is present, and in 
 advanced stages of the disease where there is wasting, unless marked 
 acidosis and symptoms of threatening coma occur. In cases where 
 there is gastritis and hyperchlorhydria the carbon dioxide set free 
 in the stomach, when sodium bicarbonate is given, may cause serious 
 discomfort, and in such cases it is advisable to substitute magnesium 
 hydroxide or calcium salts. The latter have been found by some 
 observers to influence the sugar excretion to a more marked degree 
 than sodium bicarbonate, and, since calcium salts are known to be 
 of use in many toxic conditions, it is not unhkely that their effect 
 in diabetes may be partly dependent on some such action. WiUis in 
 1679 described a case of diabetes treated \^dth lime-salts that re- 
 covered. In 1895 Griibe revived the treatment with calcium salts, 
 and insisted on its value. Glycero-phosphate of calcium and mag- 
 nesium have also been recommended by Robin, with the object 
 of making good the loss of calcium and magnesium salts in the urine 
 that occurs in diabetes. The beneficial results that follow a course 
 of treatment with Bethesda water, containing chiefly bicarbonate 
 of lime and magnesia, and Contrexeville, sulphate and bicarbonate 
 of Hme with some sodium sulphate, may possibly be dependent 
 upon the presence of calcium salts. 
 
 Lactic Acid. — According to Cantini, pure lactic acid, given in 
 water directly after meals, aids the digestion of nitrogenous food, 
 and allows a strictly nitrogenous diet to be taken for a longer 
 period without the occurrence of gastro-intestinal disturbances 
 than would otherwise be the case. It is given in the form of a 
 lemonade, 5, 15, or 20 grams being mixed with a htre of water, and 
 flavoured with peppermint or anise. Half a wineglassful of this 
 mixture, with half a gram of bicarbonate of soda added, may be 
 taken after food, and every hour or two. 
 
 Other drugs that have from time to time been prescribed and 
 advocated hj various observers are so numerous that a mere Hst 
 of them would occupy a large space. In addition to the above may 
 be mentioned chloral, sulphonal, phosphorus, phosphoric acid, 
 strychnine, cocaine, iodine, picric acid, benzoic acid, hydrogen 
 peroxide, valerian, guiacol, camphor, calcium sulphide, creosote, 
 ergot, benzosol, methylene blue, pepsin, rennet, oxygen inhala- 
 tions, &c. 
 
346 GLYCOSURIA 
 
 Electricity has been frequently tried in the treatment of diabetes, 
 but without any striking benefit. It has been recently stated that 
 high frequency currents cUminish the glycosuria experimentally pro- 
 duced in animals, but I am not aware that they have been used 
 for this purpose in human diabetes. 
 
 The very varying success that results from the employment of 
 such a varietj^ of drugs, and other forms of treatment, in persistent 
 glycosuria renders it quite jolain that no one is specific, and suggests 
 that different lines of treatment are required for different cases. 
 This is what might be expected if we accept the view that diabetes 
 is not a disease, but a series of symptoms that may arise from a 
 variety of causes. Some remedies, such as opium and its alkaloids, 
 alkahes, &c., are undoubtedly more frequently useful than others, 
 but even these cannot be employed in every case with the certainty 
 that a cure, or even an amelioration of the symptoms, will result. 
 
 General Managrement and Hygienic Treatment. — As mental 
 shock, worry, and excitement undoubtedly play a part in the pro- 
 duction of glycosuria in some cases, and tend to intensify it when 
 present in others, it is most important that the patient should be 
 placed in congenial surroundings and under conditions that will 
 soothe his nervous system. A cheerful state of mind, freedom 
 from business and other worries, and a full confidence that the 
 treatment he is undergoing wdll benefit him, are material factors 
 in deahng with every case. A fair amount of exercise, taking care 
 to avoid over-fatigue, is advisable in most cases, but in those where 
 there is marked wasting, muscular exercise is Hkely to increase the 
 glycosuria and acetonsemia, and cause depression. Massage is 
 sometimes a useful adjuvant to other forms of treatment, and more 
 particularly in diabetes associated with arterio-sclerosis, in obese 
 patients, and those who are too weak to take muscular exercise. 
 Regulated Swedish gjrmnastics may also be prescribed in suitable 
 cases. Open-air exercise is, however, to be preferred, if possible, 
 and most cases of chronic glycosuria do best when they are able 
 to obtain an abundance of fresh air and sunshine — in fact, some 
 authors contend that oj)en-air treatment is as beneficial in diabetes 
 as it is in tuberculosis. The clothing should be warm, and in 
 winter wooUen, or silk, garments should be worn next the skin. 
 The action of the skin should be promoted by warm haths fol- 
 lowed by friction, especially if the surface of the body is dry, but 
 sea-bathing and cold plunges should not be indulged in. 
 
 A warm climate is often recommended, and since a low external 
 temperature means that more energy must be produced within 
 the organism to keep up its internal temperature, a warm, sunny 
 
PERSISTENT GLYCOSURIA 347 
 
 place would seem advisable. But as the bodily and mental fatigue 
 incident on a long journey are often harmful, and in serious cases 
 may precipitate diabetic coma, the sHght advantage that would re- 
 sult from a change of environment is frequently more than counter- 
 balanced by the risks involved. The same objection also apphes to 
 the S'pa treatment of diabetes. A visit to Carlsbad, Marienbad, 
 Vichy, or Neuenahr is often very useful in mild cases, especially 
 those who are obese or gouty, for the patient is removed from his 
 usual routine, and lives a quiet, regular, and peaceful life ; he is in 
 the open air a great part of the day, he takes suitable bodily exercise, 
 and his diet is carefully regulated ; but the long journey involved, 
 and possibly the worry as to ways and means, may sometimes 
 counteract any good that would otherwise result, particularly in 
 the severer type of case. One is often asked whether the use of 
 tobacco should be given up ? Although acute nicotine poisoning 
 is known to give rise to glycosuria, owing apparently to its action 
 on the supra-renals, it is not probable that smoking in moderation 
 aggravates the condition to any appreciable extent. The abuse of 
 tobacco is, however, undoubtedly dangerous, especially when the 
 heart is not quite sound. Women who suffer from diabetes are not 
 uncommonly sterile, but it is not safe on this account to allow them 
 to marry, for should pregnancy occur it is hkely to seriously aggra- 
 vate the condition, and may rapidly bring about a fatal termina- 
 tion. Abortion frequently occurs, and pregnant diabetic women 
 are very likely to develop tuberculosis, 20 to 25 per cent, of cases, 
 according to Neumann, suffering in this way. When a married 
 woman is found to have diabetes it is necessary to instruct her 
 to avoid conception. Should she become pregnant anti-diabetic 
 treatment should be instituted with great care, all drugs being 
 avoided as far as possible in the interests of the foetus. If dietetic 
 and hygienic measures do not suffice to control the condition, it will 
 probably be necessary to induce abortion. Should this be avoided 
 in the early stages it may become necessary in the last three 
 months in the interests of both mother and child, for a strict anti- 
 diabetic diet is not favourable to healthy development, and the 
 foetus is liable to grow abnormally large, causing difficulty in de- 
 livery at term. When labour has commenced it should be ter- 
 minated as speedily as possible, whether the child is hving or not, 
 but no anaesthetic should be given. After delivery the most careful 
 antiseptic precautions should be observed, and even the smallest 
 tear of the perineum, or soft parts, be carefully attended to. The 
 condition of the urine, and the general state of the patient, must 
 be assiduously watched for evidences of severe acidosis, &c., as the 
 disease often makes very rapid progress during the l}dng-in period. 
 
348 GLYCOSURIA 
 
 Experience has shown that even in healthy women sugar tolerance is 
 lowered at this time. Hirschfeld found that glycosuria occurred in 
 10 per cent, of healthy pregnant women after 100 grams of dextrose. 
 If glycosuria only appears late in pregnancy, the sugar excretion 
 does not exceed 20 grams or so a day, and is not accompanied bj'- 
 other symptoms ; it is probable that it is transitory and will spon- 
 taneously disappear after deUvery. The patient should, however, 
 be very carefully watched. 
 
 SjTiiptoms of diabetes sometimes develop in men shortly after 
 marriage, and it is possible that sexual excitement may occasionally 
 be the determining factor. I have had one case under my care in 
 which sugar appeared in the urine six months after marriage in a 
 hsemophihac patient, but was quickly controlled by diet and rest 
 in a nursing-home. It is probably advisable that males suffering 
 from diabetes, especially the more severe forms, should not marry, 
 but each case must be judged on its merits, and, in any event, 
 excessive excitement should be studiously avoided. 
 
 Treatment of Complications. — Constipation is the compHca- 
 tion which is most commonly present, and has most frequently to 
 be guarded against in diabetes. It is best treated with saline pur- 
 gatives, mineral waters, senna, or castor-oil. Drastic purgatives 
 should be avoided. 
 
 Inflammation of the Gums. — The teeth should be regularly cleaned, 
 and a mouth-wash consisting of Hsterine or a solution of boracic 
 acid (1 dram), borax (2 drams), and pot. chlor. (I dram) in camphor 
 water (1 pint), or a 3 per cent, solution of sodium bicarbonate should 
 be used. A mixture of borax (2 drams), boracic acid (1 dram), tinct. 
 nijrrrhse (^ oz.), and water (to 6 oz.) may be prescribed. 
 
 Dyspepsia should be treated on general principles. A mixture 
 of alkaUes and hydrocyanic acid may be useful. WiUiamson 
 recommends frequent 10 grain doses of bicarbonate of soda in a 
 teaspoonful of milk. Sir W. Roberts gives a pill containing 2 or 3 
 grains of asofoetida to alleviate the craving for food and sense of 
 sinking in the epigastrium. 
 
 For flatulence and intestinal catarrh an initial j)urge followed by 
 a pill containing creosote, or thymol, and extract of belladonna may 
 be prescribed. Sahcylate of bismuth in 8 grain doses, with or 
 without opium, twice a daj^ is often useful in checking diarrhoea. 
 
 General itching of the shin may be reheved by sponging with 
 tepid water, carboUc lotion (1 : 40), hq. carb. detergens (| oz. to the 
 pint), or a mixture containing acid hydrocyan. dil. (1 dram), 
 glycerine (1 oz.), water (to 6 oz.). The best internal remedy is 
 sodium sahcylate in doses of 20 to 30 grains two or three times a 
 
PERSISTENT GLYCOSURIA 349 
 
 day. The condition is, however, most satisfactorily reHeved by 
 controUing the glycosuria. 
 
 Pruritis and eczema of the vulva, or 'prepuce, should be prevented 
 by thoroughly drying the parts after each act of micturition, and 
 reducing the sugar. Sodium sahcylate internally will often relieve 
 the pruritis. If eczema has actually developed it will be found that 
 remedies which prevent the fermentation of the sugar and urine 
 are the most serviceable. In my experience washing the parts 
 with a mixture of yeast and water, a teaspoonful of fresh yeast in 
 a pint of water, as recommended by Carnot, gives the best results. 
 Washing with boracic lotion, or van Swieten's solution (1 part 
 perchloride of mercury, 100 parts of alcohol, and 900 parts of 
 water), and the subsequent application of boracic ointment, zinc 
 ointment, or ung. conii, is often serviceable. A lotion containing 
 sodium hyposulphite (1 in 40) is a favourite remedy with some, 
 others recommend a 3 to 5 per cent, cocaine ointment, or a dusting 
 powder containing 10 per cent, of orthoform. In obstinate cases 
 the local analgesic action of X-rays has given relief. An ointment 
 containing scarlet " R " has recently been recommended for 
 diabetic ulcers and eczema. 
 
 Cystitis is best treated by washing out the bladder with a weak 
 solution of sodium sahcylate, and giving this drug, salol, urotropine, 
 or helmitol internally. 
 
 Boils and carbuncles can only be satisfactorily treated when the 
 sugar in the urine has been reduced, or caused to disappear, by the 
 ordinary dietetic and medicinal measures. As a general tonic 
 quinine, in 3 grain doses internally four times a day, may be helpful. 
 Locally the ordinary antiseptic measures should be taken, but 
 operation should be avoided if possible, especially if it involves 
 the use of a general ansesthetic. Bier's hyperamic treatment often 
 gives very satisfactory results. 
 
 Gangrene. — The treatment of dry gangrene in diabetics does not 
 differ from that usually adopted, except that the glycosuria must 
 be controlled by diet, &c., and surgical interference be undertaken 
 with care, especially if there is acetonaemia. Gangrene due to 
 arterio-sclerosis most often occurs in the lower extremities, and the 
 disease may affect but one or more toes. In these cases, if the hne 
 of demarcation forms early, if the adjacent parts are not inflamed, 
 and if there is a good pulse in the posterior tibial artery behind 
 the internal malleolus and in the dorsalis pedis, the removal of the 
 dead portion alone may be sufficient. If, however, the line of 
 demarcation forms slowly, if the foot is inflamed, and the pulse in 
 the arteries mentioned is feebly felt, if at all, operations on the foot 
 are harmful. In such cases the amputation should be above the 
 
350 GLYCOSURIA 
 
 knee. In the milder cases where amputation has been successfuUj^ 
 carried out the condition of the patient often improves to a sur- 
 prising degree, and cases have been reported by Koenig and others 
 in which the urine has become sugar-free. 
 
 As a rule, patients with moist gangrene are in an advanced stage, 
 and show weU-marked evidence of acidosis, so that the risk of opera- 
 tion is great. It frequently offers the only chance of prolonging Hf e, 
 however, and should be undertaken if there is great pain in the 
 living margin of the tissue, or the glycosuria is increasing and coma 
 threatens, or there is marked fever and rapidly spreading celluhtis. 
 In the rarer tj^e of case where acetonsemia is absent amputation 
 should be carried out as quickly as possible, and as wide of the 
 gangrenous area as can be managed. Most scrupulous care must 
 be taken with the antisej)tic details of the operation, and spinal is 
 preferable to general ansesthesia. In either case the affected part 
 must be kept dry and aseptic until surgical interference becomes 
 necessary. Herzfeld speaks highlj^ of a dressing of dry sodium 
 perborate powder for diabetic gangrene and Dieulaf oy has used hot- 
 air douches, at a temjDerature of 100° to 300° C, to prevent infection 
 and the subsequent dangers of sej)tic intoxication and septicaemia, 
 with satisfactory results. 
 
 Nephritis. — When nephritis occurs as a comphcation of diabetes, 
 and is not due to the excretion of acetone bodies in the urine, the 
 nitrogenous food in the diet should be reduced, and replaced as 
 much as possible by milk and fats. Although milk contains lactose, 
 many diabetics can take two or three pints a day without seriously 
 increasing the glycosuria. Sugar-free milk may also be used, but 
 even this, in some instances, causes the excretion of sugar to rise. 
 It is consequently often a difficult matter to arrange the diet so 
 that the albuminuria will be decreased, without at the same time 
 augmenting the glycosuria. The bowels should be kept well open 
 with saUne purges, or senna, and a bottle or so of Vichy water should 
 be taken each day. The use of opium and its derivatives is to be 
 avoided, but alkahes, and particularly sodium citrate, are often useful. 
 
 (Edema in the legs and other situations, not associated with 
 kidney lesions, is best treated bj?- rest in bed. Perchloride of iron 
 may also be given. The oedema that sometimes results from the 
 oatmeal treatment usually diminishes rapidly when the nature of 
 the diet is altered. 
 
 Nocturnal cramp is most satisfactorily treated by controlling 
 the glycosuria, but reUef may meanwhile be often obtained by the 
 administration of sodium bicarbonate in repeated doses of 10 or 
 15 grains, the last being taken at bedtime. 
 
 Neuritis and sciatica are treated on ordinary Hues. For the 
 
PERSISTENT GLYCOSURIA 351 
 
 gnawing pains in the legs of which some patients complain, 
 Wilhamson advises antipyrin in 10 grain doses three times a day. 
 
 Sleeplessness may be combated by the exhibition of opium, 
 sulphonal, and similar drugs, nervous excitement with potassium 
 bromide. 
 
 Arteriosclerosis and cardiac affections are treated by the usual 
 remedies. For the former iodides are probably the most useful 
 drugs, but attention must also be paid to the condition of the 
 bowels. Digitahs should be used with great care in gouty glycos- 
 uria, in which the heart is hypertrophied and the arterial tension 
 already too high. Its tendency to increase the tension may be 
 counteracted by the simultaneous administration of nitro-erythrite 
 or some similar drug. 
 
 Acidosis and Diabetic Coma. — Of all the complications of diabetes 
 these are the commonest and most dangerous. The latter is, as 
 we have seen, probably a result of the former, at any rate in most 
 instances, so that the treatment of the one is involved in the treat- 
 ment of the other. The use of carbohydrates has long been known 
 to restrain acidosis, and in mild cases to cause its disappearance ; 
 but such a hne of treatment has the disadvantage that it is likely 
 to ultimately do harm by making the glycosuria worse, unless it 
 is carried out with great care. The amount of carbohydrate that 
 can be utiHsed must be determined experimentally for each case in 
 the manner already suggested, and the patient must be allowed 
 this amount and no more. The assimilative powers can often be 
 thus maintained for long periods, and even be improved, especially 
 if regulated intervals of restricted diet are interposed. It is 
 evident that sugar itself is not the body that controls acidosis, 
 but that its action depends upon the formation of some derivative. 
 Many observers have therefore devoted their attention to the 
 search for this sugar product, or a substitute. A variety of sub- 
 stances, including gluconic acid, glutaric acid, alcohol, glycerine, 
 and glycerine aldehyde, &c., have been found experimentally to 
 reduce acidosis, but none have proved to be of practical cHnical 
 value, excepting perhaps alcohol. The toxic effect of alcohol has, 
 however, to be reckoned with, and its use is therefore Hmited. As 
 we have seen, the acids giving rise to acidosis originate from fats, 
 and to a less extent from the fatty acid moiety of proteins. Since 
 different fatty foods var}^ in the amount of acid thej^ produce, 
 those containing much of the lower fatty acids being especially 
 harmful, the diet should be arranged to eUminate these as much 
 as possible. With the exception of using fresh washed butter, or 
 oleo-margarine, in the place of old butter, very Httle can be accom- 
 j)lished in this direction however. With regard to the nitrogenous 
 
352 GLYCOSURIA 
 
 foods, it has been shown that they vary greatly in their influence on 
 acidosis, vegetable and egg proteins being less harmful than x^roteins 
 of animal origin. Beside increasing the acids in the circulation, 
 proteins also give rise to substances that neutrahse them — that is to 
 say, they are both ketogenic and anti-ketogenic. The latter effect 
 depends partly upon the carbohydrate group that is spht off and 
 is more easily utihsed than ingested carbohydrate by all but the 
 most serious cases, and partly upon the amino-acid fractions which 
 exert a twofold action, some increasing the acidosis by yielding the 
 fatty acid of their molecule, while others have a restraining influence 
 from the ammonia that is derived from them. Choice in the 
 selection of the nitrogenous part of the diet may therefore be used 
 to a certain extent to control acidosis, but the hmits of selection 
 are too restricted to be more than occasionally of service in practice. 
 A high intake of protein may, according to Magnus-Levy, at times 
 increase acidosis, not from the formation of oxybutyric acid from 
 the protein, but because the high protein content of the diet makes 
 an extra tax on the oxidising powers of the organism and diverts 
 these from the combustion of acetone bodies. 
 
 The first indication, therefore, in the treatment of acidosis, and 
 its sequel diabetic coma, is to prevent or postpone their onset by 
 suitable quahtative and quantitative regulation of the diet, the 
 second is to neutrahse the acids that form by the administration of 
 alkahes. The latter, although a valuable method, is, after all, only 
 palhative, for alkahes do not hmit the formation or favour the 
 combustion of acetone bodies. In many cases, on the contrary, the 
 excretion of acetone boches in the urine will increase if large qviantities 
 of alkali are given. This is not an unfavourable sign, but merely 
 indicates increased excretion. I cannot too strongly insist on the 
 alsolute necessity for adapting the diet to the individual require- 
 ments of the case at the earhest possible stage if serious disturbances 
 of metaboUsm are to be avoided or controlled. 
 
 It must be remembered that acetonuria has not always the same 
 significance, and that cases presenting this sign may be con- 
 veniently divided into three classes, as v. Noorden suggests : — 
 
 (1) Cases of shght glycosuria, readily cured by the withdrawal 
 
 of carbohydrates from the diet, but presenting shortly 
 after this diet has been instituted signs of acetone in the 
 urine. In such cases the acetonuria is physiological and 
 need cause no alarm, nor does it necessitate a withdrawal 
 of the severe diet, as it will disappear in a few weeks. 
 
 (2) Diabetes with shght glycosuria that, in spite of a partially 
 
 restricted diet, show signs of acetone, which, however, 
 disappear on a more rigid diet. The extra rigidity in 
 
PERSISTENT GLYCOSURIA 353 
 
 diet does not excite the same metabolic disturbance as 
 
 in the first class of case, where hardly any restriction of 
 
 diet had been observed. 
 
 (3) Cases where the glycosuria is marked, and where cutting 
 
 off the carbohydrates does not suffice, and reduction of 
 
 proteids is also necessary. In all such cases prolonged 
 
 administration of alkalies is strongly urged as they help 
 
 the eHmination of the acetonuric acid bodies and diminish 
 
 the acetonsemia. It is necessary to avoid too sudden 
 
 restriction of diet, and it may often be advisable to allow 
 
 a few days' liberty, and so gradually work down to a 
 
 hydrocarbon-free diet. In the worst forms even the small 
 
 amount of carbohydrate coming to the Uver via the portal 
 
 vein does not suffice to restrain the production of acetone, 
 
 and in these cases — with a persistent acetonuria^it is 
 
 quite indifferent whether we withhold the carbohydrates 
 
 from the food or not. 
 
 The alkali most commonly used in the treatment of acidosis is 
 
 bicarbonate of soda. The amount of this in excess of two drams 
 
 required to make the urine alkaline may, as we have seen, be taken 
 
 as a rough index of the intensity of the condition. As a rule it is 
 
 not advisable to render the urine distinctly alkaline, but to keep it 
 
 just short of that point, as otherwise symptoms of depression are 
 
 likely to supervene. When, however, the premonitory symptoms 
 
 of coma appear the administration of the drug should be pushed. 
 
 Hucard gives 2 to 10 drams daily, and v. Noorden half an ounce of 
 
 bicarbonate of soda each day, with 45 grams of calcium carbonate 
 
 added to make up for the loss of calcium salts in the urine. The 
 
 administration of calcium salts is probably also useful for other 
 
 reasons. Large doses of bicarbonate of soda are liable to produce 
 
 diarrhoea, and this is counteracted to a certain extent by calcium ; 
 
 moreover, calcium salts are known to combine with toxic substances 
 
 and assist in their elimination from the body. It has been proved 
 
 experimentally by Silvestin that the administration of calcium 
 
 lactate permits otherwise fatal doses of strychnine to be given to 
 
 dogs and cats, also that calcium salts increase the resistance of 
 
 animals to the effects of injections of blood serum from cases of 
 
 uremia and eclampsia, so that calcium salts may be of use in 
 
 diabetes from this point of view also. Some observers have strongly 
 
 advocated sodium citrate in conjunction with, or in place of, sodium 
 
 bicarbonate. It is pointed out that it is not neutralised by the 
 
 gastric juice, and is converted into bicarbonate in the blood where 
 
 the alkah is most needed. Its taste is not obnoxious, it does not 
 
 cause disturbances in the stomach, the appetite does not suffer, 
 
 z 
 
354 GLYCOSURIA 
 
 and even large doses, 750 grains a day, do not cause diarrhosa. 
 It is claimed by Lichtwitz that the elimination of nitrogen in the 
 urine is increased by the citrate treatment. In my own practice 
 I generally prescribe a mixture ^ containing : — ■ 
 
 Sodium bicarbonate 
 Sodiixm citrate 
 Calcium carbonate 
 Magnesium carbonate 
 Water . 
 
 gr. 10-20, up to 225 gr. daily. 
 
 „ 5-10 „ 75 
 
 „ 4-5 „ 45 
 
 „ 4-5 „ 45 
 oz. 1 ,, 10 oz. ,, 
 
 One or two bottles of Vichy, or Neuenhar, water may also be 
 taken each day. 
 
 The total amount of alkali given in the day should not exceed 
 about an ounce, as larger quantities are likel}^ to depress the heart. 
 An increase in the body weight, with oedema and ascites, is often 
 observed when large doses of sodium bicarbonate are being taken 
 by diabetics. On stopping the drug, or reducing the dose, the 
 weight falls, and the oedema disappears. It has been shown by 
 Widal, Lemierre, and Cotoni that these changes are only indirectly 
 dependent on the alkali, and are directly produced by a retention of 
 chlorides, as is the oedema of cardiac and renal disease. 
 
 Beside prescribing large doses of alkalies, the premonitory 
 symptoms of coma must be met by a thorough clearance of the 
 bowels by means of a mild purge. Castor oil, comjjound jalep 
 powder, or calomel, followed by a sahne, are probably the most 
 useful. Lavage of the stomach may be carried out, if practicable, 
 to assist in the removal of toxines. The patient should be kept in 
 bed, and all sources of bodily or mental fatigue be carefully avoided. 
 If these measures do not suffice to control the symptoms, intra- 
 venous injection of an alkaline solution must be resorted to, but 
 it is important that this should not be delayed too long, as the 
 best effects undoubtedly result before the patient becomes actually 
 comatose. Until recently a weak solution of bicarbonate of soda 
 has been almost universally used, one to three pints of a 2*5 to 5 per 
 cent, solution being injected into a vein. Blum has, however, pointed 
 out that there is a limit of safety to this method, and that alkahne 
 solutions of too great a strength may produce venous spasm and 
 various evidences of central nervous poisoning, such as con- 
 vulsions, &c. Blum believes that it is the lack of balance in the 
 solution that is responsible for its toxurity, for various ill effects 
 have been noticed after the injection of so innocuous a compound 
 as sodium chloride. To injections designed simply to introduce 
 the neutral chloride into the circulation, potassium and calcium 
 
 ^ This has been put up in compressed tablet form for me by Messrs. Allen 
 and Hanbury under the name of " Compound Neutralising Tablets." 
 
PERSISTENT GLYCOSURIA 355 
 
 chlorides can be added to balance its effect, but with alkaline 
 injections this is not possible, as insoluble calcium carbonate is 
 formed. Blum thinks that a 6 per cent, solution of bicarbonate 
 of soda is the least harmful to use, provided that this has been 
 l^reviousl}^ boiled for a quarter of an hour to sterilise the solution, 
 and to render it less toxic by converting the bicarbonate into 
 sesquicarbonate, which, according to Stadelmann, is better tolerated 
 than the carbonate, and contains more sodium than the bicarbonate. 
 Lichtwitz claims that sodium citrate is better adapted for intra- 
 venous injection than the bicarbonate, as it has only a weakly 
 alkaline reaction, and this can be neutraHsed by adding a little 
 citric acid. Some observers have been content to use only a sterile 
 normal saline solution (0*6 per cent.), but even this, unless prepared 
 with absolutely freshly made distilled water, is, according to Hort 
 and Penfold, liable to cause fever and other toxic symptoms. In 
 such a serious condition as advanced acidosis and threatening coma, 
 some risks must be taken, and, on the whole, Blum's method of 
 injecting two or three pints of a boiled 6 per cent, solution of bicar- 
 bonate of soda appears to be the least objectionable. The solution 
 should be at body temperature, and be given slowly and steadily, 
 one-half to three-quarters of an hour being taken to administer two or 
 three pints, so that a probably feeble heart shall not be overcharged. 
 As a rule, copious diuresis follows, and the pulse becomes stronger 
 .and less rapid. Sometimes a certain amount of oedema, attributed 
 by Widal to retention of chlorides, follows the injection, but it is of 
 no special import. If the patient is comatose he may return to 
 consciousness, and his mind become surprisingly clear, but usually 
 it is only a temporary rally, and after a few hours a relapse takes 
 place and death speedily follows. Very rarely the recovery is 
 permanent. Oliver, for example, has recorded a case in which the 
 patient left the hospital four weeks after an attack of diabetic coma 
 treated with intravenous injections of sodium bicarbonate, and 
 Labbe has reported one in which combined intravenous and oral 
 administration brought about a cure. The patient, who had sunk 
 into unconsciousness, received an injection of 15 grams of bi- 
 carbonate of soda, and was sufficiently restored to take 60 grams by 
 the mouth. For five consecutive days she was given injections, and 
 she was then so much better that the injections were stopped, and 
 the treatment was continued only by the mouth. Two days later 
 the coma returned ; she was then given an intravenous injection 
 of 30 grams, repeated the next da3^ and after a month's alkaline 
 treatment by the mouth she was completel}^ cured. Bleechng from 
 the opened vein to the extent of about a pint, before the injection 
 is made, may assist in overcoming the toxaemia. 
 
356 GLYCOSURIA 
 
 Sicard has recently advocated the use of intravenous injections 
 of sterihsed, concentrated (8 to 9 per cent.) solutions of sodium 
 bicarbonate in the treatment of other symptoms than acidosis, 
 when these resist the ordinary dietetic and medicinal measures. 
 In three cases, one of obstinate pruritis, another of sciatic pains, 
 and a third of optic neuritis, he obtained marked remission. He 
 injects from 100 to 250 c.c, representing about 20 grams of sodium 
 bicarbonate, into a vein in the arm, but points out that it is ab- 
 solutely necessary that the injection should be made directlj^ into 
 the interior of the vessel as the solution is caustic, and if introduced 
 into the cellular tissue would produce an inflammatory reaction. 
 The injection is extended over about five or ten minutes. He states 
 it may be repeated several times, if necessary, at intervals of a few 
 days without the least danger. 
 
 To avoid the discomfort and disturbance consequent on the 
 ingestion of large quantities of sodium bicarbonate by the mouth 
 some physicians employ rectal injections (| oz. in ^ a pint) with, 
 or without, the addition of dextrose (1 dram) in cases of impending 
 coma every four hours, but as diarrhoea and tenesmus are very 
 Hkely to result it is not a method to be recommended. 
 
 When coma is imminent, or the patient is actually unconscious, 
 cardiac stimulants such as alcohol, ether, strychnine, caffeine, 
 ammonia, and digitalis should be given by the mouth, or be ad- 
 ministered subcutaneously. Some authors have advised oxygen 
 inhalations with a view to combating the air-hunger, but, according 
 to Pembrey, this procedure is unnecessary and useless, as the 
 alveolar air contains plenty of oxygen (17 to 18 per cent.), and the 
 deep breathing is merely a response to the stimulating action of the 
 acid substances present in the blood. 
 
 Treatment of Infantile Diabetes. — Glycosuria is a rare 
 condition in young children, but it is probable that it would be 
 found more commonly than is generall}^ thought if it were sought 
 for systematically. It is particularly important that it should be 
 recognised and come under treatment as early as possible, since in 
 infants it is an acute disease that may terminate fatally in a few 
 days, and even in older children it may run its course in a few 
 months. As a rule, the onset and development of diabetes are not 
 quite the same in children as in adults. In addition to polyuria 
 there is generally incontinence ; thirst is a very prominent symptom 
 in nearly all cases ; the urine is often alkaline, and contains a high 
 proportion of sugar (30 to 80 grams per litre), and a large quantity 
 of urea (20 to 25 grams in the twenty-four hours). The onset of the 
 
PERSISTENT GLYCOSURIA 357 
 
 condition is usually insidious, and is often confused with digestive 
 or dentition disturbances, attention being first directed to the true 
 state of affairs by the thirst, polyuria, and loss of weight. 
 
 The treatment consists in reducing the ingestion of foods con- 
 taining dextrose, and improving metabolism generally as much as 
 possible. With breast-fed children nursing is continued, but a 
 small teaspoonful of Vichy water is given after each feed. With 
 artificially fed children the milk is sweetened with mannite, glycerine, 
 or saccharine, and is diluted with Vichy water. The patient may 
 also be fed with a milk specially prepared as follows : The casein 
 and fat of one and a half litres of ordinary milk are jDrecipitated by 
 adding rennet, the sugar is removed by washing the coagulated 
 mass, and the latter is then pressed through a fine sieve into a 
 mixture of 200 grams of whey and 1300 grams of waiter. The 
 mixture obtained in this way contains only 8 grams of sugar, but 
 much fat and albumen. It may be made more palatable by adding 
 saccharine. When a breast-fed child is weaned it is nourished 
 on milk, or cream, diluted with water, or Vichy water, and 
 sweetened with glycerine or saccharine. Eggs, minced meat, 
 and green vegetables should be gradually added to the diet. 
 Starchy foods must be avoided as much as possible, and as soon 
 as it is old enough the child should be given prepared vegetable 
 proteins {e.g. Roborat) and eggs. As a precaution against acidosis 
 the nitrogenous diet should not be too rigid, or be too quickly 
 introduced. Fatigue of all kinds should be guarded against. An 
 older child should have the first meal in bed at a fixed time. He 
 should remain in bed till mid-morning, and then be allowed to 
 be up and out until the hour for the noon meal, after which he 
 should be well wrapped and made to rest quietly in the open air 
 for at least two hours. He may then be given mild exercise, 
 walking or driving, if in good condition, or massage if passive 
 exercise is better borne. Supper should be not later than 6 p.m. 
 and bed at 7. The sleeping room should be ventilated as for a 
 tuberculosis case. As much care must be taken to guard against 
 waste of nervous energy, and to provide fresh air at all times, as 
 is used in regulating the food. The patient should be put on a 
 standard diet, suited to his age and condition, and as free as 
 possible from carbohydrate. A supplementarj^ diet consisting 
 principally of the carbohydrates found to be best borne should 
 be held in reserve, from which is added more or less according to 
 the varying carbohydrate tolerance shown. 
 
 The urine must be examined as often, and as thoroughly, as 
 possible, and the patient be weighed every three or four days. 
 
358 GLYCOSURIA 
 
 Drugs are to be used with care, quinine, iron, cod-liver oil, alkalies, 
 arsenic, and bromides being the most useful. Where there is 
 evidence of specific disease mercury should be prescribed. For 
 iniants, waters containing alkalies, chlorides, and arsenic are the 
 most valuable. Alkahne waters also benefit older children, but 
 when a stimulating effect is desired waters containing iron and 
 chlorides should be emploj^ed. Constipation must be carefully 
 guarded against, and diarrhrea must be promptly checked. When 
 coma is threatened the symptoms are met in the usual way with 
 purgatives, stimulants, and alkalies. Very satisfactory results 
 have been obtained in diabetic chikben with v. Noorclen's oatmeal 
 cure, especially when it has been instituted at an early stage. Abt 
 and Strouse have reported that in two cases of traumatic diabetes 
 in children that they treated by this method, carbohydrate tolerance 
 was strikingly raised and the acidosis was considerably reduced. 
 
 Prophylactic Treatment.— As our knowledge of the pathology 
 and etiology of persistent glycosui'ia grows, it becomes increasingly 
 clear that, not only can no hard and fast line be drawn between 
 simple glycosuria and glycosuria with secondary disturbances of 
 metabolism, but that in many instances sugar is intermittently 
 present in the urine before permanent glycosuria is established, and 
 also that there is probably what may be called a " pre-glycosuric 
 stage " in which carbohydrate metabolism, although disturbed, is 
 not so far interfered with that sugar appears in the urine when 
 an average amount only of carbohydrate is consumed. It is the 
 experience of all who have had to deal with many cases of diabetes, 
 that the earUer treatment is commenced the more satisfactory and 
 permanent are the results, and we may, therefore, conclude that 
 if patients could be treated in the pre-glycosuric stage the onset 
 of glycosuria might be much delayed, or even be prevented. Our 
 knowledge of the etiology of many of the pathological changes 
 that give rise to glj'-cosuria is so meagre that, at present, we can 
 accomplish little in this direction, except in a general way. It is 
 known, for instance, that sugar may appear in the urine of obese 
 and gouty individuals, so that if patients suffering in these ways were 
 watched and treated wdth the possibility of glycosuria occurring as 
 a complication in mind, it might often be prevented. Again, the 
 tendency to defective carbohydrate metabolism appears to be in- 
 herited in some families, and it is likely that much might be done to 
 hinder or prevent its development if the metabolism of such in- 
 dividuals were periodicallj- investigated, especially in early life, 
 and they were warned against conditions that are known to play a 
 
PERSISTENT GLYCOSURIA 359 
 
 part in the production of persistent glycosuria. The absorption of 
 indican, and other products of protein putrefaction, has been stated 
 to affect the supra-renals, and so may bring about a disturbance of 
 the balance between the ductless glands that normally controls 
 carbohydrate metabolism. It would, therefore, seem advisable 
 that intestinal catarrhs, and similar conditions that are associated 
 with abnormal putrefactive changes in the intestinal contents, 
 should be recognised and controlled as quickly as possible. 
 
 As a result of the work of numerous observers we are now on 
 surer ground with regard to diseases of the pancreas than was the 
 case some years ago. In a paper that I published in 1908 I pointed 
 out that : — " So long as the etiology of many lesions of the pancreas 
 was obscure, and their diagnosis a matter of extreme difficulty, 
 little could be done to deal with them before they had advanced 
 to such a stage that the functions of the gland were hopelessly 
 impaired, but now that we can readily detect the probable presence 
 of degenerative changes by means of the " pancreatic " reaction 
 in the urine, and a consideration of the clinical history and 
 symptoms, together with the results of a careful examination of 
 the urine and fseces, will usualty serve to throw light on the cause 
 of the condition, there can be no reasonable excuse for allowing 
 the disease to progress undiagnosed and untreated until diabetes 
 supervenes." My experience since these words were written has 
 not given me any reason to alter the opinion I then expressed ; in 
 fact, I feel even more confidence in stating that if pancreatic diseases, 
 and particularly those of an inflammatory type, were more generally 
 recognised in the early stages, and were thoroughly treated, the 
 number of cases of diabetes with which we have to deal would 
 be considerably reduced. As a result of the opportunities that I 
 have had during the last ten or twelve years of observing a large 
 number of individuals suffering from pancreatic disease and diabetes, 
 I have been led to divide cases of pancreatic glycosuria into three 
 classes according to the probable source of the morbid influence 
 affecting the pancreas : — 
 
 (1) Those in which the pancreatic mischief is probablj^ secondary 
 to some morbid influence reaching the gland by way of the ducts. 
 (2) Those probably secondary to blood diseases or circulatory dis- 
 turbances, including arterio-sclerosis, interacinar pancreatitis, 
 syphilis, &c. (3) Those in which the diabetes was induced by 
 destruction of the pancreas by malignant cHsease, either primary, 
 or a secondary invasion from neighbouring organs. 
 
 My object in doing so has been to make clear their etiology, 
 and so open the way for earlier and more radical treatment. In 
 
360 
 
 GLYCOSURIA 
 
 the first class there is usually a long antecedent history of 
 " dyspepsia," or gastro-intestinal trouble, probably indicating a 
 duodenal catarrh, or of attacks of jaundice of the " catarrhal 
 type," or of gall-stones. The progress of the disease is slow, and 
 it is not until the pancreatic lesion has reached an advanced stage 
 that glycosiu-ia occurs. I have met with cases in which this was 
 shown by several years intervening between the discovery of the 
 pancreatitis and the appearance of sugar in the urine. One patient 
 who had been operated on for gall-stones did not develop glycosuria 
 until eight and a half years after the operation. In another the 
 symptoms of diabetes came on four j^ears after an exploratory 
 operation for what was believed to be cancer of the pancreas. At 
 the time these operations were performed chronic pancreatitis, and 
 its j)Ossible consequences, were not fully recognised, but had means 
 been taken then to stay the progress of the disease it is hkely that 
 a cure might have been effected. The striking increase in the 
 death-rate from diabetes shown in recent years by the Registrar- 
 General's returns, and also in the mortality tables from Paris and 
 New York, is probably not unconnected with the greater prevalence 
 of digestive disturbances : — 
 
 Deafh-rate from Diabetes per 100,000 Population 
 
 
 London 
 
 Paris 
 
 New York 
 
 1880 .... 
 1890 .... 
 1900 .... 
 
 4-3 
 6-(i 
 
 7-7 
 
 5-0 
 13-0 
 15-8 
 
 5-71 
 
 8-06 
 
 11-34 
 
 A detailed study of these returns shows that the increase is 
 chiefly due to the larger number of cases occurring in older people 
 at ages from fifty-five to seventy-five, or at least not causing death 
 before the later periods of life ; that is to say, the glycosuria makes 
 its appearance at a time when we might expect the secondary effects 
 of intestinal disturbances and intestinal toxaemias on metabohsm 
 to become apparent. I would, therefore, urge that when symptoms 
 of dyspepsia are combined with the presence of the " pancreatic 
 reaction " in the urine, prompt steps should be taken to deal mth 
 the pancreatitis that is probably present, lest, in the course of time, 
 worse should befall. In addition to the routine medical treatment 
 of such cases I make use of hexamethyleneamine (urotropine), 
 sahcjdate of soda, and intestinal antiseptics, such as sulphocarbolate 
 of soda, or izal, with a view to controUing the infection of the 
 
PERSISTENT GLYCOSURIA 361 
 
 pancreatic ducts. Physiological rest is also given to the pancreas 
 by a carefuUy selected diet, and by placing the meals at the longest 
 possible intervals from each other. If, after a fair trial, these 
 methods are found to fail, and the '' pancreatic reaction " j)ersists 
 as markedly as ever, recourse must be had to surgery. The pan- 
 creatic ducts must be drained bj^ cholecyst-enterostomy, and, if 
 necessarj^, a gastro-enterostomy will give rest to the inflamed 
 duodenum. Diabetes is not a common result of obstruction of the 
 common duct by gall-stones, but it does follow in a certain pro- 
 portion of cases, and it is impossible to foretell what will result 
 in any particular instance. I have already mentioned one case in 
 which glycosm-ia developed eight and a half years after an operation 
 in which, although the gall-stones were removed, no provision was 
 made for dealing with the pancreatitis that was present ; and I have 
 met with similar cases in which sugar was found three or four j^ears 
 after an operation for gall-stones in the common duct. Chronic 
 pancreatitis is found to be associated with about 70 per cent, of 
 cases of common duct cholelithiasis, and as medical treatment has 
 little or no effect on the disease, it is important that all gall-stone 
 cases should be operated on at the earhest possible moment, par- 
 ticularly if there is a well-marked "pancreatic reaction" in the 
 urine. A small amount of sugar is no bar to 023erative inter- 
 ference ; in fact, it may result in the disappearance of the glycos- 
 Tiria. Both pancreatic calcuU and cysts are probably the result of 
 catarrhal pancreatitis, due to an infection of the ducts from the 
 duodenum, and although no radical benefit can be expected to 
 follow the treatment of the diabetes occasionally found to com- 
 phcate these conditions, much may be done to prevent their onset 
 and to avoid a consequent glycosuria by timelj^ recognition of the 
 ehronic pancreatitis that precedes and accompanies them. 
 
 The treatment of cases belonging to the second and third classes 
 calls for no special remark. The etiology of arterio-sclerosis and 
 interacinar pancreatitis is alike obscure, and, until we have further 
 information with regard to them, empirical methods of treatment 
 are all that can be adopted. Syphihtic patients with cUabetes are 
 •said to have been cured by anti-sj^Dhihtic treatment, but, as a rule, 
 the result is not satisfactory. The arterio-sclerotic tjipe of diabetes 
 is the most hopeful. It is the form most commonly met with in 
 elderly people, and is probably due to circulatory disturbances 
 in the pancreas. It is not infrequently associated with granular 
 kidney and traces of albumen in the urine. These patients usually 
 respond well to dietetic and general hygienic treatment, carefvilly 
 appHed according to the requirements of the individual case. All 
 
362 GLYCOSURIA 
 
 the cases I have had the opportunity of observing have much 
 improved under treatment. 
 
 Diabetes due to destruction of the pancreas by mahgnant 
 disease, either primary or secondary, is beyond all forms of satis- 
 factory treatment. 
 
 Surg'ieal Treatment. — It was for long one of the accepted 
 axioms of surgery that operative treatment of patients whose urine 
 contained sugar should be avoided unless absolutely necessary, but 
 with the introduction of antiseptics one of the most serious dangers 
 attending surgical interference in such cases was done away with. 
 O^^dng to their disturbed metabolism, imperfect nutrition, reduced 
 resistance, impaired reparative processes, and depressed nervous 
 sj^stem the most careful consideration is still called for before 
 operation is decided on, but if proper precautions are taken, and 
 the patient is carefully prepared beforehand, a large proportion of 
 diabetics can be operated on as safely and as satisfactorily as other 
 individuals. The three chief dangers that have to be contended 
 against are (1) sepsis, (2) coma, and (3) failure of the wound to 
 heal. By a rigid application of the principles of aseptic or, probably 
 better in these cases, of antiseptic surgery the first complication 
 can no doubt be avoided. The second is more difficult to contend 
 against. If operation becomes imperative in a joatient with 
 acetonsemia a very considerable risk is always run, especiaUy if 
 the operation is prolonged and a general anaesthetic employed. 
 It would seem advisable that, whenever possible, spinal anaesthesia- 
 should be used. When time can be given to the preparation of the 
 patient the risk of coma can be considerably reduced by regulating 
 the diet, securing thorough evacuation of the bowels, and ad- 
 ministering calcium salts and alkalies until the reaction of the urine 
 is neutral or nearly so. Should symptoms of acidosis develop 
 subsequent to the operation they should be met by the adminis- 
 tration of alkahes by the mouth, or by rectal or intravenous in- 
 jections. VonNoorden's oatmeal diet has been recommended after 
 operation as a preventive of acidosis. In my experience failure 
 of the wound to heal is often the most serious difficulty. I know of 
 several cases in which the wound showed no sign of union a week or 
 more after operation. 
 
 In adchtion to what may be termed operations of necessity, 
 such as those for diabetic gangrene, &c., operations of choice have 
 to be considered. In these there is either a tumour that needs 
 removing or some comphcation, such as duodenal ulcer, empj^ema, 
 or gaU-stones, &c., the surgical treatment of which will probably 
 
PERSISTENT GLYCOSURIA 363 
 
 give a better chance of recovery. Each case must be considered 
 from every point of view before it is decided that operation is, or is 
 not, advisable, but the dominating factor is the condition of the 
 urine, and particularly the degree of acidosis and the way it responds 
 to treatment. In my opinion the amount of sugar is not of itself 
 a question of supreme importance, for many patients with well- 
 marked glycosuria can be safely operated on, provided that secondary 
 disturbances of metabohsm are absent or are only shght. Moreover, 
 the removal of a tumour which is possibly interfering with the 
 blood supply of the pancreas, or of gall-stones that are keeping up 
 a catarrhal inflammation of the gland, may cause the sugar to dis- 
 appear or be materially reduced in amount. Evelt and Henkel, 
 for example, have reported cases of ovarian tumour in which the 
 removal of the growth was followed by a complete disappearance 
 of the sugar that had been previously present in the urine. Henkel 
 had the same experience with a case of uterine myoma. Carey 
 Evans has described a case in which the gh^cosuria disappeared 
 after gastro-enterostomy for severe indigestion lasting for eigh- 
 teen months. Mayo Robson and Mansell Moullin have published 
 records of cases in which glycosuria has been apparently cured by 
 the removal of gall-stones from the common bile duct. It must 
 not, however, be too readily assumed that because no sugar is 
 fovmd in the urine shortly after operation that a cure has been 
 effected, for it may be that it will return when the patient resumes 
 his ordinary diet and mode of life, or that the improvement is only 
 a temporary one, so that, unless the patient is treated as a potential 
 diabetic, the damage done to the pancreas will progress, and 
 eventually give rise to incurable glycosuria. I have met with two 
 cases reported as cured by operation who subsequently relapsed, 
 and one of them to my knowledge died of diabetic coma. 
 
 The need for a more general early recognition and treatment 
 by surgical means of interstitial and catarrhal pancreatitis was 
 emphasised by Mayo Robson in a paper pubhshed in 1910. He 
 pointed out that although we cannot hope to cure fibrosis of the 
 pancreas by surgical interference, it is possible to remove some of 
 the exciting causes of the antecedent inflammation and so rescue a 
 sufficient amount of gland substance in a functionally active con- 
 dition to cure an existing glycosuria, or prevent the subsequent 
 onset of diabetes. The view is an attractive one, and well worth 
 the serious consideration of both physicians and surgeons. 
 
 Prog'nosis. — Persistent glycosuria undoubtedly shortens life, 
 but different cases vary so much in their gravity that the prognosis 
 
364 GLYCOSURIA 
 
 in any particular instance cannot be based upon a statistical 
 estimate of the average duration of the disease. Each case must 
 be separately considered. The principal points to be taken into 
 account in giving a prognosis are : — 
 
 (1) The age and sex of the patient, (2) the state of nutrition, 
 (3) the cause of the glycosuria, (4) the total amount of sugar ex- 
 creted daily in the urine, (5) the presence and degree of secondary 
 disturbances of metaboHsm, (6) the response to treatment, (7) the 
 social condition of the patient, (8) the nature and extent of any 
 comphcations. 
 
 (1) Age influences the jDrognosis in diabetes to a marked degree, 
 the outlook being generally worse the earher in life the glycosiuria 
 makes its appearance. If it begins in the second half of life it is 
 often a comparatively harmless disorder, but in children it is usually 
 a very grave condition, the younger the child the shorter being the 
 duration of the illness as a rule. It must not be concluded, however, 
 that diabetes in a child is of necessity always rapidly fatal, for 
 Hedon, Abt and Strouse, Crofton, and others have reported cases 
 in which glycosuria in children as young as three months has been 
 kept in check by treatment for a period which, in one instance, 
 extended to twentj'-five years. The var^'ing course run at different 
 ages probably depends on the fact that in early hfe provision has 
 to be made, not only for the maintenance of the tissues, but also 
 for their growth, hence the energy requirement is relatively high, 
 and shoiild the organs of metabohsm be congenitally weak, or 
 diseased, they quickly give way under the double strain ; in adult 
 life sufficient energy to maintain life onlj' has to be provided, so 
 that the loss of sugar can be better borne ; in old people the 
 activities of the body are normally lower than in early life, hence 
 less energy is required and its loss in the shape of sugar in the urine 
 can be still better withstood. 
 
 Sex. — If a number of cases of diabetes are observed it will 
 usually be found that the mortahty is higher among the females 
 than among the males. Thus Laache in his experience of 122 
 cases states that oidy seventeen out of seventy-seven male patients 
 died during the time, that forty out of forty- five cases in women ter- 
 minated fatally. The greater mortality among females is probabty 
 to be explained by the fact that diabetes is more common among 
 young women than among j^oung men, and hence runs a more 
 maMgnant course. 
 
 (2) The general nutrition of the patient is an important considera- 
 tion in forming a prognosis. As a rule thin, ill-nourished individuals 
 respond badly to treatment, and are very hable to develop acidosis 
 
PERSISTENT GLYCOSURIA 365 
 
 and succumb to diabetic coma, while in stout persons, and those 
 who are not much wasted, the course of the disease is usually slower, 
 and is more readily controlled by diet, &c. The former, too, are 
 more liable to develop fatal complications such as tuberculosis. 
 
 (3) The cause of the glycosuria cannot be determined with any 
 degree of certainty in many cases, but in others there is a probable 
 explanation, and in a few it can be definitely ascertained. The 
 prognosis in traumatic glycosuria is very frequently good, for 
 the symptoms often disappear spontaneously, or are transformed 
 into those of diabetic insipidus. Even in children this form of 
 glycosuria sometimes responds well to treatment. Pancreatic 
 diabetes is generally regarded as a very grave condition. Some 
 forms, and more particularly those secondary to intestinal troubles, 
 interacinar pancreatitis, and of course malignant disease of the 
 pancreas undoubtedly run a comparatively rapid course, but the 
 variety due to interlobular fibrosis that is associated with gall-stones, 
 typhoidal, and other forms of cholangitis, pancreatic cysts and 
 calculi usually progresses slowly, particularly if the primary source 
 of the trouble can be removed. It is in the latter type that analysis 
 of the faeces shows imperfect digestion of fats, a positive pancreatic 
 insufficiency test, &c., and from the results of such an analysis it 
 is possible to gauge the extent of the cirrhotic changes in the 
 pancreas, and so form an opinion as to the probable duration of the 
 disease. Gouty diabetes, and glycosuria associated with arterio- 
 sclerosis, are usually not grave conditions, provided that the patient 
 is in a good state of nutrition for his age, and will submit to treat- 
 ment. Glycosuria in obese individuals also runs a benign course 
 as a rule, if properly treated. Inherited infantile diabetes is a very 
 grave condition, terminating fatally in a few weeks to eight months 
 in thirty-seven out of forty-three cases collected by Lion and 
 Moreau. Once installed in a family the cUsease menaces all the 
 children, but the cases on record in which one or more have escaped 
 show that it does not necessarily affect every individual of the 
 family. The prognosis for persons belonging to families in which 
 an inherited tendency to glycosuria does not show itself until 
 adult life is much more hopeful, especially if it is not apparent 
 until after middle age is reached and is associated with gout, 
 asthma, &c., in the individuals affected, or in other members of 
 the family. 
 
 (4) TheDaily Excretion of Sugar. — Thegravity of acase of diabetes- 
 is very commonly estimated by the percentage of sugar contained 
 in the urine, but this is a most fallacious guide, even when the figure 
 is determined from an analysis of the mixed daily excretion. A 
 
366 GLYCOSURIA 
 
 better, but still misleading, way is to calculate the total output 
 for the twenty-four hours. An accurate estimate of the true state 
 of carbohydrate metaboHsm can only be arrived at by determin- 
 ing the relation existing between the total output of sugar in 
 the urine and the total intake in the food, including in the latter 
 the possible maximum of sugar that can be derived from the protein 
 of the diet,&c. , in the manner already explained. It is thus recognised 
 that the amount of sugar in the urine depends both on the quantity 
 of carbohydrate ingested, and upon the height of protein meta- 
 bohsm. In severe cases where the power of utiUsing sugar, in- 
 cluding even that derived from proteins, is entirely, or almost 
 entirely, lost the figures representing the intake and the output 
 will be nearly the same, but in less serious cases the intake will 
 exceed the output by an amomit varjdng with the extent to which 
 the power of metaboHsing carbohj^drates is retained. The re- 
 lationsliip is most conveniently expressed by Falta's, or Lusk's, 
 coefficient of excretion. 
 
 Another method of estimating the gravity of diabetes has been 
 suggested by Mendel and Lusk as the result of their observations 
 on the total nitrogen and sugar contents of the urines of animals 
 with severe experimental glj-cosuria. The patient is given a 
 meat-fat diet (rich cream, meat, butter, and eggs) and the twenty- 
 four hours urine of the second day collected in such a way that the 
 final specimen shall be taken at an early morning hour (before 
 breakfast). The discovery of 3-65 grams of dextrose for each gram 
 of nitrogen in the urine indicates complete intolerance for carbo- 
 hydj-ates, and probably a quickly fatal termination. The authors 
 have consequently called this (D : N : 3-65 : 1) " the fatal ratio." 
 A lower ratio of dextrose to nitrogen on this diet shows that some 
 protein sugar is being utihsed, and the prognosis is more favourable. 
 
 (5) Secondary Disturbances of Metabolism. — Quite as significant 
 as the assimilative capacity for sugar, is the quantity of organic 
 acid in the urine. A well-marked reaction for aceto-acetic acid 
 has long been accepted as an unfavourable sign, especially if it does 
 not disappear when the patient is carefully cUeted. But the presence 
 of acetone alone does not necessarily mean there is a serious degree 
 of acidosis. A more rehable opinion as to the probable outcome 
 of the case can be arrived at by looking for and estimating the 
 oxybutyric acid. Whenever a patient regularly passes more than 
 5 grams a day the prognosis is bad, for he is liable at any time to 
 become comatose. A decUne or disappearance of the acid on a 
 suitable diet is a favourable sign, but its gradual increase, in spite 
 of cUetetic regulations, is a bad omen. A patient may five for 
 
PERSISTENT GLYCOSURIA 367 
 
 many months, however, in spite of his urine containing an amount 
 of acid equivalent to 15 or 20 grams of oxybutyric acid a day. An 
 indication of the degree of acidosis and associated secondary dis- 
 turbances of metaboHsm is more readily obtained by estimating 
 the amount of ammonia nitrogen in the twenty-four hours urine ; 
 if this is found to be 3 grams, or over, the prognosis is unfavourable, 
 and coma is probably imminent. 
 
 (6) Response to Treatment. — Many cases of chronic glycosuria 
 which at first sight appear hopeless, will, under suitable dietetic, 
 medicinal, and hygienic treatment, improve in a most remarkable 
 manner. It is, therefore, never safe to give a definite prognosis 
 until the patient has been under observation for some time. Many 
 23hysicians divide their cases into " mild " and " severe " according 
 to the way they respond to a test-diet, such as that of v. Noorden. 
 This consists of meat, eggs, bacon, butter, green vegetables, cheese, 
 lettuce, salad, coffee, and wine. At breakfast and lunch 50 grams 
 of white bread are allowed. Should the urine be sugar-free on 
 such a diet the diabetes is of a " mild " type. If sugar is present 
 the quantity of bread is gradually reduced, and if the glycosiuria 
 still persists after all the bread has been removed from the diet the 
 case is regarded as one of " severe " diabetes. 
 
 (7) The social position of the patient is chiefly of importance 
 because the wealthier classes are able to devote time and money 
 to their cure which it is impossible for persons lower in the social 
 scale to expend ; moreover, people belonging to the higher grades 
 of society are, as a rule, more intelligent, and it is consequently 
 possible to impress them with the necessity of strictly carrying out 
 the line of treatment decided on. Their surroundings, too, are 
 usually more conducive to healthy metaboHsm and the avoidance 
 of infections. Hence, other things being equal, a person of good 
 social position has a better chance of life than one belonging to 
 the lower classes of society. 
 
 (8) Complications. — Certain compHcations are particularly 
 dangerous to life in diabetics. Pneumonia, for example, is a much 
 more fatal disease than it is in healthy individuals. Pulmonary 
 tuberculosis progresses very rapidly and is hardly ever materially 
 affected by treatment. Moist gangrene is, as Ave have seen, a 
 most serious complication, and makes the prognosis much more 
 grave, but is not necessarily hopeless. In women who are diabetic 
 and become pregnant, labour is the great danger to be feared, 
 especially if it is prolonged. 
 
368 GLYCOSURIA 
 
 BIBLIOGRAPHY 
 
 Abelmann, Dissertation, 1890. 
 
 Abt and Stroiise, Amer. Journ. Med. Sci., 1911. 
 
 Allan, Lancet, 1904. 
 
 Bainbridge, Biochem. Journ, 1908. 
 
 Bainbridge and Beddard, Biochem. Journ., 1906. 
 
 Beardsley, Therap. Qaz., 1911. 
 
 Blum, Seynaine Med., 1911. 
 
 Board U.S. Dept. AgTiculture, Rep. 94, Washington, 1911. 
 
 Bruce, Practitioner, 1887-8. 
 
 Briick, Mediz Klinik., iv. 
 
 Caessaet, Semaine Med., 1875. 
 
 Cammidge, Surg. Gynec. and Obst., 1908. 
 
 Cantini, Spec. Path. u. Therap. d. Stoffwech., 1880. 
 
 Carnot, Progres Medicale, 1910. 
 
 Cavazzani, Arch. d. din. med., 1893. 
 
 Charles, Bristol Med. Chi. Journ., 1906. 
 
 Cowles, Boston Med. and Surg. Journ., 1911. 
 
 Crofton, Amer. Journ. Med. Sci., 1902 ; Philadelph. Med. Journ., 1902 ; 
 
 Amer. Med., 1902; Lancet, 1909, 1911. 
 Crow, Johns Hopk. Hosp. Bull., 1908. 
 Dickinson, Dis. of Kidneys, 1875. 
 Dieulafoy, Acad. d. Med. d. Paris, 1910. 
 Dujardin-Beaumetz, Soc. d. Therap., 1888. 
 Ebstein, Die Zuckerkrank, 1887. 
 Evans, Lancet, 1907. 
 Evelt, Monat.f. Geb. u. Gyn., 1907. 
 Falta, Arch. int. Med., 1909. 
 Feinberg, Jahresb. u. d. Leitsung., 1889. 
 Forchheiiner, Amer. Journ. Med. Sci., 1911. 
 Forschbach, Deut. med. Woch., 1909. 
 Foster, Journ. Biolog. Chem., 1907. 
 Le Gendre, Journ. d. Med., 1911. 
 Conner, Correspbl. f. Schweiz. Aertze, 1887. 
 Griibe, MUnch. med. Woch., 1895. 
 Guelpa, Auto-intox. et desintox., 1910. 
 Hedon, Physiol. Normale et Path. d. Pane, Compt. rend. d. Soc. d, 
 
 Biol., 1911. 
 Henkel, Deut. med. Woch., 1909. 
 Herter, Lectures on Chem. Path., 1902. 
 Herzfeld, Journ. Amer. Med. Assot., 1911. 
 Hirschfeld, Berl. klin. Woch., 1910. 
 Hort and Penfold, Brit. Med. Journ., 1911. 
 Hucard, Rev. gen. d. Chem. et d. Therap., 1893. 
 Jennings, Lancet, 1911. 
 Koenig, Berl. klin. Woch., 1896. 
 Laache, Mediz. Klinik., 1910. 
 
PERSISTENT GLYCOSURIA 369 
 
 Labbe, Arch. gen. d. med., 1911. 
 
 Landergren, Nord. med. ark., 1910. 
 
 Lepine, Diabete Sucre, 1909. 
 
 Leschke, MiXnch. med. Woch., 1911. 
 
 Lichtwitz, Therap. Monatsch., 1911. 
 
 Lion and Moreau, Arch. d. Med. d. Enfants, xii. 
 
 Locke, Food Values, 1911. 
 
 Lofer, Berl. klin. Woch., 1911. 
 
 Lusk, Journ. Amer. Med. Assoc, 1910. 
 
 Levy, Johns Hopk. Hosp. Bull., 1911. 
 
 Mendel and Lusk, Deut. Arch. f. klin. Med., 1904. 
 
 Minkowski, Arch. f. exp. Path. u. Pharm., 1908. 
 
 Moore, Eden, and Abrani, Biochem. Journ., 1906. 
 
 Mosse, Rev. d. med., 1902. 
 
 Mosenthal, Journ. Amer. Med. Assoc, 1912. 
 
 Moullin, Lancet, 1907. 
 
 Neumann, Zeit. f. klin. Med., 1910. 
 
 Von Noorden, Centralb.f. inn. Med., 1895 ; Handb. d. Path. d. Stoffwech.,. 
 
 1907 ; Berl. klin. Woch., 1903 ; Rif. Med., 1911. 
 Oliver, Lancet, 1898. 
 Pembrey, Brit. Med. Journ., 1910. 
 Philip, Brit. Med. Journ., 1910. 
 Ralfe, Lancet, 1892. 
 Redon, Thesis, Paris, 1877. 
 Renne and Fraser, Biochem,. Journ., 1907. 
 Richardiere, Union Med., 1895. 
 Robson, Brit. Med. Journ., 1910. 
 Rudisch, Journ. Amer. Med. Assoc, 1909. 
 Schmitz, Zuckerkrank, 1892, 
 Sewall, Amer. Journ. Med. Sci., 1911. 
 Sicard, Soc med. d. hop. d. Paris, 1911 ; Journ. Amer. Med. Assoc.,. 
 
 1911. 
 Silverstin, Gaz. d. Osp. e. d. Clin., 1911. 
 Spooner and Pratt, Journ. Amer. Med. Assoc, 1910. 
 Thompson and Wallace, Brit. Med. Journ., 1911. 
 Wallace, Journ. Amer. Med. Assoc, 1910. 
 Warden, Lancet, 1911. 
 Weintraud, Therap. Monats., 1908. 
 West, Brit. Med. Journ., 1895. 
 Widal, Lemierre, and Cotoni, Semaine Med., 1911. 
 Williams, Brit. Med. Journ., 1894. 
 Williamson, Diabetes Mellitus, 1898. 
 Zuelzer, Deut. med. Woch., xxxiv. 
 
 2a 
 
CHAPTER X 
 
 OTHER CARBOHYDRATES MET WITH IN DIABETIC URmES — 
 LEVULOSURIA, MALTOSURIA, ETC. ETC. 
 
 LevulOSUria. — Owing to reliance being placed on methods of 
 analysis which are now regarded as open to objection, many of 
 the earUer recorded cases of levulosuria are of doubtfxil character, 
 but the tests employed by recent investigators have been such 
 that there can be no doubt that levulose does appear in the urine 
 spontaneously, both alone and along with dextrose. 
 
 According to modern research three forms of levulosuria exist — 
 (1) alimentary, (2) piu-e spontaneous, (3) mixed, in which there is 
 more or less dextrose along with the levulose. 
 
 1. Alimentary levulosuria has been already considered (p. 164), 
 so that it will not be necessary to deal with it in detail. It will be 
 remembered that, according to v. Noorden, the assimilation limit for 
 levulose is about 200 grams. In some individuals, however, it is 
 lower than this, 100 grams, or less, causing levulose to appear in 
 the urine. Alimentary levulosuria is rare in health, but examples 
 have been reported in apparently healthy persons by Moritz, 
 Haycraft, and Strauss. Of much greater interest than these 
 physiological curiosities is the proved association of ahmentary 
 levulosuria with functional distm-bances of the Hver, and, although 
 levulosuria is not pathognomonic of serious hepatic mischief, its 
 presence in any particular case is of considerable diagnostic value. 
 Levulose is a sugar that is frequently weU assimilated by diabetics, 
 but alimentary levulosuria is not uncommon in diabetes, both in 
 chronic cases and also in the recent severe type. In both the 
 failure to make use of levulose is probably dependent upon func- 
 tional derangement of the hver, and in some is associated with 
 actual structural change. Cases of this description have been re- 
 corded by a number of observers, including among the more 
 recent Borchardat, Graul, Schwarz, and Schleisinger. 
 
 2. Pure spontaneous levulosuria is a rare condition. Cases 
 apparently belonging to this category have been described by 
 Ventzke, Cotton, Personne and Henniger, Seegen, Kiilz, Carles, 
 Marie and Pobinson, Rosin-Laband, Schleisinger, Lepine and 
 
LEVULOSURIA, MALTOSURIA, ETC. 371 
 
 Boulud, Schwarz, Neubauer and Moraczewski, but the evidence in 
 support of some of them at least is doubtful. 
 
 The quantity of levulose excreted has always been small, rarely 
 exceeding 1 or 2 per cent., with a total daily excretion of 20 or 30 
 grams. Seegen, for example, found up to 2 per cent., Lepine up 
 to 2-4 per cent., Neubauer 1-8 per cent., Schwarz 0*3 per cent., but 
 as they do not state the total amount of urine passed the daily 
 excretion cannot be calculated ; in Rosin's case, however, 22 grams 
 were excreted in the twenty-four hours, and in Schleisinger's 3 to 4 
 grams. 
 
 3. Mixed levulosuria and dextrosuria is a much more common 
 condition than pure levulosuria ; but there is a very wide divergence 
 of opinion among different observers as to the frequency with wliich 
 levulose is met with in the urine of patients suffering from per- 
 sistent glycosuria. According to Rosin, dextrose and levulose are 
 found together " very often," Umber considers that they are " not 
 seldom " associated, Schwarz found levulose in the urines of six 
 out of nineteen cases of diabetes, Schleisinger in three out of eighteen ; 
 but in all of them there was a large amount of sugar in the urine, 
 owing to their not having been properly treated. Umber states 
 that sHght spontaneous levulosuria is so common in cases recently 
 admitted to hospital that it is probably physiological, and, as it 
 soon disappears on a careful diet, it is probably derived from the 
 food. 
 
 In many instances statements as to the presence of levulose in 
 the urine of cases of persistent glycosuria have been based entirely 
 upon the difference between the amount of sugar shown by titration 
 and polarisation and the presence of a positive Seliwanoff reaction, 
 but such evidence is not of itself conclusive. In the first place, the 
 accuracy of the ordinary titration methods for sugar in the urine is 
 not by any means as great as is generally assumed, and a difference 
 of a I to 1, or even 2, per cent, between the figures so obtained 
 and those given by the polariscope may easily be due to experi- 
 mental errors and the presence of other reducing substances. 
 Again, it must be remembered that albumen, yS-oxybutjnric acid, 
 glucuronic acid compounds, and cystin are levo-rotatory and that 
 their presence must be excluded, or allowed for, when the polari- 
 scope is being used for the detection and estimation of sugar. 
 Sehwanoff's test, unless very carefully carried out, is very hable to 
 give misleading results, and is, moreover, not strictly specific for 
 levulose. Reahsing these sources of possible error, trustworthy 
 investigators now usually confirm their results by separating the 
 levulose as the methylphenylosazone, or as a calcium, or other 
 
372 GLYCOSURIA 
 
 comparatively insoluble, compound. When examining a urine 
 for levulose it must be borne in mind that, according to Kiilz, 
 the levo-rotatory sugar met with in diabetic urines differs from 
 ordinary levulose in being precipitated by basic lead acetate. 
 
 The presence of levulose in the urine of a diabetic does not 
 necessarily mean that the patient's powers of assimilating that 
 sugar are defective. It will be remembered that levulose can be 
 artificially formed from dextrose by gently heating a faintly 
 alkahne solution of the latter, and it would seem that a similar 
 change sometimes takes place in the body. Such a spurious 
 levulosuria may apparently occur when large quantities of alkali 
 are being taken and the urine has an alkahne reaction. According 
 to Koenigsfeld, it is also met with when there is reduced gastric 
 acidity and increased intestinal alkahnity. The apparent diminu- 
 tion in the excretion of sugar that results from a course of treatment 
 with alkahne mineral waters, with a return to the former level on 
 the completion of the course, may possibly be explained in some 
 such way ; for if the sugar is estimated throughout the treatment 
 with the polariscope, or by titration, and the results are worked out 
 on the assumption that dextrose alone is present, readings that are 
 not really comparable will be obtained. The easiest way to guard 
 against such an error is to check the results by fermentation. 
 
 The amount of dextrose present in cases of mixed dextrosuria 
 and levulosuria varies very much. They may be conveniently 
 divided into two classes — (a) those in which the dextrosuria is 
 relatively shght, such as have been described by Zimmer, Czapeks, 
 Rohmann, May, Lion, and Neubauer ; (&) those in which the levulose 
 is associated with a considerable excess of dextrose, such as those 
 reported by Rosin and Laband, Dub, Schleisinger, Schwarz, and 
 Umber. 
 
 In either type the excretion of levulose has usually been found 
 to be under 2 per cent., with a total output of 30 grams, or less, a 
 day. The case described by Zimmer and Czapeks is, however, a 
 striking exception, for as much as 4*4 per cent, of levulose, with a 
 twenty-four hours' excretion of 176 grams, was there met with. 
 On some days 5-4 per cent, of dextrose, equivalent to 216 grams 
 in the twenty-four hours, was also passed. 
 
 Source of the Sugar. — The levulose present in the urine of some 
 cases of levulosuria is apparently of alimentary origin, being derived 
 from cane-sugar, honey, fruits, vegetables, &.c., contained in the 
 diet, for if such substances are excluded it disappears. Thus in 
 Neubauer' s case the levulosuria ceased in one or two days, and in 
 the cases reported by Lepine and Schwarz after a somewhat longer 
 
LEVULOSURIA, MALTOSURIA, ETC. 373 
 
 interval, when levulose-yielding foods were excluded. In Seegen's 
 case, and also in Rosin's and SchJei singer's, the levulosuria 
 diminished, but did not quite disappear. As ingested levulose 
 is apparently converted in the hver into glycogen, which subse- 
 quently breaks down into dextrose, it would seem that the failinre 
 to assimilate levulose on the part of these patients is due to the 
 sugar not being converted into glycogen in the ordinary way. 
 Neubauer found that a definite proportion (15 to 17 per cent.) 
 of the levulose given by the mouth was excreted in the mine in 
 his case. He therefore suggests that a certain proportion of the 
 levulose of the food is, under normal conditions, directly oxidised 
 without passing through the glycogen stage, and that a failure of 
 this oxidation may be the cause of levulosm-ia. Feeding experi- 
 ments by other observers have not, however, given similar results. 
 They show that the greater part of the levulose administered is 
 retained within the organism, and only a small part is excreted in 
 the urine, also that the latter does not bear any definite relation to 
 the whole. Inuhn, hke starch in mild cases of diabetes, appears to 
 be better tolerated than levulose, owing probably to its being only 
 slowly decomposed and absorbed from the intestine. 
 
 In some cases it has been found that when levulose was taken 
 by the mouth, even in doses calculated to exceed the limit of 
 tolerance of a normal person, no increase in the urinary levulose 
 has resulted. Neubauer, for instance, has described a case in 
 which, although large doses of levulose caused no increase in the 
 output of that sugar, the administration of dextrose brought about 
 an increased excretion of both dextrose and levulose, suggesting 
 that the dextrose was in part converted into levulose within the 
 body and excreted as such in the lu-ine. That such a conversion 
 of one sugar into the other can occur within the organism is also 
 suggested by Zimmer's case, for it is not Hkely that the whole of 
 the 176 to 92 grams excreted in the urine could be entirely derived 
 as such from the food. The administration of dextrose does not, 
 however, of necessity cause an^ increase in the excretion of levu- 
 lose in the urine. Other observers have noticed that carboh5^drate 
 foods containing no levulose caused an increased output of that 
 sugar. Thus Seegen states that bread brought about this result 
 in his case, and Schwarz found that with one of his patients, whose 
 urine had been made sugar-free by diet, the use of grain-foods caused 
 both the dextrose and levulose to reappear. That levulose can be 
 formed from dextrose within the body is also suggested by the 
 experiments of Hedon, who demonstrated levulose in the blood 
 of depancreatised dogs. According to Schleisinger, injections of 
 
374 GLYCOSURIA 
 
 phlorhidzin in cases of levulosuria only cause the appearance of 
 dextrose in the urine. 
 
 Symptoms. — When levulose alone is present in the urine the 
 symptoms are of a mild type, Hke those met with in cases of shght 
 glucosuria. Polyuria is absent, the specific gravity of the urine 
 is not high, the amount of sugar is small, and there is no thirst, 
 wasting, or other characteristic sign. In one recorded case the 
 levulose was discovered by accident in the urine of a patient suffer- 
 ing from transverse myehtis, in two obesity was a concomitant 
 symptom, in some there has been a family history of diabetes, and 
 in several nervousness and mental depression have been the most 
 noticeable feature of the case. Since the presence of levulose in 
 the urine appears to depend upon interference with the functions 
 of the Hver in many instances, it is possible that the last-named 
 symptoms may be also dependent upon this, and arise from in- 
 testinal toxines having free access to the systematic circulation. 
 Levulose in association with dextrose in the urine is not apparently 
 accompanied by any particular symptoms, and may be met with 
 in mild as well as in severe cases of diabetes. The significance of 
 its presence is not certain, but it is probable that it indicates a 
 functional derangement of the liver, which may be of a temporary 
 or permanent character. 
 
 Detection. — In cases of pure levulosuria the recognition of the 
 sugar is not a difficult matter. The urine gives the reduction tests 
 with copper, bismuth, &c., is fermented by yeast, and forms with 
 phenylhydrazin an osazone with the same melting-point as 
 dextrosazone. It is, however, levo-rotatory, and gives SehwanofE's 
 reaction by either Rosin or Borchardat's modification. To exclude 
 other levo-rotatory substances, the urine should be fermented with 
 yeast and again examined with the polariscope. It is also advisable 
 to control SehwanofE's test by repeating it with the fermented urine 
 to prove that it fails after the removal of the levulose. Finally, 
 the levulose can be separated out as the methylphenylosazone, or 
 as the calcium compound, and its properties investigated. 
 
 When the urine contains both levulose and dextrose Seliwanoff's 
 reaction is a useful prehminary test, for when it is positive before, 
 and negative after, fermentation it points to the presence of a 
 ketose. If it is also found that the percentages of sugar shown 
 by fermentation, or titration, and by polarisation of the urine are 
 not the same it tends to confirm this conclusion, especially if the 
 difference is marked. 
 
 The percentage of levulose and dextrose may be calculated from 
 
LEVULOSURIA, MALTOSURIA, ETC. 375 
 
 the quantities of sugar found on fennentation and by the polariscope 
 as follows : — 
 
 D + L = " a " per cent, by fennentation. 
 
 D— L = " b " per cent, with the polariscope. 
 
 2D ={a + b) 
 
 _^ (a + b) , T- /a + b 
 
 . • . D = ^ — — - , and L = a- [ -—— 
 
 where D = dextrose, and L = levulose. 
 
 If the urine shows a levo -rotation " c " after fermentation, owing 
 to the presence of beta-oxybutyric acid, &c., this must be allowed for : — 
 
 D + L= " a " per cent, by fermentation. 
 
 D — L="6" + "c" per cent, on polarisation. 
 
 2D ={aVo + c) 
 . D =(^±|±i), and L = a-(^+|±-^) 
 
 If the sugar is estimated by titration and polarisation, the per- 
 centage of levulose is found by dividing the difference between the 
 percentages of sugar obtained by the two methods, by 2*69, provided 
 that other levo -rotatory substances are absent, since 1 gram of levulose 
 is equivalent in its reducing powers for Fehling's solution to 0-925 
 gram of dextrose, and a 1 per cent, solution of levxilose turns the ray 
 of polarised light — 0'93° to the left ; therefore a 1"76 per cent, solution 
 is as strongly levo -rotatory as a 1 per cent, solution of dextrose is 
 dextro-rotatory. Hence — 
 
 «D-f 2/ (0*93) = " a " per cent, by reduction. 
 rrD-ii/ (1-76) = " b " per cent, by polarisation. 
 2/ (0-93 + 1-76) = (a -6). 
 . Ja-b) 
 ' • ^ 2-69 
 
 Should the percentages obtained on titration and on polarisation 
 be approximately equal, it is probable that some other levo-rotatory 
 substance besides levulose is present. In this ease, or when the 
 presence of such bodies is suspected for other reasons, the optical 
 activity of the urine after complete fermentation inust be determined, 
 and allowed for as shown above. 
 
 The levulose may be isolated as the methylphenylosazone, 
 according to the method of Neuberg and Strauss : — 
 
 The urine is made acid by adding a few drops of acetic acid, boiled, 
 and filtered. It is then evaporated to a syrup at 40° C, mixed with 
 half its vohime of 98 per cent, alcohol, heated for five minutes, cooled, 
 and filtered. If the residue still possesses reducing powers the treat- 
 ment with alcohol is repeated once or twice. The alcoholic extracts 
 are filtered from any flocculent precipitate that may have formed, and 
 decolorised with animal charcoal. The quantity of sugar present is 
 then determined by titrating a sample. The remainder is evaporated 
 
376 GLYCOSURIA 
 
 to a small bulk (30 c.c), and mixed with methylphenylhydrazin, allow- 
 ing 3 molecules for each molecule of sugar. After standing for one 
 liour, any precipitate that has formed is filtered off, and the filtrate 
 mixed with an equal volume of 50 per cent, acetic acid, and sufficient 
 alcohol to give a clear solution. The mixtixre is heated in a boiling 
 water bath for three to five minutes, or left at 40° C. for twenty-four 
 hoiirs. If a considerable amoimt of levulose is present the methyl- 
 phenylosazone crystals separate out directly, or on adding a few drops 
 of water. If only a small amount is present an oily precipitate forms. 
 The osazone can be separated from this in a crystalline form by cooling 
 and treating it with carbon dioxide and ether. The crystals are purified 
 by recrystallising them from absolute alcohol in the cold, and may be 
 fvu*ther purified by dissolving in hot water, to which a Uttle pyridin 
 has been added. The solution is boiled with animal charcoal, filtered, 
 and the crystals separated out. Methylphenyl-levulosazone appears 
 as fine yellow crystals with a melting-point of 158° to 160° C. A solu- 
 tion (0-2 gram) in pyridin-alcohol (4 grams pyridin, 6 grams absolute 
 alcohol) is dextro-rotatory ( + 1° 40'). 
 
 MaltOSUria. — When maltose is present in the urine it is nearly 
 always associated with dextrose, although a few examples of pure 
 maltosuria have been described. No case of maltosuria appears to 
 have been reported previous to the year 1888, whenLe Nobel stated 
 that he had found maltose in a urine examined bj^ him. The next 
 year another case was described by v. Ackeren, and subsequently 
 others were reported by Rosenheim and Flatow, Charin and Brocard, 
 Lepine and Boulud, Kottmann, Geelmuyden, Rosenberger, and 
 Magnus- Levy. In the earlier cases the diagnosis was based entirely 
 on the difference between the results obtained on titration and on 
 examining the urine with the polariscope, and on the alteration 
 produced by heating with hydrochloric acid ; but as the differences 
 observed were always very slight, and might be accounted for by 
 experimental errors and in other ways, the true significance of the 
 results obtained in this way is doubtful. Later observers have 
 separated the phenyiosazone and based their conclusions mainly 
 on its melting-point (202° to 208° C), its nitrogen content (10-6 per 
 cent.), its solubiHties, and the effects of a solution in pyridin-alcohol 
 on polarised Hght ( -f 1° 30'). 
 
 In most of the recorded cases of maltosuria the calculated amount 
 of maltose has been under 0*5 per cent. Magnus-Levy states 
 that he met with 1*5 per cent, in association with 2 per cent, of 
 dextrose in one case, but that the maltosuria only lasted for two 
 days, and appeared to be due to the consumption of a quantity of 
 beer by the patient. 
 
 Maltose was discovered by Geelmuj^den in the urines of seven 
 
LEVULOSURIA, MALTOSURIA, ETC. 377 
 
 out of nine cases of diabetes by means of a special method of 
 separating the osazone that he employed. According to Lepine 
 and Boulud, a small quantity of maltose is not rarely present in the 
 urine of diabetics when they first come under observation. This 
 they consider is in part derived from the food, as it frequently 
 disappears when the patient is put on a strict diet. In some 
 instances, however, the maltosuria persists even when the patient 
 is taking only nitrogenous and fatty foods, so that it must also 
 have some other source of origin. Since they found from 0*2 to 
 0'3 per cent, of maltose in the inrine of depancreatised dogs, which 
 had been kept on a purely nitrogenous diet for some time prior to 
 the operation, Lepine and Boulud suggest that the presence of this 
 sugar in the urine of diabetics may be due to imperfect transforma- 
 tion of glycogen consequent on disease of the pancreas. It is true 
 that maltose has been detected in the urine during life in cases in 
 which disease of the pancreas was found post-mortem, but they are 
 too few to prove that there is any causal connection between the 
 two. Rosenheim's case passed 0*1 to 0*5 per cent, of maltose in his 
 urine, had fatty stools, lost thirty pounds in weight in nine months, 
 and after death interstitial pancreatitis was found. In Kottmann's 
 case of diabetes with maltosuria, atrophy of the pancreas was dis- 
 covered post-mortem. I have had the opportunity of examining 
 specimens of urine from a large number of cases of typical pancreatic 
 disease, over two thousand, and I have only met with two in which 
 an osazone having the characters of maltosazone was obtained in 
 sufficient quantity for an accurate investigation, and five in which 
 a small deposit of crystals, probably also maltosazone, was given. 
 One of the former was a patient in whom an operation for stone in 
 the common duct with chronic pancreatis had been performed four 
 years previousl}^ 
 
 Maltosuria has been observed by Charin and Brocard in lying-in 
 women, and Rosenberger met with a sugar resembhng maltose in a 
 case of croupous pneumonia. 
 
 Detection.— Maltose reduces alkaline solutions of copper and 
 bismuth, is strongly dextro-rotatory (-f 137°), and is easily fer- 
 mented by yeast. Its detection in the urine is usually based upon 
 the difference obtained on examining with the polariscope and by 
 titration. The optical activity of maltose is about two and a half 
 times greater than that of dextrose, while its reducing power is 
 only about two-thirds as great. As other substances, such as 
 pentoses, lactose, glucuronic acid compounds, oxybutjTic acid, and 
 amino acids can also cause a difference in the readings, it is only 
 when the urine contains a considerable quantity of maltose that a 
 
378 GLYCOSURIA 
 
 satisfactory diagnosis can be made in this way. Further evidence 
 can be obtained by preparing the phenylosazone, which is charac- 
 terised hj its solubility in hot water, its melting-point of 202° to 
 208° C, and its optical activities. Its solution in alcohol is dextro- 
 rotatory, the pyridin-alcohol solution is also dextro-rotatory 
 ( + 1° 30'), but its solution in glacial acetic acid is levo-rotatory. 
 
 Isomaltose. — Isomaltose is stated to have been recovered from 
 normal uriiie in small quantities bj^ Baisch, Lemaire, Porcher, and 
 Reinbold by the benzoyling process, and b}^ Pavy and Siau as 
 isomaltosazone. 
 
 The question as to whether it exists preformed in the urine, or 
 is derived from dextrose in the process of separation, has not yet 
 been settled, even by those who believe in its existence, and some 
 observers deny that it does. Mayer points out that the reactions 
 described as characteristic of isomaltose are also given by glucuronic 
 acid, and that to depend upon the melting-point of the osazone 
 as the chief distinguishing feature, as some observers have done, 
 is most unsatisfactor3^ According to Cremer an ahmentary 
 isomaltosuria is possible, and it may be that the traces met with, 
 in some urines are of intestinal origin. Wohl and others have 
 shown that isomaltose is very readily formed in small quantities 
 by digesting dextrose with dilute hydrochloric acid. Rosin and 
 Alfthan have found isomaltose in diabetic urines, and it is con- 
 sidered by Pavy and Siau that it is to the presence of this substance 
 that the increased reduction, shown by some diabetic urines after 
 heating with an acid, is due. 
 
 Detection. — Isomaltose, Hke maltose, reduces alkaline solutions 
 of copper and bismuth, but while maltose is fermented by yeast, 
 isomaltose is not. They are most readily distinguished by the 
 characters of their osazones, that of isomaltose melting at 150*^ 
 to 153° C, while maltosazone melts at 202° to 208° C. ; moreover, 
 isomaltosazone can be obtained from the urine after any dextrose, 
 levulose, or maltose that may be present has been removed by 
 fermentation. 
 
 Laiose. — This sugar was isolated by Leo from the urines of 
 three out of twenty-one diabetics. Its presence was inferred from 
 the quantitative estimation of dextrose by titration, being 1-2 to 
 I'O per cent, more than was shown by the polariscope. In one 
 case the urine was optically inactive, and was found on titration to 
 contam 1'8 per cent, of dextrose. Leo subsequently isolated the 
 sugar and investigated its properties. It is distinguished from other 
 sugars by its salty, rather than sweet, taste, its slight reducing 
 
LEVULOSUEIA, MALTOSURIA, ETC. 379 
 
 powers as compared with dextrose (0-4 : 1), by being imfermented 
 by yeast, by being levo-rotatory ( - 26° 7'), and by its forming 
 with phenylhydrazin a yellowish-brown, non- crystalline, oil, that 
 is insoluble in water, but is soluble in alcohol. It has been 
 variously regarded as a hexose and as a pentose (d-xylose ?). 
 
 Heptose. — From the urine of one case of diabetes Rosenberger 
 separated a sugar which in many respects resembled laiose, but was 
 considered by him to be a heptose. The isolated sugar was ob- 
 tained as a brown fluid which reduced alkahne solutions of copper, 
 and formed with phenylhydrazin an osazone with a melting-point 
 of 195° C, corresponding to that of a heptose. A solution of the 
 osazone in pyriclin was optically inactive. The urine from which 
 the sugar was derived was levo-rotatory, but the sugar itself was 
 found to be optically inactive. 
 
 Paidose. — Under this name Geelmuyden described a sugar 
 that he found in the urine of diabetic children. It was optically 
 inactive, or only very feebly active, slowly reduced Fehling's solu- 
 tion, and yielded an osazone with a melting-point of 175° to 190° C. 
 The orcin and phloroglucin reactions were negative. 
 
 Pentoses. — Some diabetic urines contain traces of a pentose, 
 but this question will be more fully dealt with when chronic pen- 
 tosuria is considered (see mixed pentosuria and dextrosuria, p. 396). 
 
 Glycog'en (Erythro-dextrin). — Reichardt found that the 
 urines of several diabetics that he examined after the complete, or 
 almost complete, disappearance of the sugar still reduced alkaline 
 solutions of copper on prolonged boiling. This he attributed to 
 the presence of a dextrin-like substance that was coloured reddish- 
 brown by iodine, which he isolated from the urines. Leube ob- 
 tained a similar substance from the urines of two cases of diabetes, 
 and considered it was glycogen. He was unable to find it in the 
 urines of healthy people, or those suffering from diabetes insipidus. 
 The exact nature of this body is not certain, but on physiological 
 grounds it is more likely to be glycogen than erythro-dextrin. 
 
 Animal Gum (Landwehrj. — This, which is probably not a 
 single substance but a mixture, is said to occur in normal urines in 
 quantities of 0*1 to 0*2 grams daily. Alfthan found that it was 
 increased in diabetes meUitus, as much as 1*2 to 36-9 grams being 
 excreted in the twenty-four hours. It is shghtly dextro-rotatory, 
 is not fermented by yeast, is not colom-ed by iodine, and gives 
 with copper a precipitate that is insoluble in alkahes and does 
 not darken on boiling. 
 
380 GLYCOSURIA 
 
 Inosite. — Inosite has the same empirical formula as the hexoses 
 (CgHjaOg), but it belongs to the aromatic series, being hexa- 
 hydroxybenzol, and is not, as was at one time thought, a carbo- 
 hydrate. It is, however, convenient to consider it here. 
 
 It was said by Neukomm, Cloetta, Gallois, and Kiilz that inosite 
 is not present in normal urine in demonstrable amounts, but Rosen- 
 berger and Starkenstein found it in every urine they examined in 
 quantities up to about 0-08 grams in the twenty-four hours. 
 According to Strauss excessive water drinking, with consequent 
 polyuria, gives rise to a varying degree of inosituria. 
 
 Inosite has been found in the urine in increased quantities in 
 three pathological conditions — namely, diabetes insipidus (Vohl, 
 Strauss, and Kiilz), nephritis (Cloetta and Kiilz), and diabetes 
 mellitus (Vohl, Gallois, Kiilz, and Lava). In the latter condition 
 it is not constantly present, and was only found by Kiilz and Lava 
 five times in thirty cases, and in all of these there was marked 
 polyuria. 
 
 Detection. — Inosite does not reduce alkaline solutions of copper 
 and bismuth. It is precipitated by lead acetate, is optically in- 
 active, and does not ferment with yeast ; it is, however, decomposed 
 by B. lactis with the formation of lactic acid, and subsequently 
 yields butyric acid. It does not form an osazone with phenylhydrazin. 
 It may be isolated and recognised as follows : — 
 
 Cooper-Lane Method. — Any albumen that may be present is re- 
 moved by boiling and filtering. The phosphates are then precipitated 
 out with baryta water, and the filtrate, after being heated, is treated 
 with lead acetate, avoiding an excess. The mixture is allowed to 
 stand for some time, and the precipitate that has formed is then 
 filtered off, washed, suspended in water, decomposed with sulphuretted 
 hydrogen, filtered, and after standing for some time to allow the uric 
 acid to separate, the filtrate is evaporated to a small bulk. The 
 creatinin is removed by mixing the residue with one to four volumes of 
 alcohol, and boiling. If a heavy precipitate which sticks to the glass 
 forms, the clear, hot, alcoholic fluid is simply decanted, but if it sepa- 
 rates out in flocculi it is filtered hot through a warm filter, and then 
 allowed to cool. The fluid is then left to stand for twenty-four hours. 
 If inosite is present in appreciable quantity it will separate out in 
 crystals which may be filtered off and washed with cold alcohol. If 
 no crystals appear, the inosite may be separated in mother-of-pearl 
 plates by adding ether, little by little, avoiding an excess, to the clear 
 alcoholic solution until a slight milkiness, that does not disappear, 
 results, and leaving in the cold for twenty-four hours. 
 
 The crystals of inosite are rhombohedral in form and melt at 
 225° C. They are soluble in water (1 : 75), but are insoluble in alcohol 
 and ether. They may be recognised by the following tests : — 
 
LEVULOSURIA, MALTOSURIA, ETC. 381 
 
 1. Scherer's Test. — A small quantity of the precipitate is mixed with 
 nitric acid on platinum foil, and evaporated almost to dryness. To 
 the residue are added a little ammonia and a drop of calciiim chloride 
 solution, and the evaporation continued to dryness. If inosite is 
 present a beautiful rose colour results. Unless the crystals are fairly 
 pure a typical reaction is not obtained. 
 
 2. SeideVs Test. — This test is carried out in the same way as the 
 preceding, except that strontium acetate is used instead of calcium 
 chloride. It gives a green colour with a violet precipitate. A positive 
 reaction is obtained with 0-3 mg. of inosite (Fick). 
 
 BIBLIOGRAPHY 
 
 Levulose 
 
 Borchardat, Zeit. /. physiol. Chem., 1908 ; Munch, med. Woch., 1909. 
 
 Carles, Chem. Zentralb., 1890. 
 
 Cotton, Bull. d. Soc. Chem., 1880. 
 
 Czapeks, Prager med. Woch., 1876. 
 
 Dub, Dissertation, Leipzig, 1902. 
 
 Graul, Centralh. f. inn. Med., 1905. 
 
 Haycraft, Zeit. f. physiol. Chem., 1894. 
 
 Koenigsfeld, Zeit. f. klin. Med., 1909. 
 
 Kiilz, Zeit.f. Biol., 1884, 1890. 
 
 Lepine and Boulud, Rev. d. Med., 1904. 
 
 Lion, MiXnch. med. Woch., 1903. 
 
 Marie and Robinson, Bull. Soc. med. d. hop. d. Paris, 1897 ; Semaine 
 
 Med., 1897. 
 May, Arch. /. klin. Med., 1896. 
 Moraczewski, Zeit. f. klin. Med., 1907. 
 Moritz, Centralh. f. inn. Med., 1891. 
 Neubauer, Munch, med. Woch., 1905. 
 Neuberg and Strauss, Zeit. f. phys. Chem.., 1902. 
 Personne and Henniger, Bull. d. Soc. d. Chem., 1880. 
 Rohmann, Centralh. J. inn. Med., 1884. 
 Rosin and Laband, Zeit.f. klin. Med., 1902 ; Ber. d. Deut. Chem. Gesellsch., 
 
 1902. 
 Schleisinger, Arch. f. exp. Path., 1903. 
 Schwarz, Arch. f. klin. Med., 1903. 
 Seegen, Centralh. f. d. med. Wiss., 1884. 
 Strauss, Deut. med. Woch., 1901. 
 Umber, Salkowski's Festschr., 1904. 
 Ventzke, Journ. f. prakt. Chem., 1842. 
 Zimmer, Deut. med. Woch., 1876. 
 
 Maltose 
 
 Ackeren, Berl. klin. Woch., 1889. 
 
 Charin and Brocard, Compt. Rend. d. Soc. d. Biol., 1898. 
 
382 GLYCOSURIA 
 
 Geelmuyden, Zeit. f. klin. Med., 1905. 
 
 Kottmann, Dissertation, Geneva, 1901. 
 
 Lepine and Boulud, Compt. Rend. d. Acad. d. Sci., 1901. 
 
 Levy, V. Noorden's Handb. d. Path. u. Stoffwech., 1907. 
 
 Le Nobel, Arch. f. klin. Med., 1888. 
 
 Rosenberger, Deut. med. Woch., 1906. 
 
 Rosenheim and Flatow, Berl. klin. Woch., 1898. 
 
 ISOMALTOSE 
 
 Alfthan, Deut. med. Woch., 1900. 
 Baisch, Zeit. f. phys. Chem., 1894. 
 Cremer, Zeit. f. phys. Chem., 1892. 
 Lemaire, Zeit. f. phys. Chem., 1895. 
 Mayer, Zeit. f. phys. Chem., 1901. 
 Pavy and Siau, Journ. of Physiol., 1901. 
 Porcher, Chem. Zeit., 1902. 
 Reinbold, Pfluger's Arch., 1902. 
 Rosin, Deut. med. Woch., 1900. 
 Wohl, Ber. d. Chem. Gesellsch., 1890. 
 
 Laiose 
 
 Fischer, Lieb. Ann., 1892. 
 Leo, Virchow's Arch., 1887. 
 
 Heptose 
 Rosenberger, Zeit. f. phys. Chem., 1906. 
 
 Paidose 
 Geelmuyden, Jahresber. f. Tiersch., 1903. 
 
 Glycogen 
 
 Kotake, Zeit. f. phys. Chem., 1910. 
 Leube, Virchow's Arch., 1888. 
 Reichardt, Arch. d. Pharm., 1874. 
 
 Animal Gum 
 
 Alfthan, Dissertation, Helsingfors, 1900, 1904. 
 
 Baisch, Zeit. f. phys. Chem., 1894. 
 
 Freund, Zentralb. f. Physiol., 1892. 
 
 Landwehr, Zeit. f. phys. Chem., 1882-3-4-5; Centralb. f. d. Med. 
 
 Wiss., 1885 ; Pfluger's Arch., 1886-7. 
 Lemaire, Zeit. /. phys. Chem., 1895. 
 Morner, Skandin. Arch., vi. 
 Reinbold, Pfluger's Arch., 1902. 
 Salkowski, Berl. klin. Woch., 1905. 
 Wedenski, Zeit. f. phys. Chem., 1888. 
 
LEVULOSURIA, MALTOSURIA, ETC. 383 
 
 Inosite 
 Cloetta, Lieb. Ann., 1856. 
 Cooper-Lane, Ann. Chem. Pharm., 1861. 
 Fick, Chem. Zentralb., 1887. 
 Oallois, De V Inositurie, 1864. 
 Kiilz, Maly's Jahresb., 1875-6. 
 Lava, Arch. f. klin. Med., 1891. 
 Maquenne, Bull. d. Soc. Chem., 1887. 
 Neiikomm, Dissertation, Zurich, 1859. 
 Rosenberger, Miinch. m,ed. Woch., 1908. 
 Scherer, Ann. d. Chem. u. Pharm., Ixxxi. 
 Starkenstein, Zeit. f. exp. Path. u. Therap., 1908, 
 Strauss, Centralb. f. inn. Med., 1872. 
 Vohl, Arch. phys. Heilk., 1858. 
 
CHAPTER XI 
 
 LACTOSUKIA, GALACTOSURIA, SACCHAROSUEIA, PENTOSURIA AND 
 GLUCURONIC ACID 
 
 LaetOSUPia. — Next to dextrose the commonest sugar met with 
 in the urine is lactose or milk-sugar. It is usually found in women 
 in connection with pregnancy, or during lactation, but may also 
 occur as an alimentary lactosmria under other conditions. 
 
 Alimentary Lactosuria. — The assimilation Hmit of a healthy adult 
 for milk-sugar is stated by Worm-Miiller to be 100 grams, but 
 Halasz has given 150 grams without producing lactosuria. For 
 healthy children the Umit of tolerance is, according to Grosz, 8-6 
 grams per kilo of body-weight. Occasionally 100 grams of lactose, 
 taken in one dose, will cause the appearance of sugar in the urine, 
 and a few cannot take 80, or even 50, grams of milk-sugar without 
 excreting same. This does not indicate any disorder of metabohsm, 
 but depends upon a defect in the lactose- splitting ferment of the 
 intestine, which allows a certain proportion of unaltered milk- 
 sugar to be absorbed into the blood, whence it is excreted into the 
 lu-ine, for lactose cannot be converted into glycogen until it has 
 been inverted, and this inversion can only take place before, or 
 during, absorption from the intestine. 
 
 It has been found that in some diseases of the gastro-intestinal 
 tract the assimilation hmit for lactose is considerably lowered, 
 Grdsz met with 2 per cent, of milk-sugar in the urine of twenty-two 
 out of twenty-three cases with gastric disorders, mostly carcinoma 
 with dilatation, when they were given 150 grams of lactose fasting. 
 Mehu states that lactose is sometimes found in the urine of the 
 patients who have taken large quantities of milk over considerable 
 periods. Grosz, and later Langstein and Steinitz, showed that 
 the reducing substance found in the urines of infants suffering 
 from gastro-intestinal disorders is lactose. The latter, also Meyer 
 and Zuelzer, have met with galactose under the same conditions 
 in some instances, the monosaccharide constituting the bulk of the 
 sugar in one instance. According to Grosz, the assimilation Hmit 
 in these cases may be as low as 2-0 to 2-9 grams per kilo. Lying- 
 in women appear to have a lowered assimilation hmit for lactose. 
 
LACTOSURIA 385 
 
 as well as for dextrose and levulose, 100 grams, and in some even 
 50 grams, causing slight lactosuria. Zuelzer and Hess state that 
 lactosuria can also be produced after abortion, and in parturient 
 women, by administering 150 grams of dextrose. 
 
 Lactose, taken as such, or in the form of milk, by diabetics is 
 not excreted unchanged but appears as dextrose, the proportion 
 varying in different cases. In the mild type a part, or even the 
 whole, is often made use of, but in severe cases an equivalent 
 amount of dextrose is passed in the urine. Thus, in a case reported 
 by Borchardat and Finkelstein, 100 grams of milk-sugar given by 
 the mouth were entirely excreted as dextrose. Bourquelot and 
 Troisier, and subsequently Voit, found lactose in the urines of 
 diabetics who had consumed a large amount of milk, but only in 
 very small quantities. 
 
 Spontaneous lactosuria is confined to women, and only occurs 
 during the later months of pregnancy and after childbirth. As far 
 back as 1849 it was noticed by Heller that a reducing substance 
 may be met with in the urine of women during the period of lacta- 
 tion, and in 1877 it was definitely proved by Hofmeister and 
 Kaltenbach that this substance is milk-sugar. Lactosuria is a 
 much more common phenomenon in nursing women than is 
 generally supposed, and milk-sugar is probably seldom entirely 
 absent from the urine at some time or other during that period. 
 It is most generally met with during the first few days after labour, 
 but may be found as late as six months subsequently. According to 
 MacCann the commonest time is between the fourth and fifth day 
 of the puerperium. Ney found lactosuria in 115 out of 148 (77 per 
 cent.) parturient women that he examined, but only in 16 per cent, 
 of those who were pregnant. Gerard also found evidence of tem- 
 porary lactosuria in only five out of forty-one (12 per cent.) pregnant 
 women approaching term, while Lemaire detected lactose in the 
 urine of eighteen out of nineteen lying-in women under his care. 
 Blumenthal states that some 80 per cent, of women who nurse 
 their children have lactosuria, but that only some 20 per cent, of 
 suckling mothers pass milk-sugar in their urine during the period 
 that the child is being fed, when, however, suckling is suddenly 
 stopped, or the breasts become engorged, or inflamed milk-sugar 
 generally makes its appearance in the urine. 
 
 As a rule the lactosuria is of short duration, but occasionally 
 it lasts for some time, persisting for as long as five months in a 
 case reported by Pavy. The quantity of sugar excreted is usually 
 small, seldom exceeding 1 per cent., and there is no polyuria. The 
 total daily output is generally under 10 to 20 grams. Naunvn, 
 
 2b 
 
386 GLYCOSURIA 
 
 however, met with cases in which the urine contained 2 to 3 per cent, 
 of lactose, Lemaire found 4 per cent., and Porcher has reported 
 cases with as much as 7 per cent. The administration of dextrose 
 increases the excretion of milk-sugar. Hess found that after 
 administering 150 grams of dextrose, from 8-4 to 19-5 grams of 
 lactose appeared in the urine. 
 
 The milk-sugar excreted in the urine in spontaneous lactosuria 
 is formed in the mamn:ary glands. Porcher has shown that if 
 these organs are removed in cows and goats immediately before 
 parturition, lactosuria does not occur, but that a glycosuria of 
 variable intensity rapidly sets in and lasts about twenty-four hours. 
 A similar result follows amputation of the breasts in lactating 
 animals, and is associated with hyperglycsemia, showing that in 
 the absence of the mammary glands dextrose is not converted into 
 lactose but is excreted, in the urine. 
 
 Diagnosis. — It is most important that lactosuria should be 
 carefully distinguished from glycosuria, since the former is temporary 
 -and does not indicate a serious disturbance of metaboHsm, whereas 
 the latter, as we have seen, is of very grave significance in pregnant 
 and parturient women, and demands prompt attention. The com- 
 plete identification of lactose in the urine necessitates its isolation 
 hy a lengthy process, but for clinical work this is not usually neces- 
 sary, as its presence can generally be established with a reasonable 
 degree of probabiHty by considering the clinical characters of the 
 case in conjunction with the results of a series of tests applied 
 directly to the urine. 
 
 Whether the urine contains lactose or glucose it will reduce 
 alkaline solutions of copper and bismuth, but with lactose the re- 
 duction is not as prompt as it is with dextrose, and does not take 
 place until the mixture is boiled. When lactose alone is present 
 the fermentation test will be negative during the first twenty-four 
 hours, and since phenyl-lactosazone does not readily crystallise out 
 except from pure solutions of the sugar, the phenylhydrazin test 
 will not yield a definite crystalhne precipitate. If, however, the 
 urine is boiled with 5 per cent, sulphuric acid for a short time, 
 neutraHsed with ammonia, and then treated with phenylhydrazin, 
 it will give crystals of glucosazone and galactosazone. The 
 reducing power of the urine will also be found to be increased by 
 boiling with a mineral acid and neutralising, but it is unchanged 
 after boihng with citric acid. Treatment with sulphuric acid does 
 not U5.ually affect the result of the fermentation test, as it might be 
 expected to do theoretically, since the large amount of sulphate 
 formed interferes with the growth of the yeast. Voit's, or Buchner's, 
 
GALACTOSURIA 387 
 
 modification of Rubner's test, and the Wohlk (Malfatti) reaction 
 may also be employed to confirm the results obtained by these 
 means. Lactosuria can be distinguished from pentosuria by the 
 phloroglucin, and orcin tests, &c. 
 
 Before leaving the subject one other cause of the presence of 
 lactose in the urine must be .mentioned, namely, malingering. 
 Hysterical persons, especially women, soldiers, and others, have 
 been known to inject milk into the bladder, or to mix it with their 
 urine after it has been passed, with a view to simulating diabetes or 
 ohyluria. If the possibihty of such a deception is borne in mind 
 it is not difficult to detect. In a case that came under my observa- 
 tion condensed milk was mixed with the urine, and the fraud was 
 exposed by discovering both lactose and cane-sugar, as well as by 
 the microscopical characters of the emulsion. 
 
 Galactosuria. — In healthy people, according to Bauer, 20 
 grams of galactose taken by the mouth do not give rise to galactos- 
 uria ; after 40 grams about 1 gram may be excreted in the urine un- 
 changed, but with 100 grams galactosuria always results. When 
 there is cirrhosis of the liver, galactose is not as well tolerated as in 
 health, 1 gram appearing unchanged in the urine after the ingestion 
 of 20 grams, and 4 grams after 40 grams have been taken. In 
 mild cases of diabetes no variation from the conditions present in 
 health was observed, but in severe cases the ingestion of 40 grams, 
 although it did not give rise to galactosuria, increased the output 
 of dextrose, and 100 grams caused both an increase in the excretion 
 of dextrose and galactosuria. With the exception of the discovery 
 by Langstein and Steinitz, already referred to, that the sugar noticed 
 by Grosz to be present in the urines of children suffering from gastro- 
 intestinal catarrhs is often a mixture of lactose and galactose, in 
 which the latter may predominate, no observations on spontaneous 
 galactosuria appear to be recorded. 
 
 The presence of galactose in the urine is recognised by the 
 results of the reduction tests, a comparison of the effects on polarised 
 light, before and after treatment with dilute acids, the melting- 
 point of the osazone (194° C), the formation of mucic acid, and the 
 results of the fermentation test. The latter is usually negative for 
 the first six hours with ordinary yeast, but later gas-formation may 
 occur. 
 
 Saccharosuria.- — The assimilation limit for cane-sugar is so 
 high that it rarely passes into the urine unchanged in healthy 
 individuals. Occasionally, however, traces may be met with in the 
 urine of patients suffering from gastro-intestinal disorders, and 
 
388 GLYCOSURIA 
 
 especially in children. Smolenski has reported a case in which the 
 ingestion of cane-sugar was followed by a marked output in the 
 urine, which gave a " Cammidge " reaction. 
 
 The recognition of cane-sugar is chiefly important from its- 
 occasional use by malingerers to simulate diabetes, the sugar being: 
 directly added to the urine. In such cases the urine will have a, 
 high specific gravity, give an atypical reduction when boiled with 
 alkahne solution of copper, ferment but slowly, and give only a few, 
 or no, crystals with phenjdhj^drazin as in Brown's case. If the 
 urine is concentrated, boiled with hydrochloric acid for half an hour, 
 neutrahsed with sodium carbonate, and re-examined, it will give 
 the typical tests for dextrose, and levulose, and will be found to be 
 levo-rotatory whereas it Avas formerly dextro-rotatory. It is stated 
 by Hirschberg that cane-sugar can be differentiated from dextrose, 
 pentoses, &c., by the following procedure. The suspected liquid, if 
 sterile, is placed in an incubator for twenty-four hours, or if in a 
 hurry or contamination is feared is boiled for forty-five minutes^ 
 with an equal quantity of deci-normal sodium hydrate solution. 
 Dilute solutions of sodium hydroxide readily affect dextrose, 
 mannite, maltose, mannose, lactose, levulose, galactose, and invert 
 sugar, complete decomposition occurring under the circumstances 
 described, but saccharose is unaffected, and may be easily detected 
 with the polariscope. 
 
 Pentosuria. — Cases of pentosuria, like those of levulosuria,. 
 can be divided into three classes — (1) an alimentary type, (2) a pure 
 spontaneous form, (3) a mixed type, in Avhich the pentose is asso- 
 ciated with dextrose. There is, however, this important difference 
 that, while the levulose is in all cases the same, the pentose found in 
 pentosuria is a different form of 5-carbon atom sugar according to 
 the circumstances that determine its excretion. 
 
 1. Alimentary Pentosuria. — The assimilation hmit of the healthy 
 organism has been proved experimentally to be much lower for 
 pentoses than it is for the hexoses ; very small doses, 0-25 gram of 
 arabinose, 0"05 gram of xylose, according to Ebstein, I'O gram of 
 rhamnose, according to Cremer, causing a previously pentose-free 
 urine to give a colour reaction for the sugar. The proportion that 
 is excreted unchanged in the urine appears to vary considerably, 
 but it is not unUkely that this depends more upon the condition 
 of the ahmentary tract and the amount that is destroyed there 
 before the sugar is absorbed, than upon individual variations in 
 assimilative power. 
 
 Since many vegetable foodstuffs contain pentosans, and some,. 
 
PENTOSURIA 389 
 
 such as cherries, phims, strawberries, whortle-berries, fruit juices, &c., 
 are comparatively rich in such substances, it is not surprising to 
 find that a number of observers, inckiding Bhimenthal, Barczewski, 
 V. Jaksch, and Johnstone, have detected a pentose in the urine of 
 persons who have consumed considerable quantities of these sub- 
 stances, and that such cases are most common in the summer when 
 iruit, vegetables, and fruit juices are largely taken. Johnstone 
 ]Droduced alimentary pentosuria in sixteen out of eighteen persons 
 by giving them from a haK to a litre and a half of apple jviice, and 
 found that the effect might persist for several days when the larger 
 quantities were consvimed. According to this observer the ad- 
 ministration of morphia increases the amount of pentose excreted. 
 
 The sugar found in the urine in ahmentary pentosuria is optically 
 active, being the 1-arabinose contained in the fruits, &c., so that 
 the urine is dextro-rotatory. The quantity excreted is always 
 small, but sufficient is present to give a reduction test which may 
 prove misleading. It is, therefore, important to bear in mind that 
 a sHght, or doubtful, reduction may be due to an ahmentary pentos- 
 uria, especially w^hen the patient is a vegetarian, in locahties where 
 fruit juices, &c., are extensively used, and in the summer time. 
 
 It is often stated that the urines of herbivorous animals usually 
 give a pentose reaction, and that this is due to the pentosans con- 
 tained in their food, since hay, for example, contains over 21 per 
 cent. According to experiments made by Cominotti, the greater 
 part of the pentose taken bj^ a fasting horse is utilised by the 
 organism and only a comparative!}^ small proportion passes in the 
 nrine. Small quantities reappear on prolonging the inanition. 
 
 Both with animal and human urines a diagnosis of pentosuria 
 should not be made, as is sometimes done, merely on the results 
 of the colour tests for furfurol, since these may also be given by 
 glucuronic acid compounds. Ebstein examined twenty-tw^o ap- 
 parenth" normal human urines with the phloroglucin test and ob- 
 tained a positive reaction with fourteen, Cremer states that almost 
 every urine gives a more or less marked reaction, and Funaro came 
 to the same conclusion. Similar results are also obtained with 
 the orcin test, unless it is very carefully carried out. Statements 
 that pentoses appear in the urine as a result of the administra- 
 tion of drugs, foods, &c., must therefore be accepted with reserve, 
 unless the proof rests on a more sure foundation than a mere 
 appUcation of these tests. 
 
 2. Spontaneous, or essential, pentosuria is quite a different con- 
 dition to the ahmentary form, having no relation to food, and 
 persisting when pentose-containing substances are excluded from 
 
390 GLYCOSURIA 
 
 the diet. Moreover, the sugar excreted in every case so far, with the 
 possible exceptions of those described bj" Luzzato, and Elliott and 
 Raper, has apparently been optically inactive arabinose, a substance 
 that is not met with otherwise in the animal or vegetable kingdoms. 
 The first case of essential pentosuria was described by Salkowski 
 and Jastromtz in 1892. The patient was a young man, a victim 
 of the morphia habit and suffering from neurasthenia. When he 
 first came under observation his urine contained traces of dextrose, 
 but after the morphia was discontinued this disappeared. It was- 
 then found that it gave the reduction tests for sugar, but did not 
 ferment with yeast, was optically inactive, and jdelded with phenjd- 
 hydrazin an osazone having a melting-point of 159° C. The melting- 
 point of the osazone suggested that the sugar was a pentose, and 
 further investigation confirmed this conjecture. Since 1892 other 
 cases of essential pentosuria have been described by Blumenthal, 
 Reale, Colombini, Bial, Meyer, Romme, Brat, Luzzato, d'Amato,. 
 Bendix, Ivlercher, Adler, Tintemann, Schiiler, Johnstone, v. Jaksch, 
 Kraft, Blum, Janeway, Kaplan, Rosenfeld, Cassiver and Bamberger, 
 Chobola, Vas, JoUes, Wall, Elhott and Raj)er, so that some thirty- 
 eight or fortj^ are now on record. Several of these, including the 
 cases reported by Reale, Colombini, d'Amato, and Kaplan, are 
 probably not to be regarded as cases of true essential pentosuria, 
 but were most likelj" of the alimentary type. 
 
 Chronic pentosuria is a rare condition. Jolles, in an examina- 
 tion of 3000 normal and pathological urines in the course of twO' 
 3^ears, only met with four undoubted cases, and in over 4000 urinary 
 analyses that I have made dm-ing the past seven years I have not 
 met with a single example, although pentoses have been systemati- 
 cally tested for. So far the great majority of the recorded cases have 
 been met with in Germany, especially at watering-places to which 
 patients with the milder forms of cUabetes are accustomed to resort,, 
 and it is noteworthy that the American cases have been of German 
 or Russian descent. A striking proportion of the cases so far met 
 with have been Jews. 
 
 The amount of pentose present in the lurine has always been 
 small, very rarely exceeding 1 per cent., and generaU}^ being under 
 that amount. In many cases the total daily excretion is not stated, 
 but, in those where the quantity of urine passed is given, it works 
 out at under 10 grams a day, and usually considerably less. The 
 quantity appears, however, to vary from time to time. In a case 
 investigated by Blumenthal, for instance, 7 grams were passed in 
 the twenty-fotu' hours, but two years later an anatyses by the same 
 method (Knapp) showed only about 3 to 4 grams. 
 
PENTOSURIA 391 
 
 The urine is acid in reaction and the specific gravity varies from 
 r025 to 1-035. Fehling's solution is reduced as it is by urines 
 containing J per cent, or so of dextrose — that is to say, only on 
 boiling the mixture. The sudden, delayed reduction, described 
 by some observers as characteristic of pentose-containing urines, 
 is attributed b}^ Bial to the specimens having been kept for some 
 time by means of a preservative. Pentoses are not fermented by 
 yeast, so that the reducing power of the urine is not impaired by 
 being incubated vs^ith yeast for twenty-four hours. This affords a 
 ready means of differentiating the sugar when it occurs in a urine 
 along with dextrose. It forms with phenylhydrazin a crystalline 
 osazone which is soluble in hot water, and, after re-crystallising, 
 melts at 156° to 160° C. The osazone yields about 17 per cent, of 
 nitrogen. The urine in essential pentosuria is optically inactive, un- 
 less dextrose is also present, and gives the usual colour reactions for 
 pentoses. Of these Bial's modification of the orcin test is the most- 
 useful clinically. According to Klercher, there is a certain paral- 
 lelism between the amount of pentose excreted and the total 
 nitrogen content of the urine, but the latter is not notably in- 
 creased, nor is the output of purin bodies, or phos]3horus, in any 
 way abnormal. 
 
 The ordinary method of estimating sugars in the urine with 
 Fehling's solution cannot be employed satisfactorily for the 
 pentoses, since the cuprous hydrate does not separate out well. 
 Even when Knapp's, or AUihn's, method is used, or the phloroglucin. 
 ]orecipitate is weighed, the results, according to Neuberg, are too 
 low, since much of the pentose is in combination with urea and 
 does not reduce until the ureide is broken down by heating with an 
 acid. Neuberg mentions 30 to 36 grams as the total amount of 
 pentose that may be excreted in a day if the sugar combined with 
 urea is taken into account. 
 
 Symptoms. — Chronic pentosuria does not give rise to any par- 
 ticular train of symptoms. Poljmria, thirst, excessive hunger,, 
 wasting, and the other characteristic symptoms of chronic clextros- 
 uria have not been present in any of the reported cases. Only in 
 one, that of Colombini, was there any affection that is usually 
 associated with diabetes. This patient was an Italian, aged fifty, 
 and appeared to be suffering from xanthoma diabeticorum. When 
 he was treated with arsenic, and put on milk and meat in place of 
 his previous vegetarian diet, the skin condition improved and the 
 pentose disappeared from the urine. As the pentosuria was a tran- 
 sitory condition occurring in a vegetarian, and the optical characters 
 of the urine were apparently not investigated, it is probable that 
 
392 GLYCOSURIA 
 
 the case was really of the alimentary type, and it is not certain 
 that there was any connection between the skin condition and the 
 j)resence of the sugar in the virine. 
 
 The etiology of chronic pentosuria is not clear. The fact that 
 Salkowski and Jastrowitz's patient, and also Reale's, were morphia 
 habitues, and that one of Bial's cases had the cocaine habit, has 
 suggested that the condition might be dependent upon the abuse 
 of drugs, but this has not been established in other cases. The 
 presence of neurasthenic symptoms and nem-algic pains, in some 
 oases (Salkow^ski and Jastrowitz, Cassiver and Bamberger), might 
 point to a nervous origin, but this too has not been proved in other 
 cases. It has been suggested that chronic pentosuria might be 
 dependent upon some lesion or abnormahty of the pancreas, but 
 there is no evidence whatever to support this \dew. In the only 
 case of pentosuria so far examined jDost-mortem the pancreas was 
 unfortunately not closely investigated, but no gross pathological 
 change was observed in it, or in any other organ, to account for 
 the sugar in the urine (Blumenthal). In some instances the sugar 
 has been found in the urine of apparently healthy individuals in the 
 course of routine examination for life insurance, or for some other 
 purpose. Some pentosurics have been members of diabetic famihes. 
 Klercher's patients were brothers, and their father and another 
 brother had died of diabetes. Schiller's patient had a brother and 
 two sisters who were diabetic, and in one of the cases described by 
 Bosenfeld the patient's father and brother had died of diabetes. 
 This patient developed pentosuria after being in a railway accident. 
 Such cases would seem to suggest that there is a relationship be- 
 tween pentosuria and diabetes, but there can be no doubt that. 
 from a metabohc standpoint, pentosuria and glycosuria are quite 
 distinct. It has been shown that the tolerance of pentosurics for 
 dextrose is not in any way diminished, and that glucose only 
 appears in the urine when taken in doses sufficient to overtax the 
 assimilative powers of a normal individual. 
 
 The most striking feature about chronic pentosuria is its tendency 
 to occur in several collaterals of a family. Brat's cases were a 
 sixtj^-two-year-old woman and her fifty-year-old brother. The 
 former had been treated for eight years as a diabetic before the 
 true state of affairs was discovered, the latter was apparently 
 quite healthy. Two of Blumenthal's cases were sisters. Of three 
 cases described by Bial two were sisters, and the third their 
 brother. Klercher's patients, who were brothers, have been 
 mientioned. Janeway's cases were also brothers. There is, as 
 yet, no record of an instance where chronic pentosuria has been 
 
PENTOSURIA 393 
 
 transmitted from parent to child, and nothing is known of con- 
 sanguinity of the parents of subjects of this condition, but the 
 tendency at the present time is to regard it as a congenital 
 abnormahty of the chemistry of the body, or, as Garrod terms 
 it, " an inborn error of metabolism," like albinism, alkaptonuria, 
 or cystinuria, and analogous to structural abnormalities such as 
 Ijolydactjdism. 
 
 The origin of the sugar found in the urine in essential pentosuria 
 has been the subject of much debate and many experiments. Its 
 most remarkable character is its optical inactivity, which marks 
 it out as a striking exception to the general rule that the animal 
 organism is built of optically active substance. In one case 
 Neuberg succeeded in isolating the pentose from the di-phenyl- 
 osazone prepared from a large volume of the urine, and was able 
 to prove that it was racemic arabinose. The optical inactivity of 
 the sugar proves that it is not derived from vegetable foodstuffs, 
 for the pentose they contain is dextro-rotatory 1- arabinose; moreover, 
 the excretion of pentose continues when all pentosans are excluded 
 from the diet. It is not likely to be derived from the pentose con- 
 tained in nucleo-proteins of the food, which is 1- xylose, for Bial and 
 Blumenthal found no increases in the excretion of pentose after 
 feeding with 500 grams of calf's thymus. Nor can it come from the 
 iSugar of the nucleo-protein of the tissues, which is also 1-xylose, 
 since, according to Griind, the total amount in the human body is 
 about 10 grams, which would be less than one day's output in some 
 ■cases of pentosuria ; further, the uric acid and phosphate excretion 
 in cases of pentosuria gives no evidence of abnormal destruction 
 of nucleo-proteins. On chemical grounds, too, the origin of r-ara- 
 binose from 1-xylose is not probable. The carbohydrate content 
 of the diet appears to bear no relation to the amount of pentose 
 appearing in the urine, and its total exclusion in no way influences 
 the output. Dextrose and levulose are both made use of by 
 pentosurics as completely as by normal individuals, and even 
 pentoses, when given by the mouth, are destroj^ed in the same way 
 as they are by healthy persons. Bial and Blumenthal found that 
 when 50 grams of 1- arabinose were given to a patient with pentos- 
 uria only 6 grams reappeared in the urine, and Tintemann showed 
 that xylose behaved as with healthy persons, about 8 grams re- 
 appearing in the urine after 20 grams had been taken by the mouth. 
 
 There appears to be some evidence thatr the excretion of 
 r-arabinose is related to the proteins of the food and the activity 
 of the metabolic processes in the body. Klercher found that the 
 output varied much during the day, and that there was a certain 
 
394 GLYCOSURIA 
 
 parallelism between its excretion and the total nitrogen content' 
 of the urine. In one of his patients the lowest figure was obtained 
 after fasting, and on that day there was an abnormally low nitrogen 
 excretion. Klercher and Janeway also observed a diminished 
 excretion on a purin-free and milk diet. Blumenthal and Meyer 
 state that meat increases the nervous disturbance, and that milk 
 is the most advantageous diet. The evidence that the excretion 
 of the pentose is influenced by the proteins of the food is not, 
 however, conclusive. Bial and Blumenthal found that the blood 
 of a patient with pentosuria gave the orcin reaction, and concluded 
 that a pentose was present, thus tending to exclude the renal 
 origin of the sugar. Blumenthal states that pentoses were absent 
 from the hydrocele fluid of one of his patients. Injection of 
 phlorhidzin gives rise only to dextrosuria in pentosurics as in nor- 
 mal individuals, while the administration of chloral and menthol^ 
 although it causes an increased output of glucuronic acid in the 
 usual way, does not influence the excretion of pentose. It may be 
 pointed out that the differentiation of glucuronic acid from pentoses 
 must be carefully made, for their chemical reactions are in many 
 respects so closely aUke that the one may be easily mistaken for 
 the other ; in fact, it is not improbable that in some cases which 
 have been described as examples of pentosuria (Caporelli, Colombini) 
 the reducing substance was really glucuronic acid. Taking all 
 the known facts into consideration, it would seem most probable 
 that the sugar found in the urine in cases of essential pentosuria 
 is derived from some substance formed within the organism, and 
 that this parent substance is not dextrose. Neuberg has suggested 
 that the most hkely mother-substance is d-galactose. Theoretically 
 such a conversion is possible, and it is known that galactose can be 
 formed within the body, for it has been shown by Theirfelder to 
 be the sugar yielded by cerebrin, and, with dextrose, forms the 
 lactose of milk. At present there is no conclusive evidence of 
 the correctness of Neuberg' s hypothesis. Tintemann observed a 
 slight increase in the amount of pentose in the urine after giving^ 
 50 grams of galactose on an empty stomach. Klercher noticed an 
 increase in the hourly output for six or seven hours after 100 grams- 
 of lactose, but the total excretion was not excessive. Blumenthal 
 and Bial found no conspicuous increase in the urinary pentose 
 after 100 grams of galactose by the mouth. But further observa- 
 tion and experiment on this point is necessary. 
 
 An optically active arabinose is stated to have been found in the 
 urine in three cases of chronic pentosuria, either alone, or with the 
 inactive sugar. In Luzatto's case 1- arabinose was believed to be 
 
PENTOSURIA 395 
 
 present, since the urine was dextro-rotatory and yielding an osazone 
 with characteristic activities and melting-point, but the evidence 
 is not very conclusive (Magnus-Levy). Blumenthal and Bial met 
 with 1-arabinose and the inactive variety together in one case, 
 but here again the proof is doubtful, and the possibility of an 
 alimentary origin for the optically active sugar was not excluded. 
 Basing his conclusion on the spectroscopical characters of the orcin 
 test. Brat concluded that a methyl- pentose (rhamnose) was present 
 in his case of pentosuria. 
 
 Rihose, a reduction product of the lactone of ribonic acid which 
 the investigations of Levene and Jacobs have shown to be contained 
 in some nucleic acids, was stated by Elhott and Raper to be the 
 pentose present in the urine of a case they examined. 
 
 Prognosis. — Essential pentosuria has been recognised for such 
 a comparatively short time, and in such a small number of cases, 
 that it cannot be definitely stated whether it does, or does not, 
 ultimately shorten hfe, but it would appear that the presence of a 
 sugar with five carbon atoms in its molecule in the urine has not 
 the serious significance that attaches to the presence of dextrose. 
 The condition may apparently persist unchanged for years without 
 there being any increased habihty to infection, or the occurrence 
 of secondary cUsturbances of metabohsm, such as result from per- 
 sistent glucosuria. Blumenthal has suggested that the excess of 
 circulating sugar may give rise to arterio-sclerosis, but there is no 
 evidence in support of this. Some pentosurics have been members 
 of diabetic famihes, small quantities of dextrose have been met 
 with in the urine of a few cases, and in some cases of diabetes a 
 pentose may be found, but all the available evidence is against 
 the view that essential pentosruia increases the habihty to chronic 
 glucosuria. The prognosis therefore is good. 
 
 Treatment. — No form of treatment has been found to materially 
 influence the condition, if we except Colombini's doubtful case. 
 An anti-diabetic diet is unnecessary, and may do more harm than 
 good, for not only is it irksome to the patient, but the inchcations 
 are that a limitation of the proteins, rather than of the carbohydrates, 
 is desirable. Morphine, and similar drugs, are contra-indicated, 
 since chronic pentosuria has been associated with a drug habit in 
 at least three cases and the administration of morphia has been 
 found to increase the tendency to alimentary pentosuria. The most 
 important reason why essential pentosuria should be watched for, and 
 recognised, is that it needs no treatment ; the patient can thus be 
 saved from the inconvenience, mental worry, and possibly financial 
 loss that an incorrect diagnosis of diabetes melhtus would entail. 
 
396 GLYCOSURIA 
 
 3. Mixed Pentosuria and Dextrosuria. — A certain number of 
 patients with undoubted essential pentosuria have excreted 
 dextrose in their urine along Avith the pentose. The original 
 patient of Salkowski and Jastrowitz passed a small quantity of 
 glucose, but, as it disappeared when the morphia he took was 
 stopped, it is possible that the morphia habit may have been the 
 exciting cause of the temporary glucosuria. One of Blumenthal's 
 patients w^as glucosuric, and so also was one of Klercher's. The 
 amount of dextrose in these cases was small (TS per cent, to 1"0 per 
 cent.). There is. however, another class in which a large amount 
 of glucose is associated with a small quantity of a pentose. The 
 frequency of this association is variously estimated by different 
 authors. Kiilz and Vogel investigated eighty diabetic urines and 
 only failed to obtain indications of the presence of a pentose in 
 foiir. With twelve a feeble, or doubtful, phloroglucin reaction 
 was obtained, but with sixty-four the colour reaction and spectro- 
 scopic appearances were distinct and characteristic. Such evidence 
 by itself would be of little value, but from several cases of severe 
 diabetes they succeeded in isolating an osazone that was soluble 
 in hot water and had the melting-point and nitrogen content of a 
 pentosazone. As only small yields of this product, at most 0-1 
 gram per htre of urine, were secured the exact nature of the pentose 
 was not determiiied. Bendix and other observers have failed to 
 detect a pentose in the diabetic urines they examined, although 
 they used the same method as Kiilz and Vogel, so that it would 
 appear that the presence of a 5-carbon atom sugar in diabetes is 
 not as common as the experience of the latter would suggest. 
 Kiilz and Vogel also found a pentose in the urine of dogs rendered 
 diabetic by removal of the pancreas, and it was noticed that, as 
 in essential pentosuria in human beings, the excretion of pentose 
 in the lurine by these animals was not dependent upon the diet. 
 The exact nature of the pentose was not determined. In only one 
 recorded case, that of d'Amato, would it appear that the presence 
 of a pentose along with dextrose in the urine was definitely asso- 
 ciated with chsease of the pancreas, but it is not unhkely that the 
 pentose was of aUmentarj' origin. Mj^ own observations on the 
 urine with the modified pancreatic reaction, would suggest that in 
 some 75 per cent, of cases of diabetes a non-fermentable reducing 
 substance, probably a pentose in some, but in others possibly 
 glucuronic acid, is present in the urine after it has been heated with 
 hydrochloric acid. 
 
 Glucuronic Acid. — A small amount of glucuronic acid is a normal 
 
GLUCURONIC ACID 397 
 
 constituent of the urine. Mayer and Neuberg state that it is 
 jDresent to the extent of 0'004 grams per 100 c.c, and Tollens and 
 Stern found an average daily output of 0-35 to 0"37 grams. It does, 
 not occur in the free state, however, but as conjugate glucuronates, 
 and it is to the presence of these that the feeble levo-rotatory power 
 and, in part, the slight reducing effects of normal urine are due 
 (Lavesson). The phloroglucin and orcin reactions given by normal 
 urines, after they have been boiled for a minute or so with 1 per 
 cent, sulphuric acid, are also referable to the presence of compound 
 glucuronates. 
 
 Physiologically glucuronic acid appears to be part of the pro- 
 tective mechanism of the body by which the organism defends 
 itself against harmful substances, formed in the tissues or intro- 
 duced from without. Poisons of various kinds are usually rendered 
 innocuous in one or more of four ways — (1) by being rapidly 
 eliminated, (2) by being dej^osited and fixed in various organs or 
 tissues, notably in the liver, (3) by being chemically altered through 
 oxidation, reduction, hydrolysis, or neutralisation, (4) by being 
 combined with substances formed, or contained, in the tissues, so 
 that compounds of a harmless, or less toxic character than the 
 original poison, result. The chemical defences employed against 
 inorganic poisons are mainly the simple processes of oxidation, 
 reduction, &c., but with the more complex organic poisons pro- 
 tective combinations are in addition very frequently formed. 
 The chief protective substances employed for this purpose are 
 alkalies, proteins, hydrogen sulphide, glycocoll, urea, bile salts, 
 acetic acid, sulphuric acid, and glucuronic acid. Although most 
 of these have special affinities for various chemical substances, 
 depending on their composition, their action is not strictly specific, 
 like the immune substances against bacteria and their product, 
 and some are capable of replacing others, so the poison may be 
 excreted partly in combination with one, and partly with another. 
 Moreover, since most at least of the protective substances are not 
 special bodies formed for the purpose of dealing with a poison 
 appearing in the circulation, but are normal products of metabolism 
 diverted to this end, when the available amount of any particular 
 one is exhausted the residue of the poison must either unite with 
 a substitute or go uncombined. We have already considered an 
 example of this protective mechanism and seen how one neutrahsing 
 substance is replaced, and augmented, by another, when deaHng 
 with the question of diabetic acidosis. In this case the acids 
 formed as a result of the abnormal metabohc changes that are met 
 with in severe cases of diabetes are first neutralised b}^ the fixed 
 
398 GLYCOSUHIA 
 
 alkalies, but later, when there is a danger of the alkalinity of the 
 blood being seriously reduced, they are combined with ammonia, 
 ■derived from the nitrogen that normally goes to form urea. 
 
 Two of the most important protective substances are sulphuric 
 
 and glucuronic acids, both of which have the power of combining 
 
 wdth a number of toxic agents to form harmless, or comparatively 
 
 harmless, compounds that are readily soluble and can be easily 
 
 •eliminated in the urine. They are most commonly met with in 
 
 combination with various aromatic bodies formed in the intestine 
 
 as a result of the cleavage of proteins. According to Herter, these 
 
 are capable of setting up marked derangements of function, and 
 
 probably even histological changes, when brought in contact with 
 
 the elements of the nervous system in an unchanged condition, 
 
 and the experiments of Woolley and Newburgh suggest that some 
 
 of them, at least, may induce hyperactivity of the chromaffin 
 
 tissue of the bodj" with resulting pathological changes. As far 
 
 back as 1877, it was shown by Baumann and Herter that when one 
 
 •of these substances, phenol, is given to animals it is excreted in 
 
 the urine as a potassium salt of the sulphuric acid derivative, and 
 
 later Magnus-Levy found that in carbolic acid poisoning, while 
 
 ,some of the phenol appears in the urine in combination with sulphuric 
 
 acid, a great part is ehminated as a glucuronic acid compound. It 
 
 has also been shown that indol and skatol are similarly rendered 
 
 innocuous by being converted into sulphuric and glucuronic acid 
 
 compounds, but in this case after a preliminary oxidation into 
 
 indoxyl and skatoxyl. It would thus seem that glucmronic acid 
 
 and sulphuric acid have similar functions. Many observers have 
 
 held that sulphuric acid is the first line of defence, and that it is 
 
 only when there is not sufficient of this to combine with all the 
 
 poison that the excess is excreted in combination with glucuronic 
 
 acid. Salkowski has pointed out, however, that the latter may begin 
 
 to be formed before the sulphuric acid is exhausted, and ToUens 
 
 has shown that the lower derivatives of protein decomposition do 
 
 not unite indifferently with sulphuric and glucuronic acid, but that 
 
 indol given by the mouth is excreted mainly in combination with 
 
 sulphuric, and j^henol with glucuronic, acid. In health some tenth 
 
 part of the total sulphuric acid of the urine is in combination with 
 
 aromatic substances, as ethereal sulphates, and according to 
 
 Tollens the excretion of glucuronic acid is, as a rule, about double 
 
 this 0-35 grams of glucuronic to about 0-18 grams of ethereal 
 
 .sulphates. The former is chiefly in combination with phenol, but 
 
 indoxyl and skatoxyl glucuronates are also present in smaller 
 
 amounts. 
 
GLUCURONIC ACID 399 
 
 The most striking demonstration of the way in which glucuronic 
 .acid is used by the organism to protect itself from the action of 
 deleterious substances, is furnished by the abundant excretion of 
 compound glucuronates that follows the administration of certain 
 drugs, and other substances that contain an hydroxyl group, and 
 are only oxicUsed with difficulty in the tissues. These include 
 chloral, camphor, bromol, naphthol, aniUne, benzine, turpentine, 
 phenol, salicylates, borneol, resorcinol, menthol, toluol, thymol, 
 antipyrin, antifebrin, and numerous other alcohols and ketones. 
 It was formerly thought that the compounds formed under these 
 ■circumstances were alcoholates, but they are now considered to be 
 of a glucosidal nature. Their relation to the glucosides is shown by 
 the action of appropriate glucoside-spHtting ferments on them; 
 thus phenol- glucuronic acid is attacked and slowly broken down 
 by emuLsin into phenol and gkicuronic acid, while other compound 
 glucuronates are gradually decomposed by invertin. Like the 
 glucosides the conjugate glucuronic acids are hydrolised by mineral 
 acids, yielding glucuronic acid and the particular alcohol from which 
 they are derived, but some more readily than others. All con- 
 jugate glucuronic acids, however, do not exhibit the characters of 
 glucosidal compounds, for some, such as urochloraHc acid (tri- 
 chlorethyl glucuronic acid) and paramidophenyl-gluciironic acid, 
 for example, reduce alkaline solutions of copper as readily as 
 dextrose, a reaction which is only obtained with most compound 
 glucuronates after the acid has been set free by hydrolysis. This 
 property appears to be due to the existence of a free aldehyde 
 group in the compound. In consequence of the reducing powers 
 possessed by some urines containing compound glucuronates, either 
 immediately, or after prolonged boihng, it was formerly thought 
 that the administration of certain drugs gives rise to temporary 
 glycosuria, and, although this appears to be true of a few, the 
 reduction obtained in most instances is dependent ujjon glucuronic 
 acid. 
 
 The compound glucuronates formed naturally with the lower 
 derivatives of protein decomposition are, like most of those re- 
 sulting from the administration of drugs, of an ethereal or glucosidal 
 nature, but show marked differences in the readiness with which 
 they are split up. Indol-glucuronic acid, for instance, reduces 
 alkaline solutions of copper on boiling for some time, but phenol- 
 glucuronic acid is not readily decomposed and therefore does not 
 reduce even after prolonged heating. The presence of an excess 
 of the former in the urine may, therefore, give rise to a false idea 
 that traces of sugar are present, unless care is taken and the results 
 
400 GLYCOSURIA 
 
 of the reduction tests are confirmed in other ways. With the 
 latter such a mistake is not hkely to arise. It will be remembered, 
 however, that indol is mainly excreted in combination with sul- 
 phuric acid, and it is probably only when more is being formed 
 than can be bound by the available sulphuric acid that a sufficient- 
 quantity to give rise to difficulty is likely to be excreted in the form 
 of a glucuronate. 
 
 Pathological Excretion. — Very Httle is known concerning the 
 excretion of glucuronic acid in disease. An increased output has 
 been observed in various pathological conditions, but reliable 
 observations are not numerous, and the significance of an increased 
 excretion is not yet agreed upon in all cases. A number of sub- 
 stances derived from the aromatic radicals of the protein molecule 
 have been found in the intestinal contents, and some of these, as 
 we have seen, make their appearance in the urine in combination 
 with glucuronic acid even in health. Thus phenol, which is probably 
 derived from tyrosin, and the closely related paracresol, para- 
 ox}^3henyl- acetic acid, and para-oxyphenyl-proj)ionic acid are 
 apparently excreted chiefly as conjugate glucuronates. Indol- 
 propionic acid, indol- acetic acid, and the better known skatol and 
 indol are derivatives of tryptophan and are partly got rid of as 
 glucuronates, although mainly as ethereal sulphates. When for 
 any reason the putrefactive changes in the intestine are increased, 
 and these substances are consequently formed in excess, the output 
 of conjugate glucm-onates is correspondingly augmented, while at 
 the same time the proportion of ethereal sulphates is increased. 
 The amount of sulphuric acid appearing in the urine in such organic 
 combination has long been considered as an index of the amount 
 of intestinal putrefaction, but very little attention has been paid 
 to the increase in conjugate glucuronic acid. Accorchng to Tollens,. 
 the ethereal sulphates and the compound glucuronates rise and fall 
 together usually, but not always, proportionally to the amount of 
 intestinal putrefaction. In peritonitis, enteritis, and other con- 
 ditions promoting abnormal putrefactive changes in the intestinal 
 contents, there is also a marked increase in both. An increased 
 excretion of glucuronic acid has also been observed in association 
 with putrefactive changes in other situations — for example, with 
 gangrene, sloughing cancer, putrid jalacenta, or decomposing, 
 exudates. 
 
 Although it does not appear probable that indol can be derived 
 from trj^tophan hberated during intracellular protein metabohsm,. 
 it is not unhkely that para-ox j^henyl- acetic acid, para-oxyphenyl- 
 propionic acid, and other derivatives of tyrosin, &c., may be 
 
GLUCURONIC ACID 401 
 
 formed in this way, and that this may account for the compound 
 glucuronates found in the urines of patients with febrile diseases, 
 respiratory difficulties, and impaired metabohsm. I have met with 
 an abnormal output of glucuronic acid in scarlet fever, smallpox, 
 and measles, also in pneumonia, chronic bronchitis, and emphysema. 
 In the course of a series of investigations with the phenylhydrazin 
 test that I carried out some years ago, I found that out of a hundred 
 urines from apparently healthy people, six gave a reaction with 
 phenylhydrazin alone, and thirteen after boiling the urine with 
 hydrochloric acid. On dividing the series into a set of forty-three 
 derived from persons Hving in the country with an abundance of 
 fresh air and exercise, and a second set of fifty-seven obtained from 
 city dwellers who live under less favourable hygienic conditions, 
 it proved that none of the former gave a reaction until the urine 
 had been boiled with the acid, but that six of the second series gave 
 a positive result before treatment, and nine after. Further in- 
 vestigation showed that in every case the reaction was due to 
 glucuronic acid. 
 
 Early in my investigations into the condition of the urine in 
 diseases of the pancreas I was struck by the marked increase in the 
 excretion of glucuronic acid that accompanied inflammatory 
 affections of the gland, and my subsequent experience has shown 
 that it is a very constant phenomenon. At first sight one might 
 be inclined to ascribe this to an excessive absorption of the 
 products of putrefaction from the intestine, owing to the pan- 
 creatic disease interfering with the normal digestion of proteins, 
 but the output of glucuronic acid does not appear to bear any 
 relation to the amount of ethereal sulphates, or indican, 
 in the urine, and is generally greater in the early stages of 
 catarrhal pancreatitis than it is when there is advanced cirrhosis 
 or mahgnant disease, where protein digestion is more seriously 
 interfered with. 
 
 Many diabetic urines contain compound glucuronates, as 
 much as 13-6 grams of the anhydride having been found by Baum- 
 garten in one case. On administering drugs such as chloral, 
 camphor, naphthol, &c., to patients with severe diabetes, as well as 
 to dogs after extirpation of the pancreas, Weintraud found that 
 these substances are eliminated in combination with glucuronic 
 acid almost as abundantly as by normal individuals. In cases of 
 chronic glucosuria where the elimination of sugar has ceased owing 
 to careful dieting, glucuronic acid compounds are still often present 
 in the urine in abnormal quantities. 
 
 Origin. — Chemically gluciu-onic acid is closely related to glucose, 
 
 2c 
 
402 GLYCOSUEIA 
 
 two atoms of hydrogen in the CHgOH group of the latter being 
 replaced by one of oxygen in the former : — 
 
 CHgOH - (CH.0H)4 - COH COOH - (CH.OH)^ - COH 
 
 Dextrose. Glucvironic acid. 
 
 Since this process of oxidation involves no distm-bance in the 
 linkage of the carbon atoms in the sugar molecule, it would seem 
 Hkely that it can be readily effected, and that the glucuronic acid 
 that is excreted in the urine in the shape of compound glucuronates 
 represents an early oxidation product of sugar, that is diverted 
 from its further natural degradation to combine with toxic sub- 
 stances for the protection of the organism. The experiments of 
 Jolles appear to support this view, for he showed that glucuronic 
 acid is one of the products of the oxidation of sugar in weakly 
 alkahne solutions. 
 
 Another hjrpothesis advanced by Sundvik, and later by Fischer 
 and Pilot, has, however, been very generally accepted. According 
 to this, the toxic substance combines directly with dextrose to 
 form an ether-hke compound. One end of the chain is thus 
 shielded from attack by the pairing substance, but the other is 
 open to chemical change. Subsequently it undergoes oxidation, 
 and glucuronic acid results, the paired, or conjugate, glucuronates 
 formed being then excreted in the urine. The faculty of removing 
 injurious substances by combining them with glucose to form in- 
 different compounds is a striking feature of plant physiology, but 
 whereas the glucosides of plants undergo no further change, and 
 exist as such in the plant cells {e.g. amygdahn in almonds, sahcin 
 in willows, sinigrin in the cruciferse, phloridzin in cherry trees, &c.), 
 the more active oxidising tissues of the animal appear to change 
 the glucose radicle into glucuronic acid : — 
 
 E.O.CH(CH.OH)2.CH.CH(OH)CH2(OH) R.0.CH(CH.0H)2CH.CH{0H)C00H 
 
 Glucoside. Paired glucuronic acid. 
 
 Whichever way glucuronic acid is formed, there can be little 
 doubt that the parent substance is dextrose, and the question then 
 arises, whence is the dextrose derived ? The first somce that 
 suggests itself is the carbohydrates of the food. Mayer has shown, 
 however, that the administration of chloral, camphor, &c., to 
 starving animals causes as great, or nearly as great, an excretion 
 of compound glucuronates as when corresjionding doses are given 
 to well-fed animals. Hildebrand also found that bases of the type 
 
GLUCURONIC ACID 403 
 
 of thjrmotinpiperidin, which are excreted in conjugation with 
 glucuronic acid, are very little less poisonous when dextrose, cane- 
 sugar, or maltose has been given beforehand than when they are 
 taken by unprepared animals. It would seem, therefore, that 
 glucuronic acid is not derived from the carbohydrate of the food, 
 but from the glycogen contained in the liver and other reservoirs, 
 or from the carbohydrate moiety of the proteins. In favour of the 
 former origin is the observation of Embden that the passage of 
 blood containing phenol through the hver of a dog, gives rise to 
 the formation of phenol-glucuronic acid, and the fact that the 
 administration of glucuronic acid with the food has been foimd to 
 be followed by a deposition of glycogen. As the results of their 
 experiments, Mendel and Jackson came to the conclusion that 
 glucuronic acid is formed solely in the intermediary metaboHsm 
 of proteins. These investigators gave camphor to fasting dogs 
 for several days, and noted the output of glucuronic acid. They 
 then gave large doses of dextrose, and found that protein metabohsm 
 fell, and with it the excretion of glucuronic acid. On adding meat 
 to the diet a rise in the excretion of campho-glucuronic acid, corre- 
 sponding to the amoimt of protein food ingested, occurred. 
 
 Most observers are agreed that the appearance of compound 
 glucuronates in the urine is generally an expression of the power 
 of the organism to deal with toxic substances, but, while many 
 consider that this is the invariable explanation, there are others 
 who maintain that it does not hold good in every case, and that in 
 some instances an increased excretion may result from a per- 
 version of the internal metaboHsm of the body. Mayer has put 
 forward the view that the oxidative capacity of the body for 
 dextrose may, under certain circumstances, be so far diminished 
 that in part the process stops short at the formation of glucuronic 
 acid, and occasionally may be insufficient to carry a portion even 
 to this stage. The larger proportion of the circulating glucuronic 
 acid then present combines with protein derivatives which would 
 otherwise unite with sulphuric acid, and, as a result, there is a 
 diminished excretion of conjugate sulphates. He accounts in this 
 way for the increased output of compound glucuronates met with 
 in febrile diseases and in conditions associated with respiratory 
 difficulties, and also the occasional appearance of sugar in the urine 
 in such cases. The excretion of sugar in febrile diseases is, however, 
 a rare phenomenon, and it seems probable that impHcation of the 
 pancreas, rather than a primary defect in oxidation, is responsible 
 for the alimentary and spontaneous glycosuria that is occasionally 
 met with. The glycosuria associated with respiratory difficulties 
 
404 " GLYCOSURIA 
 
 is of the asphyxial type, and this, as we have seen, is usually 
 ascribed to stimulation of the nerve centres by the carbon dioxide 
 present in the intensely venous blood. According to the protective 
 mechanism view, the increased excretion of glucuronates in these 
 cases is to be explained by supposing that the abnormal intermediate 
 products of metabohsm that are formed as a result of the defective 
 tissue changes, unite with dextrose in the ordinary way, and sub- 
 sequently undergo oxidation. The chief experimental evidence 
 brought forward by Mayer in support of his hypothesis is furnished 
 by observations on patients with diabetes and ahmentary glycosuria. 
 He was able to detect glucuronic acid in the urines of eleven out of 
 thirty cases of diabetes, and found that when 100 to 200 grams of 
 glucose were given to persons with ahmentary glycosm-ia, fourteen 
 excreted conjugated glucuronic acid along with the sugar, and six 
 compound glucuronates alone. Blumenthal points out that it is 
 difficult to understand how small amounts of dextrose and of 
 glucuronic acid can be excreted together solely because the oxidis- 
 ing power of the organism for these substances is diminished, and 
 thinks that it is probable that, since the conjugated acid is never 
 found in the urine, even after it has been hypo dermic ally injected, 
 it is most hkely that in such cases the formation of the substance 
 with which it combines precedes the formation of the acid. The 
 observations of Achard and Weil, and of Strauss, tend to confirm 
 Blumenthal's explanation, for the former found indoxyluria in a 
 case of ahmentarj^ glycosuria that gave similar results, which they 
 investigated, and the latter points out that indoxyluria is very 
 common in diabetes, and that the increased excretion of glucuronic 
 acid stands in very close relation to it. 
 
 Mayer explains the occurrence of oxaluria in diabetes melhtus 
 on the assumption that more glucuronic acid is formed than can 
 combine with the available quantities of protein decomposition 
 products, and that this is oxidised to oxaHc acid, which combines 
 with calcium, and is so excreted. He points out that oxaluria is 
 apt to follow the ingestion of large amounts of glucose in diabetes, 
 and that when a diabetic has so far recovered his powers of meta- 
 bohsing carbohydrates that the sugar in the urine diminishes, it 
 may be partly replaced for a time by oxalates. Mayer gave 10 
 grams of sodium glucuronate to rabbits, and found that its ad- 
 ministration was followed by the appearance of saccharic acid 
 and a large amount of oxahc acid in the urine. He also found 
 oxahc acid in the hver, and states that oxalic acid is a product of 
 the autolysis of glucuronic acid in the hver. In this connection 
 it is interesting to note that a deposit of calcium oxalate crystals, 
 
GLUCURONIC ACID 405 
 
 and an increased excretion of oxalic acid, is in my experience a 
 much too frequent occurrence to be accidental in chronic pan- 
 creatitis. It is met with most frequently in old standing cases 
 where there is considerable cirrhosis of the gland, but not yet 
 sufficient to give rise to glycosuria. 
 
 A point of considerable interest in connection with the chemistry 
 of glucuronic acid must be mentioned here, and that is its relation 
 to the pentoses. Ruff has shown that by the action of certain 
 oxidising agents, and particularly hydrogen peroxide in the presence 
 of ferric acetate, d-arabinose can be obtained from the potassium 
 salt of d-glucuronic acid, and Salkowski and Neuberg have de- 
 monstrated that 1-xylose can be derived from glucuronic acid by 
 the action of putrefactive bacteria. The latter observation is 
 particularly important, for 1-xylose is a very constant constituent 
 of the cell nuclei of the body, more especially of the pancreas, and 
 if this change can be carried out by bacteria it is not unlikely that 
 it can also be effected by animal ferments in a similar way. It is 
 worthy of note that an increased excretion of glucuronic acid is 
 very constantly associated with inflammatory affections of the 
 pancreas, and that a substance which on hydrolysis yields a body 
 giving the reactions of a pentose may, according to my experience, 
 be also obtained. 
 
 Recognition. — The presence of glucuronic acid in the urine, or 
 other fluids of the bodj^, is most satisfactorily demonstrated by de- 
 composing the conjugate glucuronates with 1 per cent, sulphuric 
 acid in the autoclave, and preparing the p-brom-phenylhydrazin 
 compound. This is characterised by its high melting-point, 236° C. 
 (200° to 216° C. in the impure form), its insolubihty in absolute 
 alcohol, and its high degree of levo-rotation in pyridin- alcohol 
 solution ( - 7° 25'). Clinically, glucuronic acid compounds may be 
 recognised in the urine by the negative result of the fermentation 
 test ; by the urine being levo-rotatory, even after fermentation ; the 
 change produced in the rotatory power by boihng with acids, the 
 levo-rotation being diminished, or replaced by a dextro-rotation ; 
 by an increase in the reducing power of the urine after it has been 
 boiled with dilute sulphuric acid and neutrahsed ; also by the fact 
 that the orcin test which was previously negative, or only given 
 after prolonged boihng, is at once positive after treating the urine 
 with dilute sulphuric acid. Tollens's naphthoresorcinol test may 
 also be used, but as it gives a positive reaction with many normal 
 urines, owing to the conjugate glucm-onic acid they contain, the 
 result must be well marked before it can be concluded that an 
 abnormal amount is present. 
 
406 GLYCOSURIA 
 
 BIBLIOGRAPHY 
 
 Lactose 
 
 Blumenthal, Path. d. Harnes, 1903. 
 Borchardat and Finkelstein, Deut. med. Woch., 1893. 
 Bourquelot and Troisier, Compt. rend. d. Soc. d. Biol., 1889. 
 Gerard, Ann. d. Gynecol., xxxvii. 
 Gr6sz, Jahrbuch. f. Kinderheilk, 1892. 
 Halasz, Orvosi Hetilap, 1906 ; Deut. med. Woch., 1908. 
 Hess, Naunyn's Diab. Mellit., 1907. 
 Hofmeister, Zeit. f. phys. Chem., 1877. 
 Kaltenbach, Zeit. f. phys. Chem., 1878. 
 Langstein and Steinitz, Hofmeister's Beitr., 1906. 
 Lemaire, Zeit. f. phys. Chem., 1895. 
 MacCann, Lancet, 1897. 
 Mehu, Chem. Centralb., 1887. 
 Meyer, Mundl. Mitteil. 
 ISTaunyn, Der Diab. Mellit., 1807. 
 Nay, Arch. f. Oyndkol, 1889. 
 Pavy, Lancet, 1897. 
 
 Porcher, Bull. d. Soc. d. m,ed., 1902 ; Biochem. Centralb., 1910. 
 Voit, Zeit.f. Biol., 1892. 
 Worm-Miiller, Pfluger's Arch., 1884. 
 
 Zuelzer, Centralb. f. med. Wissensch., 1894 ; v. Noorden's Handb. d. 
 Path. u. Stoffwech., 1907. 
 
 Galactose 
 
 Bauer, Wien. med. Woch., 1906. 
 
 Gr6sz, Jahrb. /. Kinderheilk., 1892. 
 
 Langstein and Steinitz, Hofmeister'' s Beitr., 1906. 
 
 Voit, Zeit. /. Biol., 1892. 
 
 Saccharose 
 
 Brown, Johns Hopk. Hosp. Bull., 1900. 
 
 Hirschberg, Lancet Clinic, 1912. 
 
 Levy, V. Noorden's Handb. d. Path. u. Stoffwech., 1906. 
 
 Reuss, Wein. klin. Woch., xxiii. 
 
 Smolensk!, Zeit.f. phys. Chem., 1909. 
 
 Voit, Deut. Arch.f. klin. Med., 1897. 
 
 Pentose 
 
 Adler, Pfiuger's Arch., 1905. 
 D'Amato, Eev. crit. Klin., 1902. 
 Barczewski, Gaz. Lekaraska, 1897. 
 
 Bendix, Munch, med. Woch., 1903 ; Die Pentosurie, 1903. 
 Bial, Zeit. f. klin. Med., 1900 ; Berl. klin. Woch., 1904 ; Berl. Klinik, 
 1907. 
 
LACTOSURIA, ETC. 407 
 
 Bial and Blumenthal, Deut. med. WocJi., 1901. 
 
 Blum, Zeit.f. klin. Med., 1906. 
 
 Blumenthal, Berl. klin. Woch., 1895 ; Path. d. Harnes, 1903 ; v. 
 
 Noorden's Clin. Med., 1906. 
 Brat, Zeit.f. klin. Med., 1902. 
 Cassiver and Bamberger, Deut. med. Woch., 1907. 
 Chobola, Centralb. f. inn. Med., 1907. 
 Colombini, Monats. f. prakt. Dermatol., 1897. 
 Cominotti, Biochetn. Zeit., 1909. 
 Cremer, Zeit.f. Biol., 1882, 1893. 
 Ebstein, Virchow's Archiv., 1892—3. 
 Elliott and Raper, Journ. Biol. Chem., 1912. 
 Funaro, Arch, fartnac. sperim., 1907. 
 Garrod, Inborn Errors of Metabolisin, 1909. 
 Griind, Die Pentosurie, 1903. 
 Von Jaksch, Cantralh. f. inn. Med., 1906. 
 Janeway, Am,er. Journ. of Med. Sci., 1906. 
 Johnstone, Edin. Med. Journ., 1906. 
 
 JoUes, Zeit.f. anal. Chem., 1907 ; Munch, med. Woch., 1908. 
 Kaplan, New York Med. Journ., 1906. 
 Klercher, Nord. med. Arch., 1905. 
 Kraft, Pharmaceut. Centralb., 1906. 
 Kiilz and Vogel, Zeit. f. Biol, 1895. 
 
 Luzzato, Arch, di Farmacol, 1902 ; Arch. f. exp. Path., 1908. 
 Meyer, Berl. klin. Woch., 1901. 
 Neuberg, Berich. d. deut. Chem. Gesellsch., 1900 ; Ergebenisse. d. Physiol., 
 
 1904 ; V. Noorden's Path. d. Stoffwech., 1907. 
 Reale, Riv. Clin., 1894; Centralb. f. inn. Med., 1894. 
 Romme, Presse medicale, 1901. 
 Rosenfeld, Mediz. Klinik., 1906. 
 
 Salkowski, Zeit.f. phys. Chem,., xxvii. ; Berl. klin. Woch., 1895. 
 Salkowski and Jastrowitz, Centralb. med. Wissensch, 1892. 
 Schiiler, Munch, med. Woch., 1905. 
 Thierfelder, Zeit. f. pMjs. Chem., 1890. 
 Tintemann, Zeit. f. klin. Med., 1905-6. 
 Vas, Orvosi Hetilap, 1907 ; Wien. klin. Woch., 1908. 
 Wall, Am,er. Journ. Pharm., 1909. 
 
 Glucuronic Acid 
 
 Baumann and Herter, Zeit.f. phys. Chem., 1877. 
 Baumgarten, Zeit. f. exp. Path., 1906. 
 Blimienthal, Arch. f. Phys. (supl.), 1901. 
 Cammidge, Proc. Roy. Soc, 1909. 
 Embden, Hofm,eister's Beitr. z. chem. Phys., 1901. 
 Fischer and Pilot, Ber. d. deut. chem. Oes., 1891. 
 Herter, New York Med. Journ., 1898. 
 Hildebrand, Arch. f. exp. Path., 1900. 
 JoUes, Wien. med. Woch., 1911. 
 
408 GLYCOSURIA 
 
 Lavessen, Biochem. Zeit., 1907 
 
 Levy, Munch, med. Woch., 1905. 
 
 Mayer, Deut. med. Woch., 1901 ; Zeit. f. klin. Med., 1902. 
 
 Mayer and Neuberg, Zeit. f. phys. Chem., 1900. 
 
 Mendel and Jackson, Amer. Journ. Phys., 1902. 
 
 Ruff, Ber. d. deut. Chem., 1898. 
 
 Salkowski, Zeit. f. phys. Chem., 1904. 
 
 Salkowski and Neuberg, Zeit. f. phys. Chem., 1902. 
 
 Strauss, Deut. med. Woch., 1902. 
 
 Sundvik, Jahresb. f. Tierchem,., 1886. 
 
 ToUens, Zeit. f. phys. Chem., 1910. 
 
 ToUens and Stern, Zeit. f. Phys. Chem., 1910. 
 
 Weintraud, Naunyn's Diab. Mellit., 1907. 
 
 Woolley and Newburgh, Journ. Amer. Med. ^ssoc.,' 1911. 
 
CHAPTER XII 
 
 ALKAPTONURIA AND DIABETES INSIPIDUS 
 
 Alkaptonuria.— In 1858 Bodeker detected in the urine of a 
 patient with glycosuria a second reducing substance to which, on 
 account of its behaviour with alkaHes, he gave the name alkapton. 
 This substance, in spite of its reducing powers, was found not to 
 be a sugar, but to contain nitrogen. Other observers who investi- 
 gated subsequent cases of alkaptonuria held different views as to 
 the nature of " alkapton " ; some came to the conclusion that it 
 was pyrocatechin, others thought it was protocatechutic acid, 
 Marshall named it glycosuric acid, and Kirk, who separated an 
 acid from the urines of a group of cases that he investigated, called 
 it uroleucic acid. In 1891 Wolkow and Baumann isolated and fully 
 investigated homogentisic acid, the excretion of which is un- 
 doubtedly the essential feature of alkaptonuria. The work of 
 these, and other investigators, has definitely proved that this 
 substance has the empirical formula CgHgO^, and that it has the 
 constitution of para-di-oxybenzene-acetic acid (hydroquinone- 
 acetic acid). 
 
 Alkaptonuria is a very rare condition, and, although of great 
 interest to the chemical physiologist and pathologist, would not be 
 clinically important were it not that it may be mistaken for a 
 trouble of a much graver kind if the fact of its existence, and the 
 methods of differentiating it, are not borne in mind. The copious 
 reduction that occurs when an alkaptonuric urine is heated with 
 Fehling's solution will, to the uninitiated, suggest the presence of 
 sugar, but the dark brown colour of the liquid in which the copper 
 precipitate is suspended gives it a pecuhar appearance which, to 
 the experienced eye, indicates the true cause of the reduction. 
 Nylander's solution darkens on being heated with the urine, but 
 no precipitate of reduced bismuth forms, as it does with the sugars. 
 On adding a dilute solution of ferric chloride to the urine, drop 
 by drop, a transient deep blue coloration is seen to foUow each 
 addition, and is characteristic of homogentisic acid. The urine 
 does not ferment with yeast, is optically inactive, and does not 
 form an osazone with phenylhydrazin. When freshly passed the 
 
410 GLYCOSURIA 
 
 urine of an alkaptonuric person seldom exhibits any abnormality 
 of tint, but on exposure to the air it quickly darkens, especially if 
 it is made alkaline and is gently warmed. In some instances 
 attention has been drawn to the existence of the condition by hnen, 
 and other fabrics, soiled with the urine blackening on exposure to 
 the air. Crystals of uric acid deposited from the urine are found to 
 be stained brown. Beyond the presence of homogentisic acid the 
 urine of alkaptonurics shows no striking or constant variation from 
 the normal. 
 
 In the great majority of instances alkaptonuria is present from 
 birth, and persists throughout life. It may attract attention shortly 
 after the child is born through the staining of the clothes, or it 
 may pass unnoticed until adult life is reached, when its presence 
 is discovered in the course of an examination of the urine for 
 life insurance or some other purpose. A few cases have been re- 
 corded in which it has appeared as a temporary condition, but 
 according to Garrod the evidence of its temporary nature is doubtful 
 in some, and in others the fact that the urine contained homo- 
 gentisic acid was not completely established. In only one of them 
 was a quantitative estimation carried out. This patient, who 
 was under the care of Zimnicki, had intermittent alkaptonuria and 
 suffered from hypertrophic cirrhosis of the liver. Geyger's patient 
 was a diabetic and also had intermittent alkaptonuria. Hirsch 
 described the case of a girl of seventeen, with febrile gastro-enteric 
 catarrh, who passed for three days only a xu-ine which darkened on 
 standing, contained indican, and also yielded the alkapton reactions. 
 A somewhat similar case came under my notice in 1903. The 
 patient was a woman, aged thirty-two, in the City Fever Hospital, 
 Newcastle-on-Tyne, under the care of Dr. S. G. Mostyn, who sent 
 me the urine. This was of a dark colour, specific gravity 1-012, and 
 contained a trace of albumen. It reduced Fehling's solution, but not 
 Nylander's solution, and gave no osazone with phenylhydrazin. 
 A dilute solution of ferric chloride produced a transient blue 
 coloration and darkened the fluid. A fairly well-marked reaction 
 for indican was also obtained. As the quantity of urine was too 
 small for a detailed investigation, further proof of the nature of the 
 condition was not possible, and a second specimen sent to me for 
 analysis did not give the alkapton reactions. The darkening of 
 the urine on standing was only observed for three days in all. The 
 patient was lost sight of when she left the hospital, so that I cannot 
 say whether the condition recurred or not. 
 
 Homogentisic acid is apt to be found in the urines of several 
 brothers and sisters of a family whose parents do not exhibit the 
 
ALKAPTONURIA 
 
 411 
 
 anomaly, and Garrocl has pointed out that a striking proportion of 
 alkaptonurics are the offspring of the marriage of cousins. Ac- 
 cording to Bateson and Punnett, the mode of incidence of alkapton- 
 uria finds a ready explanation if it be regarded as a rare recessive 
 character in the Mendehan sense. Like chronic pentosuria, alkaj)- 
 tonuria is now considered to be an inborn error of metabolism. 
 It is beheved to be due to a failure of the organism to deal with 
 the aromatic fractions of the proteins of the food and tissues in the 
 ordinary way, and not to an abnormal formation of homogentisic 
 acid within the body, or to its production in the intestine from 
 putrefactive changes, as was at one time thought. It appears to 
 depend upon a lack of abihty of the alkaptonuric individual to 
 split open the benzene ring of the tyrosin and phenyl- alanin formed 
 in the intermediary metabolism of proteins. Normally these 
 first suffer a splitting- off of the nitrogen radical from the alanin 
 side chain, followed by oxidation to homogentisic acid. — 
 
 OH 
 
 /\ 
 
 \/ 
 
 CH2 
 
 I 
 CHNHo 
 
 v 
 
 CH, 
 
 HO 
 
 \ 
 
 OH 
 
 CH-NHn 
 
 CHo 
 
 I 
 COOH 
 
 COOH COOH 
 
 Tyrosin. Phenyl-alanin. Homogentisic acid. 
 
 Then comes a disintegration of the benzene ring with subsequent 
 complete oxidation. Alkaptonurics can carry out the conversion 
 as far as the oxy-acid stage, but there the process stops, and the 
 acid is excreted unchanged in the urine. The relation of homo- 
 gentisic acid to the aromatic radicals of the proteins has now been 
 definitely established by an experiment carried out by Abderhalden, 
 who succeeded in inducing alkaptonuria in a healthy man who had 
 never previously exhibited the condition by feeding him with 
 50 grams of tyrosin a day, a quantity corresponding to many 
 hundred grams of protein. This experiment does not throw any 
 light on the nature of the perversion, but it is possible that the 
 catabohsm of the aromatic fractions of the proteins is carried out 
 by a series of special enzjnnes, and it has been suggested that 
 alkaptonuria arises from the absence of the ferment which normally 
 has the power of splitting the benzene ring. 
 
 Alkaptonuria gives rise to no symptoms, with the exception 
 
412 GLYCOSURIA 
 
 of occasional dysuria and undue frequency of micturition. In 
 a few instances that peculiar staining of the tissues to which Virchow 
 gave the name of ochronosis, has developed in later life. 
 
 Diabetes Insipidus. — Diabetes insipidus is a condition 
 characterised by the persistent passage of an excessive quantity 
 of urine of low specific gravity, and without any constant abnormal 
 constituent. It is probably not a disease, but a symptom that 
 results from several morbid conditions, and will be briefly con- 
 sidered here because the cHnical manifestations of diabetes insipidus 
 resemble those of diabetes melhtus in some respects, and there is 
 experimental evidence that the two conditions may be set up by 
 similar lesions. 
 
 The polyuria associated with hysteria, nervous excitement, 
 high tension, arteriosclerosis, hydronephrosis, and chronic inter- 
 stitial nephritis should be distinguished from diabetes insipidus, for, 
 although the amount of urine excreted is increased in all of them, 
 it is represented by pints rather than by quarts, and the cause of 
 the condition is obvious. 
 
 Etiology. — True diabetes insipidus is a rare condition. It may 
 exist in the new-born infant, or appear in old age, but is most 
 commonly met with in early adult life. It is about twice as fre- 
 quent in males as in females. Heredity appears to play some part 
 in its etiology, a history of polyuria, glycosuria, or albuminiu-ia in 
 previous generations being not uncommon. Gee has reported an 
 example of the transmission of diabetes insipidus through four 
 generations. A history of tuberculosis in the family is, according 
 to Haussen and Bertrand Dawson, too frequent to be a mere 
 coincidence. The onset of diabetes insipidus appears to be con- 
 nected in some cases with nervous affections, nervous excitement, 
 acromegaly, syphiUs, blows or injuries of the head, or of the trunk 
 or limbs. Sudden exposure to cold, alcohohc excess, convalescence 
 from acute febrile diseases, malnutrition, and occasionally the 
 presence of abdominal or thoracic tumours, have been stated to 
 be the exciting cause. 
 
 Symptoms. — As a rule diabetes insipidus comes on slowly and 
 insidiously, but the onset is sometimes sudden, especially after a 
 fright or injury. Unquenchable thirst and polyuria are the out- 
 standing features of the condition. The appetite is usually good, 
 but is rarely excessive, as in diabetes mellitus. The bowels are 
 often confined, and the patient sometimes complains of flatulence. 
 The mouth is dry, the sMn is harsh, and may become atrophic and 
 withered, but boils and cutaneous lesions are rare. Itching of the 
 
DIABETES INSIPIDUS 413 
 
 skin is sometimes complained of. The patient may be well 
 nourished and healthy-looking, but the loss of sleep and distress 
 consequent on the great thirst and frequent urination, sooner or 
 later interfere with the general health and the patient becomes 
 thin and debilitated, his temper is irritable, he may complain of 
 distressing headache and a dull aching pain in the back. There 
 is a loss of sexual power, the knee jerks which at first are often 
 exaggerated may disappear, and there is a subnormal temperature. 
 In some cases the blood pressure is normal, in others it is raised, 
 but it is often abnormally low. Later the appetite fails, the 
 emaciation becomes more marked, great weakness supervenes, 
 the tongue becomes dry and glazed, attacks of diarrhoea may 
 occur, and death takes place from exhaustion, a low form of 
 pneumonia, or coma, if the patient has not been meanwhile carried 
 off by some intercurrent affection, such as pulmonary tuberculosis, 
 pneumonia, &c. As a rule the progress of the condition is slower, 
 and the prognosis better, when it comes on without any obvious 
 cause, than when it is associated with organic disease or injury. 
 In the former type of case the general health of the patient may be 
 well maintained for a lengthy period, and Osier states that the 
 affection has been known to persist for fifty years. Spontaneous 
 cure may occasionally take place, and has been known to follow an 
 intercurrent disease, such as typhoid fever. A few cases have 
 yielded to treatment, but as a rule the condition is intractable. 
 
 The amount of urine passed varies from 10 to 40 pints, or more,, 
 in the twenty-four hours, according to the quantity of fluid con- 
 sumed, but it is usually much in excess of that met with in diabetes 
 mellitus. It is clear, pale, of a greenish yellow colour, faintly acid 
 in reaction, and of very low specific gravity, TOOl to 1-006, but 
 usually about 1-005. As a rule the total amount of soHd con- 
 stituents is about normal, although the percentage is of course low. 
 Meyer states that the concentration of the urine tends to remain 
 uniform, and that the amount of water is varied to regulate the 
 concentration according to the amount of sohd eliminated. If the 
 nature of the diet is taken into account the output of urea is not 
 excessive. Some observers have reported an increased excretion. 
 Gerhardt, for example, met with 70 to 80 grams a day, but this was 
 explained by the increased appetite and consequent consumption 
 of a large amount of food by his patient. In other cases the urea 
 excretion has been subnormal, and, as a daily output of less than 
 about 20 grams a day renders the patient liable to become uremic, 
 it is important that regular estimations of the total excretion of 
 urea should be made in diabetes insipidus. Uric acid, creatinin, 
 
414 GLYCOSURIA 
 
 sulphates, and phosphates are usually present in normal quantities, 
 but occasionally an excess of phosphates has been found. Teissier 
 met with 6-6 grams, and 37"5 grams of urea, in one of his cases, 
 and suggests the name " Diabete phosphatique " for this type 
 of diabetes insipidus. The urinary constituent which is most 
 commonly, although not constantly, increased is inosite, which is 
 sometimes present in considerable quantities, 18 to 20 grams in the 
 twenty-four hours. Its significance is not well understood, but 
 Strauss has shown that it is probably related to the excessive 
 consumption of water, and consequent polyuria. Acetone, aceto- 
 acetic-acid, and the other products of abnormal metabohsm met 
 with in diabetes melhtus are not seen in cases of pure diabetes 
 insipidus. Occasionally traces of albumen are found in the urine, 
 and in some cases sugar is also present, especially if there is a lesion 
 of the nervous system. 
 
 Pathology. — Diabetes insipidus, like diabetes meUitus, presents 
 no constant anatomical lesions. Sometimes it has been found 
 associated with tumours, syphilitic or tuberculous growths, or 
 aneurisms, in the pons, medulla, or cerebellum. In others there 
 has been fracture of the base of the skuU, or mechanical injury to 
 the brain. In one reported case an abdomuial tumour, and in two 
 a thoracic aneurism, was present. Many have shown no gross 
 morbid change after death, excepting those due to intercurrent 
 disease, and alterations in the urinary system, such as enlarged 
 and congested kidneys, dilated pelves, dilated ureters, and an 
 hypertrophied bladder, which might be ascribed to the passage of an 
 abnormal amount of urine. 
 
 As the result of a study of the metabohsm in diabetes insipidus, 
 Tallqvist has suggested that the polyuria may be due to defective 
 resorption of water by the renal tubules, and a nmnber of authors, 
 including Strubell, Meyer, and Seller, have stated that there is a 
 special functional disorder of the kidneys consisting of a loss of 
 abihty to concentrate the urine. If this disabihty exists it would 
 necessitate a large volume of urine being excreted to remove the 
 waste products from the body, and this in its turn would cause the 
 consumption of an abnormally large quantity of hquid. If the 
 requisite amount of water were not taken the tissue fluids would 
 be drawn upon, or a retention of the sohd urinary constituents 
 would occur, with resulting uremia. The kidneys of certain patients 
 do appear to have a difficulty in excreting concentrated solutions, 
 and this has been found to apply more especially to solutions of 
 sodium chloride, and, to a less extent, of urea. If such a person 
 is given from 10 to 20 grams of sodium chloride with his food the 
 
DIABETES INSIPIDUS 415 
 
 concentration of the salt in his urine is only slightly increased (from 
 0-1 to 0*2 per cent.), and a large amount of urine, and a considerable 
 time, are therefore necessary for the excretion of the whole of the 
 salt taken. This fact has been made use of by Minkowski to 
 differentiate between polyuria due to inability of the kidneys to 
 concentrate the urine, and an increase in the amount of urine 
 excreted arising from other causes. According to Schmidt, the 
 increased flow of urine in diabetes insipidus is dependent upon 
 dilatation of the vessels of the kidneys without increased arterial 
 pressure, and, since he found that a similar condition can be pro- 
 duced in animals by cutting the renal nerves, it is assumed that the 
 vaso- dilatation is dependent upon nervous influences, due either 
 to local irritation, as from abdominal tumours, &c., or to central 
 causes. Bernard showed that experimental puncture of the floor 
 of the fourth ventricle just above the diabetic centre, produces 
 copious diuresis in animals, and it has also been found that injuries 
 to the central lobe of the cerebellum, corresponding to the vermi- 
 form process of the human brain, will in certain animals produce a 
 similar result, so that the causal relation of central lesions in, or 
 about, these regions to diabetes insipidus receives confirmation on 
 experimental grounds. These experiments also offer an explana- 
 tion of the appearance of sugar in the urine in some cases. The 
 observations of Goetsch, Gushing, and Jacobson on the effects 
 produced by partial removal of the hypophysis cerebri in animals, 
 also throw light on the connection between tumours, or injuries, 
 of the brain and diabetes insipidus, for they show that hyper- 
 activity of the hypophysis, such as would be caused by lesions 
 stimulating the posterior lobe, is accompanied by polyuria, and, as 
 we have already seen, is hable to give rise to glycosuria. Schafer 
 showed that in dogs constant mechanical irritation of the hypo- 
 physis causes permanent diabetes insipidus and a tendency to 
 adipose-genital dystrophia. The conclusion seems inevitable that 
 the internal secretion of the hypophysis controls the activity of the 
 kidneys, and that essential diabetes insipidus may arise from 
 excessive functioning of the gland. In support of this there is also 
 considerable clinical evidence. Thus in Hagenbach's case, in 
 which polyuria and polydipsia occurred in a Httle girl, cheesy 
 tubercle of the infundibulum was found post-mortem. Rosenhaupt 
 has described a case in which fever, thirst, and polyuria came on 
 abruptly, and necropsy two weeks later revealed a sarcoma in 
 the anterior lobe of the hypophysis. Frank reports a case in which 
 disturbances similar to those produced in experimental research 
 on the hypophysis was brought about by a bullet. The patient, 
 
416 GLYCOSURIA 
 
 a man aged thirty-nine, became epileptic several years after he had 
 tried to commit suicide by firing two bullets into his right temple. 
 The balls could be seen in the head, one near to the cortex, and the 
 other close to the sella turcica. The latter evidently kept up a 
 constant mechanical irritation of the hypophysis and induced the 
 polyuria, &c., from which he suffered. The brain affections which 
 most usually accompany diabetes insipidus are those in which there 
 has been traumatic concussion, and, as the region of the hypophysis 
 is particularly liable to suffer in trauma of the skull, it is not unfair 
 to assume that the resulting cicatrical changes exert a permanent 
 irritating effect on the gland, either by direct pressure, or from 
 interfering with the flow of cerebro-spinal fluid. Tumours of the 
 brain may also act in a similar manner. A typical case of diabetes 
 insipidus resulting from injury of the skull and followed by glycosuria 
 probably brought about in this way has been described by French 
 and Ticehurst : — 
 
 A man, aged forty-four, after a fall from his bicycle remained 
 unconscious for fourteen days, with cerebro-spinal fluid dripping from 
 his nose. On recovery, bilateral temporal hemianopia and ocular 
 paresis showed that he had fractured the base of his skull near the 
 chiasma. Previous to the accident he had been perfectly well ; after 
 it he had extreme polyuria and thirst, passing up to 10,000 c. cm. 
 of iirine, specific gravity 1004*6, free from albumen and sugar. The 
 expectation that the " sugar centre " in the medulla was injured led 
 to extensive metabolism research in which all food and excreta were 
 analysed. By increasing the carbohydrates in the food, it was 
 attempted to produce glycosuria. Dextrose, starch, and cane-sugar 
 were assimilated well, up to 700 grams (dry) of carbohydrate in twenty- 
 four hours. No sugar could thus be made to appear in the urine. Two 
 years later, however, glycosuria developed spontaneously. The urine 
 remained of comparatively low specific gravity. Carefiil dieting did 
 not entirely prevent glycosuria, and both acetone and diacetic acid 
 were present as well as sugar. On ordinary diet there were no acetone 
 and diacetic acid. 
 
 Diagnosis and Treatment. — If the condition can be assigned to 
 any definite cause an attempt must be made to deal with it, but 
 if none can be discovered, and the case is apparentl}^ one of " idio- 
 pathic " diabetes insipidus, it is necessary to determine (1) whether 
 the increased excretion of urine is a primary disorder, which will 
 lead to a reduction in the water-content of the organism unless 
 large quantities of fluid are taken, or (2) whether the real cause of 
 the polyuria is an excessive thirst accompanied by excessive 
 drinking ? An answer to these questions can be obtained by ob- 
 serving the result of a reduction in the intake of water on the 
 
DIABETES INSIPIDUS 417 
 
 amount of urine passed, being careful meanwhile to watch the 
 effect on the composition of the urine, and particularly the daily 
 excretion of urea. If there is a primary polydipsia the excretion 
 of urine should be reduced by limiting the quantity of water taken, 
 but if this result does not follow very quickly the inference is that 
 the polyuria is the primary disorder. In the latter case there will 
 also be a marked loss of weight, due to removal of water from the 
 tissues. One may then proceed to attempt to differentiate between 
 a polyuria dependent upon inability of the kidneys to concentrate 
 the urine, and polyuria due to some other cause, by Minkowski's 
 method. If after giving 10 to 20 grams of sodium chloride the 
 urine of the next twenty-four hours contains practically all the 
 salt, the condition is not primarily renal, and the patient may be 
 safely placed on a limited amount of water. The limitation of the 
 fluids must, however, be carried out with care and be guided by 
 analyses of the urine, for if it is found that large quantities of urine 
 are being passed, even with a limited intake of water, it is obvious 
 that the patient must be extracting fluid from his own tissues with 
 resulting immediate and perhaps permanent injury, and the supply 
 must be increased. If, on the other hand, very little of the salt is 
 excreted without increasing the water intake, the inference is that 
 the excretory powers of the kidneys are at fault, and the best 
 treatment would appear to be to place the patient on a diet poor in 
 sodium chloride and nitrogenous foods. In one patient treated by 
 this method Minkowski obtained a reduction in the volume of urine 
 from 12 or 14 litres to 3 or 4 litres a day. Even if a restriction of 
 the chlorides does not bring about such a remarkable result as this, 
 it will tend to diminish thirst and so assist in the treatment of the 
 case. A nitrogen-poor and salt-free diet cannot, however, be per- 
 sisted in indefinitely, but by watching the weight and general 
 metabolism it may be possible to gradually relax it as the condition 
 of the kidneys improves until a more normal diet is reached, the 
 sodium chloride being still adjusted as far as possible to the tolerance 
 of the patient. 
 
 Patients suffering from diabetes insipidus should be kept free 
 from worry and anxiety. Their clothes should be warm, and the 
 food be made as nourishing and abundant as possible, with the 
 limitations mentioned. Toxaemia should be carefully guarded 
 against by regulating the bowels. The distressing thirst may be 
 assuaged somewhat by acid drinks, or by sucking ice dipped in 
 lemon-juice. Tea, coffee, and alcoholic beverages are to be avoided, 
 although the tolerance for alcohol is in some cases remarkable, a 
 couple of pints of brandy, or a dozen or more bottles of wine, 
 
 2d 
 
418 GLYCOSURIA 
 
 having been consumed in a day (Osier). A bracing climate, or a 
 sea- voyage, is a useful adjuvant to the treatment in some cases. 
 The most generally useful drugs are tonics, such as arsenic, iron, 
 quinine, and strychnine. Anti-spasmodic remedies have been much 
 used, and of these valerian was highly recommended by Trousseau, 
 who gave it in enormous doses. Antipyrin, salicylate, and tur- 
 pentine have been found useful in some cases, but they should 
 be given with caution. Opium and its alkaloids are worse than 
 useless in diabetes insipidus, for although the thirst and polyuria 
 may be diminished by their use, they greatly increase the patient's 
 discomfort, and in some instances have proved fatal. Occasional^ 
 they may be cautiously administered to combat sleeplessness and 
 nervous symptoms, but, as a rule, bromides, unless specially contra- 
 indicated, are to be preferred. Diuretics have not given favour- 
 able results. If the blood pressure is low, vaso-constrictor drugs, 
 especially ergot and extract of the pituitary gland, are indicated, 
 since these raise the blood pressure and tend to prevent cerebral 
 congestion and exudation. It is also possible that pituitary 
 extract may have a specific action in some cases, but unfortunately 
 the results are not permanent. In cases where the blood pressure 
 is high, it should be lowered by appropriate baths, massage, physical 
 exercise, change to a warm chmate, diet, and vaso-dilator drugs, 
 such as nitro-glycerine, &c. Finally, the effect of electrical treatment 
 may be tried by applying the positive pole of a constant current of 
 1 to 4 miUiamperes to the nape of the neck, and the negative pole, 
 properly insulated, to the posterior naso-pharyngeal wall. Herrick 
 has recently reported a remarkable improvement in a case of diabetes 
 insipidus of four years' standing, possibly due to a lesion of the 
 hypophysis, following the withdrawal of 5 c.c. of cerebro-spinal 
 fluid by lumbar puncture. The urine, which before the operation 
 had varied in amount from 7,500 to 11,000 c.c, with a specific 
 gravity of 1-001, never exceeded 1,800 c.c, with a specific gravity 
 averaging 1*015, during the succeeding four weeks that he was 
 under observation. The thirst at the same time disappeared. 
 
 BIBLIOGRAPHY 
 
 ALKAPTONURIA 
 
 Abderhalden, Lehrb. d. phys. Chem., 1906 ; Zeit. f. phys. Chem., 1912. 
 Bateson, Rep. Evol. Commt. Roy. Soc, 1902. 
 
 Bodeker, Zeit. f. rat. Med. 1859 ; Ann. d. Chem. u. Pharm., 1861. 
 Falta, Biochem. Centralh., 1904 ; Deut. Arch. /. klin. Med., 1904. 
 Garrod, Lancet, 1902 ; Inborn Errors of Metabolism, 1909. 
 
ALKAPTONURIA, DIABETES INSIPIDUS 419 
 
 Geyger, Pharmakeut. Zeit., 1892. 
 
 Hirsch, Berl. klin. Woch., 1897. 
 
 Kirk, Journ. Anat. and Phys., 1889 ; Brit. Med. Joiirn., 1888. 
 
 Marshall, Medical News, 1887. 
 
 Punnett, Proc. Roy. Soc, 1908. 
 
 Virchow, Virchow''s Arch., 1866. 
 
 Wolkow and Bamnann, Zeit. f. phys. Chem., 1891. 
 
 Zimmick, Abst. Centralb. f. Stojfwech. u. Verdauungsk., 1900 
 
 Diabetes Insipidus 
 
 Dawson, Allchin's Man. of Med., ii., 1900. 
 
 Frank, Berl. klin. Woch., 1912. 
 
 French and Ticehiarst, Brit. Med. Journ., 1906. 
 
 Gee, St. Barthol.'s Hosj:). Re]:)., 1877. 
 
 Gerhardt, Der Diah. Insip., 1898 ; Spec. Path. u. Therap., vii. 7, 1899. 
 
 Goetsch, Gushing, and Jacobson, Johns Hopk. Hosp. Bull., 1911. 
 
 Haussen, Norsk. Mag. f. Laegevidensk., 1912. 
 
 Herrick, Arch. Internal. Med., 1912. 
 
 Kiilz, Maly's Jahrb., 1875-6. 
 
 Meyer, Deut. Arch. f. klin. Med., 1905. 
 
 Minkowski, Die Therap. d. Gegenwart, 1910. 
 
 Osier, Princip. and Pract. of Med. 
 
 Schmidt, Wien klin. Woch., 1905. 
 
 Seller, Zeit. f. phys. Chem., 1907. 
 
 Strauss, Centralb. f. inn. Med., 1872. 
 
 Tallqvist, Zeit. f. klin. Med., 1903. 
 
 Teissier, quot. Bkmienthal, Path. d. Harnes, 1903. 
 
 Vohl, Arch. f. phys. Heilk., 1858. 
 
APPENDIX 
 
 GENERAL PROPERTIES AND REACTIONS OF THE SUGARS 
 AND RELATED SUBSTANCES 
 
 Specific G-ravity of Saccharine Solutions.^ — The specific gravity 
 of a saccharine sohition depends chiefly on the amount of dissolved 
 solid, and is approximately ec[ual for equal strengths of different 
 carbohydrates. Solutions of dextrose have a slightly lower specific 
 gravity than the corresponding solutions of cane-sugar, while solutions 
 of levulose, maltose, and invert sugar give slightly higher readings. 
 
 Fermentation. — Yeasts, and certain other lowly organisms, when 
 placed in a saccharine fluid, and kept under suitable conditions, are 
 capable of rapid multiplication and exert a peculiar chemical change, 
 known as alcoholic fermentation, in the mass of the sugar, which is 
 broken up into alcohol and carbon dioxide. This decomposition is 
 not the direct result of the changes in the plant protoplasm, but is a 
 side issue of its life processes resulting from the action of an enzyme 
 or ferment which it produces. By careful experiment it has been 
 shown that only about 1 per cent, of the sugar is used as food by the 
 yeast cells in their growth, and that the actual amount of carbon 
 dioxide they expire is verj^ small. In the case of dextrose the chemical 
 changes induced by alcoholic fermentation may be represented : — 
 
 CeHisOfi = 2C2H6O + 2CO2. 
 
 Theoretically, 100 parts of dextrose should yield 51- 11 parts by 
 weight of ethyl alcohol and 48"89 parts of carbon dioxide, but it was 
 shown by Pasteur that only 48-5 per cent, of alcohol and 46-5 per cent, 
 of carbon dioxide actually result. In addition from 2*5 to 3 per cent, 
 of glycerine, 0"4 to 0'7 per cent, of succinic acid, and 0*8 to 1-3 percent. 
 of other substances, including amyl alcohol, isobutyric alcohol, propyl 
 alcohol, and traces of fatty acids and lactic acid, are formed. 
 
 The ease with which the different sugars undergo fermentation 
 varies considerably, and is partly dependent upon the variety of yeast 
 employed, and partly on the structiu-e of the sugar. Grape-sugar,, 
 fructose, mannose, and invert sugar are fermentable by all yeasts, but 
 cane-sugar, maltose, and milk-sugar are only fermented after inversion 
 by dilute acids or appropriate enzymes, which for cane-sugar and maltose 
 are contained in ordinary yeast. The pentoses are not fermented by 
 any known yeast, but they are attacked and slowly broken down by 
 bacteria. It has been found that any species of yeast which ferments- 
 one of the three sugars, glucose, mannose, and fructose, also ferments- 
 
 420 
 
APPENDIX 421 
 
 the others, and with approximately the same readiness. These three 
 hexoses are closely related to each other in structiire, and it is possible 
 to convert them one into the other by treatment with alkalies. In 
 the transformation it is assumed that a common, or enolic, form acts 
 as an intermediate substance. Similarly in the process of fermentation 
 it is believed that the first step is the conversion of the sugar into this 
 common form by an enzyme contained in the yeast. Subsequent 
 action of this, or another, enzyme in the yeast causes simplification of 
 the molecule, the breakdown commencing at the linkage between the 
 terminal carbon atoms. Galactose is fermented with much greater 
 difficulty than glucose, and many varieties of yeast do not act upon 
 it at all. No yeast is known, however, which is capable of fermenting 
 galactose which does not also ferment glucose. It is believed that 
 this variation in susceptibility to fermentation is dependent upon a 
 difference in the configiuation {i.e. the position of the hydroxyl, -OH, 
 groups relative to the chain of carbon atoms) in the two sugars which, 
 while not sufficient to prevent fermentation altogether, makes galactose 
 far more resistant to attack. It is probable that galactose is fermented 
 by a different mechanism, and that the enzyme which converts it to 
 the enolic form is less widely distributed in yeasts than that which 
 produces the change in glucose, mannose, and fructose. This hypo- 
 thesis is supported by the fact that the isomerides of galactose, talose, 
 and tagatose are not fermented by any yeast whose action toward 
 them has up to the present been investigated, and presumably no 
 yeast exists capable of converting them into the enolic form. Many 
 other substances which are closely related to glucose, such as gluconic 
 acid, glucuronic (glyciironic) acid, the methyl glucosides, &c., are not 
 fermented by yeast, althotigh the greater part of the molecule is the 
 same. The shortening of the carbon chain to five atoms in the pentoses 
 also appears to be sufficient to place the sugar molecule out of harmony 
 with the yeast enzyme, and thus prevent its disruption. From these, 
 and other, considerations it appears highly probable that there is a 
 very close relationship between the configuration of a fermentable 
 sugar and the enzyme which causes its fermentation. Armstrong has 
 aptly compared the relationship to that which exists between the 
 fingers of a glove on the hand ; it is impossible to fit the glove if the 
 position of the fingers is altered, and, moreover, a right-handed glove 
 will not fit the left hand. 
 
 In testing a saccharine fluid for fermentation it is necessary that 
 certain points should be attended to — (1) the solution should not be 
 too concentrated, over 10 per cent. ; (2) it should be neutral, or 
 faintly acid, in reaction, never alkaline ; (3) no free antiseptic should 
 be present ; (4) the yeast should be fresh and free from starch ; (5) the 
 addition of a little yeast ash, or sodium phosphate and potassium 
 tartrate, is advisable with pure solutions to provide material for the 
 growth of the yeast ; (6) a blank experiment should be conducted to 
 prove that the yeast by itself does not give off carbon dioxide imder 
 the conditions of the experiment ; (7) the fermenting fluid should be 
 kept at a temperature of 25° to 35° C. (77° to 95° F.). 
 
422 GLYCOSURIA 
 
 Optical Characters — Polarisation. — If a piece of the semi-trans- 
 parent mineral tourmaline is cut into slices by sections parallel to its 
 axis, and one of these slices is laid upon another, it is found that in 
 certain positions they form an opaque combination, while in others 
 they are transparent. The combination is most transparent in two 
 positions, one when the slices lay in the natural relation they occupied 
 in the crystal, and the other when one of the slices is rotated through 
 180°, and is most opaque in two other positions at right angles to these. 
 The light that has passed through one such plate is in a peculiar con- 
 dition, and, since it contains rays that vibrate in one plane only, is 
 said to be plane-polarised. Polarised light cannot be distinguished 
 from ordinary light by the naked eye, but is shown by the interference 
 to its passage caused by the interposition of another " polariser," 
 called the " analyser," with its axis at right angles to the first. 
 
 In the above system the plate of tourmaline next the eye is the 
 analyser, and the other plate the polariser. One of the most con- 
 venient and effective contrivances for polarising light, or analysing it 
 when polarised, is that known, after its inventor, as Nicol's prism. 
 This is made by splitting a rhomb of Iceland-spar, or calc-spar, along 
 a diagonal plane, and cementing the two pieces together in their 
 natural position with Canada balsam. Iceland-spar, in common with 
 certain other substances, shows the phenomenon of double refraction — 
 that is to say, an incident beam of light gives rise to two refracted 
 rays which take different paths, one of these rays is termed the 
 " ordinary," the other the " extraordinary " ray. In the Nicol's 
 prism the ordinary ray is totally reflected on meeting the first surface 
 of the Canada balsam and passes out at one side of the prism, while 
 the extraordinary ray is transmitted through the balsam, and emerges 
 at the end of the prism parallel to the direction of the incident beam, 
 but polarised. This apparatus has nearly all the advantages of a 
 toTormaline plate, with the additional advantage of much greater 
 transparency and of complete polarisation. 
 
 When a plate of quartz, cut perpendicular to the axis, is inter- 
 posed between a polariser and an analyser coloiu" is exhibited, the 
 tints changing as the analyser is rotated. Similar coloiir effects are 
 produced when the solution of sugar, enclosed in the tube with plane 
 glass ends, is substituted for the c[uartz. If homogeneous light, such 
 as that from a sodivun flame, is employed, it is foLind that, if the 
 analyser is first adjiisted to produce total extinction of the polarised 
 light and the quartz or saccharine fluid is then introduced, there is 
 partial restoration of light. On rotating the analyser through a 
 certain angle, depending in the one case on the thickness of the quartz 
 plate and in the other on the length of the tube and the strength of 
 the solution, there is again complete extinction of light. The action 
 thus exerted is called " rotation of the plane of polarisation." In the 
 case of ordinary cjuartz and solutions of certain siibstances, it is neces- 
 sary to rotate the analyser in the direction of the hands of a watch 
 as seen by the observer, the rotation of the plane of polarisation is then 
 
APPENDIX 423 
 
 said to be right-handed, and the substance to be " dextro-rotatory." 
 In cases of left-handed quartz and sokitions of certain other substances, 
 the rotation of the plane of polarised light is in the ojDposite direction, 
 and the substance is said to be " levo -rotatory." 
 
 The power of rotating polarised light possessed by a particular 
 sugar is, iinder certain circumstances, a fixed quantity, known as its 
 " specific rotation," and as this property is also exerted by solutions 
 of the sugars, the angle through which rotation occurs serves for their 
 accurate estiination. 
 
 The specific rotation of a substance may be defined as the amount 
 of rotation, in degrees of a circle, of the plane of polarised light pro- 
 duced by 1 gram of the substance, dissolved in 1 c.c. of liquid, enclosed 
 in a tube 1 decimetre long. The reading is usually taken at 30° C. and 
 by homogeneous yellow (Na) light. It is necessary to make the measiire- 
 ments with monochromatic light of one particular wave-length, as the 
 apparent specific rotatory power of a substance varies greatly with the 
 wave-length of the light employed. It is most usual to refer the 
 rotation to the D line of the spectrtuii, the rotation being expressed 
 as [ajo. 
 
 The optical rotatory power of the freshly made solution of most 
 sugars undergoes a change on standing, sometimes increasing, but 
 more generally falling, until a constant value is reached. This pheno- 
 menon, which is now known as " mutarotation," or " multirotation," 
 was formerly termed birotation, because the rotatory power of dextrose 
 in a fresh solution is nearly twice as great as it eventually becomes. 
 Mutarotation is due to the fact that most sugars exist in solution as 
 a mixture of two forms in equilibrium. Thus, solid dextrose is the 
 a-modification of high rotatory power which does not persist as such 
 in the freshly made solution, but slowly passes over in part into a 
 /3-form of lower rotatory power. The change takes place very slowly 
 when highly purified materials are used, but is much accelerated by 
 impi.u"ities, and takes place almost immediately if a trace of an alkali 
 is added to the solution. 
 
 Melting-point. — The carbohydrates seldom melt sharply, be- 
 cause fusion is nearly always preceded by slight decomposition. Their 
 melting-points are therefore of minor importance as specific properties. 
 The instability of the sugars towards heat is also manifested by their 
 tendency to pass into the state of uncrystallisable sjnrups when their 
 solutions are concentrated by boiling under the ordinary atmospheric 
 pressLU-e. 
 
 The Action of Alkalies. — All the monosaccharides are readily 
 decomposed by alkalies. On being heated with a caustic alkali, such as 
 sodium, potassium, or ammonii.im hydrate, a solution of a mono- 
 saccharide sugar, such as dextrose, turns brown at about 60° C, and 
 is entirely decomposed by prolonged boiling into a variety of sub- 
 stances, including lactic acid, formic acid, and various aldehydes. 
 Alkaline salts, such as sodium or potassium carbonate, have a similar. 
 
424 GLYCOSURIA 
 
 but less intense, action. Cane-sugar is not affected by dilute solutions 
 of caustic alkalies, or alkaline carbonates, in the cold, and only very 
 slowly on heating. It is decomposed, however, by being boiled with 
 a strong alkaline solution. Fused with solid cavistic potash it gives 
 rise to potassium oxalate, acetate, &c. 
 
 The Action of Oxidising Agents. — The reducing sugars which 
 contains an aldehyde group are easily oxidised by bromine in the 
 presence of water, and give rise to monobasic acids by a transformation 
 of their terminal -CHO group into carboxyl :— - 
 
 R.CH0 + H20-fBr = R-C00H + HBr. 
 
 Thus xylose gives rise to xylonic acid, glucose to gluconic acid, 
 mannose to mannonic acid, and galactose to galactonic acid, the last 
 three being stereo-isomers of the same formula, C5H'g(OH)5.COOH. 
 In some instances (e.7. xylose) this reaction can be utilised to differen- 
 tiate the sugar. 
 
 Action of Concentrated Mineral Acids. — Concentrated nitric acid 
 in the cold combines with sugars to form nitric esters. 
 
 When heated with moderately concentrated nitric acid the sugars 
 undergo oxidation, giving rise to acids which differ accordingly to the 
 nature of the sugar and the concentration of the acid. The aldoses 
 are oxidised at each end of the chain, and yield di-basic acids with 
 the general formula, COOH(CH.OH)„COOH. 
 
 Thus when dextrose (2 grams) is heated with nitric acid of a density 
 of 1-2 (10 c.c), evaporated to a syrup, dissolved in water (5 or 6 c.c), 
 the solution saturated with potassium carbonate and acidified with 
 glacial acetic acid (3 or 4 c.c), wliite, transparent, needle-like crystals, 
 arranged in rosettes, or singly, of acid saccharate of potassium separate 
 out on cooling, owing to the formation of saccharic acid, by the action 
 of the nitric acid on the dextrose. If galactose is treated in a similar 
 way mucic acid is formed, and can be readily separated by its in- 
 solubility in water. Lactose being a di-saccharide yielding dextrose 
 and galactose, gives rise to a mixture of saccharic and mucic acid. 
 The latter can be separated from the former by its insolubihty in 
 water, appearing as short prisms, and can be distinguished from 
 calcium oxalate by its complete solubility in aiximonia. The formation 
 of mucic acid is characteristic of galactose, and substances containing 
 it, and is therefore used to detect their presence. The ketoses on being 
 heated with nitric acid break down and yield acids poorer in carbon. 
 Thus le\'ulose forms oxalic acid, tartaric acid, acetic acid, formic acid, &c. 
 
 Dextrose dissolves in cold sulphuric acid to form dextrose siolphonic 
 acid, without undergoing any colour change. Cane-sugar, on the other 
 hand, is readily carbonised by concentrated sulphuric acid, forming a 
 bulky black mass, with the evokition of sulphur dioxide and other 
 volatUe products. 
 
 Action of Dilute Mineral Acids. — All the di- and poly- 
 saccharides when heated with dilute mineral acids iindergo hydrolysis 
 
APPENDIX 425 
 
 or inversion — that is to say, they are decomposed into the simple 
 monosaccharides from which they are derived. A solution of cane- 
 sugar, for instance, when heated with dilute hydrochloric, or sulphuric 
 acid, gives rise to a mixture of dextrose and levulose. In the process 
 the specific gravity of the solution is raised, the sugar loses its power 
 of crystallising readily, and its optical activity is changed from right to 
 left-handed — that is to say, it is " inverted." This change in optical 
 activity is due to the levo -rotatory power of the levulose formed being 
 greater than the dextro-rotatory power of the dextrose, which is present 
 in equal amount. Although a similar change in rotatory power does 
 not necessarily follow the hydrolysis of other di-saccharides, &c., the 
 term inversion is frequently applied to the process generally. It is 
 better, however, to speak of it as hydrolysis, for it is attended by the 
 assimilation of the elements of water : — 
 
 CjgHoaOii + HjOq = CgHjjOQ + C^HjoOg . 
 The rate at which hydrolysis, or inversion, takes place depends 
 upon the natiu-e of the acid, its concentration, the temperature, and 
 the nature of the sugar. Cane-sugar is most readily inverted, maltose 
 much less readily, and lactose a little less readily than maltose. Cane- 
 sugar is inverted by boiling with citric acid, 2 per cent., but lactose is 
 unaffected. When the reducing powers of a sugar before and after 
 inversion are to be compared, the acid solution must be made neutral, 
 or nearly neutral, with sodium carbonate. 
 
 Action of Organic Acids. — With organic acids sugars form 
 ethereal salts, or esters. The most important of these is the benzoyl 
 ester, which is particularly usefiil in isolating dextrose and other 
 carbohydrates from physiological and pathological fluids. If a solu- 
 tion of dextrose is shaken with 6 parts of benzoyl chloride and 48 parts 
 of a 20 per cent, solution of sodiiun hydrate for each part of dextrose 
 in the solution, until the smell of benzoyl chloride has disappeared, 
 it forms dextrose pentabenzoate, which crystallises out on cooling the 
 fluid on ice and standing for twenty-fovir hours. If the benzoate is 
 then separated off, dissolved in alcohol, and recrystallised, it appears 
 as coloxorless needles with a melting-point of 179° C. 
 
 Eeducing Properties. — With the exception of cane-sugar and 
 raffinose, all the commonly occurring sugars show a strong tendency 
 to undergo oxidation, and therefore act as reducing agents. This 
 property depends upon the presence of an active carbonyl group. It 
 is consequently not specific of the sugars, but is shared by all sub- 
 stances having the properties of aldehydes and ketones. Bismuth, 
 mercury, silver, platinum, and gold salts are reduced to the metallic 
 state by hot alkaline solutions of the reducing sugars, and some reduce 
 ammoniacal silver nitrate even in the cold. Cupric and ferric salts 
 are converted into cuprous and ferrous compounds, while picric acid 
 is converted into picramic acid, ferricyanides to ferrocyanides, and 
 various dyes, such as indigotin, are reduced to colourless compoimds 
 by heating with alkaline solution of these sugars. 
 
426 GLYCOSURIA 
 
 The reaction that occurs when a sohition of a reducing sugar is 
 heated with an oxidising agent in the presence of a caustic alkali is 
 complex, and is only imperfectly understood. The principal products 
 are said to be formic, oxalic, glycollic, and carbonic acids, but the 
 products of the alkali itself on the sugar have also to be taken into 
 account. The non-reducing sugars show the same reactions after 
 hydrolysis by heating with acids, or through the action of enzymes. 
 If a solution of cane-sugar, for instance, is heated for five minutes, or 
 longer, in a boiling water bath with one-twentieth of its volume of 
 concentrated hydrochloric acid, and is then cooled, neutralised with 
 soda solution, and tested, it will be found to reduce alkaline solutions 
 of the metals, &c. 
 
 Colour Reactions. — Wlien a reducing sugar is heated with a con- 
 centrated mineral acid it yields furfurol, which can be detected by 
 the formation of coloured condensation products with various sub- 
 stances of the phenol group. The colour obtained depends on the 
 variety of sugar, and upon the phenol employed. 
 
 Molisch's Test. — About 5 mg. of the siibstance to be tested are 
 placed in a narrow test-tube, and dissolved in 10 drops of water. 
 Two drops of a 10 per cent, chloroform solution of a-naphthol are 
 added, and the contents of the test-tube well mixed. One c.c. of 
 pure concentrated sulphuric acid is allowed to flow down the lower in- 
 clined side of the tvibe so that it may form a layer beneath the aqueous 
 solution, without mixing with it. In the presence of a carbohydrate 
 a red ring will appear at the line of junction within a few seconds. 
 On standing the colour soon changes to a dark purple. If the 
 tube is shaken, and allowed to stand for one or two minutes, and 
 the contents then diluted with 5 c.c. of cold water, a dull violet 
 precipitate will immediately appear if a carbohydrate is present. 
 The addition of an excess of strong ammonia changes the coloiu" ta 
 a rusty yellowish-bro-mi. Any substance that gives dull violet and 
 rusty brown precipitate, as well as the purple coloration, under the 
 circxamstances described, may be assumed to be a carbohydrate. 
 It is essential, however, that the substance should be free from 
 all traces of filter-paper, particles of wood, dust, &c., as the test is 
 extremely delicate. The purity of the reagents employed should 
 also be beyond all qxiestion, and it is most important that the sul- 
 phuric acid should be free from all traces of nitrous acid. It is 
 advisable to condvict a blank experiment by shaking 1 drop of the 
 a-naphthol solution with 10 drops of water and 1 c.c. of the sulphuric- 
 acid, when the mixture should be of a golden yellow colour. If it 
 is dark green the reagents are not sufficiently piire. The naphthol 
 solution does not keep well, and should be prepared as required. 
 Most albumens give the violet coloration with Molisch's reaction, 
 owing to the presence of a carbohydrate gToup, but do not give the- 
 violet and rusty brown precipitates. Casein does not react with 
 Molisch's test. 
 
 Phloroglucine Test. — Shake an excess of phloroglucine with a 
 mixture of equal parts of concentrated hydrochloric acid and water 
 until the solution is sattu-ated. Boil 3 c.c. of this reagent with about 
 0"03 gram of the carbohydrate in a- small test-tube. Xote the colour 
 when it just commences to boil. Continue to boil until the colour 
 darkens considerably and the solution begins to appear slightly 
 
APPENDIX 427 
 
 tixrbid, usually within about one minute. Poiu" the hot solution, 
 without delay, on to a wet filter -paper, and rinse the scanty pre- 
 cipitate with a little cold dilute alcohol. Note the colour of the 
 precipitate while moist. With arabinose and xylose the first colora- 
 tion on heating is a pure red to violet-red, but it rapidly intensifies 
 and darkens as the heating is continued. The coloiu" of the pre- 
 cipitate varies, according to the duration of the boiling, from very 
 dark purple to black, if the heating has been continued too long. 
 If the precipitate is dissolved in alcohol, and the solution is examined 
 with the spectroscope, an absorption band is seen in the green should 
 a pentose be present. With fructose, rhamnose, and sorbinose the 
 first coloration is a yellow -orange, which quickly passes through 
 dark orange to dingy brown. The precipitate is of a rusty brown, 
 or dark shade of yellow-orange, or orange, which may be easily 
 changed to a dull black if the boiling is too long continiTed. Galac- 
 tose and lactose give a similar colour change, but on spectroscopic 
 examination they do not show the sharp absorption bands. Solu- 
 tions of glucuronic acid give a similar colovir reaction, and show the 
 same absorption bands as the pentoses. 
 
 Orcin Test. — About 0-3 to 0-4 gram of the carbohydrate, 2 or 3 e.c. 
 of piire concentrated hydrochloric acid, and a few milligrams of 
 orcin are mixed in a test-tube, and gently heated. At first a faint 
 yellow tint is seen, but in a few seconds a violet-blue coloiu" appears 
 if a pentose is present, while the methylpentoses and hexoses give 
 an orange-red coloration. On continuing the heating the colour 
 intensifies, and finally a precipitate appears. If the contents of the 
 test-tube are now cooled somewhat, and extracted with amyl alcohol, 
 the extract on being examined with the spectroscope shows a distinct 
 band between C and D (red and yellow), with often a second band 
 in the red if pentoses are present. A band in the green may be seen 
 if too much orcin has been used. Other sugars show no absorption 
 bands, but gluciironic acid gives the same results as a pentose. If 
 the solution is heated too rapidly the characteristic colour may be 
 obscTU"ed by the darkening of the liquid, and by partial destruction 
 of the sugar, with the formation of brown himious substances. This 
 reaction also takes place at ordinary temperatiires, but only after 
 standing for several hoiu-s. If the hydrochloric acid contains iron 
 the pentoses give a gxeen, instead of a violet-blue, coloration. 
 
 Aniline Acetate Test. — Dissolve about 0*3 to 0-4 gram of the 
 carbohydrate in 5 c.c. of dilute hydrochloric acid (1 vol. of HCl, 
 sp. gr. 1-2, with 3 vol. of water). Boil for one minute in a test-tube. 
 Insert a roll of thick filter-paper which has been soaked in mixture 
 of 5 c.c. of aniline and 10 c.c. of 50 per cent, acetic acid and pressed 
 between blotting-paper luitil only just moist, into the upper inch 
 or so of the test-tube, and continue the boiling. Arabinose, xylose, 
 fructose, rhamnose, and sorbinose give sufficient furfurol when 
 treated in this way to tm-n the test-paper a bright pink. The other 
 carbohydrates do not bring about any noticeable coloration. Xyli- 
 dine acetate may be substituted for aniline acetate in this test, and 
 is somewhat more sensitive. 
 
 Resorcin (Seliwanoff's) Test. — All carbohydrates when heated 
 with resorcin and strong hydrochloric acid give a red coloration, due 
 to the formation of fiirfvirol and its condensation by the resorcin, 
 but if the hydrochloric acid is diluted with its own volume of water 
 the reaction is only given by the ketoses. On this fact is based 
 the reaction of Seliwanoff which distingviishes the ketoses, and 
 
428 GLYCOSURIA 
 
 particularly le^mlose, from the aldoses. A solution of a few milli- 
 grams of resorcin in diluted hydrochloric acid (1 vol. strong HCl 
 to 2 vols, of water) is warmed with a few milligrams of the carbo- 
 hydrate, when, if levulose, or another ketose, is present a beautiful 
 red colour appears, and a red precipitate settles on standing. If 
 the solution is neutralised with sodium carbonate and extracted 
 with amyl alcohol, the alcohol takes on a red-green fluorescence, 
 which, on the addition of a little absolute alcohol, becomes pure 
 red. Examined with the spectroscope a band is seen between E 
 and b, and in higlily concentrated solutions a second band in the blue 
 near F. 
 
 Buhner's Test. — When a solution of a reducing sugar is boiled 
 for several minutes with a solution of acetate of lead and ammonia, 
 a coloiir, varying from yellow to copper red, is produced. Under 
 certain conditions the particular colour is characteristic, and may 
 be used to distinguish certain sugars, and particularly lactose. It 
 is in the first place essential that the sugar solution should not be 
 too strong (0-5 to 1 gram per litre) ; secondly, too great heat should 
 not be applied or a non- characteristic brown colour develops ; and 
 thirdly, an excess of ammonia should be avoided, as it ruins the test. 
 
 To 10 c.c. of a solution of lactose of a strength of about 0-1 per 
 cent, add about 1 gram of crj^stallised acetate of lead, and heat 
 gently to dissolve the acetate. Now add ammonia drop by drop, 
 shaking after each addition. The precipitate formed at first dis- 
 solves, but finally persists. The addition of ammonia should be 
 continued until the liquid is distinctly tm-bid (usually about 1 to 
 2 c.c. of ammonia are necessary). The mixture is now boiled for 
 two or three minutes, when a rose to orange colour develops. On 
 standing for a few seconds a bright rose-coloiored precipitate settles 
 down, and the supernatent flmd appears orange to rose-coloured. 
 
 On carrying out the test with a solution of dextrose of the same 
 strength, a wliite to yellowish precipitate is formed, and the fluid 
 appears a clear yellow. With more concentrated solutions of the 
 sugars it is necessary to increase the proportion of lead acetate, 
 and consequently of ammonia. If the solution after the addition 
 of ammonia is only warmed to 80° C. in a water -bath, dextrose gives 
 a red solution -nith a rose or sahnon-pink precipitate, lactose a yellow 
 coffee brown or red coloration but no precipitate according to the 
 concentration, maltose a slight yellow colour, and levulose no coloiir 
 at all. 
 
 Combinations with Hydrazines. — Phenylhydrazin, and the substi- 
 tuted phenylhydrazins — 
 
 Phenyl-hydrazin 
 Methyl-phenj'lhydrazin 
 Benzyl-phenylhydrazin 
 Di-phenylhyd razin 
 Para-brom-plienylhydrazin 
 
 CcH5.NH.NH2 
 
 CoH5(CH3)N.NH., 
 C„H5(C,H,)N.NH., 
 CcH5(CoH5)N.NH2 
 CcH5(CcHiBr)N.NH„ &c. 
 
 form two series of compounds with sugars containing an active alde- 
 hyde, or ketonic, group — (1) the phenylhydrazones, in which one 
 molecule of sugar combines with one molecule of the hydrazin ; and 
 (2) the phenylosazones, in which one molecule of the sugar combines 
 with two molecToIes of the hydrazin. Owing to the varying physical 
 
APPENDIX 429 
 
 and chemical properties of these conipoiinds in the case of different 
 sugars, they furnish a most important means for separating and 
 identifying them. 
 
 1. The hydrazones are prepared by digesting the sugar, dissolved 
 in not too much water, with the calculated quantity of phenylhydra- 
 zin, dissolved in its own weight of acetic acid, in the cold. The 
 amount of phenylhydrazin required is deduced from the results of an 
 approximate titration of the sugar solution. The most readily pre- 
 pared is mannose phenylhydrazone, which is formed by allowing for 
 each 180 grams of the sugar 108 grams of the base : — 
 
 CH20H.(CHOH)4.CH0 + NH2.NH.C6H5=::CH20H.(CH0H)4.CH:N - NH - C6H5+ HoO 
 
 Preparation of Mannose -phenylhydrazone.- — To 50 c.c. of a solu- 
 tion of mannose, of approximately 2 per cent, strength, add 6 to 
 7 c.c. of a solution containing 10 grams of phenylhydrazin and 
 10 c.c. of glacial acetic acid made up to 100 c.c. with water, and 
 shake. The liqiiid becomes cloudy and forms a white deposit of 
 the phenylhydrazone, which, under the microscope, is seen to consist 
 of sphsero-crystals. They have a melting-point of 186° to 188° C. 
 
 The simple phenylhydrazones of the other sugars, with the excep- 
 tion of the rare sugars rhamnose and fukose, are readily soluble in 
 water, and cannot therefore be easily prepared. 
 
 The substituted phenylhydrazins form hydrazones which are gene- 
 rally more insoluble than the corresponding simple hydrazones. They 
 are therefore of considerable value in the detection and separation of 
 the sugars. 
 
 Methyl -phenylhydrazin (CpH5(CH3).N.NH2) is chiefly useful in the 
 recovery and separation of galactose, with which it forms an insoluble 
 crystalline compound, melting at 180° C. 
 
 Benzyl-phenylhydrazin (CQH5(CyIT-).N.NH2) on being heated with 
 a 96 per cent, alcoholic solution of the sugar gives a hydrazone which 
 can be recovered by evaporating and recrystallisation from alcohol. 
 Dextrose yields a benzyl -hydrazone which is levo -rotatory, and has a 
 melting-point of 165° C. (171° to 172°). It is decomposed by boiling 
 water into its constituents, dextrose and benzyl-phenylhydrazin. 
 Levulose gives a similar compound, which is not, however, decomposed 
 by boiling water, thus affording a means of differentiating the two 
 sugars. L-arabinose and galactose both give benzyl -hydrazones, the 
 former being insoluble in alcohol and having a melting-point of 170° 
 to 174° C, the latter being feebly soluble and having a melting-point 
 of 154° C. 
 
 Di -phenylhydrazin (C'eH5(CgH5)N.NH2) is only feebly soluble in 
 water, so that in preparing the hydrazone it is necessary to dissolve 
 the reagent in alcohol. This solution is heated with the sugar for two 
 hours, or left in the cold for two to three days. The di -phenylhydrazone 
 of r-arabinose melts at 204° to 205° C, the xylose compoiuid at 107° to 
 108° C, the dextrose at 161° to 162° C, mannose at 155° C, galactose 
 at 157° C. Levulose does not yield a crystallisable hydrazone with 
 di -phenylhydrazin. 
 
430 
 
 GLYCOSURIA 
 
 Para-hrom-phenylhydrazin (C6H5(CeH4Br).]S!'.NH2) is a most im- 
 portant reagent for distingiiishing arabinose from xylose and dextrose. 
 L-arabinose forms a very feebly soluble para-brom-phenylhydrazone 
 with a melting-point of 162° C, whereas the other two sugars do not 
 yield an insoluble crystallisable hydrazone. Glucuronic acid also reacts 
 with para-brom-phenylhydrazin to form a crystalline compound, in- 
 soluble in alcohol, which in the pure condition melts at 236° C, but 
 when freshly prepared from the urine melts at 200° to 216° C. 
 
 Melting-points of the Hydrazones 
 
 
 2 
 1 
 
 2 
 
 
 1 
 
 P 
 
 C 
 > 
 
 o 
 
 <u 
 o 
 
 1 
 
 "a 
 O 
 
 0) 
 
 "a 
 130 
 
 o 
 1 
 
 Phenylhydrazone 
 
 150-153 
 
 
 
 144-146 
 
 144-146 
 
 186-188 
 
 158 
 
 Methyl-phenylhydrazoiie . 
 
 161-164 
 
 173 
 
 108-110 
 
 130 
 
 116-130 
 
 178 
 
 180-188 
 
 
 
 Benzyl-phenylhydrazone . 
 
 170 
 
 185 
 
 93 
 
 165 
 
 165 
 
 165 
 
 154-158 
 
 
 128 
 
 Di-phenylhydrazone . 
 
 216-218 
 
 204-205 
 
 107-108 
 
 161-162 
 
 
 155 
 
 157 
 
 
 
 Para-brom-phenylhydrazone 
 
 160-162 
 
 160 
 
 128 
 
 147 
 
 147 
 
 208-210 
 
 168 
 
 
 
 The hydrazones are important, not only because of their physical 
 characters, which lend themselves to the differentiation of the sugars, 
 but also from the fact that when they are treated with certain reagents 
 they are decomposed into the sugar and the hydrazin from which they 
 were derived. Thus a phenylhydrazone treated with concentrated 
 hydrochloric acid yields phenylhydrazin hydrochloride, and the sugar. 
 After the excess of hydrochloric acid has been neutraHsed with lead 
 carbonate, the sugar can be recovered from the solution. It is better, 
 however, to bring about the decomposition with benzaldehyde. 
 
 The hydrazone is placed in a flask, provided with a reflux con- 
 denser, mixed with about a quarter more benzaldehyde than is 
 theoretically required to bring about the decomposition, and four 
 times the weight of a mixture of equal parts of water and strong 
 alcohol, and heated on a water-bath for a half to one hour. It 
 is then cooled, and filtered from the precipitated benzaldehyde 
 phenylhydrazin. The filtrate is shaken with 2 volumes of ether, to 
 remove the excess of benzaldehyde, decolorised with animal char- 
 coal, and the sugar crystallised out. 
 
 CoHjA : N.NH.CoH5-hC6H5COH = C6H5CH : N.NH.CgHs + CoHiPe 
 Mannose phenyl- Benzal- Benzaldehyde Mannose. 
 
 hydrazone. dehyde. phenylhydrazin. 
 
 The substituted phenylhydrazins are similarly decomposed by 
 formaldehyde. To bring about the decomposition the hydrazone 
 is mixed with 35 per cent, formaldehyde and a little alcohol, to aid 
 solution, and heated in a flask, provided with a reflux condenser, on 
 
APPENDIX 431 
 
 a water-bath, for about an hour. It is then cooled, extracted with 
 ether several times, decolorised with animal charcoal, and evapo- 
 rated, with the addition of a little water, until the excess of for- 
 maldehyde is driven off. The sugar is then crystallised out. 
 
 C5Hio04N.N(CV,H5) + CH^O = CH2.N.N(C6H5)2 + C5H10O. 
 Arabinose de- Formal- Formaldehyde Arabinose. 
 phenylhydra- dehyde. de-phenylhy- 
 zone. drazin. 
 
 Osazones. — All the reducing sugars give with phenylhydrazin, in 
 the presence of an excess of phenylhydrazin acetate, more or less 
 insoluble yellow crystallisable compounds to which the name osazones 
 has been given. Two molecules of the base react with one molecule 
 of the sugar to form the osazone, hydrogen being liberated at the 
 same time. The latter is not, however, set free, but reacts with the 
 excess of phenylhydrazin, giving rise to aniline and ammonia. Thus 
 with dextrose : — 
 
 CgHiaOe + 3N Hg. NH.CgHg = CgHjoO^. (N.NH.CeHg)^ -I- 2 H^O + CgHjN Hg -^ NH3 
 Dextrose. Phenylhydrazin. Dextrosazone. Water. Aniline. Ammonia. 
 
 This reaction takes place in the cold, but occurs much more quickly 
 on heating. 
 
 Preparation of Qlucosazone. — Heat in a water-bath for one hour 
 a mixture of 50 c.c. of a solution of dextrose, about 1 per cent., 
 with 10 c.c. of a solution consisting of phenylhydrazin 10 grams, 
 glacial acetic acid 10 c.c, and water to 100 c.c. Cool, and pour on 
 to a moist filter. Wash the crystalline deposit with cold distilled 
 water, and afterwards with methyl alcohol several times, using 
 about 20 c.c. Dry by pressing between clean filter-paper, and 
 take the melting-point. This procedure is only applicable to those 
 osazones which are feebly soluble in methyl alcohol, such as gluc- 
 osazone and maltosazone. Those which are more soluble in water 
 and organic solvents, like maltosazone and lactosazone, must be puri- 
 fied by dissolving in boiling water and recrystallising several times. 
 
 The original method of preparing the osazones described by 
 H. Fischer [Ber. d . deutsch. chem. Ges., xvii. 579, 1884) may also be 
 followed with the pure sugars. In this 1 gram of the sugar, 2 grams 
 of phenylhydrazin hydrochloride, 3 grams of crystallised sodium 
 acetate, and 20 grams of water are mixed together, and heated in a 
 boiling water-bath for three-quarters to one and a half hours. 
 
 The rate at which osazone formation takes place is very different 
 for the various sugars. Working with the pure substances and a 
 definite mode of procedure advantage may be taken of this fact to 
 roughly differentiate them (Marquenne). (See Mulliken, Identification 
 of Pure Organic Compounds, p. 32.) 
 
 Place in a dry test-tube, having an internal diameter of 13 mm., 
 0"1 gram of the sugar, 0*2 gram of pm-e phenylhydrazin hydro- 
 chloride, 0-3 gram crystallised sodium acetate, and 2 c.c. of water. 
 Close the tube loosely with a cork, to prevent evaporation, and 
 stand in a tall narrow beaker containing two or three inches of 
 water that is already briskly boiling. Note the exact moment of 
 
432 GLYCOSURIA 
 
 immersion. Shake the tube occasionally, without, however, re- 
 moving it from the beaker. If a precipitate finally separates, note 
 the niimber of minutes that have elapsed up to the moment of its 
 appearance. Note also whether the precipitate is white, yellow, 
 or orange-yellow, and whether it is crystalline, flocculent, or tends 
 to rise to the surface in oily drops. 
 
 Under the above-mentioned conditions the monosaccharides (pen- 
 toses and hexoses) all give precipitates which separated out from the 
 hot liquid in a half to twenty minutes. Mannose gives a nearly 
 white crystalline deposit in a half minute ; levulose a. yellow pre- 
 cipitate in two minutes ; dextrose a yellow osazone which separates 
 out suddenly in four to five minutes ; xylose a light yellow osazone in 
 seven minutes ; arahinose an orange-yellow precipitate, which may, 
 however, appear partly as brownish-yellow oily drops unless the sugar 
 is very pure, in ten ixiinutes ; and galactose a yellow to orange-yellow 
 osazone in fifteen to nineteen minutes. Of the di-saccharides some, 
 like maltose and lactose, give products which do not separate out iintil 
 the hot solution has cooled, while others like saccharose are gradually 
 hydrolised to monosaccharides, which then give the corresponding 
 osazones, but naturally require a longer time for the reaction than when 
 the simple sugar was originally present. Saccharose begins to show a 
 yellow osazone formation after about thirty minutes' heating. With 
 raffinose no osazone separates out until the solution has been heated 
 for about sixty minutes. The precipitates are all phenylosazones, ex- 
 cept that from mannose, which is a simple hydrazone and is easily 
 distinguished from the others by being white instead of yellow. 
 Variations of a minute or two from the times stated will occasionally 
 occur, and must be allowed for when selecting between two species of 
 sugar whose values lie close together. 
 
 The osazones are all more or less insoluble in cold water, and some 
 are only feebly soluble in hot water. They are easily soluble in hot 
 60 per cent, alcohol, but can be recrystallised out on boiling off the 
 alcohol. They are also soluble in hot acetic acid. They are insoluble 
 in ether, chloroform, benzol, and ligroin, but are soluble in pyridine, 
 and a mixture of pyridine and alcohol. 
 
 The osazones generally crystallise well. They present considerable 
 variations in form, which, to a certain extent, is dependent upon the 
 conditions under which crystallisation occurs, but under the same 
 conditions they present forms which are more or less characteristic. 
 Examined under the microscope with a magnification of about oOO 
 diameters, glucosazone crystals are seen as sheaves and bundles of 
 yeUowish-green crystals. Levulosazone crystals are similar, but are 
 often somewhat coarser. Lactosazone appears as long, narrow, yellow 
 plates, free or grouped in various shapes. Arabinosazone forms long,. 
 thin, hair-like needles which are curved and bent. Xylosazone is seen 
 in long, straight needles. Maltosazone appears as short, leaf -like crystals- 
 which are very often grouped in rosettes. 
 
 The melting-point of the osazones of the different sugars is a con- 
 
APPENDIX 433 
 
 statit physical character which assists in their differentiation. It has 
 not, however, the great importance that is sometimes attributed to it, 
 especially in the case of osazones which melt at nearly the same tem- 
 perature. Many osazones undergo decomposition in the neighbour- 
 hood of their point of fusion, with the result that the melting-point 
 is lowered by the products of decomposition, and the more abundant 
 these are — that is to say, the longer the heating is continued — the 
 lower the melting-point appears to be. A constant reading, therefore, 
 is only obtained when the determination is carried out as rapidly as 
 possible, and at the minimimi temperature for which the fusion-point 
 is instantaneous. The presence of a very small quantity of impurity 
 also lowers the melting-point very markedly. 
 
 Purification. — As a general rule the osazones may be purified by 
 filtering off from the mother-liquor, well washing with cold distilled 
 water, removing the excess of moistm-e by pressing between filter- 
 paper, or by means of a suction-pump, dissolving in hot 60 per cent, 
 alcohol, driving off the excess of alcohol with heat, and crystallising 
 out. The osazone is then filtered off, and dried over sulphuric acid, 
 or in an oven at a low temperatvire. 
 
 Melting-point Determination. — The usual method of determining 
 the melting-point is to place a few milligrammes of the substance 
 in a thin-walled glass capillary tube, sealed at the lower end, about 
 6 to 7 cm. long and with an internal diameter of about 1 mm. This 
 is attached to a delicate thermometer by a spiral of fine platinimi 
 wire in such a way that the substance in the capillary tube is situated 
 opposite the middle of the thermometer bulb. The thermometer is 
 introduced into a small test-ttibe, with a diameter of about 15 mm., 
 containing sufficient colourless sulphuric acid (sp. gr. 1-84) to cover 
 the bulb. The test-tube is suspended by its flanged lip in a small 
 round-bottomed flask with a bulb of about 65 mm. diameter, a 
 neck 75 mm. long and a 20 mm. diameter, and a total capacity 
 of about 200 c.c, which contains sulphiu-ic acid up to the level of 
 that in the test-tube. The flask is heated rapidly to within a few 
 degrees of the approximate melting-point of the osazone, which 
 can be determined by a trial experiment if necessary. It is then 
 heated more cautiously, and the temperature just before coinplete 
 liqviefaction noted as the melting-point. It is a common practice 
 to give as the melting-point the temperature at which the first 
 drop of sufficient size to detach itself from the solid mass appears. 
 Differences of several degrees between melting-points obtained by 
 different methods of observation are therefore possible. 
 
 A method which is much used in France is the " bloc Maquenne." 
 This is a block of metal heated by a gas flame, and provided with 
 a thermometer which registers its temperature. The block is 
 rapidly heated by the gas flame so as to have a rise of 3° to 5° C. 
 per minute. At each rise of 5° C. a small Cjuantity of the dry pow- 
 dered osazone is thrown, on to the surface of the block, and the 
 temperature at which instantaneoiis fusion occurs is noted. The 
 block is then cooled to a few degrees below this point, and gently 
 reheated so as to have a rise of about 1° C. every three to four minutes. 
 At each degree a small quantity of the osazone is again thrown on 
 to the block, and the exact melting-point noted. This method 
 gives readings which it is to be noted differ considerably from those 
 obtained by the classical capillary tube method. Thus glucosazone 
 
 2e 
 
434 GLYCOSURIA 
 
 melts at 230° to 232° C. as compared with 204° to 205° C. by the 
 ordinary method, and lactosazone at 213° to 215° C. as compared with 
 200° C. ^"NTien considering the melting-point of an osazone it is 
 obvious that, unless the product was quite pure and the exact 
 method of determination is taken into account, the results may 
 prove most misleading. 
 
 Solutions of the osazones are for the most part optically active, 
 and this property serves for the further identification and separation 
 of the sugars. If 0*2 gram of the pure osazone is dissolved in a mixture 
 of 4 grams of pyridin and 6 grams of absolute alcohol, and examined 
 in a 100 mm. tube with the polariscope, the following readings, accord- 
 ing to Neuberg {Ber. d. deutsch. chem. Ges., xxxii., 579, 1884), are 
 obtained :— 
 
 r 1-arabinose phenylosazone ...... +1*10'' 
 
 I do. para-brom-phenjlosazone .... +0"28'' 
 
 Pentoses . '. l-xylose phenylosazone -0"15° 
 
 I do. para-brom-phenylosazone .... +000° 
 
 1^ Khamnose phenylosazone ...... +1'24° 
 
 [ d -glucose phenj'losazone ........ 1".30° 
 
 do. para-brom -phenylosazone .... — 0"31° 
 
 Hexoses . - fructose phenylosazone ........ 1*30° 
 
 mannose phenylosazone ....... -I'SO" 
 
 1^ d-galactose phenylosazone +0'48° 
 
 „. , ., f maltose phenylosazone . ...... -1-1 "SO" 
 
 Uisaccnanaes-^ lactose phenylosazone ±0-00° 
 
 Glucuronic acid para-brom-phenylhydrazin -7'25° 
 
 The percentage of nitrogen contained in the pure osazone affords 
 another means of differentiating the groups of sugars from each other ; 
 thus the osazones of the pentoses yield 17-7 per cent, of nitrogen, the 
 hexoses 15-7 per cent., the disaccharides 10-7 per cent., and the glucuronic 
 acid compound of phenylhydrazin 16-4 per cent. The osazones must 
 be very carefi.illy purified, however, and the nitrogen be determined 
 by Dimias' method, as the Kjeldahl process does not give satisfactory 
 resvilts. 
 
 When treated with strong hydrochloric acid the osazones are de- 
 composed with the separation of phenylhydrazin hydrochloride, and 
 the formation, not of the sugar, biit substances known as osones con- 
 taining the group - CO.CHO : — 
 
 C4H.(OH)4.C(N,HC„H5).CH.X2HC6H5 4-HCl -f 2H2O = 
 C4H5(OH)4.CO.CHO-f2C6H5NH.NH2,HCl " 
 
 The osones may be reduced to sugars by the aid of zinc dust and 
 acetic acid, but a given osazone does not necessarily yield the sugar 
 from which it was derived ; thus glucosazone, mannosazone, and 
 fructosazone all give rise to levulose : — 
 
 C4H5(OH)4.CO. CHO + 2H = C4H.(OH)4CO.CH2.0H 
 
 showing that they are not three separate osazones, and thus accounting 
 for their identical physical properties. 
 
APPENDIX 
 
 435 
 
 Glucuronic 
 acid. 
 
 W 5-1 
 
 " IM to 
 
 Ju J. '■ ■ C^ 
 — 1 0-) 
 
 6 
 
 o 
 
 1-4 
 
 (M . . . 
 C-l 
 
 1 
 
 1 
 
 c 
 
 ^ ■ • • 
 
 1 
 
 o 
 
 <M . . . 
 
 1 ■ ■ ■ 
 
 •* • • • 
 
 c 
 
 o 
 >■ 
 
 lO o 
 
 o o 
 
 Cl rt O c-1 
 1 1 c: 01 
 ■* CO T-l M 
 C O 
 
 1 
 
 C O 
 
 II • <M 
 
 — •* 
 CI ^ 
 
 >< 
 
 c 
 1^ 
 
 —1 . 00 
 
 <M ■ ■ j5 
 
 0) 
 
 o 
 n 
 
 =0 (M 
 ^ . . C-l 
 
 1 • ■ 1 
 
 C= • • C 
 
 « o 
 
 o 
 c 
 
 cs 
 
 p 
 
 o . . S5 
 
 o . . 1 
 
 l-H • • tS 
 
 C5 
 
 
 01 
 
 c 
 o 
 
 • S i> f= 
 
 O o ^ 
 
 • J 2 £ 
 a; ^C ^c f>. 
 
 g ^' t^ p^ 
 ? S S p 
 -a ^ a, P 
 'H'- "^^ ">. ■? 
 
 Oh g in PL, 
 
 Monosaccharides. — The general 
 characters of the monosaccharides, 
 and of the pentoses and hexoses, have 
 already been considered. The pro- 
 perties of the individual sugars will 
 now be briefly dealt with. 
 
 Pentoses. — Arabinose (Pectionse) is 
 most readily prepared from gum- 
 arabic, or cherry-gum, by heating it on 
 a water-bath with 2 per cent, sulphuric 
 acid for ten to fifteen minutes. The 
 product thtis prepared is the dextro- 
 rotatory 1-arabinose. The form usually 
 met with in the iirine is the racemic, 
 or optically inactive, r- arabinose. The 
 former crystallises out as glancing 
 bright needles, or prisms, which melt 
 at 150° C, and, in solution, deflects 
 polarised light to the right (aD = 
 + 104° to 105°). The latter crystal- 
 lises in rhombic plate, or hard prisms, 
 and melts at 164° to 165° C. Its solu- 
 tion has no action on polarised light. 
 
 Arabinose has a sweet taste, is 
 easily soluble in water, but is only 
 soluble with difficulty in alcohol, and 
 is insoluble in ether. It reduces 
 Fehling's and Nylander's solutions 
 somewhat better than xylose, 10 c.c. 
 of Fehling's solution being reduced by 
 43 mg. of arabinose. With phenyl- 
 hydrazin it yields, after prolonged 
 warming, an orange-yellow crystalline 
 osazone, which, however, often separ- 
 ates out in the form of oily bro-miish- 
 yellow drops, unless the sugar is very 
 pixre. Microscopically the osazone is 
 seen to consist of sheaves of fine yellow 
 flexible crystals, which when irrigated 
 with 33 per cent, sulphuric acid dis- 
 solve, and disappear, a few seconds 
 after the acid reaches them. The 
 purified product melts at 157° to 158° 
 C, but as prepared from the urine it 
 more often melts at about 150° to 
 156° C. It contains 17-07 per cent, of 
 nitrogen theoretically, but in practice 
 the nitrogen content generally works 
 
436 GLYCOSURIA 
 
 out at about 17-01 per cent. Alcoholic solutions of the osazone are at 
 first dextro-rotatory (aj,= + 18-9°.), but in a few hoiirs become optically- 
 inactive. With di-phenylhydrazin it gives coloiu'less needles, wliich are 
 insoluble in alcohol and cold water, but are soluble in glacial acetic acid 
 and pyridin, and melt at 204° to 205° C. Unlike xylose it yields a feebly 
 sohible hydrazin compound with para-brom-phenylhydrazin, which 
 melts at 200° to 202° C. Arabinose yields arabonic acid on oxidation. 
 
 I-xylose {wood-sugar) is prepared by the action of dilute acids on. 
 wood, gum, or straw. It can also be prepared from the nucleo- 
 proteid of the pancreas, liver, and other organs of the body. It- 
 crystallises out in white needles or prisms, has a sweet taste, is soluble- 
 in water and hot alcohol, but is insoluble in cold alcohol and ether. 
 The pure product melts at 153° to 154° C. It is dextro-rotatory 
 (aj,= -fl8"l°), but shows strong mutarotation, the deviation depend- 
 ing upon the concentration of the solution. Like arabinose it is 
 not fermented by yeast, but reduces Fehling's (10 c.c. = 45-6mg. xylose) 
 and Nylander's solutions, and gives an orange precipitate with Rubner's. 
 test. With phenylhydrazin it gives an orange yellow, crystalline 
 osazone, consisting of long fine needles, which are readily soluble in 
 33 per cent, sulphuric acid. The osazone is easily soluble in alcohol 
 and less easily in acetone. Its solution in alcohol is strongly levo- 
 rotatory (a^ = —43-4°), and a solution in 4 per cent, acetic acid 
 is also levo-rotatory (aD= - 1'3°), by which it is distinguished from 
 the osazone of arabinose. On being rapidly heated the osazone melts 
 at 159° to 160° C. With di-phenylhydrazin it forms a hydrazone that 
 melts at 107° to 108° C. Xylose does not give a crystalline hydrazone 
 with para-brom-phenylhydrazin. It is also distingxiished from ara- 
 binose by forming xylonic acid (QH^QOg) on being oxidised with 
 broinine. This can be separated out as a characteristic insoluble double 
 salt of cadmivim and bromine, by treating the product with cadmiunx 
 carbonate. With brucine it forms a salt which melts at 172° to 174° C, 
 
 'S.Q'^OSQ&.-^D -glucose, dextrose, or grape-sugar is met with in asso- 
 ciation with levulose (fructose) in the juices of grapes, and other 
 ripening frmts. The two hexoses are probably derived by hydrolysis- 
 from pre-existing cane-sugar, with which they usually occur. Dex- 
 trose can also be derived from other di- saccharides and polysaccharides,, 
 such as lactose or milk-sugar, maltose or malt-sugar, starch, and. 
 cellulose, by the action of dilute acids or appropriate ferments. In 
 the animal body it occurs in small quantities in the blood and lymph,, 
 and in minimal traces in nonnal urine. It is the only hexose which 
 does so exist normally ; levulose, mannose, and galactose on reaching 
 the liver being transformed, with dextrose, into the polysaccharide 
 glycogen. The urine of diabetics is characterised by a more or less- 
 marked increase of dextrose. 
 
 Dextrose separates from aqueous solutions with one molecule of 
 water of crystallisation, but this is only loosely held, as the anhydrous- 
 substance may be crystallised from dilute alcohol. Unlike cane-sugar- 
 
APPENDIX 437 
 
 it never separates in well-defined crystals, but is usually met with as 
 a crystalline powder. It is about half as sweet as cane-sugar. It is 
 dextro-rotatory ( + 52° 7' at 20° C), but a freshly made solution is 
 more markedly dextro-rotatory (mutarotation). It is fermented by 
 yeast within twelve to fifteen hours at 20° to 30° C, forming princi- 
 pally alcohol and carbon dioxide, but traces of fusel-oil, glycerol, 
 succinic acid, &c., also appear. Heated with alkalies (Moore's test) 
 a solution of dextrose turns brown, forming acetone, acetic, lactic, 
 formic acids, &c. It reduces alkaline solutions of copper (Trommer's 
 test, Fehling's test, &c.), bismuth (Nylander's test), and also acetic 
 acid solutions of copper acetate (Barfoed's test). One molecule of 
 dextrose always reduces exactly the same quantity (approximately 
 5 molecules) of cupric to cuprous oxide, a property which is generally 
 tised as the basis for estimating the quantity present in a given solu- 
 tion. With phenylhydrazin it forms a soluble hydrazone and an in- 
 soluble osazone, which suddenly separates out from the hot solution, 
 after four or five minutes' heating, if at least 5 per cent, is present. 
 The osazone forms a yellow precipitate, which, on microscopical exa- 
 mination, is seen to consist of coarse greenish-yellow crystals, and 
 which on being irrigated with 33 per cent, sulphuric acid do not readily 
 dissolve. The melting-point of the pure dry product is 204° to 205° C, 
 but the crude preparation, as obtained from the urine, usually melts 
 between 173° and 194° C. The pyridine solution of the osazone is 
 levo-rotatory ( — 1° 30'). The nitrogen content, by Dmnas' method, 
 is 15-7 per cent. With para-brom-phenylhydrazin it gives an osazone 
 (melting-point 222° C), with methyl -phenylhydrazin a hydrazone 
 (melting-point 130° C), and also hydrazones with benzyl -phenyl- 
 hydrazin (melting-point 165° C), and di -phenylhydrazin (melting- 
 point 161° C). Dextrose does not give the aniline acetate and other 
 common tests for furfiirol under ordinary conditions (c/. the pentoses). 
 On oxidation with nitric acid it yields saccharic, but not mucic acid 
 (c/. galactose). Oxidised with bromine it gives gluconic acid. 
 
 D-glucosamine or aminoglucose (CgHjoOjNHg) was the first well- 
 defined carbohydrate compound isolated from animal tissues (Ledder- 
 hose, 1878). It is of some physiological interest, as it is a constituent of 
 mucins and mucoids, and, with glycvironic acid, enters into the com- 
 position of the chondroitin sulphuric acid of cartilage. It is most readily 
 prepared by the action of concentrated hydrochloric acid on the chitin 
 in lobster shells, or fungus cellulose, which gives the hydrochloride. 
 
 D -fructose or Icevidose is found with dextrose in fruit juices, 
 honey, &c., the mixture being termed fruit- or invert-sugar. Com- 
 bined with dextrose it occurs in cane-sugar, rafflnose, &c., from which 
 it can be prepared by the action of dilute acids and ferments. The 
 polysaccharide inulin yields fructose alone when hydrolysed. It is 
 met with in pathological urines, exudates, and transudates, rarely 
 alone, generally with dextrose, and in normal and diabetic blood. 
 The form met with in diabetic urine differs from plant levulose in 
 being precipitated from its solution by basic lead acetate. 
 
438 GLYCOSURIA 
 
 Lfe\T.ilose crystallises less easily than dextrose, in long, fine, hygro- 
 scopic needles, or crusts, from alcohol, and in needles from water. 
 It is very soluble in water, is soluble in five parts of cold absolute- 
 alcohol, and, unlike other sugars, it is soluble in ether. It tastes 
 about as sweet as cane-sugar. With calcium it forms a feebly soluble, 
 coloiu-less compound (CgHj20g.Ca(OH}2), by which it can be separated 
 from the more soluble dextrose salt. It is levo -rotatory, and exhibits^ 
 ixiutarotation, but is remarkable for the very large change produced 
 in the specific rotatory power by alterations in temperature. The- 
 rotatory power falls {i.e. becomes less negative) as the temperature is 
 increased. At 20° C. a 10 per cent, solution shows a rotation of -90°' 
 to — 92°, which is rather more to the left than dextrose is to the rights 
 but at 82° C. it is equal and oppositive to that of dextrose. It gives- 
 the same reduction reactions as dextrose, but rather more rapidly. 
 Half a gram of levulose in 1 per cent, solution reduces 97*2 e.c. of 
 Fehling solution. With phenylhydrazin it forms an osazone with 
 the same melting-point, nitrogen content, &c., as dextrose. Like 
 other ketoses it gives with methyl -phenylhydrazin a characteristic 
 compound, consisting of yellow needles that melt at 158° to 160° C. 
 Tliis osazone is insoluble in cold water, but is soluble in hot alcohol. 
 A pyridin-alcohol solution is dextro-rotatory (0-2 gram in 10 c.c. = 
 + 1-40°). Le\ailose gives no crystalline compoiuid with di-phenyl- 
 hydrazin (c/. dextrose). Like dextrose, it is easily fermented by 
 brewer's yeast -within twelve to fifteen hours at 37° C, and forms the- 
 same products. Like other ketoses, it yields furfurol easily and in 
 large quantities on being heated with hydrochloric acid, and on this 
 fact is based Seliwanoff's test, in which the furfurol formed is recog- 
 nised by the coloiu' reaction it gives with phloroglucin. Arabinose and 
 xylose give a similar reaction, but, while the first coloration with 
 le\Tilose is a yellow -orange, that c[iiickly passes tlirough dark orange- 
 to a dingy brown, the pentoses give on first heating a pure red to- 
 violet -red, which rapidly intensifies and darkens. 
 
 Dextrose and levulose may be recognised in a mixture of the two 
 by the j8-naphtholhydrazin reaction of Hilger and Rothenfusser, 
 d-glucose forming a hydrazone with a melting-point of 117° C, and 
 the levulose a hydrazone with a melting-point of 162° C. Levulose 
 can also be detected in a mixtLU-e with other sugars by the blue colour 
 which it strikes with a solution of ammonium molybdate and acetic 
 acid on heating in a water-bath, other sugars giving a feeble green. 
 On oxidation with nitric acid, le^nalose gives tartaric and glycollic acid. 
 
 D-mannose, or seminose, is widely distributed in nature, occurring 
 in many plants in the form of anliydride-like condensation products, 
 known as mannosans, which are converted into mannose on hydrolysis 
 with dilute acids. In general properties it is very similar to d-glucose, 
 being fermented by the same yeasts, exhibiting mutarotation, and 
 forming the same osazone with phenylhydrazin. Its most charac- 
 teristic reaction is the formation of a very sparingly soluble hydrazone 
 with phenylhydrazin, which enables it to be easily identified. The 
 
APPENDIX 43.^ 
 
 hydrazone is precipitated out in white, sphero-crystals in the cold 
 within a few minutes, when phenylhydrazin is added to a solution of 
 mannose and acetic acid. The hydrazone has a melting-point of 
 186° to 180° C, and by treating it with benzaldehyd mannose can be 
 recovered, crystallising as rhombic crystals out of 90 per cent, alcohol. 
 A 2 per cent, solution of inannose in water is dextro-rotatory {a^^= 
 + 14-25°). 
 
 D -galactose does not occur free naturally. With dextrose, it may 
 be prepared from the disaccharide lactose, or milk-sugar, by the action 
 of dilute acids or enzymes. It also occurs in the trisaccharide rafftnose, 
 in combination with sucrose, in many gums and sea-weeds as the 
 polymeric form galactan, and in the glucoside cerebrin of the brain 
 spermatoza, pus, spleen, &c. From the brain it was isolated and 
 described under the name " cerebrose " by Thudichum. 
 
 Galactose crystallises out of alcohol in thin, brittle, six-sided water- 
 free plates, but from water in large rhombic prisms, or flat bright 
 needles, containing one molecule of water of crystallisation. It is 
 almost insoluble in absolute alcohol and ether. It is not so soluble 
 in water as dextrose, and its solution is less sweet than cane-sugar. 
 Solutions of galactose are more strongly dextro-rotatory ( + 81°) than 
 dextrose. Fresh solutions show mutarotation. In chemical pro- 
 perties it resembles dextrose. It is fermented by some yeast, but not 
 by all that ferment glucose. It reduces alkaline solutions of the 
 heavy metals. A 1 per cent, solution of galactose reduces 4-7 mole- 
 cules of copper oxide from Fehling's solution. It is only incompletely 
 precipitated from its solutions by ammoniacal lead subacetate. With 
 phenylhydrazin it forms a yellow osazone, consisting of stout needles 
 which are slightly soluble in hot water and alcohol. The pure osazone 
 melts at 196° C, but a melting-point of 193° to 194° C. is usually ob- 
 tained. The acetic acid solution, unlike that of dextrosazone, is opti- 
 cally inactive. A pyridin-alcohol solution is slightly dextro-rotatory 
 ( + 0° 48'). With di-phenylhydrazin it forms a crystalline compoiind 
 which melts at 157° C. It is, however, best recognised, and separated 
 from other sugars, by the formation of the methyl-phenylhydrazone 
 (inelting-point 180° C). Galactose is the only hexose which yields 
 mucic acid on oxidation with dilute nitric acid, and by this property 
 it may be recognised in a mixture of carbohydrates. With phloro- 
 glucin and hydrochloric acid (Seliwanoff's test) galactose gives a red 
 coloration, but no absorption bands (cf. pentoses). 
 
 Laiose (CIT20)u- — Leo's sugar is a levo-rotatory golden syrup 
 {ctB= +26° 07'), which has been separated from the lu-ine in diabetes. 
 It has a saltish, but not a sweet taste, is not fermented by yeast, either 
 before or after hydrolysis, reduces Fehling's solution, but more feebly 
 than dextrose, gives no reaction with Nylander's solution, and forms 
 with phenylhydrazin an oily compound that is insoluble in water, but 
 is soluble in alcohol. With caustic alkalies it gives, on heating, a 
 yellow, but not a brown coloration. It is not precipitated by lead 
 acetate, but separates out on adding basic lead acetate to the solution. 
 
440 GLYCOSURIA 
 
 and making it alkaline with ammonia. This precipitate does not 
 change colour on heating. Unlike glucose it is not precipitated out of 
 a methyl-alcohol solution by a methyl-alcohol solution of baryta. 
 
 Di-Saccliarides. — Maltose (0^2^2201^) is prepared by acting upon 
 starch with diastase, a ferment occiirring in malt, saliva, and pancreatic 
 juice, the only other product of the change being dextrin. It also 
 appears as an intermediate product in the action of sulphuric acid 
 on starch. 
 
 The amorphous anhydride is very hygroscopic, but it usually 
 occurs in fine crystalline needles as the hydrate {Ci2^22^n~'r^2^)- 
 It is readily soluble in water and alcohol, but is insoluble in ether. 
 Solutions of maltose are strongly dextro-rotatory (aj, =+138°), and 
 exhibit upward mutarotation. Maltose is hydrolised to two molecules 
 of dextrose when heated with dilute acids, but is far more resistant 
 than cane-sugar. It is hydrolysed more rapidly by the enzyme maltase 
 contained in many yeasts, and it is only a yeast containing this enzyme 
 that is able to ferment it, since it is necessary that it should be con- 
 verted into dextrose before the yeast can break it down into carbon 
 dioxide and alcohol. The ferments diastase, invertase, lactase, and 
 emulsin are without action on it. Maltase, prepared by extracting the 
 dried yeast with water, affords an absolute means of identifying maltose. 
 Maltose resembles dextrose in its power of reducing hot Fehling's 
 solution without previous inversion, but the amoixnt of cuprous oxide 
 precipitated is only 62 per cent, of that reduced by an equal weight 
 of dextrose. Unlike dextrose, it does not reduce a copper acetate 
 solution (Barfoed's reagent), unless the boiling is prolonged, when 
 hydrolysis of the disaccharide occurs, and reduction follows. 
 ISTylander's solution is reduced by maltose as by dextrose. On being 
 heated with phenylhydrazin at the temperature of the water-bath, 
 maltose forms an osazone which is, however, only precipitated out on 
 cooling, and after an hour's heating. The osazones of dextrose and 
 levulose are precipitated out after ten minutes' heating. Maltosazone 
 is soluble in about 75 parts of hot water, whereas glucosazone is almost 
 insoluble. Maltosazone is readily soluble in hot alcohol, and also in 
 a cold mixture of equal parts of water and acetone. These properties 
 of the osazones afford means of separating the sugars when they occur 
 in a mixtiu-e, but before testing their solubilities it is essential that 
 the osazones should be thoroughly washed with water and benzene to 
 remove products which tend to make dextrosazone appear soluble. 
 The osazone of maltose is not easily piirified, and does not show a 
 sharp melting-point, as it tends to decompose as this is reached. The 
 melting-point is usually stated to be 205° to 206° C. Microscopical 
 examination of the crystalline osazone shows yellow plates, or needles, 
 which are usually broader and shorter than those of dextrose, but 
 both the melting-point and crystalline form are greatly altered by 
 small quantities of impxirities. Theoretically the osazone should yield 
 about 10 per cent, of nitrogen. On being oxidised with bromine, maltose 
 
APPENDIX 441 
 
 yields an acid with the same number of carbon atoms, which is hydro- 
 lised to gkicose and gluconic acid by mineral acids. 
 
 Isomaltose was the name given by Fischer to a disaccharide ob- 
 tained by treating a concentrated solution of dextrose with strong 
 acids at a low temperature. It was separated out as an osazone with 
 a melting-point of 150° to 153° C, and was found to be more easily 
 soluble in hot water than maltosazone, 1 part in 4 as compared with 
 1 part in 75. Products similar to isomaltose have been repeatedly 
 described as obtained in the hydrolysis of starch, along with maltose, 
 by the action of diastase, ptyalin, amylopsin, &c., and an amyloptic 
 ferment in the blood is said to have the same, or a similar, action, 
 but definite proof of its presence in such cases is lacking. Small 
 quantities have also been stated to be present in normal blood and 
 tirine. Isomaltose is probably identical with the disaccharide obtained 
 by Crofton Hill through the action of maltase on glucose which he 
 termed " revertose." 
 
 Isomaltose is said to be readily soluble in water, to be insoluble in 
 alcohol and ether, and to have a very sweet taste. It reduces Feh- 
 ling's and Nylander's solutions, having about four-ninths the reducing 
 power of dextrose. It is not directly fermented by yeast, but under- 
 goes slow changes. It is hydrolised by emulsin, but not by maltase 
 or invertase: With a-naphthol and hydrochloric acid it gives a marked 
 furfurol reaction. Its solutions are dextro-rotatory {aj,= +139° to 
 149°). 
 
 Lactose {milk-sugar) occurs in the milk of all animals, and is occa- 
 sionally met with in the urine. It has not been found in the vege- 
 table kingdom. It is manufactured by evaporating whey, and is 
 obtained as a white crystalline powder. It has a faint sweet taste, 
 and is less soluble than other sugars (1:6 cold water, and 1 : 2 hot). 
 It is insoluble in ether and absolute alcohol. Its solutions in water 
 are dextro-rotatory («„= -1-52° 7') and exhibit mutarotation. It 
 rapidly reduces ammoniacal solutions of silver nitrate. Its reducing 
 power for Fehling's solution is intermediate between that of dex- 
 trose and maltose, being roughly half that of dextrose. With phenyl- 
 hydrazin it forms an osazone, which is soluble in 80 to 90 parts of hot 
 water, and separates out on cooling, as aggregates of yellow needles 
 with a melting-point of 200° C, and a nitrogen content of 10*76 per 
 cent. Lactosazone, like maltosazone, is difficult to purify, and does 
 not show a sharp melting-point as it decomposes on heating. Its 
 crystalline form and melting-point are also materially altered by small 
 quantities of iinpurities. Owing to the great solubility of the osazone 
 ■small quantities of the sugar, such as occvu" in the urines of nursing 
 women, cannot be detected by the phenylhydrazin test. Mineral acids 
 hydrolise lactose to glucose and galactose, but it is less readily hydro- 
 lised than cane-sugar, being unaffected by being boiled for ten minutes 
 with 2 grams of citric acid per 100 c.c. of the solution. Lactose is 
 hydrolised by a specific ferment, " lactase," found in a few yeasts and 
 in some kefir preparations, but not by maltase, invertase, diastase, nor 
 
442 GLYCOSURIA 
 
 by any of the ferments of dried brewer's yeast. Lactase is also found 
 in the secretions of the intestinal mucous membrane, particularly of 
 the new-bom. Lactose is particularly liable to undergo lactic and 
 butyric acid fermentation. Heated with hydrochloric acid and 
 pliloroglucin, it gives a precipitate soluble in alcohol to form a red 
 solution like a pentose, but does not show any bands on spectroscopic 
 examination. 
 
 Isolactose is the name given to a disaccharide obtained by Fischer 
 and Armstrong by the action of the enzyme kefir lactase on a con- 
 centrated solution of glucose and galactose, which they isolated in the- 
 form of an osazone. 
 
 Sucrose, saccharose, or cane-sugar is widely distributed in the vege- 
 table kingdom, where it acts as a store of reserve material. 
 
 It crystallises well, forming large transparent colourless monoclinic- 
 prisms, known as sugar crystals and sugar candy. It is very soluble- 
 in water, forming a sweet, viscid liquid. It is much sweeter than dex- 
 trose, but is not as sweet as invert sugar. Cane-sugar dissolves in 
 about half its weight of cold water, and is soluble in boiling water in all 
 proportions. It is ahiiost insoluble in absolute alcohol. In aqueous- 
 alcohol its solubility increases with the proportion of water. When 
 subjected to prolonged boiling, sohitions of cane-sugar acquire an 
 acid reaction and are in part inverted. The dry substance when cau- 
 tiously heated melts at about 160° C, and, on cooling, forms a trans- 
 parent amber-coloured mass known as " barley-sugar." Above 160° C 
 it decomposes, turning brown, and forming so-called " caramel." On 
 being heated with dilute mineral acids, cane-siigar is hyclrolised to 
 dextrose and le\ailose. A solution of cane-sugar is dextro-rotatory 
 (aj, = 4- 66° 5'), but does not exhibit mutarotation. Since fructose is 
 more levo -rotatory than glucose is dextro-rotatory, the products of 
 hydrolysis rotate polarised light in the opposite way to cane-sugar. 
 The process of hydrolysis is hence termed " inversion," and the pro- 
 duct " invert-sugar." A similar change is brought about by an enzyme 
 present in yeasts, moulds, and in many plants, termed "invertase or 
 sucrase." Cane-sugar is only fermented by yeasts after it has been 
 previously inverted by the invertase of the yeast, and accordingly it 
 is not fermented by yeasts that do not contain this ferment {e.g. S. 
 octosporus). Cane-sugar lacks both aldehydic and ketonic properties. 
 It does not therefore reduce Fehling's solution, does not form com- 
 pounds with phenylliydrazin, and is stable toward alkalies. On 
 treatment with moderately concentrated nitric acid it forms saccharic 
 and oxalic acids. Cane-sugar forms definite compounds with some 
 metallic oxides. Lead is attacked by sugar solutions slowly in the 
 cold, but more quickly at boiling-point, the lead passing into solution. 
 Calcium sucrate has an alkaline and bitter taste, and forms the liquor 
 calcis saccharatus of pharmacy. Barimn hydroxide forms a crystalline 
 compound, which, on recrystallisation from boiling water, forms 
 brilliant scales resembling boracic acid. It is only sparingly soluble 
 in cold water. By means of strontium hydroxide, almost the whole 
 
APPENDIX 443 
 
 of the sugar may be separated from a solution as a granular precipitate. 
 Crystalline compounds are also easily obtained with some sodiiuii 
 salts ; thus sodiiim chloride and iodide both enter into svich combina- 
 tions. These properties are made use of in the commercial separation 
 and purification of cane-sugar. 
 
 Polysaccharides. — Starch {^'^iffyr,)^ is a characteristic product 
 of the vegetable kingdom, and is found in almost every part of all 
 plants. 
 
 It is a non-crystalline, colourless powder, composed of granules, 
 which have a characteristic appearance under the microscope, and 
 particularly mider polarised light. The starch granule consists chiefly 
 of a body, called granulose, which is coloured blue by iodine, together 
 with a closely allied substance known as starch-cellulose, which gives 
 only a dirty yellow colour with iodine. Starch-cellulose occurs in 
 largest proportions in the outer layers of the granule, and probably 
 constitutes the whole of the external covering. It is owing to the 
 presence of this covering, which though slightly soluble is highly 
 colloidal, that starch is unacted on by cold water. By boiling with 
 water, starch-cellulose is mostly converted into soluble starch, and the 
 granulose is allowed to escape. Iodine solutions readily permeate the 
 outer layer of starch-cellulose, and colour the contained granulose of 
 the solid starch an intense blue. Starch is insoluble in cold water, in 
 alcohol, and in ether, but gelatinises and dissolves in hot water to form 
 an opalescent solution. This solution is strongly dextro-rotatory 
 (ac=+200°). The starch is precipitated out by ammonimn sulphate, 
 magnesimn sulphate, ammoniacal lead acetate, and tannin. It does 
 not reduce Fehling's solution, Moore's test is negative, it does not 
 form an osazone with phenylhydrazin, and is not fermented by yeast. 
 Starch is insoluble in Schweitzer's solution (ammoniacal cupric oxide). 
 Heated with dilute acids it is converted into a mixture of dextrins and 
 maltose, which ultimately yield the monosaccharide dextrose. Among 
 the intermediate products is erythro -dextrin, which is apparently the 
 first to develop. Achroo -dextrin appears later, and from this maltose, 
 and finally dextrose, are formed. During the decomposition other 
 dextrins of lower molecular weight are simultaneously produced, and 
 these also yield maltose and dextrose, but finally one dextrin, tenned 
 malto -dextrin, is obtained which undergoes no fm-ther change. Similar 
 changes are produced by the ferments diastase (amylase), pytalin, 
 and amylopsin. The first step is the formation of amylo-dextrin, or 
 soluble starch. This is decomposed into erytliro-dextrin and maltose. 
 In the third stage the erythro -dextrin is split up into achroo-dextrin 
 and a further c[uantity of maltose. Finally part of the acliroo -dextrin 
 yields maltose, and part remains as a variety of dextrin not affected 
 by the ferment. The action of acids is, however, more rapid, and is 
 carried a stage further — naixiely, to dextrose. The ferment " maltase " 
 found in the intestinal mucous membrane, and to a certain extent in 
 some plants, has the powder of converting maltose into dextrose. 
 
444 GLYCOSURIA 
 
 Inulin (CgHjdOg),! is a substance akin to starch, found in the roots 
 of various plants of the conipositse group, particularly in dahlias, 
 dandelions, and chicory. It is the only polysaccharide that can be 
 obtained in a crystalline form. It is met with as a white hygroscopic 
 powder, or as sphsero- crystals. It is slightly soluble in cold water, 
 and is readily soluble in hot water. On cooling the solution it is 
 precipitated. Its solution is levo-rotatory. It is insoluble in absolute 
 alcohol, and sparingly soluble in dilute alcohol. On being boiled with 
 dilute acids it is converted into levulose. It reduces ammoniacal 
 solutions of silver nitrate, but does not reduce Fehling's solution. It 
 does not give any colour reaction with iodine. 
 
 Glycogen (CgH^gOg)!! is as important a substance in the animal 
 economy as starch is in plant life, forming the chief reservoir of carbo- 
 hydrate material. It is present in the bodies of many protozoa, and 
 is constantly met with throughout the animal kingdom, from the 
 worms upwards. It exists as a small, but constant, constituent of the 
 protoplasm of all animal tissues. In the lower forms it is found uni- 
 formly distributed throughout the cells {e.g. taenia l"o to 4-7 per cent., 
 ascaris 4-2 to 7-1 per cent.), but in the higher animals it is more abun- 
 dantly present in certain situations, notably the muscular tissues and 
 certain portions of the aliinentary tract {e.g. the mid-gut of molluscs 
 and crustaceans) than in others. After the appearance of the liver, 
 this organ becomes one of the chief seats of the deposition of 
 glycogen. The quantity present in the liver is primarily dependent 
 upon the state of nutrition of the animal and the amount of exercise 
 that is taken. The maximal amounts, according to Kiilz, are found 14 to 
 16 hours after food. It has been calculated that in the liver of a man 
 150 grams can be stored at one time ; this would correspond to about 
 10 per cent, of an organ weighing 1500 grams. The quantity of glycogen 
 which is deposited in the muscular tissues probably represents about 
 half the total amount that is present in the entire body, and in man 
 corresponds to about 150 grams. Occasionally traces of glycogen are 
 met with in diabetic urines. 
 
 Pure glycogen is a white amorphous powder, which is both odour- 
 less and tasteless. It is slowly soluble in water, forming an opalescent 
 fluid, which can be cleared by the addition of acetic acid. Pure solu- 
 tions are strongly dextro-rotatory (oj,= -j-197°). It is insoluble in 
 alcohol and ether, and can be precipitated from watery solutions by 
 the addition of alcohol, the precipitation being promoted by the addi- 
 tion of a little sodimn chloride. It is also precipitated by baryta water, 
 and, vmlike dextrin, by basic lead acetate. Filtration through animal 
 charcoal removes it from watery solutions. With benzoyl chloride, in 
 the presence of sodium hydrate, it gives a granular precipitate of 
 benzoyl-glycogen. It is readily soluble in alkalies. Glycogen does 
 not reduce Fehling's solution, but can maintain cupric hydroxide in 
 solution. After the addition of a little sodiixm chloride, its solutions 
 are coloizred red with iodine. On boiling with dilute acids it is trans- 
 formed into dextrose. The ferments diastase and maltase produce 
 
APPENDIX 445 
 
 changes very similar to those induced in starch, a great part of the 
 glycogen being converted into dextrose, and a small part into achroo- 
 dextrin, with, in some instances, a small amount of maltose. It is 
 not fermented by yeast. Glycogen does not form an osazone with 
 phenylhydrazin. 
 
 Dextrin (CgHjQOj)!) is the name given to a nmnber of intermediate 
 products formed dm"ing the hydrolysis of starch by dilute acids and 
 diastase. The principal varieties are erythro-dextrin, which gives a 
 red colour with iodine ; achroo-dextrin, which gives no colour reaction 
 with iodine ; and malto-dextrin. 
 
 Dextrin is a white amorphous, tasteless, odoi^u-less, very delacjuescent 
 powder. It is readily soluble in water, and its solutions are strongly 
 dextro-rotatory. It is insoluble in alcohol and ether. The dextrins 
 do not reduce alkaline solutions of copper, but give a blue solution 
 with Trommer's test. On being boiled for thirty minutes with dilute 
 sulphi.iric acid, maltose, dextrose, and other reducing substances are 
 formed, and these, after neutralising the acid with a sodimn hydrate, 
 reduce Fehling's solution. The dextrins are not coagulated by heat 
 or mineral acids, and are not precipitated by tannin or baryta water. 
 Heated with nitric acid they give rise to oxalic acid. 
 
 Cellulose is a characteristic product of the vegetable kingdom, 
 forming the essential part of the solid framework of all plants. It is 
 insoluble in water and all simple solvents. It is not hydrolised by 
 boiling with dilute acids, but treatment with strong sijlphuric acid, 
 followed by dilution, gives rise to dextrose and other substances. It 
 does not react with iodine sohitions, but is turned blue after pre- 
 liminary treatment with zinc chloride. It is soluble in Schweitzer's 
 solution (amnion, cupric oxide), forming a levo -rotatory viscid solution. 
 Cellulose is not acted on by the digestive ferments of the alimentary 
 canal, but may be decomposed by intestinal bacteria into carbon 
 dioxide and methane. 
 
 Gums. — Although gioms are usually classed with carbohydrates, it 
 has now been shown that they are really glucoside derivatives of certain 
 organic acids. The acid is different in different gmns, and is to be 
 regarded as the nucleus of the particvilar gi_un. The commonest sugars 
 in gums are galactose and arabinose. 
 
 The gvuns are a peculiar class of bodies occvirring in the juices of 
 plants. They are non-volatile, have little or no taste, and are un- 
 crystallisable and eminently colloidal. They are either soluble, or 
 swell up, in contact with water, giving a levo-rotatory solution. They 
 are insoluble in alcohol, are not fermented by yeast, and do not react 
 with iodine. On boiling with dilute acids they yield the pentoses or 
 hexoses that are united to their acid nucleus. On treatment with 
 moderately concentrated nitric acid they yield mucic acid. Solutions 
 of gums give a gelatinous precipitate with copper and iron salts. 
 
 Animal Guin. — The substance separated, and first described by 
 Landwelir, im^der the name of aninial gum, is a decomposition product 
 of mucin, and is probably not a chemical entity, but a mixture of sub- 
 
446 GLYCOSURIA 
 
 stances that is precipitated from the iirine by alcohol. Chemically 
 it is not related to the vegetable gmns. 
 
 In water it gives an opalescent solution, out of which it is not 
 precipitated by alcohol. It gives no colour reaction with iodine. It 
 is not precipitated by lead acetate, but is precipitated on the addition 
 of ammonia to the solution. Boiled with dilute sulphuric acid it yields 
 a reducing, but unfermentable, sugar, which has been termed " gum- 
 mose " (CgHjoOg). The reduction of both copper and bismuth is slow 
 and incomplete. Treatment with nitric acid gives oxalic acid, and 
 with hydrochloric acid levulinic acid, leucin, tyrosin, &e. Copper and 
 iron salts give a gelatinous precipitate, like that given with the vege- 
 table gums. Boiling with hydrochloric acid and a-naphthol gives a 
 well-marked furfur ol reaction similar to that yielded by the pentoses. 
 
 Inosite is not a carbohydrate, as was at one time supposed, but 
 belongs to the aromatic series, and is commonly regarded as hexa- 
 hydroxybenzol. It is apparently a constant constituent of muscle, 
 but is also found in other tissues of the body. It is not, however, 
 peculiar to the animal world, but is also widely distributed in the 
 vegetable kingdom. Inosite occurs in the urine in some cases of 
 diabetes and albvuiiinm-ia, and is met with when polym-ia is artificially 
 produced, or results from morbid processes. According to Hoppe- 
 Seyler traces are present in all urines. 
 
 In the pvire crystalline form it appears as colourless prisms which 
 are often grouped in rosettes. It melts at 217° C. It has a sweet 
 taste, but gives none of the characteristic reactions of the sugars. It 
 is soluble in water and dilute alcohol, but is insoluble in absolute alcohol 
 and ether. Inosite does not reduce the metallic oxides in alkaline 
 solution, is optically inactive, and is not fermented by ordinary yeast. 
 Bacterium lactis decomposes it with the formation of lactic acid, and 
 it subsequently yields butyric acid. 
 
 The Carbohydrate Constituent of Proteins. — Many albumens 
 when examined with the Molisch-Udransky reaction, or Bial's modifica- 
 tion of the orcin test, give results which indicate the presence of a 
 ■carbohydrate group in the molecule. Investigation of their degrada- 
 tion products has confirmed this, and a substance having the reactions 
 -of a carbohydrate has been separated from most proteins. 
 
 Such carbohydrate complexes are most readily separated from the 
 ^lyco-proteins (mucins, cartilages, &c.), and can be prepared by simple 
 hydrolysis, a fact which distinguishes this group of substances from 
 the ordinary albumens, from which the sugar group, or the parent sub- 
 stance of the sugar group, is only eliminated by more drastic measures. 
 The carbohydrate is usually obtained in the form of chitosamine or 
 glycosamine, an amine or nitrogenous sugar (CgHj^Og.NITg). Glucos- 
 amine does not exist preformed in the albmxien molecule, but is a 
 degradation product derived from the carbohydrate constituent of the 
 molecule, which probably exists in the form of a po]3^saccharide 
 {Frankel's " albamin "). 
 
APPENDIX 
 
 447 
 
 Glucosamine-like complexes are not the only carbohydrate groups 
 that may be derived from albunnens, for it has been shown that, by 
 the hydrolysis of certain nucleic acids, substances having the reactions 
 of a pentose can be isolated. Thus frona the nucleo-proteid of the 
 pancreas, liver, &c., a carbohydrate identified as 1-xylose has been 
 prepared. It is not known whether all nucleic acids yield pentoses on 
 hydrolysis, and it is probable that hexoses can be obtained from some. 
 
 The amount of reducing substance that can be separated from 
 ■different albumens differs considerably. The largest proportion is 
 yielded by the glyco-proteids, some .30 to 40 per cent., while crystalline 
 ■egg-albumen gives from 10 to 11 per cent. Other albiunens show a 
 much smaller yield {e.g. sermn albumen 0"5 per cent.). Casein appears 
 to be the only animal albtunen which does not yield a carbohydrate 
 •complex on hydrolysis, and the vegetable albvimen appear frequently 
 to lack a carbohydrate group. 
 
 The tend of all modern research has been to show that carbohydrate 
 groups foriTL no inconsiderable part of the whole albumen molecule, 
 but much yet remains to be done before our knowledge of the subject 
 can be considered satisfactory. 
 
 Acids axd Acid-Derivatives of the Sugar Group 
 Mono-hasic Acids of the Sugar Group 
 
 GlycoUic acid. 
 
 Glyceric acid. 
 
 Tri-oxybutyric 
 acid. 
 
 Tetra-oxyvaleric 
 acid. 
 
 Gluconic, 
 
 Mannonlc, 
 
 Galactonic acids. 
 
 CH'„(OH.COOH 
 
 CH.,.OH 
 1 
 COOH 
 
 C„H„'OH)„.COOH 
 
 C.,H (OH'.,.COOH 
 
 C^H5(0H)^.C00H 
 
 C5H6(0H)5.C00H 
 
 CH., OH 
 
 1 
 
 CH.OH 
 
 COOH 
 
 CH.,.OH 
 1 
 (CH.OH)., 
 
 1 
 COOH 
 
 CH.,.OH 
 
 i 
 
 (CH.OH)., 
 
 COOH 
 
 CHo.OH 
 
 (CH.OH)^ 
 1 
 COOH 
 
 The monobasic acids result from the action of feeble oxydising 
 agents, such as bromine or dilute nitric acid, on the sugars. On 
 evaporating their solution they are partly converted into a lactone, or 
 intramolecular anhydride. The lactones are mostly less soluble, and 
 more readily crystallisable, than the corresponding acids. They form 
 characteristic compounds with strychnine, brucine, zinc salts, and with 
 phenylhydrazin. If a 10 per cent, solution of a lactone is heated with 
 a large excess of phenylhydrazin, and an equal quantity of 50 per cent, 
 acetic acid, on a water-bath for half an hour, a hydrazide of the 
 acid is formed. By heating this with baryta water the acid can be 
 recovered. Glycollic and glyceric acids do not, however, give a com- 
 pound with phenylhydrazin. 
 
 The most important members of this group are : — 
 Glycollic acid {hydroxy -acetic acid) is found in unripe grapes, and 
 in the leaves of the wild vine. It can be prepared from dextrose by 
 
448 GLYCOSURIA 
 
 oxidation with silver oxide. It appears as colourless needles, or plates, 
 which melt at 80° C. It is hygroscopic, and is readily soluble in water, 
 alcohol, and ether. Oxidised with nitric acid it yields oxalic acid. 
 
 Glycocoll (amido-acetie acid, CHo.XHg.COOH) is a constituent of 
 the glycocholic acid of the bile. It is generally prepared by treating 
 glue with acids or alkalies. It occvirs as colourless prisms, soluble in 
 water, insoluble in absolute alcohol and in ether. It has a sweet 
 taste, and is consequently known as gelatine sugar or glycocoll. 
 
 Tri-oxyhutyric acid is prepared from erythrose, or from levulose, by 
 treating them with barium hydroxide and mercuric oxide. 
 
 Gluconic Acid {Penta-oxycaproic Acid). — On treating dextrose with 
 bromine water and silver oxide, the aldehyde group is oxidised to 
 carboxyl, yielding gluconic acid : — 
 
 C.HO COOH 
 
 1 1 
 
 (CH.OH)^ (CH.OH)^ 
 
 CHo.OH CFg.OH 
 
 Mannose, galactose, and other aldoses, with their derived disaccharides, 
 on being similarly treated yield acids corresponding to gluconic acid, 
 known as mannonic, galactonic acids, &c. 
 
 Gluconic acid in sohition readily passes into the anhydride or 
 lactone form, a change which is accompanied by an alteration in the 
 rotatory power of the solution. \^'Tien heated with quinoline or 
 pyridine, gluconic acid is partly converted into mannonic acid, a re- 
 action which is reversible. This property has proved of much value 
 in the synthesis of the sugars. Similarly galactonic and talonic acids 
 are mutually interchangeable. The bacterium xylinum, or sorbose 
 bacterium, oxidises aldoses to the corresponding monobasic acids, 
 converting glucose into gluconic acid, galactose into galactonic acid, 
 and the pentoses, xylose and arabinose, into xylonic and arabonic 
 acids respectively. In all these cases the -C.HO group of the sugar 
 is oxidised to - COOH, through the agency of the organism. 
 
 Lactic acid {oxy -propionic acid, CH3.CH(0H).C00H) is the next 
 liighest homologue of glycoUic acid. It is an oxyacid, and, since it 
 contains an asymetrical carbon atom, can exist in a dextro-rotatory, 
 a levo -rotatory, and an optically inactive form. 
 
 Inactive lactic acid (a-oxy-propionic, ethylidene, or fermentation 
 lactic acid) is formed in the lactic acid fermentation of sugars, and 
 substances related to them, by lactic acid organisms if the solution is- 
 nearly neutral, and also by the action of dilute alkalies on carbohydrates.. 
 It is a thick, syrupy liquid, which on being distilled in vacuo and 
 strongly cooled separates out in a crystalline form. The crystals melt 
 at 1° C., and are very hygroscopic and delaquescent. The acid is 
 iniscible with water, alcohol, and ether. It has a strongly acid taste 
 and reaction. When heated, it is partly converted into the anliydride, 
 lactide, and partly broken up into aldehyde, carbon monoxide, and 
 water. On oxidation it gives acetic and carbonic acids. It may be^ 
 
APPENDIX 449 
 
 separated, purified, and recognised by the formation of the calcium, 
 zinc, or strychnine salts. The calcium salt fonns warty masses of 
 microscopic needles, that are easily soluble in hot water, but are much 
 less soluble in cold water (1 : 9-5), and are insoluble in cold alcohol. 
 The strychnine salt is also a comparatively insoluble compound, and 
 serves to separate it from the dextro-rotatory acid. 
 
 Dextro-rotatory lactic acid (sarco- or para-lactic acid) occurs in muscle, 
 brain, &c., and, under pathological conditions, in the urine. It forms 
 a colourless, odourless syrup which is easily soluble in water, alcohol, 
 and ether. It is dextro-rotatory (% = -f 3-5°), but its salts are levo- 
 rotatory. The most characteristic are the zinc and calcium com- 
 pounds, which differ from those of the inactive acid in crystallising out 
 with one molecule less of water, and in the zinc salt being much more 
 easily soluble and the calcium salt being much more insoluble. The 
 strychnine salt is also more soluble than that of the inactive acid. 
 Like the inactive acid, it gives Uffelmann's reaction, a canary yellow 
 colour with a 2 to 4 per cent, solution of carbolic acid and a drop of per- 
 chloride of iron solution. When heated it is converted into the lactide 
 and aldehyde. 
 
 Butyric acid (CHg.CHo.CHg.COOH) is produced in the fermentation 
 of sugars and starches, but at a later stage than lactic acid. It also 
 results when albumens are oxidised with chromic acid, and fats with 
 nitric acid. It occurs free in perspiration, the juice of flesh, the con- 
 tents of the large intestine, and in the faeces. It is met with in butter 
 as a glycerine ester to the extent of about 2 per cent. Butyric acid is 
 a thick, syrupy liquid with a rancid odoiir, miscible with water, but 
 separating out from the watery solution on the addition of salts. It 
 is only oxidised with difficulty in the laboratory. With calcium it 
 forms a salt, which separates out as glancing plates, and which is 
 reinarkable in being more soluble in cold than in hot water. It is 
 therefore deposited on warming the concentrated aqueous solution. 
 
 Beta-oxybutyric acid (CIT3.CH(OH).CH2.COOH) occiirs in the urine 
 in severe cases of diabetes, scurvy, severe infectious conditions {e.g. 
 scarlet fever and measles), and in starving insane persons, &c. It 
 also appears in the urine in health after several days on a purely 
 protein diet. 
 
 W^ith water it forms a coloiu"less syrup, but may be crystallised out 
 as transparent plates with a melting-point of 49° to 50° C. The crystals 
 are soluble in water, alcohol, and acetone. The fluid is levo-rotatory 
 (a„ = — 24° 12'). Its salts are also levo-rotatory. They are easily 
 soluble in water, feebly soluble in absolute alcohol, and are precipitated 
 from their solution by adding ether. On being heated with water, or 
 dilute sulphuric acid, oxybutyrie acid is converted into o-crotonic 
 acid (CH3.CH : CH.COOH) and water, the acid forming crystals with 
 a melting-point of 71° to 72° C. By treatment with chromic acid it 
 gives rise to acetone. The tests by which oxybutyrie acid can be 
 recognised in the urine and the methods of estimating it have been 
 dealt with elsewhere (p. 104). 
 
 2 F 
 
450 
 
 GLYCOSURIA 
 
 Aceto-aeelic. or di-acetic, acid (CH3.CO.CH2.COOH) is so called 
 because it may be considered to consist of two molecules of acetic acid 
 (CH3.COOH), ixiinus one molecule of water. 
 
 Aceto -acetic acid does not occur in the urine of healthy individuals 
 on a mixed diet, but is met with when under -nutrition and failure of 
 absorption exist. It also occurs in healthy persons after some days 
 on a purely protein diet. Pathologically it is met with in certain 
 fevers, especially in children, in gastro -intestinal diseases, particularly 
 in drunkards, and in severe cases of diabetes. It is only found when 
 acetone is present, but not always then. 
 
 Aceto-acetic acid is a syrup that is easily soluble in water, alcohol, 
 and ether. It is strongly acid in reaction. On being heated it is easUy 
 decomposed into acetone and carbon dioxide. Its salts are readily 
 soluble in water. The tests and methods of estimating aceto-acetic 
 acid are considered elsewhere (p. 104). 
 
 Acetone, or di-inethylketone (CH3.CO.CH3), occurs in normal urine 
 in small amounts, up to 10 mg. in the twenty-four hovu-s. The outpiit 
 is increased when the carbohydrates of the food are limited, and the 
 proteins are increased. Rich fat catabolism also increases the acetone 
 in the urine, but it requires about 150 grams to produce any marked 
 eSect. Acetonuria is met with in febrile conditions, especially in 
 children, in carcinoma in which inanition is not yet present, in states 
 of inanition and cachexia, psychoses and lesions of the central nervous 
 system, especially when associated with starvation, as a result of 
 auto-intoxication, in digestive disturbances, particularly gastric ulcer, 
 from chloroform narcosis, during pregnancy with a dead foetus, and 
 after certain poisons {e.g. phlorhidzin). Acetone is a colourless mobile 
 flmd, with a pleasant fruity smell, that boils at 56° C. It is miscible 
 with water, alcohol, and ether in all proportions. It can be separated 
 out from its watery solution by the addition of salts, and particularly 
 of calcium chloride. On being shaken Math a concentrated watery 
 solution of sodium bisulphite, it forms a colourless crystalline com- 
 pound, that is readily soluble in water, and is quickly decomposed by 
 dilute acids or alkalies, the acetone being regenerated. The estimation 
 and tests for acetone in the urine are considered elsewhere (p. 104). 
 
 Di-basic Acids of the Sugar Group 
 
 Oxalic Acid. 
 
 Tartronic Acid. 
 
 Tartaric Acid. 
 
 Aposorbic Acid. 
 
 Saccharic, ^lucic 
 Acids. 
 
 (COOH) a 
 
 CH.OH.'COOH)„ 
 
 (CH.OH ',,{COOH)„ 
 
 (CH.0H)3'C00H)2 
 
 (CH.0H)XC00H)2 
 
 COOH 
 COOH 
 
 COOH COOH 
 
 1 1 
 CH.OH (CH.OH).. 
 1 1 
 COOH COOH 
 
 COOH 
 1 
 
 (CH.OH), 
 1 
 COOH 
 
 COOH 
 (CH.OH)^ 
 COOH 
 
 By the action of energetic oxidising agents on the sugars both ends 
 of the carbohydrate chain are oxidising, with the formation of di-basic 
 
APPENDIX 451 
 
 acids of the general formula COOH.(CH.OH)„.COOH. Of these the 
 most important are oxalic acid, and the isomers saccharic and mucic 
 acids. 
 
 Oxalic acid is extensively produced in the physiological processes 
 of plants, and to a less extent in animals. In plants it occvirs as the 
 free acid, or as sodium, potassium, or calcium salts, the last named 
 forming crystalline deposits in the plant cells, known as " raphides." 
 It is a normal constituent of the urine, the amoiint varying from 0-2 
 to 0-5 gram in the twenty-four hours. It is supposed to be present 
 as the calcium salt, held in solution by di-acid sodium phosphate. It 
 readily separates out on standing, and is then met with in the urinary 
 sediment, and occasionally forms calculi. As certain articles of diet, 
 such as asparagus, spinach, carrots, tomatoes, grapes, rhubarb, apples, 
 plums, figs, strawberries, coffee, &c., contain a considerable amount 
 of oxalic acid it is supposed that a certain proportion of that present 
 in the urine is derived from the food, but as it does not entirely dis- 
 appear on a diet of fat and protein, or even on starvation, a part must 
 originate within the organism. From the close chemical relationship 
 of oxalic to oxaluric acid, and of the latter to uric acid and the purin 
 bodies, it is assiuiied that oxalic acid is formed from albumen, but the 
 well-known tendency to increased oxalate excretion in diabetes, and 
 the way in which a temporary diminution in the sugar output may 
 be associated with an increase in the oxalates, have suggested that it 
 may also arise from the incomplete oxidation of carbohydrates. Accord- 
 ing to Baldwin {Journ. of Exp. Med., 1900), oxaluria may be caused 
 by an excessive fermentation of carbohydrates. Oxalic acid may be 
 prepared by the oxidation of sugar, starch, wood, and other organic 
 bodies by the action of dilute nitric acid and other oxidising agents. 
 It is made in bulk commercially by melting cellulose with caustic 
 potash. It crystallises out with one molecule of water in the form 
 of prisms. It is colovu-less, odourless, has an intensely sour taste, 
 :and an acid reaction. It is intensely poisonous. One part of oxalic 
 acid is soluble in 10-46 parts of water at 14-5° C, and in 2-5 parts of 
 ■cold alcohol, but is more easily soluble in hot alcohol ; 1-266 parts are 
 soluble in 100 parts of ether at 15° C. It is insoluble in cliloroform, 
 benzene, and petrolemn spirit. On heating it volatilises without char- 
 ring at 150° to 160° C. With phenylhydrazin it forms glancing, colour- 
 less plates which soften at 170° C. The calcium salt of oxalic acid is 
 insoluble in water, ammonia, acetic and other organic acids, but is 
 soluble in dilute mineral acids {e.g. hydrochloric). It is re-precipitated 
 on making the mineral acid solution alkaline with ammonia. 
 
 The Estimation of Oxalic Acid in Urine. — Treat 600 c.c. of fresh 
 urine with a small quantity of an alcoholic solution of thjanol, 
 to prevent putrefaction. INlake neutral, or faintly alkaline, \yith 
 ammonia, and add an excess of a satiu-ated solution of calcimn 
 chloride. The disoditmi phosphate which holds the oxalic acid in 
 soKition is thus removed. The precipitate is treated with just 
 svifficient acetic acid to dissolve it. The calcimn oxalate being in- 
 soluble in acetic acid, is gradually precipitated when the mixture 
 
452 GLYCOSURIA 
 
 is allowed to stand for twenty-four hours. At the end of this time 
 the calcium oxalate is filtered off, washed with a little water, and 
 dissolved in dilute hydrochloric acid. Sufficient ammonia is added 
 to the solution to give it a feebly alkaline reaction. After standing 
 twenty-four hours the calcium oxalate will have separated out 
 again. It is then filtered off, and treated in one of the following 
 ways — (1) dried at 100° C, and weighed as calcium oxalate; (2) 
 ignited, moistened with ammonimii carbonate, again gently ignited, 
 and weighed as calcium carbonate ; (3) moistened on the filter with 
 strong sulphvu-ic acid and the whole ignited, again moistened with 
 sulphuric acid, re-ignited, and finally weighed as calcium sulphate ; 
 (4) it is ignited thoroughly and the resultant calcium oxide and 
 carbonate weighed (56 parts =128 Ca.Ox.) ; or better (5) titrated 
 with standard acid ; or (6) the filter is placed in a beaker with water 
 and dilute sulphuric acid, and the liquid titrated with standard 
 potassium permanganate. The last two methods are probably the 
 best, as they are least affected by impurity in the precipitate, but 
 in the permanganate method the precipitate must be quite free 
 from organic salts. 
 
 Blair Bell [Brit. Med. Journ., 1912, i. p. 878) has described a method 
 of estimating the calciixm in the urine with the aid of the centrifuge, 
 which he states gives accurate results. 
 
 Tartaric acid (di-oxy-succinic acid) occurs in some plant juices, but 
 its only important source is grape juice. It is met with in four forms, 
 physically isomeric : — 
 
 (1) Dextro-rotatory, or ordinary, tartaric acid is found in nature, and 
 
 particularly as the acid potassium salt, especially in grapes. 
 It forms large transparent prisms, easOy soluble in water to 
 form a dextro-rotatory solution. It is easily soluble in alcohol, 
 but is almost insoluble in ether. It reduces ammoniacal 
 solutions of silver on heating. Rochelle salt, used in the 
 preparation of Fehling's solution, is potassium sodium tartrate 
 (QH.OgKNa). 
 
 (2) Levo -rotatory form is chemically identical with the dextro- 
 
 rotatory form, but rotates the plane of polarised light in an 
 equal and opposite direction. 
 
 (3) Racemic tartaric acid is a mixture of equal parts of the dextro- 
 
 and levo -rotatory forms. It is interesting historically, as it 
 originated the idea of isomerism. 
 
 (4) Meso-tartaric acid is optically inactive, like the above, but is not 
 
 decomposable into active acids. It is produced by prolonged 
 heating of the dextro-rotatory acid with a little water at 
 165° C. 
 Saccharic acid is produced by oxidising dextrose, and substances 
 containing dextrose, such as dextrin, starch, &c., with nitric acid. 
 
 Saccharic acid is hydroscopic and easily soluble in water. It forms 
 a sparingly soluble potassium salt, which serves for its separation. 
 
 Preparation of Saccharic Acid from Dextrose. — Two grams of 
 dextrose are heated with 10 c.c. of nitric acid of a specific gravity 
 of 1-2 (made by mixing 2 parts of Cone. HNO3 ^^^ 1 P^^^ *^f water). 
 
APPENDIX 453 
 
 in a porcelain, capsule on the water-bath, until a brisk reaction 
 ensues and red vapours are given off. The heating is continued a 
 few moments, and, when the reaction has subsided, the contents of 
 the capsule are evaporated to a clear syrup to expel the excess of 
 acid. Five or 6 c.c. of water are then added. The hot liquid is 
 saturated with dry, powdered potassium carbonate, and 4 c.c. of 
 glacial acetic acid are added. The mixture is well shaken, and 
 cooled. The acid potassium saccharate, which is only slightly 
 soluble, separates out as a white crystalline deposit. Microscopical 
 examination of this shows that it consists of transparent needles 
 free or arranged in rosettes. 
 
 Mucic acid is isomeric with saccharic acid, and is produced by the 
 oxidation of galactose, and substances containing galactose, such as 
 milk-sugar, gvims, and mucilages, with nitric acid. It occvirs as a 
 sandy crystalline powder, which, unlike saccharic acid, is only feebly 
 soluble in water (1 in 300 at 14° C). It is insoluble in alcohol. It 
 melts at 206° C, and at the same tmie undergoes decomposition. It 
 is readily changed into furfiirane (C4H4O). 
 
 Preparation of Mucic Acids from Galactose. — Two grams of the 
 sugar are heated with nitric acid in exactly the same way as in the 
 preparation of saccharic acid from dextrose, but when the oxidation 
 is completed and the major part of the nitric acid has been evapo- 
 rated off, 3 or 4 c.c. of water are added, and the contents of the 
 capsule are poixred into a test-tube. The capsule is then washed 
 with water, and the washings added to the solution in the test-tube 
 until a total of 10 c.c. is reached. A rapid precipitation of the mucic 
 acid, in the form of a white powder, then takes place. This, on 
 microscopical examination, is found to consist of small short prisms. 
 The presence of mucic acid is confirmed by its complete solubility 
 in ammonia, in contrast to the calciima oxalate formed on oxidising 
 a mixture containing a calcium salt, which is insoluble in ammonia. 
 Mixed with a few drops of ammonia, evaporated to dryness, and 
 strongly ignited, mucic acid gives off pyrrol vapours, which coloiu" 
 a soft pine splinter soaked in concentrated hydrochloric acid a bright 
 red colour. 
 
 The weight of the acid, collected on a weighed filter -paper, after 
 standing for twenty-four hours, fm"nishes a means of approximately 
 estimating the galactose in a mixtiu'e, 1 gram of sugar of milk under 
 these conditions being equivalent to 0*33 gram of mucic acid dried 
 at 110° C. 
 
 Furfurol, furfurane aldehyde, or furfur aldehyde (C4H3O.CHO) is an 
 aldehyde of pyro-mucic acid (fvirfurane carboxylic acid, C4H3O.COOH), 
 which is formed by the dry distillation of mucic acid. 
 
 It is a colourless, oily liquid, with an agreeable odour resembhng 
 bitter almonds and cinnamon. It boils at 161° C, and tiu"ns brown 
 on exposure to the air. It has the general properties of an aldehyde, 
 and also shows characteristic colour reactions with certain substances 
 by which it can be recognised. Its chief importance lies in the fact 
 that it is produced on heating sugars, and substances containing a 
 carbohydrate radicle, with hydrochloric, or sulphm"ic acid, of suitable 
 strength. The readiness with which the sugars yield furfiu"ol varies 
 
454- GLYCOSURIA 
 
 with the nature of the sugar and the conditions under which the ex- 
 periment is carried out, and on this is based a number of tests for the 
 differentiation of those which, hke the pentoses and ketoses, readily 
 give much furfvirol. With bodies of the phenol series it forms colovired 
 compounds which vary in appearance with the nature of the sugar, and 
 also with that of the phenol employed. Thus with orcin, in the presence 
 of strong hydrochloric acid, the pentoses give a bkie colour, or a green 
 if iron is present, the methylpentoses and hexoses a red orange colour. 
 With phloroglucin and concentrated hydrochloric acid the colour is 
 red in all cases. The reaction of Seliwanoff, which distinguishes be- 
 tween aldoses and ketoses, is based upon the fact that the latter form 
 fiu-furol when treated with hydrochloric acid diluted with its own bulk 
 of water, whereas the former do not. The phenol employed in this 
 case is resorcin, which gives a red colour. The reaction is most usually 
 employed for the detection of levulose. Furfurol gives colour reactions 
 with other substances, such as xylidine, amyl alcohol, aceto-acetic ether, 
 acetone, brucine, a-naphthol, thymol, and aniline, all of which show^ 
 a red colour. One of the most delicate of these, showing one part in 
 a million, is aniline acetate. This reaction is particularly useful, as 
 it is peculiar to furfurol. 
 
 If equal parts of pure aniline, glacial acetic acid, and water are 
 boiled together for a few minutes, to destroy any fiarfurol that may 
 be present in the acetic acid, 1 c.c. of the reagent cooled and added 
 to 20 c.c. of a fluid containing furfurol shows a rose-pink colour in 
 ten to fifteen minutes. Filter-paper moistened with the reagent 
 held in the vapour arising from a boiling solution containing furfurol 
 gives the same colour change. 
 
 Cholic acid and sulphuric acid show with 1 part in 20,000 of furfurol 
 a crimson coloixr which forms the basis of Pettenkoffer's reaction for 
 bile acids. With urea nitrate furfurol solutions give a violet coloration 
 and deposit a black precipitate. Ammonium sulphide gives a yellow 
 crystalline precipitate. With phenylhydrazin fvirfurol forms furfurol- 
 phenylhydrazone (1 : 10,000). This is a crystalline body of a pale 
 yellow coloiir and conspicuously pearly lustre, which is insoluble in 
 water, ether, and cold alcohol, but is soluble in hot alcohol and in 
 ether. The purified product melts at 97° to 98° C. 
 
 Glucuronic (glycuronic) acid (CH0.(CH.0H)4C00H) is the only 
 important representative of a series of oxidation products intermediate 
 between the mono- and di-basic acids of the sugar group : — 
 
 CH2.OH C'HO COOH 
 
 (CH.0H)4 (CH.0H)4 (CH.OH), 
 
 I I I 
 
 COOH COOH COOH 
 
 Gluconic acid. Glucuronic acid. Saccharic acid. 
 
 It is formed in the animal economy as a product of the metabolism 
 of carbohydrates, being derived from dextrose by oxidation of the 
 
APPENDIX 455 
 
 primary alcohol group (CH2OH) to carboxyl (COOH). It is therefore 
 at once an aldehyde and an acid. It is met with in the shape of ether- 
 like derivatives, combined with substances containing an hydroxyl 
 group, in traces in the urine and blood. It has not yet been identified 
 as a plant product. 
 
 Glucuronic acid may be prepared by reducing saccharic acid. On 
 heating this substance on a water-bath for five or six hours saccharo- 
 lactonic acid (CgTTgO;) is formed. If this is reduced by sodium amalgam, 
 glucuronic acid results. Glucuronic acid is, however, most readily pre- 
 pared from Indian yellow or Purre, the magnesium salt of euxanthic 
 acid, obtained from the urine of cows fed on mango leaves, by hydro - 
 lysing it with dilute hydrochloric, or sulphuric, acid. 
 
 C'lgHigOii =Ci3Hg04+CgHj|^^,0; 
 
 Euxanthic Euxan- Glucuronic 
 acid. thon. acid. 
 
 Glucizronic acid is a syrupy liquid, soluble in water and alcohol, 
 but insoluble in ether. When its aqueous solution is boiled, evaporated, 
 or even allowed to stand, it readily loses the elements of water and 
 forms an anhydride or lactone (CgHgOg). 
 
 The anhydride form crystallises in needles, or plates, which have a 
 sweet taste, and melt at 167° C. It is insoluble in alcohol, but is 
 readUy soluble in water, forming a dextro-rotatory solution (aD = 
 -t-19° 25'). The solution prevents the precipitation of cupric salts 
 by alkalies, and exerts a powerful reducing action on Fehling's solution 
 when heated, and to a less extent in the cold. On being distilled with 
 hydrochloric acid, glucuronic anhydride yields furfvu-ol. 
 
 Glucuronic acid and its alkaline salts are dextro-rotatory {a.^ = 
 -f 35°). Most of its compounds are levo -rotatory. It is not fermented 
 by yeast, but is slowly decomposed by bacteria under suitable con- 
 ditions, yielding lactic and acetic acids. It reduces alkaline solutions 
 of copper (98-8 parts of Fehling's solution as compared with 100 by 
 dextrose), bismuth, mercury, and silver on heating, and when pure 
 in the cold. Glucuronic acid is the only substance, other than the 
 sugars, commonly occurring in the vu-ine which reacts with phenyl - 
 hydrazin. It forms a yellow crystalline compound reseiiibling gluc- 
 osazone, but the pure product has a melting-point of 114° to 115° C, 
 as compared with 204° to 205° C. for glucosazone. The impure product, 
 such as is obtained by treating urine with phenylhydrazin, does not 
 crystallise readily, and melts at about 150° C. Theoretically the 
 osazone should yield 18-4 per cent, of nitrogen, but it is difficult to 
 obtain a sufficiently pure product to make the estimation reliable. 
 With para-brom-phenylhydrazin, glucuronic acid forms a very charac- 
 teristic light yellow, crystalline compound, which, owing to its feeble 
 solubility, is most useful in separating glucuronic acid from the sugars. 
 It is readily soluble in acetic acid, but is only very slightly soluble in 
 hot water, benzol, ether, and absolute alcohol. The pimfied product 
 melts at 236° C, but the impure material prepared from the lu-ine 
 
456 GLYCOSURIA 
 
 generally melts at 200° to 216° C. A solution of 0-2 gram dissolved in 
 4 grams of pyridine and 6 grams of alcohol is levo-rotatory {ao = — 7° 25')- 
 Glucuronic acid may be set free from this compound by heating it 
 with acetic acid. On oxidising glucuronic acid with bromine, it yields 
 saccharic acid, and on being reduced with sodium amalgam it gives 
 gluconic acid. When boiled with caustic alkalies glucuronic acid forms 
 oxaHc acid, catechol, and other products. Glucuronic acid forms 
 potassium and sodium salts which crystallise in needles. The zinc, 
 eadmiiam, copper, silver, and calcium salts are uncrystallisable. Treated 
 with an excess of baryta water a solution of glucuronic acid, or one of 
 its inorganic salts, yields an insoluble yellowish-white to orange- 
 coloured precipitate of the basic barium salt, which can be employed 
 for the separation of glucuronic acid from urine. With quinine, 
 glucuronic acid forms an easily crystallisable salt which is useful in 
 preparing the acid from its organic compounds, and separating it from 
 the sugars. Lead acetate gives a white precipitate of the basic lead salt. 
 Glucuronic acid is precipitated froin acid solutions by basic lead acetate, 
 in contrast to the sugars which only separate out in an alkaline medium. 
 On being boiled with hydrochloric acid, glucuronic acid yields furfurol, 
 and hence gives the phloroglucin and orcin tests like the pentoses. 
 The furfiu-ol is, however, produced more slowly, so that a higher 
 temperature and more prolonged heating are necessary. 
 
 CgTJgOg = CjH.O^ + 2F2O + COo 
 
 Glucuronic Furfurol. Water. Carbon 
 anhydride. dioxide. 
 
 By combining the furfiu-ol with phloroglucinol the glucuronic acid 
 may be quantitatively estimated, 1 part of furfurol phloroglucide 
 corresponding to 3 parts of glucuronic anhydride. Carbon dioxide is 
 also liberated when glucuronic acid is treated with hydrochloric acid, 
 and may be used to estimate it in the presence of pentoses, 1 part of 
 carbon dioxide corresponding to 4 parts of glucuronic anhydride. 
 Glucuronic acid can be distinguished from the pentoses by the blue 
 substance, soluble in ether, formed when it is boiled with naphtho- 
 resorcinol and hydrochloric acid (ToUens). 
 
 Combined, Paired, or Conjugate Olucuronic Acids. — "When certain 
 substances that contain an hydroxyl group, and are only oxidised witli 
 difficulty, are introduced into the body they combine with dextrose to 
 form ether-like compounds. In these one end of the chain is shielded 
 from attack by the pairing substance, but the other is open to chemical 
 change. When this is oxidised gluciu-onic acid is fonned, and the 
 paired, or conjugate, glueuronate is excreted in the urine. 
 
 R.C.CH(CH.0H)2.CH.CH(0H)C00H 
 
 Paired glucuronic acid. 
 
APPENDIX 457 
 
 The nmnber of substances thus excreted in the urine in combina- 
 tion with glucuronic acid is very large. The most important are the 
 following : — 
 
 Isopropyl alcohol, methylpropyl carbinol, niethylhexyl carbinol, 
 tertiary butyl alcohol, tertiary amyl alcohol, pinacone ; 
 
 Chloral, butylchloral, bromal, dichloracetone ; 
 
 Benzene, nitrobenzene, aniline, phenol, resorcinol, thpnol, a- and 
 /3-naphthol ; 
 
 Tiirpentine oil, camphor, borneol, menthol, pinene, antipj-rin, &c. 
 
 Most of the compound glucvironates are of a glucosidal nature, 
 resembling the glucosides met with in plants. 
 
 R.O.CH(CH.OH)2.CH.CH.(OH).CH2(OH) 
 Glucoside. 
 
 This is shown by the fact that, like the latter, they are attacked, 
 and broken down, by appropriate glucoside-spHtting ferments, and 
 also that, like the glucosides, the conjugate gluciu-onates are hydro- 
 lised by mineral acids, A.-ielding gluciironic acid and the particular 
 alcohol from which they were formed. All conjugate glucuronic acids 
 do not, however, exliibit the characters of glucosidal compounds, for 
 some, such as urochloral acid and camphor-gluciu'onic acid, are capable 
 of directly reducing Fehling's solution, a reaction which is only ob- 
 tained with most conjugate glucuronic acids after the acid has been 
 set free by hydrolysis. This property appears to depend upon freedom 
 of the aldehyde group in the combined acid. Urocliloral acid (tri- 
 chlorethyl glucuronic acid) is excreted in the virine after large doses 
 of chloral hydrate (C.Clg.CHO). On being heated with a mineral acid 
 it jaelds trichlorethyl alcohol (C'.C'lgCHj.OH) and glucuronic acid. In 
 the same way camphor glucuronic acid appears in the m"ine after large 
 doses of camphor (C^uHjqO), and on hydrolysis \-ields camphoral 
 (CiqHjjO.OH) and glucuronic acid. In both cases the glucuronic acid 
 is combined with an alcohol, which, in the one instance, has been 
 derived by reduction, and, in the other, by oxidation within the body. 
 
 In addition to the non-nitrogenous conjugate glucuronic acids, a 
 nitrogen-containing variety, uramido glucuronic acid, has been de- 
 scribed. This on being heated with barium hydrate \-ields ammonia, 
 carbon dioxide, and glucuronic acid free from nitrogen. 
 
 Most conjugate gluciironic acids, and their alkaline compoiuids, 
 are easily soluble in water, and the former are also readily soluble in 
 alcohol and ether. The potassium salts can be cr^-stallised out from 
 an alcohol-ether extract of the urine after being set free from its com- 
 pounds. The gluciu'onates are as a rule readily precipitated by lead 
 acetate, basic lead acetate, and by lead acetate and ammonia. The 
 feebly soluble basic lead and barimn salts can be used to separate 
 glucuronic acid after it has been set free from its compounds, the acid 
 
458 GLYCOSURIA 
 
 being subsequently recovered by treating the compounds formed with 
 sulphuretted hydrogen, and sulphuric acid, respectively. A certain 
 amount of the conjugate glucvu-onic acid may be recovered from lorine, 
 by shaking an acidified alcoholic extract of the hydrolised lu-ine, with 
 a mixture of equal parts of alcohol and ether. Alkaloids form crystal- 
 line compounds with most conjugate glucuronic acids. 
 
 The conjugate giucuronates are levo -rotatory, and the presence of 
 small quantities in normal urine accounts for its slight levo -rotatory 
 power (a„= — 0-01° to 0-18°). The alkaline salts of glucuronic acid 
 are, however, dextro-rotatory, like the free acid. Conjugate gluciiro- 
 nates, like the acid itself, are not fermented by yeast. Many conjugate 
 glucuronic acids do not reduce alkaline sohitions of the heavy metals 
 until they have been decomposed by heating with an acid and the 
 glucuronic acid has been set free, but some, such as chloral, camphor, 
 menthol, turpentine, and indoxyl compounds reduce Fehling's solution 
 on simply boiling. The slight reducing power of normal urines is more 
 or less due to the presence of glucuronic acid compounds. Normal 
 urine also gives a reaction with phloroglucin and hydrochloric acid, 
 and the orcin reaction, after it has been boiled with 1 per cent, sulphuric 
 acid for one minute, for the same reason. The conjugate glucuronic 
 acids do not form crystalline compounds with phenyl hydrazin until 
 the glucuronic acid has first been set free. 
 
INDEX 
 
 Abdominal crises, 203 
 Abortion, 205 
 Aceto-acetic acid, 189, 450 
 
 — — conversion into, 110 
 
 — — estimation of, 115 
 — • — in urine, 107, 115 
 
 Acetone, 450 
 
 — bodies in urine, 104, 188 
 — • — source of, 208 
 
 — conversion in. 111 
 
 — estimation of, 105, 114 
 Acid, aceto-acetic, 107, 189, 450 
 
 — aposorbic, 450 
 
 — benzoic, 185 
 
 — Beta-oxybutyric, 104, 109, 113, 
 189, 449 
 
 — butyric, 449 
 — • crotonic, 110 
 
 — gluconic, 448 
 
 — glucuronic, 20, 64, 103, 396, 400. 
 454, 456 
 
 — glyceric, 447 
 — • glycollic, 447 
 
 — hippuric, 185 
 
 — homogentisic, 410 
 — • hydrochloric, 269 
 
 — intoxication, 211 
 
 — lactic, 185, 345, 448 
 
 — mucic, 58, 70, 450, 453 
 
 — oxalic, 185, 450-452 
 
 — phosphoric, 180 
 
 — picric, 37 
 
 — saccharic, 450, 452 
 
 — sulphiaric, 398 
 
 — tartaric, 159, 450, 452 
 
 — tartronic, 450 
 
 — tetra-oxyvaleric, 447 
 
 — tri-oxybutyric, 447 
 
 — uric, 184, 271 
 
 — xylonic, 64 
 
 Acidosis complicating diabetes, treat- 
 ment, 351-356 
 
 — in diabetes, 207-213 
 
 — prevention of, 301, 309 
 Acids, di-basic, of sugar group, 450 
 
 — mineral, 424 
 
 — monobasic, of sugar group, 447 
 
 — of the sugar group, 9, 10, 447 
 
 — organic, 425 
 — ■ oxidation, 17 
 
 Acini, secretion of, 142 
 
 459 
 
 Acromegaly, 235 
 
 — glycosuria in, 146 
 Addison's disease, 233 
 
 Adrenalin, glycosuria intensified by, 
 144 
 
 — hydriasis, 276 
 
 Age, influence of, in prognosis, 364 
 Air, fresh, 346 
 
 Albumen, estimation of, instrument 
 for, 97 
 
 — in diabetes, 187-188 
 
 — removal of, 80 
 Albuminuria, 204 
 
 Alcohol, energy value of, 326 ' 
 
 — influence on glycosuria, 163 
 — ■ nutritive value of, 293 
 
 Alcoholic beverages, 326 
 
 — glycosuria, 126 
 Aldoses, 4, 54 
 Alimentary dextrosuria, 162 
 
 — galactosuria, 167 
 
 — glycosuria, 155-168 
 
 — lactosuria, 167, 384 
 
 — levulosuria, 164, 370 
 
 — maltosuria, 168 
 
 — pentosuria, 168, 388 
 
 — saccharosuria, 168 
 Allialies, action of, 423 
 
 — in diabetes, 344 
 
 Alkaline earths, separation of sugar bv, 
 51 
 
 — solutions of copper, titration with, 
 79-91 
 
 — — of mercury, estimation with, 
 92 
 
 Alkaptonuria, 72, 409-411 
 Allen's method in tests for sugar, 32 
 Almen-Nylander's test for sugar, 34 
 Amblyopia, 206 
 Amenorrhcea, 205 
 Aminoglucose, 437 
 
 Ammonia in urine of healthy person, 
 183 
 
 — nitrogen in urine, 111 
 Ammoniacal copper method, Pavy's, 87 
 Amylolytic ferments, 11 
 
 Anasarca, 206 
 
 Aniline acetate test, 427 
 
 — dye tests for sugar, 38 
 Animal gum, 20, 72, 379, 445 
 Antipyrin in diabetes, 337 
 
460 
 
 GLYCOSURIA 
 
 Antiseptics, intestinal, in diabetes, 339- 
 
 342 
 Anti-syphilitic treatment of diabetes, 
 
 338 
 Aposorbic acid, 450 
 Appetite, increased, 203 
 
 — voracious, 258 
 Arabinose, 5, 6, 61, 63, 435 
 
 — optically active, 394 
 Arnold-Lipliawski test, 108 
 Arsenic in diabetes, 337 
 Arterio-sclerosis complicating diabetes, 
 
 204, 351 
 Arterio-sclerotic changes, 204 
 Asphyxial glycosuria, 125 
 Atrophy of pancreas, 220 
 Azoturia, 271 
 
 Bacteria, intestinal, 12 
 
 Bang's method of volumetric estima- 
 tion, 84-87 
 
 Barfoed's test for sugar, 45 
 
 Barium, separation of glucuronic acid 
 by, 67 
 
 Baths, warm, 346 
 
 Bauer's test for sugar, 58 
 
 Belladonna in diabetes, 336 
 
 Benedict's test for sugar, 33 
 
 Benzoic acid, 185 
 
 Benzoyl-chloride, 67 
 
 — separation of sugar by, 51 
 Benzyl-phenylhydrazin, 53, 56, 62, 70, 
 
 429 
 Beta-benzyl-phenylhydrazin, 59 
 Beta-napthyl-hydrazin, 53, 56 
 Beta-oxybutyric acid, 189, 449 
 
 — — conversions of, 110 
 
 — — estimation of, 113 
 
 — — in urine, 104, 109 
 Bial's modification test, 60 
 Bile in urine, 272 
 Bismuth test for sugar, 34 
 Black's method of conversion, 110 
 Blood, appearance of, like chocolate, 
 
 191 
 
 — in diabetes, 189-194 
 
 — injection of sugar into, 157 
 
 — normal, carbohydrates in, 18-20 
 
 — occult, in faeces, 270 
 
 — sugar in, 15, 19, 250, 254 
 
 — estimation of, 103 
 
 Boils complicating diabetes, 201, 349 
 Bondi's modification test, 109 
 Bones, diseases of, 206 
 Borchardat's test for levulose, 55 
 Bottger's test, 5 
 
 — (modified), 34 
 
 Brain, complications of, in diabetes, 
 206 
 
 — tumours causing glycosuria, 120 
 Braun's test for sugar, 37 
 
 Bread, 312 
 
 — diabetic, 320 
 
 — gluten, analysis of, 320 
 
 — white, analysis of, 320 
 Breath, sweet smell of, 202 
 Bremer's test, 193 
 Bromides in diabetes, 337 
 Brucin, 64, 67 
 
 Butyric acid, 449 
 
 Caffein glycosuria, 126 
 Calcium in diabetes, 187 
 
 — oxalate in urine, 272 
 Calculi, biliary, 173, 174, 200 
 
 — pancreatic, 221 
 Calories, carbohydrate, 290 
 Cancer of pancreas, 222 
 Cane-sugar, 442 
 
 — assimilation limit, 156 
 
 — in urine, 71 
 
 Carbohydrate constituent of proteins, 
 446 
 
 — metabolism, 14 
 
 — - — relation of pituitary to, 147 
 
 Carbohydrates, assimilation of, 10-18 
 
 — ■ chief source of energy, 289, 290 
 
 — classification of, 2 
 
 — digestion of, 10—18 
 
 — fatty acid relationships, 9 
 
 — groups of, 3 
 
 — in normal blood, 18-20 
 
 — in persistent glycosuria, 298 
 
 — percentage of, in foods, 318 
 Carbon atoms of monosaccharides, 3 
 
 — dioxide evolved, volumetric de- 
 termination, 95 
 
 — monoxide poisoning, 164 
 Carbuncles complicating diabetes, 202, 
 
 260, 349 
 
 Carlsbad salts in diabetes, 342 
 
 Castor-oil in diabetes, 342 
 
 Cataract, 206, 259 
 
 Catarrh, intestinal, complicating dia- 
 betes, 348 
 
 Cellulose, 11, 445 
 
 Cernelutti's method in tests for sugar, 
 32 
 
 Chlorides in urine, 186 
 
 Circulatory system complicating dia- 
 betes, 204 
 
 Claudication, intermittent, 202 
 
 Climate, warm, 346 
 
 Clothing, warm, 346 
 
 Codeine in treatment, 334 
 
 Cod-liver oil, use of, 326 
 
 Colour reactions of sugars, 426-428 
 
 Coma, diabetic, treatment, 351-356 
 
 — symptoms of, 208, 213-215 
 Conjugal diabetes, 340 
 Consanguinity, alkaptonuria and, 411 
 Constipation, 203, 348 
 
INDEX 
 
 461 
 
 Cooper-Lane test for inosite, 380 
 Copper, alkaline solutions of, titration 
 with, 79-91 
 
 — ammoniacal, Pavy's method, 87 
 
 — hydroxide in tests for sugar, 27 
 
 — separation of sugar by, 51 
 Cramp, 205 
 
 — nocturnal, 350 
 Creatinin in urine, 184 
 Crises, abdominal, 203 
 Crismer test for sugar, 38 
 Cromaffin tissue, 150 
 
 Crotonic acid, conversion into, 110 
 Cyanide process, Gerrard's, 83 
 Cystitis, 259, 349 
 Cysts of the pancreas, 221 
 
 Dbath-eate from diabetes, 360, 364' 
 Dermatitis, 259 
 Dextrin, 445 
 Dextrose, 4, 6, 402, 436 
 
 — assimilation limit, 156 
 
 — • excretion in depancreatised dog, 
 130 
 
 — in urine, 52 
 
 — percentage of, 374 
 
 — saccharic acid from, 452 
 Dextrosuria, 1 
 
 — diseases influencing, 1, 162 
 — - mixed levulosuria and, 371 
 
 — — pentosuria and, 396 
 
 — persistent, complicationsof, 196.257 
 
 — — • etiology of, 195 
 
 — — pathology of, 257 
 
 — — symptoms of, 194, 196, 257 
 D-fructose, 437 
 
 D-galactose, 439 
 D-glucosamine, 437 
 D-glucose, 436 
 
 Diabetes, anti-syphilitic treatment of, 
 338 
 • — bronzed, 244 
 
 — causes of, 132, 142 
 
 — coma in, 213-215 
 
 — — treatment of, 351-356 
 
 — complications of, 367 
 
 — conjugal, 340 
 
 — death-rate from, per 100,000 popu- 
 lation, 360 
 
 — dietetic treatment of, 306-330 
 
 — diseases of gastro-intestinal tract 
 in, 245-248 
 
 — — of kidneys in, 248 
 
 — — of liver in, 242-245 
 
 — disorders of nervous system in, 
 239-242 
 
 — energy requirement in, 303-304 
 
 — faeces in, 203, 263, 270 
 
 — fats in, 264 
 
 — following extirpation of pancreas, 
 132, 143 
 
 Diabetes, hepato-neurogenic theory of, 
 137 
 
 — infantile, treatment, 356-358 
 
 — insipidus, diagnosis of, 416 
 
 — • — • etiology of, 412 
 
 — — pathology of, 414 
 
 — — polyuria of, 413 
 
 — — symptoms of, 412 
 
 — — treatment of, 417 
 
 — intestinal antiseptics in, 339-342 
 
 — — fermentation and, 13 
 
 — levulosuria in, 370-376 
 
 — morbid changes in, 249 
 
 — oxaluria in, 404 
 
 — phloridzin in, 123 
 
 — pituitary gland in, 234-236 
 
 — puncture, 120-122 
 
 — relationship between pentosuria 
 and, 392 
 
 - — sex and race in, 195 
 
 — supra-renals in, 231-234 
 
 — theories of, 249 
 
 • — treatment of, by drugs, 334-338 
 Di-acetic acid in urine, 107 
 Diarrhoea, 203 
 Diastatic ferments, 11 
 Diet, carbohydrate-free, table, 311 
 
 — metabolism and, 282 
 
 — tables, 315 
 
 — test, 308 
 
 — vegetable, 329 
 
 Dietetic treatment of diabetes, 306-330 
 Differential density method, 93 
 Digestion, disordered, 269, 348 
 
 — of carbohydrates, 10-18 
 Digestive system, disorders of, compli- 
 cating diabetes, 202 
 
 Di-methylketone, 450 
 Di-phenylhydrazin, 53, 61, 63, 70, 429 
 Diphtheria, glycosuria in, 171 
 Disaccharides, chemical characters of, 7 
 
 — non-reducing, 8 
 
 — properties of, 7, 440 
 Diuretin glycosuria, 126 
 D-mannose, 438 
 Dreschel's gaunin method, 84 
 Drug, glycosuria, 122-126 
 
 Drugs in treatment of diabetes, 334— 
 338, 345 
 
 Ductless glands, glycosuria and, 120- 
 155 
 
 Dyspepsia complicating diabetes, treat- 
 ment, 269, 348 
 
 Dystrophia adij)oso-genitalis, 146 
 
 Ear, furunculosis of, 206 
 Eczema, 201, 259, 349 
 Eggs, food value of, 317 
 Einhorn's saccharimeter, 95 
 Electricity in diabetes, 346 
 Electrolysisingravimetric estimation, 91 
 
462 
 
 GLYCOSURIA 
 
 Embden's method of conversion, 110 
 Energy requirement in diabetes, 303 
 Entero -kinase, 127, 270 
 Enzymes secreted by pancreas, 133 
 Epilepsy, glycosuria in, 161 
 Erythro-dextrin, 379 
 
 — in urine, 72 
 Excitement, 346 
 Exercise, 346 
 
 Eye, accommodation of, defective, 206 
 
 F^CES, in diabetes, 203, 263 
 
 — occult blood in, 270 
 Family history, 195, 411 
 Fasting-purgation (Guelpa) treatment 
 
 of diabetes, 342-346 
 Fats, chief source of energy, 289-297 
 
 — neutral, composition, 264 
 
 — use of, 264, 297 
 Fatty acids, 9 
 
 Fehling-Soxhlet method of titration, 79 
 Fehling (Worm-Miiller) test for sugar, 
 
 5, 30, 33 
 Fehling's solution, gravimetric estima- 
 tion with, 90 
 
 — — estimations with, 93-104 
 
 — — in titration, 80-82 
 Fermentation, 6, 420 
 
 — alcoholic, 18 
 
 — test for sugar, 42 
 
 — - tests, quantitative, 93-117 
 Ferments, classes of, 11, 270 
 
 — glycolytic, 137 
 
 Ferrous thiocyanate indicator, 83 
 Fevers associated with glycosuria, 164 
 Fischer test for sugar, 39 
 Fish, food value of, 317 
 Flatulence complicating diabetes, 348 
 Folin method of estimation, 113, 114 
 Foods, carbohydrate percentage of, 10- 
 18, 318-320 
 
 — energy, 2S7 
 
 — experimental glycosuria and, 129 
 
 — fatty, value of, 317 
 ■ — levulose in, 373 
 
 — supply, sufficient, 282 
 
 — variety necessary, 300 
 Frohlich's syndrome, 146 
 Frommer's test for acetone, 106 
 Fructose in urine, 54 
 
 Fruit diet and pentosuria, 388 
 Fruits, value of, 319 
 Furfurol, 102, 103, 453 
 Furunculosis, 206, 260 
 
 Galactose, 6 
 
 — injection of, 158 
 
 — in urine, 69 
 
 — mucic acid from, 453 
 
 — separation from urine, 50 
 ^alactosuria, 387 
 
 Galactosuria, alimentary pathological, 
 
 167 
 Gall-stones, temporary glycosuria and, 
 
 173, 174, 200 
 Gangrene, 202, 259, 349 
 Gastritis, 203 
 Gastro-intestinal tract, diseases of, in 
 
 diabetes, 245-248 
 Gaunin method, Dreschel's, 84 
 Generative organs, female, in diabetes, 
 
 239, 347 
 Gerhardt's ferric chloride reaction, 107 
 Gerrard's cyanide process, 83 
 Glands, ductless, relation to glycosuria, 
 
 143-152 
 
 — — theory of correlation of, 149 
 
 — pathology of, and glycosuria, 120— 
 155 
 
 Glandular glycosuria, 126-142 
 Gluconic acid, 448 
 Glucosamine in urine, 72 
 Glucosazone, 431 
 Glucose in urine, 52 
 
 — injection of, 158 
 Glucoside, 402, 457 
 Glucosuria, 1 
 
 Glucuronates, compound, excretion of 
 
 399, 401, 403 
 Glucuronic acid, 20 
 
 — — chemistry of, 405 
 
 — — combined, conjugate, and 
 paired, 456 
 
 — — estimation of, 103 
 
 — ■ — in urine, 64, 396 
 
 — — origin of, 401 
 
 — — pathological excretion of, 400 
 
 — — recognition of, 405 
 Glyceric acid, 447 
 Glycerin aldehyde, 18 
 Glycerine, transformation of, 14 
 GlycocoU, 185, 448 
 Glycogen, 11, 13, 20, 379, 444 
 
 — in liver, 14-16 
 
 ■ — in muscle, 14, 16, 18 
 
 — in urine, 72 
 Glycollic acid, 447 
 Glycolysis, 136 
 
 Glycosuria, alimentary, 131, 160 
 - — cause of 365 
 
 — chronic, metabolic changes in, 293- 
 294 
 
 — — response to treatment in, 367 
 
 — diagnosis of, 261 
 
 — diseases influencing, 180 
 
 — drug, 122-126 
 
 — experimental, 120-154 
 
 — glandular, 126-142 
 
 — hepatic, 161 
 
 — in nervous diseases, 161 
 
 — intermittent, 172-176 
 
 — pancreatic, 128 
 
INDEX 
 
 463 
 
 Glycosuria, persistent, complications of, 
 201, 348-356 
 
 . — — general management of, 346- 
 348 
 
 — — organo-therapy in, 330-334 
 ■ — — prognosis in, 363-367 
 
 — — treatment of, hygienic, 346- 
 348 
 
 — — — preventive, 358-362 
 — ■ — — • prophylactic, 358, 362 
 
 — — — surgical, 362—363 
 
 — relation of ductless glands to, 143- 
 152 
 
 — • — of pancreas to, 143 
 
 — — of supra-renals to, 143-146 
 
 — theories of, 249 
 
 • — transitory, 169-172 
 
 — toxic, 125 
 
 — traumatic, 122 
 
 — vagabond, 170 
 
 — varieties of, 2 
 Grape-sugar, 436 
 
 Gravimetric estimation of reducing 
 
 sugars, 90 
 Guelpa treatment of diabetes, 342-346 
 Gum, animal, 20, 72, 379, 445 
 Gums, inflammation of, 348 
 
 — spongy, 202 
 
 Gunning's modification of Kjeldhahl's 
 process, 116 
 
 — — test, 107 
 
 HiEMOCHEOMATOSIS, 244 
 
 Haine's test for sugar, 33 
 Hart's method of conversion. 111 
 
 — modification of Folin's method, 115 
 Heart complications in diabetes, 351 
 
 — feeble, 204 
 
 — sugar consumption of, 135 
 Heptoses, 3, 4, 379 
 Hereditary diabetes, 195 
 Hexamethylenamine (urotropine), 342 
 Hexoses, 3, 6, 18, 436 
 
 Hippuric acid, 185 
 
 Homogentisic acid in alkaptonuria, 410 
 Hoppe-Seyler test for sugar, 37 
 Hydrazins, 6 
 
 — combinations with, 428 
 Hydrazones, 56, 429 
 
 — melting-points of, 430 
 
 — separation of sugar by, 52 
 Hydrochloric acid, 269 
 Hydrogen, reduction in, 91 
 Hyperglycsemia, 120, 251 
 
 — influence of pancreas on, 132 
 Hyperglycogenesis, 120 
 Hypochondriasis, glycosuria in, 161 
 Hypophysis in sugar metabolism, 146, 
 
 415 
 Hysteria, glycosuria in, 161 
 
 — lactosuria and, 387 
 
 I-ARABiNOSE in urine, 63 
 
 Impotence, 205 
 
 Indican in urine, 185 
 
 Indigo test for sugar, 37 
 
 Infants, diabetes in, treatment, 356- 
 
 358 
 Infections, various, in diabetes, 259 
 Infectious diseases causing glycosuria, 
 
 126, 170 
 Inosite, 446 
 
 — detection of, 380 
 Intestine, small, digestion in, 11 
 Inulin, 444 
 
 — value of, 303 
 Iodoform in diabetes, 339 
 
 — test for acetone, 106 
 lodometric method of titration, 89 
 Iron in diabetes, 187 
 Isolactose, 8, 442 
 
 Isomaltose, 8, 441 
 
 — detection of, 378 
 
 — in urine, 68 
 I-xylose, 436 
 
 Jaksch, von, test for sugar, 39 
 Jambul in diabetes, 344 
 Jaundice, temporary glycosuria in, 173 
 JoUes' method of estimation, 102 
 
 — test for pentoses, 60 
 
 Ketoses, 4, 54 
 
 Kidney disease, dextrosuria in, 163 
 
 — — in diabetes,' 204, 248, 350 
 Kidneys, polyuria and, 414 
 
 — sugar excretion of, 123 
 Kjeldhahl's process, 116 
 Knapp's method of estimation, 92 
 
 — solution, 100 
 Kowarski test for sugar, 41 
 
 Lactation, lactosuria in', 384, 385 
 Lactic acid, 185, 345 
 
 — — dextro-rotatory, 449 
 
 — — inactive, 448 
 Lactose, 7, 8, 441 
 
 — assimilation limit, 156, 158 
 
 — estimation of, 101 
 
 — in urine, 57 
 
 — separation from urine, 50 
 Lactosuria, alimentary, 384 
 
 — ■ — pathological, 167 
 
 — diagnosis of, 386 
 
 — spontaneous, 385 
 Laiose, 378, 439 
 
 — in urine, 70 
 Landwehr, 379 
 
 Lange test for acetone, 105 
 Langerhans" islands, function of. 140 
 
 — — ■ internal secretion of, 138. 142 
 
 — — pathology of, 226-230 
 L-arabinose in urine, 61 
 
464 
 
 GLYCOSUKIA 
 
 Lead, influence on glycosuria, 163 
 
 — separation of glucuronic acid by, 
 (37 
 
 — separation of sugar by, 49 
 Legal's test for acetone, 105 
 
 Legs, oedema of, complicating diabetes, 
 
 350 
 Le Nobel test for acetone, 105 
 Leucocytes, 190 
 Levulose, 4, 6, 19, 437 
 
 — assimilation limit, 156, 158 
 
 — detection of, 374 
 
 — estimation of, 99 
 
 — excretion in depancreatised dog, 
 130 
 
 — in urine, 54 
 
 — oxidation of, 17 
 
 — percentage of, 374 
 
 — source of, 372 
 
 — symptoms of, 374 
 Levulosuria, 1 
 
 — alimentary, 370 
 
 — — pathological, 164-167 
 
 — mixed dextrosuria and, 371 
 
 — spontaneous, 370 
 Lieben's iodoform reaction, 107 
 Light polarised, action of sugar, 6 
 Liver, diseases of, glycosuria in, 161 
 
 — — in diabetes, 242-245 
 
 — — levulosuria in, 165-167 
 
 — enlargement of, 204 
 
 — glycogen in, 14-16 
 
 — hypertrophy in glycosuria, 129 
 
 — theory of diabetes, 137 
 Lohenstein's saccharimeter, 95 
 Lung, gangrene of, 204, 260 
 Lymphocytes, function of, 15 
 L-xylose in urine, 63 
 
 Malaeia, glycosuria in, 170 
 Malfatti-Jager method of estimation, 
 
 112 
 Malfatti's test for sugar, 57 
 Malingering, lactosuria and, 387 
 Maltase, 157 
 Maltose, 7, 8, 11, 19^440 
 
 — detection of, 377 
 
 — estimation of, 101 
 
 — injection of, 158 
 
 — in urine, 67, 272 
 Maltosuria, 376-378 
 
 — alimentary pathological, 168 
 Mania, gtycosuria in, 161 
 Mannose, 6 
 
 — phenylhydrazone, 429 
 Massage, 346 
 
 Meat, food value of, 317 
 
 Medulla puncture causing glycosuria, 
 
 120 
 Melancholia, glycosuria in, 161 
 Mellituria, 1 
 
 Melting-points, 48, 42.3, 433, 435 
 Mental causes in diabetes, 196, 206, 346 
 Mercury, alkaline solutions of, estima- 
 tion with, 92 
 
 — tests for sugar, 36 
 Metabolism, carbohydrate, 14 
 
 — - — and ductless glands, 127 
 
 — inborn errors of, 393, 411 
 
 — in persistent glycosuria, 282 
 
 — secondary disturbances in, 366 
 Methylene blue reaction, 38 
 Methyl-phenylhydrazin, 53, 55, 62, 63, 
 
 69, 439 
 Methyl-phenyl-levulosazone, 56 
 Milk, food value of, 317 
 
 — - in lactosuria, 384 
 Milk-sugar, 441 
 
 — in urine, 57 
 
 Mineral acids, action of concentrated, 
 424 
 
 — — action of, dilute, 424 
 Minkowski's method in polyuria, 417 
 Mitscherlich's polariscope, 97 
 Molisch's test, 5 
 
 — — for colour reactions, 426 
 Monosaccharides, 3, 435 
 
 — - chemical characters of, 5 
 
 — rotatory power of, 6 
 Moore's test, 5 
 
 — — for sugar, 25 
 Morner's test for acetone, 108 
 Morphia glycosuria, 1 25 
 Morphine in treatment, 335 
 Mouth, dry, 202 
 
 Mucic acid, 70 
 
 — — from galactose, 450, 453 
 
 — — test, 58 
 Mumps, glycosuria in, 174 
 Muscle ferment, 136 
 
 — glycogen in, 14, 16, 18 
 
 Naphtho-eesoecixal test, 66 
 
 Necrosis of tissue, 259 
 
 Nephritis complicating diabetes, 350 
 
 Nerves, reflexes, 205 
 
 Nervous diseases, glycosuria in, 161 
 
 — — transitory glycosuria in, 169 
 
 — origin of glycosuria, 121 
 
 — system complicating diabetes, 205- 
 206, 239-242 
 
 Neumann's test for pentoses, 60 
 
 Neuralgia, 205 
 
 Neuritis complicating diabetes, 350 
 
 — multiple, 205 
 Neuropsychoses, glycosuria in, 161 
 Nitrogen in urine. 111, 115, 182 
 Nitro-Prusside test for acetone, 105 
 Nutrition, general, prognosis in, 364 
 
 Oatmeal cure, 301 
 
 — in persistent glycosuria, 300 
 
INDEX 
 
 465 
 
 Ochronosis, 412 
 
 Ocular changes, 206 
 
 CEdema complicating diabetes, 201, 350 
 
 Oils in food, 326 
 
 Opium in diabetes, 334 
 
 Optical characters, 421 
 
 Orcin test, 5, 427 
 
 — — for pentose, 59 
 Organic acids, action of, 425 
 Organo-therapy in persistent glycos- 
 uria, 330-334 
 
 Osazone formation, 46-48 
 Osazones, 52, 431 
 
 — melting-point determination, 433, 
 435 
 
 — purification of, 433 
 
 Osseous system, diseases of, compli- 
 cating diabetes, 206 
 Ottenberg's titration process, 117 
 Oxalic acid, 185 
 
 — — estimation of, 451 
 Oxaluria in diabetes, 404 
 Oxidation processes, 16-17 
 Oxide, cuprous, direct weighing, 91 
 Oxidising agents, action of, 424 
 Oxybutyric acid, 104, 109, 113 
 
 Paidose, 71, 379 
 Pancreas, atrophy of, 220 
 
 — calculi in, 221 
 
 — cancer of, 176, 222 
 
 — cell function, 133 
 
 — cysts of, 221 
 
 — - diseases of, dextrosuria in, 162 
 
 — enlarged, 203 
 
 — extirpation, effects of, 129-131, 142 
 
 — fatty degeneration of, 220 
 
 — function of, 132-136, 142 
 
 — glycosuric experiments and, 128 
 
 — hyperfunction of, 151 
 
 — in diabetes, 218-231 
 
 — ■ internal secretion of, 134 
 
 — lesions of, inflammatory, 175, 223, 
 401 
 
 — • relation of, to glycosuria, 143, 176 
 
 — sugar metabolism controlled by, 
 135 
 
 — transplantation of, 135 
 Pancreatitis, acute, 224 
 
 — forms of, 225 
 
 — glucuronic acid excretion in, 401 
 
 — interacinar, 226 
 
 — temporary glycosuria in, 175 
 Para-brom-phenylhydrazin, 53, 62, 63, 
 
 66, 68, 430 
 Paralysis, general, glycosuria in, 161 
 Parathyroids, 149 
 Patein-Dufau reagent, 58 
 Patient, social position of, importance 
 
 of, in diabetes, 367 
 Pavy's ammoniacal copper method, 87 
 
 Pavy's carbohydrate theory, 15 
 
 — solution, 101 
 Pentoses, 3, 6, 379, 405, 435 
 
 — assimilation of, 156 
 
 — estimation of, 102 
 
 — in urine, 59-61, 272 
 
 — varieties of, 61 
 Pentosuria, 1 
 
 — alimentary, 388 
 
 — — pathological, 168 
 
 — chronic, 390 
 
 — essential, 389 
 
 — etiology of, 392 
 
 — mixed dextrosuria and, 396 
 
 — origin of sugar in, 393 
 
 — prognosis of, 395 
 
 — spontaneous, 389 
 
 — symptoms, 391 
 
 — treatment of, 395 
 Pfliiger-Allihn's method, 90 
 Pharynx, congestion of, 202 
 Phenyl-alanin, 411 
 Phenylhydrazin, 58, 61, 63, 65, 68, 428 
 
 — test for sugar, 39 
 Phenylosazones, characters, chemical 
 
 and physical, 46-48 
 Phlegmons, 260 
 Phloridzin glycosuria, 122 
 Phloroglucin method of estimation, 102 
 
 — test, 5, 45, 426 
 Phosphates in larine, 271 
 Phosphoric acid, 186 
 Picric acid test for sugar, 37 
 Pieraerts' solution, 99 
 PinofE's test for levulose, 55 
 Pituitary gland in diabetes, 234-236 
 Pneumonia, 204 
 
 Polarisation, 421 
 Polariscope, 44 
 
 — estimation with, 96, 114 
 Polysaccharides, groups of, 8, 443 
 Polyuria in diabetes insipidus, 413 
 
 — symptom of diabetes, 258 
 Potassium, iodide of, in diabetes, 339 
 Poultry, food value of, 317 
 Pregnancy, lactosuria in, 384, 385 
 Protein energy value, 296 
 
 — requirement for average labourer, 
 286 
 
 Proteins, carbohydrate constituent, 446 
 
 — digestion of, 265 
 
 — percentage of, 294 
 
 Pruritis complicating diabetes, 201, 
 
 259, 349 
 Psoriasis complicating diabetes, 201 
 Puerperium, lactosuria in, 168 
 Pulse, regular, 204 
 Purdy's method of titration, 89 
 
 Qualitative tests of sugars in urine, 
 21-77 
 
 2g 
 
466 
 
 GLYCOSURIA 
 
 Quantitative tests of sugars in urine, 
 
 78-119 
 Quinine in diabetes, 338 
 — salt, 67 
 
 Renal system complicating diabetes, 
 204, 248, 350 
 » Resorcine test, 427 
 
 Respiratory system complicating dia- 
 betes, 204 
 
 Retinitis, 206 
 
 Ribose, 395 
 
 Rice, 303 
 
 Riegler's modification of phenyl- 
 hydrazin test, 42 
 
 — test for acetone, 109 
 
 Robert's differential density method, 93 
 Rona titration process, 117 
 Rotation, specific, of sugars, 6 
 Rothera's test, 106 
 
 Rubner's test for colour reactions, 428 
 — ■ — for sugar, 52, 57, 64 
 
 >SACCHAPac' acid from dextrose, 450, 
 
 452 
 Saccharimeters, 95 
 Saccharine solutions, specific gravity of, 
 
 420 
 
 — urines, chemical reactions of, 24 
 Saccharose, 8, 442 
 
 — injection of, 158 
 Saccharosuria. 387 
 
 — alimentary pathological, 168 
 Sachsse's method of estimation, 92 
 Safranin test for sugar, 38, 93 
 Sahli's concentrated solution, 89 
 
 — test in persistent glycosuria, 272 
 Salicylaldehyde test for acetone, 106 
 Salines, massive, or fasting-purgation, 
 
 Guelpa treatment, 342-346 
 Salkowski's method for pentoses, 60 
 
 — modification of phloroglucin test, 
 45 
 
 Salts, Carlsbad, in diabetes, 342 
 Santonin in diabetes, 344 
 Scherer's test for inosite, 381 
 Schistosoma japonica in the liver, 244 
 Schmitz method of conversion, 110 
 Schwartz's modification test, 109 
 Sciatica complicating diabetes, 350 
 Secretin, 127 
 Secretions, internal, 134, 138 
 
 — — and glycosuria, 127 
 Seegen's method in tests for sugar, 32 
 Seidel's test for inosite, 381 
 Seliwanoff's reaction test for levulose, 
 
 54 
 
 — test for colour reactions, 427 
 Seminose, 438 
 
 Senses, special, 206 
 Sex mortality, 364 
 
 Sexual excitement, 348 
 
 Shaffer's method of estimation, 113 
 
 Skin complication in diabetes, 348, 351 
 
 Sleeplessness complicating diabetes, 351 
 
 Soda salicylate in diabetes, 341 
 
 Sodium chloride glycosuria, 125 
 
 Solar plexus, diabetes following disease 
 
 of, 132 
 Spa treatment of diabetes, 347 
 Spinal cord, affections of, complicating 
 
 diabetes, 206 
 Starch, 443 
 
 — assimilation limit, 156 
 
 — digestion of, 11 
 Starvation glycosuria, 129 
 Stomach, dilatation of, 258 
 Stomatitis, 202 
 
 Stools in diabetes, 203, 263, 270 
 Stupor, glycosuria in, 161 
 Sucrose, 7, 442 
 Sugar, assimilation of, 155 
 
 — blood containing, 15, 19, 103 
 
 — destroyed by the pancreas, 133 
 
 — estimation of, instrument for, 97 
 
 — excretion of, 179, 365 
 
 — group, acids and derivatives of, 447 
 — • ■ — ■ dibasic acids of, 450 
 
 — — monobasic acids of, 447 
 
 — injected into circulation, 157 
 
 — - metabolism, controlled by pan- 
 creas, 135 
 — - origin of, in pentosuria, 393 
 
 — output in diabetes, 13 
 
 — renal excretion of, 123 
 
 — • series, acids and acid-derivations 
 of, 9-10 
 
 — source of, in organism, 14 
 
 — volumetric estimation of, 79-82 
 Sugars, alcoholic fermentation of, 18 
 
 — colour reactions of, 426-428 
 
 — division of, 2 
 
 — general properties of, 420-458 
 
 — gravimetric estimation of reducing, 
 90 
 
 — in urine, causes of, 126 
 • — — isolating, 49-52 
 
 — — normal, 21 
 
 — — qualitative tests, 21-77 
 
 — — quantitative tests, 78-119 
 
 — oxidising agents and, 17 
 
 — reactions of, 420-458 
 
 — reducing, estimation of, 78, 92 
 ■ — — properties of, 425 
 
 — rotatory power of, 6 
 Sulphates in diabetes, 186 
 Sulphuric acid in urine, 398 
 Sunshine, 346 
 
 Supra-renals, excretion of, and glycos- 
 uria, 121 
 
 — glycosuria and the, 143—146 
 
 — in diabetes, 231-234 
 
INDEX 
 
 467 
 
 Surgical transitory glycosuria, 171 
 Syphilis, 164 
 
 Taka-diastase in diabetes, 344 
 Tartaric acid, 159 
 
 — — forms of, 450, 452 
 Tartronic acid, 450 
 Teeth, carious, 202 
 
 Tests for sugar, classifying, 42-48 
 
 — — confirmatory or special, 49- 
 75, 426 
 
 — — general, 24-42 
 
 — — qualitative, 21-77 
 
 — — quantitative, 78-119 
 
 — . See also under Names of Tests 
 Tetra-oxyvaleric acid, 447 
 Tetroses, 3 
 
 Theobromine glycosuria, 120 
 Thirst, symptom of diabetes, 258 
 Thyroid gland disease, dextrosuria in, 
 163 
 
 — — in diabetes, 236-238 
 
 — — influence on carbohydrate 
 metabolism, 147 
 
 Titration quantitative fermentation 
 tests, 93 
 
 — with alkaline solutions of copper, 
 79-91 
 
 — with alkaline solutions of mercury, 
 92 
 
 Tobacco, use of, 347 
 Tonsils, abscess or gangrene of, 203 
 Toxic glycosuria, 125, 171, 211 
 Toxines inducing dextrosuria, 163 
 Traumatic glycosuria, 122, 161. 241 
 Trioses, 3 
 
 Tri-oxybutyric acid, 447 
 Trommer's test, 5, 25, 29 
 Tuberculosis, pulmonary, 204 
 Tumours, cerebral, causing glycosuria, 
 
 120 
 Tyrosine, 185, 411 
 
 Ulcers, perforating, complicatins dia- 
 betes, 202 
 Uranium nitrate in diabetes, 339 
 Urea in urine of diabetics. 183 
 Uric acid, 184 
 
 — — endogenous, 271 
 Urine, acetone bodies in, 104 
 
 — acid in, 212 
 
 — appearance of, 181 
 
 — analysis of, 270 
 
 — collection of, 23 
 
 — density of, 182 
 
 Urine, distillation of, 106 
 
 — estimation of, instrument for, 97- 
 99 
 
 — glucuronic acid in, 396-405 
 
 — in diabetes insipidus, 413 
 
 — indican in, 271 
 
 — in pentosuria, 391, 394 
 
 — isolating sugars from, 49-52 
 
 — nitrogen in, total, 115 
 
 — normal, sugar in, 21 
 
 — - physical characters of, 24 
 — • reaction of. 182 
 
 — reducing substances in, 71 
 
 — rotatory power of. 44 
 
 — sugar in. See under Sugars, &c. 
 
 — sulphates in, 271 
 
 — " pancreatic " reaction in, 273 
 
 — volume of, in persistent dextros- 
 uria, 181 
 
 Urines, abnormal, 23 
 
 — routine examination of, 73 
 
 — saccharine, chemical reactions of. 
 24 
 
 Urobilin in urine, 272 
 
 Urotropine, hexamethvlenamine. in 
 
 diabetes. 342 
 Urticaria complicating diabetes, 201 
 
 Vaccines in diabetes, 343 
 Valente's method in tests for sugar. 32 
 Vegetable diet, 299, 318, 329 
 Volumetric determination of carbon 
 dioxide evolved, 95 
 
 — estimation of sugar, 79-82 
 Vulva, pruritis t)f. complicating dia- 
 betes, 349 
 
 Weight, loss of, 258 
 
 Wender's test for sugar, 38 
 
 Williamson's test, 193 
 
 Wohlk's test for sugar, 57 
 
 Women, sterility of, in diabetes, 347 
 
 Wood-sugar, 436 
 
 Worm-Miiller test for sugar, 30, 33 
 
 Worry, mental, 346 
 
 Wouncls, failure to heal, 202, 259 
 
 Xanthin bases, 184 
 
 Xanthoma complicating diabetes, 201 
 
 Xylonic acid in urine, 64 
 
 Xylose, 5 
 
 — in urine, 63 
 
 Yeast fermentation. 6, 7 
 
 — in diabetes, 342 
 
 Printed by Ballantyne, Hanson <£r= Co. 
 Edinburgh &^ London 
 
COLUMBIA UNIVERSITY LIBRARIES 
 
 This book is due on "the date indicated below, or at the 
 expiration of a definite period after the date of borrowing, as 
 provided by the library rules or by special arrangement with 
 the Librarian in charge. 
 
 DATE BORROWED 
 
 DATE DUE 
 
 DATE BORROWED 
 
 DATE DUE 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 C28 ( lO- S3 ) lOOM 
 
 
 
 
RC660 
 C14 
 
 C • i. 
 
 Glycosuria and allied conditions, 
 
 e /"/