COLUMBIA llHHAI 
 
 HX0003178 
 
Columbia Slntoetsttp 
 
 mtljeCiipofBrmgork 
 
 College of $fjps;tctan$ anb burgeons 
 Htbrarp 
 
 %C66o 
 
Digitized by the Internet Archive 
 
 in 2010 with funding from 
 
 Open Knowledge Commons 
 
 http://www.archive.org/details/diabetesmellitusOOwill 
 
/IfJCIfc- <Zou&><r^ fcwjybs^Us 
 
 DIABETES MELLITUS. 
 
NUNQUAM ALIDD NATURA, ALIUD SAPIENTIA DIGIT. 
 
DIABETES^MELLITUS 
 
 AND 
 
 ITS TREATMENT 
 
 BY 
 
 E. T. WILLIAMSON, M.D. (Lond.), M.R.C.P. 
 
 MEDICAL REGISTRAR, MANCHESTER ROYAL INFIRMARY ; HON. MED. OFFICER, 
 
 PENDLETON DISPENSARY (SALFORD ROYAL HOSPITAL) ; ASSISTANT TO 
 
 THE PROFESSOR OF MEDICINE, OWENS COLLEGE, MANCHESTEE. 
 
 WITH EIGHTEEN ILLUSTRATIONS {TWO COLOURED). 
 
 EDINBURGH & LONDON 
 YOUNG J. PENTLAND. 
 
 1898 
 
%CMo 
 
 it 
 
 EDINBURGH : PRINTED KOR YOUNG .1. TEN'TLASD, 11 TEVIOT PLACE, AND 
 3S WEST S.MITIIFIELD, LONDON, E.C., BY MORRISON AKD G1BB LIMITED. 
 
 (All rights reserved.) 
 
PREFACE. 
 
 In the following pages I have endeavoured to present a more 
 detailed account of diabetes mellitus than is generally found 
 in text-books or systems of medicine. 
 
 The description of the symptoms and complications, and the 
 statistics with reference thereto, have been based chiefly on the 
 most severe forms of the disease, a large number of which have 
 come under my observation during the last ten years. 
 
 The literature of diabetes is so extensive that it did not 
 appear advisable to attempt to publish anything like a com- 
 plete bibliography, especially as this can be obtained so easily 
 by reference to the " Index Meclicus " and to the " Catalogue 
 of the Library of the Surgeon-General's Office, U.S. Army." 
 Nevertheless, at the end of each chapter, references have been 
 given to the more important papers. As a rule, only those 
 articles are referred to which have been actually consulted. 
 
 Since morbid anatomy has not revealed any constant or 
 characteristic changes in diabetes, and since the lesions found 
 post-mortem are (with the exception of -certain changes in the 
 pancreas and about the medulla oblongata) probably nearly 
 always secondary, or due to complications, it appeared more 
 convenient to briefly mention these lesions in the chapter on 
 pathological anatomy, and to give a detailed account of them 
 elsewhere. Thus it appeared better to describe the pathological 
 changes due to various complications just after the account 
 of the symptoms produced thereby. But the lesions found in the 
 pancreas and nervous system are fully described in the sections 
 
viii . PREFACE. 
 
 devoted to the relation between diabetes and changes in the 
 pancreas and in the nervous system. 
 
 In some parts of the work, the accounts of the conclusions 
 arrived at by various physiologists and physicians will be found 
 to be stated in almost the same words as in abstracts which 
 have appeared in several English medical journals during the 
 last nine years ; but in all such instances the abstracts have 
 been written by myself. 
 
 To the physicians of the Manchester Eoyal Infirmary — Dr. 
 J. D. Leech, Dr. J. Dreschfeld, Dr. Graham Steell, Dr. T. Harris, 
 Dr. J. S. Bury, and Dr. A. T. Wilkinson — I am greatly indebted, 
 for their kindness in allowing me to record any points of interest 
 which I may have observed in the hospital patients who have 
 been under their care, and for the opportunity of making the 
 pathological examination of the various organs in a considerable 
 number of cases. I am also indebted to Dr. E. Roberts and Dr. 
 Glascott, for kindly allowing me to examine and make sketches 
 of the fundus oculi in several cases of diabetic retinitis. In 
 the preparation of the work, I have further to acknowledge 
 my indebtedness to the excellent monographs on diabetes by 
 Frerichs, Seegen, Ebsteiu, and v. Noorden, and to the writings 
 of Pavy, Sir Win. Roberts, Dreschfeld, Minkowski, Schmitz, 
 Saundby, and many others. 
 
 To my friend, Dr. E. Brindley, I desire to express my best 
 thanks for his kindness in revising the proofs, and for many 
 valuable suggestions. 
 
 E. T. W. 
 
 Manchester, March, 1898. 
 
CONTENTS. 
 
 CHAPTER I. 
 
 INTRODUCTORY. 
 
 Definition of Diabetes Mellitus — Derivation of Name — Synonyms — His- 
 torical Note — Chemical Note on Carbohydrates . 
 
 CHAPTER II. 
 
 SUGAR TESTS. 
 
 Clinical Tests for Grape Sugar in the Urine — Moore's Test— Reduction Tests 
 — Trommer's Test— Fehling's Test — Picric Acid Test— Bismuth Tests— 
 Rubner's Test — Tests by Indigo-Carmine — Orthonitrophenyl-propiolic 
 Acid— Methylene Blue— Safframn— Fermentation Test— Phenylhydrazin 
 Test — Relative Sensitiveness of Several Tests — Detection of Lactose, 
 Leevulose, Cane Sugar, Pentose in the Urine — On the Detection of Small 
 Quantities of Grape Sugar in the Urine — Quantitative Estimation of Sugar 
 in the Urine — On the Presence of a Trace of Sugar in Normal Urine 
 
 CHAPTER III. 
 
 PHYSIOLOGICAL CONSIDERATIONS. 
 
 The Liver and Sugar Formation— Experiments of Claude Bernard, Pavy, 
 M'Donnell — Intravenous Injections of Sugar and Sugar Destruction in 
 the System — Experiments of Biedl and Kraus and F. Voit — Seegen's Views 
 on Sugar Formation — Excision of the Liver — Pavy's Recent "Views 
 
 CHAPTER IV. 
 
 EXPERIMENTAL DIABETES AND GLYCOSURIA. 
 
 Experiments of Claude Bernard, Pavy, Schiff, Arthaud and Butte on the 
 Relation of Injuries of the Nervous System to Glycosuria— Extirpation of 
 the Cceliac Plexus .....••• 63 
 
 CHAPTER V. 
 
 PIILOKIDZIN DIABETES . 
 
 70 
 
x CONTENTS. 
 
 CHAPTER VI. 
 
 EXPERIMENTAL PANCREATIC DIABETES. 
 
 PAGE 
 
 Experiments of Minkowski, v. Meriug, de Dominicis — Effects of Pancreas 
 Extirpation in Various Animals — Sugar Excretion after Total Extirpation 
 and after Partial Extirpation — Pancreatic Grafts — Effect of Various Articles 
 of Food on Sugar Excretion — Glycogen in the Organism — -Cause of Pan- 
 creatic Diabetes — Glycolytic Ferment — Pancreatic Diabetes and the 
 Nervous System ........ 73 
 
 CHAPTER VII. 
 
 GLYCOSURIA FROM VARIOUS CAUSES. 
 
 Alimentary Glycosuria — Puerperal Glycosuria — Glycosuria produced by Poisons 
 
 and Chemical Substances — Clinical Glycosuria . . . .85 
 
 CHAPTER VIII. 
 
 ETIOLOGY AND ETIOLOGICAL RELATIONS. 
 
 Etiology : General Relations — Sex — Age. Geographical Distribution : Racial 
 Influence — Rarity of the Disease — Is Diabetes becoming more Prevalent % 
 — Social Position — Heredity — Possibility of Infection — External Injury 
 — Mental Emotions — Obesity — Gout — Alcoholism — Influenza — Acute 
 Diseases — Exposure to Cold and Wet — Drinking of Cold Fluids — Malaria 
 — Lightning Stroke — Syphilis — Food — Diabetes and Pregnancy — Climac- 
 teric Diabetes. On the Relation between Diabetes Mellitus and Diabetes 
 Insipidus — Relation between Diabetes Mellitus and Diseases of the Liver 
 — Pathological Changes in the Liver. Relation between Diabetes Mellitus 
 and Affections of the Nervous System — Diabetes following Mental Dis- 
 turbances — Diabetes and Pathological Changes in the Nervous System — 
 Condition of the Nervous System in a Series of Consecutive Cases — Changes 
 found in the Nervous System — Diabetes and Diseases of the Nervous 
 System — Personal Observations — Changes in the Vagus Nucleus — Diabetes 
 and Acromegaly. Relation between Diabetes Mellitus and Lesions of the 
 Pancreas — Pancreatic Lesions — The Urine in Pancreatic Disease — Conclu- 
 sions — Diabetes Mellitus and Arterio-Sclerosis — Diabetes of Endogenous 
 Origin — Etiology of Diabetes in 100 cases. . . . .95 
 
 CHAPTER IX. 
 
 SYMPTOMATOLOGY. 
 
 Appearance of the Patient — Onset of Diabetes — The Urine — The Blood . 163 
 
 CHAPTER X. 
 
 SYMPTOMS, COMPLICATIONS, AND PATHOLOGICAL CHANGES IN CONNECTION 
 "WITH THE VARIOUS SYSTEMS. 
 
 Temperature — The Alimentary Canal, Liver, and Pancreas — The Lungs — The 
 Heart — The Kidneys — The Skin — The Eyes — The Sexual Organs and 
 Functions — The Ears — The Nervous System .... 202 
 
CONTENTS. xi 
 
 CHAPTER XI. 
 
 DIABETIC COMA. 
 
 PAGE 
 
 Exciting Causes — Analysis of Cases — Symptomatology — Forms of Coma — 
 Analysis of Symptoms — Pathology: Fat Emboli — Cardiac Failure — Uremia 
 — Acetonemia — Aceto-Acetic Acid — Acid Intoxication — Intestinal 
 Auto-Intoxication — Diagnosis — Prognosis ..... 271 
 
 CHAPTER XII. 
 
 PATHOLOGICAL ANATOMY. 
 
 Pathogenesis — Views of Bunge, Kaufmann, Seegen, Pavy, Ebstein . . 297 
 
 CHAPTER XIII. 
 
 FORMS OF DIABETES, TERMLNATIOK, PROGNOSIS, DIAGNOSIS. 
 
 Forms of Diabetes and Glycosuria — Severe. Mild, and Transitional Forms — 
 Diabetes Decipiens — Intermittent Form — Symptomatic Glycosuria — 
 Bronzed Diabetes — Phosphatic Diabetes — Termination — Duration — 
 Prognosis — Unfavourable and Favourable Indications — Diagnosis . 305 
 
 CHAPTER XIV. 
 
 THE TREATMENT OF DIABETES. 
 
 Prophylaxis — Management of a Case — General Principles of Treatment — 
 Dietetic Treatment — Mild Form — Severe Form — Articles of Diet — Bever- 
 ages — Mode of Life and Hygienic Considerations — Exercise — Massage — 
 Treatment by Alkali Mineral Waters — The Medical Treatment — Treat- 
 ment of Complications and Troublesome Symptoms — Treatment of 
 Diabetic Coma .....--■ 320 
 
 Appendix ......... 406 
 
 Indices .......... 409 
 
LIST OF ILLUSTRATIONS. 
 
 FIG. 
 
 PAGE 
 
 1. Fermentation test for sugar in the urine ; collection of gas in test-tube . 19 
 
 2. ■ , , expulsion of fluid from test-tube 20 
 
 3. U-shaped tube for detecting sugar in the urine 
 
 4. The phenyl-hydrazin test for sugar in the urine . 
 
 5. Apparatus for quantitative estimation of sugar, Pavy's method . 
 
 6. Hemorrhagic patch (shaded) in vagus nucleus in the medulla 
 
 7. Another section from the medulla of the same case . 
 
 8. Hemorrhagic patch in vagus nucleus .... 
 
 9. Section of pancreas from a case of diabetes, showing marked cirrhosis 
 
 10. Atrophied pancreas (actual size) in a case of diabetes 
 
 11. Author's methylene blue reaction for diabetic blood 
 
 12. Retinitis hemorrhagica diabetica ..... 
 
 13. Retinitis centralis punctata 
 
 1 4. Diabetic retinitis, combined form • . 
 
 15. Changes in the posterior columns of the spinal cord in diabetes . 
 16. 
 
 17. Temperature charts in diabetic coma .... 
 
 18. Granular casts in the urine in diabetic coma 
 
 22 
 24 
 43 
 135 
 135 
 136 
 144 
 147 
 192 
 232 
 233 
 234 
 246 
 248 
 284 
 286 
 
DIABETES MELLITUS. 
 
DIABETES MELLITUS. 
 
 CHAPTER I. 
 
 INTRODUCTORY. 
 
 Diabetes mellitus is a disease in which grape sugar is ex- 
 creted in the urine for a long period, — -often for months or 
 years, — and excreted in large quantity, or in a sufficient quantity 
 to give a reaction with the ordinary clinical tests for sugar. 
 But the term diabetes mellitus cannot be applied to all cases in 
 which sugar is detected in the urine. Sugar is occasionally 
 present in the urine, for a short period only, after febrile 
 attacks, acute diseases, and injuries, and as a result of the 
 action of certain toxic substances ; but these are cases of tem- 
 porary glycosuria, and not true diabetes mellitus. Then, again, 
 after a very large quantity of saccharine food has been taken, 
 a small amount of sugar appears in the urine of many apparently 
 healthy persons ; and if the sugar in the diet should exceed a 
 certain limit, a small quantity of sugar will always be found in 
 the urine. But these are instances of temporary alimentary 
 glycosuria. According to some authors, a minute trace of sugar 
 occurs in normal urine ; but, if present, it is certainly too small to 
 be detected by the usual clinical tests. In diabetes mellitus the 
 form of sugar excreted is glucose ; occasionally, however, other 
 forms of sugar are met with in the urine, such as lactose ; and 
 a few instances have recently been recorded in which pentose 
 has been found in the urine. Of course such cases are not 
 to be described as diabetes, but as cases of lactosuria or 
 pentosuria. 
 
2 INTR OD UCTOR Y. 
 
 The term glycosuria is frequently applied to those mild 
 forms of diabetes in which the presence of grape sugar in 
 the urine is almost the only indication of disease. The 
 limit between glycosuria and true diabetes is often difficult 
 to define ; but the factor of time is of the greatest importance. 
 All cases of permanent glycosuria would be regarded by many 
 authors as cases of diabetes nielli tus. Other authors would 
 apply the term chronic glycosuria to the milder forms of the 
 disease, in which sugar is present in the urine but other 
 symptoms absent, and they would reserve the term diabetes 
 for the more severe form, in which the specific gravity of the 
 urine is high, the sugar excretion abundant, and in which 
 diuresis, thirst, and other symptoms are present. 
 
 Derivation of name " Diabetes Mellitus." — Sia, through ; /?atvw, 
 I flow ; Sta^r^s, a siphon ; fxeXirra, a bee ; mellitus (L.), like honey. 
 
 Synonyms. — French, DiabHe sucre ; German, ZucJcerhrankheity 
 Zuckerliarnruhr, Harnzuckerruhr. 
 
 Historical Note. 
 
 The history of our knowledge of diabetes is of great interest. 
 According to Hirsch ( l ), the earliest account of the sweet urine of 
 the disease comes from India, and is to be found in the Ayur 
 Veda of Susruta. Certain symptoms of diabetes were known 
 to the Greek, Roman, and Arabian physicians, but Celsus first 
 described the disease, or rather its most obvious symptoms, in 
 the first century. The name diabetes (BLafirjrr)?) is said to have 
 been first used by Aretseus ( 2 ). He states that the disease 
 appears " to have got the name diabetes as if from the Greek 
 word 8ta{3r]Tr)<i, which signifies a siphon." He describes the 
 thirst and diuresis, and adds that " the emaciation is dreadful." 
 With reference to causation, he states, " some one of the acute 
 diseases may have terminated in this." References to the 
 disease also occur in the works of Galen, Avicenna, Avenzoar, 
 and other early writers. But none of these authors appear to 
 have been acquainted with the sweet taste of the urine. 
 
 An English physician, Thomas Willis ( 3 ), Sidley Professor 
 in the University of Oxford, was the first European writer who 
 mentioned the sweet taste of the urine. In his " Pharmaceutice 
 Eationalis," published in 1679, he gives such a good account 
 of the disease that several passages are worthy of quotation. 
 
HISTORICAL NOTE. 3 
 
 After alluding to the scanty references to the disease in the 
 writings of the older physicians, he states that " Galen knew 
 only two sick of it. But in our age, given to good fellowship 
 and guzzling down chiefly of unallayed wine, we meet with 
 examples and instances enough, I may say daily, of this disease." 
 
 " But yet as familiar as it is its causes and formal 
 
 reason notwithstanding is almost altogether unknown." Diabetes, 
 he tells us, is so called from Siafiaivco = transeo, or passing 
 through — too swift a passage of the matter that is drunk. 
 
 Willis states that the urine is " wonderfully sweet, as it 
 were imbued with honey or sugar," and that the patients are 
 exceedingly thirsty and quickly grow lean. He expresses his 
 opinion that diabetes is rather an " affection of the blood than one 
 of the reins " (kidneys), and he ascribes " the chiefest part of 
 the evil to the nervous juice." He attributes the disease to " an 
 ill manner of living, and chiefly an assiduous and immoderate 
 drinking of cider, beer, and sharp wines ; sometimes sadness, long 
 grief ; also, convulsive affections and other inordination and 
 depressions of the animal spirits are wont to beget and cherish 
 this morbid disposition." He refers to a patient who " con- 
 tracted an incurable diabetes," of which he died within a 
 month, and to another patient who recovered, and who was 
 treated with lime-water among other drugs. 
 
 On p. 83 he adds, in reference to the urine: "But why 
 that it is wonderfully sweet like sugar or honey, this difficulty 
 is worthy of explanation." The explanation was first given 
 one hundred years later by an English physician, Matthew 
 Dobson ( 4 ) of Liverpool. He pointed out that not only the 
 urine, but also the serum of the blood, was sweet to the taste. 
 By a number of experiments he showed that the urine contained 
 saccharine matter ; that, when evaporated down, a white cake was 
 left which tasted like sugar ; and that, when allowed to stand, 
 it fermented, and the sweet taste disappeared. He concludes 
 that the saccharine matter is a product of the animal economy, 
 and gives a list of drugs which had been used in the treatment ; 
 amongst them were Dover's powder and tinctura thebaica. 
 
 The first detailed account of diabetes was published in 1797, 
 by an English medical man, John Kollo ( 5 ), surgeon-general of 
 the Koyal Artillery. In his book entitled " An Account of Two 
 Cases of Diabetes Mellitus, etc.," he gives a general review of 
 the symptoms, and discusses the nature and treatment of the 
 
4 INTR OB UCTOR Y. 
 
 disease. Eollo was the first to point out the importance of diet 
 in the treatment. He advises "animal food, animal fats, and 
 confinement, with an entire abstinence from every kind of 
 vegetable matter," and states that the action " may be facilitated 
 by the daily use of alkalies, calcareous and testaceous substances." 
 " The quantity of animal food should be restricted, and given 
 in as small quantities as possible to satisfy the stomach " (p. 
 261). Among the "circumstances of life which have usually 
 preceded an attack of the disease," Eollo mentions " a free 
 use of fermented liquors, or a uniform participation of strong 
 vegetable food of the farinaceous kinds.' 5 
 
 Cullen ( 6 ) mentions the sweet taste of the urine, but stated 
 that he had met with one case in which the urine was insipid. 
 Latham ( 7 ), in 1811, recognised two forms of diabetes — the 
 saccharine and the serous. Gregory ( s ), in 1825, pointed out the 
 difference between diabetes mellitus and diabetes insipidus. 
 
 Bouchardt in 1841 introduced gluten bread as an article 
 of diet for diabetic patients. 
 
 In 1847, Claude Bernard commenced his series of famous 
 researches on sugar formation in the organism. These researches 
 led to the discovery of glycogen, and demonstrated the influence 
 of puncture of the floor of the fourth ventricle in the production 
 of glycosuria. 
 
 Ever since that time diabetes has been carefully studied 
 experimentally and clinically ; and the labours of Pavy, Seegen, 
 Kiilz, Brlicke, Ebstein, Cantani, Naunyn, Frerichs, Dickinson, 
 W. Eoberts, and numerous other British and continental 
 physicians and physiologists, have greatly extended our know- 
 ledge of the disease. 
 
 Eecently, the brilliant experiments of Minkowski and v. 
 Mering (recorded in 1889—1892) have proved that diabetes 
 mellitus can be produced experimentally in dogs by removal 
 of the pancreas. 
 
 Chemical Note. 
 
 The presence of sugar in the urine and excess of sugar in 
 the blood are the prominent features of diabetes mellitus. 
 
 Sugar belongs to the group of chemical substances known as 
 carbohydrates, and some of the more important members of the 
 group may be here briefly referred to. 
 
 Carbohydrates are compounds of carbon, oxygen, and 
 
CHEMICAL NOTE, 5 
 
 hydrogen ; the oxygen and hydrogen being present in the 
 same proportion as in water, i.e. two of hydrogen to one of 
 oxygen. 
 
 The following are the carbohydrates of most importance to 
 physiologists and physicians. Most of these substances have 
 active optical properties; those which deviate the plane of 
 polarised light to the right are marked + ; those which deviate 
 it to the left are marked — . 
 
 Carbohydkates. 
 I. The Amylose Group (C 6 H 10 O 5 ) n . 
 
 + Starch. + Glycogen. 
 
 Cellulose. + Dextrin. 
 
 II. The Sugar Group. 
 
 - Inulin. 
 
 Animal gum. 
 
 The Saccharoses, C 12 H 22 O n . 
 
 + Cane sugar or saccharose. 
 + Milk sugar or lactose. 
 + Maltose. 
 
 The Glucoses, C,,H 12 6 . 
 
 4- Grape sugar, dextrose, or glucose. 
 - Fruit sugar or laevulose. 
 + Galactose. 
 
 Pentoses (Pentaglucoses) (C 3 H 10 O 5 ). 
 
 + Arabinose. 
 + Xylose. 
 Mannite, formerly termed manna sugar, is not a sugar but a hexad 
 alcohol, C 6 H 8 (OH) 6 . 
 
 The sugars classed as glucoses are also known as Mono- 
 saccharides ; they have the formula C 6 H 12 6 . 
 
 By the combination of two molecules of glucoses, with the 
 loss of the elements of water, sugars are formed having the 
 formula C 12 H 22 O n ; these are known as Disaccharides — 
 
 C 6 H 12 6 + C 6 H 12 6 = C 12 H 22 O u + H 2 0. 
 
 The disaccharides, if boiled with dilute acids, take up water 
 and form two molecules of monosaccharides. 
 
 By the action of dilute sulphuric acid on proteids, or by 
 ferment action, Pavy l has been able to obtain a substance which 
 he regards as a sugar. It reduces Fehling's solution, and with 
 phenylhydrazin forms a crystalline osazone. It is optically 
 inactive, however, and does not ferment with yeast. 
 
 1 Pavy, "The Physiology of the Carbohydrates," London, 1894, pp. 28-57. 
 
INTRO D UCTOR Y. 
 
 1. HlRSCH . . . . 
 
 2. Adams, F. . . , 
 
 3. Willis, Thos. . . 
 
 4. Dobson, Matthew 
 
 5. Eollo, John . . 
 
 6. Cullen . 
 
 7. Latham, J. . . 
 
 8. Gregory . . . 
 
 REFERENCES. 
 
 " Handbook of Geographical and Historical 
 Pathology," New Syd. Soc, London, 1885/ 
 vol. ii. p. 643. 
 
 " Extant Works of Aretseus," Syd. Soc, London, 
 1856, p. 339. 
 
 " Pharmaceutics Eationalis," 1679, p. 79. 
 
 " Medical Observations and Inquiries," London, 
 1779, Second Edition, vol. v. p. 298. 
 
 "An Account of Two Cases of Diabetes," Lon- 
 don, 1797. 
 
 " First Lines of the Practice of Physic," Edin- 
 burgh, 1791, vol. iv. p. 85. 
 
 " Facts and Opinions concerning Diabetes," 
 London, 1811. 
 
 " Theory and Practice of Physic," London, 
 1825. 
 
 For a detailed account of the history of our knowledge of diabetes, 
 see Salomon, " Geschichte der Glycosurie," Deutsches Arch. f. Jclin'. Med., 
 Leipzig, Bd. viii. S. 489. 
 
CHAPTER II. 
 
 SUGAE TESTS. 
 
 I. Clinical Tests for Geape Sugar in the Urine. 
 
 The usual tests for sugar are so well known, that it may at 
 first sight seem superfluous to devote any space to their con- 
 sideration. 
 
 But, whilst the detection of grape sugar in the urine, in a 
 well-marked case of diabetes mellitus, is a very simple matter, 
 on the other hand, a very careful examination of the urine is 
 often necessary in order to decide whether a trace or small 
 quantity of sugar is or is not present. 
 
 On commencing to examine a specimen of urine, if we find 
 the specific gravity high — 1028 or more — suspicions of the 
 presence of grape sugar ought to be roused. The probability of 
 the presence of grape sugar will be great if a sample of urine 
 having the above mentioned specific gravity be clear and pale, 
 if the quantity of urine be not diminished, and if the patient 
 be not suffering from any febrile affection. If, however, the 
 quantity of urine be diminished, or if the urine be high-coloured, 
 or turbid and loaded with urates, or if the patient be suffering 
 from some febrile affection, then a high specific gravity is no 
 indication of the presence of sugar. 
 
 On the other hand, a low specific gravity does not exclude 
 the possibility of the presence of sugar. It shows, however, 
 that the case is not one presenting the typical symptoms of 
 diabetes mellitus ; but a small quantity of sugar may be present 
 when the specific gravity is only 1010 or 1015. 
 
 Of course, more reliable indications of the presence of grape 
 sugar are required, and the following are the chief chemical tests 
 which are employed clinically : — 
 
 Moore's test. — When saccharine urine and liquor potassse are 
 mixed in equal quantities and the mixture boiled, it darkens 
 
8 SUGAR TESTS. 
 
 in colour and finally becomes brown — about the colour of 
 brandy. 
 
 This test is now rarely used. As pointed out by Sir William 
 Eoberts Q-), Moore's test is wanting in delicacy : a clear reaction is 
 not obtained until the amount of sugar reaches 0'3 per cent. 
 Further, all high-coloured urines of high specific gravity darken 
 when treated with liquor potassa?. Also albuminous urines 
 darken when boiled with liquor potassa?, and the change is 
 especially marked if the liquor potassa? should contain lead as an 
 impurity. This impurity is often present when the liquor 
 potassa? has been kept in an ordinary white glass bottle for some 
 time. 
 
 Reduction tests. — Grape sugar has a powerful reducing action 
 on certain metallic salts in alkaline solution. The oxides of 
 these salts are reduced to a lower degree of oxidation, or the 
 metal itself is thrown down. Oxygen is not evolved, but it 
 unites with glucose, and various products of the oxidation of 
 sugar are formed. 
 
 Thus cupric oxide is reduced by glucose to cuprous oxide — 
 
 2 CuO = Gu 2 + 0. 
 Bismuthic oxide is reduced to metallic bismuth — 
 Bi 2 3 = Bi 2 + 3 . 
 
 Grape sugar has also a reducing action on certain organic 
 substances in hot alkaline solutions, and the colours of these 
 solutions are changed. Thus, picric acid is reduced to picramic 
 acid, and a dark brown coloration is produced ; methylene blue, 
 saffranin, and indigo - carmine are decolorised ; orthonitro- 
 phenylpropiolic acid is reduced to indigo-blue. 
 
 Trommers copijer test. — A few drops of a copper sulphate 
 solution are added to the suspected urine (about 1 drm.) in a 
 test tube. Liquor potassa? (about half a drm.) is then added, 
 and the mixture warmed. If sugar be present, a yellow pre- 
 cipitate of cuprous hydroxide, or a red precipitate of cuprous 
 oxide, will be produced. The test is now seldom employed in 
 this form in England, though it is still largely used in some 
 parts of Germany. 
 
 Fehling's test. — This is a much more satisfactory method of 
 applying the copper test for the detection of sugar. Fehling's 
 copper solution is prepared as follows : — 
 
 (a) 3 4' 6 4 grrns. of pure crystallised copper sulphate are 
 
FEHLING'S TEST. g 
 
 dissolved in warm distilled water, and the solution when cold 
 made up with distilled water to 500 c.c. 
 
 (b) 175 grms. of crystallised Eochelle salts (potassium and 
 sodium tartrate) are dissolved in about 300 c.c. of hot water ; 
 100 c.c. of caustic soda, having a specific gravity of 1'34, are 
 added, and the mixture is diluted with water up to 5~00 c.c. 
 Equal parts of the solutions (a) and (b) are mixed together. 
 Since Fehling's solution is apt to decompose in time, the two 
 solutions (a) and (b) may be kept separate, and mixed in equal 
 quantities when required for use. 
 
 In testing for sugar, a test tube is filled for three-quarters 
 of an inch with Fehling's solution, which is first boiled alone, in 
 order to see if the solution has changed by keeping. When 
 Fehling's solution has been kept for a long time, it decomposes 
 partially, and oxide of copper may be thrown down by boiling, 
 even when no urine is added. If now no oxide of copper is 
 thrown clown on boiling, the Fehling's solution has not cleterior- 
 ated by keeping, and is therefore suitable for use. A few drops 
 of the suspect urine are added, and the mixture is raised to 
 the boiling point ; yellow hydroxide, or red oxide of copper, is 
 thrown down if sugar be present. When the amount of sugar is 
 small, the precipitate is yellow ; when large, the precipitate is 
 red. If the quantity of sugar be very small, more than a few 
 drops of urine must be added before the copper oxide is pre- 
 cipitated ; but the quantity of urine should never be in excess of 
 the Fehling's solution. When only a small amount of sugar is 
 present, often there is no reduction of copper on boiling ; but a 
 greenish precipitate, or an opacity, appears when the test tube 
 cools. If no oxide of copper is thrown down on boiling, and if 
 no greenish precipitate appears when the mixture cools, then 
 sugar is absent ; or, if present, the quantity is so small as to be 
 of no practical importance. According to Sir William Eoberts, 
 no quantity above a fortieth of a grain per cent, can escape 
 detection by this method. 
 
 The three precautions in performing the test are, (1) to boil 
 the Fehling's solution alone before adding the urine, (2) to see 
 that the Fehling's solution is always in excess of the urine, (3) 
 to avoid prolonged boiling of the mixture. 
 
 If non-saccharine urine, loaded with uric acid, be boiled with 
 Fehling's solution for a considerable time, a brownish deposit may 
 be produced, and the mixture may assume a muddy, dirty fawn 
 
io SUGAR TESTS. 
 
 appearance, though no sugar be present. This turbidity is 
 probably produced by the excess of uric acid, but it only occurs 
 after prolonged boiling. 
 
 Of all the tests for sugar, Fehling's solution is the most 
 generally useful : and if the above mentioned precautions be taken, 
 it is, in most cases, very reliable. For clinical work, usually no 
 other test is required. If a negative result is obtained with 
 Fehling's test, the urine may be declared free from sugar, or free 
 from any quantity of sugar which would be of clinical import- 
 ance. If, on the other hand, a distinct precipitate of oxide of 
 copper occurs, and if there are other indications of diabetes, the 
 test is also sufficient. But there is a third class of cases, not 
 infrequently met with, which is liable to cause trouble, especi- 
 ally in the careful examination of urine for life assurance. 
 In these cases there are no indications of diabetes except the 
 reduction of Fehling's solution, and this is often slight. Fre- 
 quently, after the mixed urine and Fehling's solution have been 
 raised to the boiling point, the fluid is unchanged at first ; but 
 in a short time, as the mixture cools, a greenish precipitate 
 appears. Then the question arises, Is this slight reduction 
 due to sugar or not ? When the quantity of sugar in the 
 urine is small, its detection with certainty is not always easy, 
 since certain normal ingredients of the urine, such as uric acid, 
 when present in excess, are apt to give a slight deposit in the 
 Fehling's solution, especially if the precautions (2) and (3) above 
 mentioned are not observed. 
 
 The following is a list of some of the substances of clinical 
 importance which reduce Fehling's solution : — 
 
 Dextrose, laevulose, maltose, lactose, galactose, xylose, ara- 
 binose, pentose, glycuronic acid, glycosuria acid, hippuric acid, 
 uric acid, creatin, creatinine, xanthin, and alkapton or pyro- 
 catechin. 
 
 Glycuronic acid, C 6 H 10 O 7 , is said to occur in the urine after 
 the administration of certain drugs — morphine, chloroform, 
 chloral, butylchloral, nitrobenzene, camphor, curare, sodium sali- 
 cylate, salol, benzoic acid, etc. As Ashdown and others have 
 pointed out, the reduction of Fehling's solution by the urine, 
 which sometimes occurs after the administration of many of these 
 drugs, is probably clue to glycuronic acid, and not to sugar, as 
 was formerly supposed. Minute traces of glycuronic acid 
 compounds are said to exist in normal urine as indoxyl- and 
 
FEHLINGS TEST. n 
 
 skatoxyl-glycuronic acids, and in combination with urea as uro- 
 glycuronic acid. 
 
 Glycuronic acid gives a reduction of Fehling's solution on 
 heating, and precipitates bismuth from a hot alkaline solution of 
 the salts of this metal. It is distinguished from grape sugar by 
 not fermenting under the action of yeast. When the phenyl- 
 hyclrazin test is applied to a solution containing glycuronic acid, 
 small yellow needle shaped crystals are obtained, somewhat 
 resembling those produced by grape sugar (see p. 24). This 
 compound of phenylhydrazin and glycuronic acid has a melting 
 point of 114° to 115° C, whilst the compound of phenyl- 
 hydrazin and glucose (phenylglucosazone) has a melting point of 
 205° C. 
 
 Glycuronic acid may therefore be a source of error in urine 
 testing. If it exist in normal urine, as some believe, the quantity 
 present is so small as not to interfere with the usual clinical tests 
 for sugar. When present in sufficient quantity to give a reaction 
 with the clinical tests for sugar, it is generally due to the 
 administration of some of the drugs above mentioned, and if 
 these be discontinued the excretion ceases. A case has been 
 recorded by Ashdown ( 2 ), however, in which glycuronic acid was 
 present in considerable quantity in the urine of a healthy man, 
 who was not taking any drugs. 
 
 Alkapton or pyrocatechin ( 3 ) is another body which occa- 
 sionally, though very rarely, occurs in the urine, and is liable 
 to lead to confusion in a hasty examination, since it reduces 
 Fehling's solution. When this substance is present (alkaptonuria), 
 the urine is amber or straw coloured, and the specific gravity 
 1010 to 1020; it reduces Fehling's solution even when slightly 
 warmed, but gives no reaction with Nylander's bismuth test for 
 sugar. The urine rapidly undergoes ammoniacal fermentation ; 
 after a few hours it becomes greenish brown on the surface, and, 
 on shaking, the colour of the urine passes from dark brown to 
 black. This change occurs at once if a few drops of ammonia or 
 liquor potass^ be added to, and well mixed with the urine. 
 On the addition of yeast to urine containing alkapton, no 
 fermentation occurs. 
 
 Modifications of Fehling's test. — Since Fehling's solution may 
 be reduced by other substances besides sugar, the following 
 modifications of the test have been proposed : — 
 
 (a,) Fehling's solution in the cold.— Sugar reduces Fehling's 
 
12 SUGAR TESTS. 
 
 solution in the cold, whilst other " reducing " substances do not. 
 This is a useful control test. The suspected urine is mixed with 
 Fehling's solution, and allowed to stand for six to twenty-four hours 
 without warming ; a deposit of oxide of copper occurs if sugar be 
 present. But Fehling's solution is not reduced in the cold if only 
 a very small quantity of sugar be present. A 0*1 per cent, watery 
 solution of grape sugar gives a very slight deposit, and one 
 containing 0"05 per cent, gives no reduction (Seegen( 4 )). The 
 test is also less sensitive than the fermentation test. When the 
 urine has contained a very small quantity of sugar, I have been 
 able to obtain evidence thereof by the fermentation test, and by 
 the phenylhydrazin test when there has been no reaction with 
 Fehling's solution in the cold. 
 
 (b) Worm-Midlers method ( 5 ). — According to Worm-Muller, 
 the possibility of an excess of uric acid in the urine, causing any 
 reduction of Fehling's solution, may be excluded, by boiling the 
 two fluids separately, and mixing after twenty-five seconds, 
 whereby the temperature of the mixture sinks to 60° to 70° 0. 
 He believes that an excess of uric acid only reduces Fehling's 
 solution at boiling temperature. (Seegen's researches, however, 
 do not confirm this view.) Worm-Muller has proposed the 
 following method of performing the test : — ■ 
 
 Five c.c. of urine are boiled in a test tube. 
 
 A mixture of 1 to 3 c.c. of a 2 - 5 per cent, copper sulphate 
 solution, and 2 - 5 c.c. of an alkaline solution of Eochelle salts 
 (prepared by dissolving 100 grms. of tartrate of potash and 
 soda in 1000 c.c. of normal caustic soda solution), is also boiled 
 at the same time. 
 
 Twenty to twenty-five seconds after both have been brought 
 to the boiling point, they are mixed together. If sugar be 
 present, copper oxide is precipitated ; if small in amount, the 
 precipitate may not occur for five or ten minutes. 
 
 (c) Fehling's test after filtration through animal charcoal. — 
 Trommer's test is a very delicate one, and by this means Seegen 
 can detect with certainty 0*3 mgrms. (or 0"0046 grs.) of sugar 
 dissolved in ten thousand times the amount of fluid. This great 
 delicacy, however, only holds good so long as we have to do 
 with a watery solution of sugar. For the detection of small 
 quantities of sugar in urine, Trommer's test is not so delicate, 
 because (1) urine contains certain substances (colouring matter, 
 creatine) which tend to prevent the suboxide of copper, when 
 
MODIFICATIONS OF FEHLING'S TEST. 13 
 
 formed, from being precipitated ; (2) uric acid also reduces 
 copper salts slightly. Urine containing a large amount of uric 
 acid acts on Fehling's solution like urine containing O'l or 
 0'2 per cent, of sugar. 
 
 Seegen ( 6 ) removes the above mentioned substances from the 
 urine by repeated filtration through animal charcoal, and so 
 obtains a watery solution of sugar. Animal charcoal has the 
 property of retaining most of the constituents of urine, more 
 especially the colouring matter and uric acid. 
 
 One or two ounces of urine are filtered several times throug-h 
 charcoal, until the fluid is completely colourless. This occurs in 
 a few minutes. Then the charcoal on the filter is washed with a 
 little distilled water, and to this water, when filtered, Trommer's 
 test is applied. In this way 0"01 per cent, sugar may be 
 detected. If the urine contains a little more sugar, OT to - 2 
 per cent., the water flowing off from the second and third 
 washing acts even more energetically upon Fehling's solution 
 than the first. In normal urine no change occurs, or only a 
 slight turbidity is produced. 
 
 Seegen ( 7 ), in the last edition of his book, describes the test 
 a little differently. He filters the urine several times through 
 animal charcoal, until it is decolorised completely. Then the 
 animal charcoal is washed with distilled water, and the washings 
 and the filtrate are separately tested with Fehling's solution for 
 sugar. He points out an important precaution with regard to 
 his method, in order that success may be obtained, namely, the 
 animal charcoal used ought to be very good blood charcoal. 
 
 Sir William Eoberts ( 8 ) has recently described a modification 
 of this method. He points out that urates, uric acid, and other 
 normal constituents of the urine, which have more or less power 
 of reducing Fehling's solution, are removed by the charcoal ; but 
 sugar, if any be present in the urine, passes freely through the 
 charcoal, and is found in nearly undiminished proportions in the 
 filtrate. Hence urine, after filtration through animal charcoal, 
 is in a peculiarly favourable condition for giving a clear and 
 definite response to the copper test. He prefers the following 
 method of applying the test: — "A test tube is charged with 
 Fehling's solution to the depth of about a quarter of an inch, and 
 the filtrate is added to the depth of about two inches, and the 
 two fluids well mixed, so as to obtain a uniform coloration. The 
 flame of a lamp is then applied to the upper half of the column 
 
i 4 SUGAR TESTS. 
 
 of liquid, and this is briskly boiled for a couple of seconds. The 
 tube is then held up to the light. If sugar be present, the 
 boiled upper half of the column is soon seen to change — it loses 
 its blue colour and assumes a yellowish tinge, and the earthy 
 phosphates, which are thrown down in light flocks by the alkali 
 of the test, are tinted more or less of a gold colour by the 
 precipitation on them of the yellow suboxide of copper. Mean- 
 while the lower unheated portion of the column remains, with 
 its blue colour unchanged." 
 
 The filtration through animal charcoal is a method which 
 requires attention, at intervals, for some time, and hence is a 
 rather tedious method for the medical practitioner. Also some 
 care is required in the choice of a specimen of animal charcoal. 
 Sometimes impurities are removed from the animal charcoal, 
 which interfere greatly with the testing of the filtrate for sugar. 
 
 This method is of service when we wish to exclude excess of 
 uric acid as a source of fallacy in testing for sugar. 
 
 (d) Haines' fluid. — Haines ( 9 ) employs a copper solution of 
 the following composition in place of Fehling's solution : — 
 
 Pure copper sulphate . . .30 grs. 
 
 Distilled water . . . . £ oz. 
 
 Make a perfect solution, and add pure 
 
 glycerin . . . . ^ oz. 
 
 Mix thoroughly, and add 5 oz. of liquor potassse. 
 
 In using this solution, 1 drm. is placed in a test tube and 
 gently boiled ; six to eight drops of urine are then added, and the 
 solution again gently boiled. If sugar be present, a precipitate 
 of copper oxide is thrown down. The test solution is said to be 
 very stable, and to keep for a long time. 
 
 (e) Pavy's modification of Fehling's solution. — Another sub- 
 stitute for Fehling's solution is Pavy' fluid, which is of great 
 service in the quantitative estimation of sugar in the urine. It 
 will be found described on p. 42. 
 
 (/) Fehling's test he/ore and after fermentation ; Worm-Mullcr's 
 method. — Worm-Midler has recommended a combination of 
 fermentation with Fehling's test. 
 
 The urine is placed in a flask with yeast to ferment. If the 
 fluid reduces Fehling's solution before the addition of yeast, 
 and after the action of yeast for some time fails to do so, 
 then sugar is indicated. The test is conclusive, providing all 
 
TICR1C ACID AND BISMUTH TESTS. 15 
 
 the sugar has fermented. But Seegen points out that some- 
 times this may not occur for thirty-six or forty- eight hours ; and 
 when very minute quantities of sugar are present, fermentation 
 may not be completed even at this time. Still, the test is of 
 great service. A total disappearance or a very great diminution 
 of the reducing power of the urine, after the action of yeast, 
 indicates the presence of glucose. 
 
 Picric acid test of Braitn and Johnson ( 10 ). — To a drachm of 
 acid urine add its own volume of a saturated solution of picric 
 acid and h drm. of liquor potassa?, and boil for a few seconds. 
 An ordinary test tube, half an inch in diameter, is used in 
 performing the test. If, when the tube is held up to the light, 
 a bright red colour is visible through the middle of the column 
 of liquid, no sugar is present. As little as 2 grs. of glucose to 
 the ounce will render the liquid so dark that no light is trans- 
 mitted through the middle of the tube. The urine of patients 
 taking salicylate of sodium is darkened by boiling with picric 
 acid and potash to a degree that might indicate from 1 to 2 grs. 
 of glucose per ounce. (When the solutions are mixed before 
 heat is applied, a slight red coloration is obtained, owing to a 
 little reduction of picric acid by creatinine.) 
 
 Bismuth tests.- — Bismuth salts are reduced to black metallic 
 bismuth when heated in an alkaline solution in the presence of 
 glucose. 
 
 (a) Bottgers test. — Urine and a concentrated solution of 
 carbonate of soda are mixed in equal quantities, and a little 
 subnitrate of bismuth is added. The mixture is then boiled ; 
 if sugar be present, the bismuth subnitrate is reduced, and the 
 mixture becomes black. The test is less sensitive than Fehling's 
 test ; if the patient has taken rhubarb, a black precipitate may 
 be obtained though sugar is absent. The presence of albumin 
 also gives rise to a black deposit. 
 
 (&) Nylanders modification is said to be more accurate. 
 Almen's fluid is used. It consists of 4 grms. of Eochelle salts 
 (tartarated soda and potash), dissolved in 100 c.c. of a 10 per 
 cent, solution of caustic soda. The fluid is warmed, and 2 grms. 
 of subnitrate of bismuth added. One volume of this fluid is 
 added to ten volumes of urine, and the mixture heated ; in a 
 few minutes (three to five) it will become black if sugar be 
 present. The reaction will indicate the presence of - l per cent, 
 of sugar. 
 
1 6 SUGAR TESTS. 
 
 Nylander's test is of service chiefly as a negative test. If 
 no reaction is obtained, then sugar is absent, or the amount is 
 less than OT per cent. If the reaction is marked, then probably 
 sugar is present (providing the urine does not contain albumin 
 or any drug which reduces bismuth). If the reaction be in- 
 distinct, other confirmatory tests should be tried. 
 
 Buhner's test. — To the suspected urine an equal volume of 
 a concentrated solution of acetate of lead is added. A thick 
 white precipitate forms ; this is filtered off ; and to the filtrate, 
 in a test tube, ammonia is added drop by drop until a thick 
 curdy white precipitate forms. The test tube is then warmed 
 carefully to a temperature not exceeding 80° C. If grape sugar 
 be present in the urine, the white precipitate changes to a rose- 
 red or cherry-red colour. By heating the tube more rapidly 
 and to a higher temperature, a coffee-brown coloration is 
 obtained. In order that this reaction shall occur, the urine 
 must contain - 25 per cent, of grape sugar. 
 
 Milk sugar does not give this rose-red coloration ; only a 
 brown or yellowish-red coloration is obtained. 
 
 Indigo -car mine (Mulder's test). — When a solution of indigo- 
 carmine (sulphindigotate of sodium) is boiled with a solution 
 of carbonate of soda, the rich blue colour of the former remains 
 unchanged, but if a drop of urine containing glucose be present 
 in the mixture, then the blue colour disappears and the fluid 
 becomes yellow. If the fluid be then shaken up in the test 
 tube, the action of the oxygen of the air causes the blue colour 
 to return. The test solutions are very liable to decompose, 
 however, and hence in this form the test is seldom used. 
 Its chief advantage is that the reagents can be employed in 
 the form of dry test papers, and these have been largely used 
 by Oliver ( n ) for the bedside testing of urine. Slips of paper 
 are soaked in a solution of sodium carbonate ; they are then 
 dried and kept for future use. Similar papers are soaked in 
 a solution of indigo-carmine, and dried. (Test-papers prepared 
 according to Dr. Oliver's direction are supplied by Messrs. M. 
 Wilson & Son, of Harrogate.) 
 
 Oliver's mode of testing. — An indigo-carmine paper is dropped 
 into a half -inch test tube, then a carbonate of soda paper is 
 added, and the two papers are covered by a drachm of water. 
 Heat is applied until the water boils. If the water be soft, a 
 transparent blue fluid is obtained ; if the water be hard, then 
 
OLIVERS AND HOPPE-SE YLER'S TESTS. i 7 
 
 the solution will be turbid ; in this case a second carbonate of 
 soda paper may be added. One drop of the suspected urine is 
 allowed to fall into the tube from a pipette held vertically. 
 The contents of the tube are again boiled freely for a few 
 seconds, then the tube is raised and held an inch or two above 
 the flame for a minute. If glucose be present, the rich blue 
 colour changes to violet, then the solution becomes purple, red, 
 reddish yellow, and finally straw yellow. The time required for 
 the development of the above reaction depends on the amount 
 of glucose present. If the glucose be present in large quantity 
 (over 20 grs. to the ounce), the reaction develops in a few 
 seconds. If the quantity of glucose be small, the reaction may 
 not occur for thirty or sixty seconds. On shaking the test tube 
 after the fluid has been decolorised by glucose, the blue colour 
 returns. 
 
 Indigo -carmine is not decolorised by uric acid, albumin, 
 mucus, pus, or bile ; but the reaction is obtained with liquor 
 potassse and soda3, dextrin, and milk sugar as well as glucose. 
 Indigo-carmine is of service as a negative test when a great 
 excess of uric acid has given a slight reduction of Fehling's 
 solution. It is also useful as a positive test when lactose is 
 present in the urine. 
 
 Hoijpe-Seylers orthonitro-phenylpropiolic acid test ( 12 ). — Baeyer 
 has shown that orthonitro-phenylpropiolic acid is converted into 
 indigo when heated with alkalies and a reducing substance such 
 as grape sugar. A half per cent, solution is employed. In 100 
 c.c. of a 10 per cent, solution of caustic soda, 5*76 grms. of 
 orthonitro-phenylpropiolic acid are dissolved, and water is added 
 up to 1000 c.c. A brownish fluid is obtained, which keeps 
 fairly well. In testing for sugar, about 1 drm. of this solution 
 is placed in a test tube, and ten drops of the suspected urine are 
 added. The mixture is boiled for a quarter of a minute. The 
 fluid becomes dark blue, if the urine contains as much as 0*5 
 per cent, of sugar. A distinct blue colour is not obtained with 
 normal urine. Even when large quantities are added to the test 
 solution, only a greenish coloration is obtained. If the urine 
 contains more than 2 per cent, of albumin, which is rare, then 
 a dark red coloration is obtained. 
 
 The advantages of this test are said to be that the reagents 
 keep well, that only a small quantity of urine is required. The 
 reaction is not influenced by any ordinary amount of albumin • 
 
1 8 SUGAR TESTS. 
 
 it is less influenced by other reducing substances, such as 
 creatine and creatinine, than many sugar tests, and thus has 
 an advantage over Trommer's test, Nylander's test, etc. 
 
 The test is not so sensitive as either the phenylhydrazin or 
 the fermentation tests, and not quite so sensitive as Fehling's solu- 
 tion. It is important not to add too much urine, otherwise no 
 blue coloration is obtained, even when sugar is present. The 
 test is rapidly performed, and is a good confirmatory reaction 
 when sugar is present to the extent of 0*5 per cent. I have 
 given it a fair trial, and never obtained any reaction with 
 normal urines. Urine containing bile pigment gave no reaction ; 
 also the urine of patients taking sodium salicylate did not pro- 
 duce any blue coloration. 
 
 Methylene olue test ( 13 ). — Methylene blue is decolorised by 
 glucose in a warm alkaline solution. In performing the test, 
 the diabetic or suspected urine is diluted (1 part to 9 of water). 
 Of this diluted urine, 2 c.c. are mixed with 6 c.c. of a 1 in 3000 
 solution of methylene blue, and 2 c.c. of liquor potasste added. 
 The mixture is boiled for one or two minutes, when the blue 
 colour disappears if sugar be present. Care must be taken that 
 the fluid is shaken as little as possible, since the blue colour 
 returns easily, owing to the action of the oxygen of the air. It 
 is important to dilute the urine, as all undiluted urine dis- 
 charges the blue colour ; but normal urine diluted 1 to 9 of 
 water does not decolorise methylene blue. I have obtained a 
 distinct reaction by this method when diabetic urine was diluted, 
 until the percentage of sugar was only - 07, but when further 
 diluted until it was "014 no reaction was obtained. I have 
 found, however, that urine rich in urates gave a doubtful reaction 
 when diluted 1 to 9, and I do not think the test so satisfactory 
 as some of those already mentioned. 
 
 Saffranin test. — Saffranin when dissolved in water produces 
 a blood-red solution, which is decolorised when heated with 
 liquor potassae and grape sugar. This reaction has been sug- 
 gested as a test for glucose in the urine, and has recently been 
 strongly advocated by Allen ( 13 ). In performing the test, equal 
 parts of the suspected urine, liquor potassse, and a 1 in 1000 
 watery solution of saffranin are mixed in a test tube and boiled 
 freely, agitation being avoided as much as possible. Normal 
 urine does not decolorise saffranin, but urine containing 0'1 per 
 cent, of sugar causes the red colour to disappear. Urine con- 
 
THE FERMENTATION TEST 
 
 19 
 
 taining 0'07 per cent, almost decolorises the above mixture — 
 only a faint red tinge remains. Uric acid, creatine, creatinine, are 
 said not to decolorise saffranin. Albumin, however, decolorises 
 it slowly, and hence ought to be removed before testing for 
 sugar. I have found the test unsatisfactory, and have obtained 
 doubtful reactions when sugar has been absent. 
 
 The fermentation test. — As a positive test, this is the most 
 certain reaction for grape sugar in the urine. If yeast be 
 mixed with urine containing grape 
 sugar, and the specimen kept in a 
 warm place, fermentation occurs ; the 
 sugar is decomposed, alcohol and car- 
 bonic acids are formed : — 
 
 C 6 H 12 6 =2C 2 H 6 + 2C0 2 
 
 (glucose) (alcohol) (carbonic acid). 
 
 Other changes of minor importance 
 also occur. The liberation of carbonic 
 acid gas indicates the presence of 
 sugar. 
 
 The test can be performed in an 
 ordinary test tube, which is filled with 
 urine and a little yeast, and inverted 
 in mercury. Or one may employ a 
 U-shaped tube ( u ), or an inverted 
 test tube, the lower fourth of which 
 is filled with mercury, and the mouth 
 closed with a stopper perforated with 
 a glass tube bent twice at right angles, 
 as shown in the diagram. Now, if 
 grape sugar be present in the suspected urine, carbonic acid gas is 
 formed, and collects at the top of the test tube or U-shapecl tube 
 (see Figs. I. 1—3). It is always important to use a control test tube 
 containing yeast and normal urine, i.e. urine which does not 
 give any reaction with Fehling's solution, since the yeast 
 alone suspended in normal urine gives rise to the formation 
 of a small bubble of gas, which collects at the top of the tube 
 (see Fig. 1). 
 
 The use of mercury is not necessary, and the fermentation 
 test for the detection of grape sugar can be most conveniently 
 carried out in the following manner (Moritz): — Two ordinary 
 
 Fig. 1. — Fermentation test ibr 
 sugar, a, test tube containing 
 suspected urine+yeast; b, test 
 tube containing normal urine 
 + yeast. Clear part indicates 
 amount of gas in the tubes in 
 each case. 
 
2o SUGAR TESTS. 
 
 test tubes of equal size are employed. The same qtiantity of 
 German yeast (weighed) is placed in the bottom of each. About 
 one-tenth of each tube is filled with dry yeast. The one 
 tube is then filled to the brim with the suspected urine ; the 
 other with normal urine (i.e. urine which gives no reaction with 
 Fehling's solution). The mouth of each test tube is closed with 
 an india-rubber stopper perforated with a glass tube, bent twice 
 at right angles (see Fig. 1). A little of the fluid escapes into 
 the glass tubes when the stoppers are inserted. The two test 
 tubes are inverted and placed side by side in a warm place, 
 supported in a glass tumbler, or in some other way. The atmo- 
 
 Fig. 2. — Fermentation test for sugar ; another method. As the 
 saccharine urine ferments, gas collects in the test tube and 
 exjiels the fluid through the glass tubing into the wine glass. 
 
 spheric pressure keeps up the column of fluid in each test tube. 
 At the end of eighteen to twenty hours, often long before that 
 time, fermentation will have occurred, if sugar be present in the 
 suspected urine. At the top of the tube containing the normal 
 urine and yeast, a bubble of gas collects — this being given off 
 from the yeast itself. In the other tube containing the sus- 
 pected urine, if grape sugar be present, a greater quantity of gas 
 collects at the top than in the case of the first test tube. The 
 gas may occupy a quarter of an inch, half an inch, or an inch at 
 the top of the tube ; or if the sugar be greater in quantity, half 
 the tube or the whole tube may be filled with gas, and the fluid 
 expelled to a corresponding amount. Any excess of gas in the 
 tube containing the suspected urine, beyond the small bubble 
 
THE FERMENTATION TEST 21 
 
 given off by the yeast itself in the tube containing the normal 
 urine, indicates the presence of grape sugar or other fermentable 
 sugar. 
 
 It is important that the tube be kept in a place which is 
 not too warm ; also the india-rubber stopper ought to fit the test 
 tubes accurately. The control tube containing normal urine 
 + yeast is always necessary, since the amount of gas which 
 collects in this tube varies considerably, according to the 
 activity of the yeast. Carried out in this simple manner, the 
 fermentation test is most useful and reliable as a positive test. 
 Moritz ( 15 ' 17 ) recommends that the quantity of yeast should be 
 2 per cent. The best temperature for fermentation is 25° C. 
 
 It is pointed out by Moritz, however, that to avoid all 
 possible fallacies four tubes are useful, since the results might 
 be unreliable, owing to the activity of the yeast being greatly 
 impaired, or owing to the urine containing some substance pre- 
 venting the fermentation of yeast : — 
 
 1. A tube containing normal urine + 2 p. c. of yeast. 
 
 2. „ ,, „ ,, +0-1 p. c. of sugar. 
 
 3. ,, ,, the suspected urine ,, 
 
 Tube 1 shows the amount of gas which is given off by the 
 yeast itself. 
 
 Tube 2 shows whether the yeast is capable of giving rise to 
 fermentation, i.e. it indicates the activity of the yeast itself. If 
 the yeast be active, the gas in Tube 2 will be in excess of the 
 bubble of gas in the Tube 1. 
 
 Tube 3 is for comparison with Tube 1 ; and if an excess of 
 gas is present in 3 beyond that in 1, sugar is indicated. 
 
 Tube 4 shows whether the suspected urine is capable of 
 being fermented. If so, then the gas in 4 will be in excess of 
 that in 3. If there is no formation of gas in 4, then we may 
 conclude that there is some substance present which is checking 
 fermentation. As a rule, however, four tubes are not necessary, 
 and a conclusion can generally be arrived at from the result 
 given by two tubes, as described above. 
 
 Value of the test. — As a positive reaction, fermentation is 
 the most reliable test for sugar. If gas be given off in excess 
 of that yielded by the yeast + normal urine, then we may 
 conclude with certainty that sugar is present in the suspected 
 
SUGAR TESTS. 
 
 urine. No other substance met with in the urine ferments and 
 gives off gas when subjected to the action of yeast. 
 
 One of the chief objections to the test is the time which is 
 required, but this does not appear to me to 
 be a very serious objection. The tubes as 
 above described can be prepared in a few 
 minutes. They are then placed aside for 
 twenty-four hours, and, when examined at the 
 end of that time, at a glance one can see if 
 there is any indication of sugar. 
 
 Another objection is, that the test is not 
 so sensitive as many other reactions. 
 
 If no gas be liberated by the fermentation 
 test, and if there be no source of fallacy with 
 regard to the yeast, and if fermentation is 
 capable of occurring in the urine when sugar 
 is added (and these points can be determined 
 by the four tubes recommended by Moritz and 
 described above), we are justified in concluding 
 that sugar is either absent or only present in 
 very small quantity ; but we cannot say that 
 a minute quantity of sugar does not exist in 
 the urine. In the fermentation test the 
 fluid absorbs its own volume of carbonic acid 
 gas, hence a very small quantity of sugar may be present and 
 this may ferment, but all the gas may be absorbed. 
 
 By using very active yeast I have sometimes obtained a dis- 
 tinct reaction with the fermentation test when Fehling's solution 
 has only given a doubtful reaction, i.e. when the urine has not 
 reduced Fehling's solution at once on boiling, but has done so 
 only as the test tube cooled. According to Sir William Eoberts, 
 if the amount of sugar be less than 0'5 per cent., it cannot be 
 detected by the fermentation test. According to von Jaksch 
 and Jolles ( 16 ), by the fermentation test - l per cent, of sugar 
 can be detected. Einhorn( 18 ) states that, by boiling the urine 
 for ten minutes before applying the fermentation test, he can 
 detect 0"05 per cent, of sugar, and this statement is confirmed 
 by Kobrak ( 19 ). Of course, as Einhorn points out, it is of great 
 importance to place the same quantity of yeast in each tube. 
 Einhorn also recommends the use of a large quantity of yeast — 
 1 of yeast to 1 of fluid. 
 
 Fig. 3.— U-shaped tube 
 for detecting sugar iu 
 the urine by the fer- 
 mentation test. The 
 narrow part of the 
 tube contains mercury 
 (deep black). 
 
PHENYLHYDRAZIN TEST. 23 
 
 Seegen has pointed out that when only a minute quantity of 
 sugar is present, fermentation occurs very slowly. 
 
 Moritz ( 20 ) recommends a combination of Nylander's test and 
 fermentation, in order to be quite sure that a trace of sugar is 
 not present when no gas has been liberated. Urine which gives 
 a positive result with Nylander's test, but after fermentation no 
 longer gives Nylander's reaction, can with certainty be declared 
 to contain sugar. Urine which gives Nylander's reaction, but 
 does not ferment with yeast, and after the action of yeast still 
 gives Nylander's reaction, does not contain sugar but some other 
 reducing body. 
 
 The phenylhydrazin test. — A confirmatory test, about which 
 much has been recently written, is the phenylhydrazin test. 
 
 The researches of Emil Fischer ( 21 ) have shown that many 
 kinds of sugar are able to combine with phenylhydrazin and 
 form definite crystalline compounds — osazones. These com- 
 pounds present minor differences in their melting points, 
 solubility, etc., according to the form of sugar from which they 
 have been prepared. 
 
 By the action of phenylhydrazin on grape sugar a charac- 
 teristic crystalline compound is formed, phenylglucosazone. The 
 crystals are needle-shaped and have a bright sulphur-yellow 
 colour. This reaction was strongly recommended by von 
 Jaksch ( 22 ) some years ago as the basis of a clinical test for 
 grape sugar in the urine. The following is the method which he 
 employs ( 23 ) : — 
 
 Two parts of hydrochlorate of phenylhydrazin (twice as much 
 of the phenylhydrazin salt as will lie on the point of a knife) 
 and three of acetate of soda are placed together in a test tube 
 containing 6 to 8 c.c. of urine. If the salts do not dissolve when 
 the fluid is warmed, a little water is added. The test tube 
 containing the mixture is placed for 20 to 30 minutes in boiling 
 water. It is then taken out and placed in a vessel containing 
 cold water. If sugar be present a yellowish deposit forms, 
 which, on microscopical examination, is seen to contain yellow 
 needle-shaped crystals, often in clusters, in tufts, sheaves, or 
 rosettes. These are the crystals of phenylglucosazone, which 
 have a melting point of 205° C. 
 
 Now, if performed in this manner, it is necessary to make use 
 of a water-bath, and to keep the water boiling for twenty to thirty 
 minutes. This, of course, presents no difficulty in the laboratory, 
 
24 
 
 SUGAR TESTS. 
 
 but for clinical work is somewhat troublesome. Also, in the 
 hands of many medical men, the test has sometimes failed when 
 performed in this manner, even though sugar has been present 
 in the urine in considerable amount. One cause of failure is 
 probably because the quantity of the reagent recommended in 
 
 1 
 
 Fig. 4. — The phenylhydrazin test for sugar, a, crystals obtained with grape 
 sugar (phenylglucosazone) ; b, crystals obtained with lactose (simple method) ; 
 c, minute crystals, of doubtful nature, obtained from normal urine by follow- 
 ing the method of Moritz ; not obtained in simple method of applying test. 
 
 von Jaksch's description of the test (namely, twice as much phenyl- 
 hydrazin as will lie on the point of a knife) is somewhat 
 indefinite, and this description has been copied in many accounts 
 of the test which are given in English medical text-books. 
 
 The following is a very simjile clinical method of performing 
 
PHENYLHYDRAZIN TEST. 25 
 
 the phenylhydrazm test, which I have employed for seven 
 years with the best results — a method described in foreign 
 medical literature some years ago, but not generally referred to 
 in English medical books. 
 
 The simple method to which I refer has been described in the 
 work on urine analysis by Hoffmann and Ultzmann, and the 
 following, which I have found so useful, is a very slight modifi- 
 cation thereof. A test tube of ordinary size is filled for about 
 half an inch with hydrochlorate of phenylhydrazin 1 (in powder); 
 then acetate of soda in powder (or small crystals) is added for 
 another half inch. The test tube is then half-filled with urine 
 and boiled over a spirit-lamp. In performing the test I have 
 not attempted to dissolve the salts by shaking the tube, but 
 have simply applied the flame of the lamp to the bottom of the 
 tube, and the powders have soon passed into solution. After 
 the urine has reached the boiling point, I have always continued 
 to boil for about two minutes. The tube is then left in the test 
 stand and examined again some time afterwards. If sugar be 
 present, a yellowish deposit forms at the bottom of the tube, 
 and on microscopical examination this deposit is seen to 
 consist chiefly of beautiful needle-shaped crystals of a bright 
 sulphur-yellow colour. Generally the needles are arranged 
 in tufts, sheaves, or rosettes. The crystals are found after 
 the tube has been standing for half an hour, frequently at 
 the end of fifteen or twenty minutes ; but after boiling the 
 tube I have generally placed it in a test stand and not 
 examined the deposit for a few hours. If no sugar is present, 
 only brownish amorphous globules or yellowish scales are found 
 in the deposit. 
 
 This method is exceedingly simple, and very little time 
 and attention are required. In four minutes the first part of 
 the test can be easily completed ; the tube is then left in 
 the test stand and examined at a convenient time some hours 
 later. Permanent specimens of the crystals can be prepared by 
 drying a little of the deposit on a slide and mounting it in 
 Canada balsam. 
 
 The value of the test. — I have used this simplified method 
 for seven years in a very large number of urine examinations, 
 
 1 Hydrochlorate of phenylhydrazin is a brownish powder having a peculiar sine]]. It 
 is simply kept in tin; powdered form in a dry liottle. The acetate of soda is also kept in 
 a dry powder in another bottle. 
 
26 SUGAR TESTS. 
 
 and have never failed to find abundance of the characteristic 
 crystals on microscopical examination, if sugar has been present, 
 even when the quantity has been very minute. The reagents 
 are powders which usually keep well for months, and in this 
 respect they have a distinct advantage over Fehling's solution, 
 which so frequently decomposes. 
 
 The test can be performed, and the characteristic crystals 
 easily obtained, even when the urine contains a large quantity 
 of albumin. I have found the crystals easily in urines loaded 
 with albumin, but which contained only a very small amount 
 of sugar. In such cases the bright yellow crystals are easily 
 distinguished from the amorphous granules of coagulated albumin 
 on microscopic examination. It is always better, of course, to 
 remove the albumin first, if the quantity should be very great. 
 
 By performing the test in the simple manner just described 
 (i.e. boiling for two minutes only), I have never obtained any 
 crystals of phenylglucosazone in normal urine. Several years 
 ago I applied this test to the urines of fifty persons who were 
 either in good health, or were not suffering from any disease 
 which is accompanied by glycosuria or other diabetic symptoms. 
 (In each case no reaction for sugar was obtained with Fehling's 
 solution.) I was never able to find any trace of phenylglucosazone 
 crystals in these fifty cases. Since that time I have applied the 
 test on very numerous occasions to normal urine, but always 
 with negative results. Hence I think this simplified method 
 of performing the test may be regarded as quite reliable for 
 clinical work, and not too sensitive. 
 
 A great advantage of the phenylhydrazin test is, that it gives 
 no reaction with uric acid, creatinine, hippuric acid, pyrocatechin, 
 etc., whilst these substances may be a source of fallacy when 
 Fehling's test is applied. 
 
 Now, two objections have been raised to phenylhydrazin as 
 a clinical test for sugar in the urine — (1) It is said to give a 
 slight reaction with many normal urines and to be too sensitive for 
 clinical work. (2) It is said that glycuronic acid, if present in 
 the urine, will give rise to crystals similar to those of phenyl- 
 glucosazone. 
 
 As regards the objection that it is too sensitive for clinical 
 work, it has been pointed out by many observers that a small 
 deposit of the crystals may be obtained in many normal urines. 
 Thus, Binet ( 24 ) succeeded in obtaining the crystals in one-half of 
 
PHENYLHYDRAZIN TEST. 27 
 
 the normal urines which he examined, but the test tube was 
 kept in a hot-water oath for one hour. 
 
 Moritz ( 25 ) has pointed out that if certain proportions of the 
 reagents be employed, a slight deposit of minute yellow crystals 
 can be obtained in almost all normal urines. These proportions 
 are as follows : — To 1 c.c. of urine; 5 grm. of hydrochlorate 
 of phenylhydrazin and 1*0 grm. of sodium acetate are added. 
 The tube is placed in a water-bath, which is heated for one hour. 
 I have performed the test in this manner a good number of 
 times with perfectly normal urine, and have always succeeded in 
 obtaining a very slight deposit of the minute yellow crystals, 
 when I have followed Moritz s method strictly. But, as above 
 mentioned, I have never succeeded in obtaining the crystals from 
 normal urine when I have employed the simple method — boiling 
 in a test tube for two minutes only. 
 
 The crystals obtained from diabetic urine are generally long 
 fine needles (see Fig. 4, a), but those obtained from normal urine 
 by Moritz's method are minute very short needles, arranged 
 generally in rosettes, forming " thorn-apple " crystals (see Fig. 4, c). 
 According to Jolles ( 26 ), the crystals obtained from normal urine 
 are always thicker and more plump than in diabetes. Both 
 have the same sulphur-yellow colour. Even when diabetic 
 urine is very greatly diluted, so that the amount of sugar is 
 only 0*015 per cent., the crystals may be still long fine needles; 
 but I have always found the crystals minute and very short 
 (" thorn-apple " crystals) when obtained from normal urine by 
 Moritz's method. In cases of glycosuria, however, I have 
 occasionally found shorter forms of crystals. 
 
 Moritz ( 27 ) points out that in normal urine, even when his 
 sensitive method is used, the deposit of yellow crystals is 
 very small in amount, whilst in glycosuria and diabetes the 
 deposit of crystals is very abundant. He regards this as a point 
 of importance in the diagnosis. 
 
 Now the question arises, What is the cause of the small 
 deposit of minute short crystals in normal urine, when the test 
 is carried out according, to Moritz's sensitive method ? On this 
 point there is considerable difference of opinion. Thierf elder ( 28 ) 
 has shown that glycuronic acid forms yellow needle-shaped 
 crystals with phenylhydrazin. According to Fliickiger ( 20 ), a 
 very small amount of glycuronic acid is present in normal 
 urine, and the minute crystals obtained from normal urine 
 
28 SUGAR TESTS. 
 
 by the delicate method of performing the phenylhydrazin test 
 are due to this acid. Geyer( 30 ) states that after the action of 
 yeast on normal urine the crystals are still obtained, and he 
 regards them as due to glycuronic acid, since a trace of sugar 
 would have been destroyed by fermentation. Jolles ( 31 ) also 
 believes the crystals obtained from normal urine are clue to 
 glycuronic acid. Moritz thinks they are due partly to the 
 small amount of sugar which (according to some physiologists) 
 is present in normal urine, partly to other substances, possibly 
 glycuronic acid. 
 
 By a careful chemical process x Moritz has obtained yellow 
 needle-shaped crystals on applying the phenylhydrazin test to a 
 large quantity of normal urine, and has proved them to be due 
 to sugar and not to glycuronic acid, since their melting point 
 was 204° to 205° C. Hence he concludes that a trace of sugar is 
 found in normal urine. 
 
 I have made several experiments on this point, but have not 
 been able to confirm Geyer's statement, that these minute crystals 
 are still obtained after the action of yeast on normal urine. It 
 has been pointed out by Seegen, that if a fluid such as urine 
 contains a very small amount of sugar, it ferments very slowly 
 (thirty-eight hours may be required to complete the fermentation). 
 If the minute crystals obtained from normal urine are due to the 
 trace of sugar which some physiologists think is always present, 
 then, in order to decompose it by fermentation, it would be 
 necessary to submit the urine to the prolonged action of yeast ; 
 and I have found that, after the action of yeast for forty hours, 
 these minute crystals cannot be obtained. 
 
 A quantity of normal urine recently passed was divided into two 
 portions. One portion was placed in a bottle without the addition of 
 any yeast. To the other portion a quantity of yeast was added, and it 
 was kept for forty hoiirs in a warm place. At the end of this time, 10 c.c. 
 of the urine to which no yeast had been added were placed in a test 
 tube, A, with - 5 grm. phenylhydrazin hydrochlorate and 1 grm. sodium 
 acetate. Into a second tube, B, was placed 10 c.c. of the urine to 
 which the yeast had been added (and which had been kept warm), and 
 0'5 grm. of phenylhydrazin hydrochlorate and 1 grm. sodium acetate 
 were added. Both test tubes were placed in the same water-bath and 
 heated for one hour ; the tubes were then allowed to stand until next 
 day. The deposit in the former tube, A, contained minute yellow 
 
 1 Described in the Deutsches Arch./, klin. Med., Leipzig, Bd. xlvi. S. 257. 
 
PHENYLHYDRAZIN TEST. 29 
 
 crystals; while in the deposit in the latter tube, B — i.e. in urine which 
 had been submitted to the action of yeast for forty hours — no crystals 
 could be found. 
 
 Another experiment was performed in the same manner, hut the 
 specimens were tested at the end of twenty hours. Both in the urine to 
 which the yeast had been added, and in that to which no yeast had been 
 added, minute yellow crystals were obtained. 
 
 From these experiments we may conclude that the substance 
 in normal urine which gives rise to the minute yellow crystals, 
 when Moritz's delicate method of performing the phenyl- 
 hydrazin test is followed, is a body which is destroyed by the 
 action of yeast and warmth for forty hours ; but it is not 
 destroyed by the action of yeast for twenty hours. 
 
 Now, whatever may be the cause of these minute crystals, 
 they appear to me to form a serious objection to the test if it be 
 performed according to the delicate method, with the use of the 
 water-bath. Though the deposit of the crystal is very small in 
 normal urine, but always abundant if but a very slight degree 
 of glycosuria exists, and though the crystals are generally much 
 smaller and more plump in normal urine than in the case of 
 glycosuria, still I think that these points are insufficient to 
 enable any but those who have had great experience in the test 
 to decide whether a very small pathological amount of sugar is 
 or is not present in any suspected urine. 
 
 If, however, the water-bath be not used, and if the test be 
 performed in the simple manner above described — i.e. boiling in 
 a test tube for two minutes only — this objection does not apply. 
 By the simple method these troublesome and dubious minute 
 yellow crystals are not obtained in normal urine. 
 
 The second objection to the phenylhydrazin test is, that 
 yellow crystals similar to those of phenylglucosazone are pro- 
 duced when glycuronic acid is present in the urine. 
 
 The melting point of the crystals due to glycuronic acid is 
 114° to 115° C, whilst that of the crystals of phenylglucosazone 
 is 205° C. But it is evident that, for clinical work, the deter- 
 mining of the melting point of the crystals is too troublesome 
 and tedious a matter. 
 
 The urine of patients taking salicylate of soda and salol 
 generally gives a slight reduction of Fehling's solution, and 
 this is said to be due to glycuronic acid. In these cases 1 
 have sometimes found yellow needle-shaped crystals with the 
 
3 o SUGAR TESTS. 
 
 phenylhydrazin test, even when the simplified method has been 
 employed, but only in about one-third of the cases. 
 
 The objection that glycuronic acid may give rise to fallacy 
 is a valid one, though practically this acid is not likely 
 to give rise to much trouble, since its occurrence in any 
 quantity of practical importance is generally due to some drug, 
 and in such cases discontinuance of the medicine causes it to 
 disappear from the urine, and then crystals will no longer be 
 obtained by the simplified method of testing. Still, there is the 
 possibility of the exceedingly rare occurrence of glycuronic acid 
 apart from any drug treatment (see p. 11). 
 
 Three cases have been recorded recently by Salkowski, in 
 which pentose has been found in the urine. "With phenylhydrazin 
 it forms yellow needle-shaped crystals, which melt at 159° C. 
 But whether such crystals would be produced by the simplified 
 phenylhydrazin test (boiling for two minutes only), I do not 
 know. The occurrence of this substance in the urine is pro- 
 bably exceedingly rare, but it is now to be borne in mind as a 
 possible, though very improbable, source of fallacy, both in 
 Fehling's test and in the phenylhydrazin test. 
 
 On applying the phenylhydrazin test in its simplified form 
 (as above described, p. 25) to a solution of lactose in normal 
 urine, I have never been able to obtain any needle-shaped yellow 
 crystals like those characteristic of glucose, but I have found a 
 deposit of small yellow globules with very short spines (see 
 Fig. 4, b). 
 
 The delicacy of the simplified phenylhydrazin test (boiling 
 for two minutes only). — As above mentioned, the simplified 
 method is not too sensitive ; it does not give any reaction with 
 normal urine. Still, it is exceedingly delicate ; more delicate 
 than fermentation and than Fehling's solution, as is shown by 
 the following experiments : — 
 
 A diabetic urine containing 7 '5 per cent, of sugar was 
 greatly dilutedwith water until the percentage was only - 015. 
 This diluted urine gave (a) no reaction with Fehling's solution 
 in the cold ; (b) no reaction with Fehling's solution on boiling ; 
 (c) no indication of sugar by the fermentation test ; (d,) but a 
 distinct deposit of long yellow crystals with the phenylhydrazin 
 test (simplified method), boiling two minutes only. 
 
 Conclusions. — 1. The phenylhydrazin test, if performed ac- 
 cording to the above-described simplified method, {i.e. boiling 
 
PHENYLHYDRAZIN TEST. 
 
 3* 
 
 the reagents with the urine for two minutes only), is a most 
 valuable confirmatory test for sugar in the urine. The reagents 
 do not readily decompose, the test is easily carried out, and is 
 very suitable for clinical work. It is very sensitive — more 
 sensitive than Fehling's test and the fermentation test — and will 
 give a reaction with diluted urine containing only 0-015 per 
 cent, of sugar. But it is not too sensitive, and gives no reaction 
 with normal urine. 
 
 Relative Sensitiveness of Several Urine Tests. 
 
 + indicates reaction ; indicates no reaction. 
 
 Test for Sugar. 
 
 Amount of Glucose in Urine. 
 
 0'5 per cent. 
 
 01 per cent. 
 
 •05 per cent. 
 
 Fehling's solution + 
 heat. 
 
 Fehling's solution in 
 the cold. 
 
 Fehling's solution after 
 filtration through 
 animal charcoal. 
 
 Fermentation . 
 
 Phen ylhydrazin 
 (Simple method.) 
 
 + 
 
 + 
 + 
 
 + 
 
 + 
 
 + 
 Very slight. 
 + 
 
 + 
 Just detected. 
 
 + 
 
 + 
 
 
 
 + 
 
 05-0-01 p.c. detected, 
 
 but precipitate often 
 
 occurs only 30 to 60 
 
 seconds after heating. 
 
 O 
 
 + 
 Distinct crystals when 
 urine diluted, so that 
 only 0'015 per cent, of 
 glucose present. 
 
 2. If in performing the test a water-bath be used, and 
 boiling prolonged, the phenylhydrazin test is too sensitive, 
 and is somewhat troublesome for clinical work. 
 
 3. The great value of the phenylhydrazin test (simplified 
 method) is as a negative test. A urine which gives no reaction 
 by this method may be declared quite free from sugar for 
 practical purposes. Thus in a doubtful case, in which a slight 
 reduction of copper oxide is produced by Fehling's test, if the 
 simplified phenylhydrazin test gives no reaction, then the urine 
 may be declared free from sugar, and the reduction of copper 
 due to some other body. 
 
 4. Beside the various forms of sugar, so far as is known, 
 
32 SUGAR TESTS. 
 
 only glycuronic acid gives the phenylhydrazin reaction in the 
 urine ; but many substances which may occur in the urine 
 reduce Fehling's solution. If the patient should be taking any 
 drug, this must be discontinued, and the phenylhydrazin reaction 
 would not then be obtained, if glycuronic acid due to medicine 
 were the cause. If no drug is being taken, and the urine reacts 
 to the simplified phenylhydrazin test, it is exceedingly probable 
 that it contains sugar, since the occurrence of glycuronic acid 
 in the urine in sufficient quantities to give the reactions is 
 extremely rare. 
 
 Besides glucose, other forms of sugar are occasionally, though 
 very rarely, met with in the urine ; but at present no pathological 
 importance can be attached to them. 
 
 Milk sugar or lactose is present in the urine of women 
 during lactation, and especially when there is engorgement of 
 the breast in connection therewith (see p. 88). I have found 
 it often present in cases of abscess of the female breast. 
 
 Lactose reduces Fehling's solution on boiling. It is said not 
 to ferment with yeast ; according to some authors, however, 
 slight fermentation does occur after the 'prolonged action of yeast. 
 I have found that a strong solution of lactose in normal urine 
 showed no signs of fermentation at the end of twenty-four hours, 
 but at the end of forty hours there was a distinct but slight 
 indication of fermentation {i.e. at the top of the tube containing 
 the urine + lactose a little more gas was present than in the 
 control tube containing the normal urine). After the action of 
 yeast for twenty-four hours on urine containing lactose, the fluid 
 still reduces Fehling's solution. On the application of the phenyl- 
 hydrazin test in its simplified form (as described on p. 25) to a 
 solution of lactose in the normal urine, I have never been able 
 to obtain any needle-shaped yellow crystals like those charac- 
 teristic of glucose, but I have found a deposit of small yellow 
 globules with exceedingly short spines (see Fig. 4, b). 
 
 On the application of Buhner's test, no rose-red coloration is 
 produced (as is the case with glucose), but the fluid becomes 
 brown or yellowish reel. 
 
 According to Pavy ( 32 ), a valuable indication of the presence 
 of lactose is obtained by estimating the cupric-oxide reducing 
 power of the urine before and after boiling with sulphuric acid. 
 Taking the cupric-oxide reducing power of glucose as 100, 
 Pavy estimates that of lactose at 60, or a little over. When 
 
TESl^S FOR LsEVULOSE. 33 
 
 boiled with dilute sulphuric acid, lactose is converted into 
 dextrose and galactose — 
 
 12 H 22 O n + H 2 = C 6 H 12 6 + C 6 H 12 6 
 
 (lactose) (dextrose) (galactose) 
 
 After boiling with dilute sulphuric acid, if the phenyl- 
 hydrazin test be applied, then long straight needles and radiating 
 clusters of yellow crystals of phenyldextrosazone and galactosa- 
 zone are obtained. 
 
 Zcevulose has occasionally been found in the urine, in cases 
 of diabetes, along with grape sugar, but its occurrence is exceed- 
 ingly rare. Seegen ( 33 ) has only met with a single example in the 
 examination of a very large number of diabetic urines. Lsevulose 
 reduces Fehling's solution like grape sugar ; it ferments with 
 yeast, yielding alcohol and carbonic acid gas ; it forms yellow 
 needle-shaped crystals with phenylhydrazin, like those of phenyl- 
 glucosazone ; but it differs from glucose by rotating the plane of 
 polarised light to the left. When present along with glucose, it 
 will be found that the urine will fail to rotate polarised light to 
 the left, but the degree of rotation to the right (produced by the 
 glucose) will be greatly diminished (owing to the presence of 
 the lsevulose). In these cases the quantity of sugar indicated 
 by the polarimeter will be markedly less than that indicated by 
 the quantitative analysis with Fehling's solution. 
 
 CoAie sugar is said to be met with occasionally in the urine 
 of persons who have taken large quantities of cane sugar in 
 their food. It ferments, but does not reduce Fehling's solution, 
 and forms no compound with phenylhydrazin. 
 
 Pentose. — As previously mentioned (on p. 30), Salkowski ( 34 ) 
 has recorded three cases in which pentose was found in the 
 urine, and several have been reported since. No definite 
 pathological importance can be attached to its occurrence, how- 
 ever. The urine reduces Fehling's solution, but does not ferment 
 under the action of yeast. Pentose forms yellow needle-shaped 
 crystals with phenylhydrazin, which melt at 159° C, and which 
 fire soluble in hot water. I have not had an opportunity of 
 trying whether the phenylhydrazin test in its simplified form, 
 as described on p. 25, will give any reaction with pentose. 
 
 Urines containing pentose give a red coloration when warmed 
 with phloroglucin and hydrochloric acid. Phloroglucin is 
 dissolved in 5 or 6 c.c. of strong warm hydrochloric acid; 
 3 
 
34 
 
 SUGAR TESTS. 
 
 it is added in excess, so that a small quantity is left over 
 undissolved. The solution is divided into two portions, and 
 allowed to cool. To one-half of the solution in a test tube, 
 about half a c.c. of the suspected urine is added. To the 
 other half of the fluid a similar quantity of normal urine is 
 added. The test tubes are placed in a beaker containing boiling 
 water. In a few seconds the suspected urine presents a bright 
 red coloration if pentose be present, whilst the normal urine 
 remains unchanged. 
 
 Table showing Reactions given by Various Substances which may be 
 mistaken for Grape Sugar in the Urine. 
 
 4- indicates reaction ; indicates no reaction. 
 
 
 Reduction of 
 Fehling's 
 Solution. 
 
 Fermentation 
 with Yeast. 
 
 Phenyl- 
 hydrazin 
 test : simpli- 
 fied method. 
 
 Other Tests. 
 
 Glucose (dex- 
 
 + 
 
 + 
 
 + 
 
 Rotates the plane of polar- 
 
 trose) 
 
 
 
 Needles, 
 sheaves, 
 rosettes. 
 
 ised light to the right. 
 Reaction with Nylander's 
 and Rubuer's tests. Re- 
 action with Fehling's solu- 
 tion after filtration through 
 animal charcoal. No re- 
 action to phenylhydrazin 
 test, to Fehling's solution, 
 and other tests after fer- 
 mentation. 
 
 Ljevulose 
 
 + 
 
 ■+■ 
 
 + 
 Needles, 
 
 sheaves, 
 rosettes. 
 
 Rotates the plane of polar- 
 ised light to the left ; this 
 and other tests fail after 
 fermentation. 
 
 Lactose . . . 
 
 + 
 
 
 
 Globules 
 with very 
 short 
 spines. 
 
 No reaction with Rubner's 
 test ; brown coloration 
 only. Converted into dex- 
 trose and galactose by 
 boiling with dilute sul- 
 phuric acid. Reactions 
 for glucose then obtained. 
 
 Cane sugar . . 
 
 
 
 + 
 
 
 
 
 Pentose . . . 
 
 + 
 
 
 
 Simplified 
 method (?) 
 
 Red coloration with phlo- 
 roglucin and hydrochloric 
 acid. 
 
 Uric acid in 
 
 + 
 
 
 
 
 
 Separated by filtering urine 
 
 great excess . 
 
 Slight. 
 
 
 
 through animal charcoal. 
 Filtrate does not reduce 
 Fehling's solution. 
 
 Glycuronic acid 
 
 + 
 
 
 
 + 
 Slight de- 
 posit of 
 crystals. 
 
 Generally due to admini- 
 stration of drugs. Feh- 
 ling's test, obtained after 
 action of yeast. 
 
 Dextrin . . . 
 
 -r 
 
 
 
 
 
 After isolation, brown col- 
 oration with iodine. 
 
 Alkapton or 
 
 + 
 
 
 
 
 
 No reaction with Nylander's 
 
 pyrocatecliin. 
 
 
 
 
 test. Urine becomes dark 
 on standing, or on addi- 
 tion of liquor potassse or 
 ammonia. 
 
DETECTION OF GRAPE SUGAR IN THE URINE. 35 
 
 II. On the Detection of Small Quantities of Grape 
 Sugar in the Urine. 
 
 When sugar is present in the urine in large quantities, its 
 detection is easy ; but if only small quantities or traces be 
 present, their detection with certainty is often troublesome. In 
 the latter case, in spite of the great number of tests for sugar, 
 often a satisfactory conclusion cannot be arrived at, unless the 
 medical man be prepared to devote some time and care to the 
 urine examination. 
 
 No single sugar test is quite satisfactory in all cases. The 
 best tests we possess have some weak points when applied to the 
 detection of minute quantities of sugar in the urine. Either they 
 are not sufficiently sensitive, or they occasionally give reaction 
 with other substances besides grape sugar. 
 
 During the last eight years I have devoted a considerable 
 amount of time to the examination of urine specimens which 
 have been thought to contain small quantities of sugar, and 
 have given a fair trial to all of the sugar tests just described ; 
 but I have found Fehling's solution, the phenylhydrazin test, and 
 fermentation by far the most satisfactory. By these three tests, 
 employed and combined in the order and manner which will be 
 described, it is nearly always possible to decide with certainty 
 whether a small quantity of sugar is present or absent in a 
 suspected urine. I have found this combination exceedingly 
 useful, convenient, and reliable for clinical work, and the time 
 actually taken up in performing the several tests will not amount 
 to many minutes, though it may be necessary to wait twenty- 
 four hours before a decision can be given. 
 
 In examining a suspected urine — 
 
 1. Apply Fehling's test, taking the usual precautions. For 
 most cases no other sugar test is needed. If a negative result 
 be obtained, the urine may be declared free from sugar in the 
 clinical sense. A well-marked reduction of oxide of copper, when 
 other indications of diabetes are present, may be taken as evidence 
 of the presence of grape sugar. But, as mentioned already, it is 
 a third class of cases which is especially liable to cause trouble. 
 There are often no indications of diabetes or glycosuria, except 
 the reduction of Fehling's solution, and frequently this reduction 
 is slight. After the urine has been raised to the boiling 
 point, the fluid may be unchanged at first, but when the test 
 
3 6 SUGAR TESTS. 
 
 tube is allowed to cool, it becomes turbid, greenish, or greenish 
 yellow. 
 
 In these cases sometimes the reduction of copper is due to a 
 small quantity of grape sugar, sometimes to other substances. 
 As mentioned above, Fehling's solution may be slightly reduced, 
 if the urine contain a great excess of uric acid or urates, 1 
 especially when boiling is prolonged and when too much urine is 
 added. It is also reduced by glycuronic acid, alkapton, or 
 pyrocatechin ; slightly by creatine, creatinine, xanthine ; and 
 by the following varieties of sugar besides glucose : — lactose, 
 maltose, lsevulose, galactose, and pentose. Lactose is often present 
 in the urine during lactation and in cases of abscess of the 
 breast. The presence of glycuronic acid in the urine is generally 
 due to the administration of some drug (see p. 10), but Ashclown 
 has shown that it may be occasionally met with when no drug 
 is being taken. 
 
 When the patient is under the influence of any of the drugs 
 which are liable to cause the appearance of glycuronic acid in 
 the urine, it is well to discontinue the treatment, and to test a 
 second sample of urine. But this is not always possible, since 
 an opinion must often be given on one specimen only. 
 
 In the cases in which there are no indications of diabetes 
 beyond the reduction of Fehling's solution, and especially if the 
 reduction be slight or of an indefinite nature, it is evident, 
 therefore, that some other test is necessary before coming to a 
 decision. 
 
 2. Apply the phenylhyclrazin test, employing the simplified 
 method of boiling for two minutes, without the use of a water- 
 bath (for description, see p. 25). 
 
 Whilst many substances may reduce Fehling's solution 
 slightly, the majority of these give no reaction with the 
 phenylhyclrazin test. Thus uric acid, urates, creatine, creatinine, 
 hippuric acid, xanthine, maltose, cane sugar, give no reaction. 
 
 JSTow, phenylhyclrazin is a most useful test for sugar when 
 applied in the simple manner above described. It is exceedingly 
 
 1 When the urine contains an excess of uric acid and urates, these can be removed by 
 filtration several times through animal charcoal, whilst glucose passes through in the 
 filtrate, and may then be readily detected by Fehling's solution. When a slight reduction 
 of Fehling's solution is due to excess of uric acid or urates in the urine, the reduction 
 will not be obtained after filtration through good animal charcoal (see p. 13). 
 
 Urine which contains alkapton or pyrocatechin becomes greenish-brown on standing. 
 It changes to deep brown or black on the addition of a few drops of liquor potassse or 
 ammonia. 
 
DETECTION OF GRAPE SUGAR IN URINE. 37 
 
 sensitive, but not too sensitive for clinical work. A distinct 
 reaction is obtained even when the urine contains only 0*015 per 
 cent, of sugar, and no reaction is obtained with normal urine 
 when this simple method of performing the test is followed. 
 
 It is of value chiefly as a negative test. If no reaction be 
 obtained, we can state with certainty that glucose is entirely absent, 
 and that the reduction of the Fehling's solution was not due to 
 glucose. If, then, no deposit of crystals be found in the test tube 
 when examined half an hour (or perhaps better, one hour) after the 
 test has been performed, no further testing is required. If the 
 sulphur-yellow, needle-shaped crystals form, they generally do so 
 in fifteen or twenty minutes. These crystals would indicate 
 glucose almost with certainty, if numerous or large ; but there 
 is just the possibility of glycuronic acid being the cause, especi- 
 ally when the crystals are very scanty or small, and this is the 
 weak point of a positive reaction with the phenylhydrazin test. 
 If the crystals consist of globules with short spines as in Fig. 4, 0, 
 lactose may be suspected. If a positive reaction be obtained, 
 and we wish to be quite certain that it is due to glucose, we 
 must employ the next test. 
 
 3. Apply the fermentation test, using two test tubes, one con- 
 taining normal urine for control, as described on p. 19, and 
 taking the precautions there mentioned. If there should be a 
 greater quantity of gas at the top of the tube containing the 
 suspected urine, sugar is certainly present, and no further testing 
 is required. 
 
 A positive reaction is often obtained in one, two, or three 
 hours, but twenty-four hours may be necessary. The weak point 
 of this test is, that it will not give any reaction if the amount 1 
 of sugar should be less than - l per cent. 
 
 4. If no reaction is obtained by the fermentation test, apply 
 Fehling's test to the suspected urine, which has been submitted to the 
 ■action of yeast in the test tube for twenty-four hours. If now 
 no reaction should be obtained with Fehling's solution, the urine 
 has contained glucose. A urine which has reduced Fehling's 
 solution, however slightly, before fermentation, but which, after 
 the action of yeast for twenty-four hours in the test tube, no 
 
 1 Of course, reliable results will only be obtained when the yeast is active. There is 
 DO difficulty in obtaining active yeast, however. If the yeast should be suspected, a third 
 test tube, containing the urine + yeast and a little glucose, would show whether it is 
 capable of giving rise to fermentation. 
 
38 SUGAR TESTS. 
 
 longer reduces Fehling's solution, may with certainty be stated 
 to have contained sugar, even though no evidence thereof, by 
 excess of gas, can be obtained by the fermentation test alone. 
 G-lycuronic acid and lactose do not ferment, and therefore after 
 the action of yeast for twenty-four hours the urine still reduces 
 Fehling's solution. 
 
 I have made a number of experiments on the value of this 
 combined method, by adding a few drops of diabetic urine to 
 a large quantity of normal urine. By trial I have thus obtained 
 samples of urine which contained so little sugar that they gave 
 only a very indefinite reaction with Fehling's solution, i.e. a 
 slight greenish opacity, which did not appear at once on boil- 
 ing, but only developed after the urine was allowed to cool. 
 A well-marked reaction was obtained in these cases by the 
 phenylhyclrazin test (characteristic crystals). Now, such urines 
 gave no evidence of the presence of sugar, by excess of gas, 
 when the fermentation test was applied in a test tube in the 
 manner above described. Yet the mixed urine and yeast, taken 
 from the test tube at the end of twenty-four hours, failed to 
 reduce Fehling's solution. 
 
 In these experiments the urine, before fermentation, con- 
 tained a quantity of sugar, so small that only a very slight 
 reaction was obtained by Fehling's test (i.e. when the fluid began 
 to cool after boiling). Any quantity of sugar less than this, as 
 Sir William Eoberts points out, is of no clinical importance. 
 Hence this combination of tests is sufficient to give definite 
 results, when the urine contains the smallest quantity of sugar 
 of any clinical importance. 
 
 When the tests are performed in the above order, the phenyl- 
 hydrazin test at once eliminates the majority of the substances, 
 other than sugar, which cause the reduction of Fehling's solution. 
 Very often a negative result is obtained, and no further testing 
 is required ; no fermentation is necessary, and thus much time 
 is saved. In case a positive result should be obtained with 
 phenylhyclrazin, the application of the fermentation test may 
 be necessary. It may or may not reveal the presence of sugar. 
 In either case, if the urine no longer reduces Fehling's solution 
 (after the action of yeast), sugar is indicated. By the applica- 
 tion of the above combination of tests, in the order and manner 
 described, I believe small quantities or traces of sugar can be 
 detected with greater certainty and with less actual expenditure 
 
QUANTITATIVE ESTIMATION OF SUGAR. 39 
 
 of time than by any other means, though it may be necessary in 
 some cases to allow twenty-four hours to elapse before an 
 opinion can be given ; still the time actually expended in 
 performing the test is but little, and the methods are very 
 simple and convenient for clinical work. 
 I. — Fehling's Solution. 
 
 No reaction = 
 no sugar. 
 
 Slight reaction, other dia- A marked re- 
 
 betic indications absent action and 
 
 (glucose, lactose, pen- other indi- 
 
 tose, uric acid in excess, cations of 
 
 glycuronic acid, alkap- diabetes = 
 
 ton, etc. etc.). sugar. 
 
 II. — Phenylhydrazin Test (Simplified Method). 
 
 No reaction = no 
 sugar. (Gives 
 no reaction 
 with uric acid, 
 creatinine, etc. ) 
 
 Reaction — yellow crystals. 
 
 Glucose, crystals Glycuronic acid, crys- Lactose crys- 
 
 abundant. ' tals scanty, and smal- stals, glo- 
 
 ler as a rule. Gener- bular, with 
 
 ally patient taking very short 
 
 some drug. spines. 
 
 III.— Ferment Urine. 
 
 Distinct evidence of 
 gas = glucose. 
 
 IV. — Fehling's Test after 
 Fermentation Test. 
 
 No gas. 
 
 Glucose under 
 OT per cent. 
 
 /'After action of yeast 
 
 for twenty - four 
 
 -. hours, urine does 
 
 not reduce Fehling's 
 
 ^ solution. 
 
 Glycuronic 
 acid. 
 
 Lactose. 
 
 f Still reduces 
 \ Fehling's 
 (^ solution. 
 
 III. Quantitative Estimation of Sugar in the Urine. 
 
 By the intensity of the reaction to the ordinary tests we 
 may determine whether the urine contains merely a trace of 
 sugar, or a large quantity ; but often it is of service to determine 
 the exact amount, since the quantity of sugar excreted daily is 
 one indication of the severity of the case and of the course of the 
 disease, as will be pointed out in a subsequent chapter. 
 
 The following are the methods chiefly used for the quantitat- 
 ive estimation of glucose in the urine: — 
 
 1. Fermentation method of Sir William Roberts ( 35 ). — This is the 
 most simple method. It gives results which, though not quite 
 
40 SUGAR TESTS. 
 
 accurate, are still sufficient for clinical purposes. The whole 
 time actually required for the estimation is not more than four 
 or five minutes, but it is necessary to wait for twenty-four hours 
 for the complete destruction of the sugar by fermentation. 
 
 As a method of estimating the daily sugar excretion, in order 
 to judge of the value of various methods of treatment in reducing 
 glycosuria, the fermentation test is most convenient. The 
 method is founded on the diminution of the specific gravity 
 which occurs in saccharine urine after fermentation with yeast. 
 The sugar is decomposed, and alcohol and carbonic acid are 
 formed. The reduction of specific gravity is due to (1) the 
 destruction of glucose, (2) to the presence of alcohol formed by 
 fermentation. Sir William Eoberts pointed out, many years ago, 
 that each degree of specific gravity lost by saccharine urine, through 
 fermentation, corresponds practically to one grain of sugar to the ounce. 
 
 The following is the method of performing the test. About 
 4 oz. of the diabetic urine are placed in a 12-oz. bottle, and a 
 piece of German yeast about the size of a small walnut is added. 
 Into the mouth of the bottle is placed a cork which is nicked, in 
 order that the carbonic acid gas which is formed by fermentation 
 may escape. The bottle is put in some warm place, so 
 that fermentation may occur. A second bottle, containing 
 4 oz. of the diabetic urine, without any yeast, is placed 
 beside the first bottle. At the end of twenty-four hours the 
 urine in the first bottle will have fermented, the sugar will have 
 been decomposed, and the specific gravity will be diminished. 
 Two urine glasses are taken ; one is filled with the fermented 
 urine, the other with the unfermented urine. The specific 
 gravity of each is ascertained by the urinometer. The difference 
 of the two will indicate the number of grains of sugar per ounce. 
 
 For example — 
 
 Specific gravity of unfermented urine . 1040 
 
 of fermented ,, . . 1005 
 
 Difference . . . 35° of sp. gr. 
 .*. Urine contains 35 grs. of sugar to the ounce. 
 
 If the total quantity of urine for the twenty-four hours be 
 200 oz., for example, then the total excretion of sugar for the 
 day will be 7000 grs. 
 
 In taking the specific gravity, the ordinary urinometer may 
 
Q UANT1TA TI VE ESTIMA TION OF S UGAR. 4 1 
 
 be used, but for greater accuracy urinometers have been made 
 with very long scales, so that a difference of a quarter of a 
 degree of specific gravity can be readily detected. 
 
 There is one important precaution in employing this method : 
 it is necessary to see that the urine to which yeast has been 
 added has completely fermented. It sometimes happens that 
 the sugar has not all been decomposed, and then, of course, the 
 results are unreliable, the figures being too low. Hence, before 
 taking the specific gravity, it is well to test the urine specimen 
 to which yeast has been added, for sugar. If it should still reduce 
 Fehling's solution on boiling, then the sugar has not been com- 
 pletely decomposed, and the bottles should be put back for some 
 time in order that fermentation may be completed. Before the 
 specific gravity is taken, the urine to which yeast has been added 
 shoidd be quite free from sugar xvhen tested with Fehling's solidion. 
 
 Should it be desired to record the percentage of sugar in the 
 urine, it is only necessary to multiply the difference in the 
 specific gravity of the fermented and unfermented urines {i.e. the 
 number of grains of sugar per ounce) by 0'23. Thus, if the 
 difference of the specific gravity of the two be 35, then the 
 percentage of sugar will be 35 x 0'23 = 8 - 05. 
 
 2. Estimation by Fehling's solution. — One volume of diabetic 
 urine is well mixed with nine volumes of water. A burette 
 graduated in cubic centimetres is filled with the diluted urine. 
 Ten c.c. of Fehling's solution are placed in a flask or white 
 porcelain capsule and diluted with about twice their volume of 
 water. Now the Fehling's solution is of such a strength that 
 all the copper which is contained in 10 c.c. is reduced by 
 0"05 grms. of glucose (see p. 9 for composition of Fehling's 
 solution). In estimating the amount of sugar by this method, it 
 is simply necessary to find the quantity of urine which is needed 
 to reduce completely this 10 c.c. of Fehling's solution. Such a 
 quantity of urine must contain 0"05 grms. of sugar. 
 
 In performing the test, the diluted Fehling's solution is 
 raised to the boiling point, and then a little of the diluted 
 diabetic urine is allowed to drop in from the burette. The 
 Fehling's solution is again raised to the boiling point, when oxide 
 of copper is precipitated ; it is then allowed to stand for a few 
 seconds. If the blue colour still remains, more of the dilute 
 urine should be added, and the mixture again boiled. This 
 proceeding is repeated, time after time, until at length the blue 
 
42 SUGAR TESTS. 
 
 colour entirely disappears ; and after standing, when the oxide of 
 copper falls to the bottom, the fluid is colourless. Then we 
 know that all the copper has been reduced in the 10 c.c. of 
 Fehling's solution. From the strength of this solution (as above 
 stated) we know that 0'05 grms. of sugar are required to reduce 
 the 10 c.c. This quantity of sugar is contained, therefore, in 
 the amount of diluted urine added. Thus, for example, if the 
 amount of urine required to reduce the Fehling's solution be 
 15 c.c, we know that 15 c.c. of the diluted urine contains 0*05 
 grm. of sugar. But this 15 c.c. contains one-tenth of its volume 
 of urine, i.e. l - 5 c.c. Therefore l - 5 c.c. must contain 0*05 grm. 
 The following proportion gives the percentage of sugar : — 
 
 1-5 : 100 : : 0-05 
 100 x -05 
 
 = 3 3 per cent, of sugar. 
 
 1-5 
 
 In some cases, when the urine contains only a small quantity 
 of sugar, it is difficult to estimate it exactly by Fehling's solu- 
 tion, since the copper oxide remains suspended in the fluid, and 
 it is impossible to say when the blue colour has disappeared. But 
 in such cases I have found Pavy's method to be satisfactory. 
 
 3. Pavy's method ( 36 ) is of great service in estimating the 
 quantity of sugar in the urine. He employs a solution of copper 
 sulphate containing ammonia, in place of Fehling's solution. 
 When the copper sulphate is reduced to cuprous oxide by 
 glucose, no precipitation occurs, since the ammonia in the 
 solution dissolves the oxide formed ; but the solution loses its 
 blue colour. In making an analysis, the fluid containing the 
 glucose is simply dropped into the boiling test solution until its 
 blue colour completely disappears. Hence the troublesome wait- 
 ing until the oxide of copper subsides, in order to ascertain 
 whether the blue colour of the solution has disappeared, is no 
 longer necessary when Pavy's solution is used. 
 
 Composition of Pavy's Ammoniated Gujpric Test Solution. 
 Metrical System. 
 
 Cupric sulphate ...... 4'158 grms. 
 
 Potassic sodic tartrate (Rochelle salt) . . 20-400 ,, 
 
 Potash (caustic) 20-400 „ 
 
 Strong ammonia (specific gravity, - 880) . 300 c.c. 
 Water, 1 litre. 
 
PAVY'S METHOD. 43 
 
 English System. 
 
 Cupric sulphate 36^ grs. 
 
 Potassic sodic tartrate (Rochelle salt) . . . 178 ,, 
 
 Potash (caustic) . . . . . . . 178 „ 
 
 Strong ammonia (specific gravity, O880) . . . 6 fluid oz. 
 
 Water, to 1 pint. 
 
 " Dissolve the potassic sodic tartrate and potash together in 
 a portion of the water, and the cupric sulphate 
 with the aid of heat in another portion ; pour 
 the solution of cupric sulphate into the mixture 
 of potassic sodic tartrate and potash ; when cold, 
 add the ammonia ; and finally, with water, bring 
 the volume of the liquid to the bulk specified." 
 
 Pavy's solution does not deteriorate on keeping, 
 if preserved in a properly stoppered bottle. 
 
 Ten c.c. of the solution are decolorised by 
 0'005 grm. of glucose. In applying the test the 
 urine is diluted with water ; if only slightly 
 saccharine 1 in 10, if strongly saccharine 1 in 20 
 or 1 in 40. The diluted urine is then placed 
 into a burette tube. 
 
 The 1 c.c. of the test liquid are now measured 
 out and placed in a flask, such as that represented 
 in Fig. 5 ; the stopper being perforated by two 
 pieces of glass tubing. Then 20 c.c. of water are 
 added, and by the water thus used the c.c. measure 
 is rinsed out. The flask is now connected with 
 the burette by a piece of tubing, and heat from 
 a spirit lamp or gas burner applied underneath. FlG . 5. —Apparatus 
 When its contents have well commenced to boil, for quantitative 
 the urine is allowed to escape by drops at the "g^bypUy^ 
 rate of about 60 to 100 per minute into the method. 
 Fehling's solution, until the contents of the flask 
 are brought to the colourless state of water. Towards the end, the 
 dropping has to be conducted more slowly than at first, so as to 
 avoid going beyond the exact point required. The level of the 
 diluted urine in the burette having been read off before starting, 
 a second reading at the termination gives the amount which has 
 been required to decolorise the 10 c.c. of the test fluid, and 
 which therefore contains 0"005 grms. of glucose. 
 
44 SUGAR TESTS. 
 
 If, in performing, the analysis, the contents of the burette are 
 dropped in too slowly, and the boiling becomes too prolonged, the 
 suboxide falls, in consequence of the dissipation of ammonia 
 before the operation is completed. Should this event occur, a 
 fresh analysis must be performed, and the contents of the burette 
 dropped in a little more quickly. 
 
 The chief precautions are not to drop in the urine from the 
 burette so rapidly as to run the risk of passing beyond the point 
 required for decoloration, and at the same time not to drop it in 
 so slowly as to lead to a deposition of suboxide, owing to the 
 expulsion of the ammonia. 
 
 With everything conveniently at hand, a few minutes only 
 are required for the performance of the analysis from beginning 
 to end. 
 
 Air ought to be excluded from the flask as much as possible 
 during the analysis, since Pavy's fluid, after becoming decolorised, 
 soon absorbs oxygen and turns blue again. 
 
 From the amount of diluted urine required to decolorise the 
 10 c.c. of Pavy's solution, the percentage of sugar is easily 
 calculated. Supposing 5 c.c. of diluted urine be required to 
 decolorise the 10 c.c. of Pavy's solution, then we know (from the 
 strength of the Pavy's solution) that this quantity of urine must 
 have contained - 005 grm. of sugar. If the urine has been 
 diluted in the proportion of 1 to 10, the 5 c.c. of the diluted urine 
 corresponds to 0*5 c.c. of the undiluted urine. Therefore this 
 quantity must contain 0*005 grm. of sugar, and the percentage 
 of sugar can be easily calculated. 
 
 0-5 : 100 : : 0'005 
 100 x 005 
 
 = 1 per cent, of sugar. 
 
 0-5 
 
 4. GerrarcVs cyanocupric method. — This method, published by 
 A. W. Gerrard ( 37 ) in 1892, has been since improved, and by 
 some chemists is considered superior to the two methods just 
 described. The following account is taken from the description of 
 Allen ( 38 ), by whom it has been slightly modified and largely used. 
 
 When a solution of potassium cyanide and a cupric salt are 
 mixed together, a double cyanide of potassium and copper is 
 formed. 
 
 CuSO, + 4KCN = Cu(CN 2 )2KCN + K 2 S0 4 
 
PICRIC ACID METHOD. 45 
 
 The double cyanide is colourless or faintly yellow. If 
 Fehling's solution be used instead of sulphate of copper, the 
 colourless double cyanide is also formed, and oxide of copper 
 is not thrown down when this mixed solution is boiled with 
 grape sugar. If an excess of Fehling's solution be present, then 
 this extra quantity will be reduced by boiling with glucose, but 
 the oxide of copper formed will be dissolved. When the 
 Fehling's solution has been completely reduced, the blue colour 
 will disappear. The method, therefore, resembles that of Pavy. 
 The copper oxide is not precipitated, and the end of the re- 
 action is indicated by the disappearance of the blue colour. 
 But it has the advantage that no ammonia is given off; also 
 reoxidation of the reduced solution proceeds so slowly that, if 
 the test be performed fairly quickly, it is not necessary to 
 prevent the ingress of air, and a porcelain capsule may be 
 employed. 
 
 Ten c.c. of Fehling's solution are placed in a porcelain dish 
 with 40 c.c. of water, and the mixture boiled. A 5 per cent, 
 solution of potassium cyanide is added gradually until the blue 
 colour of the liquid just disappears, or only a slight tinge of 
 blue remains. Excess of cyanide must be carefully avoided. 
 Another 10 c.c. of Fehling's solution are now added, and the 
 mixture then becomes blue. The urine is rapidly dropped in 
 from a burette, until the blue colour disappears, the mixture 
 being kept in ebullition and constantly stirred. When this 
 occurs, we know, from the strength of the test solution, that 
 the quantity of urine added must have contained 0"05 grm. of 
 grape sugar. 
 
 5. Picric acid method. — Sir G-. Johnson ( 39 ) recommended the 
 picric acid test (p. 15) as a means of estimating the quantity 
 of sugar in the urine, since the depth of the colour produced 
 by the reaction is proportionate to the amount of sugar present. 
 A standard solution is prepared, which has the same dark red 
 colour as that produced by boiling picric acid and potash with 
 urine which contains a grain of glucose to the ounce, " four 
 times diluted by the reagents and water in the boiling tube." 
 As all saccharine solutions containing not less than a grain of 
 sugar to the ounce are similarly diluted in the process of 
 analysis, the standard represents one of a grain to the ounce. 
 
 The standard solution consists of liquor ferri perchloricli 
 fort., sp. gr. 1*42, 1 drm. ; glacial acetic acid, sp. gr. 1*058, 
 
46 SUGAR TESTS. 
 
 4 clrms. ; liquor ammonise, sp. gr. 0*959, 100 minims; distilled 
 water to 4 oz. 
 
 The suspected urine, liquor potassee, and saturation solution 
 of picric acid are boiled together ; then the mixture is diluted 
 with water and placed in a graduated glass vessel, and the 
 colour compared with that of the standard solution. From the 
 amount of dilution which is necessary before the tints are the 
 same, the quantity of sugar present is estimated. 
 
 6. Polarimetric method. — Grape sugar possesses the power of 
 rotating the plane of polarised light to the right, and this fact 
 forms the basis of a method for the quantitative estimation of 
 sugar in diabetic urine. 
 
 Various forms of apparatus — the polarimeters or sacchari- 
 meters of Lippich, Soleil, and others — have been constructed 
 for this purpose ; but the instruments are expensive, and the 
 method is open to fallacy, though it is said to give good, if 
 not quite accurate, results when the amount of sugar is not 
 too small. But when only a small quantity of sugar is present, 
 the results are said to be unreliable. 
 
 In severe cases of diabetes, /3-oxybutyric acid is often present 
 in the urine. Occasionally, though very rarely, lrevulose is 
 present in addition to glucose. Both of these bodies rotate the 
 plane of polarisation to the left, and therefore, if present in the 
 urine examined, the results of the polarimetric estimation of 
 glucose would be too small. 
 
 The method is rarely employed clinically, since others are 
 more reliable. Hence it is not necessary to devote further 
 space to the description of the apparatus and to the manner 
 of employing it. For such a description the reader may consult 
 works on physics. 
 
 IV. The Presence of a Trace of Sugar in 
 Normal Urine. 
 
 It is undisputed that normal urine gives no reaction with 
 Fehling's solution and the ordinary chemical tests for sugars, 
 when applied in the usual way. But it was stated by Briicke, 
 long ago, that a trace of sugar is present in normal urine, and 
 this view has been supported by Bence Jones, Kiihne, and by 
 numerous physiologists and physicians more recently ; whilst 
 others still hold that normal urine is free from anv trace of 
 
TRACE OF SUGAR IN NORMAL URINE. 47 
 
 sugar. Not long ago, the question whether a trace of sugar 
 is or is not present in normal urine, was the subject of con- 
 siderable correspondence in the Lancet (July— December 1894, 
 and January, February, March 1895). 
 
 Pavy ( 40 ), who investigated the subject carefully many years 
 since, supports the statement of Briicke, and still considers his 
 methods satisfactory. Briicke devised a process for removing 
 the supposed trace of sugar from normal urine by precipitating 
 it with oxide of lead (for details see Pavy's recent work 41 ). 
 Pavy, by using large quantities of normal urine (100 litres), has 
 satisfied himself of the truth of Briicke's statement. The 
 product which he obtained gave the ordinary reactions for sugar 
 with liquor potassse, nitrate of bismuth, and Fehling's solution. 
 At first no reaction was obtained with the fermentation test ; 
 the product was strongly acid, however, and on neutralising it 
 with sodium carbonate, and then applying the fermentation test, 
 a distinct reaction was obtained. Further, this product, obtained 
 from normal urine by special treatment, gave a reaction with 
 the phenylhydrazin test. 
 
 Pavy states that the amount of sugar in normal urine is 
 0-5 parts per 1000. 
 
 Benzoyl chloride precipitates carbohydrates from the urine, 
 and this forms an exceedingly delicate test. Wedenski ( 42 ), by 
 shaking up large quantities of normal urine with benzoyl 
 chloride, has obtained a white granular precipitate like that 
 yielded by glucose. By a special process applied to a large 
 quantity of normal urine, Moritz ( 43 ) has obtained yellow needle- 
 shaped crystals by the phenylhydrazin test. These crystals are 
 not due to glycuronic acid, since their melting point is 204° to 
 205° C. — the same as that of phenylglucosazone. Hence he 
 concludes that a trace of sugar is present in normal urine. 
 
 Indications of traces of sugar have been obtained only when 
 large quantities of urine have been examined and the supposed 
 sugar isolated. Certainly Pavy's estimate, 0*5 per 1000, appears 
 remarkably high. By the phenylhydrazin test, performed accord- 
 ing to the simplified method, 0-015 percent, of sugar can be easily 
 detected in saccharine urine greatly diluted with water. But 
 the test applied in the same manner to normal urine gives no 
 reaction. One would expect it to do so, if normal urine con- 
 tained the amount of sugar stated by Pavy ("05 per cent.). 
 
 G. Stillingfieet Johnson and Sir G. Johnson ( 44 ) deny that 
 
SUGAR TESTS. 
 
 normal urine contains a trace of sugar. They state that after 
 the urine has been treated with mercuric chloride, or a mixture 
 of sodium acetate and mercuric chloride (according to the method 
 proposed by G. Stilling fleet Johnson), and the creatinine thereby 
 separated, it fails to give any reaction for sugar. 
 
 Seegen ( 45 ) points out that his very sensitive test — Fehling's 
 test after nitration through animal charcoal — fails to give any 
 reaction with normal urine ; hence he concludes that normal 
 urine cannot contain any quantity of sugar above O'Ol per cent. 
 
 It is sufficient for clinical work to know that the ordinary 
 tests for sugar, performed in the usual manner, fail to indicate 
 the presence of sugar in normal urine. Whether a minute trace 
 of sugar be present or absent is a question of interest to the 
 physiologist, but is not one of any great importance to the 
 clinician, and at present opinions are divided. A full discussion 
 of the subject is given in the works of Seegen, Allen ( 46 ), and 
 Pavy. 
 
 REFERENCES. 
 
 1. Roberts, Sir Wm. . 
 
 2. ASHDOWN . . . . 
 
 3. WOLKOW AND BaU- 
 
 MANN 
 
 4. Seegen, J. . . . 
 
 5. WORM-MULLER . . 
 
 6. Seegen, J. . . . 
 
 Roberts, Sir Wm. 
 Purdy, C. . . 
 
 10. Johnson, G. . 
 
 11. Oliver, G. . 
 
 12. Hoppe-Seyler 
 
 " Urinary and Renal Diseases," Fourth edition, 
 
 London, 1885, p. 209. 
 Brit. Med. Journ., London, 1890, vol. i. p. 
 
 169. 
 Ztschr. f. physiol. Chem., Strassburg, 1891, 
 
 Ed. xv. 
 " Der Diabetes Mellitus," Berlin, 1893, S. 226. 
 Arch. f. d. ges. Physiol., Bonn, Bd. xxvii. S. 
 
 107. 
 Brit. Med. Journ., London, 1872, vol. i. p. 
 
 469. 
 "Der Diabetes Mellitus," Berlin, S. 230-232. 
 Practitioner, London, January 1896. 
 " Practical Uranalysis and Urinary Diag- 
 nosis," Philadelphia, 1894, p. 103. 
 "Medical Lectures and Essays," London, 
 
 1887, pp. 795, 806; Lancet, London, July 
 
 7, 1894. 
 "On Bed-side Urine Testing," London, 1889, 
 
 p. 151. 
 Ztschr. f. physiol. Chem,, Strassburg, Bd. 
 
 xvii. S. 83 ; Munchen. med. Wchnschr., 
 
 January 7, 1896 (in report of meeting of 
 
 Physiological Society). 
 
REFERENCES. 
 
 49 
 
 13. Allen, A. H. 
 
 14. Salkowski 
 
 Leube 
 
 15. Moritz 
 
 16. JOLLES 
 
 17. Moritz 
 
 18. Einhorn . 
 
 19. KOBRAK . 
 
 20. Moritz . 
 
 21. Fischer, E. 
 
 22. v. Jaksch . 
 
 23. do. 
 
 24. Binet . . 
 
 25. Moritz 
 
 26. Jolles . 
 
 27. Moritz . 
 
 28. Thierfelder 
 
 29. Fluckiger 
 
 30. Geyer . . 
 
 31. Jolles . . 
 
 32. Pavy, F. W. 
 
 33. Seegen . . . 
 
 34. Salkowski . . 
 
 35. Eoberts, Sir Wm. 
 
 36. Pavy .... 
 
 37. A. W. Gerrard 
 
 38. Allen, A. H. . 
 
 4 
 
 "Chemistry of Urine," London, 1895, pp. 
 
 81, 83. 
 " Die Lehre vom Harn," Berlin, 1882. 
 
 Deutsches Arch. f. klin. Med., Leipzig, Bd. 
 
 xlvi. S. 259. 
 Centralbl. f. klin. Med., Bonn, November 3 
 
 and 10, 1894. 
 Munchen. med. Wchnschr., 1891, Nos. 1 
 
 and 2. 
 Virchoio's Archiv, Bd. cii. S. 263. 
 " Zum Nachweis kleiner Zucker Mengen in 
 
 Harn," Diss., Breslau, 1887. 
 Munchen. med. Wchnschr,, 1891, Nos. 1 
 
 and 2. 
 Ber. d. deutsch. chem. Gesellsch., Berlin, 1882, 
 
 Bd. xvii. S. 579. 
 Ztschr. f. klin. Med., Berlin, 1886, Bd. xi. 
 
 S. 20. 
 " Clinical Diagnosis," translated by J. Cagney, 
 
 London, 1890, p. 226. 
 Rev. med. de la Suisse Horn., Geneve, Feb- 
 ruary 10, 1892. 
 Deutsches Arch. f. klin. Med., Leipzig, Bd. 
 
 xlvi. S. 255. 
 Centralbl. f. innere Med., Leipzig, November 
 
 3 and 10, 1894. 
 Munchen. med. Wchnschr., 1891, Nos. 1 and 2. 
 Ztschr. f. physiol. Chem , Strassburg, Bd. 
 
 xi. S. 395. 
 Jahresb. il. d. Fortschr. d. Thier-Chem., 
 
 Wiesbaden, Bd. xviii. S. 106. 
 Ztschr. f. physiol. Chem., Strassburg, Bd. ix. 
 
 S. 323. 
 Loc. cit. 
 " Physiology of the Carbohydrates," London, 
 
 1894, p. 13; Lancet, London, April 17, 
 
 1897. 
 Loc. cit., p. 229. 
 
 Berl. klin. Wchnschr., April 29, 1895. 
 Loc. cit, p. 224. 
 
 Lancet, London, 1st March 1884. 
 Pharm. Journ., London, vol. xxiii. p. 208. 
 Loc. cit., p. 74. 
 
5° 
 
 SUGAR TESTS. 
 
 39. Johnson, Sir G. 
 
 40. Pavt, F. W. . . 
 
 41. do. . . 
 
 42. Wedenski . , 
 
 43. Moeitz . . , 
 
 44. Johnson, Sir G. 
 
 45. Seegen, J. . . 
 
 46. Allen, A. H. 
 
 "Medical Lectures and Essays," pp. 795-799 ; 
 Lancet, London, 7th July 1894. 
 
 Guy's Hosp. Rep., London, 1876. 
 
 " Physiology of the Carbohydrates," London, 
 1894, p. 179. 
 
 Ztschr. f. physiol. Chem., Strassburg, Bd. xiii. 
 S. 122. 
 
 Deutsches Arch. f. Idin. Med., Leipzig, Bd. 
 xlvi. S. 257. 
 
 Lancet, London, 7th July 1894 (correspond- 
 ence for several weeks following). 
 
 "Der Diabetes mellitus," Berlin, 1893, p. 46. 
 
 "Chemistry of Urine," London, 1895, pp. 
 18-20. 
 
CHAPTEE III 
 
 PHYSIOLOGICAL CONSIDEKATIONS. 
 
 In health the blood contains a small and fairly constant quan- 
 tity of grape sugar ; but the exact source, the destiny of this 
 sugar, the nature of the glycogenic function of the liver, and 
 many physiological problems connected therewith, have been the 
 subject of endless discussion during the last fifty years. And 
 since there is still such a great divergence of opinion on these 
 physiological questions, it will only be possible, in a work 
 devoted to the clinical and practical side of diabetes mellitus, to 
 very briefly refer to some of the more important facts and 
 theories bearing on the physiology of sugar formation. 
 
 1. The carbohydrates derived from food articles are carried 
 from the alimentary canal by means of the portal vein to the 
 liver. These carbohydrates vary in quantity and nature. They 
 consist of glucose, kevulose, saccharose, lactose, maltose, and 
 traces of dextrin. Naturally the amount of carbohydrates in 
 the portal vein varies much at different times, and with different 
 articles of food. 
 
 2. The liver of man and the lower animals contains a 
 carbohydrate substance allied to starch, which has been named 
 glycogen by its discoverer, C. Bernard ( x ). If the liver of one of 
 the lower animals be removed immediately after death, rapidly 
 cut into small pieces, thrown into boiling water and broken up, 
 a decoction can be obtained, which, after filtration, is opales- 
 cent, and contains glycogen. An iodine solution gives a brownish- 
 red or port-wine colour with this decoction, owing to the presence 
 of glycogen. When tested with Fehling's solution, the decoction 
 is found to contain only a small amount of sugar. The glycogen 
 can be precipitated by alcohol, and after washing and purifica- 
 tion a white powder is obtained, which gives a port- wine colora- 
 tion with iodine, but no reaction with Fehling's solution or other 
 su^ar tests. 
 
52 PHYSIOLOGICAL CONSIDERATIONS. 
 
 Sugar is produced by the action of saliva, or some other 
 amylolytic ferment, on the separated glycogen, or on the decoc- 
 tion obtained from the liver. The opalescent solution becomes 
 clear, and the chemical reactions for sugar are obtained. 
 
 Glycogen is allied to starch ; it is often termed animal 
 starch, and its chemical formula is C 6 H 10 O 5 , or some multiple 
 thereof. 
 
 If the liver be allowed to remain in the body for several 
 hours after death, especially if the body be kept in a warm 
 place, a decoction of this organ will be no longer opalescent. It 
 will be clear, and there will be little or no port-wine coloration 
 with iodine ; but the solution will contain a very considerable 
 amount of sugar. After death the glycogen in the liver rapidly 
 diminishes, and the sugar increases. It is generally believed 
 that glycogen is gradually converted into sugar, post mortem, 
 though this view is not held by some observers. The trans- 
 formation is arrested by a high temperature (100° C), and is 
 supposed to be due either to the action of the liver cells, or to 
 some ferment not yet isolated. 
 
 The amount of glycogen in the liver is very variable ( 2 ). It 
 depends on the quantity and nature of food previously taken, 
 and it disappears during starvation. If an animal be starved so 
 that all the glycogen may be assumed to be absent from the 
 liver, and then a diet rich in carbohydrates be given, a large 
 amount of glycogen is found in this organ. Starch and various 
 kinds of sugar, therefore, cause an accumulation of hepatic 
 glycogen — sometimes a very rapid accumulation. 
 
 A nitrogenous diet also causes an accumulation of glycogen 
 in the liver, but to a much less extent than is caused by a carbo- 
 hydrate diet. Flesh meat contains a certain amount of carbo- 
 hydrate in some form ; and if animals are fed on proteid such as 
 fibrin, casein, or albumin in place of meat, the amount of glycogen 
 in the liver becomes still smaller, though it is stated to be more 
 than during starvation. 
 
 Gelatin leads to the formation of some glycogen in the liver, 
 but nitrogenous food and gelatin are far less powerful than 
 carbohydrates in producing hepatic glycogen. Eats do not give 
 rise to the formation of glycogen, and an animal fed on fat alone 
 has no more glycogen in its liver than a starving animal. 
 
 Thus the amount of glycogen in the liver depends largely on 
 the food. It is least when the animal is fed on fatty food ; it 
 
GLYCOGENIC FUNCTIONS OF THE LI FEE. 53 
 
 is greater when the diet consists of fleshy food ; it is much 
 more abundant when the diet consists chiefly of starch ; but it 
 is greatest when the diet consists of carbohydrates which can 
 be rapidly absorbed, such as sugar and dextrin (Seegen, 3 ). 
 
 Probably glycogen is derived from sugar absorbed from the 
 intestines. When sugar is injected into the jugular vein, 
 glycosuria may be produced, but a very considerable quantity 
 may be injected into the portal vein without any glycosuria 
 occurring. Apparently the liver fixes the sugar, and converts 
 it into glycogen. As already mentioned, glycogen is also formed 
 from proteids. 
 
 The liver appears to act as a storehouse for the excess of 
 glycogen ; a considerable amount is stored also in the muscles. 
 Next to the liver, the muscles form the second great storehouse 
 for glycogen. The amount of muscle glycogen is said to be 
 influenced by food ; by feeding animals on rice and cane sugar, 
 it is greater than when the diet consists of fat and fibrin. 
 During starvation the muscle glycogen disappears. Glycogen is 
 a normal constituent of muscle, though it is not invariably 
 present, and the amount varies in different animals and in 
 different muscles. The muscles of the embryo contain a large 
 amount. When the nerve to a muscle is divided, the glycogen is 
 increased ; rigor mortis causes it to diminish or to disappear 
 entirely ; apparently the muscle glycogen is then converted into 
 dextrose. 
 
 Glycogen cannot be regarded as a necessary part of ordinary 
 muscle substance, since muscles free from glycogen may be able 
 to carry on long-continued contractions. 
 
 As Michael Foster ( 2 ) puts it, the glycogen in the muscles of 
 the embryo and of the adult is probably a local branch of the 
 great carbohydrate bank ; it appears to be a store of carbo- 
 hydrate material, like vegetable starch, packed away so that it 
 can be drawn upon when required. 
 
 3. What becomes of the glycogen in the liver, and what are 
 the functions of the liver with reference to sugar formation, are 
 vexed questions. There are two chief views on the subject: — 
 {a) According to one view, which was first put forward by 
 Claude Bernard ( x ), the glycogen in the liver is gradually trans- 
 formed into sugar, which is paid out into the blood, and conveyed 
 through the hepatic vein to the right side of the heart, and then 
 to the lungs, where it is burnt up. Afterwards Bernard modi- 
 
54 PHYSIOLOGICAL CONSIDERATIONS. 
 
 fied this view, and concluded that sugar is distributed to the 
 tissues of the body, and burnt up in the capillaries thereof. The 
 liver is therefore a storehouse of reserve glycogen, which is 
 paid, as required, into the hepatic vein, and thus the amount 
 of sugar in the blood is kept within a certain normal limit. 
 (b) Pavy and many other experimenters believe, however, that 
 the liver is a sugar absorbing organ, and that instead of throw- 
 ing sugar into the system, it prevents sugar passing into the 
 general circulation. 
 
 In the first place, it is important to remember that whether 
 the liver is a sugar producing or sugar destroying organ, all 
 observers agree that normal blood docs contain a very small, 
 though fairly constant, quantity of glucose. According to Ber- 
 nard, the amount of sugar in normal blood is 1 to 3 parts per 
 1000; according to Pavy, - 6 to 1*0 or a little over I/O, per 
 1000 (in animals); according to Seegen, 0"1 to 0'2 per cent, 
 in man. 
 
 It is stated that in animals kept without food, and made to 
 perform muscular work, only traces of glycogen are found in the 
 liver and muscles, yet the arterial blood still contains the usual 
 quantity of sugar. , 
 
 Many years ago, Claude Bernard discovered that the liver of 
 animals fed entirely on flesh meat contained sugar, when examined 
 soon after death. In animals fed on flesh diet, he also found 
 that the blood of the portal vein was free from sugar, but that of 
 the hepatic vein was rich in sugar. Lehmann has since obtained 
 similar results. 
 
 Bernard afterwards discovered the substance in the liver 
 which he named glycogen. He found that the blood of the 
 hepatic vein and right side of the heart was comparatively rich 
 in sugar, whilst arterial blood only contained a trace of sugar. 
 He also found that there was a difference between arterial and 
 venous blood; that arterial blood contained 0*3 parts per 1000 
 more sugar than venous blood. Bernard therefore concluded 
 that sugar is formed in the liver from glycogen, and transmitted 
 to the general circulation to be burnt up in the tissues. 
 
 These statements have been disputed by some observers. 
 Pavy ( 4 ), by passing a catheter into the right side of the heart, 
 obtained blood from the living animal. He states that blood 
 obtained in this way contains only traces of sugar, whilst in the 
 dead animal the blood of the rio-ht side of the heart is rich 
 
SUGAR FORMATION IN THE LIVER. 55 
 
 in sugar; and that blood obtained during life from the right 
 ventricle contains only 0"4 to - 7 parts per 1000 of sugar, i.e. 
 not more than the amount in the carotid blood ; whilst blood 
 from the right ventricle after death contains 5 to 9 '4 parts per 
 1000. Hence, in the latter case, Pavy regards the excess of 
 sugar in the blood obtained from the right side of the heart 
 simply as the result of a post-mortem change. Bitter has 
 obtained similar results. 
 
 M'Donnell ( 5 ), by means of a catheter, removed blood during 
 life from the right side of the heart in twelve dogs, which had 
 been fed for some weeks exclusively on meat diet. In seven of 
 these he could not detect any sugar ; in five, traces were present. 
 In order to determine whether sugar was present in the 
 liver, Pavy removed pieces of this organ from animals immediately 
 after they were killed. The pieces of liver were placed at once 
 in a freezing mixture, broken up into fragments, and the pulp 
 thrown into boiling water. A filtered decoction gave either no 
 reaction or only a minimal reaction for sugar. 
 
 Eitter removed pieces of the liver from animals during life ; 
 the liver fragments were thrown into boiling water, but no 
 reaction for sugar was obtained. Pieces of the liver removed 
 a few minutes after death and treated in the same manner were 
 found to contain sugar. 
 
 M'Donnell ( 5 ) obtained similar results. A portion of the 
 liver removed immediately after death gave scarcely any indi- 
 cation of the presence of sugar, but a similar portion removed 
 twenty minutes after death was found to contain 12*5 grs. of 
 glucose. 
 
 M'Donnell ( 5 ) also made more conclusive experiments on the 
 hedgehog. The animal was surrounded by ice until it became 
 torpid. The application of cold was continued until the animal 
 was frozen into a solid mass. When the liver was removed, it 
 did not give the slightest reaction for sugar. The liver of a 
 frog frozen slowly was found to contain no sugar ; but if a frog 
 was frozen rapidly, the animal struggled considerably, and the 
 liver was found to contain a small amount of sugar. 
 
 Bernard in his last work stated that pieces of liver taken 
 from the body during life and thrown into boiling water con- 
 tained - 23 to 0'24 per cent, of sugar. Pavy has criticised 
 Bernard's results, and believes that too little water was used in 
 the method employed for the estimation of sugar. 
 
56 PHYSIOLOGICAL CONSIDERATIONS. 
 
 As already mentioned, Claude Bernard at first suggested that 
 the sugar in the blood was destroyed in the lungs. Afterwards 
 he modified this view, and came to the conclusion that blood 
 sugar was destroyed in the capillaries of the general circulation. 
 From eleven observations, Pavy found, however, that the average 
 amount of sugar in arterial blood was 0'941 parts per 1000, 
 in venous blood 9 3 8. But he thinks the excess in arterial 
 blood is too slight for the difference to be reliable. 
 
 According to Pavy, the liver prevents the sugar absorbed by 
 the portal circulation reaching the system. [A minute quantity 
 does reach the general circulation, however, since the blood of 
 the systemic circulation contains a very small quantity of sugar.] 
 According to Pavy and many authorities, normal urine also 
 contains a very minute trace of sugar. Pavy holds that if sugar 
 were continually formed by the liver and passed into the general 
 circulation, as Bernard asserted, it would be eliminated by the 
 kidneys, and the urine of every person would contain sugar in 
 quantity. He believes that the sugar in the blood and in the 
 urine run parallel, and that any excess of sugar in the blood is 
 eliminated in the urine. 
 
 Biedl and Kraus ( 6 ), however, have recently made intra- 
 venous injections of 2 00 to 300 grms. of grape sugar in four 
 persons ; by this proceeding they estimated that the percentage 
 of sugar in the blood was raised to more than three times the 
 normal amount, yet no glycosuria or polyuria followed. 
 
 F. Voit ( 7 ) has recorded a number of interesting facts with 
 reference to sugar destruction in the system. He points out 
 that grape sugar, which ferments easily, must be taken in large 
 quantities before it appears in the urine in a healthy person, 
 whilst sorbose, which does not ferment, is excreted in the urine 
 after small doses. Cremer has shown that the sugars which 
 ferment most easily pass into the urine with greatest difficulty 
 in health, whilst those which do not ferment appear in the 
 urine most readily. The sugars which do not ferment do not 
 form glycogen, they pass into the blood in great quantity, and 
 are, in part at least, excreted in the urine. 
 
 By the subcutaneous injection of sugar the changes produced 
 in the alimentary canal are avoided. Voit has made sub- 
 cutaneous injections in man, and introduced solutions of various 
 kinds of sugar under the skin. If the solutions used were not 
 too concentrated, and if the instruments were carefully sterilised, 
 
Injected. 
 
 Found in Urine 
 
 100-00 grms. 
 
 2-64 
 
 60-00 „ 
 
 Traces. 
 
 11-24 „ 
 
 
 10-94 „ 
 
 0-99 
 
 10-13 „ 
 
 Traces. 
 
 9-23 „ 
 
 0-16 
 
 9-58 „ 
 
 Traces. 
 
 SUBCUTANEOUS INJECTION OF SUGAR. 57 
 
 no bad effects followed. Ten per cent, solutions of sugar were 
 employed. 
 
 Of the monosaccharides, experiments were made with dextrose, 
 lsevulose, and galactose. These three substances did not appear 
 in the urine, or only appeared in very small quantities after 
 subcutaneous injection. They were therefore rapidly destroyed 
 in the organism. 
 
 Dextrose 
 
 Lsevulose 
 Galactose 
 
 Dextrose and lsevulose are therefore good glycogen-builders. 
 With galactose there is a greater excretion (in proportion to 
 amount injected). 
 
 Of the disaccharides — saccharose, lactose, and maltose were 
 tried. 
 
 Saccharose . 
 
 Lactose 
 
 Maltose ... 
 
 After the subcutaneous injection of cane and milk sugar, the 
 whole of these substances was found again in the urine, but 
 when given by mouth large quantities of cane sugar must be 
 taken before glycosuria occurs. In the intestine cane sugar 
 is inverted and dextrose and lsevulose are formed, but by 
 subcutaneous injection this change is prevented. The cells of 
 the organism have thus no power of decomposing cane sugar 
 circulating in the blood, and the liver cannot form glycogen 
 from it. 
 
 Lactose is excreted by the urine after subcutaneous in- 
 jection ; but when given by mouth it only appears in the 
 urine after considerable quantities have been administered. 
 
 Injected. 
 
 Found in Urine. 
 
 25-60 
 
 grms. 
 
 24-88 
 
 10-81 
 
 jj 
 
 10-71 
 
 9-29 
 
 )> 
 
 9-95 
 
 1-27 
 
 11 
 
 1-23 
 
 9-36 
 
 •>•> 
 
 10-06 
 
 9-05 
 
 11 
 
 9-42 
 
 1-09 
 
 5) 
 
 1-03 
 
 8-79 
 
 11 
 
 
58 PHYSIOLOGICAL CONSIDERATIONS. 
 
 When given by the mouth, milk sugar is inverted in the intestines, 
 but the cells of the organism are not able to decompose it when 
 injected subcutaneously. This explains why lactosuria occurs 
 in the puerperal state (see p. 88). Milk sugar passes directly 
 into the circulation from the mammary gland, and is therefore 
 just in the same condition as when injected subcutaneously. 
 The inverting action of the alimentary canal is avoided, hence 
 lactose appears in the urine. 
 
 Maltose does not appear in the urine when injected sub- 
 cutaneously ; it is apparently taken up by the cells of the 
 organism, and also forms glycogen in the liver. By the action 
 of certain ferments, maltose can be formed from glycogen, also 
 blood serum has the power of destroying maltose. 
 
 4. Seegen ( 8 ) of Vienna has repeated many of the experi- 
 ments of Pavy and Bernard, and has obtained results somewhat 
 different. From his observations he concludes that sugar forma- 
 tion is a normal function of the liver. He found that portions 
 of the liver (excised and thrown into boiling water) contained 
 0*4 to 0'5 per cent, of sugar. He also found that the blood of 
 the hepatic vein contained GO to 100 per cent." more sugar than 
 that of the portal vein. Whatever the nature of the food, and 
 even during starvation, the blood from the hepatic vein always 
 contained much more sugar than that of the portal vein. 
 
 Eecently Mosso ( 9 ) has made observations on the latter 
 point ; he concludes that the hepatic vein does contain more 
 sugar than the portal vein, but that the difference is very much 
 less than that stated by Seegen. 
 
 Seegen thinks that as much as 100 grms. of sugar may be 
 passed into the circulation from the liver in the twenty-four hours, 
 and he points out that by excision of the liver the amount of 
 sugar in the blood is diminished. 
 
 Minkowski has shown that when the liver is excised in 
 animals, the sugar in the blood disappears after some hours. 
 
 By ligaturing the vascular connections of the liver, and thus 
 excluding it from the circulation, Bock and Hoffmann found 
 that sugar disappeared from the blood. Seegen observed a great 
 diminution of the sugar in the blood in three similar experiments. 
 
 Tangl and V. Harley ( 10 ) have shown that by diminishing 
 the blood supply to the liver in animals, by ligaturing the three 
 intestinal arteries, the amount of sugar in blood taken from the 
 carotid artery was diminished. 
 
PAVY'S VIEWS. 59 
 
 Kaufmann ( n ) also found that the blood sugar was diminished 
 by ligaturing the vessels of the liver. 
 
 Seegen believes that albumin and fat are the materials from 
 which the liver forms sugar, and bases his conclusions on a 
 number of experiments and observations. 
 
 5. Pavy, in the Croonian lecture of 1894 ( 12 ), points out 
 that the multiplication of the yeast plant in Pasteur's fluid — 
 ammonium tartrate, 1 part; cane sugar, 10 parts; ash of yeast, 
 1 part; water, 100 parts — clearly demonstrates that sugar is 
 used in the construction of proteid material. According to 
 Pavy, a carbohydrate may also be prepared from egg albumin, 
 by a cleavage of proteid material. Pavy believes that proteid 
 matter has a glucoside constitution, and that proof has been 
 afforded that carbohydrates can be incorporated to form proteids, 
 and that it is also possible to dissociate the carbohydrate again. 
 
 Prom his experiments he concluded that — 
 
 (1) Not only can carbohydrate material be hydrated by 
 ferment and chemical action, but when in these conditions of 
 increased hydration, they can be transmuted by dehydration, 
 under the influence of protoplasmic action, to substances having 
 more complex molecules, i.e. amyloses. (2) In both the vegetable 
 and animal kingdom, carbohydrates take part in the synthesis 
 of proteids. (3) Carbohydrates are transformed into fat under 
 the influence of protoplasmic action. 
 
 Ptecently ( 1S ) a nucleo-albumin has been separated from 
 various organs — the pancreas, liver, thymus, muscles, thyroid, 
 and spleen. This nucleo-albumin has been shown to be of a 
 glucoside nature, and by special treatment a carbohydrate has 
 been separated from it, which has been proved to be a sugar — 
 pentose (Blumenthal). 
 
 6. Pavy, in his Croonian lecture of 1894, has restated his old 
 views respecting sugar formation, and has put forth a new theory 
 with reference to sugar destruction. 
 
 Briefly stated, the following are his conclusions : — 
 
 (1) The liver at death is not more saccharine than other 
 
 organs of the body ; (2) the blood flowing from the liver is not 
 
 more saccharine than that flowing to it, but the contrary ; 
 
 (3) there is no evidence of destruction of sugar in the systemic 
 
 capillaries. 
 
 After a meal of fat, the cells of the villi of the intestines 
 
 become charged with fat globules, which pass into the lacteals. 
 
60 PHYSIOLOGICAL CONSIDERATIONS. 
 
 A similar condition is met with after carbohydrate food. Oats 
 contain only 5 per cent, of fat, yet after a meal of oats the 
 lacteals in the rabbit were as fully distended with milky chyle as 
 in the dog after a meal of fat. Hence Pavy believes that the 
 carbohydrates of food are converted into fat by the protoplasm 
 of the cells of the intestinal villi. 
 
 He thinks that glycogen is formed in the liver from the 
 carbohydrates of the food which have escaped synthesis by the 
 cells of the intestinal villi. There are thus tvjo lines of defence, 
 the intestinal villi and the liver. Even in health both lines of 
 defence may be passed, and then the urine contains sugar ; this 
 happens if too much carbohydrate food is taken. 
 
 Pavy has shown that from the liver, when treated with 
 alcohol, dried, and powdered (in a special manner, which he 
 describes), a substance can be produced which may be kept 
 indefinitely in a bottle. When this powder is treated with water, 
 and placed in an incubator for two or three hours, an active 
 production of sugar takes place. He holds, therefore, that sugar 
 formation is not due to vital action, but to the presence of a 
 ferment. 
 
 The latter statement and many of Pavy's results have been 
 criticised by Noel Paton ( u ). He points out that in rats a diet 
 of fat causes the epithelium cells of the intestinal villi to be 
 loaded with fat particles, but on a diet of starch or sugar the cells 
 are free from fat globules. 
 
 Noel Paton concludes that " the early changes in the excised 
 liver are simply a continuation of the vital processes of the 
 organ — the katabolic side of metabolism being exaggerated, the 
 anabolic side in abeyance." In his last article he records the 
 results of a number of experiments, from which he comes to the 
 conclusion that " the whole weight of evidence seems to be 
 against the view that an active amylolytic zymen develops in 
 the liver immediately after death, while no material argument 
 has been adduced in opposition to the view that the conversion 
 of glycogen to glucose is simply due to katabolic changes in the 
 liver substance." 
 
 Pavy ( 15 ) has replied to the criticism of Paton, but even an 
 outline of the discussion would occupy too much space here, and 
 the reader is referred to the original articles. 
 * * * 
 
 It has only been possible to refer to some of the chief 
 
FUNCTIONS OF THE LIVER. 61 
 
 views with respect to sugar formation and the hepatic functions 
 in the above summary. 
 
 The two apparently opposite views with regard to the liver 
 functions are not, perhaps, so irreconcilable as appears at first 
 sight. Most observers admit that the liver acts to some extent 
 as a sugar absorbing or sugar converting organ, transforming 
 carbohydrates obtained from the portal circulation into glycogen. 
 Whether the liver be a sugar forming or sugar absorbing 
 organ, it is allowed by all that a small percentage of sugar 
 is present in the blood. If the function of the liver and 
 the intestinal villi be to prevent sugar entering the general 
 circulation, there can be no doubt that this action is not com- 
 plete, and that sugar does find its way into the general circulation. 
 
 Those who hold Bernard's view would say that this small 
 percentage of sugar is paid into the circulation by the liver. 
 Those who hold Pavy's view will probably admit that a small 
 quantity of sugar is allowed to leak into the circulation. After 
 all, whichever way the action of the liver is described, there is 
 no difference of opinion as to the final result, so far as the blood 
 is concerned, i.e. a small amount of sugar is present. 
 
 The diminution of sugar in the blood after excision of the 
 liver, or after exclusion of that organ from the circulation, 
 appears certainly to be in favour of the view that the liver is a 
 sugar forming organ. Then again, if the liver be an organ which 
 absorbs sugar, and which prevents sugar reaching the circulation, 
 how is it that no glycosuria occurs when the liver is seriously 
 diseased, and the liver cells destroyed to such a great extent as 
 in cases of cancer, acute yellow atrophy, and other serious 
 hepatic affections ? 
 
 KEFEKENCES. 
 
 1. Bernard, C. . . . "Legons sur le diabete," Paris, 1877. 
 
 2. Foster, M. . . . "Text-Book of Physiology," London, 1895, 
 
 pt. 2, pp. 764-765. 
 
 3. Seegex, J. . . . "Der Diabetes Mellitus," Berlin, 1893, S. 37. 
 
 4. Pavy, F. W. . . " Croonian Lectures," Brit. Med. Journ. y 
 
 London, June 1894. 
 
 5. aI'Donnell, E. . " Observations on the Functions of the 
 
 Liver," Dublin, 1865. 
 
 6. Biedl und Kraus . " Centralbl. f. innere Med., Leipzig, July 18, 
 
 1896 ; Wien. klin. Wclmschr., 1896, 
 No. 4. 
 
62 
 
 PH YSIOL O GICA L CONSIDERA TIONS. 
 
 7. Voit, F 
 
 8. Seegen, J. . . . 
 
 9. Mosso .... 
 
 10. TanglundVaughan 
 
 Ha RLE Y 
 
 11. Kaufmann . . . 
 
 12. Pavy, F. W. . . 
 
 13. Blumenthal, F. 
 
 14. Paton, Noel . . 
 
 15. Pavy, F. W. . . 
 
 Miinchen. med. Wchnsclir., 22nd September 
 
 1896. 
 Op. cit., pp. 1-45 ; also " Die Zuckerbildung 
 
 ira Thierkorper," Berlin, 1890. 
 Centralbl. f. inner e Med., Leipzig, 13tli 
 
 February 1897. 
 Arch. f. d. ges. Physiol., Bonn, 1895, 
 
 Bd. lxi. 
 Semaine med., Paris, 16th January 1895. 
 " Croonian Lectures," Brit. Med. Journ., 
 
 London, 23rd June 1894. 
 Berl. Bin. Wchnschr., 22nd Marcb 1897. 
 Edin. Med. Journ., December 1894 ; Phil. 
 
 Trans., London, 1894, vol. clxxxv. p. 233 ; 
 
 Journ. Physiol., Cambridge and London, 
 
 1897, vol. xxii. pp. 121-136. 
 Brit. Med. Journ., London, 22nd February 
 
 and 7th March 1896; ibid., 18th July 
 
 1896, pp. 147-148. 
 
CHAPTER IV 
 
 EXPERIMENTAL DIABETES AND GLYCOSURIA. 
 
 We have seen in the previous chapter that the glycogenic 
 functions of the liver, sugar formation in the system, and sugar 
 destruction, etc., are all still vexed questions. 
 
 The theories which have been put forward to explain diabetes 
 mellitus are exceedingly numerous, and the views are even more 
 varied than those with reference to the glycogenic functions of 
 the liver and sugar formation. 
 
 The riddle has not yet been solved, and probably the day is 
 far distant when a satisfactory answer will be given. Never- 
 theless, many important facts have been ascertained, and 
 experiments on animals have furnished much interesting 
 information. 
 
 Only the more important results of experimental work can 
 be referred to, however, in a book devoted to the clinical side of 
 the subject. 
 
 Claude Bernard ( x ) found, many years ago, that 'puncture of 
 the floor of the fourth ventricle in a very limited space produced 
 glycosuria in animals. The point of puncture was situated 
 between the deep origins of the vagus and auditory nerves, i.e. 
 about the region of the vasomotor centre. When this experi- 
 ment is performed on a well-fed rabbit, its urine in the course 
 of one or two hours becomes increased in quantity, and contains 
 a considerable amount of sugar. A little later the quantity of 
 sugar reaches its maximum. It then diminishes, and in the 
 course of a day or two the urine becomes normal again. The 
 better the rabbit is fed, the more glycogen its liver contains, 
 and the greater the amount of sugar in the urine. If the 
 animal be starved before the puncture of the fourth ventricle, so 
 that little or no glycogen is present in the liver, then the urine 
 will contain little or no sugar. Hence it is assumed by many 
 physiologists that, when the fourth ventricle is punctured, the 
 
64 EXPERIMENTAL DIABETES AND GLYCOSURIA. 
 
 glycogen in the liver is suddenly converted into sugar, and passed 
 into the general circulation — the liver is suddenly " emptied of 
 its glycogen." Hyperglycemia is the result, and as a conse- 
 quence glycosuria occurs. 
 
 Claude Bernard found that, after puncture of the fourth 
 ventricle, diabetes continued, even though the vagi were divided. 
 Galvanisation of the peripheral ends of the divided vagi did not 
 modify the gly oogenesis, but galvanisation of the central ends 
 gave rise to an increased formation of sugar. 
 
 The liver is supplied with nerves from the hepatic plexus, 
 which accompany the hepatic artery and portal vein, and which 
 are distributed to various parts of the organ. This plexus is an 
 extension of the great solar plexus. The right (posterior) vagus 
 sends the greater part of its fibres to the solar plexus ; and the 
 splanchnic nerves (major and minor) end in it, on both sides 
 of the body. The left (anterior) vagus forms only slight connec- 
 tions with the solar plexus, but sends off a very distinct branch 
 directly to the hepatic plexus. The liver, therefore, has nervous 
 connections with the central nervous system by the vagi, and 
 is also connected with the splanchnic nerves. 
 
 The nature of the influence or impulses started by puncture 
 of the floor of the fourth ventricle, and the paths by which they 
 reach the liver, are not definitely known. Since the " diabetic 
 centre " is close to, or almost identical with, the vasomotor 
 centre in the floor of the fourth ventricle, it is possible that the 
 changes in the liver are vasomotor in nature ; but there is no 
 direct evidence in support of this view. Section of the splanchnic 
 nerves, the channel by which the vaso-constrictor impulses pass 
 to the liver, does not produce diabetes. It seems more probable 
 that the nervous events produced in the medulla by puncture of 
 the floor of the fourth ventricle are able to bring about changes 
 in the hepatic cells. The path by which the nerve impulses 
 pass is not well defined. They do not travel by the vagus, for 
 puncture is effective after division of both vagi. They probably 
 make their way to the liver by the sympathetic system, passing 
 into the sympathetic chain in the upper thoracic region. If it 
 be true, as stated, that the puncture fails when both splanchnic 
 nerves are divided, the impulses probably travel along those 
 nerves. 
 
 Pavy ( 2 ) found that division of the medulla gives rise 
 to glycosuria when the circulation is kept up artificially; but 
 
LESIONS OF THE NERVOUS SYSTEM 65 
 
 division of the spinal cord below the origin of the phrenic nerves 
 does not produce glycosuria. Also, when the cord is divided 
 high up in the spinal canal, between the second and third 
 cervical vertebrae (artificial respiration being maintained), no 
 sugar is found in the urine. 
 
 In operating on the brain, Pavy found that many complica- 
 tions arose, but when these were least, there was no sugar in the 
 urine after division of the crura cerebri just in front of the pons. 
 
 Although no diabetic effect was observed after division 
 of the cord and vagi together, yet after division of everything 
 belonging to the nervous system in the neck, as by decapita- 
 tion, the urine became in a very short time strongly saccharine 
 (artificial respiration being performed). 
 
 Eckhard ( 3 ) has shown that mechanical injury of the vermi- 
 form process of the cerebellum produces glycosuria. 
 
 Schiff ( 4 ) has found that diabetes can be produced by 
 injury of the nervous system at various parts. In addition to 
 puncture of the floor of the fourth ventricle, he found that 
 injury to the pons also gave rise to diabetes ; but here the 
 puncture of a needle was not sufficient ; a broad instrument was 
 necessary, or the needle had to be moved about in the wound 
 right and left. Schiff also produced diabetes by division of 
 the spinal cord at various levels ; by division of the cord be- 
 fore and behind the origin of the brachial nerves ; in frogs 
 and rabbits, by division of the posterior columns. Also after 
 complete division of the spinal cord in the rat, at the level 
 of the last cervical or first two dorsal vertebrae, diabetes was 
 produced, if sinking of the bodily temperature was prevented. 
 
 Pavy found that division of the carotid sympathetic in the 
 neck did not give rise to glycosuria. But division of the 
 sympathetic fibres accompanying the vertebral artery produced 
 saccharine urine. Ligature of the two vertebrals and two 
 carotid arteries did not occasion saccharine urine ; nor did 
 simply tearing through everything traversing the vertebral 
 canal on either side of the neck ; but, by combining the two 
 operations, sugar rapidly appeared in the urine. Though 
 division of the cervical sympathetic does not produce any effect, 
 yet removal or injury of the superior cervical ganglion may 
 cause marked glycosuria in a very short time. 
 
 Pavy states that the diabetes resulting from all these opera- 
 tions on the sympathetic is quite of a temporary character, 
 5 
 
66 EXPERIMENTAL DIABETES AND GLYCOSURIA. 
 
 and is prevented by the injection of carbonate of soda (200 grs.) 
 into the circulatory system, previous to the experiment. 
 
 Arthaud and Butte ( 5 ) believe that the changes producing 
 diabetes occur in the district of distribution of the vagus, and 
 regard hypersecretion of the liver parenchyma as the most 
 probable cause. 
 
 In order to produce permanent irritation of the vagus, they 
 injected lycopodium powder, or a solution of croton-oil in ether 
 and alcohol, into the cervical part of the nerve. By experiments 
 on both vagi death occurred too rapidly. Therefore the right 
 vagus only was injected. 
 
 In a series of experiments in which the above mentioned 
 injections were made either into the uninjured nerve, or, after 
 section, into the peripheral end of the same, the following 
 changes were noted: — Wasting, polyuria, polydipsia, polyphagia, 
 glycosuria (not constant, but sometimes marked), albuminuria, 
 and azoturia. Neuritis produced in the central end of the 
 divided vagus was followed by slight glycosuria and temporary 
 azoturia. 
 
 In their further researches the authors produced neuritis in 
 both vagi by the injection of powders. One vagus was injected 
 some weeks after the other. After the injection into one vagus, 
 polyuria and slight albuminuria were observed. After the injec- 
 tion into the other vagus, gastric disturbances, thirst, wasting, 
 and glycosuria were noted. 
 
 The authors conclude that, through centrifugal vagus irrita- 
 tion in animals, the various clinical forms of human diabetes can 
 be simulated. 
 
 Pavy discovered that, after ligaturing the portal vein and 
 allowing the blood of the hepatic artery only to reach the liver, 
 the contents of the circulatory system became highly charged with 
 sugar. No sugar was found in the urine, but this was prob- 
 ably owing to the accumulation of blood to such a great extent 
 in the portal system that the supply to the kidney was insuffi- 
 cient to permit the secretion of urine. 
 
 Pavy found that the injection of defibrinated arterial or 
 oxygenated blood into the portal system induced marked 
 glycosuria, but the injection of defibrinated venous blood did not 
 give this result. He also produced glycosuria by causing dogs 
 to breathe oxygen or carbonic oxide instead of air, but this 
 result was not obtained in all cases. Pavy believes that blood 
 
PA FY'S EXPERIMENTS. 67 
 
 containing an excess of oxygen or blood containing carbon 
 monoxide acts upon the amyloid substance of the organism, 
 and leads to its abnormal transformation into sugar. 
 
 Schiff asserted, years ago, that Bernard's puncture produced 
 dilatation of the small vessels of the intestine and liver — a 
 paralytic hyperemia of the organs, and he suggested that in the 
 hypercemic state a ferment probably developed which acted upon 
 the amyloid substance of the liver and gave rise to glycosuria. 
 
 Pavy believes that, owing to a vasomotor paralysis produced 
 by a lesion of the nervous system, an imperfectly de-arterialised 
 blood finds its way into the portal vein, and this determines the 
 escape of sugar from the liver. But hyperemia of the liver 
 alone is not sufficient to cause glycosuria. Division of all the 
 nerves going to the liver, which might be expected to produce 
 great hyperemia, is not followed by glycosuria. 
 
 Pavy thinks the above explanation is probably true for 
 natural diabetes. He believes that diabetes is produced by a 
 loss of power in the vasomotor centres, or by a lesion of some 
 part of the cerebro-spinal system which leads to an inhibitory 
 influence being exerted upon these centres. 
 
 In the last Croonian lecture, and in his recent work on the 
 Physiology of the Carbohydrates, Pavy ( 6 ) points out that in 
 diabetes the two lines of defence — the villi of intestine and the 
 liver — are inadequate to accomplish their function of synthesising 
 the carbohydrates, and thus carbohydrates reached the general 
 circulation in excessive quantity, and appear in the urine as 
 sugar. A diabetic person has not the protoplasmic power 
 sufficient to dispose of the carbohydrates in his diet by synthesis 
 to proteids, by transmutation into fat, and by dehydration into 
 glycogen. Pavy believes that this power is influenced by the 
 state of the blood vessels, which again is determined by the con- 
 dition of the nervous system. A hyper-oxygenated state of the 
 blood favours the passage of carbohydrate matter into glucose, 
 and is capable of producing saccharine urine. Dilatation of the 
 hepatic arterioles permits the blood to pass too rapidly through 
 the capillaries, so that the blood in the veins is not sufficiently 
 de-arterialised. Under such circumstances, the passage of carbo- 
 hydrates into glucose is favoured. 
 
 Pavy thinks that the whole history of diabetes goes to prove 
 that it is frequently due primarily to changes in the nervous 
 system. But in many severe cases there is some other abnor- 
 
63 EXPERIMENTAL DIABETES AND GLYCOSURIA. 
 
 mality ; the most rigid diet does not check the glycosuria. Also 
 complete absention from all food does not suffice altogether to 
 prevent the appearance of sugar in the urine. The tissues there- 
 fore must produce sugar, and since they consist mainly of pro- 
 teid matter, which has a glucoside constitution, Pavy thinks it 
 is only necessary to suppose the existence of a ferment action to 
 explain the pathogenesis. 
 
 Extirpation of the cceliac plexus. — The coeliac plexus has 
 been regarded by some writers as the seat of the lesion in 
 diabetes. This plexus has been extirpated by several observers, 
 but the results have varied somewhat. The following are those 
 obtained by Lustig and Peiper. 
 
 After the removal of the cceliac plexus from dogs and 
 rabbits, Lustig ( 7 ) observed the following symptoms : — The 
 digestive system was not affected ; there was nothing abnormal 
 as regards the appetite ; the faeces had their usual appearance ; 
 a temporary glycosuria occurred frequently ; there was no 
 atrophy of the pancreas ; acetonuria was a constant result, and 
 it continued until the death of the animal. Other symptoms 
 were — increasing emaciation, sinking of the temperature below 
 normal, and slowing of respiration ; after a time albumin appeared 
 in the urine. The animals died in most cases, some survived ; 
 but in the rest, after one or more weeks, death by coma occurred. 
 
 Lustig regards the occurrence of acetone in the urine as 
 a marked symptom of the extirpation of the cceliac plexus. 
 
 Peiper ( s ) gives the results of experiments on fifteen rabbits. 
 Eleven animals survived at least three or four weeks ; the other 
 four died partly in consequence of the ether narcosis, partly from 
 hemorrhage and peritonitis. In the animals which survived, 
 great emaciation followed the operation, without the occurrence 
 of any other marked symptom. Diarrhcea did not occur in any 
 case. There were no signs of hyperemia of the abdominal 
 organs at the post-mortem examination. Diabetes insipidus 
 was not observed. Glycosuria occurred frequently soon after 
 the operation. In one case, in which, however, an extensive 
 resection of the splanchnic nerves was undertaken, the sugar 
 in the urine amounted at various times to 2, 3, and 4 per 
 cent. Atrophy of the pancreas did not occur in any case. 
 Acetone (as tested by the methods of Lieben and Reynolds) 
 was only discovered a few times. Albuminuria was found only 
 in two cases. 
 
REFERENCES. 
 
 69 
 
 Four rabbits, which had lived two to four months, were killed, 
 but no special changes were found at the autopsy. Seven rabbits 
 showed symptoms of marked marasmus. 
 
 Thus the experiments of Peiper are opposed to the view that 
 diabetes insipidus depends upon a derangement of the coeliac 
 plexus. They also are opposed to the view that diarrhoea or 
 diabetes mellitus is produced by extirpation of the plexus. 
 Acetonuria and albuminuria are not characteristic symptoms 
 after extirpation of the plexus. 
 
 1. Bernard, C. . . 
 
 2. Pavy, F. W. . . 
 
 3. EcKHARD . . . . 
 
 4. Schifp, M. . . . 
 
 5. Arthaud et Butte 
 
 6. PAvr, F. W. . . 
 
 7. Lustig, A. . . . 
 
 8. Peiper . . . . 
 
 REFERENCES. 
 
 " Lecons sur le diabete," Paris, 1877, p. 370. 
 " Researches on the Nature and^ Treatment 
 
 of Diabetes," London, 1862. 
 Beitr. z. Anat. u. Physiol. (Eckhard), Giessen, 
 
 1867, Bd. iv. 
 " Untersuchungen ueber Zuckerbildung in der 
 
 Leber," Wiirzburg, 1859, 
 Arch, de pliysiol. norm, et path., Paris, 1888, 
 
 tome i. p. 344; Compt. rend. Soc. de biol., 
 
 Paris, tome 108, No. 4. 
 " The Physiology of the Carbohydrates," 
 
 London, 1894. 
 Arch, per le sc. med., Torino, 1889, No. 6 ; 
 
 Abstr., Centralbl. f. Idin. Med., Bonn, 
 
 May 10, 1890. 
 Miinchen. med. Wchnschr., May 29, 1890 
 
 (Congress fiir innere Medicin, Wien). 
 
CHAPTER V. 
 
 PHLOKIDZIN DIABETES. 
 
 In 1886, v. Mering ( x ) discovered that the administration of 
 phloridzin produced diabetes in dogs. 
 
 Phloridzin (phlorhizine or phloorhizine) is a substance 
 met with in the bark of the trunk and root of apple, pear, plum, 
 and cherry trees. Its chemical formula is given as C 21 H 24 O 10 + 
 2H 2 0. It consists of fine silky needles which are very soluble 
 in water at 50° C, but dissolve only in 1000 parts of cold 
 water. Dilute acids decompose it on boiling into phlorose and 
 phloretin. 
 
 C 21 H 24 O 10 + H 2 = C 15 H 14 5 + C 6 H 12 6 
 
 (phloridzin) (phloretin) (phlorose) 
 
 Phlorose is a form of sugar which reduces copper solutions. 
 By boiling with caustic potash, phloretin is converted into 
 phloretin acid, C 9 H 10 O 3 , and phloroglucin, CgH^Oy. 
 
 Given internally or hypodermically, phloridzin produces 
 glycosuria, but the hypodermic method is the more powerful. 
 For hypodermic injection phloridzin may be dissolved in a weak 
 solution of sodium carbonate, or suspended in olive-oil ; or it 
 may be given in mucilage of gum-arabic. The latter form is 
 said to be the best (Coolen 2 ). 
 
 After the administration of phloridzin, the quantity of urine 
 is increased, the specific gravity is raised, and sugar is present. 
 The form of sugar is glucose. It reduces Fehling's solution, 
 ferments with yeast, and behaves like glucose on examination 
 with the polariscope. Phloridzin does not give rise to a per- 
 manent diabetes. When the drug is discontinued, the glycosuria 
 and other symptoms disappear. If the drug be continued, the 
 animal loses weight, and by its prolonged administration death is 
 caused. Occasionally, after the drug has been given for a long 
 time, acetone and /3-oxybutyric acid have been found in the urine, 
 and symptoms like those of diabetic coma have been observed. 
 
PHL ORIDZIN DIABE TES. 7 1 
 
 v. Mering has also produced a temporary diabetes in the 
 human subject, in three cases, by the administration of phloridzin. 
 The amount of sugar in the urine was considerable. In one 
 case 1 grm. of phloridzin was given night and morning for a 
 month, and the average daily excretion of sugar was 97 grms. 
 The glycosuria disappeared the day after the phloridzin was 
 discontinued, and no bad effects or disturbance of the general 
 condition were observed. 
 
 The sugar excretion after the administration of phloridzin 
 is not dependent on the nature of the food. When dogs were 
 fed on an exclusively nitrogenous diet, glycosuria was still 
 produced by phloridzin. When large quantities of amylaceous 
 food were given, the sugar excretion was not greater than when 
 the animal was fed only on nitrogenous food. After starvation 
 for eighteen days, when it may be assumed that all the glycogen 
 had disappeared from the liver and muscle, phloridzin still pro- 
 duced glycosuria ; also glycosuria was produced even after the 
 liver had been removed in geese and frogs (v. Mering). 
 
 According to v. Mering, Minkowski, and others, in phloridzin 
 diabetes there is no excess of sugar in the blood, whilst in 
 ordinary diabetes in man the blood sugar is generally much 
 increased in quantity. Levene ( 3 ), however, has found the amount 
 of sugar in the blood often diminished, but in some cases 
 increased. Coolen states that in rabbits rendered diabetic by 
 the hypodermic injection of large quantities of phloridzin, he 
 has detected an increase of sugar in the blood (i.e. there is a 
 hyperglycemia, just as occurs in ordinary diabetes in man). 
 Pavy ( 4 ) has recently criticised the methods of analysis employed 
 by v. Mering and Levene. He finds that in cats rendered diabetic 
 by phloridzin the amount of sugar in the blood is increased. 
 There is, then, a considerable difference of opinion with regard to 
 the condition of the blood in this form of diabetes. 
 
 Minkowski ( 5 ) has shown that phloridzin diabetes is not 
 caused by any special action of the drug on the pancreas. It 
 can be produced in those animals in which diabetes does not 
 follow pancreas extirpation. Minkowski is of opinion that this 
 drug acts on the kidneys in some way, and thereby produces 
 diabetes. He found that extirpation of the kidney in animals 
 suffering from pancreatic diabetes caused an increase of the 
 sugar in the blood, whilst kidney extirpation in animals suffer- 
 ing from phloridzin diabetes did not materially affect the amount 
 
7 2 PHL ORIDZIN DIA BE PES. 
 
 of sugar in the blood. Also, in the diabetes produced in animals 
 by pancreas extirpation, phloridzin increased the amount of 
 sugar in the urine ; hence we may assume that phloridzin 
 diabetes is not the result of the action of the drug on the 
 pancreas. Levene states that the venous blood of the kidney 
 contains more sugar in some cases than the arterial blood. 
 
 Of the decomposition products of phloridzin, only phloretin 
 produces diabetes ; phlorose, phloretin acid, and phloro-glucin 
 have no effect. 
 
 v. Mering concludes, from the results of a number of ex- 
 periments, that the sugar which is excreted after phloridzin 
 administration is formed from the albumin of the body. 
 
 Kosenfeld ( 6 ) has shown that, under certain conditions, phlorid- 
 zin produces fatty liver as well as glycosuria, and he records 
 the results of a number of careful experiments on dogs. The 
 dogs were kept without food for five days ; then on the sixth 
 and seventh days 10 grms. of phloridzin were given. The 
 animals were killed on the eighth day, forty-eight hours after 
 the first close of phloridzin. Pathologically, fatty liver was 
 found, and microscopical examination showed that the condition 
 was one of fatty infiltration and not fatty degeneration, also 
 that the fat was situated chiefly around the central veins of the 
 lobules. 
 
 REFERENCES. 
 
 1. v. Mering . . . Ztschr. f. klin. Med., Berlin, Ed. xiv. S. 406. 
 
 2. Coolen .... Arch, de pharmacodynamic, vol. i. p. 267, 
 
 Fasc. 4. 
 
 3. Levene .... Journ. Physiol., Cambridge and London, Oct. 
 
 1894, 
 
 4. Pavy, F. W. . . Ibid., 1897, vol. xxi. 
 
 5. Minkowski . . . Arch. f. exper. Path. u. Pharmakol., Leipzig, Ed. 
 
 xxxi. S. 85-189. 
 
 6. Rosenfeld . . . Ztschr. f. Idin. Med., Berlin, Bd. xxviii. 
 
CHAPTEE VI. 
 
 EXPERIMENTAL PANCREATIC DIABETES. 
 
 The theory that lesions of the pancreas play some part in the 
 causation of diabetes mellitus is by no means a recent one. In 
 the year 1788, Thomas Cawley ( 1 ) found, on making the post- 
 mortem examination in a case of diabetes, that the pancreas 
 was full of calculi ; its right extremity was very hard, and 
 " appeared to be scirrhous." On looking over the literature 
 of the subject, it is interesting to note that, since the days 
 of Cawley, various physicians have, from time to time, called 
 attention to the connection between diabetes and lesions of 
 the pancreas. Thus Chopart in 1821, and Eecklinghausen in 
 1864, especially drew attention to the pancreatic lesions in 
 diabetes ; Kokitansky found pathological changes of various 
 kinds in the pancreas, in thirteen out of thirty cases of diabetes ; 
 also Lancereaux, Baumel, Cantani, and many others, have pointed 
 out the importance of these pancreatic lesions. Baumel ( 2 ) 
 in 1882 went so far as to attribute diabetes to the absence 
 of diastatic pancreatic ferment in the alimentary canal, and 
 recommended the use of pancreatic preparations in the treatment 
 of the disease. Many years ago, Lancereaux especially insisted 
 that diabetes with wasting is associated with pathological changes 
 in the pancreas. 
 
 But it was not until 1889 that Minkowski, v. Mering, 
 and de Dominicis showed, by experiments on animals, that 
 pancreatic lesions may give rise to diabetes. 
 
 With respect to the relation of pancreatic lesions to diabetes 
 mellitus, both pathological anatomy and experiments on animals 
 furnish important evidence. Apart from all theoretical con- 
 siderations, there are two facts of great interest : — 
 
 1. Total extirpation of the pancreas gives rise to diabetes 
 mellitus in dogs and many other animals. 
 
 2. The records of pathological examination show that the 
 
74 EXPERIMENTAL PANCREATIC DIABETES. 
 
 pancreas is diseased in a considerable proportion of cases of 
 diabetes mellitus in man. 
 
 In this section only the experimental evidence will be 
 considered ; the pathological evidence will be discussed in the 
 section devoted to etiology and etiological relations. 
 
 The older experimenters did not succeed in producing 
 diabetes by operations on the pancreas. Claude Bernard ob- 
 structed the pancreatic duct by injecting various substances, but 
 failed to produce diabetes thereby. Heidenhain and Finkler 
 obtained similar results ; Arnozan, Vaillard, and others liga- 
 tured the pancreatic duct in rabbits, but did not produce 
 diabetes. 
 
 In 1889, however, Minkowski and v. Mering ( 3 ) in 
 Germany, and de Dominicis in Italy, succeeded in pro- 
 ducing permanent diabetes mellitus, resembling that which 
 occurs in man, by total extirpation of the pancreas in clogs. 
 The papers recording these results were published about the 
 same time, but to the two German experimenters belongs the 
 honour of proving that total extirpation of the pancreas invariably 
 produces diabetes ; they also showed that if a small portion of 
 the pancreas be left behind at the time of the operation, diabetes 
 will not occur, even though this portion be separated from the 
 duodenum, and the pancreatic duct be ligatured. 
 
 Some experimenters have found sugar in the urine in the 
 majority of cases after extirpation of the pancreas, but not in all 
 cases. Thus de Dominicis ( 4 ) produced glycosuria in twenty-one 
 out of thirty-four cases ; Eenzi and Reale ( 5 ) in 75 per cent, 
 of the cases. De Dominicis found that the other symptoms, 
 such as polyphagia, polydipsia, polyuria, emaciation, falling off 
 of the hair, etc., invariably followed extirpation of the pancreas, 
 but that glycosuria was absent in over one-third of the cases 
 (thirteen out of thirty-four cases). When pancreas extirpation 
 has not been followed by diabetes, it appears probable, however, 
 that a small portion of the gland has been accidentally left 
 behind, and that this has prevented the appearance of sugar 
 in the urine. 
 
 The statement of Minkowski and v. Mering, that complete 
 removal of the pancreas is invariably followed by diabetes 
 mellitus, has since been confirmed by Hedon, Lepine, Lancereaux, 
 Thiroloix, Gley, Vaughan Harley, and others. 
 
 In 1893, Minkowski published a long article, giving a very 
 
EFFECTS OF PANCREAS EXTIRPATION. 75 
 
 full account of his brilliant and numerous experiments on 
 pancreatic diabetes ; from this and from his previous articles 
 most of the following details are taken ( 6 ). 
 
 (a) Effects of 'pancreas extirpation in various animals. — In 
 dogs and cats, diabetes mellitus of a most severe form was 
 invariably produced by total removal of the pancreas. In 
 rabbits, owing to the anatomical position of the gland, total 
 extirpation was scarcely possible. In pigs, total extirpation also 
 produced diabetes. Minkowski failed to produce diabetes in 
 frogs by this experiment, but Aldehoff' has been successful. 
 
 (b) Sugar excretion after total extirpation. — In dogs, as a 
 rule, the day after the operation the urine contains only 1 per 
 cent, of sugar, on the second day 4 to 6 per cent., and on the 
 third day 8 to 10 per cent., which is the highest sugar excretion. 
 If no food be given, the sugar excretion gradually diminishes, 
 but does not disappear, even after seven days' starvation. If 
 the animals be placed on a free diet, the sugar excretions 
 remain very high, and show great variations, according to the 
 nature of the diet. The sugar excretion has a definite relation 
 to the urea excretion, on the average 2 '8 to 1, when carbo- 
 hydrates are excluded from the diet. 
 
 The result of pancreas extirpation in dogs was to produce 
 not a temporary glycosuria, but a genuine permanent diabetes, 
 corresponding to the most severe form of the disease in man. 
 The sugar excreted in the urine was grape sugar. The dogs 
 were abnormally voracious, and suffered from great thirst. If 
 allowed to drink as much water as they desired, polyuria 
 occurred. In spite of the large amount of food which they 
 took, they soon became thin and feeble. The amount of sugar 
 in the blood was considerably increased. 
 
 In course of time, when the dogs became emaciated, the 
 quantity of sugar in the urine diminished. Shortly before death 
 it became very small, and occasionally disappeared completely, 
 just as sometimes occurs in cases of diabetes in man. 
 
 (c) Partial extirpation. — Diabetes does not occur after partial 
 extirpation of the pancreas. Minkowski found that when one- 
 quarter or one-fifth of the gland was left behind, diabetes did 
 not develop. Even when the pancreatic duct was ligatured, and 
 the remaining piece of the pancreas had no connection with the 
 duodenum, still it was sufficient to prevent glycosuria. In these 
 cases, after large quantities of carbohydrates (500 to 1000 grms. 
 
76 EXPERIMENTAL PANCREATIC DIABETES. 
 
 of bread and 100 to 200 grms. of cane sugar), sugar did not 
 appear in the urine, as a rule. 
 
 Vaughan Harley ( 7 ) states that if one-sixteenth part of the 
 gland be left behind, even though it may have no excretory duct 
 opening into the intestine, it is sufficient te prevent glycosuria. 
 Minkowski ( s ), however, points out that it is not possible to say 
 exactly what the size of the remaining pancreas fragment must 
 be to prevent the occurrence of diabetes, since much depends on 
 the condition of the nutrition and blood supply of the fragment. 
 In some cases a slight or temporary glycosuria was produced by 
 partial extirpation, the sugar disappearing when the animal was 
 fed on nitrogenous food only. 
 
 (d) Transplantation of pieces of pancreas under the skin of the 
 abdomen. — Minkowski has shown the relation of the pancreas 
 to diabetes by the most striking experiment of transplanting a 
 piece of the gland. He has succeeded (in dogs) in transplant- 
 ing a portion of the pancreas, and grafting it under the skin of 
 the abdominal wall external to the abdominal cavity. If the 
 transplanted portion of pancreas (or graft) does not necrose, but 
 maintains its nutrition, then diabetes will not occur when the 
 whole of the remaining intra-abdominal part of the gland is 
 removed. But if the transplanted portion of pancreas — i.e. the 
 graft under the skin of the abdomen — be subsequently removed, 
 then diabetes occurs. Hence a portion of pancreas grafted under 
 the skin of the abdomen is sufficient to prevent the occurrence 
 of diabetes when the whole of the gland is removed from the 
 interior of the abdomen, but, on the removal of this graft, 
 diabetes in its most severe form follows. The disease is thus 
 finally produced by an extraperitoneal operation of very short 
 . duration — a few minutes only — in which all secondary influences 
 can be excluded. Minkowski's results with respect to these graft 
 experiments have been confirmed by Heclon, Thiroloix, Gley, and 
 Lancereaux. 
 
 In Hedon's experiments ( 9 ), the descending portion of the 
 pancreas of the dog was grafted under the skin of the abdomen, 
 care being taken not to destroy a vascular connection (in order 
 that the nutrition of the grafted piece might be maintained), until 
 adhesions had formed between the pancreas graft and the sub- 
 cutaneous tissue. The grafted piece of gland communicated, 
 therefore, with the abdominal cavity by slender vessels only. 
 These were ligatured at a later date, without destroying the 
 
EFFECTS OF VARIOUS ARTICLES OF FOOD. 77 
 
 vitality of the graft. In dogs having pancreas grafts, as just 
 described, removal of the whole of the pancreas remaining in the 
 abdomen did not produce glycosuria. But when the extra- 
 abdominal pancreatic graft was subsequently removed, very 
 intense glycosuria developed, and persisted up to the death of 
 the animal. 
 
 (e) The effects of various articles of food, etc., on sugar excretion 
 in pancreatic diabetes. — In the pancreatic diabetes of dogs the 
 amount of sugar in the blood is considerably increased, and the 
 glycogen of the organs disappears early. The whole of the 
 grape sugar given in the food is excreted in the urine. Min- 
 kowski thinks this would not occur if another organ besides the 
 pancreas could bring about the destruction of sugar. 
 
 After pancreas extirpation, ordinary starch passes away 
 unchanged in the fasces. Soluble starch and dextrin -produce 
 an increased excretion of grape sugar ; feeding with bread also 
 increases the sugar in the urine. After feeding with maltose, 
 the sugar excretion is increased, but grape sugar only appears 
 in the urine With regard to the lasvorotatory carbohydrates, 
 Minkowski has shown (1) that for the most part they are burnt 
 up in the organism ; (2) in part they are changed into grape 
 sugar, and excreted as such; (3) after the administration of 
 large quantities of lajvulose, a portion passes off unchanged in 
 the urine ; but when the animal is fed with inulin, lsevulose is 
 not found in the urine. After feeding with cane sugar, neither 
 this body nor lasvulose are found in the urine, but the grape 
 sugar is considerably increased. When lactose is given, the 
 sugar in the urine is increased, but only grape sugar is excreted. 
 Subcutaneous injection of freshly prepared pancreas extract, and 
 also feeding with fresh pancreas, has no influence on the sugar 
 excretion (Minkowski). The use of jambul gave negative 
 results. 
 
 (/) G-lycogen in the organism in pancreatic diabetes. — v. Mering 
 and Minkowski have shown that after pancreas extirpation the 
 glycogen rapidly diminished in the liver, until only traces of it 
 were left, and yet in these cases the amount of sugar in the 
 blood was abnormally high. But in animals in which a portion 
 of the pancreas was left in the abdomen, and in which, therefore, 
 the intensity of the diabetes was slight, a considerable amount 
 of glycogen was found in the liver. 
 
 After extirpation of the pancreas, Abelmann has shown that 
 
7 8 EXPERIMENTAL PANCREATIC DIABETES. 
 
 absorption of fat from the intestines ceased entirely, and only 
 about half of the albuminous material was absorbed. 
 
 Acetone, aceto-acetic acid, and oxy butyric acid, were some- 
 times met with in the urine, but they were not always present. 
 The glycogen in pus cells, and in the white blood corpuscles in 
 cases of pancreatic diabetes, was found to be increased. Compli- 
 cations, such as peritonitis, necrosis of the duodenum, abscesses, 
 etc., prevented the excretion of sugar, just as sometimes occurs 
 in diabetes in man. Possibly by the influence of pathogenic 
 micro-organisms the sugar of the blood is destroyed. 
 
 (g) The cause of diabetes after complete removal of the pancreas. 
 — (1) In the first place, diabetes following extirpation of the 
 pancreas is not due to absence of pancreatic juice in the intes- 
 tine, nor to the diminished elimination of certain substances 
 formed in the pancreas. As above mentioned, the older experi- 
 menters failed to produce diabetes by ligature and obstruction of 
 the pancreatic duct. More recently, Heclon has obtained the 
 same negative results. Then, as already mentioned, Minkowski 
 and v. Mering have shown that if a small portion of pancreas be 
 left behind in the abdomen, no diabetes follows, even when the 
 pancreatic duct is ligatured, and when the portion of pancreas 
 remaining has no connections with the duodenum. When a 
 small portion of the tail of the gland is left behind, diabetes 
 does not occur. Further, the fact that diabetes can be prevented 
 by a pancreas graft under the skin of the abdominal wall, when 
 the whole of the gland is removed from within the abdomen, is 
 conclusive evidence that it is not the absence of pancreatic juice 
 from the intestinal canal which is accountable for the develop- 
 ment of the disease. (2) Neither is diabetes after pancreas extir- 
 pation the result of lesion of the solar plexus of the abdominal 
 sympathetic nerves. As Minkowski and v. Mering point out, the 
 solar plexus would be injured as much by partial extirpation 
 (when only a fragment of the gland is left behind) as by total 
 extirpation, and yet in the former case diabetes does not occur. 
 Then, again, the fact that an extra-abdominal pancreas graft can 
 prevent the occurrence of diabetes, when the whole of the gland 
 is removed from the abdominal cavity, shows that the injury of 
 the solar plexus is not responsible for diabetes after pancreas 
 extirpation. Further, Peiper ( 10 ) and Lustig ( n ), by removal of 
 the coeliac plexus, were not able to produce true diabetes (see p. 
 78). (3) Neither is pancreatic diabetes due to an increased 
 
THE CAUSE OF PANCREATIC DIABETES. 79 
 
 flow of arterial blood to the liver, owing to the ligature of the 
 vessels supplying the pancreas on removal of the gland. Though 
 Arthaud and Butte ( 12 ) have produced glycosuria by ligature of all 
 the branches of the cceliac axis, with the exception of the hepatic 
 artery, their results have not been confirmed by others ; and the 
 effects of partial extirpation of the pancreas and of subcutaneous 
 pancreas grafting, as above described, are opposed to this view. 
 One would expect the supply of arterial blood to the liver to be 
 increased practically to the same extent in partial pancreas 
 extirpation (when only a fragment is left behind) as in total 
 extirpation, yet diabetes occurs in the latter case, but not in the 
 former. Again, the arterial supply to the liver would be the 
 same in an animal from which the pancreas had been totally 
 removed, as in the case of an animal in which all the gland had 
 been removed from the abdomen, but a pancreatic graft left 
 under the skin of the abdominal wall. Yet diabetes occurs in 
 the former case, but not in the latter. (4) Pancreatic diabetes 
 is not due to a renal disturbance, such as is supposed by 
 some to be the cause of phloridzin diabetes, since removal of 
 the kidneys produces an increase of the sugar in the blood, 
 whilst it does not affect the amount of sugar in the blood in 
 phloridzin diabetes. (5) Minkowski's remarkable experiments, 
 and especially the fact that a subcutaneous graft of pancreas 
 will prevent pancreatic diabetes from occurring when the whole 
 of the gland is removed from the abdominal cavity, all point 
 to a special or specific function of the pancreas in preventing 
 diabetes. A " something " is formed in the pancreas which 
 passes into the circulation, and brings about sugar destruction, 
 or prevents the accumulation of sugar in the blood (Minkowski). 
 
 By many this " something " is regarded as an internal secretion. 
 Lepine believes that in the normal condition the pancreas forms 
 a sugar destroying ferment — the glycolytic ferment — which is 
 absorbed by the pancreatic lymphatics and veins, and, passing 
 into the general circulation, prevents the accumulation of sugar 
 in the blood. According to Lepine, the absence of this ferment 
 from the blood is the cause of diabetes when the pancreas is 
 removed completely. 
 
 Lepine and Barral ( 13 ) have shown that in blood with- 
 drawn from the body, and kept for some time at the body 
 temperature, the sugar pretty rapidly disappears. This loss 
 of sugar they believe to be due to the presence of a ferment 
 
So EXPERIMENTAL PANCREATIC DIABETES. 
 
 in the blood — the glycolytic ferment, and they regard the pan- 
 creas as one source of this ferment. The loss of sugar during 
 a given time, and under certain conditions, is taken as the 
 indication of the glycolytic power of the blood, and they have 
 shown that in the pancreatic diabetes of dogs this glycolytic 
 power is diminished. The loss of sugar is less, in a certain 
 time and under certain conditions, in the blood taken from dogs 
 suffering from pancreatic diabetes than in the case of healthy 
 dogs ; it is also less in the blood of diabetic patients than in 
 healthy persons. 
 
 According to Lepine ( u ), the pancreatic cells perform their 
 functions in two directions. From the inner portions of the 
 cell, pancreatic juice is poured into the branches of the excretory 
 duct. From the outer portion of the cell, the base, or the part 
 in relation to the vessels, the glycolytic ferment is poured out 
 into the venous blood and lymphatics, i.e. an internal secretion 
 occurs. According to Lepine, the pancreas is not the only source 
 of the glycolytic ferment, since the blood possesses a certain 
 glycolytic power after the removal of this gland. 
 
 The experiments above mentioned, in which a graft of 
 pancreatic tissue in the abdominal wall prevented the occurrence 
 of diabetes, when the whole of the pancreas was removed from 
 within the abdomen, certainly seem to indicate that some sub- 
 stance or " internal secretion " passes from the gland tissue into 
 the circulation, and in some way prevents the occurrence of 
 diabetes. But Lupine's views respecting the so-called glycolytic 
 ferment have been much criticised. 
 
 Arthus ( 15 ) has made a number of experiments on the 
 subject, and concludes that the glycolytic ferment does not exist 
 in the circulating blood, but is formed outside the body, and 
 that the glycolysis or sugar destruction observed is a cadaveric 
 phenomenon as in the coagulation of the blood. 
 
 Kraus ( 16 ) has shown that the so-called glycolytic power of 
 blood, obtained by venesection, varies in different individuals, but 
 the same variations are found in diabetic patients. Seegen ( 17 ) 
 thinks that the disappearance of sugar from the blood removed 
 from the body is a post-mortem change. He found that if 
 chloroform be added to the blood, and its protoplasm thereby 
 killed, the same disappearance of sugar occurs, and that sugar 
 added to the blood undergoing putrefactive changes gradually 
 disappears. He also found that if blood be heated to a tern- 
 
THE CAUSE OF PANCREATIC DIABETES. 81 
 
 perature sufficient to destroy any enzymes present, and if sugar 
 then be added, in time it disappears totally or partially ; and 
 he therefore concludes that there is no definite evidence of a 
 glycolytic ferment in the blood. Kaufmann ( 18 ) and Chauveau 
 have made a series of analyses of the arterial and venous blood 
 in healthy persons and in cases of diabetes, but they have never 
 found any difference between the normal sugar destruction and 
 that in diabetes. 
 
 Minkowski ( 19 ) has also recorded an experiment in which 
 pancreatic diabetes of a severe form was produced in a dog, and 
 yet the blood lost a greater percentage of sugar in one hour, 
 after withdrawal from the body, than the highest limit observed 
 iu the case of normal blood, i.e. the so-called glycolytic power 
 was not at all diminished. Minkowski thinks that it is not 
 possible at present to give a satisfactory explanation of the 
 exact manner in which extirpation of the pancreas produces 
 diabetes. Nevertheless the graft experiments so often referred 
 to show that the pancreas has a specific function, even if only 
 a small fragment of the gland be present, in preventing the 
 occurrence of diabetes. 
 
 Baldi ( 20 ) has performed experiments on dogs, in order to 
 ascertain the changes produced in the blood during its passage 
 through the pancreas. Blood, to which grape sugar had been 
 added, was passed through the vessels of the gland in situ. The 
 amount of sugar was found to be less in the blood of the 
 pancreatic veins than in the blood sent to the pancreas, also the 
 blood of the pancreatic veins contained a little less sugar than 
 the carotid blood. Baldi concludes that the living pancreas 
 parenchyma, and not a ferment, causes the destruction of sugar 
 in the blood. 
 
 Marcus ( 21 ) has shown that in frogs extirpation of the 
 liver checks the sugar excretion after removal of the pan- 
 creas. By extirpating the liver, the source of the formation 
 of sugar is removed, and hence diabetes cannot occur in these 
 cases. 
 
 Ht'don ( 22 ) found that puncture of the floor of the fourth 
 ventricle in animals rendered diabetic by extirpation of the pan- 
 creas, considerably increased the glycosuria — to the extent of 
 ■jO to 40 per cent., sometimes more. Hence puncture of the 
 floor of the fourth ventricle does not produce diabetes by its 
 action on the pancreas. 
 6 
 
82 EXPERIMENTAL PANCREATIC DIABETES. 
 
 Thiroloix ( 23 ) has produced slow atrophy and destruction of 
 the pancreas in animals by injecting foreign substances into the 
 pancreatic excretory duct. The results were as follows: — (1) 
 The animals wasted ; (2) then they recovered weight ; (3) 
 when the pancreas had become markedly atrophied and sclerosed, 
 glycosuria occurred, if the diet consisted of amylaceous and 
 saccharine material, but ceased when the diet was nitrogenous ; 
 (4) when the atrophy of the pancreas had become so great that 
 the gland was practically destroyed, diabetes developed, and 
 sugar was then present in the urine, whatever the nature of the 
 food. From the results of a number of experiments, Thiroloix 
 concludes that in the normal condition the pancreas produces a 
 substance which is carried by the portal vein to the liver, and 
 the action of this substance on the liver cells transforms sugar 
 into glycogen, and fixes it in the cells. When the pancreas is 
 completely destroyed, the liver produces sugar largely, and 
 glycosuria follows. 
 
 ill) Pancreatic diabetes and the nervous system. — Kaufmann 
 ( 2i - 25 > and 26 ), during the last four years, has performed a large 
 number of experiments on animals with respect to the relation 
 between the pancreas, the nervous system, and the liver in the 
 control of sugar formation. From these experiments he con- 
 cludes that the pancreas, by its internal secretion, plays the 
 part of inhibitor of sugar formation in the liver, indirectly by 
 means of the central nervous mechanism of the latter, and 
 directly by its influence on the liver cells. 
 
 Section of all the nerves to the liver, in an animal having 
 the pancreas still active, is followed either by hypoglycemia, or 
 the amount of sugar in the blood remains normal. But removal 
 of the pancreas in a dog in which all the nerves to the liver have 
 been divided, is followed by hyperglycemia. Hence the internal 
 secretion of the pancreas must have a direct action on the liver 
 by means of the blood — an action hindering sugar formation. 
 Further, Kaufmann has shown that in animals which have no 
 excess of sugar in the blood, if the nerves of the liver and 
 pancreas be completely destroyed, puncture of the fourth ventricle 
 does not produce its usual effect of hyperglycemia and glycosuria. 
 But if hyperglycemia and glycosuria be present, the nerves of the 
 liver and pancreas being destroyed, or the nerves and the liver 
 being destroyed and the pancreas removed, puncture of the 
 fourth ventricle always greatly increases the hyperglycemia and 
 
DIABETES AND THE NERVOUS SYSTEM. 83 
 
 glycosuria. Hence Kaufmann concludes that puncture of the 
 fourth ventricle acts not only on the liver and pancreas, but 
 also on the general histolytic work of the various tissues of the 
 organism ; and that the histolysis is regulated botli by the 
 pancreatic products thrown into the blood, and also by the 
 nervous system. The internal pancreatic secretion has an in- 
 hibitory action, by means of the blood, on the tissue changes, 
 and on sugar formation in the liver. 
 
 From Kaufmann's experiments, it appears, therefore, that the 
 internal secretion of the pancreas has an inhibitory action on 
 sugar formation — 
 
 (1) Indirectly, through the central nervous system and 
 hepatic nerves. 
 
 (2) Directly, through the blood, which carries the internal 
 secretion of the pancreas to the liver cells. 
 
 (3) By its influence on the tissues of the body. 
 
 It has been shown that the pancreatic internal secretion is 
 also under the control of the central nervous system. 
 
 In addition to the above-described experimental proceedings, 
 glycosuria may be produced in animals in numerous other ways ; 
 but it has only been possible to refer to those which are of chief 
 interest to the practitioner of medicine. 
 
 REFERENCES. 
 
 1. Cawley, Thomas . Land. Med. Journ., 1788, p. 286. 
 
 2. Baumel .... Montpel. med., January and May 1882; 
 
 Abstr., Jahresb. il. d. Leistung. ... d. 
 ges. Med., Berlin, 1882, Bd. ii. S. 223. 
 
 3. Minkowski und von Arch. f. exper. Path. u. Pharmakol., Leipzig, 
 
 Merino 1889, Bd. xxvi. 
 
 4. de Dominicis . . Giorn. internaz. d. sc. med., Napoli, 1889, 
 
 p. 801 ; Centralbl. f. Tdin. Med., Bonn, 
 June 7, 1890; Munchen. med. Wchnsclir., 
 13th and 20th October 1891. 
 
 5. Reale und de Renzi Wien.med. Wchnschr., 15th August 1891. 
 
 6. Minkowski . . . Arch. f. exper. Path. u. Pharmakol., Leipzig, 
 
 1893, Bd. xxxi. S. 85-189. 
 
 7. Harley, Vaughan . Med. Chron., Manchester, August 1895, 
 
 p. 323. 
 
 8. Minkowski . . . Loc. cit. ; also Centralbl. f. klin. Med., Bonn, 
 
 1st February 1890. 
 
 9. Hedon .... Gaz. mkl. de Pari*, 13th August 1892. 
 
8 4 
 
 EXPERIMENTAL PANCREATIC DIABETES. 
 
 10. Peiper 
 
 11. Lustig, A. 
 
 1 2. Arthaud et Butte 
 
 13. Lepixe et Barral . 
 
 14. Lepixe . . . . 
 
 15. Arthus . . . . 
 
 16. Kraus . . . . 
 
 17. Seegex, J. . . . 
 
 18. Kaufmanx . . . 
 
 19. Mixkowski . 
 
 20. Baldi 
 
 21. Marcus . . . . 
 
 22. Hedox, E. . . . 
 
 23. Thiroloix . . . 
 
 24. Chauveau et Kauf- 
 
 maxx 
 
 25. Kaufmaxx . . . 
 
 26. Do. . . . 
 
 Miinchen. med. Wchnschr., Bonn, 29th April 
 
 1890 (Report of Congress f. innere Medicin, 
 
 Wien). 
 Arch, per le sc. med., Torino, 1889, No. 6 ; 
 
 Centralbl, f. Min. Med., Bonn, 10th May 
 
 1890. 
 Semaine med., Paris, 5th February 1890 ; 
 
 Abstr., Annual Univ. Med. Sciences, 1891 
 Gaz. med. de Paris, 31st Januarj'' 1891. 
 Berl. Min. Wchnschr., 11th May 1891. 
 Arch, de physiol. norm, et path., Paris, July 
 
 1891. 
 Ztschr. f. Min. Med., Berlin, Bd. xxi. 
 " Der Diabetes Mellitus," Berlin, 1893, S. 85. 
 Semaine med., Paris, 16th January 1895. 
 Arch. f. exper. Path. u. Pharmacol., Leipzig, 
 
 1893, Bd. xxxi. S. 176. 
 Arch, di farm, e terap., Palermo, tome iii. 
 
 fasc. 4, p. 173 ; and abstract, Jahresb. ii. 
 
 d. Leistung. . . . d. ges. Med., Berlin, 1895, 
 
 Bd. xi. S. 42. 
 Ztschr. f. Min. Med., Berlin, Bd. xxvi. S. 225. 
 Arch, de physiol. norm, et path., Paris, 1894, 
 
 Ser. 5, tome ri. p. 269. 
 Gaz. hebd. de med., Paris, 2nd March 1895. 
 Compt. rend. Soc. de bid., Paris, February 
 
 and March 1893. 
 Ibid., 1894, Ser. 10, tome i. p. 254. 
 Ibid., 1895, Ser. 10, tome ii. p. 5. 
 
CHAPTER VII. 
 
 GLYCOSURIA FROM VARIOUS CAUSES. 
 
 Alimentary Glycosuria. 
 
 In health, according to some authors, the urine contains a 
 minute trace of sugar, but it is so small (if present) that it is not 
 detected by the usual clinical method of testing, and for practical 
 purposes the urine may be considered to be free from sugar. It 
 has been demonstrated, however, by a number of observers that 
 when a very large quantity of saccharine food is taken, a small 
 amount of sugar is found in the urine even in health. 
 
 Moritz (*) found sugar in the urine in five out of eleven 
 healthy persons, after a meal very rich in saccharine materials ; 
 two to four hours after food 0*1 to 0'25 per cent, of sugar was 
 present : but the glycosuria soon disappeared. 
 
 The sugar found in the urine is said to be the same as that 
 taken in excess — glucose producing glycosuria ; lactose, lactosuria ; 
 laevulose, Levulosuria ; saccharose, saccharosuria. 
 
 According to von JSToorden ( 2 ), the amounts of the various 
 forms of sugar which must be taken before the sugar is found in 
 the urine are as follows : — 
 
 Milk sugar, over 120 grms. 
 Cane sugar, ,, 150 to 200 grms. 
 Fruit sugar, about 200 grms. 
 Grape sugar, ,, 180 to 250 grms. 
 
 These figures represent, therefore, the amounts of the various 
 forms of sugar required to produce alimentary glycosuria in 
 healthy persons. They indicate the average quantities only, and 
 naturally they vary somewhat in different individuals. To 
 produce this alimentary glycosuria the sugar should be given 
 all at one dose, early in the morning, when the stomach is 
 empty, and the urine should be examined every hour for sugar. 
 
86 GLYCOSURIA FROM VARIOUS CAUSES. 
 
 The occurrence of alimentary glycosuria depends not only on the 
 quantity of glucose taken, but also on the rapidity of absorption. 
 For example, in a case examined by Striimpell ( 3 ), 150 grms. of 
 glucose given all at once on an empty stomach caused distinct 
 glycosuria, but when the same quantity was divided into three 
 portions, and taken separately in the course of an hour, the 
 urine remained free from sugar. 
 
 The sugar appears in the urine from three-quarters to one 
 hour after the administration of the glucose, and the excretion 
 thereof continues one to three hours. 
 
 The total amount of the various forms of sugar excreted is as 
 follows : — 
 
 2 - 8 per cent, of the cane sugar taken, 1 per cent, of the grape 
 sugar, 0'8 per cent, of the milk sugar. 
 
 The power of assimilation for starch in healthy persons 
 appears to be unlimited ; probably so much time is required for 
 its absorption and digestion, that over-saturation of the blood 
 with carbohydrates does not occur. Those persons whose urine 
 contains sugar after a diet containing large quantities of starch, 
 have a low and abnormal assimilation limit, and the condition 
 approaches diabetes. 
 
 In healthy persons, then, slight glycosuria can only be 
 produced by very great excess of sugar in the diet, and not by 
 excess of starch ; in mild forms of diabetes, starchy food as well as 
 saccharine food produces glycosuria ; in the most severe forms, 
 glycosuria is present even when the diet consists of nitrogenous 
 materials only. 
 
 By an excess of carbohydrates in the food, the glycogen 
 stored in the liver and muscles is increased, but the storing 
 space is apparently not unlimited. When there is an excess of 
 carbohydrates absorbed day by day, the limit of the storing- 
 capacity of the liver and muscles seems to be reached, and the 
 excess of carbohydrates is transformed into fat, which is deposited 
 in the connective tissue in various parts of the body. When the 
 excess of carbohydrates cannot be disposed of in either of these 
 ways, because the glycogen depot is too limited, or the quantity of 
 carbohydrates too great permanently or temporarily, there is an 
 excess of sugar in the blood, and glycosuria results. 
 
 A number of observations have been made during the last 
 ten years, with reference to power of sugar assimilation in various 
 
ALIMENTARY GLYCOSURIA. 87 
 
 diseased states, in order to ascertain whether^in these conditions 
 glycosuria can be produced with smaller quantities of sugar than 
 in health; but in a few conditions only has this power been 
 found diminished. In most diseases, 150 to 200 grms. of 
 grape sugar can be taken without glycosuria being produced 
 (v. Noorden) ; but larger doses — as in health — produced 
 glycosuria. 
 
 Lanz ( 4 ) has found, however, that the power of assimilation of 
 grape sugar is diminished during pregnancy. To thirty pregnant 
 women he administered 100 grms. each of pure grape sugar, and 
 in nineteen cases he was able to detect sugar in the urine. 
 (As already stated, normally, 180 to 250 grms. of grape sugar 
 can be assimilated.) 
 
 According to Kraus and Ludwig ( 5 ), Chvostek ( 6 ) and 
 v. Noorden, there is a diminished power of sugar assimilation 
 in Graves' disease, i.e. alimentary glycosuria is more easily pro- 
 duced than in health. A similar condition was found by 
 Striimpell in cases of neurasthenia, or traumatic neurosis. In 
 other diseases of the brain, spinal cord, nerves, and muscles, no 
 diminished power of sugar destruction has been detected. 
 
 In some cases of cirrhosis of the liver glycosuria has been 
 produced more readily than in health, but in most cases, as 
 well as in other forms of liver disease, there has been no 
 diminished power of sugar destruction. This is somewhat re- 
 markable, since in many of these diseases the liver tissue is 
 destroyed to a great extent. 
 
 v. Jaksch has shown that in phosphorus poisoning, however, 
 when the symptoms are far advanced, and when the protoplasm 
 of the muscles and liver cells is markedly affected, alimentary 
 glycosuria is produced more readily than in health. 
 
 In diseases of the heart, lungs, and blood, in arterio-sclerosis 
 and marasmus, the power of sugar destruction does not appear to be 
 altered. In many, but not in all, cases of chronic alcoholism, or, 
 more correctly, cases of habitual drinkers of large quantities of 
 beer, Striimpell found that glycosuria could be readily induced 
 by 100, 75, or even 50 grms. of grape sugar. The same 
 condition the author discovered in some persons after they 
 had taken an excessive quantity of beer (1^ to 2 litres) very 
 rapidly. Alimentary glycosuria does not occur in all great 
 beer drinkers, however, and much depends upon individual 
 peculiarity. 
 
88 GLYCOSURIA FROM VARIOUS CAUSES. 
 
 • Puerperal Glycosuria. 
 
 It has long been known that a substance occurs in the 
 urine in the puerperal state, which reduces Fehling's solution 
 like grape sugar. Blot ( 7 ) drew attention to this fact many- 
 years ago. He concluded that a physiological glycosuria existed 
 in women during labour, in women during lactation, and in 
 almost 50 per cent, of women during pregnancy. No diabetic 
 symptoms are present in this physiological glycosuria. After 
 delivery, and on the cessation of lactation, the sugar disappears. 
 The observations of Blot were disputed at first, but have now 
 been abundantly confirmed. Matthews Duncan ( 8 ) stated that 
 in the Edinburgh Maternity he found a slight amount of sugar 
 in the urine of every woman during lactation. 
 
 Ney ( 9 ) examined the urine of 148 women during lactation, 
 and found sugar present in 77 per cent. 
 
 M'Cann and Turner ( 10 ) have investigated one hundred cases, 
 in order to decide whether sugar is always present in the urine 
 of nursing women at some period of lactation. The following 
 are the results of their observations : — 
 
 1. That sugar is present in the urine of women during 
 lactation. 
 
 2. That sugar is present at some period in every case. 
 
 3. That in the majority of cases the largest amount occurs 
 on the fourth and fifth days of the puerperium. 
 
 4. That the quantity depends on (a) the condition of the 
 breasts, (b) the quantity and quality of the milk, and (c) the 
 sucking of the child. In one hundred cases the average quantity 
 found was 0*35 per cent., i.e. 14- grs. per ounce. 
 
 5. That when lactation is diminished or suppressed the 
 amount of sugar diminishes or disappears. 
 
 6. That when the production and exhaustion of the milk 
 are equal, the amount of sugar is very small. 
 
 Kaltenbach ( u ) has demonstrated that the form of sugar in 
 puerperal glycosuria is lactose. He bases this conclusion on the 
 inability of the sugar to ferment directly, on its rotatory power, 
 on the fact that it is transformed directly into fermentable 
 sugar by boiling with sulphuric acid, and on the mucic acid 
 reaction. (The other reactions given by lactose in the urine are 
 pointed out on p. 32.) 
 
 The presence of lactose in the urine appears to be connected 
 
PUERPERAL GLYCOSURIA. 89 
 
 with engorgement of the breast. The lactose in the urine is 
 abundant in women whose children have been born dead or have 
 died during the period of lactation ; it is also abundant when, 
 from mastitis, clefts of the nipples, or other causes, the child 
 cannot be applied to the breast, and engorgement of the gland 
 occurs. 
 
 Sinclair ( 12 ) states that he has never failed to find sugar in 
 the urine, when the breasts have been engorged from any cause 
 during the period of functional activity. He believes that the 
 sugar in the urine is due to absorption from the engorged 
 breasts ; and he has met with a large number of cases in which 
 sugar has appeared in the urine, on the occurrence of still birth, 
 or when the child has been suddenly weaned. (See also p. 58.) 
 
 In cases of abscess of the breast, I have often found that the 
 urine gives a slight reduction of Fehling's solution, but no 
 indication of sugar by the fermentation test ; also filtration 
 through animal charcoal does not remove the reducing body, 
 which in all probability is lactose. 
 
 grlycosukia produced by poisons and chemical 
 Substances. 
 
 Numerous drugs have been stated to be capable of producing 
 glycosuria, but it is probable that very frequently the substance 
 in the urine which has reduced Fehling's solution has not been 
 sugar, but some other reducing body — often glycuronic acid. 
 When the urine reduces Fehling's solution after the adminis- 
 tration of any drug, it is always necessary to try some con- 
 firmatory test before deciding that the reducing substance is 
 sugar. 
 
 The following are the drugs and chemical substances which 
 have been said to sometimes produce glycosuria in man or 
 animals : — Curare, orthonitro-phenylpropiolic acid, carbon mon- 
 oxide (in eleven out of sixteen cases observed by Frerichs ( 13 ) 
 — the sugar disappearing in two to four days), amyl nitrite (by 
 injection in dogs), and methyldelphinin. The following sub- 
 stances are said to occasionally produce glycosuria, when taken 
 or inhaled in very large quantities : — Opium, morphia, chloral 
 hydrate, prussic acid, mineral acids, coal gas (in one out of four 
 cases — Frerichs), poisonous doses of mercury, arsenic, and phos- 
 phorus. Strychnia poisoning is stated to have produced 
 
9 o GLYCOSURIA FROM VARIOUS CAUSES. 
 
 glycosuria in frogs, but Frerichs has failed to detect sugar in 
 the urine in a well-marked case of strychnia poisoning in 
 man. 
 
 According to Jacoby ( u ), caffeine, theobromin, benzoate of 
 caffeine and soda, diuretin, and caffeine sulphonic acid, give rise to 
 glycosuria with diuresis in rabbits, when the animals are fed 
 on roots, carrots, and turnips. Ether and chloroform narcosis 
 have been stated to occasionally give rise to slight glycosuria, 
 but in cases examined by Frerichs no sugar was detected. I 
 have examined the urine of a number of surgical patients, after 
 chloroform narcosis, without ever being able to obtain even 
 the slightest reduction of Fehling's solution. 
 
 I have recently examined the urine in three cases of opium 
 poisoning, with the following results. In a case which terminated 
 fatally in less than twelve hours, I was unable to obtain any 
 evidence of sugar in the urine drawn off from the bladder 
 immediately after death. In another case the urine reduced 
 Fehling's solution slightly, gave the characteristic yellow crystals 
 when the phenylhydrazin test for sugar was applied, and also 
 a distinct reaction for sugar with the fermentation test. In 
 a third case of opium poisoning, the urine gave no reaction for 
 sugar when the fermentation test was employed. 
 
 It was pointed out by Leconte ( 15 ), in 1851, that the 
 prolonged administration of small doses of uranium salts caused 
 glycosuria in animals. Chittenden, Lambert, Woroschilsky, and 
 others, have published the results of observations on the physio- 
 logical action of these salts, and the subject has been recently 
 reviewed by Cartier ( 16 ). In large doses, nitrate of uranium 
 is an irritant poison, but small doses act as a diuretic. The 
 specific gravity of the urine is increased, and albuminuria and 
 glycosuria are produced : other symptoms are loss of appetite, 
 thirst, general torpor, paresis of the posterior limbs, paralysis, 
 and emaciation. Finally, the animals become comatose, and 
 death occurs. Cartier points out that the chief changes found 
 pathologically are in the liver and kidneys. Small doses cause 
 simply hepatic congestion. After larger doses (in rabbits), 
 hyaline masses have been found in the interior of the liver 
 cells, and the nuclei have been found pushed to the side of 
 the cell ; at some parts the cells are necrotic. The hyaline 
 masses do not contain glycogen. The changes in the kidney 
 have varied from congestion up to necrosis. The chief lesion 
 
GL YCOSURIA PR OD UCED B Y POISONS. 9 1 
 
 has been found in the convoluted tubes where necrosis of the 
 epithelium has been observed. 
 
 Pavy ( 17 ) has shown that phosphoric acid, injected into the 
 jugular vein of animals, produces glycosuria. 
 
 It is stated that the intravenous injections of large quantities 
 of a 1 per cent, sodium chloride solution, or of solutions of 
 carbonate, sulphate, acetate, and succinate of soda, have produced 
 glycosuria in animals. 
 
 Claude Bernard's diabetic puncture of the fourth ventricle is 
 said to be unsuccessful in producing glycosuria, if sodium car- 
 bonate be previously injected into the blood, and the glycosuria 
 which is produced by lesions of the sympathetic nerves can also 
 be arrested by the injection of sodium carbonate. 
 
 In animals suffering from phosphorus or arsenical poisoning, 
 Bernard's puncture is said to have only very slight effect. 
 
 Harley ( 18 ) produced glycosuria by injecting ether, chloro- 
 form, ammonia, alcohol, and methylated spirits into the portal 
 vein. 
 
 Clinical Glycosuria. 
 
 In a number of varied diseased conditions, a small 
 quantity of sugar is occasionally met with in the urine, 
 though polyuria or other diabetic symptoms are absent. But 
 it is well to remember, in enumerating these conditions, that 
 in all of them glycosuria is much more frequently absent than 
 present, and that when detected it is temporary, and continues 
 usually for a few hours or days only. 
 
 Glycosuria is sometimes present in cases of train lesions, and 
 injury to the head is an occasional cause. It will be pointed out 
 in the section on etiology that true diabetes sometimes follows, 
 and is apparently excited by, a blow on the head. These cases 
 are very rare, but cases of temporary glycosuria after an injury 
 to the head are not very infrequent. 
 
 Higgin and Ogden ( 19 ) have carefully examined the urine in 
 212 cases of injury to the head, and have found sugar present in 
 twenty. In cases of scalp wounds and minor head injuries, 
 glycosuria was met with in 5*95 per cent. ; in cases of scalp 
 wounds with exposure of bone, glycosuria was present in 9 3 
 per cent. ; in cases of concussion, in 2 "5 per cent. ; in fractures 
 of the vault of the skull, in 20 '8 per cent. ; in fracture of the 
 base, in 23*8 per cent. 
 
92 GLYCOSURIA FROM VARIOUS CAUSES. 
 
 From the examination of these 212 cases, the authors draw 
 the following conclusions : — 
 
 1. That sugar may appear in the urine as early as six hours 
 after a head injury, and disappear within twenty-four ; the 
 average time for its appearance being eight to twelve hours ; the 
 average time for the disappearance being from the fifth to the 
 ninth day. 
 
 2. That a small proportion of cases exhibit a permanent 
 glycosuria from the date of the injury to the head. 
 
 3. That acetone and diacetic acid are rarely, if ever, found 
 in such cases, excepting where the condition becomes a permanent 
 glycosuria, and even then probably only after a number of months 
 or years. 
 
 Brain lesion, such as cerebral haemorrhage, meningitis, 
 cerebral aneurysm, brain tumours, tumours of the pituitary body, 
 as well as neuralgia, sciatica, and Graves' disease, have all been 
 occasionally associated with glycosuria. Occasionally, though 
 very rarely, glycosuria has been found in cases of tetanus, 
 paralysis agitans, epilepsy, and atheroma of arteries about the floor 
 of the fourth ventricle ; but the glycosuria is usually temporary. 
 A few cases of locomotor ataxia, associated with glycosuria, or 
 much more rarely with true diabetes, have also been recorded. 
 
 Glycosuria is occasionally associated with mental diseases. 
 Bond ( 20 ) has carefully examined the urine of 355 cases of mental 
 disease, but has only found sugar present in 5 '3 5 per cent. 
 Only two suffered from true diabetes ; the rest were cases of 
 simple glycosuria. In the majority of the cases in which 
 glycosuria was present, the patient suffered from melancholia. 
 Maniacal patients, epileptics, and congenital cases were appa- 
 rently exempt. 
 
 Naunyn ( 21 ) has drawn attention to the occasional association 
 of glycosuria, or mild forms of diabetes, with general paralysis of 
 the insane. 
 
 Glycosuria has been met with, though very rarely, after the 
 following febrile affections : — Typhoid, scarlet fever, malaria, 
 diphtheria, cholera, anthrax. It has been observed occasionally, 
 though very rarely, in gallstone colic ; and several instances of 
 glycosuria associated with appendicitis have been recorded. 
 
 In most of the above-mentioned cases the glycosuria is 
 temporary, and persists only for a few hours or days ; occasionally 
 it is permanent, and all the symptoms of diabetes may develop. 
 
CLINICAL GLYCOSURIA. 93 
 
 In other cases the glycosuria disappears for a time ; but may 
 afterwards reappear, and occasionally true diabetes may ultimately 
 develop. Loeb has recently recorded cases in which this has 
 occurred (see p. 166). 
 
 v. Noorden found that in four out of fourteen obese 
 individuals, temporary glycosuria was produced by the adminis- 
 tration of only 100 grms. of grape-sugar, i.e. there was a 
 diminished power of sugar destruction in the system. Of these 
 four persons, two became diabetic some years later ; in the other 
 two cases, the lapse of time is still too short to allow any very 
 definite opinion to be given with regard to the probability of 
 diabetes developing. 
 
 The association of glycosuria with the above-mentioned 
 affections is usually exceedingly rare. Whilst a slight glyco- 
 suria is not very uncommon in the ailments of stout and wealthy 
 elderly persons seen in private practice, it is rare amongst the 
 poor and badly-nourished patients of large hospitals. 
 
 In addition to the diseases mentioned, I have recently met 
 with the following affections associated with glycosuria : — 
 
 Cerebral haemorrhage and fractured skull (recorded on p. 294). 
 
 Two cases in which there was marked double optic neuritis, 
 with severe headache. Other indications of brain lesion were 
 absent ; the sugar disappeared from the urine in the course of 
 some days, and gradual improvement took place. 
 
 One case of chronic lead-poisoning, with double wrist-drop. 
 
 A case of ataxia, of sudden onset, accompanied by diplopia 
 and paresis of both external recti. The optic discs were normal, 
 no other nervous symptoms were present, and in the course of a 
 few weeks the glycosuria and nervous symptoms disappeared. 
 
 Two cases of hemiplegia. 
 
 One case of optic atrophy. 
 
 One case of chronic parenchymatous nephritis, with enlarged 
 liver. 
 
 A case of fracture of the tibia. On admission to the 
 Hospital, sugar was present in the urine in large quantity, but 
 disappeared in a few days. There were no other indications of 
 diabetes. 
 
 REFERENCES. 
 
 1. Moritz, F. . . . Mwnchen. med. Wchnschr., 1891, Nos. 1 and 2. 
 
 2. v. Noorden . . . "Die Zuckerkrankheit und ihre Behandlimg," 
 
 Berlin, 1895, S. 11-13. 
 
94 
 
 GLYCOSURIA FROM VARIOUS CAUSES. 
 
 3. 
 
 4. 
 
 5. 
 
 6. 
 7. 
 8. 
 9. 
 10. 
 
 11. 
 
 12. 
 13. 
 14. 
 15. 
 16. 
 
 17. 
 
 18. 
 
 19. 
 
 20. 
 21. 
 
 Strumpell, A. . . Berl. Idin. Wclmschr., 1896, No. 46. 
 
 Lanz Centralbl. f. innere Med., Leipzig, 1896, 
 
 No. 29 ; Wien. med. Presse, 1895, No. 49. 
 Kraus und Ludwig Wien. Idin. Wclmschr., 1891, Nos. 46 and 48. 
 Chvostek. . . . Ibid., 1892, No. 17. 
 
 Blot Gaz. d. hop., Paris, 1856, No. 121. 
 
 Duncan, M. . . . Trans. Obst. Soc. London, 1882. 
 
 Xey Arch. f. Gynaek., Berlin, Bd. xxxv. S. 239. 
 
 M'Cann and Turner Trans. Obst. Soc. London, 1892, vol. xxxiv. 
 
 p. 473. 
 Kaltenbach . . . Ztschr. f. Geburtsh. u. Gyndk., Stuttgart, 
 
 1879, Bd. iv. S. 161. 
 Sinclair, J. . . . Med. Chron., Manchester, 1885-86, vol. iii. 
 
 p. 276. 
 Frerichs, F. T. . . "Ueber den Diabetes," Berlin, 1884, 
 
 S. 25-33, 38. 
 Jacoby, C. . . . Arch. f. exper. Path.u. Pliarmalwl., Leipzig, 
 
 Bd. xxxv. Hefte 2 and 3. 
 West, S Brit. Med. Journ., London, 24th August 
 
 1895. 
 Cartier, F. . . . " Glycosuries toxiques et en particulier in- 
 toxication par le nitrate d'urane," These, 
 
 Paris, 1891. 
 Pavy, F. \Y. . . Guy's Hosp. Pep., London, 1861. 
 Harley, G. . . . "Diabetes," London, 1866; also Brit, and 
 
 For. Med.-Cliir. Rev., London, July 1857. 
 Higgins, F. A., and Boston Med. and S. Journ., 28th February 
 
 Ogden, J. B. 1895. 
 
 Bond, C. G. . . . Lancet, London, 1896, vol. ii. p. 1606. 
 Naunyn .... Neurol. Centralbl., Leipzig, 1896, S. 606. 
 
CHAPTER VIII. 
 
 ETIOLOGY AND ETIOLOGICAL RELATIONS. 
 
 A. Etiology. 
 
 1. General relations. — Sex. — Diabetes mellitus is a disease 
 which is more common in males than females. In the early 
 period of life, however, the liability of the two sexes is about 
 equal. Afterwards, especially after 30, males are affected more 
 frequently than females. Thus, in 100 cases of diabetes at the 
 Manchester Eoyal Infirmary, the proportion was as follows : — 
 
 Males 
 
 Females 
 Under the age of 30. 
 Males . . 18 cases. 
 Females . . 19 ,, 
 
 62 cases. 
 38 „ 
 
 Over 30. 
 
 44 cases. 
 
 19 „ 
 
 Age. — The following table shows the number of patients at 
 various ages in 100 consecutive cases of diabetes. At the time 
 of admission to the Hospital the majority were suffering from a 
 severe form of the disease. 
 
 One Hundred Cases of Diabetes {chiefly Hospital Patients). 
 
 Age. 
 
 Years. 
 
 Total. 
 
 10-20. 
 
 20-30. 
 
 30-40. 
 
 40-50. 
 
 18 
 3 
 
 50-60. 
 
 60-70. 
 
 Males 
 
 6 
 6 
 
 12 
 13 
 
 14 
 8 
 
 9 
 3 
 
 3 
 
 5 
 
 62 Males. 
 38 Females. 
 
 Females 
 
 
 Total . . 
 
 12 
 
 25 
 
 22 
 
 21 
 
 12 
 
 8 
 
 100 
 
 In private practice the proportion of cases at various ages is 
 somewhat different to that given in the above table. Thus 
 
96 ETIOLOGY AND ETIOLOGICAL RELALLONS. 
 
 Karl Grube ( a ) of Neuenahr — a spa much frequented by diabetic 
 patients — found the highest percentage of cases between the 
 ages of 50 and 60. 
 
 The following table, which I have prepared from the pub- 
 lished records of several authors, shows that there is considerable 
 difference with regard to the age of the patients in various series 
 of cases. Probably the difference is due to the fact that in 
 private consulting practice a large number of cases of diabetes of 
 the milder form, in well-to-do elderly people, are met with, whilst 
 these cases are much more rare amongst the poor and hard- 
 working classes that chiefly supply the diabetic hospital patients ; 
 or, at least, if these mild forms do occur in elderly people 
 amongst the poor, they more rarely come under treatment. 
 
 Percentage of Cases of Diabetes at various Ages. 
 
 
 Under 
 10 yrs. 
 
 10-20. 
 
 20-30. 
 
 30-40. 
 
 40-50. 
 
 50-60. 
 
 00-70. 
 
 70-80. 
 
 Over 
 
 80. 
 
 Grube 
 
 
 17 
 
 2-8 
 
 11-2 
 
 23-1 
 
 39-5 
 
 18-1 
 
 3-4 
 
 
 Seegen 
 
 0-5 
 
 3 
 
 16 
 
 16 
 
 24 
 
 30 
 
 10 
 
 0-5 
 
 
 
 0-58 
 
 4-19 
 
 7-13 
 
 16-4 
 
 24-92 
 
 30-73 
 
 13-37 
 
 2-49 
 
 •07 
 
 
 1 
 
 7 
 
 10 
 
 18 
 
 25 
 
 26 
 
 11 
 
 1 
 
 
 Manchester Royal In- '( 
 firmary . . . . J 
 
 
 12 
 
 25 
 
 22 
 
 21 
 
 12 
 
 8 
 
 
 
 The mortality (per million living), as given in the Eeport of 
 the Eegistrar-General, is greatest between the ages of 65 and 
 75 ; whilst, amongst hospital patients in Manchester, the table 
 given above shows that 68 per cent, are between the ages of 20 
 and 50 when they first come under treatment. 
 
 The table on page 97 is taken from the " Supplement to the 
 Fifty-fifth Annual Eeport of the Eegistrar-General" (Part I., 
 1895, p. cxi.). 
 
 Death from diabetes is very rare in extreme old age. In 
 young persons and children the disease is rare, as shown by the 
 following table. Eecently two patients each aged 12 years, 
 and one patient aged 11, were admitted to the Manchester 
 Eoyal Infirmary, suffering from diabetes mellitus. Bell ( 2 ) has 
 recorded a fatal case of diabetes mellitus in a child set. 3 
 months. 
 
GENERAL RELATIONS— AGE. 
 
 97 
 
 Mortality from Diabetes Mellitus %)er million living. 
 
 
 
 Years. 
 
 < SP 
 < 
 
 0- 
 
 5- 
 
 10- 
 
 15- 
 
 20- 
 
 25- 
 
 35- 
 
 45- 
 
 55- 
 
 65- 
 
 i — 
 
 1861-70"! 
 
 1871-80 ,l Persons. . . . 
 1881-90 J 
 
 30 
 
 3 
 
 5 
 
 9 
 
 17 
 
 19 
 
 32 
 
 42 
 
 59 
 
 95 
 
 117 
 
 74 
 
 38 
 
 1 
 
 4 
 
 10 
 
 19 
 
 25 
 
 37 
 
 51 
 
 72 
 
 132 
 
 171 
 
 113 , 
 
 57 
 
 4 
 
 7 
 
 15 
 
 24 
 
 30 
 
 46 
 
 64 
 
 107 
 
 217 
 
 293 
 
 238 
 
 1861 -70 ^i 
 
 41 
 
 3 
 
 5 
 
 10 
 
 22 
 
 24 
 
 44 
 
 55 
 
 83 
 
 136 
 
 180 
 
 120 
 
 1871-80 \ Males .... 
 1881-90 J 
 
 50 
 
 1 
 
 4 
 
 10 
 
 24 
 
 34 
 
 48 
 
 68 
 
 96 
 
 181 
 
 247 
 
 172 
 
 69 
 
 5 
 
 7 
 
 14 
 
 26 
 
 35 
 
 58 
 
 78 
 
 134 
 
 282 
 
 397 
 
 314 
 
 1861-70 \ 
 
 20 
 
 2 
 
 4 
 
 8 
 
 13 
 
 14 
 
 22 
 
 30 
 
 37 
 
 58 
 
 62 
 
 38 
 
 1871-80 I Females . . . 
 
 27 
 
 2 
 
 4 
 
 10 
 
 14 
 
 18 
 
 27 
 
 35 
 
 50 
 
 87 
 
 106 
 
 67 
 
 1881-90 J 
 
 45 
 
 3 
 
 6 
 
 15 
 
 22 
 
 26 
 
 35 
 
 51 
 
 82 
 
 161 
 
 206 
 
 180 
 
 Wegeli ( 3 ) gives the following table with reference to the 
 ages of 1 2 cases of diabetes mellitus in children : — 
 
 Under 1 year 
 1-5 years 
 5-10 „ 
 10-16 „ 
 
 3 
 
 26 
 31 
 
 42 
 
 102 
 
 In children and young people the disease often runs a very 
 rapid course though occasionally there are exceptions to this 
 general rule. In elderly persons the disease is frequently 
 associated with gout and obesity ; it is then generally of a mild 
 form, and often runs a very chronic course. 
 
 2. Geographical distribution, racial influence, and rarity of the 
 disease. — In most countries diabetes is a rare disease. A 
 medical man engaged in a busy general private practice in 
 England only occasionally meets with a well-marked case. In 
 hospital practice more cases are collected, and here the disease 
 can be best studied. At the Manchester Eoyal Infirmary, during 
 the twenty years 1875-1895, the number of medical in-patients 
 was 27,721, and of these only 272 suffered from diabetes, not 
 quite 1 per cent. ( — 0'9 per cent.). 
 
 The death-rate from diabetes in England, per million living, 
 in the ten years 1881-1890 was 57. In Manchester the 
 census of 1891 gave the population as 505,368; in Salford, 
 7 
 
9 8 
 
 ETIOLOGY AND ETIOLOGICAL RELATIONS. 
 
 198,717. The number of deaths from diabetes in the same 
 year was 29 in Manchester, and 8 in Salford — total, 37. 
 
 Osier ( 4 ) states that the disease is more rare in America 
 (United States) than in Europe. The last census for the 
 United States gave a mortality of 2 "8 per 100,000, whilst in 
 European countries the mortality is from 5 to 9 per 100,000. 
 Among 35,000 patients treated at the Johns Hopkins Hospital, 
 Baltimore, there were only ten cases of diabetes mellitus. 
 
 The following table, prepared by Ebstein ( 5 ), shows the pro- 
 portion of cases of diabetes treated in various University clinics 
 in Germany : — 
 
 ■ Year. 
 
 
 c 
 c 
 o 
 pq 
 
 3 
 0) 
 
 H 
 
 5 
 
 to 
 
 c 
 
 3 
 
 7 
 
 10 
 7 
 
 "c3 
 1 
 
 '53 
 
 5 
 
 1 
 
 1 
 
 "3 
 
 5 
 5 
 8 
 
 2 
 2 
 
 S? 
 
 .a 
 
 2c 
 
 '3 
 
 :o 
 
 4 
 5 
 6 
 
 si 
 
 ~ 
 
 * 
 1 
 
 3 
 1 
 
 ol 
 
 49 
 53 
 39 
 
 Total No. of all 
 Oases treated. 
 
 1887-88 
 1888-89 
 1889-90 
 
 18 
 
 15 
 
 Vac at. 
 
 8 
 
 4 
 
 11 
 
 4 
 8 
 5 
 
 54,494 
 :!!>,'.i73 
 41,470 
 
 Total 
 
 33 
 
 23 
 
 2 
 
 1 
 2 
 
 5 
 
 17 
 
 2 
 5 
 2 
 
 9 
 
 24 
 
 11 
 10 
 10 
 
 31 
 
 •i 
 
 4 
 1 
 
 5 
 
 is 
 
 4 
 
 15 
 
 5 
 
 141 
 
 135,937 
 
 1887-88 
 1888-89 
 1889-90 
 
 Vacat. 
 12 
 13 
 
 6 
 4 
 3 
 
 13 
 
 10 
 
 8 
 2 
 
 20 
 
 15 
 
 9 
 
 20 
 
 44 
 
 1 
 
 1 
 •J 
 
 7 
 
 51 
 53 
 55 
 
 159 
 
 7,919 
 11,965 
 14,346 
 
 34,230 
 
 Total 
 
 25 
 
 In the 
 
 Medical 
 Polikliniks 
 
 arid 
 
 Iu the 
 Medical 
 Clinics 
 
 In the Medical Clinic of the Berlin Charite Hospital there were in the 
 
 Year 1874, amongst 326S medical cases, 9 diabetic patients, 1 : 363. 
 „ 1875, ,, 3076 ,, 7 „ 1:439. 
 
 „ 1888-89, ., 3605 ,, 12 ,, 1:300. 
 
 „ 1889-90, „ 5239 „ 13 „ 1:402. 
 
 Of the 
 Prussian 
 
 Universities. 
 
 Ebstein also mentions that, in the Eppendorff Hamburg 
 Hospital, amongst 7710 medical cases there were only four cases 
 of diabetes, i.e. one in 1927. 
 
 The disease is therefore met with more frequently in the 
 Manchester Eoyal Infirmary than in German hospitals. 
 
 In India, Ceylon, South Italy, and Malta, diabetes is much 
 more common than in most other countries. 
 
 Bose ( 6 ) has pointed out that of the various races of India it 
 is the Hindus who chiefly suffer. Amongst 250 cases which he 
 has collected, only two were Mohammedans, seventeen were 
 Europeans, and 231 Hindus. All the 250 cases were males. 
 
GEOGRAPHICAL DISTRIBUTION. 
 
 99 
 
 Bose points out that in India the disease is chiefly one of middle 
 life, but it also occurs, though rarely, at an early age. He 
 believes that the disease is increasing greatly. 
 
 Sen ( 7 ) also has shown that in Lower Bengal the Hindus 
 suffer more than other races ; he gives the following table of 
 the mortality from diabetes in Calcutta : — 
 
 
 1876-80. 
 
 1881-85. 
 
 1886-90. 
 
 Male. 
 
 Female. 
 
 Total. 
 
 Male. 
 
 Female. 
 
 Total. 
 
 Male. 
 
 Female. 
 
 Total. 
 
 Hindus . . 
 
 Mohammed- 
 ans 
 Non-Asiatics 
 
 98 
 3 
 
 13 
 
 Ill 
 3 
 
 142 
 2 
 
 17 
 1 
 
 159 
 3 
 
 167 
 3 
 3 
 
 36 
 2 
 
 203 
 3 
 5 
 
 The following table shows the proportion of races amongst 
 the inhabitants of Calcutta (Sen) : — 
 
 
 Hindus. 
 
 Mohammedans. 
 
 Christians. 
 
 Other Religions. 
 
 Male. 
 
 Female. 
 
 Male. 
 
 Female. 
 
 Male. 
 
 Female. 
 
 Male. 
 
 Female. 
 
 1881 . . . 
 1891 . . . 
 
 168,107 
 
 272,432 
 
 98,538 
 156,330 
 
 75,731 
 125,591 
 
 33,722 
 63,635 
 
 11,581 
 13,690 
 
 10,484 
 12,716 
 
 2,359 
 3,326 
 
 1,149 
 1,681 
 
 
 Total— 1881, 401,671 
 
 ; 1891, 649,401. 
 
 
 The following table shows the mortality from diabetes in 
 Calcutta per 100,000 deaths, from 1876-1890 (Sen) : — 
 
 1876 . . 
 
 263 
 
 1881 . . 
 
 215 
 
 1877 . . 
 
 173 
 
 1882 . . 
 
 291 
 
 1878 . . 
 
 107 
 
 1883 . . 
 
 263 
 
 1879 . . 
 
 . • 190 
 
 1884 . . 
 
 276 
 
 1880 . . 
 
 196 
 
 1885 . . 
 
 314 
 
 1886 
 1887 
 1888 
 1889 
 1890 
 
 384 
 382 
 445 
 325 
 379 
 
 In Sweden, East Prussia, and the middle Bhine provinces, 
 diabetes is said to be more common than in other parts of the 
 German Empire (v. Noorden). According to Striimpell, in 
 Germany the disease is most common in "Wiirtemberg and 
 Thiiringen. Seegen ( 8 ) states that a large proportion of the 
 diabetic patients who visit Carlsbad come from Frankfurt-am- 
 Main and Thiiringen. He points out that his diabetic patients 
 from Thiiringen were mostly poor, hard-working people, and that 
 
i oo E TIOL OGY AND E TIOL GICAL EEL A TIONS. 
 
 a large proportion of the patients from Frankfurt were Israelites. 
 Twenty-five per cent, of his cases were Israelites. Making 
 allowance for the frequency of the visits of wealthy Jews to 
 Carlsbad and other spas, he thinks that 10 per cent, would 
 indicate fairly the proportion of Israelites amongst his diabetic 
 patients. Seegen attributes this frequency of the disease 
 amongst the Israelites to a less stable condition of the Hebrew 
 nervous system. 
 
 Other writers have drawn attention to the frequency of the 
 disease in the Hebrew race. Frerichs states that of his 400 
 cases 102 were Jews. 
 
 But amongst the wealthy Jews it is probable that the con- 
 ditions of life play an important part in the causation of the 
 disease, as well as racial peculiarity. Purely states that the 
 Israelites generally suffer from a mild form of the disease, and 
 that the patients have usually been great eaters. 
 
 At the Manchester Eoyal Infirmary a considerable number 
 of very poor Russian, Polish, and German Jews are admitted 
 as in-patients, but, as far as one can judge, diabetes certainly 
 does not appear to be more common amongst these poor Jewish 
 patients than amongst the British patients. 
 
 Wallach ( 9 ), however, has clearly demonstrated the greater 
 frequency of the disease amongst the Jews in Frankfurt. During 
 nineteen years, 1872—1890, there were 171 deaths from 
 diabetes in Frankfurt. The proportion of deaths from diabetes 
 to the number of deaths from all causes was six times greater 
 amongst the Jews than amongst the rest of the inhabitants. 
 Carefully prepared tables show that the mortality from diabetes 
 per 1000 was greater at all ages amongst the Jews than 
 amongst the rest of the inhabitants of Frankfurt : — 
 
 Age. 
 
 Deaths from 
 Diabetes per 1000 
 amongst the Jews. 
 
 Deaths from 
 Diabetes per 1000 
 amongst people of 
 
 other Faiths. 
 
 General mortality 
 from Diabetes per 
 1000 Inhabitants. 
 
 0-14 years .... 
 
 02 
 
 0-005 
 
 007 
 
 14-19 „ 
 
 
 
 
 0-007 
 
 0-005 
 
 19-29 „ 
 
 
 
 011 
 
 03 
 
 0-04 
 
 29-39 ., 
 
 
 
 004 
 
 02 
 
 02 
 
 39-49 „ 
 
 
 
 019 
 
 09 
 
 010 
 
 49-59 „ 
 
 
 
 1-01 
 
 20 
 
 38 
 
 59-69 „ 
 
 
 
 2-40 
 
 0-69 
 
 0-92 
 
 69-79 ,, 
 
 
 
 2-86 
 
 0-54 
 
 94 
 
 79-89 „ 
 
 
 
 2 52 
 
 031 
 
 0/5 
 
 89-99 „ 
 
 
 
 
 
 
RACIAL INFLUENCE. 101 
 
 Purely ( 10 ) draws attention to the remarkable fact that the 
 mortality reports of the census for 1880 in the United States 
 of America do not furnish a single death from diabetes in either 
 the Indian or Chinese population of the country. Indians are 
 spare eaters, and subsist almost exclusively on nitrogenous food. 
 With regard to the Chinese, the explanation is by no means 
 easy, but the fact is interesting ; diabetes does occur in China, 
 however. The apparent exemption may be due to race 
 peculiarity. Diabetes is said to be rare in Japan. 
 
 From a careful study of the mortality from diabetes in 
 various regions of the United States, Purdy ( u ) concludes that it 
 is highest in the regions where there is " the lowest range of tem- 
 perature, in conjunction with the higher altitudes, and vice versd." 
 Dickinson ( 12 ) concludes that diabetes is more common in 
 the colder counties of England. He also thinks that the statistics 
 show that the disease is more prevalent in agricultural than in 
 urban districts. 
 
 Purdy has shown that in the United States the relative 
 mortality of diabetes in rural and urban populations is chiefly 
 determined by temperature ; in the colder regions the mortality 
 being decidedly higher in the country, whilst in the warmer 
 regions it is higher in the cities. 
 
 Henniker, in the Forty-eighth Annual Eeport of Eegistrar- 
 General (1885), draws attention (at pages xix.— xxi.) to the differ- 
 ence in the mortality from diabetes in various localities of 
 England. Ho thinks there is an inverse relation between the 
 prevalence of diabetes and the rainfall ; that where the rainfall 
 is least, as in eastern counties, the mortality is greatest. 
 
 The mean annual mortality from diabetes per million, from 
 1874 to 1883, was forty-two in England and Wales, thirty-one in 
 Scotland, twenty-five in Ireland. In Scotland and Ireland the 
 rainfall is much in excess of that of England. Of the various 
 English counties, the death-rate per million living from diabetes 
 during 1885—86 was highest in Norfolk; Berkshire came next, 
 and then followed Suffolk, Derbyshire, Sussex, Ptutland, whilst 
 London and Lancashire came low down on the list. 
 
 3. Is diabetes mellitus becoming more prevalent? — An inspec- 
 tion of the reports of the Eegistrar-General for the last forty- 
 five years shows that the number of deaths registered as due to 
 diabetes, and therefore the mortality per million living, has 
 steadily increased. 
 
i o 2 E TIOL OGY AND E TIOL GICAL RELA TIONS. 
 
 Deaths from Diabetes Mellitus per million living ( 
 
 in England) 
 
 1850 . 
 
 . 24 
 
 1873 ... 35 
 
 1851 . 
 
 • . 23 
 
 1874 . 
 
 . 37 
 
 1852 . 
 
 . 21 
 
 1875 . 
 
 '. 39 
 
 1853 . 
 
 . 23 
 
 1876 . 
 
 . 37 
 
 1854 . 
 
 . 24 
 
 1877 . 
 
 . 41 
 
 1855 . 
 
 . 24 
 
 1878 . 
 
 . 42 
 
 1856 . 
 
 . 23 
 
 1879 . 
 
 . 41 
 
 1857 . 
 
 . 25 
 
 1880 . 
 
 . 41 
 
 1858 . 
 
 . 27 
 
 1881 . 
 
 . 47 
 
 1859 . 
 
 . 25 
 
 1882 . 
 
 . 47 
 
 1860 . 
 
 . 27 
 
 1883 . 
 
 . 51 
 
 1861 . 
 
 . 27 
 
 1884 . 
 
 . 54 
 
 1862 . 
 
 . 29 
 
 1885 . 
 
 . 55 
 
 1863 . 
 
 . 27 
 
 1886 . 
 
 . 59 
 
 1864 . 
 
 . 32 
 
 1887 . 
 
 . 62 
 
 1865 . 
 
 . 32 
 
 1888 . 
 
 . 62 
 
 1866 . 
 
 . 32 
 
 1889 . 
 
 . 60 
 
 1867 . 
 
 . 32 
 
 1890 . 
 
 . 65 
 
 1868 . 
 
 . 31 
 
 1891 . 
 
 . 66 
 
 1869 . 
 
 . 34 
 
 1892 . 
 
 . 68 
 
 1870 . 
 
 . 33 
 
 1893 . 
 
 . 70 
 
 1871 . 
 
 . 35 
 
 1894 . 
 
 . 68 
 
 1872 . 
 
 . 33 
 
 1895 . 
 
 . 75 
 
 No doubt the increase of the number of deaths registered as 
 diabetes mellitus is due in some degree to the improved training 
 of the medical man of the present day ; cases are not so frequently 
 overlooked at the present time as they were many years ago. 
 But the increase is much too great and too constant to be 
 explained in this way. Thus, in 1873 the death-rate was 35 
 per million, in 1895 it was 75 per million. The sugar tests 
 usually employed at the present time are much the same as 
 those of twenty years ago, and it cannot be maintained that 
 medical men are now twice as careful in diagnosis and urine 
 testing as they were then. 
 
 It is interesting to note that several Kegistrar-Generals have 
 drawn attention to this increasing death-rate from diabetes. 
 
 Dr. Tatham, in the report for the ten years 1881—1890, 
 especially draws attention to the increase. As he points out, 
 the mortality has already become considerable, and is increasing 
 year by year. 
 
INCREASE OF THE DISEASE. 
 
 10 3 
 
 Thus for the ten years 1871—1880 it was 38 per million 
 living, whilst in 1881-1890 it had risen to 57 per million. 
 This increased mortality has been noticed in other countries also. 
 Thus Purely gives the following figures with respect to the 
 United States : — 
 
 Rates of Deaths from Diabetes in the United States. 
 
 Year. Eatio. 
 
 1850 . . . . 72 to 100,000 deaths. 
 
 1860 .... 98 „ 
 
 1870 . . . . 170 „ 
 
 1880 ...-. 191 „ 
 
 Purdy attributes this remarkable increase to more extrava- 
 gant and luxurious living. 
 
 Lepine ( 13 ) has pointed out that in Paris during the last 
 thirty years the mortality has increased six fold — from 2-3 per 
 100,000 to 13 per 100,000 inhabitants. 
 
 In Denmark, Caroe ( 14 ) has shown that the same increase has 
 occurred. Thus, from 1860—69 the death-rate from diabetes in 
 Copenhagen and the towns of Denmark was 2 per 100,000: 
 whilst in 1890-94 it had risen to 8 per 100,000 inhabitants. 
 
 Diabetes Mellitas amongst the In-patients at the Manchester Royal 
 Infirmary, 1875-95. 
 
 
 Total No. of 
 
 Medical 
 In-patients. 
 
 No. of Cases of 
 
 Diabetes Mellitus 
 
 among'st Medical 
 
 Cases. 
 
 1875-76 .... 
 
 981 
 
 8 
 
 1876-77 
 
 
 
 
 1003 
 
 6 
 
 1877-78 
 
 
 
 
 1277 
 
 10 
 
 1878-79 
 
 
 
 
 1280 
 
 9 
 
 1879-80 
 
 
 
 
 1351 
 
 22 
 
 1880-81 
 
 
 
 
 1433 
 
 26 
 
 1881-82 
 
 
 
 
 1428 
 
 13 
 
 1882-83 
 
 
 
 
 1519 
 
 19 
 
 1883-84 
 
 
 
 
 1543 
 
 10 
 
 1884-85 
 
 
 
 
 1658 
 
 16 
 
 1885-86 
 
 
 
 
 1474 
 
 17 
 
 1886-87 
 
 
 
 
 1448 
 
 14 
 
 1887-88 
 
 
 
 
 1477 
 
 11 
 
 1888-89 
 
 
 
 
 1511 
 
 17 
 
 1889-90 
 
 
 
 
 1525 
 
 16 
 
 1890-91 
 
 
 
 
 1528 
 
 11 
 
 1892 
 
 
 
 
 1376 
 
 12 
 
 1893 
 
 
 
 
 1374 
 
 18 
 
 1894 
 
 
 
 
 1286 
 
 5 
 
 1895 
 
 
 
 
 1249 
 
 12 
 
io 4 ETIOLOGY AND ETIOLOGICAL RELATIONS. 
 
 There is no definite increase to be noted, however, in the 
 proportion of cases of diabetes amongst the medical patients at 
 the Manchester Boyal Infirmary. This is shown by the table on 
 p. 103. Perhaps the table is not quite a reliable indication 
 of the frequency of the disease during the various years. During 
 the years 1875—80 Sir William Eoberts was physician to the 
 hospital, and his great interest in urinary diseases no doubt 
 caused some increase in the number of diabetic patients who 
 sought relief at the Manchester Infirmary at that time. During 
 the last ten years, however, the figures may be taken as a fairly 
 reliable indication of the frequency of the disease. 
 
 During the ten years 1875—85 there were, therefore, 139 
 cases of diabetes mellitus amongst 13,483 medical in-patients; 
 during the ten years 1885-95 there were 133 cases amongst 
 14,238 medical in-patients. 
 
 4. Social position. — v. Noorden thinks that the poorer 
 classes of society suffer less from the disease. Frerichs found 
 that only sixty out of his 400 cases belonged to the poorer 
 classes, v. Noorden thinks that riches, culture, and good social 
 position increase the liability to the disease ten fold. Diabetes has 
 been thought to be more common amongst persons whose occupa- 
 tion demands much mental work and is associated with mental 
 worry and irritation. It has been stated that it is more prevalent 
 amongst learned men, musicians, poets, schoolmasters, merchants, 
 and politicians. Also it is stated to be frequently met with 
 amongst stout, wealthy, indolent persons. 
 
 In spite of these statements, however, it must be remem- 
 bered that the disease, often in its most severe form, is met 
 with amongst poor, hard-working people ; whilst amongst the 
 wealthy and better educated classes many of the cases are 
 examples of the mild form of the disease. 
 
 5. Heredity. — The influence of heredity in the causation of 
 the disease has been variously estimated. As above stated, it 
 appears probable that certain races are a little more prone to 
 suffer from the disease, as, for example, the Jews and Hindus. 
 
 Sometimes there is a distinct family history of diabetes. 
 Several members of the same family, children of the same 
 parents, occasionally suffer from the disease ; sometimes an uncle 
 or an aunt of the diabetic patient has been similarly affected ; 
 occasionally, but very rarely, the father or mother have been 
 diabetic. But this family history of diabetes is not common. 
 
SOCIAL POSITION AND HEREDITY. 105 
 
 In the first 100 cases of diabetes in which I took careful notes 
 with respect to the family history, I found that evidence of 
 heredity could be obtained in thirteen only. Most of these cases 
 were hospital patients, and therefore 13 per cent, will probably 
 be a minimum estimate of evidence of heredity. 
 
 The following were the instances amongst 100 consecutive 
 cases in which a family history of the disease could be 
 obtained : — 
 
 Case 1. — Female, set. 20 : two sisters also suffer from a severe form 
 of diabetes. 
 
 Cases 2, 3, 4. — In each case a brother of the patient has died from 
 diabetes. 
 
 Case 5. — A brother of the patient is said to have suffered from 
 diabetes, and to have recovered. 
 
 Cases 6, 7, 8. — In each case an uncle of the patient had suffered from 
 the disease. 
 
 Case 9. — Patient's aunt had suffered from diabetes. 
 
 Case 10. — The patient's mother is said to have suffered from 
 diabetes many years ago, but to have recovered; her urine is now free 
 from sugar. 
 
 Case 11. — Patient's mother is said to have suffered from glycosuria. 
 
 Case 12. — Patient's grandfather died of diabetes. 
 
 Case 13. — Patient's father and uncle died of diabetes. 
 
 In twenty-eight cases of diabetes in children, collected by 
 Wegeli ( 15 ), there was a family history of the disease in eight. 
 In seven of these cases a brother or sister had suffered from 
 diabetes. 
 
 Often there is a history of phthisis or mental disease in the 
 family of a diabetic patient, and in the mild form of the disease 
 not infrequently a family history of gout or obesity is obtained. 
 
 6. The possibility of infection. — Schmitz ( 16 ) in 1890 drew 
 attention to this subject. He recorded twenty-six instances 
 amongst 2320 cases of diabetes, in which apparently healthy 
 persons, living in intimate association with diabetics, had 
 developed the disease. Most cases were instances of married 
 females, who had become diabetic after nursing a diabetic 
 husband. In these cases there were no indications of hereditary 
 predisposition ; there was no family relationship between the 
 patients; no excess of sugar had been taken in the food; and 
 the patients had not suffered from gout. Hence Schmitz raised 
 the question of the possibility of the transmission of the disease. 
 
106 ETIOLOGY AND ETIOLOGICAL RELATIONS. 
 
 This view seems very improbable, however, and has been much 
 disputed. Nevertheless the association is interesting, and seems 
 to point to some common cause, at least in some of the cases, as 
 it is scarcely likely that all are mere coincidences. It is well 
 known that the nursing of a patient during a long illness, and 
 the anxiety connected therewith, have occasionally been the excit- 
 ing causes of diabetes. Now, if a patient be suffering from diabetes, 
 and be nursed by his wife, and if the anxiety and strain should 
 bring on diabetes in the wife, then we should have the occur- 
 rence of the disease in two persons living together. If to these 
 cases we acid those due to accidental coincidence, an explanation 
 may perhaps be offered for the facts above mentioned. I have 
 never met with any instances of people living together who have 
 both suffered from the disease, but I have recently been some- 
 what surprised to find that three diabetic patients at the 
 Manchester lloyal Infirmary all came from one small suburb of 
 Manchester — Didsbury. One patient was a child, ast. 11, who 
 died of diabetes after an illness of six weeks only. The other 
 was a girl, a?t. 12, and the third was a young adult female. 
 Considering the rarity of diabetes, the occurrence of three cases 
 in this small suburb is somewhat remarkable. 
 
 Oppler and Kiilz ( 17 ) have recently given a critical review of 
 the cases of supposed infection hitherto published, and have 
 recorded ten additional examples, taken from 900 cases of 
 diabetes seen by the late Professor Kiilz. 
 
 Besides the ten (duplicate) cases they record, they have also 
 found sixty-seven cases reported in medical literature. The 
 numerical relation between diabetic married couples and other 
 diabetic cases is shown in the following table, giving the results 
 of the observations of several authors : — 
 
 Betz 
 
 Hertzka 
 
 Lecorche 
 
 Schmitz 
 
 Seegen 
 
 Kiilz 
 
 1 married couple amongst 31 diabetic patients. 
 
 1 „ 86 
 
 6 „ 114 
 
 26 ., 2320 
 
 3 „ 938 
 
 10 „ 900 
 
 Total . . 47 married couples amongst 4389 diabetic patient?, 
 or 1 : 93|, or T08 per cent. 
 
 Since the proportion of cases is so small, and since the 
 influence of heredity and other predisposing causes cannot be 
 
POSSIBILITY OF INFECTION. 107 
 
 excluded with certainty, Oppler and Ktilz do not think there 
 is sufficient evidence for accepting the view that diabetes is 
 contagious. 
 
 Senator ( 18 ) states that amongst 770 cases of diabetes that 
 have come under his observation, there have been nine instances 
 of man and wife suffering from the disease. This proportion — 
 1-19 per cent. — corresponds pretty closely with the statistics of 
 Oppler and Kiilz. 
 
 But Senator points out that when all the cases are excluded 
 in which there is a family history of the disease, or a history of 
 any of the well-known etiological antecedents, the cases remain- 
 ing (of diabetes in man and wife) are so few that it seems 
 probable that the occurrence is accidental, or that both man and 
 wife have been subjected to the same antecedents. 
 
 7. External Injury. — In the section devoted to glycosuria, it 
 has been pointed out that after injuries to the head a small quantity 
 of sugar is sometimes found in the urine for a short time. But, 
 long ago, attention was drawn by Goolden ( 19 ) and Fischer ( 20 ) 
 to the fact that true diabetes occasionally follows an external 
 injury. When we consider how many people suffer from the 
 effects of external injury, and in what a very small proportion 
 of cases diabetes follows, it is quite evident that there must be 
 some other factor beside trauma in the causation of the disease 
 in these cases. Moreover, in a series of cases of diabetes a 
 history of trauma is obtained only in a small proportion. 
 Nevertheless, cases are occasionally met with in which there 
 appears to be little doubt that the injury has been the exciting 
 cause, or at least an important factor in the causation of the 
 disease. The following is an example of diabetes following and 
 apparently due to an injury to the head ( 21 ) : — 
 
 M. A. E., set. 18 (under the care of Dr. Steell at the Manchester 
 Infirmary). The patient was a strong, healthy, stout girl up to the 
 time of an accident seven months before admission. On going down- 
 stairs she suddenly slipped and fell down thirteen steps, the top 
 of her head coming violently in contact with a door at the bottom 
 of the steps. She was stunned by the fall, but did not lose con- 
 sciousness. A dish which she had in her hand at the time was 
 broken in the fall, and her face was cut by the broken porcelain just 
 below the right eye. She received a deep cut on the flexor surface 
 of the right forearm just below the elbow. She had great pain in 
 the head for many hours, and very frequently suffered from severe 
 
i o S E TIOL OGY AND E TIO L O GICAL RE LA TIONS. 
 
 headache afterwards. Two weeks after the accident she first noticed 
 thirst, which was soon followed by wasting. On admission, the patient 
 was emaciated; knee-jerks absent; urine acid, 1045, no albumin; 
 large amount of sugar present (23 grs. to the ounce) ; marked 
 brownish-red coloration with Fe 2 Cl 6 . During the time the patient 
 was in the hospital the amount of urine at first varied from 80 to 90 
 oz. daily, afterwards from 56 to" 68 oz. daily, with increase of sugar 
 to 33 grs. per ounce. The disease terminated fatally about eighteen 
 months later. 
 
 I have met with a few other cases in which diabetes has 
 followed an injury to the head or some other part of the body, 
 but in these cases the connection of the disease with the injury 
 has been less evident than in the case recorded above. In 100 
 consecutive cases of diabetes mellitus in which I made careful 
 inquiries, a history of external injury was obtained in six only. 
 In two of these six cases the connection between the injury and 
 the causation of the disease seemed doubtful. Ebstein obtained 
 a history of external injury in six cases out of 116 diabetic 
 patients. 
 
 The earlier the diabetic symptoms follow the accident, 
 and the greater the probability that there were no diabetic 
 symptoms previously, the more likely it is that diabetes is due 
 to the accident. Ebstein ( 22 ) has recently collected fifty cases 
 of traumatic diabetes from his own clinic and from literature. 
 From a review of the cases, he concludes that individual pre- 
 disposition is an important factor. Occupation, age, and sex 
 appear to be of no importance. Injuries to the head appear to 
 be followed by diabetes more frequently than injuries to other 
 parts. In one-half of the fifty cases of traumatic diabetes col- 
 lected by Ebstein the head was the seat of injury. In other 
 cases, injuries to the neck, liver, region of kidney, pubes, etc., 
 have been followed by the disease. 
 
 From the fact that only an exceedingly small proportion of 
 injured persons become diabetic, we must conclude that there 
 exists some peculiarity in the nature of the injury, or some 
 peculiarity in the individual, in cases of traumatic diabetes. 
 The second view is the more probable. 
 
 Ebstein and Asher ( 23 ) have reported cases of traumatic 
 neurosis which has been followed by diabetes. 
 
 8. Mental Emotions. — It has long been known that diabetes 
 sometimes follows, and appears to be caused by, mental emotions, 
 
MENTAL EMOTIONS, OBESITY, GOUT. 109 
 
 such as fright, anxiety, mental worry, etc. These antecedents 
 will be discussed later, in the section on the Relation of Diabetes 
 to Affections of the Nervous System, p. 122. 
 
 9. Obesity. — It is well known that diabetes frequently occurs 
 in the obese. Frerichs states that one-seventh of diabetics suffer 
 from obesity (59 in 400). Women at the climacteric period, 
 especially when they are stout, are liable to suffer from diabetes 
 (Frerichs). In hospital practice, obesity is rarely connected 
 with diabetes, but the association is much more common in 
 private practice. Seegen points out that often obesity is present 
 first ; and generally diabetes does not develop for years after 
 the obesity has become marked. In these cases the development 
 of diabetes is slow ; the disease generally commences between 
 the age of 40 and 60, and it is almost always of a mild 
 form ; sugar usually disappears from the urine after a rigid diet, 
 and the prognosis is favourable. But in another group of cases 
 obesity occurs in young individuals, at an age when obesity 
 is rare ; it develops rapidly and to a marked extent, and then 
 symptoms of diabetes in its severe form appear. In many of 
 these cases there is a predisposition to diabetes in consequence 
 of heredity or brain affection. In cases of the first group, 
 obesity predisposes to diabetes. In cases of the second group, 
 there seems to be a more intimate connection, and the obesity 
 appears to be a forerunner of the diabetes, but the exact relation 
 has not been determined (Seegen 2i ). 
 
 The following case is an example of the latter variety : — 
 
 Mary A. L., set. 29, in-patient at the Manchester Royal Infirmary 
 (under the care of Dr. Leech). As a girl she was exceedingly fat, and 
 was known in the country district where she lived as "fat Polly." She 
 had influenza two years before admission to the hospital, and states that 
 she has "never been well since." Ten months before coming under 
 observation she began to suffer from thirst and diuresis. She has lost 
 flesh and become weak. Weight previous to onset of diabetes, 200 lb. 
 Present weight, 15th October 1893, 107 J lb. She is suffering from a 
 severe form of diabetes. Urine 1038, large quantity of sugar present; 
 marked reaction with perchloride of iron. 
 
 1 0. Gout. — Diabetes and gout are frequently associated. Grube 
 ( 25 ), of Neuenahr, found that sixteen out of 177 diabetic patients 
 suffered from gout, and twenty-three had gouty parents. But 
 it is to be remembered that treatment by the Neuenahr mineral 
 waters is especially recommended for the mild forms of diabetes 
 
no ETIOLOGY AND ETIOLOGICAL RELATIONS. 
 
 occurring in gouty patients, and this fact probably accounts, in 
 part, for the frequency of the association of gout and diabetes 
 in Grube's cases. In the diabetic cases at the Manchester 
 Eoyal Infirmary (mostly poor patients suffering from a severe 
 form of the disease), there is very seldom any association of the 
 two affections. 
 
 v. Noorden ( 26 ) points out that gout and diabetes may be 
 associated in three ways — 
 
 (1) Patients who are gouty in middle life may suffer from 
 glycosuria at a later period. 
 
 (2) Gout and glycosuria may alternate. 
 
 (3) Gout and glycosuria may be present together. 
 
 In the cases in which the two diseases are associated, the 
 diabetes is usually of a mild form, the general condition is 
 good, and the prognosis is favourable. But if diabetes is present 
 first, and gouty symptoms develop later, the prognosis is more 
 unfavourable. 
 
 11. Alcoholism. — It has already been pointed out that in 
 great beer drinkers alimentary glycosuria can often be produced 
 more readily (i.e. with a smaller amount of glucose) than in 
 healthy persons. In Manchester a history of alcoholism is not 
 infrequently obtained from diabetic patients. In 100 consecut- 
 ive cases of diabetes mellitus I obtained a history of very great 
 alcoholic excess in seventeen. All of the seventeen patients 
 were males, and the form of alcohol was generally English beer, 
 of which enormous quantities had been taken daily for long 
 periods. 
 
 Strumpell ( 27 ) points out that he has frequently met with 
 diabetes in persons who have been great beer drinkers. Often 
 these were cases of " diabetes with obesity." Frequently they 
 were mild forms of the disease, but sometimes severe forms, 
 complicated with gangrene and tuberculosis of the lungs, and 
 often they have been complicated with other results of 
 alcoholism — alcoholic neuritis, chronic nephritis, etc. Strongly 
 in favour of the view T of a " beer diabetes " are the results of 
 Striimpeli's observations on alimentary glycosuria in great beer 
 drinkers. No sharp line can be drawn between alimentary 
 glycosuria and diabetes. At first, as a result of excessive beer 
 drinking, the sugar destroying function of the system is weakened, 
 and alimentary glycosuria occurs ; later, the sugar destroying 
 function is lost, and permanent diabetes follows. 
 
A L CO HOLISM AND INFECTIO US DISEASES. 1 1 1 
 
 Striimpell believes that it is excessive beer drinking, rather 
 , than simple alcoholism, which is the cause of certain cases of 
 diabetes. 
 
 12. Influenza. — During the last six years influenza appears to 
 have been the exciting cause in a considerable number of cases 
 of diabetes. But when we consider how many persons have 
 suffered from influenza during the recent epidemics, how very 
 seldom diabetes has followed, and how rarely diabetes can be 
 traced to a previous attack of influenza, it is quite evident that 
 there is always some other factor in the causation of the disease. 
 Nevertheless, in a few cases, an attack of influenza does appear to 
 have had some influence as an exciting cause. 
 
 In 100 cases of diabetes, in which I have taken careful 
 notes on the previous history, I find that in six influenza appears 
 to have played some part as an exciting cause. 
 
 In two of these cases the symptoms of diabetes first appeared 
 whilst the patient was suffering from an attack of influenza, and 
 continued from that date. In two the diabetic symptoms first 
 appeared a few days after an attack of influenza. In two cases 
 there was a considerable interval of time between the attack of 
 influenza and the onset of diabetic symptoms ; but in both cases 
 the patients declared that they " were never well after the 
 influenza " ; the influenza attacks were followed by a low state of 
 health, and finally diabetes developed. In one case a patient had 
 suffered from diabetes, but the symptoms had subsided; an 
 attack of influenza was followed, however, by a return of the 
 diabetic symptoms, which persisted. 
 
 13. Acute infectious diseases. — Symptoms of diabetes are 
 sometimes first noticed soon after an attack of one of the specific 
 fevers, such as typhoid fever, scarlet fever, cholera, diphtheria, 
 etc. ; and occasionally diabetes follows an acute illness such as 
 acute rheumatism, acute tonsillitis. I have met with one 
 instance in which the diabetic symptoms first appeared when the 
 patient was recovering from an attack of pleurisy ; in another 
 case they first appeared as the patient was recovering from an 
 attack of pneumonia. 
 
 Probably, in' some instances, the association of the two 
 diseases is accidental. Certainly there must be some other 
 factor in the causation besides the previous febrile affection ; 
 though it is very probable that an acute illness may act 
 occasionally as an exciting cause. We know that in acute 
 
1 1 2 ETIOLOGY AND E TIOLOGICAL RELA TIONS. 
 
 diseases secondary changes often occur in the heart, lungs, kidneys, 
 and other organs, and it is possible that occasionally the pancreas 
 may be affected and diabetes produced. 
 
 14. Exposure to cold and wet. — Occasionally diabetic 
 symptoms have been attributed by the patient to exposure to 
 cold and wet. Sometimes diabetic symptoms have first been 
 noticed when the patient has been suffering from, or just 
 recovering from, an attack of nasal catarrh and bronchitis. 
 
 In three cases out of 100 in Manchester, the diabetic 
 symptoms first developed whilst the patients were suffering from 
 a severe cold. 
 
 15. Drinking of cold fluids. — A number of cases are on 
 record, in which apparently healthy persons have developed 
 diabetic symptoms directly after drinking large quantities of 
 cold fluid, or after taking ices, when the body has been very hot 
 and covered with perspiration. 
 
 In one instance which I have met with, the symptoms were 
 first noticed after drinking a large quantity of cold water one 
 very hot summer's day. Diabetic symptoms were definitely 
 stated to have been absent previously. In two cases the thirst 
 was said to have commenced suddenly whilst the patients were 
 drinking beer. Peiper ( 2S ) reports the case of a person who, 
 whilst very hot at a ball, drank a very large quantity of cold 
 water, and immediately diabetic symptoms developed. 
 
 16. Malaria. — The relation between malaria and diabetes is 
 disputed. Some authors regard malaria as an occasional exciting 
 cause of diabetes, but by others this relationship is denied. 
 Cantani ( 29 ) is not inclined to attach any importance to malaria 
 as a cause of diabetes, except perhaps in so far as it increases 
 the vulnerability of the organism generally, and diminishes its 
 resisting power. In 2 1 8 of his cases, only thirteen had suffered 
 from malaria — a percentage which is very small considering the 
 frequency of malaria in Italy. 
 
 17. Lightning stroke. — Hermanides ( 30 ) reports a case in 
 which a man, after being struck by lightning, suffered from 
 severe headache. One year later diabetes developed, and two 
 years later hemianopsia was detected. The disease terminated 
 fatally. Post-mortem examination showed that the dura mater 
 over the hemispheres was markedly thickened in many places, and 
 adherent to the pia. The latter was turbid and adherent to the 
 brain cortex, most markedly in the region of the right parietal lobe. 
 
SYPHILIS, FOOD. 113 
 
 18. Syphilis. — Several cases have been recorded, by Ord ( 31 ) 
 and others, in which glycosuria has been found in patients 
 presenting the ordinary general symptoms of constitutional 
 syphilis. A few cases are also on record, in which true diabetes 
 has been met with in syphilitic patients, and has very probably 
 been due to a syphilitic lesion. 
 
 Feinberg ( 32 ) has reported three cases of diabetes, and one of 
 glycosuria, which were apparently of syphilitic origin. In all of 
 these cases there were signs of syphilitic affection of the nervous 
 system, and under anti-syphilitic treatment recovery occurred in 
 one of the cases of diabetes, great improvement in the other 
 two, but in the case of glycosuria no definite improvement 
 followed. It is quite possible that syphilis may cause diabetes 
 by producing a specific lesion_ of the brain, especially of the 
 medulla. Syphilitic arterial disease, with secondary nervous 
 lesions in the region of the fourth ventricle, or a gumma in this 
 part, would be capable of producing diabetes. Also it is possible 
 that syphilis may give rise to diabetes by causing disease of the 
 blood vessels of the pancreas and secondary disease of the gland 
 tissue. 
 
 Though syphilis may be an occasional cause of diabetes, 
 there can be no doubt that such cases are exceedingly rare. In 
 100 cases of diabetes I obtained a history or indications of 
 previous syphilis in six only, and in these cases there was no 
 reason to regard syphilis as the cause — probably in all of the six 
 cases the association with syphilis was accidental. 
 
 19. Food. — It has already been pointed out, that if a very 
 large quantity of saccharine material be taken by a healthy 
 person, a trace of sugar appears in the urine for a short time. 
 But glycosuria cannot be produced by amylaceous food, however 
 great the quantity taken. Diabetes has often been attributed to 
 the excessive use of saccharine or amylaceous food. I have never 
 met with any case, however, in which I could satisfy myself that 
 there was any reason for attributing the disease to excess of 
 carbohydrate food, or to a special form of diet. 
 
 Cantani ( 33 ) thought that the excessive consumption of fari- 
 naceous and saccharine food predisposes to diabetes. Amongst 
 his Italian patients he seldom met with a diabetic who had not 
 taken excess of these articles of food. In the 210 cases which 
 came under his observation, generally the patient's diet had con- 
 sisted chiefly of farinaceous and saccharine food, and but little 
 
ii 4 ETIOLOG Y AND ETIOLOGICAL RELATIONS. 
 
 albumin. Many of his wealthy patients had eaten flesh meat 
 only once or twice a week, his poor patients only once or twice 
 a year. 
 
 In Tlniringen diabetes is more frequent than in other parts 
 of Germany, and Cantani points out that the people of that 
 district live chiefly on farinaceous food. 
 
 In Ceylon, where diabetes is so common, a large quantity of 
 saccharine food is taken. But the Chinese rarely suffer from 
 diabetes, and yet their diet consists chiefly of farinaceous food. 
 
 20. Diabetes and pregnancy. — It has been pointed out on 
 p. 88 that during the puerperal state lactose is present in the 
 urine. Apart from this condition of physiological lactosuria, 
 there appears to be occasionally some connection between preg- 
 nancy and the onset of true diabetes. Matthews Duncan ( 34 ) has 
 studied this association, and has collected twenty-two cases from 
 his own practice, and from medical literature. He draws the 
 following conclusions with respect to the connection between 
 diabetes and pregnancy : — 
 
 (1) Diabetes may come on during pregnancy. 
 
 (2) Diabetes may occur only during pregnancy, being absent 
 at other times. 
 
 (3) Diabetes may cease with the termination of pregnancy, 
 recurring some time afterwards. 
 
 (4) Diabetes may come on soon after parturition. 
 
 (5) Diabetes may not return in a pregnancy occurring after 
 its cure. 
 
 (6) Pregnancy may occur during diabetes. 
 
 (7) Pregnancy and parturition may be, apparently, unaffected 
 in its healthy progress by diabetes. 
 
 (8) Pregnancy is very liable to be interrupted in its course, 
 and probably always by the death of the foetus. 
 
 Amongst the cases of diabetes which have come under my 
 observation, the following have been associated with the puer- 
 peral state : — The symptoms first developed in one case just 
 after confinement ; in another case after a miscarriage, the fifth 
 which had occurred ; in another case after a fourth miscarriage ; 
 in another case the patient had had a miscarriage, after which 
 she had never been well, and soon symptoms of diabetes 
 developed. In three cases the diabetic symptoms developed 
 directly after the patient had suffered from an abscess of the 
 breast. In one case the child would not suckle, and three 
 
DIABE TES AND PREGNANC Y. 115 
 
 months later diabetic symptoms appeared. It is possible that 
 in some of these cases, pregnancy, the puerperal state, or the 
 condition of the breast, may have been exciting causes of the 
 disease. 
 
 21. Climacteric diabetes. — It has been thought by some writers 
 that the climacteric period favours the occurrence of diabetes in 
 women. 
 
 Tait( 35 ) believes that there is a climacteric glycosuria which 
 ends in recovery. 
 
 Imlach ( 36 ) reports an interesting and apparently well-marked 
 case of diabetes in which pyosalpinx was present. Eemoval of 
 the diseased uterine appendages was followed by disappearance 
 of the diabetic symptoms, which did not return, even when 
 ordinary diet was taken. 
 
 22. Sexual excess. — Occasionally the diabetic symptoms first 
 appear in previously healthy men not long after marriage, and the 
 question of the part played by sexual excess as an exciting cause 
 has been sometimes raised by the patients. It has not yet been 
 clearly shown, however, that sexual excess acts as an exciting 
 cause. 
 
 B. On the Belation between Diabetes Mellitus and 
 Diabetes Insipidus. 
 
 Experiments on animals have shown that there is a close 
 relation between polyuria, and polyuria with glycosuria. Clinic- 
 ally, also, there appears to be a close relation between diabetes 
 mellitus and diabetes insipidus in a few rare cases. With regard 
 to the etiology, it is well known that both forms of diabetes are 
 often connected with altered conditions of the nervous system. 
 Gross lesions are only found in a few cases, but there are fre- 
 quently indications of some minute or functional changes in the 
 nervous system in each affection. Also diabetes mellitus and 
 diabetes insipidus occasionally occur in near blood relatives. 
 Senator ( 37 ) has recorded a case of diabetes insipidus occurring 
 in a person whose mother died of diabetes mellitus. Another 
 indication of the relationship of the two diseases is the occa- 
 sional transition of one disease into the other, or the alternation 
 of the diseases in the same patient. The transition of diabetes 
 mellitus into diabetes insipidus is the more common. This is 
 a favourable change, and sometimes complete recovery finally 
 
1 1 6 E TIOL OGY AND ETIOL O GICAL RE LA TIONS. 
 
 occurs. Senator has recorded two cases of diabetes mellitus in 
 which the sugar disappeared and was replaced by albumin ; then 
 the albuminuria disappeared, and polyuria remained. Finally, 
 the patients recovered completely. He also reports a case of 
 diabetes insipidus in which glycosuria occurred ; the patient 
 became wasted, and the disease terminated fatally. West- 
 phal ( 3S ) has recorded the alternation of diabetes mellitus and 
 diabetes insipidus in the same patient. 
 
 All these facts indicate that occasionally the two diseases 
 are closely related. 
 
 C. Relation between Diabetes Mellitus and Diseases 
 of the Liver. 
 
 From the results of physiological experiments, one would 
 expect that there would be some clear relationship between 
 diabetes and pathological changes in the liver. If the function 
 of the liver be to prevent sugar entering the general circulation, 
 as Pavy believes, then one would expect that in serious hepatic 
 diseases this function would not be performed, and that diabetes 
 would follow. But from clinical experience and pathology we 
 know that extensive destruction of the liver parenchyma may 
 occur, in cases of cancer and cirrhosis of that organ, in phos- 
 phorus poisoning and other diseases, without any sugar appearing 
 in the urine. Glycosuria is very rare, even in advanced diseases 
 of the liver. Frerichs has recorded a case of cirrhosis of the 
 liver, in which subsequent post-mortem examination showed that 
 there was almost complete degeneration of the liver parenchyma, 
 yet no glycosuria occurred, even after a large quantity of sugar 
 had been taken by the mouth. Again, in nineteen cases of 
 phosphorus poisoning, large quantities of sugar (200 grms.) were 
 given by Frerichs, but in seventeen of the cases no glycosuria 
 was produced ; in two cases only did a small quantity of sugar 
 appear in the urine. Frerichs also found that there was no 
 sugar or glycogen in the liver in phosphorus poisoning when 
 fatty degeneration of that organ had occurred. 
 
 No definite or constant pathological change is met with in 
 the liver in diabetes, though this organ is sometimes diseased. 
 In the cases which have come under my observation patho- 
 logically, apart from hypertrophy, there have been generally 
 no definite naked eye changes in the liver. In a few cases, 
 
DIABETES AND DISEASES OF THE LIVER. 117 
 
 however, changes were present. In one ease there was cirrhosis 
 with fatty infiltration, and the liver was much enlarged ; 
 hut during life the diabetic symptoms were only slight. In 
 another case cirrhosis was present, but the patient had been a 
 great beer drinker. In one case multiple abscesses of the liver and 
 recent purulent cerebro-spinal meningitis were found post mortem. 
 The following are the notes of the case: — 
 
 J. R. came under my care on 28th September 1890. History 
 of an attack of vertigo and vomiting in February 1890. Five weeks 
 later he began to be troubled with thirst and polyuria. He had been 
 unemployed for eleven months, had lost a large sum of money, and been 
 subject to great mental anxiety before the onset of diabetes. He had 
 obtained employment, however, at a bleaching works about one month 
 before he first noticed diabetic symptoms. A brother died of diabetes. 
 
 When first seen the patient was considerably wasted. He suffered 
 from thirst and polyuria. Urine 1032, acid ; contained 5T7 per cent, of 
 sugar; no albumin. Knee-jerks absent. Commencing cataract in each 
 eye. Bowels regular. Heart, lungs, liver, and spleen normal. Four 
 months later the amount of sugar varied from 2500 to 4800 grs. daily, 
 and the amount of urine from 90 to 130 oz. From 21st January 1891 
 until 24th February the temperature was normal. At the latter date the 
 evening temperature was 102° F. 
 
 25th February. — Patient complained of pain in the epigastrium. 
 Urine contained for the first time a trace of albumin, and also gave for 
 the first time a dark reddish-brown coloration with Fe. 2 Cl 6 . Bowels 
 constipated. For the next three days the evening temperature was 
 between 102° and 102"8° ; then for four days the evening temperature 
 was between 99° and 100°. 
 
 ith March. — -Evening temperature 102°; pulse 96; respirations 25. 
 5th March. — Morning temperature 99 - 6° ; evening lOS^ . 
 6th March. — Morning temperature 101 "2°; evening 100-2°. Patient 
 complained of pain in the right hypochondriac region. Edge of liver 
 felt about one inch and a half below the ribs. No jaundice ; no rigors. 
 Patient became very restless and semi-conscious, but he could be roused 
 to answer questions. Enee-jerks absent ; calf muscles tender ; no 
 paralysis; no anaesthesia. Death at 2 a.m. on 7th March. Shortly 
 before death the pulse was 224 ; respirations 72. jS t o discharge from 
 ears or symptoms of suppurative otitis media during the patient's illness. 
 Necropsy (abstract). — Peritoneum, pericardium, heart and aorta nor- 
 mal. Lungs : adhesions at right apex ; small patch of broncho-pneumonia 
 in left lower lobe and another at right apex. Left apex puckered ; 
 contained several small calcareous nodules and cicatrices. The kidneys 
 were enlarged, each weighed 11 oz. Suprarenals : medulla deeply con- 
 
1 1 8 ETIOLOG Y AND ETIOLOGICAL RELA TIONS. 
 
 gested ; two small yellowish nodules, each about the size of a pea, in 
 the medulla of the right suprarenal. Spleen slightly enlarged, very, 
 soft, almost semifluid, congested ; weight 1 1 oz. Pancreas : weight 2^ 
 oz., somewhat small, firm to the touch. Liver : weight 5 lb., capsule 
 normal ; on the under surface of the right lobe a few small irregular 
 yellowish patches, each about the size of a pea. On section, scattered 
 throughout the right lobe irregularly, but most numerous in the outer 
 half of the lobe, were a number of small points of suppuration, varying 
 in size from a pea to a marble. About the middle of the outer half 
 of the right lobe the points of suppuration were clustered together so 
 as to form a large, irregularly-shaped abscess, about 4 in. in the trans- 
 verse and 2-3 in. in vertical diameter. The pus could be easily washed 
 away from the points of suppuration, leaving cavities bounded by a dis- 
 tinct limiting wall. In only a few places was there any zone of con- 
 gestion around the abscesses. The pus was of a greyish-yellow colour. 
 Under the microscope, pus cells and a few scattered granules of bile 
 pigment were seen. The gall bladder contained a small amount of pale 
 orange-coloured bile. The mucous membrane was congested, but no 
 ulceration could be detected. There was no obstruction of the cystic 
 duct, common bile duct, or hepatic duct. A number of the large bile 
 ducts were traced into the liver substance with a probe and slit open, 
 but no obstruction was met with ; the ducts were not dilated. They 
 did not contain any pus. Xo thrombosis in the portal vein, nor in any 
 of its larger branches. Stomach : no ulceration. Rectum : no ulceration, 
 nor any other abnormality detected. Bladder contained a large amount 
 of urine. Prostate normal. Spinal dura mater normal to the naked 
 eye. Anterior surface of the spinal cord and arachnoid normal. Arach- 
 noid of posterior surface of a yellowish turbid appearance in the dorsal 
 region. Only slight turbidity in the cervical and lumbar regions. Vessels 
 on the posterior aspect of the cord much distended. Only slightly dis- 
 tended on the anterior aspect. On section, the cord presented a normal 
 appearance. 
 
 Skull cap, dura mater, and blood sinuses normal. 
 
 Brain. — Membranes of each cortex presented a turbid, yellowish-white 
 appearance. Distinct suppuration along the course of the large vessels, 
 and at the sulci of the convolutions. At the anterior end of the first 
 left frontal convolution was a hasmorrhage about the size of a halfpenny. 
 It was irregular in shape, situated between the arachnoid and the cortex, 
 and passed down between the convolutions, but did not enter the brain 
 substance. The membranes at the upper surface of the cerebellum pre- 
 sented the same appearance as those of the cortex cerebri. Membranes 
 of the base of the brain, pons, and under surface of cerebellum of normal 
 appearance. The lateral ventricles contained a small amount of turbid 
 fluid. In the posterior horn of the left lateral ventricle Avas a small 
 
DIABETES AND DISEASES OE THE LIVER. itq 
 
 amount of pus. Vessels of velum interpositum and choroidal plexuses 
 much distended. In the fourth ventricle nothing abnormal detected. 
 
 On section, basal ganglia, pons, medulla, and cerebellum normal to 
 the naked eye. 
 
 Microscopical examination of the medulla and cord reveal purulent 
 meningitis, but no other changes. Microscopically the pancreas was 
 normal. 
 
 Pathological diagnosis. — Diabetes mellitus ; multiple abscesses in the 
 liver, spinal and cerebral meningitis. 
 
 It is difficult to say what was the relation of the liver 
 abscesses to the diabetes. A similar case is published by 
 Frerichs ( 39 ), and two are mentioned by Saundby ( 40 ). 
 
 Condition of the liver in a series of consecutive cases of diabetes 
 mellitus. — In discussing the relation of diabetes to liver diseases, 
 it will be well to consider next the conditions of this organ in a 
 series of consecutive cases. Taking for this purpose the records 
 of the Pathological Institute of Vienna, we meet with the follow- 
 ing results in 122 cases of diabetes (recorded by Seegen). 
 
 During the professorship of Eokitansky thirty cases were 
 examined : — 
 
 In 15 the liver was enlarged, firm, and hyperaemic. 
 ,,2 ,, small and ansemic. 
 
 ,, 1 medullary cancer of the liver was present. 
 ,, 1 tuberculosis ,, ,, 
 
 During the professorship of Kundrat ninety-two cases were 
 examined : — 
 
 In 9 the liver was fatty. 
 
 „ 7 parenchymatous degeneration was present. 
 
 ,, 8 there was fatty infiltration of the liver. 
 
 ,, 4 the liver was atrophic. 
 
 ,,4 ,, cirrhotic. 
 
 ,, 5 ,, congested. 
 
 „ 2 tuberculosis of the liver was present. 
 
 ,, 1 gall stones were present. 
 
 „ la fistula was present between the gall bladder and colon. 
 
 In the last twenty cases of diabetes in which I have seen 
 or made the autopsy, the condition of the liver has been as 
 follows : — 
 
 Liver enlarged in . . . .11 cases. 
 
 Normal in size in . . . . 5 ,, 
 
 Diminished in size in . . 4 ,, 
 
1 20 E TIOLOG Y AND ETIOLOGICAL RELA TIONS. 
 
 In 1 case multiple abscesses were present. 
 
 ,, there was marked alcoholic cirrhosis. 
 
 „ „ cirrhosis and fatty infiltration. 
 
 ,, the liver was congested and fatty. 
 
 In the rest, beyond variation in size, the only other change 
 detected was congestion, which was often present. 
 
 From the above lists it will be seen that no constant macro- 
 scopic pathological changes are met with in the liver in diabetes. 
 Also, microscopically, no definite or constant change has been 
 detected. 
 
 The most common abnormalities, as shown by the results 
 given above, are enlargement, hyperemia, fatty infiltration and 
 degeneration, and cirrhosis. Eecently attention has been drawn 
 by a number of authors to the association of a peculiar, rare 
 form of diabetes, — so-called " bronzed diabetes," — with pigment- 
 ary hypertrophic cirrhosis of the liver (this variety is described 
 on p. 308). Excluding the last pathological condition, it may be 
 pointed out that the above mentioned changes are very often 
 found unassociated with diabetes mellitus, and that none of 
 them are sufficiently constant to be regarded as playing any 
 part in the causation of the disease. 
 
 Gk'nard ( 41 ) states that he has found changes in the liver during 
 life, by examination of the abdomen, in 60 per cent, of diabetic 
 patients; in 34*5 per cent, hypertrophy was present. But it is 
 well to remember that he practises at Vichy, which is visited by 
 a large number of gouty and obese diabetic patients. 
 
 It has been pointed out on p. 110 that in some cases of 
 diabetes there is a history of marked alcoholism, and that it 
 appears probable that occasionally alcoholism plays a part as an 
 exciting cause. Triboulet ( 42 ) regards diabetes in these cases 
 as the result of a liver affection — a pre-cirrhotic change. 
 
 Glycogen in the liver of diabetic patients. — By means of a fine 
 trocar, Ehrlich ( 43 ) removed a few liver cells, during life, from 
 an alcoholic patient, four and a half to five and a half hours 
 after a meal rich in amylaceous substances. Microscopically, 
 glycogen was found in the cells in considerable quantity. In a 
 case of diabetes, the liver cells, removed in the same way, were 
 almost free from glycogen ; only in a few isolated cells was 
 glycogen found : but in another case of diabetes glycogen was 
 present in the cells in considerable quantity. 
 
 Kiilz ( 44 ) estimated the amount of glycogen in the liver of 
 
DIABETES AND DISEASES OF THE LIVER. 121 
 
 a diabetic patient, who had been restricted to a meat diet for a 
 long time. The patient had taken his last meal thirty-four 
 hours before death, and the autopsy was made twelve hours 
 after death. He found that the total amount of glycogen 
 present was 10 to 15 grms. The liver also contained a large 
 quantity of sugar, which had been transformed from the glycogen. 
 Hence the amount of glycogen during life must have been very 
 considerable. v. Mering( 45 ) found sugar and glycogen in 
 abundance in the liver of two diabetic patients who died 
 suddenly. In two other cases, in which sugar had disappeared 
 from the urine eighteen and twenty hours respectively before 
 death, neither glycogen nor sugar was found in the liver. 
 
 Whatever conclusions we might be inclined to draw from 
 physiological experiments, with respect to the relation of the 
 glycogenic or other functions of the liver to diabetes in man, it 
 is clear, from a consideration of the above facts, that at present 
 pathological anatomy alone does not furnish any clear evidence 
 that diabetes is related to hepatic changes. 
 
 D. The Kelation between Diabetes Mellitus and 
 Affections of the Neevous System. 
 
 It is well known that diabetes sometimes immediately 
 follows, and appears to be caused by, severe mental anxiety or 
 shock. Long ago, Thomas Willis attributed the disease to " sad- 
 ness or long sorrow " ; Prout attributed it, among other causes, 
 to mental anxiety or distress ; and in medical literature many 
 cases have been recorded which afford the strongest evidence of 
 the importance of mental shock, anxiety, etc., in the causation 
 of the disease. Seegen, Pavy, Frerichs, and Dickinson all regard 
 changes in the nervous system as the most important factor in 
 the causation of diabetes. Indications of gross lesion of the 
 nervous system (brain or spinal cord) are comparatively rare ; 
 but numerous cases are on record in which the disease appears 
 to have been due to some change in the nervous system, though 
 it may not be referable to any definite anatomical lesion. The 
 fact that diabetes and glycosuria can be produced experimentally 
 by injuries to the nervous system — floor of the fourth ventricle 
 and other parts — is also of great importance. Then again, it 
 is interesting to note that diabetes sometimes appears quite 
 suddenly, the patient being able to give the exact hour at which 
 
1 2 2 ETIOLOG Y AND ETIOLOGICAL RELA TIONS. 
 
 the symptoms commenced, (see p. 164) ; this sudden onset is very 
 suggestive of a lesion of the nervous system. 
 
 Diabetes following mental disturbances. — Numerous cases have 
 been recorded in which diabetes has followed fright, anger, or 
 depressing mental emotions, mental over-strain, mental anxiety 
 or worry owing to loss of money, loss of employment, etc., 
 mental anxiety and overwork associated with the nursing of a 
 sick relative, etc. 
 
 In 100 cases of diabetes (mostly patients at the Manchester 
 Eoyal Infirmary), a careful inquiry into the history revealed the 
 following facts with regard to mental disturbances : — 
 
 In eight cases the disease followed the mental anxiety, over- 
 work, and loss of rest, etc., connected with the nursing of a sick 
 relative — generally husband or wife — during a long and fatal 
 illness. 
 
 In four cases the diabetic symptoms followed great mental 
 worry and anxiety in connection with loss of money. 
 
 In two cases the disease followed mental anxiety from other 
 causes. 
 
 In one case diabetes followed mental anxiety and worry 
 connected with loss of employment for a long period. 
 
 In another case the disease followed great mental strain and 
 overwork in connection with the preparation for the final B.A. 
 examination of the London University. The patient, who had 
 failed at the examination previously, decided to devote almost 
 the whole of his time to reading (except that required for eating 
 and sleeping). This he did for some months, until diabetes of 
 a severe form developed. 
 
 In two cases the disease was attributed to fright. 
 
 The following are brief notes of one of these cases : — 
 
 Ellen B., cet. 20 (under the care of Dr. Steell at the 
 Manchester Eoyal Infirmary), enjoyed good health until January 
 1892, when she suffered severely from mental anxiety and 
 fright. She was left in charge of her sister's baby, but in order 
 to attend, for a few minutes, to some household work, she placed 
 the baby on the table, and went a few yards away. The child 
 fell off the table, the head came in contact with the ground, 
 and a scalp wound was produced. The patient was greatly 
 frightened ; at the time she feared the skull was fractured, and 
 was in the greatest state of anxiety. A medical man was sent 
 for at once, and stated that the patient Ellen B. was suffering 
 
DIABETES AND THE NERVOUS SYSTEM. 123 
 
 much more — owing to the mental shock — than the baby. In 
 a short time the child recovered completely, but the mental 
 distress and anxiety from which Ellen B. suffered were so great 
 that she was never able to do her usual work again. Prior to the 
 fright she had been in good health, and had followed her employ- 
 ment (that of a dressmaker) regularly. For many days after the 
 accident she remained in a nervous and excited condition, and 
 was never able to follow her employment again. Thirst and 
 diuresis were noted four weeks after the fright. The urine was 
 examined, and diabetes diagnosed. A severe form of the disease 
 developed, and terminated fatally in about seventeen months. 
 (For further notes, see p. 134.) 
 
 In eighteen out of 100 cases analysed, it appeared probable 
 that mental disturbances played some part as an exciting cause, 
 though no doubt there was generally some additional etiological 
 factor. 
 
 Dickinson ( 50 ) reports a case similar to that of Ellen B. " A 
 child fell from a third-floor window, and was smashed upon the 
 pavement, to all appearance, hopelessly. But the accident was 
 more fatal to its mother than itself. The child survived. The 
 mother, who was abruptly made aware of the accident, never 
 recovered from the shock. For three weeks she, to use her own 
 words, could neither eat nor sleep. Within two months she 
 became much emaciated, was consumed with thirst, and was 
 passing water in great quantities, which incrusted upon and 
 stiffened any garment it touched. She died of diabetes within 
 ten months of the occurrence upon which it had succeeded." 
 
 Garrod ( 51 ) relates the following case : — " Two gentlemen 
 fought a duel in Holland ; after the first had fired he remained 
 for some time in a state of suspense, from his adversary's pistol 
 once or twice missing fire. He was uninjured, but, a day or so 
 after, became diabetic." 
 
 Seegen ( 52 ) mentions the following cases : — Baron K.. an 
 officer, aet. 22, was present at a duel. One of the combatants, a 
 friend of the patient, was killed on the spot. From that day 
 Baron K., who had previously been a cheerful young man, 
 became markedly unwell and depressed ; he soon began to lose 
 flesh ; two months later it was found that he was suffering from 
 diabetes of a severe form. 
 
 Seegen refers to another case, in which a previously healthy 
 man received a violent mental shock; next day marked thirst 
 
i2 4 E TIOLOG Y AND ETIOL O GICAL HE LA TIOA^S. 
 
 and diuresis commenced, and all the symptoms of diabetes 
 followed. 
 
 In another case the symptoms followed a railway accident. 
 Though the patient was uninjured, he was greatly agitated, and 
 trembled for twenty-four hours. A few minutes after the acci- 
 dent he became very thirsty, and these symptoms continued 
 for some weeks. He consulted a medical man, and a large 
 amount of sugar was found in the urine. 
 
 Dickinson also cites the following case, communicated by 
 Dr. Herman Weber : — A. C, a merchant, was for two months, in 
 1857, under constant excessive anxiety on account of business 
 troubles. After many sleepless nights he became delirious, or 
 rather insane, — for there was no pyrexia, — and almost suddenly 
 began to pass urine very frequently and in large quantities. It 
 was found to contain abundance of sugar. The daily quantity 
 of urine varied from 7 to 9 pints, and the specific gravity from 
 1036 to 1044. His business passed safely through the crisis. 
 Within three weeks the urine was free from sugar, and remained 
 so, as ascertained by several annual examinations, until the year 
 1866. During the mercantile crisis of that year, A. C. again 
 became restless, sleepless, and diabetic. After the crisis he 
 again recovered completely, and remained well, with perhaps a 
 single exception, up to 1870. At the outbreak of the Franco- 
 Prussian war in this year, he became much excited, and died of 
 apoplexy. 
 
 Pavy ( 53 ) relates the case of a gentleman who became diabetic 
 under the influence of mental excitement, and ceased to be so 
 when retirement from town and professional duties had enabled 
 his mind to recover its equipoise. 
 
 Purdy ( 5i ) reports a fatal case due to over-anxiety in con- 
 ducting extensive transactions on the produce exchange. The 
 patient accumulated a large fortune at the expense of contract- 
 ing diabetes, which proved fatal in less than twelve months. 
 He also reports another case brought on by business worry ; 
 recovery occurred after a complete rest. 
 
 Frerichs ( 55 ) mentions a case in which the symptoms followed 
 great loss of money. By removal from business excitement, 
 and under treatment with opium, every trace of sugar disap- 
 peared in three weeks. Several months later, after a second 
 business loss, diabetes again developed. In course of time the 
 patient improved, but a third business misfortune caused the 
 
DIABE TES AND THE NER VO US S YSTEM. 1 2 5 
 
 disease to return. He also mentions the case of a man who 
 became diabetic immediately after discovering that his wife had 
 been unfaithful. 
 
 Eoberts ( 56 ) states that in one of his patients the disease 
 followed distress of mind, caused by an unjust suspicion of 
 theft ; in another, it followed the anxiety occasioned by the 
 burning down of the patient's place of business ; in a third, it 
 was attributed to anxiety attendant on a Chancery suit. 
 
 Eayer ( 57 ) mentions a case of diabetes coining on after a 
 violent lit of anger. 
 
 Diabetes and, pathological changes in the nervous system. — 
 Though there are so many points in favour of the connection 
 of diabetes with some changes in the nervous system, it must be 
 confessed that in a large number of cases pathological examination 
 does not reveal any marked or definite abnormality in the brain 
 or spinal cord. In a small portion of cases, however, changes 
 have been met with, which are of great interest. 
 
 In discussing the relation of diabetes to pathological changes 
 in the nervous system, it will be well to consider — (a) the 
 condition of the brain in a series of consecutive cases of diabetes, 
 and (5) the more important pathological changes which have 
 been met with in the brain. 
 
 (a) Condition of the nervous system in a series of consecutive 
 cases of diabetes mellitus. — Seegen ( 58 ) records the condition of the 
 brain in 122 cases of diabetes mellitus, in which a post-mortem 
 examination was made in the pathological department of the 
 Allgemeine Krankenhaus, Vienna. Thirty examinations were 
 made during the time Eokitansky was professor, and ninety- 
 two during Kundrat's professorship. 
 
 In Eokitansky 's thirty cases no changes of any importance 
 were found. Cerebral oedema was present in three cases, oedema 
 of the brain and chronic hydrocephalus in one case ; in one case 
 in which there was general tuberculosis, tubercular meningitis 
 was present. 
 
 In fifty-four of Kundrat's ninety-two cases the brain was not 
 quite normal. The following were the changes recorded : — 
 
 Slight General Changes. 
 
 Cases. 
 Anaemia . . . . . . . . . - 5 
 
 Marked hyperemia ....... 8 
 
 (Edema of meninges . . . . . .4 
 
i26 E TIOLOG Y AND ETIOLOGICAL RE LA TIONS. 
 
 Cases. 
 (Edema of the brain . . . . . . .10 
 
 ,, of the brain and meninges .... 8 
 
 Atrophy of the brain (one case with hydrocephalus) . 8 
 
 More Localised Lesions. 
 
 Dilatation of the fourth ventricle with cerebral hyperemia 3 
 
 Posterior horns of ventricle closed by a cicatrix ; ependyma 
 of septum granular . . . . . .1 
 
 Cerebral haemorrhages — 
 
 (1) Inter meningeal haemorrhage over frontal and 
 temporal lobes, optic chiasma, pons, and into the ven- 
 tricles ; softening of gyrus fornicatus. (2) Haemorrhages 
 into the floor of fourth ventricle. (3) Haemorrhages 
 into lenticular nucleus breaking into ventricles (lateral 
 and fourth). (4) Haemorrhages into optic thalamus . 4 
 
 Depression in the apex of the frontal lobe and at the 
 top of the occipital lobes ; cicatricial condition, with 
 calcareous plates, in the arachnoid over the frontal lobes 1 
 
 Dura mater thick, and covered with purulent exudation 
 at the vertex, corresponding to a necrotic patch in the 
 skull ; region of calamus scriptorius of fourth ventricle 
 much depressed . ...... 1 
 
 Cysticercus the size of a cherry in the telachoroida, at the 
 inferior and lower part of the fourth ventricle . . 1 
 
 Total, 54 
 
 In eleven only of the 122 diabetic cases, i.e. 9 per cent., 
 was there any naked-eye change of much importance. In many 
 of these, the significance of the changes with respect to the 
 causation of the disease was very doubtful. 
 
 The spinal cord presented grey degeneration in the lateral 
 and posterior columns in one of Kundrat's cases (how many were 
 examined is not stated). 
 
 To consider the subject from another standpoint, in 485 
 cases of brain tumour — collected from literature — Bernardt ( 59 ) 
 found that sugar was present in the urine in five cases only. 
 In twenty-one cases of tumour of the medulla, sugar was found 
 in the urine in one case only. In fifteen cases of tumour of the 
 hypophysis, sugar was found in the urine twice ; in ninety cases 
 of cerebellar tumour, once; and in 124 cases of tumour of the 
 cerebral hemispheres, once. 
 
CHANGES IN THE NERVOUS SYSTEM. 
 
 127 
 
 (b) The more important changes which have been found in 
 the nervous system. — A number of cases have been recorded 
 in medical literature in which well-marked and important 
 changes have been met with in the nervous system, but the 
 proportion of such cases, to those in which the nervous system 
 presents no macroscopic or microscopic changes, is probably 
 small. 
 
 The following is a tabulated group of thirty-four cases, 
 collected from medical literature, in which definite changes of 
 some importance have been found in the nervous system. In 
 some of these possibly the association of diabetes mellitus with 
 the nervous lesions has been accidental, but in many there can 
 be little doubt that the lesions played a very important part in 
 the causation of the disease : — 
 
 Tumours of the Medulla. 
 
 1. Cystic sarcoma, right half of 
 medulla. 
 
 2. Tumour of each pyramid of the 
 
 medulla close to the pons. 
 
 3. Tubercle of the medulla be- 
 
 tween the posterior border 
 of the olivary body and the 
 origin of the first cervical 
 
 Dompeling — Nederl. Arch. v. Gen- 
 ees-en NatuurL, Utrecht, 1868 
 (quoted by Seegen, " Der Dia- 
 betes mellitus," Berlin, 1893, 
 S. 208 and 425). 
 
 Frerichs — "Ueber den Diabetes 
 mellitus," Berlin, 1884, S. 202. 
 
 De Jonge — Arch, f. Psychiat, 
 Berlin, 1882, Bd. xiii. S. 663. 
 
 Tumours in the Floor of the Fourth Ventricle. 
 
 4. Tumour in region of choroidal 
 
 plexus of fourth ventricle. 
 
 5. Colloid tumour in fourth 
 
 ventricle. 
 
 6. Two tumours of choroidal 
 
 plexus. 
 
 7. Tumour in fourth ventricle at 
 
 the calamus scriptorius ; 
 glioma in floor of fourth 
 ventricle. 
 
 Recklinghausen — Virclwios 
 
 Archiv, Bd. xxx. 
 Perroton — These de Paris, 1859. 
 
 Lionville — " Verrons etudes sur 
 les tumeurs du ventricle quarte/' 
 These de Paris, 1874. 
 
 Reimer — Jahrb.f. Kinderh., 1876. 
 s. 306. 
 
128 E TIOL OGY AND E TIOL O GICAL RELA T10NS. 
 
 Other Changes in the Floor of the Fourth Ventricle. 
 
 8. Patch of softening, floor of 
 
 fourth ventricle. 
 
 9. Softening due to fatty de- 
 
 generation, floor of fourth 
 ventricle. 
 
 10. Similar case. 
 
 11. Cysticercus in fourth ventricle. 
 
 12. Cysticercus in fourth ventricle. 
 
 13. Disappearance of grey sub- 
 
 stance and of nerve cells 
 from floor of fourth ventricle. 
 
 14. Sclerosis in floor of fourth 
 
 ventricle. 
 
 15. Patches of disseminated 
 
 sclerosis in the floor of the 
 fourth ventricle ; multiple 
 sclerosis of cord and brain. 
 
 16. Disseminated sclerosis, with 
 
 lesion in the floor of the 
 fourth ventricle. 
 
 1 7. Patch of disseminated sclerosis 
 
 in floor of fourth ventricle. 
 
 18. Cerebro - spinal meningitis ; 
 
 purulent fluid in fourth 
 ventricle. 
 
 19. Old and recent haemorrhages 
 
 in the floor of fourth ventricle 
 and medulla. 
 
 20. Haemorrhage, floor of fourth 
 
 ventricle. 
 
 21. Patch of sclerosis in floor of 
 
 fourth ventricle, left side ; 
 connective tissue rich in 
 nuclei ; traversed by arteries 
 with very thick walls. 
 
 Richardson — Med. Times and 
 
 Gaz., London, 1866. 
 Luys — Gaz. vied, de Paris, 1860, 
 
 No. 24, p. 384. 
 
 Luys — Ibid. 
 
 Ivan, Michael — Deutsches Arch, 
 f. Tdiii. Med., Leipzig, Ed. 
 xliv. 
 
 Case under the care of Riess 
 of Berlin, mentioned by Osler, 
 " Principles and Practice of 
 Medicine," Edinburgh and Lon- 
 don, 1895, 2nd edition, p. 
 320. 
 
 Zenker — Quoted by Weichsel- 
 baum, Wien. med. Wehnsclir., 
 1881, No. 32. 
 
 Frerichs — Loc. eit., s. 195. 
 
 Weichselbaum — Wien. med. 
 Wchnschr., 1881, No. 32. 
 
 Edwards — Rev. de med., Paris, 
 1886, p. 703. 
 
 Richardiere — Ibid., Paris, 1886, 
 
 p. 623. 
 Frerichs — Loc. eit., S. 198. 
 
 Frerichs — Loc. eit., S. 203. 
 
 Dutrait — Quoted by Weichsel- 
 baum, loc. eit., 1884, No. 32. 
 
 Thiroloix — Gaz. de med. et chir., 
 16th April 1892. 
 
TUMOURS, LESIONS, AND CHANGES. 
 
 129 
 
 Lesion about the Base of the Brain. 
 
 22. Calcareous tumour compress- 
 
 ing the under surface of 
 pons; also abscess in posterior 
 cerebral lobes. 
 
 23. Sarcoma of the pituitary body. 
 
 (Two similar cases, coming 
 under my observation in 
 Manchester, are recorded on 
 pp. 137-38.) 
 
 24. Tumour at the base of the 
 
 brain, limited in front by 
 the anterior frontal convolu- 
 tion, behind by the pons. 
 Pia mater in the fourth 
 ventricle thickened. 
 
 25. Sclerosis of arteries of the base 
 
 of the brain ; old thrombus 
 in the basilar artery, soften- 
 ing in the pons, left cms, 
 and left cerebellar cortex. 
 
 Richardson — Med. Times and 
 Gas., London, 8th March 1862. 
 
 Eosenthal — " Klinik der Nerven- 
 krankheiten," Stuttgart, 1875. 
 
 Grossmann — Berl. Idin. Wchnschr., 
 1879, No. 10. 
 
 Frerichs — Loc. cit., S. 200. 
 
 26. Patch of softening in nucleus 
 
 dentatus of left cerebellar 
 hemisphere. 
 
 27. Cysticercus in cerebellum. 
 
 Cerebellum. 
 
 Mosler — Deutsches Arch. f. Jdin. 
 Med., Leipzig, Bd. xv. S. 229. 
 
 Prerichs — Loc. cit., 8. 193. 
 Changes in Cerebral Hemispheres. 
 
 28. Tumour, size of a hazel-nut, in 
 
 the posterior part of the right 
 temporal lobe. 
 
 29. Pachymeningitis in left occi- 
 
 pital lobe, with calcareous 
 tumour the size of a pea. 
 
 Frerichs — Loc. cit., S. 135. 
 
 Prerichs — Loc. cit. 
 
 30. Tumour compressing the spinal 
 cord in the cervical region. 
 
 31. Haemorrhage and softening in 
 the spinal cord in the cervical 
 and upper dorsal region. 
 9 
 
 Lesions of the Spinal Cord. 
 
 Shingleton Smith — Brit. Med. 
 
 Journ., London, 1883, vol. i. 
 
 p. 657. 
 Silver and Irvine — Trans. Path. 
 
 Soc. London, vol. xxix. 
 
no E TIOL OGY AND E TIOL O Gl CAL RE LA TIONS. 
 
 32. 
 
 Atrophy of the anterior horns 
 of the upper part of the cord 
 and lower part of the medulla 
 (from the fifth cervical nerve 
 up to the pyramidal crossing). 
 
 Frerichs — -hoc. tit., S. 136. 
 
 Vagus Nerve. 
 
 33. Calcareous tumour, the size of 
 
 a hazel-nut, pressing on the 
 right vagus nerve. 
 
 34. Tumour on the right vagus 
 
 nerve, close to medulla. 
 
 Harlet — Quoted by Sir William 
 Eoberts, " Eenal and Urinary 
 Diseases,' - " London, 1885, p. 279. 
 
 Frerichs — Loc. tit., S. 91. 
 
 Summary of Cases Tabulated. 
 
 Tumours of the medulla oblongata 
 
 ,, floor of the fourth ventricle . 
 
 Other changes in the floor of the fourth ventricle 
 Lesions at tbe base of the brain 
 Tumour of the pituitary body 
 Lesions of cerebellum 
 
 ,, cerebral hemispheres 
 
 ,, the spinal cord 
 Tumour compressing the right vagus 
 
 Cases. 
 3 
 4 
 14 
 3 
 1 
 2 
 2 
 
 3 
 
 2 
 
 34 
 
 Microscopical changes. — Dickinson ( 60 ) has drawn attention to 
 the presence of minute excavations, chiefly around the arteries, 
 in the brains of diabetic patients. The importance of these 
 changes was much discussed many years ago. In some cases of 
 diabetes they have not been found by other observers. Then, 
 again, perivascular excavations have been sometimes found in 
 other diseases, and it is now generally believed that they have 
 no relation to diabetes, but are due to the effects of hardening. 
 
 Frerichs ( 61 ) devoted much attention to the microscopical 
 examination of the medulla in diabetes. He has described a 
 marked dilatation of the fine vessels of the medulla, and this 
 change he regards as the most important and constant in the 
 nervous system in diabetes. Often the dilatation of the vessels is 
 accompanied by small haemorrhages, partly old and partly recent. 
 To the dilatation of the perivascular spaces and thickening of 
 the ependyma, Frerichs did not attach any importance, but he 
 
MICROSCOPICAL CHANGES. 131 
 
 has drawn attention to the great dilatation of a vein found, as 
 a rule, just at the outer border of the vagus nucleus. 
 
 Apart from these vascular changes, no other constant or 
 well-defined changes have been met with on microscopical exam- 
 ination of the medulla. The nerve cells of the vagus and other 
 nuclei have not been found to present any abnormality. 
 
 Saundby( 62 ) mentions that he has found the capillaries of 
 the vagus nucleus in one case abnormally numerous and full 
 of blood. In other cases examined, no changes could be detected 
 in the medulla, basal ganglia, or cerebral cortex. 
 
 The two cases above referred to, in which tumours were 
 found pressing on the vagus nerve, are interesting in connection 
 with the experiments of Arthaucl and Butte, on the production 
 of diabetes by chronic irritation of the vagus nerve (see p. 66). 
 
 Slight changes in the posterior columns of the spinal cord 
 have been found in a few cases, but they have probably been 
 secondary in nature, and will be referred to under the complica- 
 tion of diabetes on p. 66. 
 
 The sympathetic nerves and ganglia. — From time to time 
 changes in the abdominal sympathetic nerves and ganglia 
 (chiefly in the cceliac plexus and the semilunar ganglia) have 
 been described by various authors — by Klebs and Munk ( 63 ), 
 Poniklo ( 64 ), Cavazzani ( 65 ), Lubrinoff ( 66 ), Saundby ( 67 ), Hale 
 White ( 68 ), and others ( 69 ). 
 
 The changes have consisted chiefly of atrophy, sclerosis, or 
 cell infiltration of the ganglia, of increased vascularity, and 
 excess of connective tissue in the nerves ; sometimes the 
 ganglia have been greatly enlarged. But in other cases the 
 histological examination of these nerves and ganglia has only 
 yielded negative results. Then, again, marked changes, similar 
 to those met with in diabetic patients, have been found in cases 
 which had not presented any symptoms of diabetes during life. 
 
 As already mentioned, the cceliac plexus has been extirpated 
 by Lustig and Peiper without producing a permanent diabetes. 
 
 Windle( 70 ) has collected the results of the pathological 
 examination of the sympathetic nerves and ganglia in seventeen 
 cases ; in eight of these the results were negative. 
 
 The relation of changes in the sympathetic nervous system, 
 to a number of pathological conditions has been carefully studied 
 by Hale White, who has made histological examinations of the 
 sympathetic ganglia in various diseases. Hale White has 
 
1 3 2 E TIOL G Y AND E TIOL O GICAL RE LA TIONS. 
 
 examined the superior cervical ganglia of seven patients who 
 died of diabetes, but no single condition was " ever present that 
 is not found, and even more marked, in many patients who have 
 never had glycosuria." He has also examined the semilunar 
 ganglia. Sometimes the ganglia were well developed and 
 healthy looking, and the changes found in the others were only 
 such as are frequently observed when there have been no signs 
 of diabetes during life. Hence he concluded that the ganglia 
 in the cases he has examined showed no changes to which the 
 disease could be attributed. 
 
 Diabetes and diseases of the nervous system. — In various well- 
 defined diseases of the nervous system, glycosuria is occasionally 
 met with, and sometimes, though very rarely, true diabetes ; but 
 the latter is generally of a mild form. 
 
 Thus diabetes has been met with occasionally, though very 
 rarely, in association with disseminated sclerosis, locomotor 
 ataxia ( 71) 72 ' 73 ), chronic anterior poliomyelitis, Graves' disease, etc. 
 In Graves' disease, it has been already pointed out (see p. 87) that 
 alimentary glycosuria is produced more readily than in healthy 
 persons. Occasionally Graves' disease is complicated with 
 glycosuria ; much more rarely with true diabetes. A well-marked 
 example of the latter association has recently been recorded 
 by Bettmann ( 74 ). Occasionally diabetes is associated with 
 symptoms pointing to a cerebral tumour or other gross cerebral 
 lesion ; but in such cases the diabetes is generally of a mild 
 form. 
 
 Pcrsoncd observations. — A number of cases of well-marked 
 temporary glycosuria, associated with various nervous lesions, 
 have come under my observation, but only a very few cases 
 of true diabetes associated with evidences of gross lesion of 
 the brain. In not more than 3 per cent, of cases of diabetes 
 observed in Manchester, was there any evidence of gross 
 lesion of the nervous system (i.e. excluding nervous complica- 
 tions due to the disease). The following are notes on an 
 interesting case of diabetes associated with symptoms of gross 
 lesion of the brain which came under my observation at the 
 Manchester Eoyal Infirmary : — 
 
 Marked cdaxia of sudden onset ; %>avcdysis of right internal rectus ; 
 tremor of right arm; onset of diabetes seven lueeks after commencement 
 of symptoms. — J. W., set. 48, was admitted under the care of Dr. Steell 
 on 7th May 1890. Patient was quite well until 26th April 1890. 
 
PERSONAL OBSERVATIONS. 133 
 
 On the morning of that day, when going to his work, he suddenly 
 became ataxic and fell to the ground. There was no loss of conscious- 
 ness, no paralysis of limbs. He has been unable to walk or stand since, 
 owing to ataxia. History of alcoholism; no syphilitic history. On 
 admission, patient is unable to stand alone ; when well supported, he is 
 able to advance a few steps, and throws forward his right leg in a jerky, 
 irregular manner. Marked ataxia. No paralysis of limbs ; no rigidity ; 
 knee-jerks present; no ankle-clonus. Eight palpebral fissure smaller 
 than left ; apparent ptosis of right eye ; paralysis of right internal 
 rectus; left pupil larger than the right. Slight nystagmus; slight 
 tremor of the right arm even when the limb is supported, but this is 
 more marked on voluntary movement. Optic discs normal. Old 
 perforation of right tympanic membrane. Temperature normal. Urine 
 1015, acid, no albumin, no sugar, quantity normal. 
 
 \Wi June. — Amount of urine increased to 92 oz. daily. 
 
 15th June. — Urine pale, clear, 1024, acid, contains a large amount 
 of sugar. 
 
 23rcZ June. — Great thirst, voracious appetite. In the first three 
 weeks during which the diabetic symptoms were noted, the amount of 
 urine varied from 160 to 190 oz.; and the sugar from 3200 to 4200 grs. 
 daily. The excretion of urine and sugar were both diminished by 
 restricted diet. A few months later several large carbuncles developed 
 on the gluteal region, but in course of time these, and the paralysis of 
 the right internal rectus, the ptosis and nystagmus, disappeared. The 
 diabetic symptoms, the tremor of the right hand, and the ataxia 
 persisted. 
 
 In two cases diabetes was associated with typical symptoms 
 of acromegaly, and post-mortem examination revealed a tumour 
 of the pituitary body. These cases will be described subsequently. 
 
 I have only met with a few cases in which post-mortem 
 examination has revealed any naked-eye lesion of the brain or 
 nervous system. In fourteen cases in which I have recently had 
 the opportunity of making a careful examination of the brain 
 post-mortem, the results have been as follows : — 
 
 In nine cases the medulla oblongata and other parts of the 
 brain appeared normal to the naked eye. 
 
 In five cases naked eye changes were found. 
 
 In two of these cases a tumour (round-celled sarcoma) of 
 tin; pituitary body was present, and during life there had been 
 typical symptoms of acromegaly. In these two cases the 
 medulla appeared normal. 
 
 In one case there was well-marked cerebro-spinal meningitis, 
 
i34 E TIOL OGY AND E TIOL O GICAL RE LA TIOJVS. 
 
 and multiple abscesses in the liver, but sections of the medulla 
 showed no other change beyond the inflamed meninges (see 
 p. 117). 
 
 In another case, a cyst about the size of a small pigeon's 
 egg was found in the right lateral hemisphere of the cerebellum 
 on its under surface, close to the pons and anterior part of the 
 medulla. Clinically, the case had been one of severe diabetes, 
 finally complicated by phthisis. Death occurred from coma. 
 During life there were no indications of brain tumour or other 
 cerebral lesion. The cyst had a thin smooth wall, and the 
 cerebellum was excavated by it. There was no indication of 
 any tumour growth in connection with the cyst. During life 
 the cyst would probably have pressed upon the nerves arising 
 from the right side of the pons and medulla. 
 
 In one case, described below, a minute haeruorrhagic patch 
 could be detected by the naked eye in the left vagus nucleus 
 after the medulla had been hardened in Muller's fluid. 
 
 Microscojncal examination of the medulla. — In ten cases of 
 diabetes mellitus I have made a microscopical examination of 
 the medulla, especially in the region of Claude Bernard's 
 " diabetic centre." In all of these cases the medulla was 
 hardened in Muller's fluid and embedded in celloidin for section 
 cutting. The sections were stained with aniline blue-black, 
 logwood, and according to the methods of Pal or Weigert. 
 
 In the following case distinct changes were found in the 
 vagus nucleus : — 
 
 E. B., set. 20. The clinical history of this case is recorded on p. 122. 
 The symptoms followed a severe fright, which apparently was the ex- 
 citing cause of the disease. 
 
 On making a transverse section through the medulla (after 
 hardening in Muller's fluid) at the region of auditory striae, a small 
 dark red patch like a small haemorrhage, was seen, just in the region 
 of the left vagus and glosso-pharyngeal nuclei. It extended one-eighth 
 of an inch above and one-eighth of an inch below the point of 
 section. Altogether, therefore, its antero - posterior extent was a 
 quarter of an inch. In transverse section it appeared as a very small 
 dark red point. In sections examined under a low power of the micro- 
 scope, it was seen that the haemorrhagic patch extended at some parts 
 almost to the surface of the floor of the fourth ventricle, whilst at other 
 parts it was some distance below it, and was situated near the centre 
 of the left vagus nucleus. In some sections there were two haemorrhagic 
 
MICROSCOPICAL EXAMINATION. 135 
 
 patches in the region, of the vagus nucleus. In one section there 
 were a number of minute haemorrhagic patches, just around the vessels, 
 in the region of the left vagus nucleus. Under a high power it was seen 
 that the haemorrhagic patch consisted of red blood corpuscles, and the 
 
 Fig. 6. — Hemorrhagic patch (shaded) in vagus nucleus in the 
 medulla; case E. B., under the care of Dr. Steell (low 
 power of microscope). 
 
 margins of the patch were well defined and regular. In fact it appeared 
 as if the blood corpuscles were contained in well-defined spaces or 
 capillary vessels. The patch resembled a capillary angioma situated in 
 
 Fig. 7. — Another section from the medulla of the same case (low 
 power of microscope). 
 
 the vagus nucleus, though distinct vessel walls could not be detected 
 (see Figs. 6. 7, and 8). 
 
 The blood vessels of both vagus nuclei were dilated and full of red 
 corpuscles. The perivascular spaces were also dilated, and in many places 
 in the left vagus nucleus contained red blood corpuscles. The nerve 
 
136 ETIOLOGY AND ETIOLOGICAL RELATIONS. 
 
 cells of the vagus and other nuclei in the medulla, the nerve fibre and 
 other structures, were of normal appearance. The changes in the region 
 of the vagus nucleus appeared to indicate a condition of localised in- 
 
 - 
 
 4 * 
 
 ~4 
 
 . Fig. 8. — Hemorrhagic patch in vagus nucleus of medulla ; case E. B. (high 
 power of microscope, Zeiss, D.). 
 
 creased vascularity, and may in some way have been connected with 
 the causation of the disease. 
 
 In a second case, J. D., the clinical history of which is 
 recorded on p. 243, small haemorrhagic patches, similar to those 
 in the case just reported, only much smaller, were found in 
 the vagus nuclei. 
 
 In a third case, M'M., the vessels of the vagus nuclei were 
 dilated, and minute haemorrhages, mostly perivascular, were found 
 in one vagus nucleus. In two other cases the vessels of the 
 vagus nuclei were dilated, but no haemorrhages were found. 
 
 In four cases, sections of the medulla were normal. In one 
 case inflammation' of the meninges was discovered (case of 
 cerebro-spinal meningitis, recorded on p. 117). 
 
 Diabetes and acromegaly. — Diabetes has been associated with 
 symptoms of acromegaly in two cases in which I have made a 
 pathological examination of the brain. In both cases a tumour 
 of the pituitary body was present. The optic chiasma, and 
 
DIABETES AND A CROMEGAL Y. 137 
 
 afterwards the optic nerves, had been compressed by the. tumour 
 growth in each case. The following are the notes : — 
 
 1. Acromegaly; double optic atrophy; diabetes mellitus ; tumour 
 of the pituitary body. — Mary W., set. 36. Seven years before the 
 patient came under ray observation, she had noticed that the hands 
 were becoming large. She was seen by Dr. Hill Griffith on account 
 of impaired vision three years later. At that time there was bi-temporal 
 hemianopsia, and Dr. Griffith recognised the case as one of acromegaly. 
 The hands and feet became very large, the lips prominent and thick ; 
 the nose became enlarged, and the features altered very much. The 
 patient was troubled with frequent pain in the head, chiefly in the fore- 
 head, at the early part of the illness, but there was no vomiting. Her 
 vision gradually became worse, and for two years before she came under 
 my observation she had been quite blind. For nine months she had 
 suffered from thirst and diuresis, and had been under the care of my 
 friend Dr. E. Somers, who had found a large quantity of sugar in the 
 urine. After the onset of diabetic symptoms there was considerable 
 wasting. 
 
 During the last six days of her life the patient was under the care 
 of Dr. Dreschfeld at the Manchester Royal Infirmary, where, as medical 
 registrar, I had the opportunity of carefully examining her. 
 
 The hands and feet were still enlarged, in spite of considerable 
 wasting, which had been produced by the diabetes. The jaw, nose, 
 and lips were enlarged, and the typical symptoms of acromegaly were 
 present. The patient was quite blind, and ophthalmoscopic examination 
 revealed double optic atrophy (primary). There was no paralysis of 
 facial or ocular muscles. The knee-jerks were absent. Examination 
 of the chest and abdomen revealed nothing of importance. The urine 
 had a specific gravity of 1040; it contained a large amount of sugar, 
 but no albumin on admission (2nd May 1895). There was a marked 
 brownish red reaction with perchloride of iron, and also a distinct 
 reaction for acetone (Legal's test). 
 
 The daily quantity of urine was as follows : — 
 
 3rd May. — 76 oz. 
 
 ith May. — 164 oz. ; total quantity of sugar excreted in twenty-four 
 hours = 4960 grs. 
 
 5th May.— 160 oz. 
 
 During the night of 6th May the patient became very restless. On 
 7th May the pulse was rapid, 140 per minute, and very feeble; the 
 respirations deep, 32 per minute. The patient was semi-comatose. The 
 urine contained a small amount of albumin ; it gave a distinct brownish- 
 red coloration with perchloride of iron, and contained numerous hyaline 
 and granular casts. Casts were not present in the urine on 4th, 5th, 
 
1 3 8 E TIOL OGY AND E TIOL O GICAL RE LA TIONS. 
 
 and 6 th May. Death occurred with all the symptoms of diabetic coma 
 on 7 th May. 
 
 At the autopsy there were the characteristic changes in the limbs — 
 enlargement of the hands, feet, etc., but no abnormalities of importance 
 were detected in the thorax or abdomen. The pancreas weighed 2 oz., 
 and was normal macroscopically and microscopically. On examination 
 of the brain, a tumour of the pituitary body about the size of a pigeon's 
 egg Avas found. The under surface of the growth was smooth, and the 
 sphenoid bone was not invaded by it. The optic chiasma and optic 
 nerves were embedded in the growth. Microscopically, the tumour 
 was a round-celled sarcoma. The medulla and other parts of the brain 
 were normal. 
 
 2. Acromegaly; double optic atrophy; diabetes mellitus ; tumour of 
 the pituitary body. — A. H., set. 23, was admitted under the care of 
 Dr. Boss to the Manchester Infirmary on 16th November 1890. 1 
 History of headache and gradual enlargement of the hands and feet 
 — symptoms of acromegaly. Temporal hemianopsia of the right field 
 of vision ; in the left field, vision entirely lost, with the exception 
 of a very small area near, but a little to the temporal side of, the 
 centre of the field. Urine 1035, no albumin, no sugar, loaded with 
 urates. After remaining in the hospital for several weeks, the specific 
 gravity of the urine rose to 1045, and it was found to contain sugar by 
 Fehling's test, by the phenylhydrazin test, and also by fermentation. 
 The amount of sugar at first was 14 grs. to the ounce; the amount of 
 urine was then normal. The sugar gradually increased. The patient 
 left the hospital for several months, and then returned on account of an 
 abscess which had developed in the left breast. She was admitted on 
 the surgical side under the care of Mr. Southam. The patient com- 
 plained of thirst, and on examination of the urine a large amount of 
 sugar was found. The abscess was opened and treated antiseptically, 
 and the patient put on diabetic diet. The thirst, however, increased, 
 and five days later the patient died of diabetic coma. The temperature 
 remained practically normal. At the post-mortem examination a round- 
 celled sarcoma of the pituitary body was found, which pressed on the 
 optic chiasma. The left optic nerve was embedded in the growth. The 
 rest of the brain was normal. The thyroid was a little enlarged, and 
 contained a colloid cyst. There Avere changes in the pancreas, which 
 will be afterwards described (see p. 146). Liver enlarged, weight 5 lb. 
 2 oz. ; signs of fatty degeneration. 
 
 Glycosuria has been observed in other cases of acromegaly 
 which have been recorded in medical literature ; sometimes 
 
 1 This patient's symptoms were the subject of a clinical lecture published by Dr. Ross 
 iu the International Clinics, vol. i. p. 1 . 
 
DIABETES AND ACROMEGALY. 139 
 
 polyuria has been present also, and in some cases the symptoms 
 have been those of well-morked diabetes, as in the first case 
 just reported. 
 
 Thus sugar was found in the urine in six out of twenty-one 
 recently reported cases of acromegaly ; in the other fifteen the 
 urine was free from sugar. In four of the cases in which 
 glycosuria was present the symptoms were those of true 
 diabetes. In two of the six cases a tumour of the pituitary 
 body was found at the autopsy, just as in the two cases recorded 
 above ; the other four patients were still alive when the cases 
 were reported. It is worthy of note, that in both of the cases 
 recorded above, diabetes only came on at a late period of the 
 disease, and probably some of the fifteen cases just mentioned, 
 in which the urine was free from sugar, would develop glycosuria 
 before death. In a case of acromegaly with diabetes, which 
 has been recently reported by Hansemann ( 75 ), post-mortem ex- 
 amination revealed considerable increase in the connective tissue 
 of the pancreas ; a tumour of the pituitary body was also present. 
 Out of ninety-seven cases of acromegaly which Hansemann has 
 collected from medical literature, diabetes or glycosuria was 
 present in twelve. 
 
 The association of diabetes with acromegaly does not appear 
 to be constant, however, even at a later period of the disease. 
 * * * 
 
 From a consideration of the facts recorded in this section, 
 we may draw the following conclusions respecting the relation 
 of diabetes to affections of the nervous system : — 
 
 1. Emotional disturbances (mental anxiety, worry, shock, or 
 fright) are antecedents, and apparently play some part as 
 exciting causes in certain cases of diabetes. In 1 6 per cent, of 
 the cases at the Manchester Eoyal Infirmary, a history of such 
 emotional disturbances was obtained. 
 
 2. On post-mortem examination, in the majority of cases of 
 diabetes, to the naked eye the nervous system is either normal, 
 or presents only slight and unimportant changes. 
 
 3. In a minority of cases, definite macroscopic changes, of 
 diverse nature, are found ; they are most commonly localised in 
 the floor of the fourth ventricle or in the medulla; in all 
 probability they have played an important part in the causa- 
 tion of the disease. 
 
 4. In case of tumour of the pituitary body, sometimes there 
 
1 40 E TIOL OGY AND E TIOL O GICAL RE LA TIONS. 
 
 have been symptoms of diabetes, often associated with those of 
 acromegaly, during life. 
 
 5. Microscopically, the small vessels of the medulla are 
 dilated in many cases of diabetes, especially in the region of 
 the vagus nuclei ; in other cases they are normal. The nerve 
 cells and other structures in the medulla appear normal, or at 
 least do not present any definite or constant change. Occa- 
 sionally minute hemorrhagic patches are found in the vagus 
 nucleus. 
 
 In spite of the negative result of microscopical examination 
 of the nerve cells and other structures of the medulla in the 
 majority of cases, it is still quite possible that diabetes is often 
 dependent on some minute or functional changes in this region. 
 The microscopical examination of the medulla in other diseases 
 teaches us that very slight changes in the nerve cells may be 
 productive of most serious symptoms. In bulbar paralysis, for 
 example, if the disease does not run a very chronic course, often 
 the changes in the medulla are slight. During recent years a 
 number of cases of bulbar paralysis and of ophthalmoplegia have 
 been recorded, which have terminated fatally, and in which 
 careful microscopical examination of the medulla, pons, etc., has 
 failed to reveal any pathological changes ; and yet there can be 
 little doubt that disease has been present in these parts. (I have 
 examined two such cases myself during the last four years.) 
 Hence it is quite possible that some slight functional change in 
 the nerve cells of the nuclei of the medulla may be the cause, 
 or may play some important part in the causation, of certain 
 cases of diabetes. 
 
 E. Belation between Diabetes Mellitus and Lesions 
 of the Pancreas. 
 
 Changes in the pancreas in diabetes have been described, 
 from time to time, by various physicians and pathologists, ever 
 since Thomas Cawley ( 76 ), in 1788, recorded a case of diabetes in 
 which the pancreas was atrophied and contained calculi. Thus 
 attention has been called to the importance of pancreatic lesions 
 in diabetes by Chopart, Bright, Eecklinghausen, Frerichs, Silver, 
 Munk, Mackenzie, Taylor, Bond, Windle, Duffey, and many others. 
 Lancereaux, many years ago, put forward the view that diabetes 
 with wasting is particularly associated with pancreatic lesions. 
 
PANCREA TIC LESIONS IN DIABE TES. 1 4 r 
 
 Baumel ( 77 ), in 1882, recorded cases of diabetes with pancreatic 
 changes, and went so far as to attribute all forms of diabetes to 
 the absence of diastatic pancreatic ferment in the intestine. 
 
 But the importance of these pancreatic changes was not 
 fully recognised until 1889, when Minkowski and v. Mering ( 78 ) 
 showed, by experiments on animals, that permanent diabetes 
 mellitus could be produced by complete removal of the pancreas. 
 (These experiments have been already discussed, see p. 74.) 
 Since that time many cases of diabetes associated with pan- 
 creatic changes have been recorded (by Lepine, Williamson ( 79 j, 
 Vaughan Harley ( 80 ), Hoppe-Seyler ( 8:I ), Fleiner, Hansemann, and 
 many others). 
 
 There can be no doubt, however, that whilst in some cases 
 of diabetes the pancreas presents marked changes, in other cases 
 no pathological changes can be detected by our present methods 
 of macroscopical and microscopical examinations. Some writers 
 refer all cases of diabetes to disease of the pancreas; some refer 
 only diabetes with wasting to pancreatic affections ( S2 ) ; whilst 
 others believe that there is no connection between diabetes and 
 pancreatic lesions ( 83 ). Hence some caution is necessary before 
 coming to a definite conclusion, and, in the following pages, 
 evidence for and against the pancreatic theory of diabetes will 
 be put forward. 
 
 1. Frequency and nature of pancreatic lesions in diabetes 
 mellitus. — In looking over old post-mortem reports of cases of 
 diabetes, very frequently one finds that there is no mention of the 
 condition of the pancreas. Before the publication of the results 
 of the experiments of Minkowski and v. Mering on pancreatic 
 diabetes, frequently the state of the pancreas was overlooked in 
 the post-mortem examinations of diabetic subjects. Eokitansky, 
 however, found naked-eye changes in the pancreas in thirteen 
 out of thirty cases of diabetes. 
 
 During the last seven years much attention has been paid 
 to the macroscopical and microscopical condition of the pancreas 
 in diabetes, and many cases have been recorded in which 
 pathological changes have been found, but few in which the 
 pancreas has been normal — microscopically, as well as macro- 
 scopically. 
 
 Lepine ( 84 ) lias pointed out that the pancreas may appear 
 normal to the naked eye, but may present distinct changes on 
 microscopical examination. Lemoine and Lannois ( 85 ) have de- 
 
1 4 2 E TIOL OGY AND E TIOL O GICAL RE LA TIONS. 
 
 scribed a peri-acinous sclerosis of vascular origin, which can only 
 be recognised on microscopical examination — though the import- 
 ance of these changes has been disputed. 
 
 Hence, in order to form a reliable estimate of the proportion 
 of cases in which the gland is diseased, and the proportion of 
 cases in which it is normal, and also of the nature and frequency 
 of the various pathological changes, it is necessary to examine 
 the pancreas, in a series of consecutive cases, microscopically, as 
 well as macroscopically. The following are brief notes of the 
 condition of the pancreas in twenty-three consecutive cases of 
 diabetes mellitus. In twenty-two cases I have examined the 
 pancreas microscopically, as well as macroscopically. A few of 
 these were my own cases, but for the opportunity of making the 
 pathological examination in the others, I am indebted to the 
 kindness of Drs. Leech, Dreschfeld, Steell, Harris, and Wilkinson 
 (of the Manchester Infirmary), to the late Dr. Boss, and to my 
 friends Dr. Mackenzie and Mr. Milner : — 
 
 Case 1. — Diabetes mellitus ; history of alcoholism ; marked cirrhosis 
 of the 'pancreas, with small calculi. — Thomas L., a?t. 45. 
 
 Previous history. — Patient enjoyed fairly good health until the 
 summer of 1890. In April 1890 he lost his employment (that of a door 
 porter at a music hall), and was unable to obtain employment again 
 until December 1890. During this period he had much mental worry, 
 and was not able to obtain sufficient food. In September 1890 he 
 began to lose flesh, and in November of the same year he began to be 
 troubled with great thirst and diuresis, together with increased appetite. 
 By January 1891 the patient had become much weaker, and the thirst 
 and diuresis had increased. The patient's weight was 11 st. 8 lb. in 
 April 1890 ; in January 1891, 8 st. 1\ lb. From the age of 29, until 
 five weeks before admission, the patient had taken large quantities of 
 alcohol (8 to 10 pints of beer) daily. There is no history of injury to 
 the head or back. At the age of 19 he suffered from gonorrhoea, but 
 no history of syphilis could be obtained. His wife has had one mis- 
 carriage. He has been fond of sweet food. There is no family history 
 of diabetes. 
 
 Present state. — Patient complains of great thirst and diuresis. There 
 is marked wasting ; the skin is harsh, dry, and scaly ; the temperature 
 is normal. The urine is clear, sp. gr. 1037, acid, contains a large amount 
 of sugar, but not a trace of albumin ; there is no deposit. It has not, 
 however, the pale straw colour of diabetic urine, but has a slight reddish 
 or pink tinge. With perchloride of iron no brown-red coloration pro- 
 duced. The bowels are not constipated. Distinct signs of phthisis 
 
PANCREATIC LESIONS IN DIABETES. 143 
 
 present. Heart normal; pulse 84. Respirations 18. Hepatic and 
 splenic clulness normal. Knee-jerks absent. 
 
 During the time the patient was in the hospital (22nd to 27th 
 April), the amount of urine varied from 82 to 176 oz. daily; the 
 specific gravity from 1035 to 1038 ; the amount of sugar from 27 to 
 36 grs. to the ounce, or from 2664 to 4752 grs. daily. Death from 
 coma, 27th April. 
 
 Autopsy (Abstract). — Body wasted. Pericardium normal. Heart 
 soft and flabby ; valves normal. 
 
 Lungs. — In the uppermost lobe of the right lung was a large 
 phthisical cavity, surrounded by nodules of caseous pneumonia. The 
 lung tissue was also studded with miliary tubercles. 
 
 Brain. — Meninges normal. Slight thickening of the choroidal 
 plexuses of the fourth ventricle ; otherwise fourth ventricle, medulla, 
 pons cerebellum, and cerebrum quite normal. 
 
 Liver. — Weight 3 lb. 10 oz. ; capsule normal. On section, firm 
 and somewhat congested ; no evidence of cirrhosis. 
 
 Spleen congested, soft, and pulpy ; weight 1 1 oz. 
 
 Stomach and intestines normal. 
 
 Kidneys slightly congested. 
 
 Bladder normal. 
 
 Pancreas very firm and hard ; firmly adherent to, and very difficult to 
 separate from, adjacent parts; the tail was adherent to the spleen. Weight 
 4J oz. On section it was pale, very firm, and cut with considerable diffi- 
 culty. The pancreatic duct was dilated at several places, forming cyst-like 
 dilatations (about the size of a small pea), which contained mucous fluid 
 and small calculi. In one of these dilatations, about the centre of the 
 pancreas, was an irregular oblong calculus, about the size of a small pea. 
 At the tail extremity of the pancreas were several small cysts, contain- 
 ing calculi, each about the size of a pin's head. The calculi had a pale, 
 yellowish colour, and were of the consistence of mortar. The pancreatic 
 tissue (on section) was seen to be broken up into very small round or 
 irregular masses, surrounded by broad tracts of fibrous tissue. After 
 hardening in Midler's fluid, and then washing well in water, these 
 changes in the pancreas were seen very clearly. 
 
 Microscopical Examination. — Sections of the pancreas, stained with 
 logwood and eosin, showed most marked cirrhosis of the gland. In 
 many parts, only a few small clusters of pancreatic cells could be seen, 
 almost the whole of the field . of the microscope being occupied with 
 dense fibrous tissue. In other parts, there were well-stained small masses 
 of pancreatic tissue, with one or two small ducts, in transverse or 
 longitudinal sections. These masses of pancreatic tissue were separated 
 and surrounded by a large amount of dense fibrous tissue; also fibrous 
 bands were prolonged between the gland cells. In some parts of the 
 
i44 -E TIOL OGY AND E T10L O GICAL RE LA TIONS. 
 
 gland there were larger masses of pancreatic tissue, surrounded and 
 infiltrated by dense fibrous tissue. Many of the pancreatic cells were 
 broken down in these patches, and did not stain well with logwood. 
 In many parts the dense fibrous tissue was well supplied by fine blood 
 
 Fig. 9. — Drawiug of section of pancreas, showing marked cirrhosis, 
 Case I. (Zeiss, D). 
 
 vessels. Microscopical examination showed that the condition was 
 certainly one of marked cirrhosis, and not scirrhus (see Fig. 9). On 
 microscopical examination, the liver appeared normal. After hardening 
 in Midler's fluid, careful macroscopic examination of the medulla and 
 pons failed to reveal any abnormality. Also on microscopical examina- 
 
PANCREATIC LESIONS IN DIABETES. 145 
 
 tion, sections of the medulla appeared quite normal, with the exception 
 of very slight dilatation of the blood vessels at the posterior part, near 
 the floor of the fourth ventricle. 
 
 In the case recorded, the urine had all the characteristics 
 of diabetic urine, with the exception of the colour. Instead 
 of the usual pale straw colour, it had a rose or pinkish tinge, 
 and yet not a trace of albumin was present ; there was no 
 deposit of urates or other substances, and the urine was quite 
 clear. Unfortunately, a thorough chemical examination was not 
 made. Fichtner x records a similar case. In his case the urine 
 became-rose coloured on standing. 
 
 Case 2. — Mary H., eet. 34. Considerable wasting. History of 
 thirst and diuresis for four weeks. Urine 1029 ; large amount of sugar 
 present ; acetone and cliacetic acid present ; trace of albumin. Death 
 from coma on the second day after admission to the hospital. 
 
 Necropsy. — Pancreas small, weight If oz. The gland, especially 
 at the head, was exceedingly firm and hard. Microscopical examination 
 showed most extensive cirrhosis. There was no fatty degeneration of 
 the pancreas. The brain was normal to the naked eye ; microscopically, 
 the medulla was normal. 
 
 Case 3. 2 — Mary H., set. 41. Duration of illness only about three 
 months. Chief symptoms : epigastric pain, swelling of abdomen, ascites ; 
 ill-defined, resisting ridge felt in the epigastrium. Patient did not become 
 very much emaciated. Urine not increased in quantity, sp. gr. 1047, 
 acid, no albumin, large amount of sugar (20 grs. to the ounce), marked 
 deep brown-red coloration with perchloride of iron (Gerhardt's reaction), 
 also marked reaction for acetone (Legal's test). During the last few 
 days of life, vomiting was frequent. Hsematemesis and death from 
 syncope. 
 
 Necropsy. — Peritoneal cavity contained a large quantity of bloody 
 fluid ; omental sac full of blood clot. The following is the condition 
 which I found on dissection : — Pancreas densely infiltrated by tumour 
 growth. At some parts the pancreatic duct contained a thick mortar- 
 like material; three branched, irregular, friable white calculi, each 
 a little larger than a pea, were found in the head of the gland. !Near 
 the middle of the gland were two larger calculi of similar colour and 
 consistence, one about the size and shape of a date-stone, the other 
 about I in. long, and bent at a right angle at one end ; a large 
 
 1 /1'iUsckes Arch.f. klin. Med., Leipzig, Bel. xlv. S. 117. 
 
 z 'I'h is case was reported by Dr. Dresehfulrl in a paper in the Med. Chron., Manchester, 
 April 1895, p. 14. 
 io 
 
' 1 46 E TIOL OGY AND E TIOL O GICAL RE LA TIONS. 
 
 number of smaller calculi were also present. Little or no normal 
 pancreatic tissue could be detected. The growth extended from the 
 pancreas in the most irregular manner into the surrounding parts at the 
 posterior wall of the abdomen. On the serous surface of the posterior 
 wall of the stomach was a diffuse mass of new growth, which extended 
 towards the cardiac end of the organ. The growth was infiltrated with 
 blood. The pancreatic tumour also extended from the spleen to the 
 portal vein, which it compressed but did not infiltrate. Superficially 
 the growth was very soft and vascular ; the pylorus and duodenum were 
 not invaded by it. The large omentum was much infiltrated with fat, 
 and presented numerous secondary deposits. There were also secondary 
 nodules in the gastro-hepatic omentum. The intestines were not impli- 
 cated by the growth. The liver was small, and the surface smooth. 
 There was no evidence of cirrhosis, but several small secondary deposits 
 of new growth were found. The gall bladder was normal. The spleen 
 was slightly enlarged. The kidneys were pale, but normal on section. 
 On the serous surface of the uterus were several nodules of new growth. 
 Microscopical examination showed that the growth was carcinoma, the 
 firmer parts presenting the appearance of scirrhus. 
 
 Case 4, — 0., tet. 67. Emaciated, but not to a great extent. Urine 
 1033, contained a large quantity of sugar. Severe thirst and diuresis. 
 
 Necropsy. — Pancreas hard and firm on section, weight 2 oz. 
 Microscopical examination revealed distinct cirrhosis. In the inter- 
 acinous connective tissue were many groups of fat cells (lipomatosis). 
 
 Case 5. — J)^ aet. 52. Much emaciated. Urine 1040; sugar varied 
 from 20 to 26 grs. to the ounce. The daily quantity of urine varied from 
 140 to 170 oz. Arteries very atheromatous. 
 
 Necropsy. — Pancreas weighed 3 oz. It was very fjrm, and appeared 
 cirrhotic. Microscopical examination revealed a moderate degree of 
 cirrhosis, but no fatty degeneration. Brain normal. Extensive tuber- 
 culosis of both lungs. Arteries markedly atheromatous, in many parts 
 calcified. 
 
 Case 6. — H., set. 23. Acromegaly. Optic atrophy. Glycosurin 
 without thirst. At a later date, thirst marked, diuresis. Patient well 
 nourished. Death from coma. (For other notes, see p. 138.) 
 
 Necropsy. — Sarcoma of the pituitary body, involving the optic 
 chiasma. Pancreas very large, infiltrated at many parts with fat ; 
 weight 7 oz. Microscopically, well-marked lipomatosis detected. 
 
 Case 7. — B., set. about 35. Emaciated. Urine, sp. gr. 1040 ; daily 
 quantity 80 to 150 oz. ; large amount of sugar present. 
 
PANCREATIC LESIONS IN DIABETES. 
 
 H7 
 
 Necropsy. — Pancreas much atrophied, exceedingly soft and flabby, 
 weight \\ oz. Scattered throughout the gland (after hardening in 
 Midler's fluid) were seen numerous small pale patches. Microscopical 
 examination revealed fatty degeneration and fatty infiltration (lipomat- 
 osis) at these parts. 
 
 Case 8. 1 — J. C, set. 46. Wasted markedly. Urine 
 oz. daily; sp. gr. 1035 to 1045; sugar 4000 to 5300 
 grs. daily. JS T o reaction with perchloride of iron. 
 
 Necropsy. — Pancreas very small, weight not quite a 
 quarter of an ounce. Pancreatic duct of normal size. All 
 organs wasted. Heart only weighed 6 oz.; but the wasting 
 of the pancreas was altogether out of proportion to that 
 of the other organs. The pancreas was firm, and pre- 
 sented no other change beyond atrophy. 
 
 100 to 136 
 
 Case 9. — J., set. 65. Much emaciated. Urine, sp. 
 gr. varied from 1028 to 1034 ; the daily quantity of iirine 
 from 74 to 104 oz. ; the amount of sugar from 22 to 28 
 grs. to the ounce. 
 
 Necropsy. — Pancreas weighed 10 drms., and was 
 atrophied and flabby. Microscopical examination revealed 
 slight fatty degeneration with slight lipomatosis. 
 
 Case 10. — ~N., aet. 44. Very marked wasting. Urine, 
 daily amount varied from 100 to 204 oz. ; the sp. gr. was 
 1034 to 1035 ; a large amount of sugar present. Death 
 from coma. 
 
 Necropsy. — The pancreas was markedly atrophied, and 
 weighed 5-| drms. only ; but, apart from the atrophy, 
 microscopical examination did not reveal cirrhosis, fatty 
 degeneration, or any other change. 
 
 Fig. 10.— Actual 
 size of atro- 
 phied pancreas 
 in Case 8. 1 
 
 Case 11. — W., set. 19. The daily amount of urine 
 varied from 70 to 90 oz. ; the sp. gr. from 1035 to 1038 ; 
 a large quantity of sugar was present. The disease ran a very rapid 
 course, and death occurred from coma about fourteen weeks after the 
 symptoms were first noticed. 
 
 Necropsy. — The pancreas was atrophied, and weighed 10 drms.; but 
 microscopical examination did not reveal any other pathological 
 change. 
 
 1 Case under the care of Dr. Harris. Described by him at a meeting of the Manchester 
 Pathological Society. See Brit. Med. Journ., London, 28th November 1890. 
 
148 ETIOLOGY AND ETIOLOGICAL RELATIONS. 
 
 Case 12. — Ellen B., set. 20. Marked emaciation. Tuberculous lung 
 disease. The diabetic symptoms came on soon after a severe fright. 
 Urine, amount varied from 90 to 116 oz. daily; sp. gr. 1032 to 1036 ; 
 sugar 28 to 30 grs. to the ounce. 
 
 Necropsy. — Pancreas weighed 1 1 drms. (but all the organs were very 
 much wasted) ; microscopically it appeared normal. 
 
 Case 13. — R, set. 29. Marked wasting. Urine, daily amount 
 varied from 100 to 150 oz. ; sp. gr. 1030 to 1042 ; and the sugar from 
 20 to 36 grs. to the ounce. 
 
 Necropsy. — Pancreas weighed 1^ oz. ; macroscopically and micro- 
 scopically it appeared normal. 
 
 Case 14. — T., set. 26. Duration of disease, nine months. Emacia- 
 tion. Urine 90 to 120 oz. daily; sp. gr. 1035; sugar 30 to 38 grs. 
 per ounce. 
 
 Necropsy. — Pancreas atrophic, weight lh oz. ; microscopically 
 
 normal. 
 
 Case 15. — W., set. 31. Marked history of alcoholism. Great 
 emaciation. Death from coma. Urine 255 to 293 oz. daily ; sp. gr. 
 1036 ; sugar 22 to 23 grs. to the ounce. 
 
 Necropsy. — Pancreas, weight H oz. Macroscopically, normal. 
 Microscopically, no cirrhosis, no fatty degeneration, or infiltration ; in 
 some places nuclei of cells stained badly with logwood, otherwise the 
 sections appeared normal. Lungs presented early tubercular changes. 
 
 Case 16. — B., set. 52. History of alcoholism. Liver enlarged. At 
 first urine 154 oz. ; sp. gr. 1017; 10 grs. of sugar to the ounce. After- 
 wards the diabetic symptoms increased rapidly, the sp. gr. of the urine 
 reached 1031, and the amount of sugar was 26 grs. to the ounce. 
 
 Necropsy. — Pancreas weighed 3^ oz. On microscopical examination 
 the gland appeared normal. Marked cirrhosis of the liver ; tuberculous 
 disease of the lungs. 
 
 Case 17. — M., set. 46. Emaciated slightly. Urine, daily amount 
 varied from 170 to 190 oz. ; sp. gr. 1035 ; sugar 30 grs. to the ounce. 
 Tuberculous disease of the lungs. Death from coma. 
 
 Necropsy. — Pancreas weighed 3 oz. ; macroscopically and micro- 
 scopically it appeared normal. 
 
 Case 18. — F., set. 21. Much emaciated. Urine, daily amount 
 varied from 130 to 140 oz. ; sp. gr. from 1030 to 1040; sugar 33 to 35 
 grs. to the ounce. 
 
PANCREATIC LESIONS IN DIABETES. 149 
 
 Necropsy. — Pancreas weighed 2| oz. ; macroscopically and micro- 
 scopically it appeared normal. 
 
 Case 19. — E., set. 42. Great emaciation. Urine, amount varied 
 from 90 to 130 oz. daily; sp. gr. 1032 to 1045; amount of sugar, 
 5 to 7 per cent. 
 
 Necropsy.- — Pancreas weighed 2| oz. On microscopical examination 
 it appeared normal. Multiple abscesses in the right lobe of the liver. 
 Cerebro-spinal meningitis (recent). 
 
 Case 20. — R., set. 35. Marked emaciation. Amount of urine 
 varied from 117 to 125 oz. daily; the amount of sugar was about 30 
 grs. to the ounce. Tuberculous lung disease. 
 
 Necropsy. — Pancreas weighed 2|- oz. ; macroscopically and micro- 
 scopically it appeared normal. 
 
 Case 21. — W., set. 36. Bi-temporal hemianopsia, afterwards double 
 optic atrophy. Acromegaly. At a later date, marked wasting. Symp- 
 toms of diabetes mellitus. Urine 160 oz. ; sp. gr. 1032; sugar 31 
 grs. to the ounce ; trace of albumin ; marked brown-red coloration with 
 perchloride of iron. Death from coma. 
 
 Necropsy. — Sarcoma of pituitary body, involving optic chiasma. 
 Macroscopically, pancreas appeared normal, weight 2 oz. Micro- 
 scopically it also appeared normal. 
 
 Case 22. — M'M., set. 19. Duration of symptoms five weeks only. 
 Sudden onset from thirst. Marked wasting. Urine 180 oz. ; sp. gr. 
 1032 to 1034; sugar 4400 to 4860 grs. daily. 
 
 Necropsy. — Pancreas weighed If oz., atrophied somewhat, but 
 scarcely out of proportion to the general wasting of the body. Other- 
 wise, macroscopically and microscopically, the gland appeared normal. 
 Tubercular disease of the apex of the left lung. 
 
 Case 23. — P., set. 26. Admitted into hospital comatose. Diabetes 
 then first discovered. Urine, strongly acid, 1025; large amount of 
 sugar present. Marked brownish -red coloration with perchloride 
 of iron. Good reaction for acetone (Legal's test). Patient 
 emaciated. 
 
 Necroj/sy. — Pancreas weighed 1\ oz. ; normal macroscopically and 
 microscopically, except that a few scattered patches were seen in which 
 the nuclei of the cells did not stain well. 
 
 Results. — Condition of the pancreas in twenty-three con- 
 secutive cases of diabetes mellitus : — 
 
1 5 o E TIOL OGY AND E TIOL O GICAL RE LA TLONS. 
 
 (1) Extensive changes : — Cases. 
 
 Very marked cirrhosis . . . . .2 
 
 Cancer ........ 1 
 
 (2) Fairly well-marked changes : — 
 
 Cirrhosis ........ 2 
 
 Lipomatosis ....... 1 
 
 Atrophy, fatty degeneration, and infiltration . 1 
 Yery advanced atrophy, gland weighing less than 
 
 ioz 1 
 
 (3) Slight changes : — 
 
 Atrophy, with slight fatty degeneration . . 1 
 
 Atrophy (without any other changes) out of pro- 
 portion to the general wasting . . .2 
 
 (4) Atrophy, but only in proportion to general wasting ; 
 
 no other change ...... 
 
 (5) Pancreas normal, macroscopically and microscopic- 
 
 ally 
 
 Total 
 
 4 
 
 8 
 23 
 
 In twelve out of twenty-three cases, therefore, the pancreas 
 was either normal, or only atrophied in proportion to the 
 general wasting. 
 
 Some of the above cases were published by the writer ( s6 ) in 
 a paper in The Lancet, 1894 (14th April), and soon afterwards 
 an article appeared by Hansemann ( 87 ) of Berlin, in which the 
 relation of the pancreas to diabetes is carefully discussed. A 
 survey of the pathological reports of the Berlin Pathological 
 Institute for ten years gave the following results : — 
 
 Cases. 
 
 (1) Diabetes without pancreatic changes . . .8 
 
 (2) Diabetes without any information respecting the 
 
 pancreas ........ 6 
 
 (3) Diabetes with pancreatic affections . . .40 
 
 (4) Pancreatic disease without diabetes . . .19 
 
 As it is not stated that a microscopical examination was 
 made in every case, these results probably refer to the macro- 
 scopical changes only. It is remarkable, that of the forty cases 
 in which pancreatic changes were found, thirty-six presented 
 simple atrophy ; three cases presented fibrous induration ; one 
 was a case of pancreatic cyst. 
 
 Nature of pancreatic lesions in diabetes. — The above table 
 of twenty-three cases along with the results of Hansemann, 
 
PANCREATIC LESIONS IN DIABETES. 
 
 I5 1 
 
 shows the nature of the pancreatic changes. But in order 
 to obtain further information on this point, I have collected and 
 classified the results in 100 cases of diabetes, recorded in medical 
 literature, in which the pancreas was abnormal. 
 
 In very few of the cases was a microscopical examination 
 made, hence this table represents generally the positive result 
 of macroscopical examination only. . 
 
 One Hundred Cases of Pancreatic Lesions in Diabetes. 
 
 No. of Cases. 
 
 Atrophy of the pancreas (more or less marked) 
 Very marked atrophy, gland almost absent 
 
 ,, „ gland not recognised by naked 
 
 eye 
 
 Very marked atrophy, with cystic dilatation of duct 
 
 ,, ,, ,, and induration 
 
 Calculi present, other conditions not stated 
 Marked fatty degeneration .... 
 
 ,, ,, ,, with calculi 
 
 Fatty degeneration, with increase of connective tissue 
 Complete fatty degeneration and marked atrophy 
 
 ,, transformation of pancreas into a fatty mass 
 Cystic disease ...... 
 
 Large pancreatic cysts .... 
 
 Cyst of pancreas, with necrosis . 
 
 Transformation into a firm mass of fibrous tissue, pan 
 
 creatic tissue being almost absent . 
 Marked cirrhosis ..... 
 Cirrhotic and cystic pancreas 
 Calcified pancreas ..... 
 Peripancreatitis and pancreatitis hemorrhagica 
 Abscess . . . . 
 Cancer ....... 
 
 39 
 
 2 
 2 
 1 
 1 
 10 
 3 
 1 
 1 
 2 
 
 3 
 3 
 
 1 
 
 10 
 3 
 
 1 
 1 
 2 
 3 
 
 61 
 100 
 
 Hansemann has given the following results from an analysis 
 of seventy-two cases recorded in medical literature : — 
 
 Cases. 
 Pancreatic calculus . . . . . . .11 
 
 Cancer with obstruction of the duct . . . .5 
 
 Simple atrophy with interstitial inflammation . . 38 
 
 Other changes ........ 15 
 
152 E TIOL OGY AND E TIOL O GICAL RE LA TIONS. 
 
 The changes in the pancreas in diabetes are therefore varied 
 in nature. Atrophy is the most common change, but it is 
 probable that in some of those cases, at least, the atrophy may be 
 really only a part of the general wasting. Marked atrophy of 
 the pancreas has been observed, however, in diabetes when the 
 general wasting has been absent or trifling (as in Case 11), 
 and it has also been observed in stout diabetics. 
 
 Hansemann states that the atrophy of the pancreas in 
 diabetes differs from atrophy of the pancreas which is simply a 
 part of general wasting. In the former condition, the stroma of 
 the gland is not atrophied as in the latter ; and in the former the 
 stroma has filled more or less the spaces left by the atrophy 
 of the gland parenchyma, i.e. a kind of interstitial inflammation 
 has taken place. He compares the condition to granular 
 atrophy of the kidney. I have often been struck by this 
 appearance of the atrophied pancreas in diabetes. 
 
 2. Condition of the pancreas in a series of consecutive autopsies. 
 — In considering the relation of pancreatic lesions to diabetes, it is 
 important to know what is the condition of the gland in a series 
 of unselected autopsies on persons dying from various ailments. 
 On this point Kasahara ( 8S ) has published the records of the 
 pathological examination of the gland in eighty-three consecutive 
 cases. The following table I have drawn up from his records : — 
 
 Pancreas normal, or presenting only trifling changes 
 Increase of fat — 
 
 Interstitial fat markedly increased 
 
 Interstitial fat moderately increased . 
 
 Fatty metamorphosis (one case of diabetes) 
 
 Marked localised fatty changes . 
 
 Partial fat necrosis .... 
 
 Interstitial fat greatly increased 
 fncrease of connective tissue — 
 
 In a marked degree .... 
 
 In a moderate degree 
 
 Slightly increased .... 
 
 Increased locally .... 
 
 Marked atrophy (both cases of diabetes) . 
 Cancer ....... 
 
 Marked turbidity of cells 
 
 Cases. 
 
 42 
 
 41 
 
 83 
 
URINE IN PANCREA TIC DISEA SE. 153 
 
 Kasahara points out that in these eighty-three unselected 
 cases (of various diseases) met with in the post-mortem room, 
 marked atrophy was found in two only, and both were dia- 
 betic patients. In no class of cases, accompanied by wasting, 
 has he found such marked atrophy of the pancreas as that met 
 with in some cases of diabetes. 
 
 3. The condition of the urine in diseases of the pancreas.- — 
 In discussing the relation of diabetes to disease of the 
 pancreas, it is interesting to consider the subject from yet 
 another standpoint. In a series of cases of definite pancreatic 
 disease, what is the condition of the urine ? Undoubtedly 
 the urine will be found free from sugar in a large proportion 
 of such cases. 
 
 Numerous cases of pancreatic disease are on record in which 
 the urine did not contain sugar. But it is important to bear in 
 mind the results of Minkowski's experiments on partial ex- 
 tirpation of the pancreas. If a small part of the pancreas be 
 left behind (one-fifth even has sufficed), diabetes does not occur. 
 Now, if this small piece of pancreas remaining in the abdomen 
 is sufficient to prevent the occurrence of diabetes in animals, 
 one would expect that in pancreatic affections in man, if the 
 whole of the gland be not diseased, i.e. if a small piece should 
 retain its function, that diabetes would not occur. Certainly, in 
 a very large number of cases of pancreatic disease in man, a 
 portion of the gland has remained unaffected. Thus, in ten 
 cases of pancreatic disease (confirmed by autopsy) at the 
 Manchester Eoyal Infirmary, the urine was free from sugar 
 in every one. But post-mortem examination showed that in 
 five of the cases only the head of the pancreas was affected, and 
 the rest of the gland was normal. In the other five cases the 
 post-mortem records state that the lesion was in the head of the 
 pancreas, and no mention is made of the condition of the rest of 
 the gland, so that we are not justified in concluding that the 
 whole of the gland was affected ; probably the rest of the gland 
 was normal. 
 
 In many cases of pancreatic cyst, diabetes has been absent ; 
 but this is explained by the fact that there has been some 
 normal gland tissue remaining. 
 
 On the other hand, there are on record some cases of 
 extensive necrosis, and some cases of diffuse cancer infiltrating 
 the whole of the pancreas, which have not been associated with 
 
154 E TIOL OGY AND E TIOL O GICAL RE LA TIONS. 
 
 diabetes. With both of these conditions diabetes is sometimes 
 associated, however. 
 
 4. Conclusions. — It is very improbable that all the varied 
 pancreatic changes which have been recorded in diabetes are the 
 result of the disease. It is scarcely credible that one disease — 
 diabetes — should produce such a number of varied affections of 
 the pancreas as cancer, cirrhosis, atrophy, abscess, etc. Then 
 again, from the frequency of pancreatic lesions in diabetes, and 
 from the fact that often the pancreas presents marked changes, 
 whilst in the rest of the organs none of importance are 
 detected, it is improbable that the lesions are accidental. 
 Further, when we consider the remarkable results produced by 
 extirpation of the pancreas, it appears very probable that, in 
 certain cases, diabetes is directly due to pancreatic disease. 
 
 Whilst experiments on animals appear to have conclusively 
 proved that total removal of the pancreas produces diabetes, 
 there are two objections to the pancreatic theory of the origin 
 of diabetes in man, as was pointed out by Minkowski himself. 
 In the first place, pathological changes are not found in all cases 
 of diabetes in man ; secondly, sugar is not found in all cases of 
 disease of the pancreas. This second objection has just been 
 discussed, and partly met. As regards the first objection, it was 
 pointed out by Minkowski that the number of cases in which 
 the pancreas had been carefully examined, microscopically as 
 well as macroscopically, was comparatively small. But during 
 the last five years great attention has been paid to the condition 
 of the pancreas in diabetes, and there can now be no doubt that, 
 microscopically as well as macroscopically, the pancreas not 
 infrequently appears normal, as is shown in the records of the 
 cases reported above. The proportion of cases in which the 
 pancreas is diseased can only be correctly estimated by the 
 examination of a series of consecutive cases, and the records of 
 the twenty-three cases which I have described above, furnish, 
 I believe, a fairly correct representation. From these records 
 we see that in eight out of twenty-three cases the pancreas 
 appeared normal, both microscopically and macroscopically; and in 
 four other cases the only change was atrophy, which was scarcely 
 out of proportion to the general wasting of the body. Hence, in 
 twelve out of twenty-three cases diabetes was due to some other 
 cause than pancreatic disease, or the pancreatic affection must 
 have been one giving rise to changes which cannot be detected 
 
CONCLUSIONS. 155 
 
 microscopically or microscopically, i.e. a functional disease. It 
 may be pointed out, that not only the pancreas but also the 
 brain, liver, and other organs are very often normal, or present 
 changes only which are undoubtedly of a secondary nature. On 
 the other hand, the importance of the pancreas is evident, when 
 we bear in mind the results of experiments on animals, which 
 show that total removal of the pancreas invariably produces 
 diabetes, and that a small portion of pancreatic tissue, grafted 
 under the skin of the abdominal wall, is sufficient to prevent 
 diabetes, when the whole of the gland is removed from within 
 the abdomen. Considering also the number of cases recorded 
 in which marked pathological changes in the gland — often almost 
 complete destruction — have been found at the post-mortem 
 examination of diabetic patients, it appears exceedingly probable 
 that the disease has been due to a pancreatic lesion when 
 extensive changes have been found in that gland. 
 
 But it must be borne in mind that 'partial extirpation of 
 the pancreas in animals is not followed by diabetes. A small 
 portion of the gland left behind is sufficient to prevent the 
 occurrence of glycosuria. Hence it is very questionable whether 
 diabetes in man can be attributed to pancreatic disease when 
 the gland presents only slight changes ( 89 ). 
 
 There is another point to consider with reference to the 
 subject. The pancreas is a very vascular organ, and a slight 
 lesion in the nervous system might produce vasomotor changes 
 in the gland, and cause an excess of arterial blood to flow 
 through it. This might lead to alteration in the internal 
 secretion, and yet, post-mortem, no definite pathological changes 
 might be detected. 
 
 In a considerable number of cases of diabetes — twelve out 
 of the above twenty-three cases — it seems probable, therefore, 
 that diabetes was either due to some other cause than pancreatic 
 lesion, or to some functional or vasomotor change in the 
 pancreas. 
 
 In two of the cases a tumour of the pituitary body was 
 probably the cause of the disease. 
 
 In eight or nine of the twenty-three cases recorded above, it 
 appears probable that the disease found in the pancreas was the 
 cause of the diabetes — Cases 1, 2, 3, 4, 5, 7, 8, 9 (?), and 10. 
 
 That diabetes is produced in various ways seems very 
 probable. Just as fever is produced in many ways, so it is 
 
156 ETIOLOGY AND ETIOLOGICAL RELATIONS. 
 
 probable that glycosuria and the accompanying symptoms of 
 diabetes mellitus are produced by several pathological conditions. 
 
 Lancereaux ( s2 ) recognises three varieties of diabetes — (1) 
 Traumatic or spontaneous diabetes through lesion of the nervous 
 system; (2) diabetes with wasting, or pancreatic diabetes; 
 (3) diabetes with obesity, or constitutional diabetes. Diabetes 
 with wasting, however, is not always associated with pancreatic 
 disease. Lepine ( 8i ) has pointed out that the pancreas may be 
 normal in diabetes with wasting, and he has found this gland 
 diseased in diabetes with obesity. And in the cases recorded 
 above, 18, 19, 20, 22, and 23, there was marked wasting, and 
 yet the pancreas was of normal weight and appearance. 
 
 Hoppe-Seyler ( s:l ), Fleiner ( 90 ), and others have recorded 
 cases of diabetes associated with pancreatic disease and arterio- 
 sclerosis. In these cases it appeared probable that arterio- 
 sclerosis of the pancreatic arteries was the cause of the affection 
 of the pancreas, and therefore indirectly of the diabetes. 
 
 f. k elation between diabetes mellitus and 
 Arteriosclerosis. 
 
 Diabetes and arterio-sclerosis are both diseases most com- 
 monly met with after the age of 45. It is not surprising, 
 therefore, that the two affections should be frequently associated 
 in elderly people. But not infrequently diabetics under middle 
 age present marked signs of arterial sclerosis, and it seems prob- 
 able that in some cases there is an indirect connection between 
 the two diseases. 
 
 Grube ( 46 ) of Neuenahr found arterio-sclerosis present in sixty- 
 three out of 137 male patients, and in three out of forty female 
 patients. But the majority of his patients were over the age of 45, 
 and most of them suffered from the milder forms of the disease. 
 
 In a case of diabetes in a male ret. 52, with very advanced 
 atheroma of the arteries, I had the opportunity of making an 
 autopsy some time ago. Apart from the tubercular lung mis- 
 chief (which had developed late in the disease), the most striking 
 pathological condition was the very advanced atheroma of the 
 arteries in almost all parts of the body. Many of the arteries 
 were calcined, and many of their small branches ramifying in 
 the muscles were markedly atheromatous and rigid. I have 
 never seen a case which has presented more extensive atheroma 
 
DIABETES MELLITUS AND ARTERIOSCLEROSIS. 157 
 
 of the arteries on post-mortem examination. Definite symptoms 
 of diabetes had been present during the last eleven months of 
 life only, whilst the arterial changes were evidently of much 
 longer standing. In this case it is by no means improbable 
 (though not capable of proof) that the arterial disease was the 
 starting-point of the diabetes, possibly through secondary changes 
 in the pancreas, since cirrhosis of that organ was present. 
 
 Frerichs ( 47 ) drew attention to the frequency of arterial 
 changes in chronic forms of diabetes, especially when associ- 
 ated with gout, and thought that probably some connection 
 existed between diabetes and the arterial changes. Other writers 
 (Perraro, Laache, Grube, Fleiner) also regard arterio-sclerosis as 
 a cause of diabetes in certain cases. 
 
 It is conceivable that arterio-sclerosis may cause diabetes by 
 producing changes in the pancreas or in the nervous system 
 (medulla). 
 
 Hoppe-Seyler ( 4S ) records an instructive case of diabetes 
 mellitus, in which post-mortem examination showed that the 
 pancreas was transformed into a mass of adipose tissue. Micro- 
 scopically, it was found to consist chiefly of fat, with small scattered 
 islands of degenerated pancreatic tissue. There was atheroma 
 of the aorta, and the coeliac, gastro-duodenal, and splenic arteries, 
 especially the latter, were markedly calcified. Also the smallest 
 branches of the splenic artery entering the pancreatic substance 
 were thickened. Hoppe-Seyler attributes diabetes in this case 
 to the pancreatic changes, and believes that these were pro- 
 duced by arterial sclerosis. 
 
 Fleiner ( 49 ) has also recorded a case of cirrhosis of the pan- 
 creas, apparently the result of marked arterio-sclerosis. 
 
 Arterio-sclerosis often produces cirrhosis of the kidney ; 
 myocarditis may follow arterio-sclerosis of the coronary arteries ; 
 also valvular lesions of the heart, and various cerebral affections 
 may be produced by the same arterial changes. Hence it is 
 quite conceivable that arterio-sclerosis may occasionally lead to 
 cirrhosis, or other lesions of the pancreas. 
 
 When the changes produced in the pancreas by arterio- 
 sclerosis only affect a portion of the gland, then the remaining 
 normal gland tissue may be sufficient to prevent the occurrence 
 of diabetes; but when the whole of the gland is affected, it is 
 probable that diabetes will follow. 
 
 Kleiner thinks that a great number of the milder forms of 
 
158 E TIOL OGY AND E TIOL O GICAL RE LA TIONS. 
 
 diabetes in elderly persons are probably indirectly due to 
 arterio-sclerosis. 
 
 G. Diabetes of Endogenous Origin. 
 
 In many cases of diabetes, the most careful inquiry into 
 the previous history of the patient fails to reveal indications 
 of any external exciting cause. There is often no history of 
 injury, mental shock or anxiety, over-work, alcoholism, syphilis, 
 heredity, or any of the supposed exciting causes of the disease 
 which have been already referred to. In 100 cases at the 
 Manchester Eoyal Infirmary, there was no history of any of 
 the so-called exciting causes in fifteen at least. It is quite 
 possible, as has been pointed out by Strumpell and others, 
 that some cases are entirely or almost entirely of endogenous 
 origin, i.e. they are due to some developmental abnormality ; 
 and in most of the cases which appear to be due to the above- 
 mentioned causes, it is probable that there is an endogenous 
 in addition to the exogenous factor (or external exciting 
 cause) in the production of the disease. In very many cases 
 in which a history of some supposed exciting cause is given, it 
 is difficult to say whether this has really played any part in 
 the production of the disease. Sometimes the most careful 
 examination of the brain, pancreas, liver, or other organs fails 
 to reveal any pathological change, or any change to which im- 
 portance can be attached, and there is also no history of any of 
 the supposed exciting causes. Such cases show how very meagre 
 is our knowledge respecting the origin of the disease, in spite of 
 the enormous mass of literature which exists in relation to this 
 subject. 
 
 Etiology of Diabetes. 
 
 In 100 consecutive cases of diabetes, I have made careful 
 inquiries with respect to the usual etiological factors, with the 
 following results : — 
 
 Heredity : other members of family affected . in 13 cases. 
 
 History of marked mental worry, anxiety, 
 
 mental shock . . . . . . „ 10 ,, 
 
 Mental anxiety and over-work, etc., connected 
 with the nursing of a sick relative for a long 
 period ....... ,, 8 ., 
 
REFERENCES. 
 
 J 59 
 
 Disease associated with acromegaly 
 
 „ followed overwork 
 
 ,, ,, external injury 
 
 History of great alcoholic excess 
 Definite history of syphilis 
 Disease followed soon after pregnancy, 
 
 carriage, or abscess of breast 
 Disease developed directly after influenza 
 
 „ ,, directly after an acute 
 
 pleurisy .... 
 
 ,, ,, directly after pneumonia . 
 
 „ ,, during an attack of nasal 
 
 catarrh and bronchitis .... 
 Patient exceedingly stout before diabetes 
 
 developed ...... 
 
 Very sudden onset of thirst — exact time given 
 No history of any definite exciting cause 
 
 „ „ gout obtained in any of the cases. 
 
 in 2 cases. 
 
 5 
 
 6 
 
 17 
 
 6 
 
 4 
 
 9 
 
 15 
 
 PREFERENCES. 
 
 Grube, Karl 
 Bell . . . 
 Wegeli . 
 Osler, W. 
 
 5. Ebstein, W. 
 
 9. 
 
 10. 
 
 11. 
 12. 
 
 Ztsclir. f. Min. Med., Berlin, Bd. xxvii. 
 Edin. Med. J own., February 1896. 
 Arch./. Kinderh., Stuttgart, 1895, S. 14. 
 " The Principles and Practice of Medicine," 
 
 Edinburgh and London, 2nd edition, 1895, 
 
 p. 320. 
 " Ueber die Lebensweise der Zuckerkranken," 
 
 Wiesbaden, 1892, S. 100. 
 Bose, K. C. . . . Indian Med. Gaz., Calcutta, April 1895. 
 Sen, B. C. ... Ibid., July 1893. 
 
 "Der Diabetes mellitus," Berlin, 1893, S. 
 
 125, 130. 
 Deutsche med. Wchnschr., Leipzig, 1893, S. 
 
 779. 
 " Diabetes," Philadelphia and London, 1890, 
 
 p. 17. 
 Ibid., p. 12. 
 " Diseases of the Kidney and Urinary 
 
 Derangements," Part I.; "Diabetes," 
 
 London, 1875, p. 72. 
 Rev. de med., Paris, 1895, p. 1036; Ibid., 
 
 June 1896. 
 
 Caroe Ibid., June 1896. 
 
 Wegeli .... Loc. ci/., S. 57. 
 
 Seegen, J. 
 
 Wallach . 
 
 Purdy, C. . . . 
 
 ., ... 
 Dickinson, W. H. 
 
 13. Lepine 
 
 14. 
 15. 
 
1 60 E TIOL OGY AND E TIOL O GICA L RE LA TIONS. 
 
 ] 8. Senator, H. . . 
 
 19. Goolden . . . 
 
 20. Fischer . . . 
 
 21. Williamson, E. T. 
 
 22. Ebstein, W. . . 
 
 23. Asher . . . 
 
 24. Seegen . . 
 
 25. Grube, K. 
 
 26. v. Noorden, C 
 
 27. Strumpell . 
 
 28. Peiper. . . 
 
 16. Schmitz, E. . . . Berl. Min. Wchnsclir., 19tli May 1890. 
 
 17. Oppler, B. ttnd Ibid., 1896, Nos. 26 and 27. 
 
 Kllz, C. 
 
 Ibid., 27th July 1896. 
 
 Lancet, London, 1854, vol. i. p. 656, and vol. 
 
 ii. p. 29. 
 Arch. gen. de med., Paris, 1862, tome ii. pp. 
 
 257 and 413. 
 Lancet, London, 9th July 1892. 
 Deutsches Arch. f. Min. Med., Leipzig, Bd. 
 
 liv. 
 Quoted by Ebstein, loc. cit. 
 Loc. cit., S. 126. 
 Loc. cit. 
 
 "Die Zuckerkrankheit," Berlin, 1895, S. 48. 
 Berl. Min. Wchnschr., No. 46, 1896. 
 Deutsche med. Wchnschr., Leipzig, 1887, No. 
 
 17. 
 "Der Diabetes mellitus," Berlin, 1880 
 
 (German trans, from the Italian, by Dr. S. 
 
 Hahn), S. 294 ; " Specielle Pathologie und 
 
 Therapie der Stoffwechselkrankheiten," 
 
 Bd. i. 
 (Abstract), Deutsche med. Wchnschr., Leipzig, 
 
 1888, S. 825. 
 Brit. Med. Jo-urn., London, 1889, vol. ii. p. 
 
 965. 
 Berl. Min. Wchnschr., 1892, No. 6. 
 Loc. cit., S. 288. 
 
 34. Duncan, Matthews Trans. Obst. Soc. London, 1882, p. 285. 
 
 35. Tait, Lawson . . Practitioner, London, 1886, p. 401. 
 . . Brit. Med. Journ., London, 11th July 1885. 
 
 . Deutsche med. Wchnschr., Leipzig, 10th June 
 
 1897. 
 Arch. f. Psychiat., Berlin, 1878, Bd. viii. 
 
 S. 510. 
 "Ueber den Diabetes," Berlin, 1884. 
 " Lectures on Eenal and Urinary Diseases," 
 
 Bristol, 1896, pp. 43, 272. 
 Lyon med., tome lxiii. Nos. 16-25. 
 Gaz. hebd. de med., Paris, 16th April 1896. 
 
 43. Ehrlich-Frerichs . Loc. cit., S. 272. 
 
 44. KtiLZ Arch. f. d. yes. Physiol., Bonn, 1876, Bd. 
 
 xiii.'s. 267. 
 
 29. Cantani, A. 
 
 30. Hermanides 
 
 31. Ord, W. M. 
 
 32. Feinberg . 
 
 33. Cantani, A. 
 
 36. Imlach, F. . 
 
 37. Senator, H. . 
 
 38. Westphal, C. 
 
 39. Frerichs, F. T. 
 
 40. Saundby, E. . 
 
 41. Glenard, F. . 
 
 42. Triboulet 
 
REFERENCES. 
 
 161 
 
 45. von Mering 
 
 46. Grube . . . 
 
 47. Frerichs, F. T. 
 
 48. Hoppe-Setler 
 
 49. Fleiner, W. . 
 
 50. Dickinson, W. H. 
 
 51. Garrod, A. . 
 
 52. Seegen, J. 
 
 53. Pavy, F. W. . 
 
 54. Purdy, C. W. 
 
 55. Frerichs, F. T. 
 
 56. Roberts, Sir ¥m., 
 
 and P. Maguire 
 
 57. Payer . . . . 
 
 58. Seegen, J. . . . 
 
 59. Bernardt . . . 
 
 60. Dickinson . . . 
 
 61. Frerichs, T. H. 
 
 62. Saundby, P. . . . 
 
 63. Klebs and Munk . 
 
 64. Poniklo, S. . . . 
 
 65. Cavazzani . . . 
 
 66. Lubrinoff . . . 
 
 67. Saundby, P. . . . 
 
 68. Hale White . . 
 
 69. Smith, P. Shingle- 
 
 ton 
 
 70. Windle, B. C. A. . 
 
 71. eulexberg . . . 
 
 72. Oppenheim . . . 
 
 73. Peumont . . . . 
 
 74. Bettmann . . . 
 
 Ibid., 1877, Bd. xiv. S. 284 (quoted by 
 
 Bunge, G., " Physiological and Pathological 
 
 Chemistry," translated by L. C. Wool- 
 
 dridge, London, 1890, p. 430). 
 Ztschr.f. Min. Med., Berlin, Bd. xxvii. 
 Loc. cit., S. 77. 
 Deutsches Arch. f. Min. Med., Leipzig, Bd. 
 
 hi. S. 171. 
 Berl. Min. Wchnsclir., 1894, Xos. 1 and 2. 
 "Diseases of the Kidney and Urinary 
 
 Derangements," Part I. ; " Diabetes," 
 
 London, 1875, p. 75. 
 Quoted by Dickinson, loe. cit., p. 76. 
 Loc. cit, S. 121. 
 Quoted by Dickinson, loc. cit. 
 Loc. cit., p. 37. 
 Loc. cit., S. 213 (footnote). 
 " Urinary and Renal Diseases," London, 1885, 
 
 p. 257. 
 Quoted by Roberts, loc. cit. 
 Loc. cit., S. 208-215. 
 " Symptomatologie und Diagnostik der 
 
 Hirngeschwiilste," Wiirzburg, 1865. 
 Loc. cit. 
 Loc. cit. 
 
 Loc. cit., p. 264. 
 Guy's Hosp. Rep., London, vol. xlvi. p. 42 ; 
 
 quoted by Hale White. 
 Lancet, London, 1878, vol. i. p. 268. 
 Centralbl. f. d. med. Wissensch., Berlin, 1894, 
 
 S. 623. 
 Virchow's Arcliiv, Bd. lxi. ; quoted by Hale 
 
 White, loc. cit. 
 Loc cit., p. 267. 
 
 Guy's Hosp. Rep., London, vol. xlvi. 
 Brit. Med. Journ., London, 1883, vol. i. 
 
 p. 657. 
 Dublin Journ. Med. Sc, 1883, vol. lxxvi. 
 
 p. 112. 
 Virchow's Arcliiv, Bd. xxviii. S. 26. 
 Berl. Min. Wchnschr., 1885, No. 49. 
 Ibid., 29th March 1886. 
 Munchen. med. Wchnschr., 1896, Nos. 49 
 
 and 50. 
 
i62 ETIOLOGY AND ETIOLOGICAL RELATIONS. 
 
 75. Hansemann . 
 
 76. Cawley, Thos. 
 
 77. Baumel . . 
 
 78. Minkowski und v. 
 
 Mering 
 
 79. Williamson, B. T. . 
 
 80. Harley, Vaughan . 
 
 81. Hoppe-Seyler . . 
 
 82. Lancereaux . . . 
 
 83. Tylden . . . . 
 
 84. Lepine . . . . 
 
 85. Lannois et Le- 
 
 MOINE 
 
 86. Williamson, B. T. . 
 
 87. Hansemann . . . 
 
 88. Kasahara . . . 
 
 89. Freyhan . . . . 
 
 90. Fleiner, W. . . . 
 
 Berl. Min. WcMsckr., 17th May 1897. 
 
 Lond. Med. Journ,, 1788, p. 286. 
 
 Montpel. vied., January and May 1882 ; 
 
 abstract in Jahresb. ii. d. Leistung . . . 
 
 d. rjes. Med., Berlin, 1882, Bd. ii. S. 223. 
 Arch. f. exper. Path. u. Pharmakol., Leipzig, 
 
 Bd. xxvi. ; Bd. xxxi. S. 85. 
 Med. Chron., Manchester, March 1892. 
 Brit. Med. Journ., London, 1892, vol. ii. 
 
 Med. Chron., Manchester, August 1895. 
 Deutsches Arch. f. Min. Med., Leipzig, Bd. 
 
 Hi. S. 171. 
 Union med., Paris, 1880, Kos. 13 and 16. 
 St. Barth Hosp. Rep., London, 1892. 
 Lyon med., 1890, No. 19. 
 Arch, de med. exper. et d'anat. path., Paris, 
 
 1891, No. 1. 
 Med. Chron., Manchester, March 1892 ; 
 
 Lancet, London, 14th April 1894. 
 Ztschr.f. Min. Med., Berlin, 1894, Bd. xxvi. 
 
 S. 191. 
 Virchow's Archiv, 1896, Bd. cxliii. S. 111. 
 Berl Min. Wchnschr., 1893, No. 6. 
 Ibid., 1894, Nos. 1 and 2. 
 
CHAPTER IX. 
 
 SYMPTOMATOLOGY. 
 
 The Appearance of the Patient. 
 
 There are two well-marked forms of diabetes, the severe 
 and the mild, besides intermediate varieties ; and the appear- 
 ance of the patient generally differs in the two forms of 
 the disease. In well-marked examples of the severe form the 
 patients are very often under middle age ; frequently they 
 are young persons ; they are thin, they have lost flesh, some- 
 times to a very great extent, and the skin is dry and harsh. 
 The appearance of the face is sunk and wasted, and yet the 
 facial expression differs from that of other diseases associated 
 with great wasting, such as phthisis and malignant disease. In 
 the severe forms of diabetes, the wrinkles of the face, about the 
 mouth, are often well-marked; the naso-labial fold is deep 
 or sharply denned, and it is frequently prolonged downwards, 
 well round the angle of the mouth. Sometimes the face is 
 pale ; but there is rarely any marked anaemia, unless the disease 
 be complicated with tuberculosis. In other cases, a very slight 
 degree of congestion or cyanosis is observed about the tip of the 
 nose, the cheeks, and the lips ; and sometimes the conjunctivas 
 appear abnormally moist. Though the face is wasted, there is 
 not the great ansemia of advanced phthisis. The lips are not 
 anaemic, and the pink flush of the cheek and the refined delicate 
 expression of the face, so common in phthisis, are absent in 
 the diabetic ; also the earthy cachetic appearance of the face of 
 malignant disease is absent. Patients suffering from the severe 
 form of diabetes often look older than their years, and the 
 appearance of the face is frequently suggestive of nervousness, 
 anxiety, or grief ( 1 ). The face in many cases of the severe form 
 of diabetes may be briefly described as anxious or sad, wasted, 
 but not ancemic, or not markedly anaemic. To define the facial 
 
1 64 SYMPTOMATOLOGY. 
 
 expression in diabetes of the severe type is difficult, yet I am 
 inclined to think there is often something peculiar in the 
 appearance of these cases. Whether this is so or not, on seeing 
 the new cases admitted to the wards of the Manchester Eoyal 
 Infirmary, the facial expression has on several occasions caused 
 a diagnosis of diabetes to flash across my mind before any ques- 
 tions had been asked, and examination has proved the suspicion 
 to be correct. 
 
 The facial appearance above described is not always observed, 
 even in severe cases of diabetes ; and in the mild forms of the 
 disease the patient is often well nourished, sometimes very 
 stout, and the facial expression is not in the least characteristic. 
 In these cases, though the patient is still well nourished, or even 
 stout, on inquiry we often find that there has nevertheless been 
 considerable loss of flesh. 
 
 In the severe cases with wasting there is often very great 
 loss of weight, and Saundby points out that there are often 
 marked and sudden fluctuations in the weight. 
 
 In any case of marked wasting in a young person, or of great 
 obesity, especially in an old person, an examination of the urine 
 for sugar should never be omitted. 
 
 Hanot and others have described a rare form of diabetes, 
 associated with pigmentation of the skin and hypertrophy of 
 the liver. The pigmentation is most marked on the face, limbs, 
 and genital organs. To this peculiar form Hanot has given the 
 name of cliaMte bronze 1 . It is described on p. 308. 
 
 Onset of Diabetes. 
 
 The patient sometimes first seeks medical advice on account 
 of loss of flesh, or of gradually increasing weakness. Pains in 
 the legs, especially cramps in the calf muscles, are sometimes the 
 earliest signs of the disease which attract attention. In many 
 cases, increase in the quantity of urine is the first symptom 
 the patient notices. He finds that he has to get up during the 
 night to micturate, whilst previously he had always been able 
 to sleep undisturbed. But in the majority of cases great thirst 
 is the first symptom, and it is a curious fact that sometimes the 
 thirst comes on very suddenly. The patient may be able to 
 state the date on which the thirst commenced, often the time of 
 the day, or the exact hour. (Nine out 100 cases, see p. 159.) 
 
ONSE T OF DIABE TES. i 65 
 
 Thus a most intelligent patient stated that the thirst com- 
 menced on September the 24th, in the afternoon. Another 
 diabetic stated that the thirst commenced suddenly one night 
 at 12 o'clock, whilst he was engaged on night duty at his work. 
 In a third case, it began suddenly in the evening, whilst the 
 patient was working unusually late. In a fourth, it began 
 suddenly between 8 and 9 o'clock in the morning. In two 
 other cases, the thirst came on very suddenly after drinking a 
 glass of beer. 
 
 A diabetic patient recently told me that he had been per- 
 fectly well until one evening when he took a great quantity of 
 beer ; he was carried to bed intoxicated, and next morning on 
 awaking he experienced a most intense thirst, which he has 
 never been able to quench since. 
 
 Another diabetic stated that he had gone to bed quite well 
 one evening, but next morning the " legs had shrunk," the " flesh 
 had shrunk," and he experienced an intense thirst, which he 
 was quite certain had not been present the day previously. 
 The thirst persisted ever afterwards, and the patient suffered 
 from a very severe form of the disease. 
 
 Sometimes, especially in the mild forms of the disease, the 
 patient first comes under treatment on account of some of the 
 complications of diabetes, such as soft cataract, carbuncle, 
 gangrene, irritation about the genitals, eczema of the vulva, or 
 skin diseases of various kinds. Occasionally diabetic patients 
 first seek advice on account of nervous depression, failure of 
 health, loss of strength, and inability to perform their daily 
 duties, the diuresis or thirst not having attracted their atten- 
 tion. These symptoms have often followed mental shock, or 
 great anxiety, or occasionally some bodily injury (see p. 107). 
 In the former cases, the state of nervous excitement or mental 
 depression has persisted for some time, and, on examination of 
 the urine, sugar has been detected. 
 
 One of my patients first consulted me on account of a sore- 
 ness of the throat and a salty taste in the mouth. Two weeks 
 previously, immediately after drinking half a glass of beer, he 
 had experienced a salty taste in the mouth, and this had con- 
 tinued ever since, and the throat had become dry. On examina- 
 tion of the urine, I found he was suffering from a severe form 
 of diabetes. 
 
 In certain cases the urine has contained sugar in small 
 
1 66 £ YMPTOMA TOL OGY. 
 
 quantities, or there has been a temporary glycosuria months 
 or years previous to the onset of severe diabetic symptoms. 
 Loeb ( 2 ) has recently drawn attention to this point. In one 
 of his cases, the urine contained 5 "3 per cent, of sugar, the 
 specific gravity was 1038, and there was a history of thirst 
 of only fourteen days' duration. But he happened to have 
 examined the urine two years previously, whilst the patient was 
 suffering from an attack of intercostal neuralgia. At that time 
 - 25 per cent, of sugar was present, but this had disappeared 
 at the end of nine months. Hence, long before the well-marked 
 symptoms of diabetes had developed, there was a diminished 
 power of utilising carbohydrates in the system, and slight 
 temporary glycosuria as a result. In a second case of diabetes, 
 the urine contained 7' 9 per cent, of sugar, and had a specific 
 gravity of 1042. The patient stated definitely that symptoms 
 of the disease had been present for four weeks only. But Loeb 
 happened to have made an examination of the urine five months 
 previously, and at that time had found traces of sugar present. 
 In a third case of diabetes, the urine contained 4 per cent, of 
 sugar. Loeb had found traces of sugar in the urine two years 
 and a quarter previously. Several similar cases have come 
 under my own observation. 
 
 On the other hand, it has also been clearly shown that 
 diabetes in its severe form is not always preceded by slight 
 glycosuria. Thus, Wallach ( 3 ) records a case of acute diabetes 
 in which the urine of the patient happened to have been 
 examined chemically a little more than five weeks before death. 
 No sugar was present at that time ; but after an attack of 
 bronchial catarrh, diabetes developed suddenly, and terminated 
 fatally five weeks after the urine had been examined. 
 
 The Ukine. 
 
 Quantity. — An increase in the amount of urine passed is 
 often the most obvious sign of diabetes ; and diuresis, and 
 especially the necessity for micturition during the night, are 
 frequently the first symptoms which attract the patient's atten- 
 tion to the disease. The daily quantity of urine varies con- 
 siderably, according to the severity of the case. The amount 
 may be 100, 150, or 300 oz. in the twenty-four hours. It 
 has been known to reach 640 oz., but in most cases the daily 
 
THE URINE. 
 
 167 
 
 quantity is between 100 and 300 oz. In many of the mild 
 cases, in old stout persons, the increase is not great (60 to 100 
 oz.), and in the mildest cases (chronic glycosuria) the urine is 
 often normal in quantity. Then again, in some of the severe 
 cases with marked wasting, which have had a very rapid course, 
 I have often found that the amount has not been more than 
 90 or 100 oz. 
 
 In most cases, but not in all, the night urine is less than 
 the day urine (whilst the opposite condition is met with in 
 granular kidney). 
 
 Generally the amount of urine increases with the percentage 
 of sugar which it contains. The more sugar is excreted, the 
 greater the quantity of urine, and vice versd. 
 
 This is well shown by the following figures from a case in 
 which the sugar and urine were very carefully estimated : — 
 
 Date. 
 
 Daily 
 
 Amount of 
 Urine. 
 
 Daily 
 
 Amount of 
 
 Sugar. 
 
 1 
 Remarks. 
 
 June 1 
 „ 2 
 „ 3 
 „ 4 
 „ 5 
 „ 6 
 » 7 
 „ 8 
 „ 9 
 „ 10 
 ,.- 11 
 „ 12 
 
 
 
 
 
 Oz. 
 230 
 240 
 200 
 150 
 112 
 150 
 150 
 159 
 188 
 258 
 220 
 220 
 
 Grs. 
 4600 
 4800 
 4200 
 2850 
 1792 
 1800 
 2400 
 1908 
 3384 
 5676 
 4840 
 4400 
 
 [-Ordinary diet. 
 
 I Very rigid nitrogenous 
 f and fatty diet. 
 
 VOrdinary diet. 
 
 But, as v. Noorden points out, there are exceptions to these 
 general rules, and the amounts of sugar and urine do not always 
 run parallel. 
 
 As the glycosuria increases, the amount of urine also 
 increases, but at a slower pace ; hence the urines which are 
 greatest in quantity have generally the highest percentage of 
 sugar. The urine may sometimes increase in quantity, however, 
 without the amount of sugar increasing, and again the urine 
 may diminish without a corresponding diminution of sugar 
 taking place. As Ktilz points out, two patients may excrete the 
 same amount of sugar, and very different quantities of urine. 
 
 The amount of urine is about equal to the quantity of fluid 
 taken. It is reduced by a rigid diet of nitrogenous food ; it is 
 
1 68 SYMPTOMATOLOGY. 
 
 also reduced by an intercurrent disease, as, for example, by any 
 febrile affection ; and it often diminishes just before a fatal 
 termination. There is frequently a diminution of urine just 
 before the onset of diabetic coma ; and I have often noticed, 
 when phthisis has developed as a complication, that towards the 
 termination of life the whole clinical aspect of the illness has 
 changed, the thirst has entirely disappeared, and the quantity of 
 urine has become quite normal. The spontaneous diminution of 
 thirst and diuresis, in an advanced case, is therefore by no 
 means a favourable sign, though it is very pleasing to the 
 patient. Apart from this final diminution, in some cases, from 
 time to time, there is a temporary return of the urine to the 
 normal quantity. Naturally, the amount of urine is less if 
 diarrhoea be present. 
 
 The colour of the urine is very pale — generally light yellow 
 or straw-coloured. Very often it has a slight pale greenish- 
 yellow tint. This greenish tint I have never noticed in any 
 other disease ; it may be detected best by placing the urine glass 
 on a sheet of white paper, and looking down from above ; it is 
 not always present, however, even in severe cases. Diabetic 
 urine is bright and clear, and froths in the urine vessel more 
 than normal urine. 
 
 When the quantity of urine excreted is not increased, or is 
 only a little increased, then the colour and appearance may be 
 those of normal urine. A normal colour generally indicates a 
 mild form of the disease ; also I have often observed a pale diabetic 
 urine to become darker, almost normal in tint, as fatal coma has 
 developed. In one case I noticed a slight pinkish tinge of the 
 urine during coma, and yet it contained no blood pigment, and 
 no deposit of urates (see p. 142). 
 
 When a mucous cloud is present in diabetic urine, it is 
 frequently seen, not at the bottom, but at the upper part of the 
 urine glass. 
 
 On standing, diabetic urine speedily becomes opalescent, 
 owing to the rapid development of yeast spores and other fungi. 
 
 Sometimes cases of glycosuria are met with in which fre- 
 quent micturition is a troublesome symptom, and yet, when the 
 total twenty-four hours' urine is collected, it is found that it is 
 not above normal ; the frequent micturition being due to the 
 irritation of the mucous membrane of the bladder by the sacchar- 
 ine urine. 
 
THE URINE. 169 
 
 The odour. — Diabetic urine has often a peculiar sweet or 
 aromatic smell, and when coma occurs, or when this termination 
 is threatening, generally the urine has a peculiar sweet or 
 chloroform-like smell — a smell supposed to be due to acetone. 
 
 The taste of the urine was formerly frequently noted in 
 diabetes. As Willis pointed out long ago, it is " wonderfully 
 sweet, as it were imbued with honey or sugar." 
 
 The reaction is usually acid. It is often strongly acid, and 
 the acidity increases at the onset of diabetic coma. When 
 allowed to stand, diabetic urine remains acid for many days, 
 and it may even increase in acidity, owing to the development 
 of lactic acid by fermentation. 
 
 The specific gravity is increased. It generally ranges between 
 1030 and 1045 or 1050; sometimes it is even higher still. In 
 any urine which is clear and not high coloured, if the specific 
 gravity be over 1025, there is the strongest probability that 
 sugar will be present. Whilst the specific gravity is raised if a 
 large quantity of sugar be present, we cannot conclude, if it be 
 low, that sugar will be absent. Urines are sometimes met with 
 in which the specific gravity is low, and yet a small amount of 
 sugar is present. 
 
 The specific gravity corresponds broadly to the amount of 
 sugar present, and is higher the greater the percentage of sugar, 
 and vice versd: but to this general rule there are many excep- 
 tions. A high specific gravity must not be regarded as a proof 
 of the presence of sugar, nor as an exact measure of the per- 
 centage of sugar, since it depends not only on sugar, but on the 
 urea and other solids of the urine. A high-coloured urine has 
 often a high specific gravity, though sugar is absent. 
 
 The presence of sugar is, of course, the most important 
 change in the urine in diabetes. Sometimes before the patient 
 has noticed any symptoms of the disease, he has been struck by 
 the fact that flies have been attracted to his urine. This has 
 been due to the sugar which it contained. In other cases the 
 pitient has noticed that if a drop of urine has fallen on his 
 boot, or on to any adjacent object, and has been allowed to dry, 
 a white salt-like deposit has been left. 
 
 The sugar present in diabetic urine is grape sugar (glucose, 
 dextrose). The percentage varies according to the nature of 
 the case, from 0'5 up to 8 or 12. The daily quantity of sugar 
 excreted also varies. It is often 3000 to 4000 ffrs. in the 
 
1 7 o S YMPTOMA TO LOG Y. 
 
 twenty-four hours, but may rise up to 12 or even 25 oz. It 
 has been stated to have reached 2-g- lb. in the twenty-four 
 hours. On the other hand, it may fall to 1 oz. daily. 
 
 Besides dextrose, lsevulose and other forms of sugar are 
 occasionally met with in the diabetic urine : but Seegen states that 
 the presence of ltevulose in the urine, with or without dextrose, 
 is exceedingly rare; he has found it only once in 1000 cases. 
 
 To estimate the amount of sugar, the total quantity of urine 
 for the twenty-four hours must be collected and mixed well, and 
 then a sample submitted to examination. 
 
 The sugar increases, in the mild cases, after food, and 
 diminishes during fasting ; and hence during the night the sugar 
 excretion is at its lowest. Kiilz has shown that in mild cases 
 the sugar excretion reaches its highest level between two and 
 three hours after a meal. 
 
 In very mild cases, sugar may be absent or greatly diminished 
 in the urine which has collected in the bladder during the night, 
 i.e. in the urine passed early before breakfast. 
 
 The sugar excretion is increased by starchy or saccharine 
 food, and diminished by nitrogenous diet. Hence, in comparing 
 the severity of two cases, or of the same case at different 
 periods of the disease, it is of importance to know the exact diet, 
 as a small amount of sugar, when the patient is on a nitrogenous 
 diet, is of much graver import than a large amount when the 
 patient is on a mixed diet. 
 
 The two well-marked types of diabetes, the mild and the 
 severe forms, present points of difference with respect to sugar 
 excretion. 
 
 In the mild cases, when carbohydrates are withdrawn from 
 the diet, the sugar disappears in a few days, and only when 
 carbohydrates are added again does sugar reappear in the urine. 
 In some cases complete withdrawal of carbohydrates is necessary 
 to cause the sugar to disappear entirely from the urine, but 
 other patients can take a small quantity of carbohydrates without 
 the glycosuria returning. In testing the tolerance in these cases, 
 a diet practically free from carbohydrates is given ; and when 
 sugar has disappeared from the urine, a little bread is allowed, 
 and this is gradually increased, the quantity being accurately 
 noted, until glycosuria reappears. The amount of bread added 
 is an indication of the tolerance for carbohydrates. It often 
 varies from time to time. 
 
THE URINE 171 
 
 These are points of practical importance with respect to 
 the examination of the urine in mild cases of diabetes, and 
 especially in the examination of the urine in cases of life assur- 
 ance. If the sample be taken from the morning urine, just 
 before breakfast, it may be free from sugar, and yet the patient 
 may suffer from glycosuria. Also, before testing the urine, it 
 ought to be known whether the diet is mixed or entirely nitro- 
 genous, as the latter diet may have caused a glycosuria to 
 disappear. If the urine passed two or three hours after a meal 
 of mixed diet be free from sugar, then glycosuria of any degree 
 may be excluded. 
 
 In the severe cases of diabetes sugar is present in the urine, 
 in spite of the withdrawal of all carbohydrate food. In some 
 cases, when the patient is taking a restricted diet, containing, 
 however, a large quantity of flesh meat (1000 grms. daily), 
 sugar is present in the urine ; but when the flesh meat is 
 reduced to 500 grms., the glycosuria disappears (v. Noorden). 
 
 When a diet is prescribed free from carbohydrates, the amount 
 of sugar remaining in the urine is a measure of the severity 
 of the disease. In severe cases, sugar is present in the urine 
 even during fasting. Schmitz ( 4 ) points out that in the severe 
 forms the night urine {i.e. that passed early in the morning 
 before breakfast) may contain the greatest amount of sugar 
 (see p. 307). 
 
 In the mild forms of diabetes only the sugar derived from 
 the food appears in the urine ; in the severe cases sugar appears 
 to be formed from albuminous bodies in the system. 
 
 Influence of food on sugar excretion- — No case is so severe 
 that no portion of the carbohydrates of the food is burnt up in 
 the system ; but some carbohydrates are utilised better than others. 
 Grape sugar (dextrose) is the carbohydrate which causes sugar 
 to be eliminated in the greatest quantity in the urine ; whilst 
 starch and other carbohydrates are less injurious. Fruit sugar 
 (kevulose) is only about half as injurious as grape sugar ; milk 
 sugar and cane sugar stand intermediate between grape sugar 
 and kevulose. Fats never increase the sugar excretion, either 
 in human diabetes, or in the pancreatic diabetes, or phloridzin 
 diabetes of animals. Alcohol in moderate quantity does not 
 increase the amount of sugar in the urine. 
 
 Muscular exercise diminishes the sugar excretion in well- 
 nourished patients at an early stage of the disease; the sugar 
 
172 S YMPTOMA TOL OGY. 
 
 being apparently burnt up in the muscles ; and Finkler ( 5 ) has 
 shown that massage has the same action. But in other cases, 
 when the patient is wasted and the disease chronic, exercise 
 increases the sugar excretion (Kiilz). 
 
 It has been pointed out that mental shock, etc., may act as 
 an exciting cause of the disease ; clinical experience also shows 
 that cases of diabetes are made worse, and the sugar excretion 
 greatly increased, by nervous excitement, mental worry, etc. 
 The sugar excretion is increased by a long journey, and 
 after admission to a hospital, frequently it is higher on the 
 first day than a few days later. Gastro-intestinal symptoms, 
 loss of appetite, and diarrhoea cause a diminution of the 
 glycosuria. 
 
 Diminution or arrest of sugar excretion by intercurrent affection. 
 — It has long been known that the sugar excretion may 
 be diminished or entirely checked by certain intercurrent 
 diseases. 
 
 (a) Febrile affections often have this effect, as, for example, 
 typhoid fever, pneumonia, influenza, and chronic febrile con- 
 ditions. Thus, in a case of diabetes of the mild form, in a very 
 stout elderly woman at the Manchester Eoyal Infirmary, the 
 sugar disappeared from the urine during an attack of influenza, 
 to the great delight of the patient, but returned again when the 
 symptoms of influenza subsided. 
 
 (b) When phthisis occurs as a complication of diabetes, the 
 sugar excretion often diminishes, and sometimes ceases entirely 
 for a short time before death (see cases on p. 211). 
 
 (c) When ascites, due to cirrhosis of the liver, is present in 
 diabetic patients, sugar sometimes disappears from the urine. 
 
 Pusinelli ( 6 ) records an instructive case of diabetes associated 
 with cirrhosis of the liver. The amount of sugar in the urine 
 was from 2 to 5 per cent. ; but when ascites developed, the glycos- 
 uria entirely ceased. After paracentesis for the fourth time, the 
 ascites disappeared, and then glycosuria returned. Two and a 
 half years later the ascites reappeared, and then the glycosuria 
 diminished. 
 
 (cl) During an attack of gout, sugar may disappear from the 
 urine. 
 
 (e) Also during an attack of jaundice, the sugar has been 
 known to disappear. 
 
 (./) When granular kidney is associated with diabetes 
 
DIMINUTION OF SUGAR EXCRETION 173 
 
 mellitus, at the terminal stage the sugar excretion may entirely 
 cease (see p. 218). 
 
 (g) Pneumaturia, an exceedingly rare complication in diabetes 
 mellitus, may cause the glycosuria to disappear. Both in diabetic 
 and non-diabetic patients, pneumaturia may be due to the en- 
 trance of air into the bladder from the intestine, owing to a 
 fistulous connection between the two. In these cases, if the 
 patient be diabetic, the sugar is still present. But pneumaturia 
 is occasionally the result of the decomposition of sugar in the 
 bladder, owing to the presence of micro-organisms or yeast fungi. 
 In such cases the sugar disappears from the urine, and gases 
 (carbonic acid, hydrogen, and carburetted hydrogen) are formed 
 in the bladder. 
 
 (h) In diabetic coma, often the amount of sugar diminishes 
 considerably. 
 
 It is important to bear in mind that a diminution of the 
 thirst, diuresis, and sugar excretion is not necessarily a favour- 
 able sign in the most severe forms of diabetes ; on the contrary, 
 it is sometimes a serious indication. In such cases, if the general 
 condition is becoming worse, and the weight of the patient 
 diminishing, then the prognosis is bad whatever may be the 
 change in the sugar excretion, diuresis, and thirst. Very often 
 the chart shows a pleasing diminution of sugar and urine, and 
 yet the general condition may be gradually becoming worse. On 
 the other hand, in severe cases at an advanced stage, by a less 
 rigid diet the sugar may be increased, but often the general 
 condition is improved. 
 
 Glycogen. — According to Leube ( 7 ), the urine of diabetic 
 patients contains a small quantity of a substance which gives 
 a dark brown reaction with iodine and iodide of potash, and 
 which is converted into grape sugar by boiling with dilute 
 sulphuric acid. This substance he regards as glycogen. In 
 normal urine, and in the urine of a case of diabetes insipidus, 
 no glycogen was found. 
 
 Urea. — The amount of urea in the urine is generally stated 
 to be increased in diabetes. The increase, however, is largely 
 due to the great quantity of nitrogenous food taken. In order to 
 decide whether the diabetic patient excretes more urea than a 
 healthy person, it is necessary to keep both on exactly similar 
 diet. The experiments of Pettenkofer and Voit have shown that, 
 even when fasting, a diabetic patient excreted about 8 per cent. 
 
i74 £ Y MP TO MA TO LOGY. 
 
 more urea than a healthy person. When both were on the 
 same medium diet, once the nitrogen excreted was the same in 
 each case, three times it was higher in the diabetic patient. 
 When both were fed on a non-nitrogenous diet, the nitrogen 
 excreted was considerably greater in the case of the healthy 
 person than in the diabetic patient. 
 
 Gaethgens also found that diabetics excrete more nitrogen 
 than healthy persons, when both are taking the same diet. 
 
 Seegen ( 8 ), from his own observations, draws the following 
 conclusions : — (a) The urea excretion is increased (but generally 
 not markedly) in almost all cases of diabetes, (b) There is no 
 relation between the excretion of urea and sugar, (c) The urea 
 excretion is chiefly dependent upon the nitrogen of the food : 
 the richer the diet in nitrogenous matter, the greater is the 
 excretion of urea. In only a few cases is the urea excreted so 
 abundantly that it is necessary to refer its origin to the destruc- 
 tion of albumin of the body. 
 
 Leo ( 9 ) has shown that carbohydrates diminish the nitro- 
 genous metabolism. In two severe cases of diabetes, the patients 
 were kept for some days on uniform diets, rich in albumin, but 
 containing very little carbohydrate. The nitrogenous excretion 
 was estimated, and, after this had become stationary, a definite 
 quantity of carbohydrates was allowed in addition to the previous 
 diet. The nitrogen in the urine and faeces was again carefully 
 estimated, and the results showed that the nitrogenous excretion 
 in the urine was diminished distinctly when carbohydrates were 
 added to the diet. 
 
 Uric acid. — Not infrequently a small deposit of uric acid 
 crystals is seen at the bottom of the urine glass, especially in 
 mild cases of diabetes. Seegen and Pavy both point out that a 
 dark urine and a deposit of uric acid are indications of a mild 
 form of the disease. In the severe cases it is somewhat difficult 
 to estimate the amount of uric acid, since it is dissolved in a 
 very large quantity of fluid, and, as Seegen points out, the usual 
 method of estimation is not satisfactory. 
 
 Klilz has found that, in the severe cases of diabetes, the uric 
 acid (as estimated by a method of his own) is a little less than 
 the normal quantity. 
 
 Ammonia. — In health, according to Hallervorden ( 10 ), 0*5 to 
 1*0 grm. of ammonia is excreted in the urine daily, and on a 
 meat diet the amount may increase to 1*2 or 1'5 grm. (v. Noorden). 
 
URIC ACID, AMMONIA, OXALATES. 175 
 
 In many cases of diabetes, but not in all, the excretion of 
 ammonia is increased, and may equal from 3 to G grins. A 
 moderate increase may be present for weeks or months (v. Noor- 
 den). In diabetic coma the amount is greatly increased, but the 
 urine has still an acid reaction ; and, as Stadelmann ( n ) points 
 out, since the basic elements are out of proportion to the acids 
 known to us, it is probable that some unusual acid is present in 
 these cases. Stadelmann thought that this acid was crotonic 
 acid ; Minkowsky and Kiilz believed it to be /S-oxy butyric 
 acid. 
 
 The kreatine of the urine is greatly increased in diabetes, 
 owing to the excess of nitrogenous food, and to the increased 
 muscular wasting. 
 
 Oxalates. — Sometimes an abundant deposit of oxalate of lime 
 is found in diabetic urine, both in mild and severe cases. 
 
 Inorganic salts. — Sodium chloride, the sulphates and phos- 
 phates, are increased, owing to the large amount of nitrogenous 
 food and to the great destruction of albuminoids. Many 
 observers ( 12, 13 ) have shown that the excretion of lime salts 
 in the urine is increased in diabetes. Eecently Teubaum ( 14 ) 
 has reinvestigated this point. From the results of his observa- 
 tions on twelve cases, he concludes that in the severe form of 
 the disease the excretion of lime salts is considerably increased, 
 whilst in mild forms the excretion is normal, or only a little 
 above the normal amount. 
 
 The proportion of total solids to the water is diminished. 
 Indican is often present in excess. 
 
 In diabetic urine generally there is no deposit at the bottom 
 of the urine glass, but, as already mentioned, occasionally there 
 is a deposit of uric acid, or oxalates, or a mixture of mucus 
 and phosphates. 
 
 Albuminuria. — Sometimes albumin as well as sugar is found 
 in the urine of diabetic patients. The proportion of such cases 
 is estimated differently by various authors — from 5 to 66 per 
 cent. Schmitz ( 15 ) found albuminuria present in 824 out of 
 1200 diabetic cases. But some authors include in their statis- 
 tics of albuminuria only those cases in which there is distinct 
 evidence of nephritis ; whilst others include all cases in which 
 there is the slightest trace of albumin. Much depends also on 
 the class of cases from which the statistics are drawn. 
 
 The albuminuria of diabetes may exist in two forms — (1) 
 
176 £ YMPTOMA TOL OGY. 
 
 In the most common form only a small quantity or a trace of 
 albumin is detected, and there are no other indications of kidney 
 mischief. (2) In the second form, in addition to the albumin- 
 uria, there are other signs of Bright's disease (oedema, headache, 
 albuminuric retinitis, cardio-vascular changes, etc.). There may 
 be all the indications of parenchymatous nephritis, and a large 
 quantity of albumin along with casts may be found in the urine ; 
 the albumin may gradually increase, and the sugar gradually 
 diminish, until finally the sugar disappears, and the albumin only 
 remains. In other cases, symptoms of granular kidney are 
 present ; the specific gravity of the urine diminishes, and finally 
 the sugar often disappears. The disappearance of the sugar and 
 the change of the general symptoms to those of chronic Bright's 
 disease are grave prognostic signs. 
 
 In the second group of cases, the kidneys, on post-mortem 
 examination, present well-marked signs of chronic parenchy- 
 matous or interstitial nephritis, but in the first group only 
 slight and microscopical changes are met with. These are 
 described on p. 220. 
 
 Bouchard points out that a slight albuminuria is often 
 observed when phthisis occurs as a complication of diabetes. 
 Also in elderly diabetics a small quantity of albumin is not 
 infrequently met with in the urine, and is often accompanied by 
 an excess of uric acid. 
 
 In the majority of cases of diabetes met with in hospital 
 practice in Manchester (most of which belong to the severe 
 form of the disease, accompanied by marked wasting), not a 
 trace of albumin can be found in the urine when the patient 
 comes under treatment, but, as the disease advances, albumin 
 appears. At first only the slightest trace can be detected, but 
 in time the amount of albumin often increases a little ; it nearly 
 always remains small in quantity, however, and other indications 
 of nephritis do not develop. This small quantity of albumin often 
 disappears and reappears. In the last stage of these severe 
 cases, a trace or small quantity of albumin is nearly always 
 present. 
 
 A large amount of albumin and other indications of nephritis 
 are not common in cases of diabetes seen in hospitals, but in 
 private practice, where many cases of the mild form of diabetes 
 in elderly people are met with, this complication is more 
 frequent. 
 
ALBUMINURIA. 
 
 177 
 
 Albuminuria in 100 Cases of Diabetes. 
 
 (Mostly hospital patients, suffering from a severe form of the disease, and often at an 
 
 advanced stage.) 
 
 
 Remarks. 
 
 Albumin absent when urine first 
 examined in . 
 
 Cases. 
 70 
 
 In fourteen of these a trace of albumin 
 appeared at a later date. 
 
 Small quantity or a mere trace of 
 albumin in 
 
 26 
 
 Seven of these were at a very advanced 
 stage ; one was comatose when first 
 examined. In two the albumin dis- 
 appeared at a later date. 
 
 Large or considerable quantity of 
 albumin in 
 
 4 
 100 
 
 
 Albumin present in 30 per 
 ,, ,, 42 pei 
 
 cent, of patients when first examined, 
 cent, at a later stage. 
 
 In ninety-six of the hundred cases there were no indications 
 of nephritis or organic disease of the kidney. 
 
 With respect to the traces of albumin, it is well to bear in 
 mind that not infrequently the saccharine urine gives rise -to 
 irritation of the prepuce, and balanitis results ; and in the female, 
 inflammation of the vulva may arise from the same cause. A 
 little purulent discharge becomes mixed with the urine in either 
 case, and occasionally this is the cause of a trace of albumin in 
 diabetic urine ; but no doubt in many cases the albumin is of 
 renal origin. 
 
 In all the cases of diabetic coma which I have examined, 
 the urine has contained a trace or small quantity of albumin 
 (see p. 285), and Maguire ( 16 ) states that albuminuria is always 
 present in diabetic coma. The urine may be quite free from 
 albumin shortly before the onset of coma. Sometimes the 
 appearance of albumin is the forerunner of coma ; but, on the 
 other hand, a trace of albumin is often present for a long 
 period, and coma may never develop, and the patient may die 
 of phthisis or some complication. 
 
 Casts. — With the exception of the cases in which diabetes 
 is complicated by parenchymatous or interstitial nephritis, casts 
 are not usually detected in the urine, even in the severe forms 
 1 2 
 
i/8 SYMPTOMATOLOGY. 
 
 when a trace or small quantity of albumin is present. But when 
 diabetic coma occurs, as Klilz( 1,,ls ) has pointed out, an abundant 
 deposit of casts is met with. In these cases the urine is often 
 slightly turbid when first passed, from the presence of minute 
 floating particles. On standing, a greyish-white sediment is 
 deposited at the bottom of the urine glass, and if a little be 
 examined under the microscope, it is found to consist of enormous 
 numbers of casts. Often a remarkable appearance is presented, 
 the whole field of the microscope being crowded with tube-casts. 
 The casts are granular or hyaline, studded with fine granules. It 
 is important to watch the urine in advanced or severe cases of 
 diabetes for the appearance of this greyish-white or flocculent 
 deposit. It has been already pointed out that there is occasion- 
 ally a deposit of mucus and phosphates in diabetic urine ; but 
 usually it does not sink quite to the bottom of the urine glass. 
 The deposit of casts, however, lies exactly at the bottom of the 
 glass, it is more of a greyish colour, and more opaque than the 
 mucus deposit. If a deposit should appear in the urine in severe 
 or advanced diabetes, it ought to be examined carefully for casts ; 
 when detected they are generally, though not invariably, a sign 
 of approaching coma (see p. 285). 
 
 Diabetic mine, when allowed to stand, soon becomes opalescent, 
 owing to the development of yeast fungi; and when kept for 
 some time a white deposit may form, which, under the micro- 
 scope, is found to consist of fungus spores. But occasionally, 
 even when recently passed, the urine contains fungus spores and 
 mycelia with a few pus cells and epithelial scales. As will be 
 mentioned on p. 224, in diabetes sometimes an inflammation of 
 the prepuce in the male, or of the vulva in the female, is 
 produced by the constant irritation of saccharine urine. In 
 these cases there is often a growth of fungus on the inflamed 
 mucous membranes of the genital organs, and from these inflamed 
 parts, fungus spores and mycelia, pus cells and epithelial scales 
 may be washed away, and give rise to a slight depositor turbidity 
 in the urine. 
 
 Thus, in a case of diabetes of four years' duration, the urine, 
 when recently passed, presented a slight turbidity, and numerous 
 flocculent masses were seen suspended in it. Under the micro- 
 scope I found these flocculent masses to consist of epithelial 
 scales, a few pus cells, fungus spores, and mycelia. The urine 
 gave the very slightest reaction for albumin. An examination 
 
CASTS IN THE URINE. 179 
 
 of the genital organs revealed oedema of the prepuce and penis, 
 with marked phimosis and balanitis. The end of the prepuce 
 was much inflamed, and coated with a number of white patches. 
 When scraped away and examined microscopically, these white 
 patches were seen to consist of epithelial scales, mixed with 
 fungus spores and mycelia. The epithelial scales, fungus spores, 
 and mj'celia in the urine had evidently been washed away from 
 the penis. 
 
 In another case (a youth of 19) the recently passed urine 
 contained a small white sediment. On examination microscopic- 
 ally, I found it to consist of epithelial squames, with fungus 
 spores and mycelia. This patient was suffering from balanitis, 
 and on the mucous membrane of the glans penis a few 
 white patches were seen ; when scraped away and examined 
 microscopically, they were found to consist of epithelial squames, 
 pus cells, fungus spores, and mycelia. The fungus spores and 
 mycelia in the urine were evidently derived from these patches. 
 
 In a third case a diabetic urine was slightly turbid and 
 a trace of albumin was present. Microscopical examination 
 revealed the presence of a considerable number of pus cells and 
 epithelial scales. On examination of the penis, a slight balanitis 
 was found, and there were a few white patches on the glans penis. 
 By strict cleanliness the balanitis disappeared in a few days, and 
 the urine became quite free from turbidity and albumin. 
 
 I have sometimes found the urine of female diabetic patients 
 slightly turbid, owing to the presence of numerous epithelial and 
 pus cells. In these cases there has been great irritation of the 
 vulva by the saccharine urine. 
 
 Gerhardt's reaction ; brownish-red coloration with perchloride 
 of iron ; diacetic acid. — In the severe forms of diabetes, especi- 
 ally in young patients, or in patients who are markedly wasted, 
 the urine often gives a dark brownish-red coloration with a 
 solution of perchloride of iron. In these cases a few drops of the 
 liquor ferri perchloridi (B.P.) added to the diabetic urine, generally 
 causes a slight white precipitate of phosphates at first, which 
 disappears on adding a little more of the perchloride solution, 
 whilst the fluid becomes of a dark brownish-red or Bordeaux-reel 
 colour (Burgundy wine colour). The coloration is usually 
 attributed to the presence of diacetic acid. On this point 
 there is a little difference of opinion, however, and some authors 
 believe that it is not due to diacetic acid, but to the presence 
 
i So SYMPTOMATOLOGY. 
 
 of an organic compound closely allied to it. The exact cause of 
 the perchloride of iron reaction is fully discussed by Macmunn in 
 his work on the " Clinical Chemistry of Urine " ( 19 ). 
 
 Salkowski could not obtain the perchloride of iron reaction 
 in the ether extract from the urine, but could do so in the ether 
 extract when (for a control experiment) diacetic acid was added 
 to normal urine. 
 
 The urine has frequently a peculiar smell resembling chloro- 
 form — the so-called acetone smell — when the reaction with 
 perchloride of iron is obtained. 
 
 It is often stated that if the urine be boiled before the 
 perchloride of iron solution is added, then no reaction occurs ; 
 also that the colour produced by perchloride of iron disappears 
 on boiling. But boiling for a few minutes is not sufficient. 
 In many cases which I have examined, I have found, after 
 vigorous boiling for several minutes, that the urine gave a much 
 fainter reaction when perchloride of iron was added, but a slight 
 brown colour was still obtained. Also by vigorously boiling the 
 urine, for a few minutes only, after the perchloride of iron has 
 been added, I have found that the dark brown colour diminishes, 
 but does not entirely disappear. For the detection of diacetic 
 acid in the urine, v. Jaksch ( 20 ) gives the following process :- — 
 
 " To the urine a fairly concentrated solution of perchloride 
 of iron is cautiously added, and if a phosphatic precipitate 
 forms, this is removed by nitration and more of the perchloride 
 of iron solution supplied. If the Bordeaux-red coloration appears, 
 one portion of the urine is boiled, whilst another is treated with 
 sulphuric acid and extracted with ether. If, now, the urine which 
 has been boiled shows little or no change, whilst the perchloride 
 of iron reaction in the ethereal extract is no longer evident after 
 twenty-four to forty- eight hours ; and if at the same time (on 
 testing the urine directly and its distillate) it is found to 
 be rich in acetone, the condition may be inferred to be that of 
 diaceturia." 
 
 This coloration with perchloride of iron is not found in all 
 cases of diabetes ; it is not present in all severe cases, but it is 
 present in a large proportion of such cases at a late period 
 of the disease ; whilst in the mild form of diabetes, and in 
 the earlier stages of the severe form, it is generally absent. 
 It is usually, though not invariably, associated with marked 
 constipation. The reaction is always an indication of a seveie 
 
GERHARDT'S REACTION. 181 
 
 form of the disease, and in cases in which it is obtained it is 
 important that the diet should not be too rigid, i.e. that carbo- 
 hydrates should not be excluded entirely. If this be done 
 there is considerable risk of diabetic coma developing. The 
 reaction with perchloride of iron often varies very much from 
 time to time. In severe forms of diabetes, when phthisis has 
 developed as a complication, sometimes the sugar excretion 
 finally ceases, and a reaction with perchloride of iron can no 
 longer be obtained. In diabetic coma the urine generally gives 
 the reaction with perchloride of iron, but not invariably. I have 
 often found that the reaction has become more and more marked 
 for a few days before the onset of coma, but when the comatose 
 symptoms have developed the reaction has become a little less 
 marked. 
 
 This brown-reel coloration with perchloride of iron is occa- 
 sionally met with in other diseases besides diabetes, though 
 such cases are rare. It has been observed sometimes in febrile 
 affections, and in acute diseases, in measles, scarlet fever, pneu- 
 monia, etc. It has also been met with occasionally in cases of 
 cancer of the stomach, and cancer of other organs, in Bright's 
 disease, perityphlitis, strangulated hernia, hysterical anorexia 
 with vomiting, and, according to Dreschfeld, occasionally 
 in Graves' disease. A similar coloration is obtained with 
 /3-oxybutyric acid. The urine of patients who are taking 
 salicylic acid, salicylate of soda, or salol, contains salicyluric acid 
 and gives a purplish-brown coloration with perchloride of iron. 
 This resembles somewhat the above mentioned brownish -red 
 coloration obtained in diabetic urine, and often it is inferred that 
 the urine of these patients contains cliacetic acid. But if the 
 coloration be carefully noted, a difference will be readily detected 
 by the naked eye ; that given by perchloride of iron in the urine 
 of patients taking sodium salicylate has much more of a violet 
 or purple tint, that given with diabetic urine has more of a 
 brownish colour. The difference in tint is very marked if the 
 contents of the test tube be placed in a urine glass, and well 
 diluted with water in eacli case. 
 
 A coloration resembling Gerhardt's reaction is obtained with 
 perchloride of iron in the urine of patients taking other drugs — 
 antipyrin, thalline, phenocoll, salipyrin. 
 
 Acetone. — Urine which contains diacetic acid, or which 
 gives a brownish-red coloration with perchloride of iron, generally 
 
1 8 2 S YMPTOMA TOL OGY. 
 
 contains a considerable amount of acetone also. The following 
 are the chief clinical tests for this substance : — 
 
 Legal's test. — To several c.c. of urine in a test tube a few 
 drops of a concentrated, freshly made solution of sodium nitro- 
 prusside are added. The mixture is then made alkaline with 
 liquor potassee. A red coloration is produced which soon dis- 
 appears ; but when a little acetic acid is added, a deep violet-red 
 coloration is the result, if acetone be present. If acetone be 
 absent, the urine has its usual yellow colour after adding the 
 acetic acid. 
 
 Le Nobel's test is also useful ( 21 ). To an ounce of urine add a 
 drachm or two of a solution of nitroprussicle of sodium (5 grs. 
 to the ounce), and then a few drops of liquor ammonia? ; dilute 
 with water, and allow to stand. After a short time a pinkish 
 violet coloration will develop. 
 
 The production of iodoform from the distillate of the urine is 
 the most delicate, test for acetone. About half a litre of urine is 
 placed in a retort and distilled, a little dilute phosphoric acid 
 being added to prevent too great effervescence of the liquid. 
 (The reaction may also be obtained from a few ounces of urine.) 
 To about 20 c.c. of the distillate a few drops of a solution of 
 iodine in potassium iodide are added, and a few drops of a 
 solution of caustic potash. When acetone is present, a precipi- 
 tate of iodoform will occur. It is light yellow in colour, and has 
 the characteristic smell of iodoform, which can be detected 
 readily. Under the microscope this precipitate is seen to consist 
 of hexagonal plates, or stars or rosettes of the hexagonal system. 
 
 Other substances in the urine, such as alcohol and lactic acid, 
 give a similar reaction, however. 
 
 Kalfe has proposed the following test for acetone : — Dis- 
 solve 4 drms. of potassium iodide in 1 oz. of liquor potassa?. 
 Place 1 drm. of this solution in a test tube, and boil. Then 
 very carefully add a drachm of urine so as to float on the surface 
 of the alkaline iodide solution. At the line of junction a white 
 cloud of phosphates will form, which after a short time will 
 become yellow if acetone be present. The feathery phosphates 
 will be seen tipped with yellow points where iodoform is 
 deposited ; these after a time fall through the cloud, and deposit 
 at the bottom of the test tube. 
 
 v. JSToorden ( 22 ) gives the following method for the quantitat- 
 ive estimation of acetone : — 
 
TESTS FOR ACETONE. 183 
 
 To 100 c.c. of urine, 1 c.c. of concentrated glacial acetic 
 acid is added, and the mixture distilled until three-fourths of 
 the fluid have passed over. To the distillate five drops of dilute 
 sulphuric acid (25 per cent.) are added. Urea crystals are 
 removed, and after cooling well the fluid is again distilled. To 
 the second distillate caustic potash is added, and then an excess 
 of a solution of iodine in potassium iodide. A precipitate of 
 iodoform forms. This is collected, after six hours, on a weighed 
 filter and washed with distilled water. After the filter and 
 precipitate have become dry, they are placed in an exsiccator 
 with sulphuric acid, and weighed after one and a half to two 
 hours. One grm. of iodoform corresponds to 0*147 grm. of 
 acetone. 
 
 The urine and the breath of patients suffering from the 
 severe forms of diabetes have often a peculiar sweet smell — 
 sometimes described as a chloroform smell. This smell re- 
 sembles that of acetone ; and acetone may be detected chemically 
 both in the urine and in the breath in such cases. This acetone 
 smell of breath and urine is generally noticed in diabetic coma. 
 A similar smell of the breath is sometimes met with in febrile 
 affections. 
 
 During the last ten years the occurrence of acetone and 
 diacetic acid in the urine (acetonuria and diacetonuria) has been 
 carefully studied by v. Jaksch, Biermer, Hirschfeld, Eosenfeld, 
 and many others. According to v. Jaksch, normal urine con- 
 tains minute traces of acetone. It is very often present in large 
 quantity in diabetic urine, especially in the severe forms of the 
 disease and in diabetic coma. In the mild forms of the disease 
 both acetone and diacetic acid are absent. Acetone is also 
 present in some mental disorders, and in these cases may be 
 the result of a gastro-intestinal auto-intoxication. Acetonuria 
 occurs sometimes in febrile affections, such as typhoid, scarlet 
 fever, measles, in some cases of cancer of the stomach and 
 other organs, in some cases of gastric disorders, and in 
 starvation. It has been attributed to increased albumin de- 
 struction in the system, and to fermentative changes in the 
 alimentary canal. 
 
 Hirschfeld ( 23 ) has shown that in healthy persons acetone is 
 excreted in quantity in the urine, if the diet consist only of 
 albuminous and fatty food ; but it is at once reduced (to the 
 normal trace) if carbohydrates be added to the diet. The action 
 
1 84 SYMPTOMATOLOGY. 
 
 of carbohydrates is to spare the destruction of albumins ; but 
 a quantity of carbohydrate food, too small to influence the 
 albumin metabolism very much, has a marked action on the 
 acetone excretion. Also, when the destruction of albumin is 
 spared, and the loss of albumin has ceased, owing to the addition 
 of fatty food to a nitrogenous diet, Hirschfeld has found that the 
 abundant acetone excretion still continues. It appears, therefore, 
 that the absence of carbohydrates from the diet is a cause of 
 acetonuria. The acetonuria of fevers, cancer, or gastritis dis- 
 appears when carbohydrates are taken and assimilated (Hirsch- 
 feld). 
 
 Rosenfeld ( 24 ) has shown that if a healthy person be put on 
 an exclusively nitrogenous diet, which is commenced in the 
 morning, then acetonuria occurs by the evening of the second or 
 by the morning of the third day. He found that the acetonuria 
 produced by a flesh diet was removed by the addition of 
 carbohydrates, and he points out, also, that whenever there is 
 great destruction of the albumin of the body, it occurs if 
 carbohydrates be excluded from the food. It is present, for 
 example, in starvation, or when a purely nitrogenous diet is 
 taken. In the mild form of diabetes, acetonuria is only pro- 
 duced by a nitrogenous diet, just as in healthy persons ; but 
 in the most severe cases, it occurs in spite of carbohydrates 
 being present in the diet. In severe cases of diabetes there is 
 then a true pathological acetonuria, — an acetonuria which cannot 
 be removed by the addition of carbohydrates to the food, — and 
 often there is also diacetonuria. A strictly nitrogenous diet 
 generally produces acetonuria in healthy persons, but in such 
 cases diacetonuria rarely occurs, and, if present, it is very slight. 
 In diabetic coma, acetone is almost invariably present in the 
 urine ; and it has often been observed that, after giving 
 an exclusively nitrogenous diet, in severe cases of diabetes, 
 acetonuria and coma have developed. 
 
 According to Hirschfeld, a high degree of acetonuria, or an 
 increase of the acetone in the urine, are indications of approach- 
 ing coma. 
 
 A number of observations, recorded during the last five years, 
 appear to show that, in patients who have shown signs of 
 commencing coma, these symptoms have sometimes been arrested 
 by the addition of a small quantity of carbohydrates to the diet 
 (see p. 329). 
 
OXYBUTYRIC ACID. 185 
 
 fi-oxybutyric acid. — This acid is found in the urine in certain 
 cases of diabetes of the severe form, and its presence in consider- 
 able quantity {i.e. more than several grammes per diem) is of 
 grave prognostic significance. In most of these cases, if the 
 excretion of /3-oxybutyric acid continues for a few days or weeks, 
 diabetic coma develops. Though this is the general rule, there 
 are exceptions, and occasionally the acid is excreted for a long 
 period without the development of coma. 
 
 According to v. Noorden ( 25 ), the presence of /3-oxybutyric 
 acid may be inferred — 
 
 (1) When the amount of sugar indicated by the titration 
 method with Fehling's solution is considerably higher than the 
 estimation by polarisation. Grape sugar rotates the plane of 
 polarised light to the right, oxybutyria acid to the left ; hence, 
 when the latter is present, part of the dextro-rotatory power 
 of the grape sugar is neutralised, or, if a great quantity of 
 oxybutyric acid is present, the urine may even rotate the plane of 
 polarised light to the left. 
 
 (2) When the urine, after complete fermentation with yeast, 
 rotates the plane of polarised light to the left. 
 
 (3) When the urine, after precipitation with basic acetate of 
 lead and ammonia, rotates polarised light to the left. By this 
 treatment the grape sugar is precipitated whilst oxybutyric acid 
 remains in solution and passes over in the filtrate. 
 
 v. Noorden points out that positive results with these tests 
 render the presence of oxybutyric acid exceedingly probable. 
 The patient ought not to be taking benzosol, however, since this 
 substance renders the urine levo-rotatory. The presence of 
 leevulose in the urine does not lead to any fallacy, if the second 
 and third tests give a positive result. 
 
 In order to detect /3-oxybutyric acid with absolute certainty, 
 however, other more complicated methods must be employed. 1 
 
 The Blood. 
 
 A drop of blood, obtained by pricking the finger of a diabetic 
 patient, does not generally present any change to the naked eye. 
 It has sometimes appeared to me, however, that diabetic blood 
 was slightly darker in colour than normal blood. 
 
 1 See Minkowski, Arch. f. exper. Path. u. Pharmakol., Leipzig, 1886, B<1. xxi. S. 140 ; 
 and Kiilz, Ztschr.f. Biol, Miinehen, 1887, Bd. xxiii. S. 321. 
 
1 8 6 S YMPTOMA TO LOGY. 
 
 Reaction. — The blood is alkaline in diabetes. Even during- 
 coma the reaction is distinctly alkaline. Nevertheless, careful 
 estimations by many observers have shown that the alkalinity 
 is diminished in severe forms of diabetes with loss of flesh, 
 whilst it is normal, according to v. Noorden ( 20 ), when the 
 general condition is good. In cases of diabetes in which 
 oxybutyric acid is present in the urine, and in cases of diabetic 
 coma, the alkalinity of the blood is said to be greatly diminished, 
 and to be lower than in any other disease (v. Xoorden). But 
 the reaction of the blood is never acid, and by using litmus 
 test papers containing varying quantities of oxalic acid (specially 
 prepared for determining the alkalinity of the blood, according 
 to the method proposed by Haycraft and myself ( 27 )), I have 
 always obtained a very decided alkaline reaction, even in diabetic 
 coma shortly before death. 
 
 Percentage of water. — According to Leichtenstein and v. 
 Jaksch, the percentage of water in diabetic blood is slightly 
 diminished, and according to the latter author ( 2S ) the percentage 
 of albumin is increased. But v. Noorden points out that this 
 slight diminution of the percentage of water and increase in that of 
 the solids is not constant ; in some cases, the proportion is normal, 
 or there may be even an increase of the percentage of water. 
 
 The specific gravity of diabetic blood has been stated to be 
 increased, but James ( 20 ) has found it to be within the normal 
 limits in ten cases, which he has recently examined (according 
 to Eoy's method). 
 
 Number of reel corpuscles and amount of hcemoglobin. — Many 
 observers have found that in diabetes mellitus both the number 
 of red corpuscles and the amount of haemoglobin in the blood 
 are often actually greater than normal. Leichtenstein and others 
 have suggested that the cause of this condition is the poverty of 
 the blood in water, and the consequent concentration thereof, owing 
 to the polyuria. James has shown that, in thirteen cases, the 
 red corpuscles were over 6,000,000 per c.mm. in 5 ; 5,000,000 
 {i.e. normal) in 5; 4,000,000 in 2; 3,000,000 in 1. The 
 haemoglobin was over 100 per cent. {i.e. above normal) in 3 ; 60 
 per cent, in 8 ; 50 per cent, in 2. In these cases the specific 
 gravity, as ascertained by Eoy's method, showed no distinct 
 increase, as would have been expected if the increase of blood 
 corpuscles had simply been clue to deficiency in the percentage 
 of water. 
 
THE BLOOD. 
 
 187 
 
 The following are the results of the enumeration of red 
 blood corpuscles in a few cases of diabetes which I have recently 
 examined, the greatest care being taken to avoid any source of 
 fallacy in the examination : — 
 
 
 Age. 
 
 Form. 
 
 Wasting. 
 
 Appearance of 
 Face. 
 
 Red 
 Corpuscles 
 per C.mm. 
 
 Ada 0' N. . . . 
 
 32 
 
 Severe. 
 
 Considerable. 
 
 Rather pale. 
 
 6,730,000 
 
 Examination at 
 
 
 
 
 
 5,945,000 
 
 another date . 
 
 
 
 
 
 
 JRobertH. . . . 
 
 18 
 
 
 )f 
 
 Pale. 
 
 5,550,000 
 
 Alfred S. . . . 
 
 43 
 
 
 Slight. 
 
 Not anaemic. 
 
 5,760,000 
 
 Fred D 
 
 63 
 
 Mild. 
 
 Very slight. 
 
 Sallow, slight 
 jaundice tinge. 
 
 6,150,000 
 
 Keuben B. . . . 
 
 50 
 
 Severe. 
 
 Slight. 
 
 Pale. 
 
 4,760,000 
 
 James M. . . . 
 
 35 
 
 j, 
 
 Very slight. 
 
 Not ansemic. 
 
 3,600,000 
 
 Patrick M. . . . 
 
 42 
 
 
 Slight. 
 
 ,, 
 
 5,160,000 
 
 Herbert J. . . . 
 
 23 
 
 
 Considerable. 
 
 Pale. 
 
 5,340,000 
 
 Louisa C. . . . 
 
 18 
 
 )5 
 
 Slight, thin girl. 
 
 Not ansemic. 
 
 5,910,000 
 
 Leah S 
 
 55 
 
 Mild. 
 
 Stout. 
 
 ,, 
 
 5,340,000 
 
 Thos. J. . . . 
 
 56 
 
 Severe. 
 
 Very slight. 
 
 
 5,090,000 
 
 Edward G. . . . 
 
 40 
 
 
 Moderate. 
 
 ,, 
 
 5.180,000 
 
 John R. . . . 
 
 33 
 
 " 
 
 Marked. 
 
 " 
 
 5,440,000 
 
 The following table shows the number of red blood cor- 
 puscles, the percentage of haemoglobin, and the specific gravity 
 of the blood in fourteen cases of diabetes, examined by 
 James : — 
 
 Diabetic Blood. 
 
 Red Corpuscles 
 
 Haemoglobin 
 
 Specific Gravity 
 
 per c.mm. 
 
 per cent. 
 
 of Blood. 
 
 1. 6,730,000 .... 
 
 66 
 
 1056 
 
 2. 6,100,000 
 
 
 
 
 
 61 
 
 1059 
 
 3. 4,800,000 
 
 
 
 
 
 58 
 
 
 4. 5,250,000 
 
 
 
 
 
 60 
 
 
 5. 5,600,000 
 
 
 
 
 
 65 
 
 
 6. 3,550.000 
 
 
 
 
 
 52 
 
 1054 
 
 7. 
 
 
 
 
 
 
 1060 
 
 8. 5,300,000 
 
 
 
 
 
 75 
 
 1056 
 
 9. 6,280,000 
 
 
 
 
 
 118 
 
 1055 
 
 10. 5,380,000 
 
 
 
 
 
 96 
 
 1055 
 
 11. 5,564,000 
 
 
 
 
 
 112 
 
 1056 
 
 12. 6,200,000 
 
 
 
 
 
 112 
 
 1057 
 
 13. 4,460,000 
 
 
 
 
 
 55 
 
 1054 
 
 14. 6,000,000 
 
 
 
 
 
 96 
 
 
 The above figures show that often there is an increase in the 
 number of red corpuscles in diabetes; but this condition is not 
 constant, and sometimes they are diminished. Diabetes mellitus 
 appears to differ, therefore, from many chronic ailments, in the 
 
1 8 8 5 YMPTOMA TOL O G Y. 
 
 fact that it is not necessarily associated with any diminution of 
 the number of red corpuscles, the number per cubic millimetre 
 being frequently actually increased. The amount of haemoglobin 
 is sometimes normal, sometimes diminished, sometimes increased, 
 as shown by the above table. In cases examined at the 
 Manchester Eoyal Infirmary, it has been sometimes normal and 
 sometimes increased slightly. 
 
 Leucocytes. — There is no definite change in the proportion 
 or the number of white corpuscles of the blood in diabetes, but, 
 according to v. Limbeck ( 3,) ), the leucocytosis accompanying diges- 
 tion is frequently very well marked in severe cases. 
 
 Glycogen in the Mood, — G-abrischewsky ( 31 ) has drawn 
 attention to the excess of glycogen, which can be detected by 
 microscopical examination of a cover-glass preparation of the 
 blood in diabetes. A small drop of diabetic blood is placed on 
 a cover-glass, another cover-glass is placed on the top of it, 
 and then lightly drawn over the first one — care being taken to 
 avoid pressure as much as possible. In a few seconds the thin 
 film of blood is dry, and the cover-glass may then be mounted 
 in iodine gum of the following composition : — 
 
 Iodine ........ 1 
 
 Potassium iodide ...... 3 
 
 Water, to which an excess of pure gum-arabic 
 
 lias been added . . . . . 100 
 
 The glycogen of the blood can be detected in two forms — (1) 
 In the multinuclear neutrophile leucocytes, as intracellular 
 glycogen ; (2) as free extracellular glycogen, which arises from 
 the degeneration of leucocytes. In normal blood only the extra- 
 cellular glycogen can be recognised with certainty by the 
 action of iodine on cover-glass preparations ; but in diabetic 
 blood, minute specks of glycogen, stained deep brown with the 
 iodine, may be seen distinctly in some of the leucocytes, and 
 the amount of extracellular glycogen is two or three times 
 more than that seen in cover-glass preparations of normal 
 blood. 
 
 Lvpcemia. — A milky appearance of diabetic blood has occa- 
 sionally been observed on post-mortem examination. In other- 
 cases it has been reported to have had a pink colour, and on 
 standing a milky or cream-like serum has separated on the sur- 
 face. This cream-like condition of the serum has been shown, 
 
THE BLOOD. 
 
 189 
 
 by microscopical and chemical examination to be due to the 
 presence of fat globules. 
 
 Analysis of the blood in a number of cases has shown that 
 sometimes the percentage of fat has been greatly increased. The 
 mean amount of fat in human blood (including under this term 
 all the constituents of the ethereal extract of blood, namely, 
 neutral fats, cholesterin and lecithin) is stated by Becquerel and 
 Eodier to be 1*6 parts per 1000 of blood, the maximum being 
 3'25 and the minimum 1-00. Gamgee ( 32 ) obtained the follow- 
 ing results, however, on examination of the blood from two 
 cases of diabetes terminating in coma:- — 
 
 Case 1. 
 
 
 Blood drawn during 
 Life. 
 
 Blood collected after 
 Death. 
 
 Water in 1000 parts . . 
 Total solids .... 
 
 Ethereal (Central fats) 
 
 extract \ Leclthm J 
 extract ( cholesterin 
 
 744-6 
 255-4 
 
 10-8 
 1-96 
 
 757*7 
 
 242-3 
 
 (9 86) 
 11-55 -13-35 
 
 2 14) 
 
 Case 2. — Blood collected after death. — Ethereal extract of 1000 
 parts of blood, 1'88 parts ; cholesterin contained in ethereal extract, 0-642 
 parts. 
 
 Schmidt found the amount of fat in the blood of two cases 
 of diabetes l - 82 and 2-13 per 1000 respectively. Hoppe- 
 Seyler found the fat considerably increased in the blood in four 
 cases of diabetes. 
 
 During life I have examined drops of the blood from 
 numerous diabetic patients microscopically, but have never met 
 with any indication of fat globules in the serum ; also, the 
 addition of perosmic acid has not revealed the presence of any 
 fat globules. Frerichs has obtained similar negative results. 
 
 Acetonemia. — On post-mortem examination the blood of 
 diabetic patients has sometimes the same peculiar smell as that 
 which is often noticed in the breath of severe cases during life — 
 the so-called acetone smell. By distillation, acetone has been 
 obtained from the blood of patients who have died of diabetic 
 coma, by several observers, but has not been obtained by others. 
 
 Glycolytic ferment. — The observations and the views with 
 
1 9 o S YMPTOMA TOL OGY. 
 
 respect to the presence of a glycolytic ferment and the amount 
 thereof, in normal and in diabetic blood, are discussed on p. 79. 
 
 Sugar in the blood. — It was shown long ago that the blood of 
 diabetics contains an excess of sugar, and the sugar in the urine, 
 in most cases, if not in all, is the consequence of this hyper- 
 glycemia. According to Pavy ( 33 ), there is a constant relation 
 between the sugar in the urine and that in the blood, any 
 increase in the latter is shown by an increase in the glycosuria. 
 
 Normal blood contains a small quantity of sugar. In the 
 blood from the general circulation this sugar is glucose, as 
 shown by its reaction to fermentation, polarisation, Fehling's 
 solution, and phenlhydrazin ; but in the blood of the portal 
 system the sugar has a lower capric-oxide reducing power than 
 that of glucose (Pavy). 
 
 From a collection of upwards of a hundred observations upon 
 the dog, cat, rabbit, sheep, ox, horse, and pig, Pavy concludes 
 that the amount of sugar normally present in the blood of the 
 general circulation may be stated to range from about - 6 to 1*0, 
 or a little over 10 per 1000. 
 
 Seegen ( 34 ) has obtained results somewhat similar in animals 
 (the figures being a little higher, however). He has also 
 estimated the quantity of sugar in normal human blood ob- 
 tained by wet cupping. In the blood from ten healthy men the 
 average quantity was 0'17 per cent. 
 
 In diabetes the amount of sugar is often greatly increased ; 
 it may reach 2 # 7 to 5*7 per 1000 (Pavy) ; Seegen has found it 
 as high as - 47 per cent. Seegen states that whilst the blood, 
 in the severe form of the disease, contains a great excess of 
 sugar, in the mild forms sometimes the percentage of sugar has 
 not been above the normal limit. It is interesting to note that 
 in phloridzin diabetes it is stated that the blood sugar is not 
 increased, though this has been disputed. 
 
 The estimation of the amount of sugar in normal and 
 diabetic blood requires considerable care and some experience of 
 delicate chemical analysis. In a work devoted to the considera- 
 tion of diabetes from the clinical standpoint, it is scarcely 
 necessary to describe the usual method of quantitative estima- 
 tion, since they can only be applied to large quantities of blood ; 
 and as venesection is not now performed in cases of diabetes, 
 such quantities can rarely be obtained in medical practice. 
 
 The reader may be referred to the following works for 
 
SUGAR AND GLUCOSE IN THE BLOOD. 191 
 
 details as to the method of quantitative estimation of sugar in 
 the blood : — 
 
 Pavy, F. W. — " Croonian Lecture on Certain Points connected with 
 Diabetes," London, 1878 ; "Physiology of the Carbohydrates," London, 
 1894, pp. 58-80. Seegen, J., "Die Zuckerbildung im Thierkorper," 
 Berlin, 1890, S. 11. v. Jaksch. — "Clinical Diagnosis," translation by 
 Cagney. 
 
 For clinical methods of estimation of the amount of glucose 
 in the blood, see pp. 195—96. 
 
 A simple metliod of distinguishing diabetic from non-diabetic blood 
 — The decolorisation of methylene blue. — In order to detect the 
 difference between diabetic blood and non-diabetic blood, from the 
 amount of sugar present in each, it is necessary to examine a 
 considerable quantity of blood, and I am not aware that any method 
 has hitherto been proposed by which the excess of sugar can be 
 detected by the examination of a drop of blood obtained by 
 pricking the patient's finger. But by the following method the 
 difference between diabetic and non-diabetic blood can be easily 
 demonstrated clinically ( 35 and 36 ). 
 
 Whilst making a number of blood examinations, it occurred 
 to me to try the effect of adding a drop of diabetic blood to 
 a solution of methylene blue. 1 I was surprised to find that 
 diabetic blood was much more powerful than non-diabetic blood 
 in removing the colour from a warm alkaline solution of 
 methylene blue. I found that, on heating a small quantity of 
 blood and methylene blue solution in certain proportions, the 
 colour was removed in the case of diabetic blood, but not when 
 non-diabetic blood was employed. The reaction is so sensitive, 
 that the difference between non-diabetic blood and the blood 
 from a well-marked case of diabetes can be easily demonstrated 
 in a drop of blood obtained by pricking the finger, and I have 
 frequently demonstrated this difference to medical students and 
 friends. 
 
 The following is the exact method which I have used : — 
 
 A small narrow test tube 2 is well cleaned, and at the bottom 
 of the tube are placed 40 c.mm. of water. To measure this I 
 
 1 By mistake the term methyl blue was used in place of methylene blue in my first 
 description of the above method, published in the Brit. Med. Journ., London Sept 
 19, 1896. 
 
 2 it is important to use a narrow test tube, so that the upper surface of the fluid 
 with which the air comes in contact, may be as small as possible. 
 
£ YMPTOMA TOL OGY. 
 
 have used the capillary tube of a Gowers' hsemoglobinometer, 
 which is graduated for 20 c.mm. The tip of one of the patient's 
 fingers is cleaned and dried, then pricked, and when a large 
 drop of blood has escaped, it is sucked up into the small 
 capillary hamioglobinonieter tnbe ; 20 c.mm. of blood are taken 
 up from the finger. The blood is then blown gently into the 
 water at the bottom of the small test tube. If it should adhere 
 
 to the side of the tube, it must 
 be carefully shaken to the bottom. 
 Then 1 c.cm. of a 1 in 6000 
 watery solution of methylene blue 
 is added. (To measure this I have 
 used the 1 c.cm. tube supplied 
 with Southall's ureometer.) To 
 the mixture, finally 40 c.mm. of 
 liquor potassre are added. The 
 contents of the tube are then 
 well mixed by shaking. As a. 
 control experiment, a second test 
 tube of similar size is taken, and 
 
 ■ into this is placed the same 
 
 quantity of non-diabetic blood, with 
 the same proportion of water, 
 methylene blue, and liquor potassa?. 
 The fluid in each tube has a 
 fairly deep blue colour. Both 
 \^_J) tubes are then placed 'in a beaker, 
 capsule, or very wide test tube 
 containing water. Heat is applied 
 by a spirit-lamp until the water 
 boils ; it is allowed to continue 
 boiling for about four minutes. 
 By the end of this time the fluid in the tube containing the 
 diabetic blood changes its colour from fairly deep blue to 
 a dirty pale yellow (almost the colour of normal urine). Whilst 
 the fluid in the tube containing the non-diabetic blood remains 
 blue, occasionally it becomes bluish green, sometimes pale violet, 
 but it is never decolorised, that is, it never loses its blue colour. 
 The tubes should be kept quite still whilst in the water bath, 
 as, by shaking, the decolorised methylene blue is oxidised by the 
 oxygen of the atmosphere, and a blue tint may then return to 
 
 Fig. 11.— Author's test for diabetic 
 blood : end of reaction : fluid in 
 tube containing normal blood, 
 blue ; that in tube containing 
 diabetic blood, yellow. 
 
AUTHOR'S DIABETIC BLOOD REACTION. 193 
 
 the fluid. This is the reason why it is necessary to use a water- 
 bath, since, if the test tubes be heated directly over the spirit- 
 lamp, it is difficult to avoid shaking of the fluid. 
 
 I have made fifty examinations of diabetic blood according 
 to this method, in twenty cases of diabetes mellitus. Seventeen 
 were suffering from a very severe form of the disease. In every 
 examination, in all of the twenty cases, a 1 in 6000 solution 
 of methylene blue was readily decolorised when the test was 
 carried out, as described above. In very severe cases a solu- 
 tion of methylene blue of double strength (1 in 3000) was also 
 decolorised. I have made a large number of examinations of 
 normal blood, but have never found the methylene blue solution 
 to be decolorised when the above-described proportions of fluid 
 were used. I have also examined the blood in 100 patients 
 who were not suffering from diabetes. These 100 cases included 
 the most varied ailments, often at a very advanced stage, such 
 as diseases of the heart, lungs, stomach, liver, nervous system, 
 the various forms of antemia and Bright's disease, purpura, 
 leucocy thsemia, gout, rheumatism, lead-poisoning, cancer of various 
 organs, jaundice, and many other affections. I have never met 
 with any case amongst the non-diabetic patients, in which a 1 in 
 6000 solution of methylene blue was decolorised when the test 
 was carried out as above described. On the other hand, diabetic 
 blood has always decolorised it readily. 
 
 With respect to the cause of this marked difference between 
 diabetic and non-diabetic blood, it is most natural to attribute it 
 to the excess of sugar present in the former. It is known that 
 grape sugar readily removes the colour from a warm alkaline 
 solution of methylene blue. I have found that diabetic urine 
 decolorises a 1 in 3000 solution, even when diluted to such an 
 extent that the amount of sugar is only 0'07 per cent. 
 
 If the decolorisation produced by diabetic blood be due to 
 the excess of sugar present, then one would expect that by using 
 a larger proportion of normal blood a similar decolorisation would 
 result, since the percentage of sugar in many cases of diabetes is 
 not more than three times the normal percentage. Thus one 
 would expect that the amount of sugar would be almost the 
 same in a mixture of 20 c.mm. of blood from a severe case of 
 diabetes + 1 c.cm. of methylene blue solution + 40 c.mm. of 
 liquor potassae, as in a mixture of 60 c.mm. of normal blood + 
 1 c.crn. of methylene blue solution and 40 c.mm. of liquor 
 13 
 
1 9 4 S YMPTOMA TO LOG Y. 
 
 potassse. I have found that when the last-mentioned proportions 
 of normal blood and solution are used, the blue colour is removed, 
 as in the case of diabetic blood. But since diabetic blood and 
 normal blood differ in other respects, in addition to the difference 
 in the percentage of sugar, I have not been able to prove that 
 the above reaction, in the case of diabetic blood, is due simply 
 to the excess of sugar, and it is quite possible that there may be 
 some other causes. 
 
 In one case which I examined, the patient was suffering from 
 a severe form of the disease, the urine contained a large amount 
 of sugar, and the methylene blue solution was readily decolorised 
 by the blood. In this case phthisis developed ; and when the 
 lung symptoms became very advanced, the amount of sugar in 
 the urine greatly diminished, until finally the urine did not give 
 any immediate reaction on boiling with Fehling's solution ; oxide 
 of copper only being thrown down when the solution was cooling. 
 Nevertheless the methylene blue solution was still decolorised 
 fairly well by 20 c.mm. of the patient's blood, though it is pro- 
 bable that at this late stage the blood only contained a small 
 excess of sugar. In another mild case of diabetes (in which 
 there was no thirst, no diuresis, and in which the total quantity 
 of sugar excreted per diem was only 1008 grs.) this blood 
 reaction was obtained distinctly. In a very mild case of 
 glycosuria, in which there was no excess of urine, I obtained 
 the reaction quite well. The test is exceedingly delicate, but it 
 is possible that in the mildest forms of glycosuria the reaction 
 may not always be obtained, or may only be obtained when 
 different proportions of the fluid are used. 
 
 I believe by the above-described simple method — by the 
 decolorisation of a warm alkaline, 1 in 6000 solution of methylene 
 bl ue — it is possible to easily distinguish the blood of an ordinary 
 case of diabetes mellitus from that of a healthy person or non- 
 diabetic patient ; and the test is so sensitive that it can be 
 employed when only a drop of blood can be obtained, as by 
 pricking the patient's finger. 
 
 In cases in which the urine cannot be obtained for examina- 
 tion, this blood test will be of service in diagnosis. 
 
 The reaction is interesting pathologically, and it might be of 
 value clinically if a diabetic patient were seen for the first time 
 in a comatose condition, and no urine could be obtained for 
 examination. 
 
DIABETIC BLOOD. 195 
 
 My observations have been since confirmed by Lupine and 
 Lyonnet ( 37 and 38 ), by P. Marie and Le Goff ( 39 and 40 ), Loewy, 
 Goldscheider, and others ( 41 ). 
 
 Le Goff ( 42 ) has made a number of observations on the 
 amount of methylene blue solution which is decolorised by 
 diabetic blood. His researches show that not only will diabetic 
 blood decolorise a larger quantity of alkaline methylene blue 
 solution than normal blood, but also the rapidity of the reaction 
 is greater. Thus Le Goff found that 20 c.mm. of the blood of 
 a diabetic patient was able to decolorise 1 c.c. of a methylene 
 blue solution in one to two minutes, whilst 20 c.mm. of normal 
 blood decolorised the same quantity only at the end of ten 
 minutes. 
 
 Two clinical methods of estimating the amount of sugar in 
 the blood, based on the methylene blue reaction, have recently 
 been worked out by Lyonnet and Le Goff respectively. 
 
 Lyonnet's ( 38 ) method is as follows : — 
 
 The small tube of a Malassez's hamioglobinometer is filled 
 with blood up to the mark 1, and with distilled water up to 100. 
 The mixture is then placed in a glass tube, and two drops of 
 potash solution are added. The tube is placed in a small vessel 
 of boiling water. Then drop by drop, from a graduated tube, 
 a 1 in 4000 solution of methylene blue is added, until the 
 mixture acquires a persistent blue colour. When the blood of 
 the dog is employed, it is necessary to add ten drops, in order 
 to obtain a persistent blue colour. In man the results are 
 practically the same. 
 
 In cases of cardiac disease, dyspepsia, cancer of the uterus, 
 and gangrene, the number of drops of methylene blue solution 
 required were 11, 9, 8, 9, respectively. 
 
 In the case of a dog rendered diabetic by the removal of the 
 pancreas, thirteen to fourteen drops were required. 
 
 The results obtained in three diabetic patients were as 
 follows : — 
 
 Case 1. — Urine, 50 grms. of sugar per litre . . 20 drops of methylene 
 
 blue required for 
 above reaction. 
 >> 2. ,, 36 ,, ,, 13 ,, 
 
 »> "• >> *-o ,, ,, !_/ ,, 
 
 Lyonnet thinks that there is a certain relation between the 
 
1 96 S YMPTOMA TO LOG Y. 
 
 glycosuria and the quantity of methylene blue decolorised by the 
 blood of diabetic patients, and that this method, even if not exact, 
 will indicate roughly whether a small or large quantity of sugar 
 is present in the blood. 
 
 Le Goff ( 43 ) estimates the amount of sugar in the blood 
 roughly, by ascertaining the quantity of an alkaline solution of 
 methylene blue which must be added to a definite quantity of 
 blood before a persistent blue colour is obtained when the fluid 
 is heated. The method is as follows : — 
 
 Twenty c.mm. of blood are taken from the finger and placed 
 in a small tube, with 1 c.c. of a 1 in 5000 solution of methylene 
 blue (containing 1'5 mgrms. of potash per c.c). After 
 mixing well, the tube is placed in boiling water. The blue 
 colour of the mixture disappears. More of the alkaline 
 methylene blue solution is added, until a blue colour is obtained, 
 which persists for at least a quarter of an hour. 
 
 With normal blood, the mean quantity of methylene blue 
 solution required was found by Le Goff to be 1256 c.mm. 
 With diabetic blood the quantity was always great — 2550 to 
 3000. 
 
 P. Marie and Le Goff ( 39 and u ) have employed the following 
 clinical method for estimating the quantity of sugar in the blood 
 more exactly : — 
 
 By means of a graduated Pravaz syringe, 1 c.c. of blood is 
 taken from one of the veins in front of the elbow, and added 
 drop by drop to 6 or 8 c.c. of alcohol (96°) in a small glass 
 tube. The mixture is well shaken for ten minutes, and then 
 allowed to stand for twenty-four hours. The albumins, haemo- 
 globin, and glycogen are precipitated. The precipitate is filtered 
 off, washed in alcohol, and the sugar in the alcoholic extract 
 estimated by an alkaline solution of methylene blue of the 
 strength given above. 6500 c.mm. of the solution correspond 
 to 1 mgrm. of glucose. The tube containing the alcoholic 
 extract is not placed in boiling water, but in hot water having 
 a temperature of 80° to 9 0°. 
 
 The following is an example of this method : — 
 
 One c.c. of blood was taken from a diabetic patient, and 
 treated as above described. The alcoholic extract amounted to 
 6 c.c. One c.c. of this extract was mixed with 3000 c.mm. of 
 alkaline methylene blue solution, and the tube placed in hot 
 water at 90° at 3.15 A:M. 
 
250 
 
 3 3 
 
 250 
 
 3 ) 
 
 250 
 
 ) 5 
 
 100 
 
 3 3 
 
 150 
 
 J 3 
 
 250 
 
 3 3 
 
 STAINING OF RED CORPUSCLES. 197 
 
 At 3.25 the fluid, completely decolorised ; 250 c.mm. of m.-blue added. 
 
 33 3-30 ,, ,, 
 
 ,, 6.61 ,, ,, 
 
 „ 3.44 
 
 „ 3.45 
 
 ., 3.47 
 
 ,, 3.55 ,, ,, 
 
 ,, 4.15 the blue colour persists. 
 
 Total quantity of methylene blue solution added to 1 c.c. of the 
 
 alcohol extract = 4500 c.mm. 
 
 Quantity required for the 6 c.c. = 6 x 4500 = 27 c.c. 
 
 The strength of the blue solution is such that 6 '5 c.c. corresponds 
 
 to 1 mgrm. of glucose. Hence the sugar in the alcoholic extract is 
 
 27 
 
 — = 4T5 mgrms. Therefore the blood contained 4T5 grins, of sugar 
 
 6-5 ° 
 
 per litre. 
 
 Staining of red corpuscles in diabetes. — Bremer ( 45 ) has dis- 
 covered that the red corpuscles of diabetic blood stain green when 
 a cover-glass preparation is treated according to a certain method, 
 with a special solution of eosin and methylene blue, whilst the 
 red corpuscles of non-diabetic blood stain purple or madder 
 colour when treated in the same manner. Lepine and Lyonnet 
 ( 46 ) have confirmed Bremer's observations, but have also found 
 that the red corpuscles of the blood in a case of leucocythcemia 
 stained green just in the same manner as the red corpuscles of 
 diabetic blood. 
 
 The following are the details of the Bremer method : — 
 A cover-glass preparation of blood is made in the usual 
 manner, and a thin, even film obtained. For comparison, a 
 similar preparation of normal blood is made. Both are placed in 
 a wide-mouthed bottle or glass, containing equal parts of alcohol 
 and ether (10 grms. of each). This vessel is then placed in hot 
 water, and the ether-alcohol allowed to boil for four minutes. 
 The cover-glasses are then stained in a solution prepared in the 
 following manner : — Saturated watery solutions of eosin and 
 methylene blue are mixed in equal parts. A precipitate forms ; 
 this is filtered off, washed, and dried. It is then reduced to a 
 powder, and one twenty-fourth part of eosin and one-sixth of 
 methylene blue are added. From 0'025 grm. to 0"05 grm. of this 
 mixture is dissolved in 10 grms. of a 33 per cent, solution of 
 alcohol. This staining fluid does not keep well, and must be 
 freshly prepared shortly before the cover-glasses are stained. 
 
1 9 8 £ YMPTOMA TOL OGY. 
 
 After remaining in the above fluid for four minutes, the cover- 
 glasses are washed in water. The film of diabetic or glycosuria 
 blood is stained sap or bluish green, the non-diabetic blood 
 reddish violet. Under the microscope the red corpuscles of 
 diabetic or glycosuria blood are stained green, whilst those of 
 non-diabetic blood are stained of purple or madder colour. 
 Bremer found this reaction in fifty cases of diabetes mellitus or 
 glycosuria, and believes that in these diseases a diagnosis can be 
 made by examination of a drop of blood. In a postscript he 
 adds that in one case recently examined he has obtained negative 
 results. 
 
 Le Goff ( 47 ) has carefully investigated the colour reactions 
 of diabetic blood. He has employed eosin-methylene-blue for 
 this purpose. A saturated watery solution of eosin is mixed 
 with a saturated watery solution of methylene blue (the propor- 
 tion of the two solutions is not stated — probably the author 
 intends equal volumes to be used). The precipitate which 
 forms is washed in water, then dried. Five cgrms. of this 
 substance are dissolved in 20 to 25 grms. of alcohol (30°). The 
 solution is then filtered. Cover-glass preparations of blood 
 (normal and diabetic) are heated in a hot chamber for two 
 hours, at a temperature of 120° C, and then stained in the 
 above solution, washed in distilled water, dried with filter paper, 
 and mounted in xylol balsam. The red corpuscles of normal 
 blood are stained, the colour varying from a clear purplish rose 
 colour to a dark maroon, whilst the red corpuscles of diabetic 
 blood are pale green, yellowish green, yellowish or unstained. 
 The nuclei of the white corpuscles are stained blue, and in other 
 respects are the same, both in normal and diabetic blood. 
 
 Le Goff has confirmed the results of Bremer, and has 
 recorded a number of interesting observations on the colour 
 reactions of diabetic blood. 
 
 More recently Bremer ( 48 ) has pointed out that the blood 
 of patients suffering from diabetes mellitus can be distin- 
 guished from normal blood by the following reactions to aniline 
 stains : — 
 
 A drop of blood is smeared by means of a slide over a second 
 slide, so as to cover about one-third or one-half of the surface. 
 A number of slides are prepared from diabetic and non-diabetic 
 blood (the latter for control staining). As the reaction can be 
 seen by the naked eye, tolerably thick layers of blood are 
 
STAINING OF RED CORPUSCLES. 199 
 
 employed, and these have, as nearly as possible, the same 
 thickness. The blood preparations are now heated in a hot-air 
 chamber for six to ten minutes, at a temperature of 125° C. 
 It is important to avoid heating the preparations over 140° C. ; 
 also they should not be heated for more than ten minutes. 
 The slides are then stained in a 1 per cent, watery solution of one 
 of the following stains : — Congo red, methylene blue, Biebrich 
 scarlet, or in Ehrlich-Biondi's staining fluid. Slides of diabetic 
 and non-diabetic blood are treated exactly in the same manner 
 for comparison. The results are as follows — Congo red : stain for 
 one and a half to two minutes, then wash rapidly in water, and dry ; 
 the non-diabetic blood is stained, whilst the diabetic preparation 
 is not stained, or only indifferently stained. Methylene blue 
 stains non-diabetic, but not diabetic blood. With Biebrich 
 scarlet the diabetic blood is stained deeply, the non-diabetic is 
 unstained. If preparations be placed in Ehrlich-Biondi's fluid 
 two to three minutes, the diabetic blood is stained orange, whilst 
 the non-diabetic blood is stained violet. Beautiful contrasts are 
 obtained by double staining. Preparations of diabetic and non- 
 diabetic blood are stained in a 1 per cent, watery solution of 
 methylene green for one and a half to two minutes and then 
 washed. Both preparations are green, the diabetic being more 
 deeply stained. They are next placed in a \ per cent, watery 
 solution of eosin for eight to ten seconds. The diabetic 
 preparations remain green ; the non-diabetic take the eosin 
 stain. Similar results are obtained with methyene blue and 
 eosin. 
 
 REFERENCES. 
 
 1. Seegen, J. . . . " Der Diabetes mellitus," Berlin, 1893, p. 167. 
 
 2. Loeb Centralbl. f. inner e Med., Leipzig, 21st 
 
 November 1896. 
 
 3. Wallach .... Vircho'w's Archiv, Bd. xxxvi. S. 297. 
 
 4. Schmitz, R, . . . Deutsche mecl. Wchnschr., Leipzig, 27th 
 
 November 1893. 
 
 5. Finkler .... Verhandl. d. Cong.f. innereMed., Wiesbaden, 
 
 1886, S. 190. 
 
 6. Ptjsinelli . . . Berl. Idin. Wchnschr., 17th August 1896. 
 
 7. Leube, W. . . . Virchoiv's Archiv, Bd. cxiii. S. 391. 
 
 8. Seegen .... Loc. cit., S. 147-156. 
 
 9v Leo, H Deutsche med. Wchnschr., Leipzig, 1892, 
 
 No. 33. 
 
S YMPTOMA TOL O G Y. 
 
 10. Hallervorden . . 
 
 11. Stadelmann . . . 
 
 12. TORALBO . . . . 
 
 13. jMoraczewski, W. v, 
 
 14. Teubaum, E. . . . 
 
 15. Schmitz, E. 
 
 16. Maguire, E. 
 
 17. Kulz . . 
 
 18. 
 
 19. 
 
 20. 
 
 21. 
 
 22. 
 23. 
 
 24. 
 
 25. 
 26. 
 
 27. 
 
 28. 
 
 29. 
 30. 
 
 31. 
 32. 
 
 Saxdmeyer . . 
 Macmunn, C. A. 
 v. Jaksch. . . 
 Le Nobel . . 
 v. Noorden, C. . 
 
 HlRSCHFELD . . 
 
 RoSENFELD . . 
 V. NoORDEN . . 
 
 do. . . . 
 
 Haycraft, J. B., AXD 
 Williamson, E. T. 
 v. Jaksch .... 
 
 James, A. . 
 v. Limbeck 
 
 Gabritchewsky, G. 
 Gamgee, A. . . 
 
 Arch. f. exper. Path. u. Pharmakol., Leipzig, 
 
 1880, Bd. xii. S. 237. 
 Ibid., 1883, Bd. xvii. S. 419. 
 Abstract, Centralbl, f. Min. Med., Bonn, 1890, 
 
 S. 19. 
 Centralbl. f. inner e Med., Leipzig, 1897, 
 
 p. 921. 
 Ztschr. f. Biol, Miinchen, 1896, S. 379- 
 
 403 ; Abstract, Fortschr. d. Med., Berlin, 
 
 1897, S. 342. 
 Berl. Min. Wchnschr., 1891, No. 15. 
 Article, " Diabetes Mellitus," " Fowler's Dic- 
 tionary of Medicine," London, 1890. 
 Lyon med., 1892, No. 23; Abstract, Oesterr. 
 
 -ungar Centralbl. f. d. med. Wissensch., 
 
 Wien, 1st January 1893. 
 Centralbl. f. Min. Med., Bonn, 1890, No. 28, 
 
 Beilage. 
 " Clinical Chemistry of Urine," London, 
 
 1889, pp. 194, 200. 
 " Clinical Diagnosis," translated by Cagney, 
 
 London, 1890, p. 250. 
 Arch. f. exper. Path. u. Pharmakol., Leipzig, 
 
 Bd. xviii. S. 6. 
 "' Die Zuckerkrankkeit," Berlin, 1895, S. 199. 
 Deutsche med. Wchnschr., 1893, No. 38; 1895, 
 
 No. 26 ; Ztschr. f. Min. Med., Berlin, Bd. 
 
 xxxi. ; ibid., Bd. xxxviii. ; Centralbl. f. 
 
 innere Med., Leipzig, 1896, No. 24. 
 Centralbl. f. innere Med. , Leipzig, 1895, No. 5 1 . 
 Loc. cit., S. 201. 
 hoc. cit, S. 90. 
 Proc. Roy. Soc. Ed in., vol. xv. ; v. Jaksch 's 
 
 "Clinical Diagnosis," London, 1890, p. 4. 
 Miinchen. med. Wchnschr., 25th April 1893, 
 
 S. 328 (Congress Report). 
 Edin. Med. Journ., September 1896. 
 " Grundriss einer klin. Pathologie des Blutes," 
 
 Jena, 1896, S. 347 (quoted by Bettmann, 
 
 Miinchen, med. Wchnschr., 1896, S. 1231). 
 Arch. f. exper. Path, u. Pharmakol., Leipzig, 
 
 Bd. xxviii. S. 272. 
 ''Physiological Chemistry of the Animal 
 
 Body," London, 1880, vol. i. pp. 170-172. 
 
REFERENCES. 
 
 33. Pavy, F. W. . . . 
 
 34. Seegen . . . . 
 
 35. Williamson, K. T. . 
 
 36. 
 
 do. 
 
 37. Lepine et Lyonnet 
 
 38. Lyonnet .... 
 
 39. Marie, P. et Le 
 
 GOFF, J. 
 
 40. Le Goff, J. . . . 
 
 41. Loewy 
 
 42. Le Goff, J. 
 
 43. do. 
 
 44. do. 
 
 45. Bremer 
 
 46. Lepine et Lyonnet 
 
 47. Le Goff, J. . . . 
 
 48. Bremer . . . . 
 
 " Physiology of the Carbohydrates," London, 
 
 1894, p. 158. 
 Loc. cit., S. 29, 89, 90. 
 Brit. Med. Journ., London, 19th September 
 
 1896. 
 Centralbl. f. inner e Med., Leipzig, 21st 
 
 August 1897 (original contribution). 
 Lyon med., 1897, p. 20. 
 Ibid., 1897, p. 137. 
 Bull, et mem. Soc. med. d. hop. de Paris, 
 
 1897, Nos. 16 and 17. 
 " Sur certaines reactions chromatiques du 
 
 sang dans le diabete sucre," Paris, 1897, 
 
 pp. 44-81. 
 Berl. Uin. Wchnschr., 1897, Eo. 46, S. 1016. 
 
 (Report of Verein f. innere Med.) 
 Loc. cit., p. 71. 
 Loc. cit., p. 66. 
 Loc. cit., p. 77. 
 New York Med. Journ., 7th March 1896; 
 
 Centred!)!, f. d. med. Wissensch., Berlin, 
 
 1894, No. 49; Med. News, Philadelphia, 
 
 9th February 1895. 
 Lyon med., 1896, tome Ixxxii. p. 187. 
 Loc cit. 
 Centrcdbl. f. innere Med., Leipzig, 1897, 
 
 No. 22. 
 
CHAPTER X. 
 
 SYMPTOMS, COMPLICATIONS, AND PATHOLOGICAL CHANGES 
 IN CONNECTION WITH THE VARIOUS SYSTEMS. 
 
 Besides the changes in the urine and blood, there are a number 
 of prominent symptoms. The most important are : — Thirst and 
 increased appetite ; great weakness and emaciation in the severe 
 forms of the disease ; a harsh, dry skin ; often a red, raw-looking 
 tongue. These will be best considered, however, under the 
 abnormalities met with in the various systems. 
 
 Temperature. — In diabetes the temperature is generally normal 
 or subnormal. 
 
 In diabetic coma it is usually very low — 95°, 94°, or lower. 
 Occasionally, though exceedingly rarely, it is high, just before 
 death in diabetic coma. A high temperature may be also 
 caused in diabetes by the occurrence of various complications. 
 
 The Alimentary Canal, Liver, and Pancreas. 
 
 Saliva. — The mouth of the diabetic patient is dry, and the 
 saliva as a rule scanty ; occasionally, though very rarely, 
 pytalism occurs. The reaction of the saliva, even the parotid 
 fluid collected from Steno's duct, is usually acid. Almost with- 
 out exception, the saliva is free from sugar — Frerichs ( 1 ), v. 
 Noorden ( 2 ). 
 
 v. Noorden states that sometimes he has failed to obtain the 
 sulphocyanide reaction. 
 
 Gums and teeth. — On the gums and soft palate white 
 patches are sometimes seen in advanced cases. These are due 
 to the growth of fungus spores and mycelium on the superficial 
 epithelium, and may be prevented by careful cleanliness. 
 
 The gums often become red and inflamed ; they are some- 
 times spongy and swollen, and bleed easily. The teeth often 
 become loose and fall out ; they are frequently carious ; and 
 
THE ALIMENTARY SYSTEM. 203 
 
 sometimes these dental troubles occur at an early stage of the 
 disease. Alveolar periostitis occasionally occurs. 
 
 The tongue is often very red and abnormally clean ; 
 frequently the surface is cracked and raw in appearance, the 
 epithelium being deficient. In mild cases, however, the tongue 
 is often moist, and coated with a thin yellowish-white fur. 
 
 Thirst is one of the most characteristic symptoms, and often 
 the first indication of the disease. The onset of thirst is occa- 
 sionally quite sudden, and the patient is able to state definitely 
 the time of the day or the hour when it first commenced (see p. 
 164). The patient takes enormous quantities of liquids, 10 to 
 15 pints or even more per diem, and still the thirst continues. 
 Some diabetics complain more of dryness of the mouth than of 
 thirst. In a case already referred to, the patient complained of 
 a salty taste. Sometimes the thirst is so great that sleep is 
 prevented or much disturbed ; as a rule it runs parallel with 
 the sugar excretion, and diminishes when this is reduced or 
 arrested by a restricted diet (see p. 167). In some of the very 
 mild cases of diabetes, thirst is absent or very slight, and even 
 in severe cases, which run a fairly rapid course, occasionally 
 thirst is not a very prominent symptom ; the patient does not 
 complain of it, but when questioned admits that he is more 
 thirsty than formerly. At the last stage of the disease, when 
 phthisis occurs as a complication, the thirst may diminish and 
 finally disappear just before death. 
 
 The appetite is generally increased, often enormously so. 
 As Seegen puts it, many patients describe their condition as if 
 they had " a hole in the stomach." This great increase in the 
 appetite is met with chiefly in the severe forms of the disease, or 
 in the mild cases, so long as carbohydrate food is taken in large 
 quantities. In the latter class of cases, if carbohydrate food be 
 excluded from the diet the hunger subsides. 
 
 Thirst is a much more constant symptom than hunger. 
 Sometimes, in the very severe form of diabetes, the appetite is 
 not increased or is only slightly increased, especially at a late 
 stage when there is great wasting, when advanced phthisis is 
 a complication, or when catarrh of the stomach is present, owing 
 to a prolonged nitrogenous diet, or to other causes ; and in 
 certain mild cases there is no increased appetite. In the last 
 stage of the disease there is often marked anorexia. 
 
 The digestive power of diabetic patients is generally very 
 
2o 4 SYMPTOMS AND PATHOLOGICAL CHANGES. 
 
 good, and persons whose digestion has previously been feeble 
 often improve markedly in this respect after diabetes develops. 
 Dyspeptic individuals who have previously found it necessary to 
 be very cautious in the choice of their diet, often lose all their 
 dyspeptic troubles after the onset of diabetes, and can take 
 large quantities of food with impunity. 
 
 Sometimes signs of gastric catarrh, impaired digestion, and 
 loss of appetite are met with, when the patient has been kept 
 for some time on a very rigid diet. Occasionally dilatation of 
 the stomach occurs as a complication, especially when the treat- 
 ment has been neglected, and when the patient has taken 
 enormous quantities of food. Gastric ulcer has also been recorded 
 as a complication, but it is exceedingly rare. 
 
 The gastric juice has been examined by Grans ( 3 ), Honig- 
 mann ( 4 ), Eosenstein ( 5 ), and others, but the variations in the 
 amount of hydrochloric acid have been mostly within the normal 
 limits ; sometimes, however, free hydrochloric acid has been 
 absent. Eosenstein has shown that free hydrochloric is sometimes 
 temporarily, sometimes permanently, absent from the gastric 
 juice. He believes the former cases to be due to gastric neurosis, 
 and the latter to interstitial gastritis, with extensive atrophy of 
 the mucous membrane. 
 
 Karl Grube ( 6 ) has observed gastric crises, resembling those 
 of tabes dorsalis. The patient is suddenly seized with violent 
 pain in the abdomen, especially at the pit of the stomach ; this 
 is accompanied by abdominal distension and eructations. Nausea 
 and vomiting of acid material occur, and sometimes diarrhoea 
 and cramps in the legs follow. 
 
 Intestines. — Diabetic patients very frequently suffer from 
 constipation, which is sometimes very obstinate, and which is 
 increased if a nitrogenous diet and opium are prescribed. The 
 constipation is in part the result of the nitrogenous diet ; but 
 this is not the only cause, since constipation is a common 
 symptom, when the patient's diet has not been restricted ; it 
 is common also amongst diabetic patients on admission into 
 hospital, even when there has been no restriction of diet and 
 when no opiates have been taken. In severe cases of diabetes, 
 when the bowels are markedly constipated, frequently the urine 
 gives a dark brownish-red coloration with perchloride of iron 
 (diacetic acid reaction), but if the constipation be relieved by 
 purgatives, I have frequently found that the perchloride of iron 
 
THE ALIMENTARY SYSTEM. 205 
 
 reaction has become less marked or has disappeared. Occasion- 
 ally, but rarely, I have found the perchloride of iron reaction 
 when the bowels have not been constipated. 
 
 Sometimes, especially in chronic cases, the patient suffers 
 from diarrhcea. This is a serious complication, and rapidly 
 undermines the patient's strength, since the absorption of food 
 is hindered, and the amount of fatty food must then be limited. 
 Hence diarrhcea ought to be energetically treated by opium, 
 chalk, etc. Sometimes the diarrhcea is of a dysenteric 
 character. 
 
 In diabetic coma the bowels are generally constipated 
 (seventeen out of twenty-seven cases, see p. 275), and obstinate 
 constipation often precedes coma, and appears to be a predispos- 
 ing factor. Hence it is especially necessary to prevent prolonged 
 constipation in advanced or severe cases of diabetes. Schmitz 
 drew attention to this point. He thought it probable that when 
 the bowels of diabetic patients remained constipated for a pro- 
 longed period, toxic substances were formed in the intestines, 
 and to the absorption of these he attributed the coma. Some- 
 times, though very rarely, diarrhoea is present in diabetic coma, 
 and an attack of diarrhoea at an advanced stage of diabetes 
 occasionally appears to be the exciting cause of coma. 
 
 Fatty motions — steatorrhea. — It is well known that some- 
 times in diseases of the pancreas, a great excess of fat is found 
 in the motions, though this is by no means a constant symptom ; 
 also Abelmann has shown that after complete pancreas extirpa- 
 tion in animals, the fats of the food are not absorbed. Natur- 
 ally one would expect that, in cases of diabetes associated with 
 disease of the pancreas, analysis of the fasces would show an 
 excess of fat. I do not know of any series of cases, of diabetes 
 associated with pancreatic disease, in which such an analysis has 
 been made. 
 
 Hirschfeld ( 7 ) has drawn attention to a class of cases of 
 diabetes in which the assimilation of albumins and fats is very 
 considerably diminished, and these he regards as forming a new 
 clinical variety of the disease. 
 
 Marked pathological changes in the stomach and intestines are 
 not frequent and not characteristic. 
 
 In ninety-two cases examined at the Vienna Pathological 
 Institute, and recorded by Seegen ( 8 ), the following changes were 
 noted : — 
 
2o6 SYMPTOMS AND PATHOLOGICAL CHANGES. 
 
 Cases. 
 
 Dilatation of the stomach . . . . . . in 6 
 
 Ecchymoses in the gastric mucous membrane . . ,, 5 
 
 Marked swelling and redness of the gastric mucous 
 
 membrane . . . . . . . . ,, 3 
 
 Marked acute intestinal catarrh . . . . . ., 4 
 
 „ chronic catarrh of the large intestine . . „ 1 
 
 Tubercular ulceration of the intestine, associated with 
 
 tuberculosis of the lungs . . . . . „ 2 
 
 In the post-mortem records of fifty-five cases recorded by 
 Frerichs ( 9 ), the stomach is often reported to be normal, often 
 dilated, and often thickening of the walls is recorded ; but no 
 characteristic or important change was observed. In diabetes, 
 complicated by phthisis, tubercular ulceration in the intestines 
 was sometimes recorded, and in two cases a dysenteric condition 
 of the large intestine was observed. A thick layer of fungus 
 growth was often present in the mucous membrane of the 
 oesophagus and fauces. In many cases the mesenteric glands 
 were enlarged. 
 
 The liter. — Sometimes during life the liver is found to be 
 enlarged, owing to hyperemia, fatty infiltration, or cirrhosis. 
 This enlargement' is met with chiefly in obese or gouty diabetics 
 who suffer from a mild form of the disease. Amongst hospital 
 patients suffering from the most severe form of diabetes with 
 wasting, generally no enlargement of any importance can be 
 detected during life. 
 
 Glenard ( 10 ) states that he has found changes in the liver, on 
 examination of the abdomen, in 60 per cent, of diabetic patients. 
 Hypertrophy was the most common change (34 - o per cent, of 
 the cases). But Glenard's statistics were based on 324 cases 
 met with in private practice at Vichy, and these cases would no 
 doubt include a large number of the mild forms of diabetes in 
 gouty and elderly people. Certainly, in the severe cases of 
 diabetes which have come under my observation, it has been 
 rare to detect any noteworthy enlargement of the liver during 
 life. 
 
 Obese diabetic patients sometimes suffer from gall stones ; 
 but in hospital patients this complication is exceedingly rare. 
 Probably it does not occur in more than 1 per cent, of the 
 cases. 
 
 The relation of liver affections to diabetes has been discussed 
 
THE LUNGS. 207 
 
 already on p. 116, and the diseases of the liver occasionally met 
 with are there referred to. 
 
 Pancreas. — Changes in this gland have been already described 
 (p. 140) ; but during life it is not yet possible to diagnose, with 
 certainty, whether a case of diabetes is, or is not, due to pan- 
 creatic disease. 
 
 Lancereaux has drawn attention to the association of disease 
 of the pancreas with a severe form of diabetes, in which there is 
 marked wasting, and in which the disease runs a rapid course. 
 But in diabetes of this form the pancreas is sometimes normal 
 (as is shown on p. 156); also the pancreas has sometimes been 
 found diseased when the patient has not been wasted. 
 
 As a rule, no evidence of pancreatic disease can be detected 
 in the diabetic patient on examination of the abdomen during life. 
 
 v. Noorden ( n ) is inclined to attach some importance to 
 the following symptoms as indications of pancreatic diabetes, 
 though their occurrence, and especially their combination, is 
 no doubt exceedingly rare : — 
 
 1. Proof of a tumour of the pancreas (carcinoma, cyst, 
 hydatid, etc.). 
 
 2. History of severe colic, which cannot be referred either 
 to the kidneys or liver, but which, from its position, is suggestive 
 of a calculus in the pancreatic duct. 
 
 3. The occurrence of maltose in the urine (found twice ; 
 importance not clear). 
 
 4. Steatorrhea or excess of fat in the motions, unassociated 
 with jaundice. 
 
 5. Presence of great quantities of nitrogenous material in 
 the fseces. 
 
 The Lungs. 
 
 The most important and most frequent lung complication 
 in diabetes is tubercular •phthisis. 
 
 As to the frequency of tubercular disease of the lungs in 
 diabetic patients, there is considerable difference of opinion. 
 v. Noorden ( 12 ) states that at least one-fourth of all diabetic 
 patients in Germany suffer from phthisis. According to some 
 writers, especially physicians who practise at the continental spas 
 which are much frequented by diabetic patients, tuberculosis is 
 not a common complication. But the discrepancies can be easily 
 explained. 
 
2 o8 SYMPTOMS AND PATHOLOGICAL CHANGES. 
 
 As has been pointed out by Seegen ( 13 ) and others, tuber- 
 cular disease is most common in poor, hard-working people. 
 These patients often do not come under treatment until a late 
 stage of the disease, and their general conditions of life and 
 surroundings usually predispose to phthisis. Hence phthisis is 
 common in diabetic hospital patients. Age is also of importance. 
 In young diabetic patients tuberculosis is frequent, and these 
 cases usually terminate either in diabetic coma, or in phthisis ; 
 whilst in elderly persons, especially obese or gouty patients, 
 tuberculosis is comparatively rare. 
 
 Physicians who practise at the various continental spas are 
 consulted chiefly by patients in good circumstances, and see a 
 large number of cases of the mild forms of the disease, such as 
 occur in stout or gouty elderly persons ; and these are just the 
 patients who suffer least from tubercular complications. 
 
 In Manchester, tubercular disease of the lung is a common 
 complication amongst hospital patients. 
 
 In the last 100 consecutive cases of diabetes, in which I 
 have made a careful examination of the lungs, I have obtained 
 the following results : — 
 
 Cases. 
 Physical signs of advanced phthisis . . . .14 
 (In twelve of these the signs were most marked at the 
 apex of the lungs ; in two, at the base.) 
 Slight signs of phthisis at the apex . . . .14 
 Tubercular phthisis found post-mortem, not detected 
 during life ........ 1 
 
 In 100 consecutive cases phthisis present in . . . 29 
 
 The majority of the cases were hospital patients, and most 
 were suffering from an advanced or severe form of diabetes. In 
 twenty-seven the signs were most marked at the apices of the 
 lung. Many of these were proved to be tubercular, either by 
 subsequent post-mortem examination, or by the detection of 
 tubercle bacilli in the sputum. Probably all of the twenty- 
 seven were tubercular. Of the two cases in which the physical 
 signs were at the base, one was proved to be non-tubercular (see 
 p. 212); in the other case the question was not definitely settled. 
 
 Next to coma, tubercular phthisis is the most common ter- 
 mination of diabetes. It is well known that when marked 
 tubercular phthisis has once developed in a diabetic patient, as- 
 
THE LUNGS. 
 
 209 
 
 a rule, death does not occur from coma. To this general rule 
 there are occasional exceptions, however. I recently saw a case 
 of diabetes in which there was very advanced tuberculosis of the 
 lungs, and yet death occurred from coma ; but such cases are 
 rare. In young patients, if death does not occur at a com- 
 paratively early stage from coma or some accidental inter- 
 current disease, tuberculosis of the lungs generally develops. 
 
 In the last forty-two fatal cases of diabetes which have come 
 under my observation, the termination has been as follows : — 
 
 Coma ......... 
 
 Asthenia from phthisis and lung disease (six tubercular 
 one non-tubercular) . 
 
 Acute double pneumonia 
 
 Cardiac failure 
 
 Gangrene of foot . 
 
 Asthenia from cirrhosis of the liver 
 
 Pysemic condition (multiple abscesses of the liver, cerebro- 
 spinal meningitis) ....... 
 
 Cases. 
 
 29 
 
 7 
 1 
 2 
 1 
 1 
 
 42 
 
 In the last twenty-four cases of diabetes which I have seen 
 or have made the autopsy, there has been — 
 
 Cases. 
 
 Tubercular disease of the lungs in . . 12 | _ . 
 
 \ 54 per cent. 
 
 . 11 
 
 jSTon-tubercular phthisis in . 
 Lungs not affected by phthisis in . 
 
 24 
 
 Tuberculosis of the lungs was thus found post-mortem in 
 about 50 per cent, of the cases; lung affections in 54 per cent. 
 [These cases were mostly hospital patients.] 
 
 Seegen ( u ) gives the following results of the post-mortem 
 examination of the lungs in ninety-two cases examined at the 
 Pathological Institute of the Vienna General Hospital during 
 the time Kundrat was professor of pathology 
 
 Tuberculosis of the lungs in 
 Croupous pneumonia 
 Lobular „ 
 
 (Edema of the lungs 
 
 '4 
 
 Carry forwt 
 
 ird 
 
 Cases. 
 
 40 
 
 9 
 
 8 
 13 
 
 70 
 
2 to SYMPTOMS AND PATHOLOGICAL CHANGES. 
 
 Cases. 
 Brought forward . .70 
 Hyperemia ......... 4 
 
 Diffuse gangrene of the lungs ..... 2 
 
 Localised gangrene through the rupture of an extensive 
 ulcerated cancer of the oesophagus .... 1 
 
 Hemorrhagic infarct ....... 1 
 
 11 
 Tubercular phthisis is then by far the most common form 
 of lung disease in diabetes. It was formerly stated that phthisis 
 in diabetes was frequently non-tubercular, but certainly this is 
 not correct. In all the cases of which I have seen the autopsy, 
 the lung affection has been tubercular with one exception (this 
 case will be referred to later). Frerichs ( 15 ) is very dogmatic 
 on this point. The following are the changes in the lung found 
 post-mortem in fifty-five of his diabetic cases : — 
 
 Tubercular disease (with pneumothorax in three) 
 
 Croupous pneumonia 
 
 Gangrenous pneumonia . ... 
 
 Lobular pneumonia with suppuration 
 
 Lungs normal in . 
 
 Cases. 
 
 21 
 
 1 
 
 7 
 
 1 
 17 
 
 It has been stated that there is often difficulty in detecting 
 tubercle bacilli in the sputum in the tubercular phthisis of 
 diabetic patients. I have not found this difficulty myself. 
 Gabbett's method of staining is the one I have always employed. 
 In seven consecutive cases which I recently examined, I found 
 tubercle bacilli readily in six ; in the other case the sputum 
 was frequently examined by myself and others with negative 
 results, and the case proved to be non-tubercular post-mortem. 
 Saundby ( 16 ) also states that he has not had any difficulty in 
 detecting tubercle bacilli in the sputum. 
 
 There are several peculiarities with respect to tubercular 
 phthisis occurring in diabetic patients, as was pointed out long 
 ago by Jaccoud ( 17 ), Leyden ( 18 ), Seegen, and others ( 19 ). Tuber- 
 cular lung disease in diabetes usually runs a comparatively 
 latent course. Frequently tubercular phthisis is found post- 
 mortem when the disease has not been diagnosed during life ; 
 and the post-mortem changes are nearly always much more 
 extensive than has been suspected from the symptoms and 
 physical signs. 
 
THE LUNGS. 211 
 
 Cough and expectoration are comparatively slight, as a rule. 
 The temperature is not much raised ; it may even be normal 
 (Leyden). Haemoptysis is very rare. In all the cases of tuber- 
 cular phthisis which I have seen, haemoptysis has been absent 
 even up to the last. 
 
 To these general rules there are exceptions. I have occa- 
 sionally seen abundant expectoration in later tubercular phthisis 
 of diabetic patients, and Leyden mentions that he has seen slight 
 haemoptysis in a few cases. 
 
 In two cases which I saw a short time ago, the phthisical 
 symptoms had chiefly attracted the attention of the patient and 
 his medical attendant, and the primary diabetes had not been 
 diagnosed. 
 
 Tubercular disease of the lung is a most serious complication 
 of diabetes. The disease often runs a rapid course, and I am 
 not aware that diabetic phthisis ever heals. 
 
 When phthisis, tubercular or non- tubercular, complicates 
 diabetes, often the glycosuria finally diminishes, and it sometimes 
 happens that sugar disappears from the urine shortly before 
 death. Three such cases have come under my observation : the 
 sugar disappeared a few days before death in one case, one week 
 before death in the second, and six weeks before death in the 
 third. All three had been most severe cases of diabetes, with 
 marked thirst and diuresis, and the urine had given a distinct 
 brownish-red coloration with perchloride of iron. With this 
 disappearance of sugar the thirst and diuresis also ceased, and 
 the whole clinical aspect of the cases, in the last stage of the 
 disease, gradually changed from diabetes to phthisis with marked 
 wasting. 
 
 The pathological changes in the lungs in diabetes have been 
 very carefully described by Dreschfeld ( 20 ). Pathologically, 
 the tubercular phthisis of diabetes is found to be due to a 
 chronic caseous tubercular broncho -pneumonia. It generally 
 runs a more rapid course than the tubercular phthisis of non- 
 diabetic persons. Caseation occurs rapidly, and the diseased 
 parts soon break down and form cavities. There is no tendency 
 to cicatrisation. Anatomically, the usual phthisical changes are 
 met with in the lungs — caseation, cavities, isolated tubercular 
 masses, and sometimes pneumothorax also (Dreschfeld). Micro- 
 scopically, the usual appearances are met with : peribronchial 
 infiltration, interstitial changes, broncho-pneumonic infiltration 
 
2i2 SYMPTOMS AND PATHOLOGICAL CHANGES. 
 
 of the alveoli, giant-cells, and endarteritis. Thickening of the 
 pleurse, with adhesions and occasionally pleuritic tubercles, is 
 also met with. 
 
 Leyden points out three pathological peculiarities, in a 
 number of cases of the tubercular phthisis of diabetes which 
 he reports. 
 
 1. That miliary tubercles were absent. 
 
 2. That giant-cells rarely occurred in the tubercular parts. 
 
 3. That obliterating arteritis was much more extensive than 
 in ordinary cases in non-diabetic subjects. 
 
 Tubercular ulcers are sometimes present in the intestines in 
 cases of diabetes complicated with pulmonary tuberculosis ; but 
 with the exception of the intestines, tuberculosis is generally 
 limited to the lungs, and tubercular affection of other organs is 
 very rare. Occasionally, however, tubercles have been found in 
 the larynx, kidneys, and liver. 
 
 Chronic 'pneumonic (non-tubercular) phthisis. — Whilst phthisis 
 in diabetic subjects is undoubtedly tubercular in the majority 
 of cases, still occasionally, though very rarely, a non-tubercular 
 form is met with — a chronic fibroid phthisis or chronic 
 pneumonic non - tubercular phthisis. As above mentioned, 
 in twenty-four cases of diabetes, at the autopsy tubercular 
 phthisis was present in twelve, non-tubercular in one. This is 
 the only case of non-tubercular phthisis I have ever seen 
 in a diabetic subject. It was under the care of Dr. Harris at 
 the Manchester Eoyal Infirmary, and is described by him in the 
 British Medical Journal, 28th November 1896. In this case 
 the sputum was repeatedly examined for tubercle bacilli with 
 negative results. Guinea-pigs were inoculated with the sputum 
 by Dr. Harris, but no tubercular disease was produced. Post- 
 mortem examination showed that the disease was non-tubercular. 
 
 Similar cases of non-tubercular phthisis in diabetic patients 
 have been recorded by Dreschfeld, and by Eoque, Devic, and 
 Hugounenq ( 21 ), in which no tubercle bacilli could be found, 
 either in the sputum during life, or in sections of the lung tissue 
 on pathological examination. 
 
 Gangrene of the lungs sometimes occurs in diabetic patients. 
 It is a very rare complication, and generally follows broncho- 
 pneumonia ; occasionally it follows croupous pneumonia and 
 severe forms of bronchitis, and occasionally it is due to trauma. 
 The expectoration has an acid reaction, and, as Frerichs has 
 
THE LUNGS. 213 
 
 pointed out, is generally not foetid. Dreschfeld states, how- 
 ever, that he has met with some cases of diabetic gangrene of 
 the lung in which the expectoration has had a very foetid 
 character. According to Frerichs, foreign bodies are frequently 
 found in the gangrenous patches ; he believes that these 
 bodies have been aspirated, and have given rise to gangrenous 
 pneumonia. 
 
 Broncho-pneumonia sometimes occurs. It may terminate in 
 caseation and lead to tubercular phthisis ; or it may terminate 
 in gangrene. According to Frerichs, it almost always leads to 
 gangrenous patches of greater or less extent. 
 
 Acute croupous pneumonia is a rare complication. It occurred 
 once in forty-two fatal cases in Manchester, nine times in 
 ninety-two cases reported by Seegen, and once in fifty-five of 
 Frerichs's cases. 
 
 According to Dreschfeld, " It runs a very acute course, and 
 is very fatal. It resembles the pneumonia seen in alcoholics ; 
 commences insidiously without a rigor; runs its course, as a 
 rule, without great rise of temperature, without perspiration, 
 and without expectoration, except where the case goes on to 
 gangrene." 
 
 During the attack of pneumonia, the amount of sugar ex- 
 creted in the urine has been found to diminish, but it increases 
 again at the crisis when the temperature falls. In thirteen 
 cases of pneumonia in diabetic patients, Senator never found 
 the sugar to disappear entirely from the urine during the course 
 of this complication. Sugar has been found in the expectoration 
 (0'25 per cent., Bussenius 22 ). 
 
 Occasionally other lung affections, such as emphysema and 
 chronic bronchitis, are met with. Bronchitis occurs chiefly in 
 the chronic forms of the disease in elderly persons. Pleurisy 
 and empyema are rare. Fat emboli have been found in the 
 lungs of patients who have died of diabetic coma, by Sanders 
 and Hamilton and others, and have been regarded as playing 
 an important part in the pathology of this complication ; but 
 at present this view has not many supporters. 
 
 The breath of diabetic patients has often a peculiar smell, 
 like that of decomposing sweet fruit (Seegen), rotten apples or 
 pears. In diabetic coma the breath has a peculiar smell like 
 chloroform or acetone, and Le Nobel has detected acetone in the 
 expired air. 
 
214 SYMPTOMS AND PATHOLOGICAL CHANGES. 
 
 Oxalates in the sputum. — -Fiirbringer ( 23 ) has recorded a case 
 of diabetes in which numerous octahedral crystals of oxalate of 
 lime, or amorphous conglomerations of that salt, were found in 
 the sputum. 
 
 The Heart. 
 
 The statements of various writers show that there is con- 
 siderable difference of opinion as to the frequency of cardiac 
 disease in diabetes mellitus. Possibly these discrepancies, like 
 many others, may be explained in part by the fact that some 
 writers have based their statistics on a series of cases which 
 included a large number of the mild forms of the disease in 
 elderly persons, whilst other writers have based their statistics 
 chiefly on severe forms of the disease, such as are met with in 
 young persons and in hospital patients. 
 
 In a large number of cases in which I have carefully 
 examined the heart clinically, I have generally failed to detect 
 any abnormality, or at least any abnormality of importance. 
 These cases have been chiefly hospital patients suffering from 
 a severe form of the disease. Thus, in the last 100 cases 
 which I have examined, only seven have presented symptoms or 
 physical signs of cardiac disease. In the remaining ninety-three 
 cases the heart has been practically normal until the disease has 
 reached a very advanced stage, when a feeble apex impulse and 
 feeble pulse have been noted. Also, when diabetic coma has 
 developed, the heart's action has become rapid and feeble. 
 
 The following are the changes which I have met with in the 
 series of 1 consecutive cases : — 
 
 1. Systolic apical murmur; slightly accentuated aortic second 
 sound; no cardiac enlargement. No other signs of cardiac 
 disease. 
 
 2. Eeduplicated first sound. No other signs of cardiac 
 disease. 
 
 3. Alcoholic cardiac dilatation, with signs of cardiac failure. 
 
 4. Systolic apical murmur. No other signs of cardiac 
 disease. 
 
 5. Cardiac dilatation and cardiac failure. (Mild cases of 
 diabetes.) 
 
 6. Dilated left ventricle ; apical systolic murmur. 
 
 7. Paroxysmal tachycardia. 
 
 Seegen ( 2i ) states that, with few exceptions, he has found the 
 
THE HEART. 215 
 
 heart normal, both as regards its size and the condition of the 
 valves. 
 
 There can be no doubt that in the majority of patients who 
 suffer from the severe forms of the disease, no abnormality of the 
 heart can be detected during life, with the exception of the weak 
 cardiac action at the last stage. Certain writers, however, have 
 found cardiac changes in a small proportion of the cases of 
 diabetes which they have examined. 
 
 The following are the cardiac changes which have been 
 described : — 
 
 Cardiac enlargement — hypertrophy and dilatation. — J. Mayer 
 ( 25 ) of Carlsbad found cardiac enlargement — hypertrophy or dila- 
 tation — in eighty-two out of 380 cases which he examined (i.e. 
 in 21 '6 per cent.) ; and in these cases there was no other disease 
 to account for the cardiac trouble except diabetes. Mayer also 
 found cardiac affections recorded in 13 per cent, of the cases of 
 diabetes examined in the Pathological Institute of Berlin. 
 Israel ( 26 ) found cardiac hypertrophy in 1 per cent, of cases of 
 diabetes examined pathologically. When the heart has been 
 hypertrophied, pathological examination has generally revealed 
 hypertrophy of the kidney also ; and this condition was present 
 in all of the 13 per cent, of cases of cardiac hypertrophy 
 referred to by Mayer. But, as will be pointed out on p. 219, 
 hypertrophy of the kidneys is often present when the heart is 
 normal or atrophied. 
 
 Israel and Mayer, from experiments and clinical observations, 
 conclude that the cardiac hypertrophy is due to the circulation 
 of some irritating chemical substance in the blood. 
 
 Cardiac weakness — atrophy and degeneration. — At the last 
 stage in many severe cases of diabetes, the pulse, the apex 
 impulse, and the heart's action become very feeble. This is not 
 surprising, considering that marked emaciation and pulmonary 
 tuberculosis are so often present. On post-mortem examination 
 in these cases, the heart is often very small, and the weight 
 greatly diminished. The muscle wall is sometimes pale, some- 
 times brownish red in colour. Fatty degeneration and glycogenic 
 degeneration of the cardiac muscle are sometimes met with. 
 
 Dreschfeld ( 27 ), Frerichs ( 28 ), Schmitz ( 29 ), and others have 
 drawn attention to the termination of diabetes with symptoms 
 of comparatively rapid cardiac failure and collapse. 
 
 The pulse is small and quick; the apex impulse is scarcely 
 
2 1 6 6" YMPTOMS AND PA THOL GICAL CHANGES. 
 
 perceptible ; the cardiac dulness is increased towards the right ; 
 and the first sound at the apex is very feeble (Schmitz). The 
 extremities are cold ; the patient becomes drowsy, and finally 
 comatose. In these cases the urine does not contain diacetic 
 acid. The patient is sometimes greatly wasted, sometimes well 
 nourished, sometimes even obese. The above symptoms are 
 those of one of the varieties of diabetic coma, and will be referred 
 to subsequently. They are often excited by some unusual 
 physical exertion, by over-strain, mental exertion, fright, anger, 
 etc., or by some error in diet. Post-mortem examination has 
 revealed atrophy of the heart, with fatty degeneration, and 
 sometimes glycogenic degeneration, and to these changes in the 
 cardiac muscles the symptoms have been attributed. 
 
 In the other two varieties of diabetic coma, as will be 
 described subsequently, the heart's action and pulse are very 
 rapid and feeble. 
 
 Valvular disease of the heart is sometimes met with as an 
 accidental complication, the patient generally having suffered 
 from rheumatism before or after the onset of the diabetic 
 symptoms. Sometimes aortic valvular disease is associated with 
 arterio-sclerosis in old-standing cases. 
 
 Functional disturbances — such as palpitation and paroxysmal 
 tachycardia — have been occasionally met with, and angina 
 pectoris is a rare complication. 
 
 In a case which recently came under my observation, the 
 patient suffered from very sudden attacks of tachycardia. 
 
 Pathological condition of the heart in twenty consecutive cases. 
 — In hospital patients suffering from a severe form of the disease, 
 the heart is usually atrophied, and the valves normal. 
 
 In the last twenty cases of diabetes in which I have seen or 
 made the autopsy, the heart has been enlarged in one only. 
 This was a very mild form of diabetes, or rather chronic glycos- 
 uria, and the cardiac dilatation was probably due to alcoholism. 
 The subject was a female, and the heart weighed 16 oz. In 
 fifteen out of the twenty cases the heart was very small, and its 
 weight much diminished. In one male subject the weight was 
 only 6h oz. (normal weight of male heart 11 oz.) ; in one female 
 subject the heart only weighed 4 oz. 30 grs. (instead of 9 oz., the 
 normal weight of the female heart). The heart muscle was often 
 soft and flabby. In one case there was atheromatous thicken- 
 ing of the aortic valves, and in another the same condition of 
 
THE HEART AND KIDNE YS. 217 
 
 both aortic and mitral valves. In the other eighteen cases the 
 valves were normal. 
 
 All the cases were hospital patients, and nineteen suffered 
 from a severe form of the disease with marked wasting. 
 
 The pulse. — The pulse is usually regular and of normal 
 frequency. The tension is often normal, occasionally a little below 
 normal ; in other cases the pulse is hard, large, and of high tension. 
 I have sometimes found the pulse remarkably hard, and the 
 tension exceedingly high when there has been no kidney mischief, 
 and the patient has been under middle life. 
 
 Arterio-sclerosis is often a complication of diabetes. The 
 radial and other arteries frequently feel thickened. In many 
 cases of diabetes, if three fingers be placed on the radial artery, 
 and firm pressure be made with the finger nearest the elbow, so 
 that the pulsation is arrested, then the artery can be rolled 
 beneath the two fingers, as a hard cord, on the peripheral side of 
 the point of pressure. In other superficial arteries the same 
 condition is frequently present also. In some patients, especi- 
 ally in elderly persons, very marked atheroma can be detected ; 
 but even in young diabetics I have occasionally met with 
 very marked arterio-sclerosis. Now, as diabetes and arterio- 
 sclerosis are both diseases most commonly occurring after the 
 age of 45, and as arterio-sclerosis is a common affection, 
 it is not surprising that in many elderly diabetics the arteries 
 should present this pathological condition. But the frequency 
 of arterial thickening in diabetics under 45 appears to indicate 
 that the association is not always accidental, and that probably 
 in certain cases there is some connection between the two 
 diseases. The relation between arterio-sclerosis and diabetes is 
 discussed on p. 156. 
 
 The Kidneys. 
 
 It has already been pointed out (p. 175) that slight albu- 
 minuria is frequently met with in diabetes ( 30 ), but that it is 
 somewhat rare to find albumin present in the urine in large 
 quantities. In most of the cases in which albumin is present in 
 small quantity, there are no other indications of kidney lesions, 
 and post-mortem there are no signs of actual nephritis, though 
 slight changes may be present in the renal epithelium. 
 
 This slight albuminuria has been attributed to excess of 
 nitrogenous food in some cases, to catarrh of the bladder in 
 
2 1 8 S YMPTOMS A ND PA THOL O GICAL CHANGES. 
 
 others (E. Schmitz). I have found a trace of albumin present, 
 however, when the patient has been taking ordinary diet. As 
 mentioned on p. 177, it appears probable that occasionally the 
 trace of albumin is the result of a balanitis and the mixture 
 of a little pus from the inflamed parts with the urine. 
 
 In only four out of 100 consecutive cases of diabetes (mostly 
 hospital patients suffering from a severe form of the disease) 
 were there indications of nephritis and abundant albuminuria 
 (see p. 176). 
 
 One Hundred Consecutive Cases. 
 
 Cases. 
 
 Albumin absent when urine first ex- 
 amined in 70 
 
 Small quantity or a mere trace of 
 albumin in 26 
 
 Large or considerable amount of albu- 
 
 In fourteen of these a trace of albumin 
 
 appeared at a later date. 
 Seven of these were at a very advanced 
 
 stage. One was comatose when first 
 
 examined. In two the albuminuria 
 
 disappeared later. 
 
 100 
 
 In the chronic mild form of diabetes in elderly people, the 
 proportion of cases in which a large amount of albuminuria is 
 -present is probably much greater. 
 
 For other points with reference to the albuminuria of 
 diabetes, see p. 176. 
 
 In diabetes, therefore, two forms of albuminuria may occur : 
 (1) very slight albuminuria not associated with nephritis; (2) 
 albuminuria due to nephritis. In these cases the albumin is 
 abundant, or there are other indications of Bright's disease. 
 When there are signs of interstitial nephritis, the urine is 
 abundant, but the specific gravity diminishes, though it often 
 still remains above normal. When the kidney condition 
 is one of parenchymatous nephritis, the quantity of urine is 
 less than in the interstitial variety, but almost always above 
 the normal amount. 
 
 In cases of diabetes presenting signs of nephritis, often the 
 albumin and sugar bear no relation to each other ; but in some 
 cases, as the kidney changes advance, the albumin increases, and 
 the sugar decreases. 
 
 When diabetes is complicated with granular kidney after 
 the renal changes reach a certain stage, the glycosuria gradually 
 diminishes, and finally disappears, and the symptoms remaining 
 are those of chronic interstitial nephritis. This rare termina- 
 
THE KIDNEYS. 219 
 
 tion may occur in cases of diabetes complicated with obesity or 
 with gout, and, according to v. Noorden, scarcely ever in any 
 other forms. The disappearance of the diabetic symptoms, and 
 their gradual replacement by those of chronic nephritis of either 
 form, is a bad prognostic sign. The cases of diabetes compli- 
 cated by actual chronic nephritis — as indicated by symptoms or 
 post-mortem evidence — are rare in hospital patients and in the 
 more severe forms of the disease ; they are more common in 
 elderly persons suffering from the mild forms, and in the dia- 
 betes of obese or gouty persons. 
 
 (Edema occasionally occurs in the feet in diabetes when 
 albumin and signs of nephritis are absent (see p. 227). 
 
 Pathological changes in the kidneys. — The following are the 
 changes which were met with in the kidneys, post-mortem, in 
 ninety-two cases of diabetes examined at the Vienna Pathological 
 Institute, from 1870—1892, during the professorship of Dr. 
 Kunclrat (quoted by Seegen 31 ) : — 
 
 Cases. 
 
 Parenchymatous and fatty degeneration . . .in 24 
 
 Acute hsemorrhagic nephritis . . . . . ,, 1 
 
 Granular kidney . . . . . . . ,, 9 
 
 Chronic tuberculosis of the kidney . . . . ,, 2 
 
 Hypertrophy (in one of these cases small cysts were 
 
 present) . . . . . . . . ., 4 
 
 In twenty consecutive cases of diabetes amongst hospital 
 patients in which I have recently seen or made the autopsy, the 
 condition of the kidneys (on macroscopic examination) was as 
 follows : — 
 
 Eenal hypertrophy was present in eight. (In one male 
 subject the kidneys weighed 11 oz. each; in another, 10| and 
 12 oz. respectively.) 
 
 The kidneys were normal or slightly diminished in size in 
 twelve. 
 
 In one case there were tubercles in the kidneys. The only 
 other change met with was hyperemia. 
 
 In none was there any naked-eye evidence of nephritis 
 (parenchymatous or interstitial), and in none was there any 
 indication of nephritis during life. 
 
 The macroscopical changes ( 32 ). — Hypertrophy of the kidneys. — 
 On post-mortem examination the kidneys are often enlarged in 
 diabetes. As already mentioned, Mayer, Israel, and others have 
 
220 SYMPTOMS AND PATHOLOGICAL CHANGES. 
 
 found cardiac hypertrophy associated with hypertrophy of the 
 kidneys in some cases. Cardiac enlargement, however, is cer- 
 tainly very rare in the severe forms of diabetes which have 
 come under my observation in hospital practice, and yet in 
 these forms the kidneys were often enlarged (eight out of 
 twenty cases). In none of these cases was there any cardiac 
 hypertrophy ; in six the heart was diminished in size. 
 
 Though the kidneys are often congested and enlarged, on 
 the other hand they are sometimes small and pale ; but increase 
 in the connective tissue is rare in the severe forms of diabetes. 
 
 Nephritis. — Occasionally chronic interstitial nephritis is 
 met with, but almost exclusively in obese or gouty patients. 
 Parenchymatous nephritis is also an occasional complication. 
 A diffuse nephritis, in which the parenchyma is chiefly affected, 
 but in which there is a certain degree of interstitial sclerosis 
 also, has been found in a few cases. 
 
 Abscesses, amyloid degeneration, tubercles, and fat emboli 
 are changes which have been occasionally, though very rarely, 
 observed. 
 
 Microscopical changes in the renal epithelium. — Often the renal 
 epithelium cells present microscopical changes. 
 
 (a) Hyaline degeneration of renal epithelium. — Cantani has 
 recorded the changes observed by Armanni ( 33 ) in three of his 
 cases. These consisted of a hyaline degeneration of the 
 epithelial cells. The cells affected presented a swollen trans- 
 lucent appearance, as if transformed into large hyaline vesicles 
 with distinct cell walls. The nuclei stained well, and were often 
 seen pushed towards the periphery of the cells. 
 
 Armanni described these changes in the cells of the collecting 
 tubules and the tubuli recti of the medulla. Cantani attributes 
 them to a dropsy of the cells. 
 
 Other observers have recorded similar changes. Stephen 
 Mackenzie ( 34 ) has found this condition of epithelium only in the 
 cells of the collecting tubes ; Saundby found the changes con- 
 fined to the cells of Henle's tubes ; Ebstein detected them in 
 the loop of Henle. 
 
 (b) Necrosis of epithelium. — Ebstein ( 35 ) has drawn attention 
 to a necrosis of the renal epithelium cells in diabetes, similar to 
 the " coagulation necrosis " described by Weigert, in other 
 diseases. The nuclei of the cells gradually atrophy ; they do 
 not stain or stain badly with the usual staining agents ; the 
 
THE KIDNEYS. 221 
 
 protoplasm of the cells degenerates and becomes granular, and, 
 finally, simple or fatty detritus is found in. many places. 
 
 (c) Fatty degeneration of renal epithelium. — Fichtner ( 36 ) has 
 described a form of fatty degeneration of the renal epithelium of 
 the greater part of the kidney cortex in cases of diabetic coma. 
 He regards the following points as characteristic : (1) the 
 arrangement of fat globules in a row at the periphery or attached 
 part of the cells lining the tubules ; (2) the affection of those 
 renal tubules only which are lined by the so-called " cloudy " 
 epithelium. 
 
 Fatty degeneration of the renal epithelium is frequently met 
 with, however, in other chronic diseases, such as cancer, phthisis, 
 pernicious anaemia, and in many febrile affections. 
 
 (d) Glycogenic degeneration of renal epithelium. — Ehrlich and 
 Frerichs ( 37 ' 38 ) have described a glycogenic degeneration of the 
 renal epithelium of Henle's tube. This they state to be a con- 
 stant change. The glycogenic degeneration can be detected 
 macroscopically by treating the section of the kidney with 
 Lugol's iodine and iodide of potassium solution. By the 
 action of this reagent, small brown streaks are produced about 
 the junction of the medullary and cortical parts of the kidney. 
 They correspond to the markedly dilated tubules at the isthmus 
 of Henle's loop. Under the microscope the epithelium cells 
 are found to be enlarged at this point ; they are polygonal, 
 clear, and, when stained with iodine gum, they are seen to 
 contain large or small masses which are coloured more darkly 
 than the protoplasm of the cells. The protoplasm is stained pale 
 yellow ; the darkly stained parts vary in tint from yellow up to 
 pure mahogany colour. 
 
 Ehrlich and Frerichs think that this change is due to the 
 presence of glycogen in the epithelium cells at the most narrow 
 part of the uriniferous canal ; they believe that absorption of 
 sugar probably occurs, and that this is transformed afterwards 
 into glycogen. 
 
 Straus ( 39 ) believes that the hyaline changes described by 
 Armanni, and the glycogenic changes of Ehrlich, are really of the 
 same nature. He thinks they occur very frequently, though not 
 constantly ; and, when present, he regards them as characteristic. 
 In a later paper ( 40 ) he points out that in some cases in which the 
 hyaline changes described by Armanni are particularly distinct, 
 glycogen cannot be detected in the cells. In such cases he believes 
 
222 £ YMPTOMS AND PA THOLOGICAL CHANGES. 
 
 that an infiltration of the cells with glycogen had existed at one 
 time, but that the glycogen had disappeared during the life of 
 the patient, leaving only the hyaline changes. Straus believes 
 the hyaline changes to be more frequent than the glycogenic, 
 but thinks both have the same significance, and are equally 
 characteristic. 
 
 Cystitis occasionally occurs in diabetes. It is due to the 
 irritation of the mucous membrane of the bladder by the 
 saccharine urine. Schmitz ( 41 ) recognises three forms of chronic 
 cystitis in diabetic patients. In the mildest form, subjective 
 symptoms are absent ; the urine is faintly acid ; it contains a 
 few pus corpuscles, calcium phosphate, and bacteria. In the 
 second form the urine is turbid, the smell objectionable, the 
 reaction very faintly acid or neutral ; pus corpuscles, triple 
 phosphates, calcium phosphate, and bacteria are present ; mic- 
 turition is frequent. In the third form the ordinary subjective 
 symptoms of cystitis are present ; the smell of the urine is 
 offensive ; the reaction is ammoniacal ; pus cells, triple phos- 
 phates, calcium phosphate, urate of ammonia, fungi, and bacteria 
 are present. A portion of the sugar in the urine undergoes 
 fermentation, carbonic acid gas is formed and often collects 
 within the bladder, and is passed with the urine. Cystitis has 
 been present twice only amongst the last 140 cases which have 
 come under my observation. 
 
 The Skin. 
 
 In severe cases of diabetes the skin is dry, and feels rough 
 when touched ; but sometimes it appears and feels normal ; 
 and occasionally diabetic patients perspire freely, even when 
 lung disease or other complications, liable to give rise to per- 
 spiration, are absent. Unilateral sweating has been recorded 
 in diabetes, but of course it is exceedingly rare. The sweat 
 of diabetic patients has been examined for the presence of 
 sugar, with varying results. Griesinger, Vogel, Fubringer, and 
 Forster have been able to detect sugar in the sweat ; bub 
 Frerichs and others have not been able to detect a trace, 
 v. Noorden examined the sweat of six diabetic patients 
 after the injection of pilocarpine, but even with the phenyl- 
 hyclrazin test could not obtain any evidence of sugar. The 
 skin of diabetic patients, especially in the severe forms, often 
 
THE SKIN. 223 
 
 feels cold when touched. During diabetic coma, coldness of the 
 skin is a common symptom. 
 
 Itching of the skin, pruritus, is sometimes troublesome ; 
 occasionally it is one of the first symptoms. It is diminished 
 by a diet or treatment which causes a reduction in the sugar 
 excretion. 
 
 This pruritus may be general or local. General pruritus is 
 very rare ; it has been attributed to irritation of the peripheral 
 cutaneous nerves owing to dryness of the epidermis, or to the 
 circulation of fluids containing an excess of sugar. Another 
 explanation is, that central irritation is projected towards the 
 periphery (v. Noorden). Local pruritus, in the genital organs, is 
 a common symptom. It is more frequent in females ; and some- 
 times medical advice is sought on account of pruritus or eczema 
 of the vulva, this being the first symptom of the disease which 
 has attracted the patient's attention. Hence, in all cases of 
 pruritus or eczema of the vulva, the urine ought to be examined 
 for sugar. The parts affected first are the labia minora, then 
 the labia majora, and, finally, the skin of the thighs adjacent to 
 the genital organs. All those parts with which the urine may 
 come in contact are liable to be affected. The itching is excited 
 by the irritation of the sugar in the urine and by the growth of 
 fungus on the genital organs. Pruritus is usually accompanied 
 by congestion and redness of the vulva, and may be followed by 
 general eczema. 
 
 In the male, pruritus of the glans penis and prepuce some- 
 times occurs, and occasionally the scrotum is affected ; but 
 these conditions are much more rare than pruritus of the 
 vulva. 
 
 Eczema and erythema. — The genital organs are most com- 
 monly affected owing by their irritation by saccharine urine and 
 fungus growth. But sometimes eczema and erythema occur in 
 parts with which saccharine urine does not come in contact. 
 
 Eczema of the vulva is the most common affection. The 
 irritation of saccharine urine causes pruritus and a burning 
 sensation in the region of the vulva. The parts become 
 congested, and, owing to the irritation, excoriations are pro- 
 duced ; also fungus spores and mycelia develop in the mucous 
 membrane, which, being frequently wet with the saccharine 
 urine, forms a suitable soil for their growth. Both the con- 
 ditions lead to the development of eczema. On the affected 
 
224 SYMPTOMS AND PATHOLOGICAL CHANGES. 
 
 parts whitish crusts and scabs are found, which consist of 
 epithelial scales and dried secretion mixed with fungus spores 
 and mycelia. Eczema of the vulva may lead to dermatitis 
 or hypertrophic vulvitis ; or, if septic organisms penetrate 
 deeply into the affected parts, boils and phlegmonous vulvitis 
 may develop ; sometimes the eczema extends to the mucous 
 membrane of the vagina. Pruritus and eczema of the vulva are 
 much more common in dirty patients than in clean patients. 
 
 The majority of cases of eczema of the vulva occurring 
 about the climacteric period of life are due to diabetes. 
 
 The amount of sugar in the urine is sometimes only small, 
 and often there is great improvement under treatment, especially 
 in females at the climacteric period. 
 
 In the male, pruritus of the penis or scrotum may occur, 
 owing to the irritation of the saccharine urine. Erythema 
 around the meatus is liable to develop, and this may spread 
 over the glans penis ; finally, a balanitis may be produced, and 
 fungus spores and mycelia may develop under the prepuce. 
 This condition is met with chiefly in patients who have a long 
 prepuce. The mucous membrane may become fissured and 
 thickened, the prepuce may lose its elasticity, and finally 
 phimosis may develop. The integument of the prepuce may 
 also become thickened ; sometimes the prepuce becomes very 
 (edematous, and even sloughing and gangrene have been observed. 
 Pus cells and sometimes fungus spores and mycelia may be 
 washed away from the penis, and cause a trace of albumin and 
 a slight deposit in the urine. 
 
 Eczema of the male genital organs is occasionally met with : 
 it starts from the glans and prepuce, and may extend to the 
 neighbouring parts. Sometimes white patches of fungus growth 
 are found on the glans penis and prepuce. 
 
 Thus in a case of diabetes of four years' duration, the end 
 of the prepuce was red, thickened, inflamed, and cedematous, and 
 there was also phimosis. The mucous membrane at the end of 
 the prepuce was studded with small white patches, which, on 
 microscopical examination, were found to consist of fungus 
 growth and epithelial scales. 
 
 In another case, a severe form of diabetes in a youth of 
 19, the prepuce was* inflamed and thickened, and when retracted, 
 a number of small white patches were seen on the glans penis. 
 Scrapings from these patches, when examined microscopically, 
 
E CZEMA—B OILS— CA RB UNCLES. 2 2 5 
 
 were found to consist of epithelial cells, a few pus corpuscles, 
 and fungus spores and mycelia. 
 
 Boils. — Boils are amongst the most common of the skin 
 lesions in diabetes. They often occur at an early stage, and 
 sometimes they are the first symptoms noticed by the patient. 
 Seegen states that he has never seen boils in an advanced 
 stage of the disease. In the advanced cases of diabetes admitted 
 into the Manchester Eoyal Infirmary, certainly they are very 
 rare. Boils generally occur in stout patients ; they may be 
 single or multiple ; they usually develop upon the neck, back of 
 the shoulders or buttocks, but they may occur at any part. In 
 females, sometimes they form on the labia of the vulva. 
 
 Boils are, of course, due to other causes besides diabetes ; 
 but a large proportion of all cases (one-fourth according to 
 v. Noorden, one-third according to Marechal) are due to this 
 disease. In diabetic patients, as in other subjects, infection 
 with micro-organisms is the exciting cause, and pure cultures of 
 the staphylococcus aureus have been obtained from the boils of 
 diabetic patients. 
 
 Carbuncles are less frequent than boils ; they may be 
 amongst the earliest symptoms, however, which have attracted 
 the patient's attention, and he may come under treatment for 
 this complication before diabetes has been diagnosed. Car- 
 buncles may also appear at an advanced stage of the disease. 
 They occur mostly in the neck, occasionally on the face or other 
 parts of the body. They have a tendency to extend and become 
 gangrenous, or to give rise to surrounding cellulitis, or to great 
 destruction of tissue ; and very often they lead to a fatal ter- 
 mination. Like boils, carbuncles are excited by the action of 
 micro-organisms. They occur in other diseases besides diabetes, 
 but the two affections are so frequently associated, that it is 
 important to examine the urine of every person who suffers 
 from carbuncles. 
 
 Gangrene is a complication sometimes met with. It may be 
 primary, or secondary to wounds, contusions, boils, carbuncles, 
 or cellulitis. 
 
 The lower limbs are most often affected, the gangrene com- 
 mencing in the toes, and not infrequently at a part which is 
 subject to the pressure of a tight boot. Just as in the case of 
 boils and carbuncles, gangrene is sometimes the first symptom to 
 attract attention to diabetes in an apparently healthy person, — 
 IS 
 
226 SYMPTOMS AND PATHOLOGICAL CHANGES. 
 
 hence the necessity of examining the urine for sugar in all cases 
 of gangrene. 
 
 Diabetic gangrene may be moist or dry. Before early 
 middle age, spontaneous gangrene does not occur, or occurs only 
 very rarely in diabetic patients. I have never seen gangrene 
 amongst the numerous cases of the severe forms of the disease 
 under the age of 40 at the Manchester Royal Infirmary. All 
 the cases have been over that age. 
 
 G-odlee ( 42 ) points out that there are three exciting causes of 
 gangrene in diabetic patients: (1) Inflammatory conditions; (2) 
 atheroma of the vessels; (3) neuritis in the peripheral nerves. 
 
 A large proportion of cases of diabetic gangrene of the leg are 
 due either to atheroma or neuritis. 
 
 Koenig( 43 ) attaches great importance to the association of 
 extensive atheroma of the arteries with diabetic gangrene. He 
 states that thrombosis of the larger vessels is often present, and 
 generally there is marked atheromatous degeneration of the 
 smaller arteries (nine out of eleven cases examined). Some- 
 times both atheroma and neuritis play a part in the development 
 of diabetic gangrene. 
 
 When gangrenous inflammation appears in a diabetic patient, 
 severe symptoms develop; the patient becomes drowsy, and 
 delirium, loss of appetite, and coma may occur. These 
 symptoms are more marked when the gangrene is moist than 
 when it is dry. 
 
 Wounds of the skin heal badly in diabetic patients, and 
 sometimes become gangrenous. Operation incision wounds 
 generally heal badly also, and though the results are now much 
 better than formerly, surgeons still avoid operations on diabetic 
 patients if possible. The operation for phimosis is said to be 
 particularly unsatisfactory (v. Noorden). There is one exception 
 to the above statement, i.e. the operation for diabetic cataract, 
 which is now very often successful. 
 
 Perforating ulcers, chiefly on the soles of the feet, and about 
 the toes, especially about the big toe, are occasionally met with 
 in diabetes. The ulcers resemble those seen in locomotor ataxia, 
 and the knee-jerks are frequently absent in these cases. Hence, 
 if the urine be not examined, a diagnosis of early locomotor 
 ataxia is liable to be made. Often the patients complain of 
 pains in the legs, and the calf muscles are tender. Sometimes 
 there are other signs of peripheral neuritis. 
 
PERFORATING ULCERS— VASOMOTOR CHANGES. 227 
 
 Pathologically, parenchymatous neuritis of the peripheral 
 nerves of the legs has been found in some of the cases of 
 diabetic perforating ulcer, and it is probable that the ulcers are 
 due to neuritic changes ( 44 ). 
 
 Often the starting-point of a perforating ulcer is a large 
 corn or bunion, which is cut by the patient, the skin around and 
 beneath being injured in the process. In other cases a patch of 
 superficial necrosis develops ; this separates, and an ulcer forms. 
 Sometimes a hemorrhage, just beneath the skin, appears to be 
 the starting-point of a superficial necrosis and ulcer ( 45 ). 
 
 It is important to examine the urine for sugar in all patients 
 who suffer from perforating ulcers of the feet. 
 
 Amongst the last 140 cases of diabetes which have come 
 under my observation, perforating ulcers have been present in 
 four only. The ages of the four patients were 52, 55, 55, and 
 67 respectively. All were suffering from a mild form of the 
 disease. In the first case, distinct symptoms of peripheral 
 neuritis were also present, and the knee-jerks absent ; in the 
 second case, the knee-jerks were absent ; in the third case, one 
 knee-jerk was absent, the other very feeble ; and in the fourth 
 case, both knee-jerks were present. In one case the ulcer was 
 just below the external malleolus ; in one case in the centre 
 of the sole ; and in two cases near the toes. 
 
 Vasomotor changes — bulbous fingers. — In three cases of 
 diabetes I have observed a peculiar condition of the fingers 
 The extremities of the fingers were much swollen, bulbous, 
 hyperemia, and of a bright red colour ; and the patients 
 complained of burning and tingling of the finger-tips. Some- 
 times there was also hyperemia of the palms of the hands. The 
 toes were affected in a similar manner, but to a less degree. One 
 of these patients was suffering from advanced phthisis, another 
 from early phthisis, and in the third, signs of phthisis appeared 
 soon after the bulbous condition of the fingers was noted. 
 
 Spontaneous shedding of the nails has been recorded in a 
 few cases of diabetes ( 4G ). 
 
 Anasarca without any signs of cardiac failure, and without 
 the presence of albumin in the urine, is an occasional complica- 
 tion of diabetes. It has not been well marked in more than 
 5 per cent, of cases which have come under my observation. 
 The oedema is chiefly in the legs, and there is pitting of the skin 
 about the ankle, on the dorsum of the foot, and over the tibia. 
 
228 SYMPTOMS AND PATHOLOGICAL CHANGES. 
 
 Occasionally oedema affects the hands, face, and other parts. 
 Thus in the case of a diabetic patient, ret. 34, under the care 
 of Dr. Leech at the Manchester Infirmary, after the disease 
 had been present for five years, there was well-marked oedema 
 of the feet, scrotum, and penis. The oedema varied from time 
 to time ; some days it was almost absent ; on other occasions, 
 after the patient had been walking about for some time, it was 
 very well marked. There was no albumin present in the urine, 
 and no indication of cardiac or vascular trouble. At a later 
 date, there was also great oedema of the dorsum of each hand, and 
 marked pitting on pressure, but this disappeared rapidly after 
 rest in bed. Sir William Roberts ( 47 ) draws attention to a slight 
 pitting over the tibiae, which he attributes to the soft atonic 
 state of the subcutaneous tissues rather than to true oedema. 
 Frerichs ( 4S ) found oedema, without kidney mischief, in twenty- 
 five out of 400 cases. Ascites with oedema of the arms and 
 hands is mentioned by Roberts, who also quotes a similar case 
 recorded by Fischer. I have seen two cases of slight ascites in 
 diabetes and one marked case, but in the latter the liver was 
 cirrhosed ; in the other two there were no signs of liver disease. 
 
 Xanthoma diabeticorum. — This is a very rare complication. 
 Morris 19 ' 50 ' 51 ) has collected twenty-one cases from literature, 
 and the following account is based chiefly on his description. 
 The affection is evidently directly due to the diabetic conditions. 
 When the excretion of urine and sugar diminishes the eruption 
 disappears, but reappears with the return of the glycosuria. The 
 skin eruption consists of small papules or nodules about the size 
 of a pea ; they have a rounded or conical form, and may be 
 discrete or confluent. Some are surmounted by a yellowish 
 apex, which gives them a deceptive resemblance to pustules of 
 solid consistence. They are firm to the feel, and are essentially 
 inflammatory. Subjective sensations of heat, burning, and itching- 
 are present. Sometimes the nodules have a reddish tinge, and 
 only appear yellow when the skin is stretched. They are found 
 on the forearms, buttocks, and knees, and may extend to the 
 scalp, face, and trunk ; but in some cases the eruption is universal. 
 The papules may remain stationary for months or years, and 
 then undergo rapid involution and leave no trace behind. They 
 occasionally re-develop two or three times. The presence of this 
 eruption may be the first indication of diabetes. 
 
 Morris gives the following points of difference between 
 
XANTHOMA DIABETICORUM. 229 
 
 diabetic xanthoma and the non-diabetic varieties: — (1) While 
 ordinary xanthoma planum and multiplex is slow in evolution 
 and generally permanent, xanthoma diabeticorum appears 
 suddenly and subsides almost as suddenly. (2) While xanthoma 
 planum almost invariably begins on the eyelids, and is usually 
 confined to them, and whilst xanthoma multiplex chiefly affects 
 the flexor and extensor surfaces of the limbs, the eyelids, and the 
 palms, the diabetic variety for the most part attacks the neck, 
 trunk, and extensor surfaces of the limbs. (3) While jaundice is 
 a very frequent accompaniment of ordinary xanthoma in adults, 
 and diabetes mellitus has never been recorded as occurring in 
 association with it, the reverse is the case as regards xanthoma 
 diabeticorum, in which sugar in the urine is a constant feature, 
 and jaundice an unknown complication. 
 
 But Morris thinks that xanthoma planum and xanthoma 
 multiplex are members of the same group, and that anatomically 
 the three varieties are practically identical. 
 
 In addition to the above-mentioned skin complications, the 
 following affections have also been described in association with 
 diabetes : — Psoriasis, acne cachecticorum, dermatitis herpetiformis, 
 herpes zoster with persistent neuralgia, eczema tous impetigo, 
 lichenoid eruptions, chronic papular urticaria, purpura hamior- 
 rhagica, erysipelas, and dermatitis diabetica papillomatosa 
 (Kaposi). 
 
 Eecently attention has been drawn to a form of diabetes 
 associated with bronzing of the skin, and the disease has been 
 described as diabete bronzL An account of this affection is given 
 on p. 308. 
 
 The Eyes. 
 
 Defects of vision are not uncommon in diabetes, and 
 sometimes the patient first seeks medical advice on account 
 of ocular symptoms. Some of these ocular affections are 
 merely accidental complications, or are only very indirectly 
 the results of the diabetes, whilst others are directly caused by 
 the disease. The latter group includes — (1) Cataract; (2) pure 
 accommodation paralysis in middle age; (3) short-sightedness, 
 developing later in life, between the ages of 40 and GO years, 
 without any opacity of the lens (Hirschberg 52 ) ; (4) vitreous 
 opacities; (5) retinitis; (6) amblyopia, like tobacco amblyopia. 
 Retinitis and amblyopia are very rare. 
 
2 3 o SYMPTOMS AND PATHOLOGICAL CHANGES. 
 
 Cataract. — This is the most common ocular affection, which 
 gives rise to marked defect of vision in diabetic patients. 
 
 Diabetic cataract is usually bilateral ; it is met with in 
 children, as well as in adults and old persons. I have seen it in 
 a girl set. 11, in another set. 15, and in a youth of 20. It is 
 generally of the soft variety, but not invariably, and in old 
 diabetic patients it may present the same characters as in 
 non-cliabetic patients. It does not disappear, as a rule, under 
 anti-diabetic treatment (Hirschberg). Seegen ( 5S ), however, men- 
 tions two cases in which the opacity of the lens diminished 
 with improvement of the symptoms and the diminution of the 
 sugar excretion, but increased when the patient became worse 
 again. A case of spontaneous disappearance has been reported 
 by Nettleship ( 54 ). 
 
 In persons under middle age, diabetic cataract generally 
 develops quickly ; and the rapid development of double cataract 
 in a young person ought always to raise the suspicion of diabetes. 
 Diabetic cataract can now be removed successfully, and the 
 results are almost as good as in non-diabetic cases. The opera- 
 tion is sometimes followed by diabetic coma, however. 
 
 In the last 100 consecutive cases of diabetes (mostly at an 
 advanced stage of the disease), in which I have examined the 
 eyes, cataract was present in nine. The ages of the patients 
 were— 12, 15, 20, 40, 40, 40, 47, 56, and 59 years re- 
 spectively. 
 
 Cataract is not. confined to markedly wasted patients; the 
 patients may even be well nourished. Seegen points out that 
 a large amount of sugar is always present in the urine, and 
 states that he has never seen diabetic cataract when the urine 
 contained only a small quantity of sugar, though in these cases 
 other visual symptoms are often present. A satisfactory expla- 
 nation of the origin of diabetic cataract has not been given. It 
 has been attributed to general marasmus, to the presence of 
 sugar in the lens, to the abstraction of water from the lens 
 owing to the sugar in the adjacent media, and to the conversion 
 of grape sugar in the aqueous humour into milk sugar. But 
 there are objections to all these views. Eecently vascular disease 
 in the ciliary processes has been regarded as the cause of nutri- 
 tional changes in the lens ( 55 ). 
 
 Diabetic retinitis. — Retinal changes in diabetes mellitus were 
 recorded first by Edward Jaeger, and have since been described by 
 
CATARACT— DIABETIC RETINITIS. 231 
 
 various observers during the last forty years. In a large pro- 
 portion of the cases both sugar and albumin have been present 
 in the urine, and in these cases, therefore, it is somewhat difficult 
 to say, in the absence of a pathological examination of the 
 kidneys, whether the retinal changes were not entirely or in 
 part the result of renal disease. But in other cases retinal 
 changes have been found, when there has been no albumin 
 in the urine. There is undoubtedly, then, a form of retinitis 
 occurring in diabetic patients, which is not an " albuminuric : ' 
 retinitis. The retinitis of diabetes also differs somewhat from 
 that of chronic Bright's disease. 
 
 It appears to me that the frequency of retinitis in diabetic 
 patients has been over-estimated, however. Many text-books 
 of medicine simply make the statement that retinitis is met 
 with in diabetes, and give no information or hint as to the 
 frequency of this complication. Hence the medical student 
 often receives the impression from such book, that retinitis 
 occurs as commonly in diabetes as in chronic Bright's disease ; 
 but, as a matter of fact, diabetic retinitis is rare. The per- 
 centage of cases of diabetes in which retinal changes are met 
 with is very small — much smaller than the proportion of cases 
 of retinitis in chronic Bright's disease. Also, when present, the 
 retinal changes appear to be less extensive, as a rule. Eales has 
 estimated that retinitis is met with in 28 per cent, of the cases 
 of chronic Bright's disease ; Gowers ( 59 ) also estimates the pro- 
 portion to be about 1 to 3|. In the first fifty consecutive cases 
 of diabetes mellitus, however, which I carefully examined at the 
 Manchester Eoyal Infirmary (always using the direct method 
 of ophthalmoscopic examination), I did not meet with a single 
 instance of diabetic retinitis, and yet the majority of these 
 patients were suffering from the most severe form of diabetes 
 — often at an advanced stage. I have examined altogether 100 
 diabetic patients ophthalmoscopically, with considerable care, but 
 have only met with retinal changes in seven cases. Five cases 
 were specially sent to me by my friend Dr. Edward Eoberts, 
 whom they had consulted on account of ocular troubles. In 
 three of the seven cases the urine contained so much albumin, 
 that kidney disease could not be excluded as a cause of the 
 retinal changes : in two there was only a trace of albumin pre- 
 sent, but no other signs of nephritis. In two cases albumin 
 was absent. Those figures show that retinitis is a very rare 
 
2 32 SYMPTOMS AND PATHOLOGICAL CHANGES. 
 
 complication in diabetes, and it presents in this respect a 
 marked contrast to the retinitis of chronic Bright's disease. 
 
 Diabetic retinitis is met with almost invariably in middle- 
 aged and elderly patients — over the age of 45. As a rule, 
 retinal changes only occur when diabetes has existed for a long 
 period. 
 
 A good number of the patients I examined were under 
 the age of 45, and in many cases the disease was of an acute 
 form. Had there been a greater proportion of elderly persons 
 amongst these cases, probably retinal changes would have been 
 
 Fig. 12. — Retinitis hemorrhagica diabetica. 
 
 found more frequent. From the fact that retinitis is so rare 
 in the severe forms of diabetes, in persons under 45, I am 
 inclined to believe that there is some other factor necessary for 
 its production, in addition to the diabetic condition of the blood. 
 
 The changes in diabetic retinitis consist of haemorrhages and 
 white patches, both of which are often so small that they cannot 
 be detected unless the direct method of ophthalmoscopic examina- 
 tion be employed. Hirschberg ( 56 ) recognises three varieties of 
 diabetic retinal disease — (1) Eetinitis rnemorrhagica diabetica; 
 (2) retinitis centralis punctata; (3) combined form. Both 
 eyes are generally affected. 
 
 1. Retinitis hcemorrhagica diabetica. — In this form small dark 
 
DIABETIC RETINITIS. 233 
 
 red dots or punctiform patches of haemorrhage are scattered over 
 the retina ; occasionally they are striated. They are generally 
 situated behind the vessels ; they may occur alone, but more 
 frequently are associated with white patches, as in the third 
 variety — the combined form. 
 
 2. Retinitis centralis -punctata. — The retinal changes may 
 develop gradually or suddenly. They consist of small bright 
 white patches, which are situated in the deeper retinal layers. 
 They are found chiefly near the centre of the retina, between 
 the upper and lower temporal branches of the central artery. 
 
 Fig. 13. — Retinitis centralis punctata. 
 
 They are also found near the optic disc, and to the nasal side 
 thereof. The smaller patches consist of little white dots or 
 specks, which have been described as " curdy." If larger, they 
 may appear like white stripes ; or they may be clustered together 
 in the form of a semicircle or incomplete circle around the 
 macula ; occasionally there are two incomplete concentric rings 
 of white patches. But the white patches are never arranged, 
 like those in the retinitis of chronic Bright's disease, in a star- 
 shaped or fan-shaped form, radiating from the yellow spot. 
 There is no xjigmentation around the retinal patches. 
 
 3. In the combined form, both haemorrhages and white patches 
 are met with. 
 
234 SYMPTOMS AND PATHOLOGICAL CHANGES. 
 
 Blurring of the margin of the disc and optic neuritis are 
 absent in diabetic retinitis, whilst blurring of the margin 
 of the disc or slight neuritis is very frequent in albumin- 
 uric retinitis. It is stated that occasionally the retinal 
 changes are associated with primary optic atrophy ; whilst 
 optic atrophy is very rare in albuminuric retinitis, and, if 
 present, generally follows neuritis. Frequently, diabetic retinitis 
 is associated with opacities of the vitreous, the latter being due 
 to minute haemorrhages : the association is always suggestive of 
 diabetes. Occasionally diabetic retinitis has been followed by 
 
 Fig. 14. — Ketinitis, combined form. 
 
 hemorrhagic glaucoma. Sometimes, though very rarely, the 
 patient comes under treatment on account of the defect of 
 vision produced by the retinal changes. Often, however, well- 
 marked symptoms of diabetes have caused the patient to seek 
 medical advice, long before the retinal changes have developed. 
 
 The symptoms produced by diabetic retinitis are failure of 
 sight and difficulty in reading. The patient may complain of 
 a haze, or of a mist, or of dark specks before the eyes. The 
 retinal changes are progressive, and recovery never occurs. 
 
 Saunclby ( 60 ) believes diabetic retinitis to be of grave prog- 
 nostic significance, though he is not able to support this view by 
 statistics, and, owing to the rarity of the complication, statistics 
 
DIABETIC AND ALBUMINURIC RETINITIS. 235 
 
 would be difficult to obtain. When diabetic retinitis develops 
 in any given case of diabetes, the prognosis in that case is 
 naturally worse than if retinitis were absent. Nevertheless the 
 prognosis in diabetes is so much influenced by the form of the 
 disease, the age of the patient, and other conditions, that the 
 prognostic significance of such a rare complication as diabetic 
 retinitis is very difficult to estimate. Eetinal changes are 
 not found in young persons and in acute cases, but are 
 met with chiefly in elderly patients, and in these cases the 
 prognosis is best. Hence the prognosis will be probably worse 
 in a young patient without retinal changes than in an old 
 person with retinal changes. But of elderly persons the prog- 
 nosis would naturally be worse, other things being equal, when 
 retinal changes were present, than when they were absent. 
 
 The following are some of the points of difference between 
 diabetic and albuminuric retinitis: — 
 
 Diabetic Retinitis. 
 
 1. Patches of retinitis distributed 
 
 irregularly over the central 
 part of the retina ; not speci- 
 ally localised to the region of 
 the macula ; no tendency to 
 grouping of the patches in a 
 star-shaped or fan-shaped man- 
 ner around the macula ; some- 
 times patches arranged in an 
 incomplete circle around this 
 region. 
 
 2. Not associated with optic neuri- 
 
 tis. 
 
 '■'>. Optic disc not affected, or, if 
 affected, condition is one of 
 primary optic atrophy. Atro- 
 phy is exceedingly rare. 
 
 Albuminuric Retinitis. 
 
 Patches frequently most numerous 
 at the macula ; often localised 
 to this region ; often grouped 
 in a star-shaped or fan-shaped 
 cluster around macula. 
 
 Often associated with optic neur- 
 itis ; and when marked optic 
 neuritis is absent, the disc is 
 often cloudy and the margins 
 indistinct, even at an early 
 stage. 
 
 Optic atrophy rare ; occurs only 
 very late in the disease, and 
 follows optic neuritis ; is gene- 
 rally associated with typical 
 albuminuric retinitis. 
 
236 SYMPTOMS AND PATHOLOGICAL CHANGES. 
 
 Haemorrhages 
 pvmctiform. 
 
 as a rule are 
 
 5. Eetinal arteries and veins not 
 
 much changed in appearance. 
 
 6. jS"o diffuse retinitis. 
 
 7. Often associated with haemor- 
 
 rhages into the vitreous. 
 
 Haemorrhages generally striated or 
 irregular in shape, not puncti- 
 forni. 
 
 Arteries small, veins often dilated 
 and slightly tortuous. 
 
 Often diffuse retinitis present. 
 
 Not associated with vitreous opa- 
 cities. 
 
 Nettleship ( 57 ), in a case of diabetic retinitis, in which he 
 examined the eye microscopically, found oedema of the retina and 
 hyaline degeneration of the intima of the small arteries. He 
 also found that the capillaries were distended, and that minute 
 globular aneurysms were connected therewith. The vessels of 
 the brain, kidney, and spleen also presented similar minute 
 aneurysmal dilatations. 
 
 Occasionally, though exceedingly rarely, a cerebral tumour 
 has given rise to optic neuritis and glycosuria in the same 
 patient. 
 
 Sometimes, though very rarely, albuminuric retinitis occurs 
 in a diabetic patient who is also suffering from granular 
 kidney. Primary optic atrophy has been met with in a few 
 cases, but it is exceedingly rare, and probably only indirectly 
 connected with diabetes. In the two cases of diabetes associated 
 with acromegaly, recorded pp. 137-8, there was primary optic 
 atrophy. In both cases a sarcoma of the pituitary body was 
 found ; the growth had compressed the optic chiasma and optic 
 nerves, and so given rise to optic atrophy. 
 
 Diabetic amblyopia. — Occasionally a defect of vision, resem- 
 bling tobacco amblyopia, is met with in diabetes. In many of 
 these cases naturally the patients have been tobacco smokers, 
 and the visual defects have probably been clue to tobacco 
 amblyopia in diabetic patients. Also, cases have been recorded 
 in which anti-diabetic treatment was ineffectual until smoking 
 was discontinued. But the same kind of amblyopia has been 
 met with in diabetic patients who were not smokers. 
 
 In these cases the vision is impaired, but ophthalmoscopic 
 examination reveals no changes in the disc or retina. A peri- 
 metric tracing shows that peripheral vision is normal, but that 
 
DIABE TIC AMBL YOPIA. 2 3 7 
 
 there is a central scotoma for colours, or for both white and 
 colours. Diabetic amblyopia is thought to be clue to a retro- 
 bulbar neuritis. In the two following cases amblyopia of this 
 nature was present : — 
 
 Case 1. — A. B. ; the patient suffered from a severe form of diabetes. 
 Sight had been gradually failing for fourteen weeks. He had been 
 a smoker, but when I examined him had not smoked for three Aveeks, 
 nevertheless the vision had not improved but had become worse. 
 Distant vision was 4\ in each eye. The media, optic disc, and retina 
 of each eye were normal, but there was a central scotoma. Eed and 
 green were not recognised at the centre of the field, but were recognised 
 at the periphery. 
 
 Case 2. — J. B., set. 30, suffered from a severe form of diabetes. 
 Vision was greatly impaired, E. V. = /^-, L. V. = ^. Ophthalmoscopic 
 examination showed that there were no opacities of the cornea, lens, 
 or vitreous, and the retina and optic disc were normal in each eye. 
 There Avas no hemianopsia, the pupils reacted to light and accommoda- 
 tion, and there Avas no paralysis of the ocular muscles. Examination 
 shoAved a central scotoma for red and green. Blue and yellow were 
 seen in all parts of the field. The patient had smoked half an ounce 
 of tobacco daily. 
 
 Schmidt-Eimpler, in the examination of 140 diabetic patients, 
 obtained evidence of retrobulbar neuritis (which could not be 
 traced to alcohol or tobacco) in thirty-four. In one case of 
 diabetes with central colour scotoma, distinct atrophy of the 
 macular fibres of the optic nerve was found on microscopical 
 examination. 
 
 The following ocular affections have been also met with in 
 diabetes, but probably most if not all of these are mere 
 accidental complications : — Diplopia clue to paralysis and paresis 
 of ocular muscles, loss of the power of convergence, iritis, corneal 
 inflammation, conjunctival haemorrhages, furuncles and eczema 
 of the eyelids, hemiopia. 
 
 The Sexual Okgans and Functions. 
 
 The affections of the skin of the genital organs have been 
 described already on p. 223. 
 
 In addition to tliese cutaneous affections, the sexual functions 
 are often markedly altered in diabetes. In the male, diminution 
 
23 S SYMPTOMS AND PATHOLOGICAL CHANGES 
 
 or loss of the sexual power is not infrequent. The loss of sexual 
 power varies in degree from a defective power of erection to 
 total extinction of sexual desire. It may occur not only in 
 advanced cases, but sometimes it is one of the early symptoms, 
 and occasionally the patient first seeks medical advice on account 
 of impotence. 
 
 Frerichs ( 61 ) mentions the case of a diabetic gentleman, 
 whose illness dated from a violent fit of passion, brought about 
 by the discovery of the deception of his steward. The next day 
 sugar was found in the urine, and at once impotence was noticed. 
 Seegen records similar cases. Usually, however, the sexual 
 weakness is only noticed after the disease has been present a 
 long time. 
 
 In the milder forms of diabetes, when the symptoms improve 
 under treatment, often the sexual functions are restored. 
 Seegen ( 62 ) points out that diabetic patients frequently note that, 
 with an improvement of other symptoms under treatment, 
 erections again occur. 
 
 Loss of sexual power does not always occur, however ; even 
 when the disease is advanced, the sexual function may be 
 unimpaired. In some cases, both increased sexual power and 
 increased sexual desire have been noted. 
 
 In females the sexual desire is said to diminish greatly in 
 severe cases, whilst in elderly women who suffer from the milder 
 forms of diabetes it is said to be increased ( 63 ). 
 
 Amenorrhea is sometimes a symptom at an early stage ; in 
 other cases, menstruation occurs at regular or irregular intervals, 
 until a later period of the disease. 
 
 Conception may occur in diabetic women, and pregnancy 
 and parturition may be apparently unaffected by the disease, 
 but there is a great tendency to abortion. According to 
 Gaudard ( e4 ), 33 per cent, of pregnant diabetic women abort. 
 During pregnancy and the puerperal state the disease often 
 advances markedly. According to G-audard, 41 per cent, of 
 children born of diabetic women die. 
 
 (Matthews Duncan's ( 65 ) conclusions with reference to the 
 relation between diabetes and pregnancy are given on p. 114.) 
 
 The Ears. 
 Diseases of the ear are not common in diabetic patients. 
 
SEXUAL ORGANS— EARS— NERVO US SYSTEM. 239 
 
 Many authors ( 66 ), however, regard diabetes as an occasional 
 predisposing cause of furunculosis of the external auditory 
 meatus. Also a number of cases have been recorded of acute 
 inflammation of the middle ear in diabetic patients. The course 
 is said to differ from that of ordinary acute middle-ear catarrh, 
 by the rapid onset, the abundant suppuration, the tendency to 
 severe haemorrhage, and the early extension of the disease to the 
 mastoid process. 
 
 Pathological examination has shown that the petrous and 
 mastoid portions of the temporal bone are implicated early, and 
 that often extensive caries or necrosis occurs. Freriehs ( 67 ) 
 records a case in which otitis media and caries of the mastoid 
 process were followed by thrombosis of the lateral sinus. 
 
 Acute otitis media is not common in diabetes. It has 
 been present twice only amongst 140 patients who have come 
 under my observation in Manchester. In one case the affection 
 was bilateral and the suppuration profuse. 
 
 The Nervous System. 
 
 1. Mental Condition. 
 
 Often there is a marked change in the mental state in 
 diabetes. Mental dulness, apathy, or drowsiness, are frequently 
 noticed, especially in advanced cases. The patient is often 
 disinclined to perforin mental or bodily work ; he is frequently 
 sorrowful, despondent, melancholy, and depressed ; sometimes 
 hypochondriacal, and occasionally has suicidal tendencies. He 
 often speaks and looks like a man who recognises his condition 
 to be hopeless ; and it is somewhat rare amongst the severe 
 forms of the disease, such as are met with in hospital practice, 
 to find a cheerful, hopeful patient ; and in this respect the wasted 
 diabetic differs from the wasted phthisical patient. 
 
 It is not to be wondered at that the patient's character often 
 alters as the disease advances. He frequently becomes dis- 
 agreeable, bad tempered, and sulky. He may become deceitful, 
 untruthful, and cunning. A cheerful, good-natured, well-behaved, 
 and agreeable man may gradually become dull, sulky, disagreeable, 
 rude, and deceitful, as was the case with a diabetic patient whom 
 I had the opportunity of carefully observing for five years. 
 
 Weakness of mental power, of judgment, of memory and 
 
2 4 o SYMPTOMS AND PATHOLOGICAL CHANGES. 
 
 will, may occur, but, as a rule, the intellect remains clear up to 
 the last. 
 
 Mania has been recorded in a very few cases. Insomnia 
 is sometimes a troublesome symptom, but periodic attacks of 
 somnolence have also been described. 
 
 Epilepsy and diabetes. — Diabetes and epilepsy are very rarely 
 associated in the same subject, though epilepsy, diabetes, and 
 mental disease are sometimes met with amongst various members 
 of the family of a diabetic patient. In only one case of diabetes 
 which has come under my observation was the patient subject to 
 fits, but he had suffered from epilepsy for some years before the 
 onset of diabetes. 
 
 As regards the relationship of the two diseases, Ebstein ( 68 ) 
 points out that glycosuria may be (1) the cause of epilepsy, (2) 
 the result, (3) or both conditions may arise independently of 
 each other, but may be due to a common cause. 
 
 (1) Cases in which epilepsy is caused by diabetes may be divided 
 into two classes : («) Epilepsy may be due to cerebral lesions which 
 sometimes occur in the course of diabetes (as in a case recorded 
 by Lepine and Blanc, in which hemiplegia and epileptic attacks 
 occurred in a diabetic patient, and both were due to a micro- 
 scopical lesion of the cerebral cortex), (b) It is possible that 
 epilepsy may be occasionally due to the action of toxic products, 
 formed in the system of the diabetic patient. But the con- 
 nection is certainly very rare. Thus Dreschfeld ( 69 ) states that 
 convulsions were present once on]y out of sixteen cases of 
 diabetic coma, and in eighty cases collected from literature he 
 found convulsions recorded in six only. Jacoby ( 70 ) has reported 
 three cases of epileptic attacks in diabetic patients, which he 
 regards as the result of acetonemia ; but certainly convulsions in 
 diabetic coma are quite exceptional. In twenty-seven cases of 
 diabetic coma which have come under my observation in 
 Manchester, convulsions were absent in all. 
 
 (2) Epilepsy is thought by some authors to occasionally cause 
 glycosuria. Some of the older observers state that glycosuria 
 may be produced by an epileptic fit. Ebstein states, however, 
 that he has never found sugar in the urine as the result of a fit. 
 I may also add that many years ago I frequently examined the 
 urine of epileptic patients, when house physician to the National 
 Hospital for the Paralysed and Epileptic, Queen's Square, 
 London, but never met with any case in which sugar was present 
 
LOCALISED BRAIN LESIONS. 241 
 
 in quantity sufficient to give a reaction with Fehling's solution ; 
 neither have I met with any instance since then. Probably 
 glycosuria is only very rarely produced by an epileptic lit. 
 
 When diabetes is detected after an epileptic attack, it is 
 difficult or impossible to prove that the former affection is the 
 result of the latter. The coexistence of the two affections may 
 be merely accidental ; and Ebstein concludes that, from the few 
 facts recorded, it appears at present impossible to prove that 
 epilepsy produces diabetes, or even acts as a predisposing cause. 
 
 (3) With respect to the occurrence of diabetes and epilepsy 
 simultaneously, Ebstein points out that, since epilepsy frequently 
 alternates with diabetes and mental disorders in neuropathic 
 families, it would not be in the least surprising if an individual 
 affected with hereditary neuropathy developed simultaneously 
 epilepsy and diabetes. Also, in other cases, there may be 
 predisposing causes determining simultaneously diabetes and 
 epilepsy. 
 
 Diabetes and insanity. — In cases of insanity, sometimes sugar 
 is found in the urine ; but in these cases the condition is 
 generally one of slight glycosuria, or a mild form of diabetes. 
 The mental condition is primary, and the glycosuria a secondary 
 complication (see p. 92). 
 
 Landenheimer ( 71 ) has recently very carefully reviewed the 
 relationship of diabetes to general paralysis of the insane. He 
 concludes that there is no clear evidence that diabetes can 
 produce general paralysis, but that in a few rare cases symptoms 
 resembling those of general paralysis have been met with in 
 diabetic patients, and in such cases improvement has occurred 
 under anti-diabetic diet. 
 
 2. Localised Brain Lesions. 
 
 A few cases are on record of cerebral symptoms, such as 
 hemiplegia, monoplegia, aphasia, hemianopsia, localised epileptic 
 convulsions, etc., occurring in the course of diabetes, and yet on 
 post-mortem examination the brain has been found apparently 
 normal. These symptoms appear to be due to the action of 
 some toxic substance, and such cases resemble those of paralysis, 
 etc., occurring in uraemia, in which brain lesions are not dis- 
 covered post-mortem. 
 
 Occasionally diabetes is associated with symptoms of disease 
 16 
 
242 SYMPTOMS AND PATHOLOGICAL CHANGES. 
 
 of the brain during life, and post-mortem the usual changes, 
 such as softening, haemorrhage, etc., are found. In some cases 
 the association is merely accidental ; in other cases the brain 
 lesions play some part in the causation of the diabetes ; whilst 
 in a third group of cases they are connected with diabetes 
 indirectly. Thus it is possible that arterio -sclerosis may be the 
 cause of diabetes, by producing pancreatic changes, or some 
 lesions in the medulla ; and in such cases the arterio-sclerosis 
 may also lead to cerebral haemorrhage or thrombosis. Again, 
 it is possible that syphilis may be the cause of diabetes, by 
 producing changes in the medulla ; and in these cases other 
 syphilitic lesions of the nervous system may develop. The 
 condition of the brain in diabetic patients, and the significance 
 of the various cerebral lesions detected, has been already 
 discussed. 
 
 3. Diabetes and Lesions of the Spinal Corel. 
 
 G-lycosuria or diabetes is sometimes associated with symp- 
 toms indicating disease of the spinal cord ; in other cases, changes 
 in the spinal cord are discovered post-mortem, though there 
 have been no indications of spinal disease during life. In by 
 far the majority of cases, however, there are no indications of 
 spinal disease, either during life or on post-mortem examination. 
 
 (1) Bearing in mind the results of physiological experiments 
 on the spinal cord, recorded on p. 65, it is by no means 
 improbable that spinal cord disease is occasionally, though very 
 rarely, the cause of glycosuria or diabetes. 
 
 A number of cases of locomotor ataxia and disseminated 
 sclerosis, associated with glycosuria, and occasionally with true 
 diabetes, have been recorded ( 72 ~ 76 ). It is probable that in some 
 of these cases the glycosuria or diabetes has been produced by 
 sclerosis, extending to the medulla. Such cases are exceedingly 
 rare, and in most of them the nervous disease has been associated 
 simply with glycosuria, very seldom with true diabetes. 
 
 (2) In another group of cases of diabetes or glycosuria 
 associated with spinal disease, it is not clear whether the spinal 
 disease is the cause of the diabetes, whether both are due to 
 some common cause, or whether the association is merely 
 accidental. 
 
 Cases of diabetes associated with chronic anterior polio- 
 
DIABETES AND LESIONS OF SPINAL CORD. 243 
 
 myelitis — such are those recorded by Nonne ( 77 ) and Strumpell ( 78 ) 
 — would belong to this class. 
 
 (3) In a third group of cases, changes are found in the spinal 
 cord, which are probably the result of diabetes, and due to some 
 toxic substance in the blood. 
 
 The following are abstracts of notes in two such cases of 
 diabetes, in which I found pathological changes in the posterior 
 columns of the spinal cord :• — 
 
 [The first case was under the care of Mr. Milner, M.B., at 
 the Salford Union Hospital ; the second was under the care of 
 Dr. Steell, at the Manchester Eoyal Infirmary. To the kindness 
 of these gentlemen I am indebted for the opportunity of making 
 the clinical and pathological examinations.] 
 
 Case 1. — J. D., eet. 52 — Paresis and toasting of right deltoid, 
 pectorals, biceps and triceps muscles; on left side the same muscles 
 affected, but to a less extent. — Previous health good until nine months 
 ago, when he suffered from a severe cold. Whilst recovering he began 
 to he troubled with great thirst. ISTo family history of diabetes. No 
 history of injury or alcoholism. Six months ago the left arm became 
 weak. At that time no weakness of the right arm had been noticed. 
 Three months later the patient regained power in the left arm, but the 
 right arm then became weak, and has continued weaker than the left up 
 to the present time. 
 
 302% May 1892. — Present state. — Patient is much wasted. Urine 
 1040, no albumin ; contains a large amount of sugar, 20 to 26 grs. to the 
 ounce ; quantity of urine, 140 to 170 oz. daily. Arteries atheromatous, 
 radials calcareous. 1S0 affection of heart or lungs can be detected. The 
 patient cannot raise the hands to the mouth, but can flex at the elbows 
 and perform movements of fingers. Abduction and other movements at 
 shoulders very feeble. Slight foot-drop. Patient can dorsiflex feet, 
 though feebly. He complains of numbness of the legs, and frequent 
 cramp in the calves of the legs at night. Knee-jerks both absent. 
 
 29th June. — Patient is able to walk, but drags his legs and walks 
 with his feet widely apart. No ataxia. Slight dropping of the feet. 
 When in bed he can raise his legs in the extended position quite well. 
 The thigh muscles are in a fairly good state of nutrition. Slight 
 oedema of the feet. Both knee-jerks are now present, but very feeble. 
 Frequent cramps in the calf muscles at night. ]S"o other sensory symp- 
 toms in the legs. Wasting of biceps, triceps, deltoid, pectorals, and 
 scapular muscles of each side, especially the right. The forearms and 
 hands are only very slightly wasted, and present a well-marked contrast 
 to the shoulder and upper arm muscles. The posterior borders and 
 
244 SYMPTOMS AND PATHOLOGICAL CHANGES. 
 
 lower angles of the scapulae project, especially the left. This projection 
 is more marked when the arms are raised. When the patient shrugs his 
 shoulders, the upper part of each trapezius is felt to contract. Patient 
 has difficulty in placing right hand on left shoulder, but can place left 
 hand on right shoulder fairly well. The clavicular part of the right 
 pectoral is markedly wasted, and there is a deep depression below the 
 clavicle. When the hands are clasped together firmly between the 
 knees, the left pectoral muscles become much more prominent than the 
 right. Patient is only able to abduct at the shoulder to a slight extent ; 
 he cannot raise the arms into a horizontal position. Flexion (forwards) 
 at the shoulder is much more feeble on the right side than the left ; 
 the arm cannot be brought into horizontal position ; it can only be raised 
 to an angle of about 45°. The left arm can be flexed (forwards) at 
 the shoulder into the horizontal position. Patient can extend at the 
 shoulders (backwards) fairly well on both sides, and he is able to place 
 each hand on the sacrum. Extension of both elbows feeble ; flexion at 
 the right elbow very feeble ; flexion at the left elbow somewhat feeble, 
 but much better than on the right side, and much better than extension 
 at the same joint. Patient has great difficulty in raising the right 
 hand to his mouth. He is obliged to feed himself with the left hand. 
 He cannot raise a pot of water to his mouth with the right hand, but 
 is able to do so with the left. (Patient is a rightdianded man.) He 
 complains of a cold feeling in the fingers and hands, also of numbness 
 and tingling in the fingers. He has often to place them in Avarm 
 water to "get the feeling back." The fingers are swollen, especially 
 the terminal phalanges, which are bulbous, but the skin is pale. He is 
 able to feel slight tactile sensations, and to distinguish between the head 
 and point of a pin on each arm and hand. 
 
 20th July. — Knee-jerks now normal, no ankle clonus ; right plantar 
 reflex present, left absent. Abdominal epigastric and cremasteric 
 reflexes normal. Patient is able to feel and localise touch with a pin's- 
 head, and to distinguish between the head and point of a pin quite well 
 all over the arms and legs. Flexion at the right elbow exceedingly 
 feeble, and only the slightest prominence of the belly of the supinator 
 longus can be felt when the elbow is flexed, the forearm being midway 
 between pronation and supination. The left supinator longus is felt 
 readily, when same movement performed. Right biceps and supinator 
 longus greatly wasted. 
 
 There is no ataxy in walking, and no symptoms of tabes are present. 
 Death occurred from phthisis, about eleven months after the onset of 
 the disease. 
 
 Necropsy. — Body wasted. A considerable amount of serous fluid in 
 the abdominal cavity. Tuberculous disease of the lungs. Pancreas 
 firm ; weight, 3 oz. ; microscopical examination, after hardening, 
 
DIABETES AND LESIONS OE SPINAL CORD. 245 
 
 revealed cirrhosis. No changes of importance in the heart, liver, 
 kidneys, or spleen. Atheroma of the aorta. Attached to the left side 
 of the lumbar vertebrae (first and second) was a firm oval tumour the 
 size of an egg. Microscopical examination showed that it was a fibroma. 
 The brain (including medulla and floor of fourth ventricle) and the 
 spinal cord aj)peared normal to the naked eye. The arteries and their 
 smallest branches to the biceps muscles on each side were completely 
 calcified. 
 
 The spinal cord was hardened in Miiller's fluid. On section 
 of the hardened cord, in the lumbar region, the cut surface 
 appeared quite normal, but in the cervical and dorsal regions 
 there were marked naked-eye changes in the posterior columns. 
 Portions of these columns were much paler than the rest of the 
 wdiite matter of the cord. The colour of the affected parts 
 resembled closely the colour of tracts of ascending degeneration 
 — -in cases of transverse lesion of the cord — after hardening in 
 Miiller's fluid. In the lowest dorsal region this change (marked 
 paleness) was seen only at the posterior half of Goll's column 
 and the posterior third of Burdach's column on each side. A 
 little higher, the whole of both posterior columns, with the 
 exception of a narrow area in front, was much paler than the 
 rest of the white matter (see Fig. 15, c). Higher still, about mid- 
 dorsal region, these changes were seen in Goll's columns and the 
 median halves of Burdach's columns. At one spot in the upper 
 dorsal region the cord was slightly bruised in removal ; but 
 examination of the bruised part showed the absence of compound 
 granular cells or other evidence of myelitis. In the uppermost 
 portion of the dorsal region, just above the bruised part, the 
 changes (marked paleness) affected Goll's columns chiefly, but 
 extended slightly into Burdach's columns. In the lowest cervical 
 region the changes were very well marked, and affected Goll's 
 columns only. They gradually diminished at the anterior 
 parts of these columns, and in the highest cervical region only 
 the posterior two-thirds of Goll's columns were affected, but the 
 alteration in colour was quite distinct (see Fig. 15, a). With the 
 exception of the posterior columns, the cord appeared quite 
 normal to the naked eye. 
 
 On microscopical examination the changes found were very 
 slight, and this was surprising, considering the well-marked 
 naked-eye appearances. In sections stained according to 
 Weigert's method, the posterior median columns in the cervical 
 
246 SYMPTOMS AND PATHOLOGICAL CHANGES. 
 
 region were paler to the naked eye than the rest of the white 
 matter. In the dorsal region they were very slightly paler than 
 the other tracts of white matter, but the difference was less 
 marked than in the cervical region. 
 
 In sections stained with aniline blue-black, the posterior 
 median columns of the cervical region were slightly darker in 
 colour than other parts of the white matter ; but in the dorsal 
 region the change could only just be detected. 
 
 Under the microscope, sections stained both according to 
 Weigert's method and with aniline blue- black showed slight 
 excess of neuroglia connective tissue in the posterior median 
 columns of the cervical region. 
 
 In the posterior median columns (cervical region) many of 
 the nerve fibres were swollen, and the part of myelin which, in 
 normal nerve fibres, is stained black in Weigert's method (that 
 
 a be 
 
 Fig. 15. — Naked-eye appearances of the spinal cord on section, after hardening 
 in Midler's fluid ; normal white matter is shaded. The portions of the 
 posterior columns affected are pale, a, npper cervical region ; b, lower 
 cervical region ; c, dorsal region. 
 
 is, outer part) was reduced to a very narrow rim. In sections 
 stained with aniline blue-black, many of the axis cylinders of nerve 
 fibres in the posterior median columns were seen to be swollen. 
 
 In sections stained according to Marchi's method, scattered 
 degenerated fibres (stained black) were seen in the posterior 
 columns (median and external) of the dorsal region and -in the 
 posterior median columns of the cervical region ; they were 
 more numerous in the latter region. Sections of the lumbar 
 part of the cord appeared normal. At no part of the cord was 
 there any evidence of myelitis. The pyramidal tracts (direct 
 and crossed), the direct cerebellar tracts, and all other parts 
 of the white matter were normal. In the lower part of the 
 cervical region most of the nerve cells of the anterior horns of 
 grey matter were pigmented, and some — chiefly those of the 
 inner group — were perhaps slightly atrophied. 
 
 The finest nerve fibres entering the biceps muscles (branches 
 of the musculo-cutaneous) were teased in osmic acid, but 
 
DIABETES AND LESIONS OF SPINAL CORD. 247 
 
 microscopically they appeared quite normal ; also sections of 
 nerve fibres to the biceps muscles, hardened and stained accord- 
 ing to Marchi's method, appeared normal. 
 
 Case. 2. — E. B., set. 21, gave a history of severe fright, followed by 
 great mental distress and anxiety, twelve months previous to admission 
 to the hospital. Prior to that date the patient had been in good health, 
 but she has not been able to follow her employment (that of a dress- 
 maker) since. Thirst and diuresis noted first about four weeks after the 
 fright. The patient was considerably emaciated. The urine had a specific 
 gravity of 1038; it contained a large amount of sugar, but no albumin; 
 it gave a distinct (Gerhardt's) reaction with perchloride of iron, and a 
 well-marked reaction for acetone (Legal's test). There were signs of 
 tuberculous disease of the apex of the left lung, and the sputum con- 
 tained numerous tubercle bacilli. The knee-jerks were present, but 
 feeble. The pupils reacted to light and accommodation. A few days 
 after admission the knee-jerks disappeared ; they remained absent up to 
 death, which occurred from asthenia about seven months after admis- 
 sion. During the last few days of life there were slight pains in the legs, 
 chiefly about the ankles and soles of the feet, but there were no symptoms 
 of locomotor ataxy or other cord lesion during the illness. Both lungs 
 were extensively affected with tuberculous disease during the latter 
 three months of the patient's life. The quantity of urine varied from 
 about 90 to 116 oz. daily during the time the patient was in the 
 hospital, and the amount of sugar from 28 to 32 grs. to the ounce. 
 
 At the necropsy there was extensive tuberculous disease of both 
 lungs. The pancreas was small, and weighed 10 drms. 55 grs., but 
 the heart and other organs were proportionally atrophied. On micro- 
 scopical examination the pancreas appeared normal. 
 
 The spinal cord appeared normal at the necropsy, but after 
 hardening in Midler's fluid, marked naked-eye changes were 
 seen in the posterior columns, on transverse section. The 
 affected parts appeared much paler than the rest of the cut 
 surface, the colour resembling that of degenerated tracts above 
 a transverse lesion of the cord, when the specimen has been 
 hardened in Midler's fluid. In the lowest lumbar region no 
 changes could be detected, but in the middle lumbar region the 
 median thirds of both posterior columns (external and median) 
 were slightly paler than the rest of the cut surface of the cord. 
 In the upper lumbar region the change in colour was more 
 distinct and the pale area more extensive (see Fig. 16, c). In the 
 lower dorsal region there was a very pale streak in each postero- 
 
2 4 S SYMPTOMS AND PATHOLOGICAL CHANGES. 
 
 external column. This was especially well marked in the upper 
 dorsal region, and extended over the whole of each postero- 
 external column (see Fig. 16, b). In the lower cervical region 
 the postero-median columns presented this pale appearance. In 
 the upper cervical region only the posterior halves of these 
 columns were affected (see Fig. 16, a). The direct cerebellar 
 tracts and other parts of the white matter appeared normal in 
 all the regions of the cord. 
 
 Microscopical examination. — Sections were stained according 
 to the same methods as in Case 1. In the pale tracts in the 
 posterior columns numerous swollen nerve fibres were seen. 
 Only a very narrow rim of the myelin of these fibres was stained 
 black in the Weigert's specimens ; no other changes were 
 detected. Nerve fibres (branches of the anterior crural nerve) 
 
 a be 
 
 Fig. 16. — Naked-eye appearances of the spinal cord on section after hardening 
 in Midler's fluid. Changes in the posterior columns; normal white matter 
 shaded, parts of white matter affected are pale, a, cervical region ; b, upper 
 dorsal region ; c, upper lumbar region. 
 
 to the rectus femoris (left) appeared normal on microscopical 
 examination. 
 
 In both cases the changes were detected much better by 
 the naked eye than on microscopical examination. 
 
 I have examined the spinal cord in six other cases of diabetes, 
 but the appearances above described were not met with. 
 
 Slight changes in the posterior columns of the cord, in diabetes 
 mellitus, have been described by Sandmeyer ( 79 ), Minor ( 80 ), and 
 Leyclen ( 81 ). Since the above cases were published, Kalmus ( s2 ) 
 has recorded two additional cases in which similar changes were 
 found. In the first of his cases, on naked-eye examination of the 
 spinal cord, after hardening in Mliller's fluid, degeneration was 
 found in the posterior columns. This was most marked in the 
 cervical and lumbar enlargements. In the lower cervical and 
 dorsal regions the lesion was confined to Goll's columns ; above 
 and below it extended laterally into Burdach's column. The 
 sacral region was unaffected. In the lower dorsal region the 
 
DIABETES AND LESIONS OF SPINAL CORD. 249 
 
 right posterior column was distinctly more markedly affected. 
 In the second case, degeneration of the posterior columns was 
 also found. It was limited to Goll's columns in the upper 
 cervical region. In the lower cervical region it spread to 
 Burdach's column, and was most extensive in the lower cervical 
 and middle dorsal regions. Below the lumbar enlargement the 
 degeneration ceased. 
 
 Somewhat similar changes in the posterior columns, more 
 marked to the naked eye than on microscopical examination, 
 have been reported by Minnich ( 83 ), Bowman, and others in cases 
 of pernicious anaemia. Tooth ( Si ) has pointed out that on the 
 sixth day after section of the cord in animals, tracts of secondary 
 degeneration are well marked to the naked eye (by their light 
 colour) on hardening in bichromates, whilst the microscopical 
 changes at this date after section are very slight, and consist 
 chiefly in swelling of the axis cylinders and nerve fibres. Prob- 
 ably tracts of degeneration can be recognised by the naked eye, 
 after hardening in bichromates, before any microscopical changes 
 occur ; and Tooth considers that the naked-eye changes are due 
 to a preliminary chemical alteration in the nerve fibres. The con- 
 dition of the posterior columns in the cases of diabetes recorded 
 above, would appear to correspond closely to the early stage of 
 degeneration described by Tooth. The changes in the posterior 
 columns in the cases 1 and 2 were probably the result of the 
 toxic diabetic blood condition. 
 
 Kalmus also regards the spinal changes in his cases as the 
 result of the action of some toxic substance in the blood. 
 
 In case 2, p. 247, the knee-jerks were absent, and slight 
 naked-eye changes were found in the lumbar region of the cord. 
 As the peripheral nerves (fibres of the anterior crural to the 
 rectus femoris) were normal, it is possible that these changes 
 were the cause of the loss of knee-jerks. Eichhorst ( 85 ) has 
 found degenerative changes in the peripheral nerves in two 
 cases of diabetes in which the knee-jerks were lost ; but in other 
 cases the peripheral nerves have been normal. The absence of 
 knee-jerks in some cases of diabetes may be due to slight changes 
 in the lumbar region of the cord, such as those described above. 
 
 The spinal cord has always been normal in the cases which 
 I have examined, with the exception of the two above described, 
 and the case in which cerebro-spinal meningitis was present 
 Csee p. 117). 
 
250 SYMPTOMS AND PATHOLOGICAL CHANGES. 
 
 Considering the question of spinal cord complications from 
 the clinical side, we find their occurrence exceedingly rare. 
 Thus, in 140 cases of diabetes in Manchester, indications of 
 spinal cord disease were present during life once only. This 
 patient suffered from paresis of the legs with bladder symptoms, 
 probably due to a transverse myelitis. 
 
 4. Affections of the Peripheral Nerves. 
 
 Sometimes one of the earliest symptoms of which a diabetic 
 patient complains is cramp in the calves of the legs, especially 
 at night; and often there are pains in other parts of the legs 
 also. These symptoms are very common in the severe forms 
 of the disease, and attention has been drawn to them by the late 
 Sir B. W. Richardson, Unschuld, and others. In some cases the 
 legs are very tender, so that the patient cannot bear one leg to lie 
 on the top of the other in bed. Neuralgia and sciatica sometimes 
 occur, but they are very rare in the severe forms of the disease, 
 such as are met with in hospital practice and amongst young 
 patients. Both neuralgia and sciatica are often bilateral ; they 
 have been known to appear at the same time as the glycosuria, 
 and to increase along with the increase of the sugar in the urine, 
 and to diminish or disappear when the sugar excretion diminishes 
 or ceases. (Diabetes neuritis is described on pp. 259—66.) 
 
 Analgesia and tlicrmo-ancesthesia. — Yergely has recorded cases 
 of diabetes which have presented sensory disturbances like those 
 which are so frequently met with in syringomyelia, i.e. loss of 
 sensation to pain and temperature, whilst tactile sensation has 
 been unaffected. I have examined a good number of diabetics 
 for this symptom, but have not met with it. 
 
 5. The Knee-jerks in Diabetes Mellitus. 
 
 Many years ago Bouchard pointed out that the knee-jerks 
 may disappear in diabetes mellitus ( 86 ). In a paper published 
 in 1892, I recorded the condition of the knee-jerks in fifty cases 
 of diabetes mellitus ; in one-half of the cases both the knee- 
 jerks were absent ( 87 ). I pointed out that previous statistics had 
 varied considerably, and that much difference of opinion existed 
 respecting the prognostic value of this sign. Thus Bouchard 
 found the knee-jerks absent in 36 - 9 per cent, of his cases, 
 
KNEE-JERKS IN DIABETES MELIITUS. 251 
 
 Auerbach ( 88 ) in from 35 to 40 per cent., Maschka ( 8!) ) in 30'6 
 per cent., and Eichhorst ( 90 ) in 20'9 per cent. Soon afterwards, 
 Karl Grube, of Neuenahr, recorded ( 91 ) the results of the 
 examination of the knee-jerks in a series of cases of diabetes ( 92 ). 
 Grube pointed out that in his cases (131 in number) the knee- 
 jerks were absent in 7 '6 per cent. only. In a second paper ( £3 ) 
 Grube adds other cases, and states that in the combined series 
 the knee-jerks were absent in 13 "5 per cent. 
 
 The great difference between 7'G per cent. (Grube) and my 
 own results, 50 per cent., caused me to examine carefully the 
 condition of the knee-jerks in a second series of cases, in order 
 to find out, if possible, the reason for this remarkable discrepancy 
 in statistics. My results in the second series of fifty cases 
 were, however, practically the same ; the knee-jerks were both 
 absent in twenty-four, both were present in twenty-three, and 
 one was absent and one present in three cases. I may point 
 out that the cases have been most carefully examined. In taking 
 the reflexes, the skin over the patellar ligament has been well 
 exposed and the ligament struck directly, no clothing interven- 
 ing. If no knee-jerk has been obtained in the ordinary way, 
 then Jendrassik's and Buzzard's methods have always been 
 employed before the knee-jerks have been recorded as absent. 
 (The patient is seated on a chair with his feet w T ell in contact 
 with the floor, and the knees bent at an angle just a little more 
 than a right angle. He is made to look upwards, to link his 
 fingers together and to pull tightly. The observer places one 
 hand on the rectus femoris muscle, and strikes the patellar liga- 
 ment with a percussion hammer or stethoscope. If the slightest 
 knee-jerk be present, a contraction of the rectus muscle will be 
 felt 9i .) In many cases the knee-jerks appeared to be absent 
 when an examination was made in the ordinary manner, but by 
 the above method they were just obtained ; and these cases, 
 however feeble the reflexes, were classed amongst those in which 
 the knee-jerks were present. Most of the cases examined were 
 in-patients at the Manchester Tioyal Infirmary ; they were 
 examined repeatedly, often for weeks or months, and were thus 
 under the most favourable conditions for examination. Hence I 
 feel justified in believing that my figures represent fairly accur- 
 ately the condition of the knee-jerks amongst hospital patients 
 who suffer from diabetes mellitus in Manchester. The follow- 
 ing are my results in 100 cases: — 
 
252 SYMPTOMS AND PATHOLOGICAL CHANGES. 
 
 Cases. 
 Both knee-jerks lost in . .49 
 
 One present, one absent in . . 6 
 
 Both present in .... 45 
 
 100 
 
 One of the cases in which the knee-jerks were present at 
 first, came under observation for a second time, after an interval 
 of eighteen months, and both knee-jerks were then absent. If 
 this be added to the forty-nine cases, we have then the knee- 
 jerks absent in 50 per cent, of the cases. 
 
 As above stated, Grube found the knee-jerks absent in only 
 7 '6 per cent, of his first series of cases. What is the cause 
 of this remarkable difference in statistics ? I know Dr. Karl 
 Grube to be such a careful and reliable observer, that I can 
 place the greatest confidence in his observations, and, of course, 
 am convinced that his statistics are correct with respect to the 
 cases which he has examined. But I feel justified in believing 
 my own statistics to be also accurate. The difference in results 
 is interesting, and I believe the following to be the explanation. 
 The 100 cases, on which I have based my statistics, have been 
 mostly hospital patients in Manchester, who have been suffering 
 usually from a very severe form of the disease. In a large 
 proportion of cases the disease has been in a very advanced 
 condition, and often the observations have been continued 
 practically up to the last — in twenty-six out of the 100 
 cases up to the last few days, and in a large number of 
 cases up to the last few weeks of life. The majority of the 
 patients have been poor, hardworking people, and amongst 
 the 100 cases there were very few examples of the mild 
 form of the disease which is met with in stout or gouty 
 old people. As regards age, eighty-one of the 100 cases 
 were under the age of 50 years. On the other hand, Dr. 
 Karl Grube states ( 95 ) that the majority of his patients were 
 over the age of 50 years, and it is well known that the 
 severity of the disease is much less in patients over that 
 age. Then, again, Dr. Grube is engaged in private consulting 
 practice at Neuenahr, a small town in West Germany, much 
 frequented by diabetic patients on account of its alkaline 
 mineral waters, which are regarded as being of great value in 
 diabetes, especially in the diabetes of gouty persons. Also his 
 
KNEE-JERKS IN DIABETES MEIIITUS. 253 
 
 patients .have probably been in a better social position, and 
 have lived under more favourable conditions of life, than the 
 hospital patients on whom my observations were made. Prob- 
 ably a large proportion of his cases would not be in the very 
 last stage of the disease, otherwise they would not have been 
 allowed to travel to jSTeuenahr. These I believe to be the causes 
 of the difference of our results. I think it is a point of some 
 interest that the knee-jerks should be absent so much more 
 frequently amongst diabetic hospital patients, who usually suffer 
 from the disease in its severest form. Eichhorst ( 96 ) has pre- 
 viously drawn attention to a difference between his hospital 
 and private patients,- the knee-jerks being absent in 42 - 9 per 
 cent, of the former and 16 - 7 per cent, of the latter; but his 
 observations were based on thirty-six private cases and only 
 seven hospital patients. 
 
 In the earliest stage of diabetes, even in its most severe 
 form, I believe the knee-jerks would be found present as a rule ; 
 but in these cases they are very often lost later. I have fre- 
 quently observed that in the severe cases in which the knee- 
 jerks have been present when first examined, they have finally 
 disappeared if the cases have been under observation for a 
 long period. Sometimes I have found the knee-jerks pre- 
 sent in mild cases of long duration ; once at the end of five 
 years, once after the disease had been present seventeen years. 
 In the 100 cases which I have examined, the knee-jerks, as 
 above stated, were lost in 50 per cent., but in many cases I 
 was not able to continue my observations up to the last. In 
 twenty-six cases, however, in which I examined the knee-jerks 
 up to the last day or last few days of life, I found they were 
 both absent in nineteen ( = 73 per cent.), and present in seven; 
 but in four out of these seven cases they were exceedingly feeble. 
 In the majority of cases I have found the knee-jerks to be absent 
 during diabetic coma. Thus in twenty-one cases of diabetic 
 coma which have come under my observation in Manchester, the 
 knee-jerks were absent in eighteen, but in three cases they were 
 present (in one case, half an hour before death ; in another, three 
 hours before death ; and in the third case, at the commencement 
 of coma). Generally, however, the knee-jerks were absent for a 
 long period before the onset of coma. In one case they were pre- 
 sent the day before the commencement of coma, but disappeared 
 when the early symptoms of that complication were observed. 
 
254 SYMPTOMS AND PATHOLOGICAL CHANGES. 
 
 The following table (of 100 cases) shows that the knee-jerks 
 are lost in a much greater proportion of cases under the age of 
 30 years than in cases over 30. 
 
 Table shotting the Relation between the Presence or Absence of Knee-jerk 
 and the age at which Diabetes occurs. 1 
 
 Between the ages of 
 
 Knee-jerks 
 absent. 
 
 Knee-jerks 
 present. 
 
 Remarks. 
 
 10 and 20 years . . . 
 20 and 25 years . . 
 25 and 30 years . . . 
 
 13 
 9 
 
 8 
 
 3 
 2 
 
 4 
 
 Under the age of 25, knee-jerks 
 were lost in 81 per cent. 
 
 Under the age of 30, knee-jerks 
 were lost in 76 '9 per cent. 
 
 Totals .... 
 
 30 
 
 9 
 
 30 and 40 years . . . 
 40 and 50 years . . . 
 50 and 60 years . . . 
 Over 60 years . . . 
 
 8 
 5 
 9 
 3 
 
 10 
 
 14 
 
 4 
 
 8 
 
 Over the age of 30, knee-jerks 
 were absent in 40 9 per cent. 
 
 Totals .... 
 
 25 
 
 36 
 
 1 This table is based on another series of 100 cases. 
 
 It is well known that in young persons diabetes is of a 
 much more severe type than in middle or advanced life ; and 
 this may be the reason why the knee-jerks are absent in a 
 greater proportion of cases under the age of 30. 
 
 Variability of the reflex. — A knee-jerk which has disappeared 
 in a patient suffering from diabetes, may return again, and vary 
 very much in the course of time. Thus in a case, J. L., ret. 36, 
 in September 1890, the knee-jerks were absent. Urine, daily 
 quantity 210 to 280 oz. ; sp. gr. 1030; acid, no albumin; sugar 
 = 27 to 30 grs. to the ounce. 
 
 loth November. — Knee-jerks present ; urine, sp. gr. 1032 ; 26 
 grs. sugar to ounce ; trace of albumin, marked reaction with Fe 2 Cl . 
 
 30 th June 1891. — Knee-jerks absent. 
 
 17th September 1891. — Patient improved; right knee-jerk 
 present, but very feeble ; left knee-jerk absent. 
 
 In another case, J. D., ret. 52 (urine, 1030 ; 26 grs. of sugar 
 to ounce; amount of urine, 150 to 170 oz. daily), the knee- 
 jerks were absent (Jendrassik's method), May 1892. 
 
 29th June. — Knee-jerks present. 
 
 20th July. — Knee-jerks well marked. 
 
KNEE-JERKS IN DIABE TES MELLITUS. 2 5 5 
 
 Relation to general nutrition. — In diabetic patients who are 
 obese, as well as in those who are markedly wasted, the knee- 
 jerks are sometimes absent, sometimes present ; but they are 
 more frequently absent in the latter cases than in the former. 
 Thus in fifteen fairly stout diabetic patients, the knee-jerks were 
 both present in twelve, absent in one ; and in two cases one 
 reflex was present, the other absent. In fifteen patients who 
 were markedly wasted, the knee-jsrks were absent in ten. 
 
 Relation to the condition of the urine. — Though the knee-jerks 
 are more frequently lost in the severe than in the mild forms 
 of the disease, there is no close relation between the state of 
 the reflexes and the condition of the urine as regards specific 
 gravity, quantity of sugar, perchloride of iron reaction. Occa- 
 sionally they are absent when the urine only contains a moderate 
 amount of sugar ; sometimes they are still present when the 
 urine contains a very large percentage of sugar. 
 
 When the perchloride of iron reaction is intense, the knee- 
 jerks are sometimes present, sometimes absent. 
 
 Relation to pathological conditions. — The loss of the tendon 
 reflex cannot be regarded as evidence of the nervous origin of 
 the disease in any particular case, but must be looked upon 
 simply as a complication. Thus in the case of diabetes 
 presenting symptoms of gross lesion of the nervous system, 
 recorded on p. 132, the knee-jerks were present. 
 
 In another case of mild diabetes, in which a tumour of the 
 pituitary body was found post-mortem, the knee-jerks were 
 present. On the other hand, they were absent in a case 
 following a blow on the head, and in cases having a previous 
 history of great mental worry. 
 
 Subcutaneous injection of strychnia does not cause the knee- 
 jerks to return (Eosenstein). 
 
 As regards the cause of the loss of knee-jerks in diabetes, 
 the frequent association of pain, numbness, muscular tenderness, 
 and cramps points to an early peripheral neuritis ( 97 ) ; and a few 
 cases have been recorded in which well-marked typical motor 
 and sensory symptoms of peripheral neuritis have been present( 98 ); 
 also in a few cases examined pathologically, parenchymatous 
 neuritis of the peripheral nerves (branches of the anterior crural) 
 has been found ("). But frequently when the knee-jerks are 
 absent there are no other indications of slight peripheral neuritis, 
 and the peripheral nerves may be normal on pathological exainin- 
 
256 SYMPTOMS AND PATHOLOGICAL CHANGES. 
 
 ation. I have examined the small peripheral nerves in three 
 cases of diabetes mellitus in which the knee-jerks were absent 
 — in two cases the branches of the anterior crural in the thigh, 
 in the third case branches of the anterior tibial. In all three 
 cases the nerve fibres appeared normal. I have found slight 
 changes in the posterior columns of the spinal cord in two cases 
 of diabetes ( 10 °). Similar changes have also been recorded by 
 Sandmeyer ( 101 ), Leyden ( 102 ), and Minor ( 103 ) (see p. 246). In one 
 of these two cases the knee-jerks were absent, the peripheral 
 nerves examined (branches of the anterior crural) were normal, 
 but changes were found in the posterior columns extending down 
 to the upper lumbar region. In other cases I have found the 
 spinal cord normal, and many cases have also been recorded in 
 which the knee-jerks have been absent, and yet the cord has been 
 normal on pathological examination. Nonne has reported a case 
 of diabetes complicated with progressive muscular atrophy ( 104 ). 
 Pathological examination revealed chronic anterior poliomyelitis. 
 The nerve cells and fibres of the anterior horns of grey matter 
 were markedly degenerated at all parts of the cord ; at a 
 late stage of the disease the knee-jerks were absent. In one 
 ease in which the knee-jerks were absent during life, I found 
 both the smallest peripheral nerves (branches of anterior crural) 
 and the spinal cord normal on pathological examination, and 
 others have recorded similar results ( 105 ). 
 
 The changes just mentioned (in the peripheral nerves or spinal 
 cord) have been sufficient to explain the loss of knee-jerks in 
 certain cases. But in other cases, in which both cord and small 
 peripheral nerve present no recognisable changes histologically 
 (and probably these are more numerous than the former), some 
 other explanation is needed. Either changes (not yet described) 
 are present in the most minute nerve fibres and motor end- 
 plates, or the loss of the knee-jerk is due to some functional 
 alteration in the spinal cord, peripheral nerves, or muscles. 
 Marinesco ( 106 ) has recently recorded a case of diabetes in which 
 both knee-jerks were absent, but after an attack of hemiplegia 
 the knee-jerk returned on the paralysed side, the case being- 
 analogous to that of tabes reported by Dr. Hughlings Jackson 
 and Dr. Taylor, in which, after an attack of hemiplegia, the 
 knee-jerk returned on the paralysed side. 
 
 Prognostic value of absent knee-jerks. — Whilst some writers 
 do not attach any prognostic value to the loss of knee-jerks, 
 
KNEE-JERKS IN DIABETES MEIIITUS. 257 
 
 others consider that the prognosis is distinctly more unfavour- 
 able in cases in which the knee-jerks are absent. Bouchard, 
 Niviere, Marie, and Guinon ( 107 ) hold the latter view, and give 
 statistics in favour thereof. As already mentioned, the knee- 
 jerks occasionally persist up to the last. On the other hand, 
 they are sometimes absent in mild cases. Nevertheless the 
 knee-jerks are more frequently lost in severe than in mild cases, 
 and though other symptoms give much more important prognostic 
 indications, still the absence of knee-jerks must be regarded as 
 an unfavourable sign. 
 
 The following table shows the duration of life in ten patients 
 whose knee-jerks were absent, and in ten patients whose knee- 
 jerks were present, on first coming under observation : — 
 
 Duration of Life from Date on which the Condition of the Knee-jerks 
 was first ascertained. 
 
 When patient first came under observation. 
 Knee-jerks absent Knee-jerks present 
 in ten cases. in ten cases. 
 
 Died within two weeks .... 4 1 
 
 ,, between two weeks and six months . 2 3 
 
 „ „ six months and twelve months 2 1 
 
 . •. died within twelve months . . 8 5 
 
 Died between one and two years ... 1 3 
 
 Alive two years afterwards .... 1 2 
 
 . ". survived over twelve months . . 2 5 
 
 Conclusions. — (1) In Manchester, amongst hospital patients 
 suffering from diabetes mellitus, the knee-jerks are lost in from 
 49 to 50 per cent, of the cases. These patients mostly suffer 
 from a severe form of the disease; 81 per cent, are under the 
 age of 50 years; frequently there is great emaciation, and the 
 cases are often at an advanced stage. (2) In private practice, 
 amongst patients who live under more favourable conditions, and 
 in the milder forms of the disease occurring in gouty or well- 
 nourished people over the age of 50 years, I believe the pro- 
 portion of cases in which the knee-jerks are absent will be much 
 less (knee-jerks were absent in 1G'7 per cent, of private 
 patients, Eichhorst; 7'G per cent., Grube of Neuenahr — the 
 
 n 
 
25 3 SYMPTOMS AND PATHOLOGICAL CHANGES. 
 
 patients being mostly over 50 years). (3) The knee-jerks 
 when present at an early period are frequently lost or dimin- 
 ished later. Dining the last few days of life the knee-jerks are 
 lost in 73 per cent, of hospital diabetic patients in Manchester. 
 (4) They were lost in eighteen out of twenty-one cases of dia- 
 betic coma (86 per cent.). (5) Amongst diabetic hospital 
 patients the knee-jerks are more frequently lost under the age 
 of 30 years than over 30. (6) Since the course of diabetes 
 mellitus depends on so many circumstances, it is somewhat 
 difficult to estimate the exact prognostic value of one symptom, 
 which is occasionally absent even up to the last ; but the above 
 facts and considerations seem to show clearly that the loss of 
 knee-jerks is more frequently associated with unfavourable prog- 
 nostic indications. 
 
 6. Condition of other Reflexes in Diabetes. 
 
 In severe cases of diabetes the wrist-jerks are frequently 
 absent. When the knee-jerks are lost in diabetic patients, I 
 have usually found the wrist-jerks absent also ; though this is 
 not invariably the case. The following results are taken from 
 my notes in fifty cases of diabetes (mostly severe forms) : — 
 
 Cases. 
 
 Both wrist-jerks absent in . . . .30 
 
 Both wrist-jerks present in . . ... 19 
 
 One absent, one present ...... 1 
 
 50 
 
 [It is to be remembered, however, that the wrist-jerks are 
 sometimes absent in health. A few years ago I examined the 
 condition of the wrist-jerks in 110 persons, who were either 
 in good health or were suffering from some local surgical 
 affection not likely to have any effect on the reflexes. In this 
 series of cases I found the wrist-jerks present in practically 
 7 5 per cent, and absent in 2 5 per cent.] 
 
 Of the nineteen diabetic cases in which the wrist-jerks were 
 present, the knee-jerks were also present in eighteen ; in the other 
 case one knee-jerk was absent and one present. In the thirty 
 diabetic cases in which the wrist-jerks were lost, the knee-jerks 
 were also absent in twenty-five ; in the other five cases the 
 
PERIPHERAL NEURITIS IN DIABETES. 259 
 
 knee-jerks were present, but in three of these they were very 
 feeble. In the case in which one wrist-jerk was present and 
 the other absent, both knee-jerks were lost. 
 
 As already stated, the figures given above indicate the 
 condition in patients suffering mostly from the severe form of 
 diabetes. I have not had the opportunity of making observa- 
 tions on an equal number of cases of the mild form, but I 
 believe that in such a series the wrist-jerk would not be found 
 absent much more frequently than in healthy persons. 
 
 The superficial reflexes — plantar, abdominal, and epigastric — 
 are present as frequently as in health, whatever may be the con- 
 dition of the deep reflexes. In the severe forms of the disease 
 the superficial reflexes are usually well marked ; and when the 
 knee-jerks and wrist-jerks are absent, generally the superficial 
 reflexes are readily obtained. 
 
 7. Peripheral Neuritis in Diabetes. 
 
 The occasional occurrence of various forms of paresis and 
 paralysis, and the frequency of neuralgic pain in the limbs in 
 diabetic patients, have been mentioned by many of the older 
 medical writers, but it is only within recent years that any of 
 these symptoms have been attributed to peripheral neuritis. 
 
 Some of the nervous symptoms met with in diabetic patients 
 strongly resemble those of alcoholic and other forms of peripheral 
 neuritis. 
 
 v. Ziemssen ( 108 ) was the first, in 1885, to attribute the 
 neuralgia so often observed in diabetes, to peripheral neuritis. 
 Soon afterwards v. Hoesslin ( 109 ), and at a later date Eichhorst, 
 supported this view. 
 
 Pryce ( no ) has reported a case of diabetes with ataxic 
 symptoms, in which the peripheral nerves showed evidences of 
 neuritis. 
 
 Leyden ( m ), Althaus ( 112 ), Charcot ( 113 ), Buzzard ( m ), 
 Auche ( 115 ), Brims ( U6 ), have reported cases of peripheral neur- 
 itis in diabetes, or, to be more exact, cases presenting symptoms 
 similar to those of neuritis from alcohol or other causes. 
 
 More recently other cases have been recorded by Eich- 
 horst ( nr ), Pryce ( 11S ), and Eraser and Bruce ( Hl) ), in which 
 changes have been found in the peripheral nerves. 
 
 The proportion of cases of diabetes presenting marked 
 
2 6o SYMPTOMS AND PATHOLOGICAL CHANGES. 
 
 symptoms of peripheral neuritis is very small, but slight symptoms 
 are common. 
 
 The following are brief notes of two cases presenting slight 
 symptoms of peripheral neuritis, such as are frequently met 
 with : — 
 
 Case 1. — T. R, ret. 26. Urine 1040; no albumin; loaded with 
 sugar. When patient first came under my care, the knee-jerks were 
 present, and there were frequent cramps in the legs. Ten months later 
 he complained that he had to drag his legs in walking. There was 
 slight tenderness of the calf muscles, and frequent cramps ; he com- 
 plained of great pain of a gnawing character in the calf muscles and in 
 the front of the thighs ; also he had pain over the tibiae. Left knee- 
 jerk absent ; right, very feeble (Jendrassik's method). At a later date 
 he also complained of tingling and numbness in the legs. Ko real 
 paralysis, no anaesthesia. 
 
 Case 2. — J. R, set. 47. "When the patient first came under my 
 care, tbe knee-jerks were absent, but there were no other signs of 
 neuritis. Urine 1032, pale, acid, no albumin, contained a large amount 
 of sugar. At a later date both legs became numb ; there was also 
 tingling and pain in the legs, and the calf muscles were tender. The 
 legs became so tender that patient " could not bear one leg on the top of 
 the other in bed." He could feel and localise a touch with a pin's head, 
 and could distinguish between the point and the head of a pin. Move- 
 ments legs weak, but no real paralysis. 
 
 In both of these cases the symptoms resemble those met 
 with in the early stages of alcoholic peripheral neuritis. In 
 neither case was there a history of alcohol, nor of anything to 
 which the symptoms could be attributed except diabetes. 
 
 Many years ago, when house physician to Dr. Buzzard at the 
 National Hospital for the Paralysed and Epileptic, Queen Square, 
 London, I had the opportunity of observing a well-marked case of 
 diabetic neuritis. Buzzard ( m ) has since recorded the case, and 
 as it is one of the most marked on record, I acid a brief abstract, 
 taken from his report : — 
 
 J. K.j set. 55. Ten months before the patient came under Dr. 
 Buzzard's observation, he began to suffer from severe pain and tender 
 ness on the front of the right thigh, which soon extended down to the 
 foot. A month later the left leg was affected in the same way. Both 
 logs became weak, and at the end of three months he was unable to 
 walk. He then began to suffer from "pins and needles " in the soles of 
 
PERIPHERAL NEURITIS IN DIABETES. 261 
 
 the feet, and at a later date had numbness and tingling in the tips of 
 the fingers. Soon after the onset of illness, thirst, diuresis, and wasting 
 became prominent symptoms. 
 
 During the time the patient was under treatment, the chief 
 symptoms were — Loss of power in the legs ; inability to dorsiflex the 
 feet; dropped toes of left foot; wasting of muscles; reaction of de- 
 generation in anterior tibial muscles ; absence of knee-jerks and plantar 
 reflexes ; continuous pains in the legs and feet ; tenderness of the 
 legs and soles of the feet; numbness and tingling in the legs and 
 diminished cutaneous sensibility; pains in the fingers and diminution 
 of tactile sensation in the hands. The bladder was not affected. 
 On the outer side of the right foot, just below the malleolus, was a deep 
 ulcer surrounded by an area of congestion, and there was also a cicatrix 
 of an old ulcer just below the metatarso - phalangeal joint of the 
 great toe. 
 
 Urine, 57 to 84 oz. daily; sp. gr. 1042 to 1045. Sugar, 25 to 35 
 grs. per oz., according to diet. By rigid diet the sugar was reduced to 
 16 grs. to the ounce ; the ulcers healed, and patient improved. 
 
 Such cases as the one just recorded are exceedingly rare. 
 Amongst 140 diabetic patients who have come under my obser- 
 vation in Manchester, I have never seen any similar case pre- 
 senting such distinct motor symptoms of peripheral neuritis. 
 
 I have met with two cases of diabetes, in which there 
 has been neuralgic pain down the front and inner side of the 
 right thigh (along the course of the anterior crural nerve), 
 accompanied by paresis of the muscles which flex the thigh 
 on the abdomen. In one case both knee-jerks were present, 
 in the other the knee-jerk was absent on the affected side 
 but present on the opposite side. Probably both were cases 
 of localised neuritis in the anterior crural nerve. Brans has 
 recorded similar cases. 
 
 Laseque, Charcot, and others have described a number of 
 cases of monoplegia and paralysis of single groups of muscles in 
 diabetic patients. Lecorche ( 12 °) points out the rarity of these 
 cases, and states that they do not appear to be connected with 
 any profound lesion of the nervous system ; that they are 
 characterised by the paralysis being incomplete, localised, and 
 transitory ; and, generally, sensory symptoms, hyperesthesia, or 
 anaesthesia are present. It seems probable that these cases 
 are due to an affection of the peripheral nerves or peripheral 
 neuritis. 
 
 Ulcus on the feet, especially about the toes, " perforating 
 
262 SYMPTOMS AND PATHOLOGICAL CHANGES. 
 
 ulcers," are sometimes met with in diabetic patients. Probably 
 they are due to peripheral neuritis (see p. 226). 
 
 Re'sume of symptoms in diabetic neuritis. — Slight symptoms of 
 neuritis, such as pain in the legs, cramps, numbness, tingling, 
 tenderness, and absence of knee-jerks are not infrequent, but 
 marked paresis or paralysis is rare. The onset of symptoms is 
 gradual or subacute. 
 
 On an analysis of sixteen cases, recorded by various authors 
 during the last four years, the following description is based : — 
 
 Motor symptoms. — Paresis or paralysis, when present, most 
 frequently affects the legs — diabetic neuritic paraplegia (Buzzard). 
 In some cases — Buzzard ( m ) and Charcot ( n3 ) — the anterior 
 tibial muscles are chiefly affected. There is dropping of the 
 toes and feet, and the patient is unable to dorsiHex the feet. 
 In other cases, as in the two mentioned on p. 261, and in three 
 recorded by Bruns ( 116 ), the paralysis affects chiefly the muscles 
 supplied by the anterior crural and obturator nerves on the front 
 of the thighs, and the symptoms are sometimes much more 
 marked on one side than the other. In a case recorded by 
 Auche these muscles were affected on one side only, and the 
 patient was unable to go upstairs. In some cases — Leyden ( m ), 
 Salomonsen ( m ) — the motor symptoms are described simply as 
 weakness in the legs. The arms may be affected without the 
 presence of any paralysis in the legs. One arm only may be 
 paralysed, or one arm may be affected at first, but at a later 
 date both may be affected (Buzzard 114 ). In other cases the 
 muscles of the shoulder and upper arm may be affected on one 
 or both sides (see also case recorded, p. 243); or the paralysis 
 may be localised to a group of muscles, as the muscles supplied 
 by the ulnar nerve (Ziemssen), or to a single muscle, as the 
 deltoid (Althaus 112 ). 
 
 The affected muscles are generally wasted. In one case 
 (Charcot) they are said not to have been wasted. The knee- 
 jerks are absent when the legs are affected. Diminished 
 excitability to electricity, and partial or complete reaction of 
 degeneration, have been often observed in the affected muscles. 
 
 The sensory symptoms are often more marked than the 
 motor, and may be present when the latter are very slight 
 or absent. When motor symptoms are present, the sensory 
 are generally localised in the same region ; in the legs below 
 the knees, when the anterior tibials are chiefly affected ; in 
 
PERIPHERAL NEURITIS IN DIABETES. 263 
 
 the front of the thighs, when the muscles of this region are 
 chiefly affected (Brans) ; in the arm (Buzzard) ; or in the region 
 of the ulnar (Ziemssen) or the circumflex nerves (Althaus). In 
 some cases in which the legs are chiefly affected, sensory 
 symptoms are also present in the hands, and the grasp is weak. 
 
 The following are the chief sensory symptoms : — Pain, 
 neuralgic in character, often described as intense or violent, 
 shooting or tearing; hypersesthesia, tenderness, and pain on 
 pressure of muscles ; tingling, numbness ; sometimes a sensation 
 of coldness in the hands and feet. Diminished tactile sensation 
 is rare, but even anesthesia may occur ( m ) ; also diminished 
 sensation to pain is recorded. Neuralgia unassociated with motor 
 symptoms often occurs in diabetes, and is characterised, ac- 
 cording to Berger, by its spontaneous origin, by its frequent 
 localisation in the branches of the sciatic, sural, and plantar 
 nerves, by the violence and long duration of the attacks, by the 
 occurrence of vasomotor disturbances in the district of the 
 affected nerves, by the resistance to ordinary treatment for 
 neuralgia, and by the improvement under anti-diabetic treatment. 
 The cause of neuralgia, at least in a certain number of cases, 
 is probably neuritis. In some cases the nerves affected have 
 been noted to be tender on pressure. 
 
 The condition of bladder and rectum has frequently not been 
 stated, and therefore probably it has been normal. In other 
 cases both have been definitely stated to be normal. In one 
 case (Charcot) there was slight incontinence of urine. 
 
 Small ulcerations about the toes, perforating ulcers like those 
 of tabes ( no and m ), shining and glossy skin, ecchymoses, shedding 
 of the nails, and oedema, have been recorded. Herpes zoster has 
 also been described ( m ). 
 
 Ataxia has been occasionally noted ( 118 ). 
 
 Leyden recognises three forms of peripheral neuritis in 
 diabetes : — 
 
 1. The Ju/percesthdic or neuralgic variety, in which there is 
 more or less severe pain. This variety may occur in the form 
 of neuralgia (trigeminal neuralgia, sciatica, etc.), or as a multiple 
 neuritis in the feet, legs, and hands. Usually there is weak- 
 ness of the affected parts. Pain stands out as the prominent 
 .symptom. 
 
 2. The motor or 'paralytic form. In this form there is more 
 <>v loss marked paralysis of muscles of the legs or of other 
 
264 SYMPTOMS AND PATHOLOGICAL CHANGES. 
 
 parts. The knee-jerks are lost. Electrical changes are some- 
 times found. Often in this form there are neuritic pains. 
 
 3. The ataxic form, so-called pseudo-tabes. In this variety, 
 besides ataxia, there are sensory symptoms, numbness, formica- 
 tion in the feet. The muscular power for coarse movements is 
 maintained, or not essentially diminished. The tendon reflexes 
 are lost. But the pupils react to light and accommodation, and 
 in this respect, therefore, the cases differ from many cases of 
 true locomotor ataxia. Some neurologists, however, doubt the 
 occurrence of real ataxia. 
 
 As Charcot points out, however, occasionally, though very 
 rarely, a patient may present symptoms of true tabes dorsalis 
 along with diabetes. He recognises two groups of possible cases — 
 
 (1) In cases of tabes dorsalis, symptoms of diabetes may occur, 
 the latter being a complication, and due to the extension of the 
 lesion to the floor of the fourth ventricle. In these cases gastric 
 and laryngeal crises are often observed. Such cases, however, 
 are exceedingly rare. Marie and Guinon examined the urine in 
 iifty cases of tabes without finding glycosuria in a single case. 
 Gillas found glycosuria three times only in 100 cases of tabes. 
 
 (2) There is the possibility of the occurrence of tabes and diabetes 
 in the same patient as a coincidence. 
 
 Pathological evidence. — In most of the cases of peripheral 
 neuritis recorded, the diagnosis rests on clinical evidence — the 
 similarity of the symptoms to those of neuritis produced by 
 other causes ; the absence of any symptoms definitely referable 
 to the spinal cord or brain; and the absence of any of the 
 other recognised causes of neuritis, such as alcoholism, plumbism, 
 diphtheria, etc. 
 
 In a few cases the diagnosis has been confirmed by post- 
 mortem examination — in three cases reported by Pryce ( n0 and 118 ), 
 in two cases reported by Eichhorst ( 117 ), in three reported by 
 Auche ( 115 ), and in one reported by Fraser and Bruce ( m ). 
 
 In all these cases microscopical examination revealed the 
 existence of parenchymatous neuritis in the nerves of the affected 
 parts. In the case reported by Fraser and Bruce the affected 
 nerves presented changes similar to those described by Gombault, 
 as ndurite scgmentaire periaxile. 
 
 In one of Auche's cases there was paresis of the legs, 
 muscular cramps, and gangrene of the right foot. In another 
 case there were itching, tingling, and pricking pains in the feet, 
 
PERIPHERAL NEURITIS IN DIABETES. 265 
 
 legs, and hands, and the knee-jerks were absent. In a third 
 case there were violent cramps in the calves at night, slightly 
 diminished sensation to a pin-prick on the dorsal surface of the 
 forearms, subinguinal htemorrhages, shedding of the nails. 
 
 In one of Eichhorst's cases the knee-jerks were absent, but 
 there was no paralysis and no disturbance of the sensation. In 
 the second case the knee-jerks were absent ; the patient was 
 able to move the arms and legs, but there was marked muscular 
 weakness of the limbs. Examination of the anterior crural 
 nerves in both cases revealed parenchymatous neuritis. 
 
 In the first case, reported by Pryce, there were perforating 
 ulcers on both feet, diminished cutaneous sensibility of both 
 feet, and of the lower thirds of both legs. The knee-jerks were 
 absent. It is also stated that the patient had ataxic symptoms. 
 The examination of the peripheral nerves (by Mr. Bowlby) re- 
 vealed parenchymatous neuritis, but the ganglion cells of the 
 lumbar region of the cord were atrophied also. In the second 
 and third cases ataxic symptoms were present, cutaneous sensi- 
 bility was diminished, the patients suffered from pain in the legs, 
 and finally gangrene occurred. 
 
 In the case recorded by Eraser and Bruce the knee-jerks 
 were absent, and the patient suffered from pains in the legs and 
 tenderness of the calf muscles. 
 
 In most of the cases recorded the neuritis has occurred in 
 patients over the age of 50 (in twelve out of sixteen cases). 
 
 All writers on the subject agree that the neuritis does not 
 bear any relation to the amount of sugar in the urine. In many 
 of the cases recorded the amount of sugar has been small ; in 
 one reported by Brims it was not quite 1 per cent., and in 
 another case it was between 1 and 2 per cent. Further, the 
 symptoms continue if, by strict diet, the sugar can be made to 
 disappear from the urine. 
 
 The above facts seem to indicate that the neuritis is not due 
 directly to the presence of an excess of sugar in the blood. 
 Auche has made experiments on the lower animals, and exposed 
 the sciatic nerve to the action of different fluids containing sugar. 
 His experiments show the sugar has only a slight action on the 
 nerves, similar to that produced by water, and the changes in 
 the nerves in diabetes are probably due to some other cause — 
 the poverty of the tissues in water, the general disturbance of 
 nutrition, acetone, or some unknown chemical substance in the 
 
266 S YMPTOMS AND PA TH OL O GICAL CHANGES. 
 
 blood. Growers believes that the neuritis is due to some toxic 
 substance comparable to acetone, but not acetone, and he thinks 
 the fact that the reduction of the sugar excretion has but little 
 influence on the condition, suggests that the poison is not a 
 product of the decomposition of sugar, but a material formed in 
 place of sugar by some modification of the chemical processes 
 that lead to the increased sugar production. 
 
 Changes in the muscles have been described by Fraser and 
 Bruce ( 119 ) in the case in which the nerves showed the signs of 
 peripheral neuritis. 
 
 The muscles, on microscopical examination, presented a slight 
 increase in the distinctness of the longitudinal striation. This 
 was due to the presence of rows of fine fat granules between the 
 fibrillae of the muscle. The granules were all extremely minute, 
 and seemed to be developed from the cement substance rather 
 than the muscle fibres. In the portions of the muscles where 
 this defeneration was found, the transverse striation of the fibres 
 had disappeared. 
 
 REFERENCES. 
 
 1. Frerichs, F. T. 
 
 2. V. XoORDEX, C. 
 
 3. Gans .... 
 
 4. honigmann 
 
 5. eosenstein . . 
 
 6. CtRUBE, K. . . 
 
 7. HlRSCHFELD, F. 
 
 8. Seegen, J. . . 
 
 9. Frerichs, F. T. 
 
 10. Glenard, F. 
 
 11. v. Noorden, C. 
 12 
 
 13. Seegen. J. . . 
 
 14. „ . . 
 
 15. Frerichs, F. T. 
 
 16. Saundby, R. 
 
 " Ueber den Diabetes mellitus," Berlin, 1884, 
 
 S. 67. 
 "Die Zuckerkrankheit unci ihre Behand- 
 
 lung," Berlin, 1895, S. 87. 
 Verhandl. d. Cong.f. innere Med., Wiesbaden, 
 
 1890, S. 286. 
 Deutsche med. Wchnschr., Leipzig, 1890, 
 
 No. 43. 
 Berl. Jdin. Wchnschr., 1890, No. 13. 
 MiincJien. med. Wchnschr., 1895, S. 136. 
 Ztschr. f. Jdin. Med., Berlin, Bd. xix. 
 "Der Diabetes mellitus," Berlin, 1893, S. 
 
 214-218. 
 Loc. cit., S. 140. 
 Lyon med., tome lxiii. Nos. 16, 18, 20, 21, 
 
 23, 25. 
 "Die Zuckerkrankheit," Berlin, 1895, S. 99. 
 Ibid., Berlin, 1895, S. 101. 
 " Der Diabetes mellitus," Berlin, 1893, S. 171. 
 Ibid., p. 217. 
 
 '•Ueber den diabetes," Berlin, 1884, S. 140. 
 "Lectures on Renal and Urinary Diseases," 
 
 Bristol, 1896, p. 316. 
 
REFERENCES. 
 
 267 
 
 17. Jaccoud . . . 
 
 18. 
 
 Letden . . . . 
 
 19. 
 
 BOUCHARDAT 
 
 20. 
 
 Dreschfeld, J. 
 
 21. 
 
 Eoque, Devic, and 
 
 
 HuGOUNENQ 
 
 22. 
 
 BuSSENIUS . . . 
 
 23. 
 
 FlJRBRINGER . . 
 
 24. 
 
 Seegen, J. . . . 
 
 25. 
 
 Mayer, J. . . . 
 
 26. Israel, 0. . . . 
 
 27. Dreschfeld, J. 
 
 28. Frerichs, F. T. . 
 
 29. Schmitz, B. . . 
 
 30. „ . . 
 
 31. Seegen, J. . . . 
 
 32. Jacobson, G. . . 
 
 33. Armanni and Can- 
 
 TANI, A. 
 
 34. M'Kenzie, Stephen 
 
 35. Ebstein, W. . . 
 
 36. FlCHTNERj E. . . 
 
 37. Frerichs . . . 
 
 38. Frerichs and Ehr- 
 
 lich 
 
 39. Straus, J. . . . 
 
 40. 
 
 41. Schmitz, E. 
 
 42. Godlee, E. J. 
 
 43. Koexjo . . 
 
 44. Prtcb . . 
 
 " Nouveau clictionnaire de med. et chirurgie," 
 
 Paris, 1869, tome xi. p. 283. 
 ZtscJir. f. Irtin. Med., Berlin, Bd. iv. 
 " De la glycosurie ou diabete sucre," Paris, 
 
 1878. 
 Med. Chron., Manchester, 1884, vol. i. p. 5. 
 Rev. de vied,, Paris, 1892, p. 995. 
 
 Berl. Jdin. Wchnsclir., 6th April 1896. 
 Deutches Arch. f. Jdin. Med., Leipzig, 1895, 
 
 Bd. xvi. S. 499. 
 " Der Diabetes mellitus," Berlin, 1893, S. 182. 
 Ztschr. f. Jdin. Med., Berlin, Bd. xiv. S. 212 ; 
 
 Berl. Jdin. Wchnschr., 1890, S. 457. 
 Verhandl. d. Cong. f. innere Med, (Eleventh 
 
 Congress), Wiesbaden, 1892, S. 353 ; and 
 
 Vir chow's ArcJiiv. Bd. lxxxvi. S. 299. 
 Brit. Med, Journ., London, 31st August 
 
 1886. 
 " "Leber den Diabetes," Berlin, 1884, S. 81, 
 
 120. 
 "Prognose und Therapie der Zuckerkrankheit," 
 
 Bonn, 1892, S. 51. 
 Berl. Jdin. Wchnschr., 1891, Xo. 15. 
 " Der Diabetes mellitus," Berlin, 1893, S. 215. 
 Gaz. d, hdp., Paris, 25th August 1894. 
 " Der Diabetes mellitus," aus dem Italien- 
 
 ischen, von Dr. Hahn, Berlin, 1890, S. 315, 
 
 319. 
 Trans. PatJi. Soc. London, 1883, vol. xxxiv. 
 
 p. 355. 
 Deutsches ArcJi. f. Jdin. Med., Leipzig, Bd. 
 
 xxviii. S. 143. 
 VircJioio's ArcJiiv, Bd. cxiv. S. 400. 
 " Ueber den Diabetes," Berlin, 1884, S. 142. 
 Ztschr. f. Jdin. Med., Berlin, Bd. vi. S. 33. 
 
 Arch, de pJvysiol. norm, et patJx., Paris, 1885, 
 
 tome vi. p. 322. 
 Ibid., 1887, vol. x. p. 76. 
 Berl. Jdin. WcJmscJir., 1890, No. 23. 
 Med.-Chir. Trans., London, vol. lxxvi. 
 Berl Jdin. WcJmscJir., 22nd June 1896. 
 Lancet, London, 2nd July 1887. 
 
268 SYMPTOMS AND PATHOLOGICAL CHANGES. 
 
 45. Buzzard, T. . . 
 
 46. Auche . . . . 
 
 47. Eoberts, Sir Wm., 
 
 AND MaGUIRE, E. 
 
 48. Frerichs, P. T. . 
 
 49. Morris, M. . . . 
 
 50. Crocker, E. . . 
 
 51. „ . . 
 
 52. HlRSCHBERG . . 
 
 53. Seegen .... 
 
 54. Nettleshu' . 
 
 55. v. noorden . . 
 
 56. HlRSCHBERG . . 
 
 57. Mackenzie and 
 
 Nettleship 
 
 58. Nettleship . . . 
 
 59. Gowers . . . 
 
 60. Saundby . . 
 
 61. Frerichs, F. T. 
 
 62. Seegen, J. . . 
 
 63. v. Noordex 
 
 64. Gaudard . . 
 
 65. Duncan, Matthews 
 
 66. Kuhn .... 
 
 67. Frerichs . . 
 
 68. Ebstein, W. . 
 
 69. Dreschfeld, J. 
 
 70. Jacoby, G. W. 
 
 71. Landenheimer 
 
 72. Eedlich . . . 
 
 Brit. Med. Journ., London, 21st June 1890. 
 Journ. de vied, de Bordeaux, 11th January 
 
 1891. 
 " Eenal and Urinary Diseases," London, 1885, 
 
 p. 264. 
 "Ueber den Diabetes," Berlin, 1884, S. 69. 
 Brit Journ. Dermal, London, 1892, p. 250. 
 Ibid., 1892, p. 285. 
 "Diseases of the Skin," London, 1893, 
 
 p. 459. 
 Centralbl. f. praltf. Augenli., Leipzig, 1891, 
 
 S. 174; and Deutsche med. Wchnschr. t 
 
 Leipzig, 1891, No. 13. 
 "Der Diabetes mellitus," Berlin, 1893, S. 172. 
 Lancet, London, 1885, vol. i. p. 938. 
 "Die Zuckerkrankheit," Berlin, 1895, S. 108. 
 Deutsche med. Wchnschr., Leipzig, 1890, Nos. 
 
 51, 52. 
 Ojphth. Hosp. Rep., London, 1879, vol. ix. 
 
 Trans. Ophth. Soc. IT. Kingdom, London, 
 vol. vi. p. 331 ; vol. viii. p. 159. 
 
 "Medical Ophthalmoscopy," London, 1895. 
 
 "Lectures on Eenal and Urinary Diseases," 
 Bristol, 1896, p. 304. 
 
 "Ueber den Diabetes," Berlin, 1884, S. 72, 
 107. 
 
 "Der Diabetes mellitus, Berlin, 1893, S. 178. 
 
 "Die Zuckerkrankheit," Berlin, 1895, S. 107. 
 
 " Essai sur le diabete sucre dans l'etat puer- 
 peral," Paris, 1889 (quoted by Seegen, loc. 
 cit., S. 181). 
 
 Trans. Obst. Soc. London, 1882. 
 
 Ann. d. mal. de Voreille, du larynx, etc., 
 Paris, June 1890; Abstract, Med. Chron., 
 Manchester, 1890, vol. xii. p. 428. 
 
 " Ueber den Diabetes," Berlin, 1884, S. 149 
 
 Semaine med,., Paris, 6th May 1896. 
 
 Brit. Med. Journ., London, 21st August 
 1886. 
 
 New York Med. Journ., 1895, vol. lxii. 
 p. 585. 
 
 Arch. f. Psychiat., Berlin, Bd. xxix. 
 
 Wien. med. Wchnschr., 1892, Nos. 37-40. 
 
REFERENCES. 
 
 269 
 
 73. Oppenheim . . . 
 
 74. Maeie and Guinon 
 
 75. ElCHAKDIERE AND 
 
 Edwards 
 
 76. Weichselbaum 
 
 77. Nonne . 
 
 78. Strumpell 
 
 79. Sandmeyer 
 
 80. Minor . 
 
 81. Leyden . 
 
 82. Kalmus, E. 
 
 83. Minnich 
 
 84. Tooth . 
 
 85. Eichhorst (and 
 
 Nonne, quoted 
 by Eichhorst) 
 
 86. Bouchard . . . 
 
 87. Williamson, E. T. 
 
 88. auerbach 
 
 89. Maschka 
 
 90. Eichhorst 
 
 91. Grube . 
 
 92. „ 
 
 93. „ 
 
 94. Buzzard 
 
 95. Grube . 
 
 96. Eichhorst 
 
 97. Buzzard 
 98. 
 
 99. Prtce 
 
 Berl. Min. Wchnschr., 1885, S. 49. 
 Rev. de med., Paris, 1886-87. 
 Ibid., 1886, No. 3. 
 
 Wien. med. Wchnschr., 1880, No. 32. 
 Berl. Min. Wchnschr., 1896, No. 10. 
 Ibid., 1896, No. 46. 
 Deutsches Arch. f. Min. Med., Leipzig, 
 
 Bd. 1. 
 Arch, de neural., Paris, tome xvii. p. 391. 
 Wien. med. Wchnschr., 1893, No. 21 (Society 
 
 Report, S. 925). 
 Ztschr. f. Min. Med., Berlin, Bd. xxx. 
 Ibid., Berlin, 1893, Bd. xxii. 
 Brit. Med. Journ., London, 1889, vol. i. 
 
 p. 754. 
 Vir chow's Archiv, Bd. cxxvii. S. 1. 
 
 Pr ogres med., Paris, 1884, No. 41. 
 
 Med. Chron., Manchester, November 1892. 
 
 This article was afterwards included in the 
 
 work on " Peripheral Neuritis," by Dr. J. 
 
 Eoss and Dr. J. S. Bury. 
 Deutsches Arch, f. Min. Med., Leipzig, 1887, 
 
 Bd. xli. S. 484. 
 Wien, med. Presse, 1885, No. 3. 
 Virchoufs Archiv, Bd. cxxvii. S. 1. 
 Neurol. Centrcdbl., Leipzig, 1893, No. 22. 
 Lancet, London, 17th March 1894, p. 689. 
 Deutsche med. Wchnschr., Leipzig, 6th June 
 
 1895. 
 Lancet, London, 28th January 1888. 
 Deutsche med, Wchnschr., Leipzig, 6th June 
 
 1895. 
 Loc. cit,, S. 1. 
 
 Lancet, London, 28th January 1888. 
 Eor summary of cases see Med, Chron., Man- 
 chester, loc. cit, ; also Buzzard, Brit. Med. 
 
 Journ., London, 21st June 1890. 
 Lancet, London, 2nd July 1887; and Brain, 
 
 London, 1893, p. 416; Eichhorst, loc. cit.; 
 
 Eraser and Bruce, Edin. Med. Journ., 1896, 
 
 vol. xlii. p. 300. 
 
SYMPTOMS AND PATHOLOGICAL CHANGES. 
 
 Brit. Med. Journ., London, 24th February 
 
 1894. 
 Deutsches Arch. f. Jdin. Med., Leipzig, Ed. i. 
 Wien. med. Wchnschr., 1893, No. 21, S. 925. 
 
 (Society Eeports.) 
 Arch, de neurol., Paris, tome xvii. p. 391. 
 Berl. 1dm. Wchnschr., 9th March 1896. 
 Quoted by Eichhorst, loc. cit. 
 Compt. rend. Soc. de biol, Paris, 1895, 
 
 p. 691. 
 Rev. de med., Paris, 1886, p. 640. 
 Milnchen. med. Wchnschr., 1885, No. 44. 
 Ibid., 1886. 
 
 Lancet, London, 2nd July 1887. 
 " Die Entziindung der peripheren Nerven,"' 
 
 Berlin, 1888. 
 Lancet, London, 1890, vol. i. p. 455. 
 Arch, de neurol., Paris, 1890, tome xix. p. 
 
 305. 
 Brit. Med. Journ., London, 21st June 1890; 
 
 Lancet, London, 28th January 1888. 
 Arch, de med. exper. et d'anat. path., Paris,. 
 
 1890, No. 5. 
 Berl Uin. Wchnschr., 1890, No. 23. 
 Virchoiv's Archiv, Ed. cxxvii. S. 1. 
 Brain, London, 1893, vol. xvi. p. 416. 
 Edin. Med. Journ., 1896, vol. xlii. p. 300. 
 
 Arch, de neurol., Paris, 1885, tome x. p. 395 ; 
 
 1886, tome xi. 
 Arch, de neurol., Paris, tome iv. 
 Quoted in the Annual of the Universal Medical 
 
 Sciences, Sajous, Phila,, 1892. 
 "Diabetic Zona," Progres med., Paris, 26th 
 
 September 1891. 
 
 270 
 
 SYMPTOMS Ai 
 
 100. 
 
 Williamson, E. T. 
 
 101. 
 
 Sandmeyer . . 
 
 102. 
 
 Leyden .... 
 
 103. 
 
 Minor . . . . 
 
 104. 
 
 NONNE .... 
 
 105. 
 
 ,, .... 
 
 106. 
 
 Marinescu . . . 
 
 107. 
 
 Marie et Guinon 
 
 108. 
 
 V. ZlEMSSEX . . . 
 
 109. 
 
 V. HOSSLIN . . . 
 
 110. 
 
 Pryce .... 
 
 111. 
 
 Leyden .... 
 
 112. 
 
 Althaus . . . 
 
 113. 
 
 Charcot . . . 
 
 114. 
 
 Buzzard .... 
 
 115. 
 
 Auche . . . . 
 
 116. 
 
 Bruns . . . . 
 
 117. 
 
 Eichhorst . . . 
 
 118. 
 
 Pryce .... 
 
 119. 
 
 Eraser, T. E., and 
 
 
 Bruce, A. 
 
 120. 
 
 Lecorche . . . 
 
 121. 
 
 Bernard and Fere* 
 
 122. 
 
 Salomonsen . . 
 
 123. Y ERG ELY 
 
CHAPTER XL 
 
 DIABETIC COMA. 
 
 A diabetic patient may become comatose owing to various com- 
 plications. Thus cerebral haemorrhage or softening, meningitis, 
 and interstitial or parenchymatous nephritis, may give rise to 
 coma, when occurring as complications in a diabetic subject. 
 
 But apart from coma, produced by these and other complica- 
 tions, there is a special group of symptoms ending in coma, 
 which is a frequent termination of diabetes. These symptoms 
 are unaccompanied by any gross lesions of the organs, and are 
 apparently due to the toxic condition of the diabetic blood. The 
 name of diabetic coma has been given to this peculiar group of 
 symptoms, ending in unconsciousness. It was first described in 
 Germany by Kiissmaul in 1874 ; and in England, Sir W. B. Foster 
 drew attention to it in 1877. Diabetic coma has also been 
 carefully studied by Dreschfelcl, Frerichs, and numerous other 
 observers more recently. 
 
 Generally, the patient comes under treatment for other 
 symptoms of diabetes before the onset of coma ; but occasionally, 
 especially in acute cases amongst the poor, the patient is first 
 seen during the comatose stage. Thus a man was recently 
 brought to the Manchester Infirmary in a semi-comatose state. 
 The symptoms and the condition of the urine showed the case 
 to be one of diabetic coma, but the tubercular disease of the 
 lungs from which he suffered had led the medical attendant to 
 overlook the primary disease — diabetes. It is somewhat 
 surprising that diabetic coma has not given rise to medico-legal 
 difficulties more frequently ; and this no doubt would happen if 
 diabetes were a more common disease. It is possible that a few 
 of the unsatisfactory cases of coma which have not been cleared up 
 by post-mortem examination — cases which have been regarded 
 as serous or simple apoplexy, or of sunstroke, etc. — have really 
 been due to diabetic coma ; and since there are no pathological 
 
272 DIABETIC COMA. 
 
 appearances which are characteristic of diabetes or of diabetic 
 coma, a pathologist would not be able to arrive at a correct 
 diagnosis, in the absence of a chemical analysis of the urine or 
 blood. 
 
 Frequency of diabetic coma. — Coma is the most frequent 
 termination of diabetes. Thus, in twenty-eight of the last forty 
 fatal cases of diabetes which have come under my observation, 
 the cause of death was coma. 
 
 Relation to phthisis. — When advanced phthisis is present as 
 a complication of diabetes, death by coma does not usually 
 occur ; in most cases of diabetic coma, phthisis is absent, or the 
 lung changes are slight. But to these general rules exceptions 
 occasionally occur. Thus, in twenty-eight cases of diabetic coma, 
 which have come under my observation during the last six years, 
 there was extensive tuberculosis of the lungs (verified patho- 
 logically) in four. 
 
 Exciting Cause, etc. — Diabetic coma occurs both in the 
 severe and mild forms of diabetes. It may terminate life in 
 mild cases associated with obesity, but it is especially common in 
 the severe forms of the disease. It may occur at any age, 
 but it is very common in young persons and in persons under 
 middle life. 
 
 Dreschfeld ( x ) found, from an analysis of eighty cases of 
 diabetic coma, that 70 per cent, occurred between the ages of 
 20 and 40, and only 24 per cent, between 40 and 60, though 
 death from diabetes is more common at the latter-mentioned 
 period, as is shown by the tables on p. 97. Coma may terminate 
 the life of a patient at an early date after the first appearance 
 of the diabetic symptoms, or it may not occur for years after 
 the onset of the disease. Examples of the early onset of coma, 
 from two weeks to six months after the commencement of the 
 disease, will be found mentioned on p. 275. 
 
 Dreschfeld refers to a case communicated to him by Charcot, 
 in which the urine was examined and found free from sugar two 
 months before death from diabetic coma. 
 
 Sometimes several members of a family suffer from diabetes, 
 and die of diabetic coma. Quincke records the case of a girl 
 who died of diabetic coma at the age of 16, the patient's brother 
 died of diabetic coma at the age of 19, and an uncle at the 
 asje of 29. 
 
 It has loner been known that the exciting cause is sometimes 
 
EXCITING CAUSES. 273 
 
 a long railway journey — such as is necessitated in a visit to a 
 town physician by a patient living in a distant country place. 
 Frequently the journey to a continental spa, such as Carlsbad, 
 has been the exciting cause of diabetic coma. Excessive 
 muscular exertion is another exciting cause. Again, coma has 
 often developed in diabetic patients somewhat suddenly after 
 severe mental shock, anger, fright, disappointment, emotional 
 disturbances, mental strain, and worry. 
 
 A sudden change of diet — from a mixed diet to a rigid 
 — often appears to be the exciting cause of diabetic coma ; and 
 it is said that a sudden change from a rigid diet' to a mixed 
 diet has occasionally been immediately followed by coma. The 
 opinion is gradually gaining ground, that a rigid nitrogenous 
 favours the development of coma, especially in severe cases in 
 which the urine gives a dark brown coloration with perchloride 
 of iron. Ebstein ( 2 ), Hirschfeld ( 3 ), Schmitz ( 4 ), and Grube ( 5 ) have 
 especially drawn attention to this point, and the three latter 
 observers have advocated an increase of the carbohydrates in the 
 diet when coma appears to be threatening. Diabetic coma may 
 develop, however, when the patient's diet is not much restricted, 
 and even when there is no restriction whatsoever. 
 
 In a large number of cases, prolonged constipation has 
 appeared to play some part as an exciting cause, but it 
 is not invariably present. On the other hand, an attack of 
 severe diarrhoea has occasionally appeared to act as an exciting 
 cause. 
 
 Often coma has developed very soon after a patient has been 
 admitted into hospital. Of course this is sometimes due to the 
 fact that he has been brought to the hospital because he was 
 becoming rapidly worse ; but in other cases the change of diet 
 (sudden restriction) may have had some effect, though in most 
 cases at the Manchester Eoyal Infirmary coma has developed 
 when the patient's diet has been by no means rigid. Constipa- 
 tion produced by confinement to bed may also have aided in 
 producing the comatose termination soon after the patient's 
 admission into hospital. Exposure to cold, and alcoholic and 
 sexual excess, have appeared to act as exciting causes in certain 
 cases. 
 
 Diabetic coma appears to be occasionally excited by inter- 
 current affections or complications, as, for example, bronchitis, 
 pleurisy, croupous or catarrhal pneumonia, gastritis, hepatic colic, 
 
274 
 
 DIABETIC COMA. 
 
 strangulated hernia, influenza, tonsillitis, carbuncles, and pharyn- 
 geal, ischio-rectal, or alveolar abscesses. The administration of 
 anEesthetics, the performance of surgical operations, even cataract 
 operations, have all appeared to occasionally act as exciting 
 causes of coma. 
 
 Hirschfeld has drawn attention to rapid and marked loss of 
 weight, which often precedes the onset of coma (see p. 321), and 
 he believes that inanition is an important exciting cause. 
 
 Analysis of Exciting Causes in Twenty -seven Cases of 
 Diabetic Coma. 
 
 These cases were mostly hospital patients at the Manchester 
 Eoyal Infirmary. A few were cases in private practice. The 
 following table shows the age and sex of these twenty-seven cases 
 of diabetic coma : — 
 
 
 Age in Years. 
 
 No. of Cases 
 under 30. 
 
 Age in Years. 
 
 No of Cases 
 over 30. 
 
 13 Males . . . 
 
 14 Females . . 
 
 18, 19, 20, 26 
 
 11, 17, 19, 20, 31, 
 21, 23, 26, 26, 29 
 
 4 
 10 
 
 31, 34, 35, 35, 43, 
 
 43, 45, 46, 51 
 34, 36, 44, 66 
 
 9 
 4 
 
 27 cases. 
 
 Under the age of 30 
 
 14 
 
 Over the age of 30 
 
 13 
 
 Thus, under the age of 30, coma occurred more frequently 
 in females; over 30, in males. Half the cases of coma occurred 
 under the age of 30, though 63 per cent, of the diabetic 
 patients at the Manchester Infirmary are over that age. All 
 these twenty-seven patients suffered from a severe form of 
 diabetes with wasting, which was often very marked. 
 
 It is difficult to obtain reliable evidence as to the exact time 
 of the onset of the disease in many diabetic patients, but the 
 table of cases on the next page shows that the most prominent 
 symptoms, thirst and diuresis, were often noticed only a short 
 time before death occurred from diabetic coma. In two of the 
 twenty-seven cases, coma appeared to have been excited by 
 sudden mental shock and anxiety. Too great restriction of diet 
 could not be regarded as the exciting cause of coma in any of 
 the twenty-seven cases. In one case the patient was admitted 
 into the hospital in a state of coma, and diabetes had not 
 been previously recognised. 
 
ANAL YSIS OF CASES. 
 
 2 75 
 
 1. Boy set. 18, death from coma within 2 weeks v 
 
 2 # 
 
 Man 
 
 , 43, 
 
 3. 
 
 Woman 
 
 , 34, 
 
 4. 
 
 Boy 
 
 , 19, 
 
 5. 
 
 Girl 
 
 , 11, 
 
 6. 
 
 Woman 
 
 , 26, 
 
 7. 
 
 Girl 
 
 , 19, 
 
 8. 
 
 Man 
 
 20 
 
 9. 
 
 Man 
 
 , 45, 
 
 10. 
 
 Man 
 
 „ 31, 
 
 11. 
 
 Girl 
 
 „ 17, 
 
 12. 
 
 Man 
 
 „ 35, 
 
 13. 
 
 Man 
 
 , 51, 
 
 3 women (set. 19, 21, and 26) ,, 
 
 3 women (set. 21, 23, and 26) | 
 
 2 men (set. 34 and 43) j about 
 
 1 woman, set. 66, death from coma within 
 
 1 woman ,, 44, ,, ,, 
 
 1 month 
 
 1 
 
 jj 
 
 6 
 
 weeks 
 
 2 
 
 months 
 
 2 
 
 )» 
 
 3 
 
 j? 
 
 5 
 
 j; 
 
 6 
 
 ), 
 
 6 
 
 5) 
 
 6 
 
 5) 
 
 6 
 
 : J 
 
 6 
 
 !) 
 
 12 
 
 !) 
 
 12 
 
 ;> 
 
 18 
 
 >> 
 
 2 
 
 years 
 
 i> 
 
 II 
 
 Often coma developed very shortly after admission to the 
 hospital. No doubt this was, in part, explained by the fact that 
 the rapid advance of the disease had caused the patient to seek 
 admission. But in many cases it appeared probable that the 
 sudden change in the mode of life had played some part in 
 exciting coma. Confinement to bed and constipation produced 
 thereby, change in diet and the journey to the hospital, may have 
 had an injurious effect. Amongst twenty-one hospital cases 
 terminating by coma (cases in which there were no signs of 
 coma on admission), in twelve death occurred within one week ; 
 in two cases coma developed on the second day after admission 
 to the hospital ; in one case on the third day ; in six cases on 
 the fourth day ; in one case on the fifth day ; and in one case 
 on the sixth day ; in one case within seven days. 
 
 As regards the condition of the bowels, in seventeen of the 
 twenty-seven cases coma was preceded by obstinate constipation. 
 Sometimes the bowels had been fairly regular until the patient 
 was admitted into the hospital. Then for several clays after 
 admission there had been obstinate constipation, and coma had 
 developed. The onset of coma was not invariably preceded by 
 constipation, however. In several cases this symptom was absent, 
 and in one case there was diarrhoea, but in this case the 
 symptoms were those of diabetic collapse (see p. 281). In one 
 
276 DIABETIC COMA. 
 
 case the onset of coma followed the appearance of an ischio- 
 rectal abscess. 
 
 Symptomatology. — The symptoms often commence with 
 lassitude, epigastric pain, nausea, and occasional vomiting. In 
 other cases shortness of breath is one of the earliest symptoms. 
 Headache is also sometimes an early symptom. The patient 
 becomes anxious, restless, or excited ; he tosses about in bed, turns 
 from side to side, and sometimes the arms are thrown about 
 wildly at frequent intervals. Speech becomes thick and 
 incoherent ; the patient becomes drowsy, and the drowsiness 
 gradually develops into coma. 
 
 The pulse increases in frequency. Often the number of beats 
 rises to 120 or 130, at the last to 160 or more, per minute, or 
 the pulse, becomes so rapid and feeble that it cannot he counted 
 at the wrist. The tension is low, and the beats feeble, but gener- 
 ally regular. The heart's action is rapid and feeble, but cardiac 
 murmurs are not usually heard. Lepine regards rapidity of the 
 pulse as an important early indication of commencing coma. 
 
 Dyspnoea is a prominent feature in the majority of cases, and 
 it may be the first symptom, but it is not always present. The 
 dyspnoea is inspiratory at first ; later, both inspiration and expira- 
 tion are deep and prolonged, and the breathing has a peculiar 
 panting or sighing character. When the respirations are counted, 
 it is found very often that there is no increase, or only a slight 
 increase, in the number per minute. The respiratory difficulties 
 are indicated by deep inspiration and expiration rather than by 
 increased frequency of respirations. Not infrequently there is 
 a moan or groan with each expiration, even when the patient is 
 deeply comatose. The thorax expands well, and there is no 
 obstruction to the entrance of air into the chest. Physical 
 examination reveals nothing of importance, as a rule, in the 
 heart or lungs to account for the great dyspnoea, and there is no 
 distension of the veins of the neck. This peculiar dyspnoea has 
 been described by Kussmaul as " air hunger." 
 
 The tongue is dry and red. The bowels are generally con- 
 stipated, often markedly so, and frequently the onset of coma 
 has been preceded by a long period of constipation. Occasion- 
 ally diarrhoea has been present. 
 
 The face becomes pale and cold. In many cases there is 
 slight cyanosis. The nose, lips, and ears appear slightly 
 cyanotic ; they feel very cold to the touch, and the circulation 
 
5 YMPTOMA TOL OGY. 277 
 
 in these parts is feeble. The limbs and trunk are often cold 
 also, and usually the hands and feet are slightly cyanosed. 
 
 The temperature is generally subnormal; it may finally sink 
 to 95° or 94° F., but occasionally at the last it rises to a great 
 height, without any cause being detected clinically or at the 
 autopsy for this pyrexia (see p. 284). The expired air feels 
 colder than in the case of a healthy person. 
 
 Generally the breath has a peculiar oclonr. The smell is 
 noticed all around the patient, but it is particularly strong in 
 the breath. The urine also has the same smell. It is variously 
 described — most frequently, perhaps, as a smell resembling chloro- 
 form. It has been compared also to the smell of ether, sour 
 beer, apples, hay, or acetone. By allowing the patient to breathe 
 into water, Dreschfeld has been able to detect acetone in the 
 expired air, but no aceto-acetic acid was present. 
 
 This odour is sometimes detected for a considerable period 
 before the onset of coma ; but if not already developed, it is 
 noticed when coma commences. The smell varies in intensity. 
 Sometimes it is marked, at other times very slight. Some 
 people can detect it very readily. A few physicians state, 
 however, that they have never been able to recognise it. 
 Nevertheless the odour undoubtedly exists. Junior students 
 who see diabetic coma for the first time, and have never before 
 heard of the symptom, often notice the peculiar smell. A short 
 time ago I sent a junior student to the bedside of a patient who 
 was suffering from diabetic coma, and asked him to make a 
 diagnosis. The student had never heard nor read of diabetic 
 coma. He returned rapidly, and stated that, from the smell 
 about the patient, he believed he was still under the influence 
 of chloroform, and that probably some operation had just been 
 performed, for which chloroform had been given as the anaes- 
 thetic, though the patient was not in a surgical ward. 
 
 The acidity of the urine increases in diabetic coma ; and a 
 marked increase in the acidity of any diabetic urine is said to 
 point to the onset of coma. The quantity of urine often dimin- 
 ishes a little, and the sugar frequently diminishes also. It is said 
 that the sugar sometimes disappears, but certainly this is very 
 rare. I have never met with such a case. The colour of the 
 urine is often not so pale as before the onset of comatose 
 symptoms; and in one case I noticed a slight pinkish tinge, 
 though there was no indication of blood pigment, or any excess 
 
278 DIABETIC COMA. 
 
 of urates. The urine, in almost every case, has a peculiar 
 chloroform or sweet smell, like that of the breath. This smell 
 is often detected also in late stages of diabetes. The urine con- 
 tains a small amount of albumin. Often in the late stage of 
 diabetes, albumin is present ; but if it has not already appeared, 
 it does so just before, or just at, the onset of the comatose 
 symptoms. The urine when passed is generally (if not always) 
 slightly cloudy, and very minute floating particles can be recog- 
 nised by the naked eye. On standing, there is nearly always 
 a small deposit, greyish white or yellowish white in colour, 
 which sinks quite to the bottom of the urine glass. 
 
 On examination of the deposit, it is found to consist of 
 enormous numbers of casts. The field of the microscope is 
 usually crowded with them, and a few degenerated epithelial 
 cells are often present also. The casts are composed of a 
 hyaline material with fine granules. In some cases the granules 
 are few, whilst in others the casts appear to be composed of a 
 mass of closely-packed granules. Occasionally the granular 
 casts contain one or two degenerated epithelial cells. The 
 deposit of casts appears shortly before the onset of comatose 
 symptoms. 
 
 According to Maguire, albumin, in small quantity, is present 
 in every case of diabetic coma, and Kiilz states that casts are 
 always present also. Since becoming acquainted with their 
 observations, I have examined the urine in diabetic coma for 
 casts and albumin in every case which has come under my 
 observation (sixteen in number), and I have always found both 
 to be present. As regards casts, I have generally found enormous 
 numbers present, so that the field of the microscope has been 
 crowded with them. 
 
 Kiilz regarded the appearance of casts as a valuable pre- 
 monitory sign of the onset of coma. I have frequently examined 
 the urine of advanced cases of diabetes for casts or deposit, but 
 generally both have been absent up to the last few days of life. 
 Finally, however, a yellowish white deposit has appeared at the 
 bottom of the urine glass. Microscopical examination has 
 shown it to be composed of casts in enormous numbers, and 
 coma has followed. Hence I believe that when a small amount 
 of dense yellowish white deposit appears exactly at the bottom 
 of the urine glass, and when this deposit consists of casts in great 
 numbers, coma nearly always follows. To this general rule I 
 
SYMPTOMATOLOGY. 279 
 
 have found a few exceptions. Thus in two severe cases of 
 diabetes with wasting, a deposit of casts appeared in the urine, 
 though previously casts had been absent. The patient in each 
 case became drowsy ; the urine gave a deep reddish brown 
 coloration with perchloride of iron, and contained a small quan- 
 tity of albumin. The symptoms were very suggestive of the 
 onset of diabetic coma, but in each case the drowsiness passed 
 away, the casts disappeared from the urine in the course of a 
 day or two, and coma never developed. Death occurred in each 
 case some time afterwards from phthisis. Possibly these may 
 have been cases of abortive coma. They show, however, that 
 when casts appear, the cases do not invariably terminate in 
 fatal coma. 
 
 It has been already mentioned that the urine gives a deep 
 brownish red coloration with perchloride of iron (so-called 
 diacetic acid reaction) in many advanced cases of diabetes. This 
 reaction is nearly always obtained during coma, but exceptions 
 occasionally occur. 
 
 In coma the urine also gives the reaction for acetone. It 
 is often present in advanced cases of diabetes, but Hirschfeld has 
 shown that the amount of acetone increases greatly just before 
 death from coma. Acetonuria is no doubt increased by the 
 withdrawal of carbohydrate food in the severe forms of diabetes, 
 but it occurs also when carbohydrates are allowed, and may even 
 occur when the patient is taking ordinary diet. 
 
 Stadelmann ( 6 ) states that diabetic coma occurs only when 
 the urine contains fi-oxybutyric acid (see p. 185). He attaches 
 almost equal importance to the amount of ammonia in the urine, 
 whilst its estimation is far easier. Diabetic patients excreting 
 2, 4, 6, or more grms. of ammonia need constant watching, and 
 are in danger of coma. 
 
 If ammonia and oxybutyria acid cannot be determined, the 
 urine should at least be examined with perchloride of iron. 
 
 Stadelmann points out the danger of a strict diet in the 
 severe cases when a great amount of ammonia is excreted, or 
 when oxybutyria acid is, present, or when the urine gives a 
 reaction with perchloride of iron. 
 
 The alkalinity of the blood is said to be much decreased in 
 diabetic coma, but I have obtained decided alkaline reactions. 
 even shortly before death (see p. 186). 
 
 The blood presents no noteworthy change on microscopic 
 
280 DIABETIC COMA. 
 
 or spectroscopic examination. I have found that the blood in 
 coma, as in all cases of diabetes which I have examined, 
 decolorises a warm alkaline solution of methyl blue more 
 readily than normal blood (see p. 192). It has already been 
 mentioned that an excess of fat has been found in the blood in 
 a few cases (see p. 188). 
 
 Convulsions as a rule do not occur, and in this respect 
 diabetic coma differs markedly from ursemia. But there are 
 few rules in clinical medicine which are absolute, and ' though 
 their occurrence would be evidence against diabetic coma, 
 still a few cases have been recorded in which convulsions 
 have been noted (once in sixteen cases, Dreschfeld). Jacoby, 
 Finlayson, and others have recorded cases in which convulsions 
 occurred (see p. 240). 
 
 As already mentioned, the knee-jerks are frequently absent 
 in the severe form of diabetes. (In the hospital patients at the 
 Manchester Eoyal Infirmary they are absent in 50 per cent.) 
 If the knee-jerks have not already disappeared, they usually do 
 so when coma develops, but in a few cases they remain present 
 up to the last. In twenty-one cases of diabetic coma in Man- 
 chester, the knee-jerks were absent in eighteen, present in three ; 
 in one case the knee-jerks were present half an hour before death. 
 The condition of the pupils varies; they react to light, 
 though sluggishly. The eyelids are half closed as a rule. The 
 fundus oculi appears normal, with the exception of those rare 
 cases in which retinitis has been present. 
 
 As the coma increases it becomes more and more difficult 
 to rouse the patient, and finally he becomes totally unconscious ; 
 but in some cases the coma does not become complete, and 
 though the patient becomes more and more drowsy, and takes 
 no notice of his surroundings, he can be roused to take his 
 medicine up to the last. 
 
 Diabetic coma generally ends fatally within forty-eight 
 hours ; but abortive cases are common, according to Hirschfeld. 
 I have met with three cases in which early symptoms of coma 
 developed, but disappeared again. 
 
 The symptoms just recorded are those met with in the most 
 common variety of diabetic coma, Kussmaul's " air hunger." 
 But there are two other varieties in which the symptoms differ 
 from those described above ; these are (1) the alcoholic form of 
 diabetic coma, and (2) diabetic collapse. 
 
S YMPTOMA TOL OGY. 281 
 
 (1) Alcoholic form. — In this variety the symptoms resemble 
 those of alcoholic intoxication. In a case recorded by Ktilz 
 (quoted by Dreschfeld), a large amount of alcohol was found in 
 the urine ; the patient was excited ; he sang and swore ; the 
 speech became thick, and he staggered like a drunken man ; 
 the pulse was quick; the conjunctiva; congested; coma soon 
 developed, and terminated fatally. 
 
 An interesting case of this variety was met with by Dr. 
 Culling worth in Manchester, and is also quoted by Dr. Dreschfeld. 
 
 Dr. Cullingworth was consulted by a servant girl on account 
 of great thirst and weakness ; the urine had a specific gravity of 
 1035, and contained a considerable amount of sugar. He pre- 
 scribed opium and mix vomica, and advised her to go to her home 
 in Wales. On her journey home she left the train at one of the 
 small stations, and seemed so excited that she was believed to 
 be intoxicated. She was refused admission at one hotel, owing 
 to her apparently drunken condition. " At last she was taken 
 up by a policeman, who took her to a small inn, where she was 
 seen by Mr. Eeecl, a local medical man, late at night." She was 
 found to be in' a semiconscious state ; her face was very pale ; 
 the pupils were dilated and reacted slowly to light ; her pulse 
 was rapid, and the breathing quick. Next morning she was 
 quite unconscious, the breathing rapid and noisy, and the ex- 
 tremities cold. The urine contained sugar. The coma deepened, 
 and death occurred in the evening. 
 
 Frerichs has described similar cases. 
 
 In this form there is not the dyspnoea and feeling of 
 anxiety which characterises the first variety. The patient has 
 a feeling of intoxication, his gait becomes unsteady, he becomes 
 drowsy, and the drowsiness gradually passes into deep coma. 
 The breath has the characteristic smell, and the urine becomes 
 reddish brown on the addition of perchloride of iron. 
 
 (2) Diabetic collapse. — This condition was first mentioned by 
 Prout, and has since been carefully described by Dreschfeld, 
 Frerichs ( 7 ), and others. The following is Dreschfeld's account. 
 In these cases the patient suddenly begins to suffer from drowsi- 
 ness and great weakness. The extremities become cold; the 
 hands, feet, and face become livid; the pulse becomes quick, 
 small, and threadlike, soon reaching 120 to 130. The respira- 
 tions are only slightly quickened, they are shallow, and there is 
 not much dyspnoea. The temperature gradually falls, the skin 
 
232 DIABETIC COMA. 
 
 in some cases becomes covered with perspiration, the patient 
 becomes more drowsy and finally comatose, and death occurs from 
 collapse in ten to twenty hours. The urine contains sugar, but no 
 acetone or aceto-acetic acid, and the expired air has not the 
 peculiar chloroform-like smell. There is no delirium. 
 
 Dreschfeld points out that diabetic collapse occurs chiefly, 
 though not exclusively, in patients over the age of 40, and that 
 it attacks persons who have suffered from diabetes for some time, 
 and who are, as a rule, still stout and well nourished. In such 
 cases the disease has run a chronic course, and is often associated 
 with gout or nephritis. The exciting cause of diabetic collapse 
 is some extra physical exertion or some sudden shock, in a few 
 cases errors of diet or immoderate drinking. 
 
 Frerichs and Dreschfeld believe that the cause of this 
 variety of coma is cardiac failure, owing to degeneration and 
 atrophy of the cardiac muscle. 
 
 Analysis of Symptoms in Twenty-six Consecutive Cases 
 of Diabetic Coma. 
 
 The following is an analysis of my notes on the symptoms 
 of twenty-six cases of diabetic coma, most of which have come 
 under my observation at the Manchester Infirmary ; a few have 
 been cases in private practice. 
 
 In most of the twenty-six cases, epigastric pain, nausea, 
 and drowsiness were the first symptoms, and were generally 
 followed by, or associated with, dyspnoea. The first symptom, 
 in one case, was pain in the right lumbar and hypochondriac 
 regions. 
 
 In most of the twenty-six cases the patients were restless, 
 anxious, and somewhat excited at the onset ; in one case the 
 excitement was very great, and resembled that of early alcoholic 
 intoxication. 
 
 In one case dyspnoea was the first and only symptom for 
 several days. The patient had been under hospital treatment 
 in Belfast. He had left the hospital, and crossed by steamboat 
 to Liverpool, and had taken train to Manchester, where he 
 arrived the day after leaving Belfast. His breathing had become 
 difficult, somewhat suddenly, before reaching Manchester, and 
 he came to the accident room of the Royal Infirmary on account 
 of this symptom. I happened to see the patient in the acci- 
 
ANAL YSrS OF S YMPTOMS. 2 8 3 
 
 dent room, and as the breathing was very laboured (both inspira- 
 tion and expiration being very deep), and examination of the 
 chest revealed no cause for the dyspnoea, commencing diabetic 
 coma was diagnosed. There was then no coma or drowsiness 
 whatsoever, and dyspnoea was the only symptom of which 
 the patient complained. The urine was very acid ; the specific 
 gravity was 1036; sugar was present in large quantity, and 
 there was a small amount of albumin present also. The urine 
 gave a dark brownish red coloration with perchloride of iron. 
 The man was admitted as an in-patient under the care of Dr. 
 Steell, and the dyspnoea continued for five clays, when death 
 occurred, the other symptoms of diabetic coma having in the 
 meantime developed. The respirations varied during the time the 
 patient was in the hospital from 18 to 30 per minute ; but the 
 breathing was always laboured, inspiration and expiration being 
 both very deep. The patient was only comatose during the last 
 six hours of life. The interesting point about this case was the 
 fact that dyspnoea was the prominent symptom, and persisted for 
 five clays. For four days the patient was quite conscious and 
 intelligent, and actual coma only developed on the fifth day. 
 
 Dyspnoea was a very prominent symptom in nearly all of 
 the twenty-six cases ; in one, however, it was slight until the 
 last few hours of life. The number of respirations per minute 
 was often not much increased, frequently being only 18 to 20 ; 
 in other cases the respirations were 25 to 35 or more per 
 minute ; but whether the number of respirations was normal or 
 increased, the breathing was laboured, the inspirations and 
 expirations were very deep, and the breathing had the peculiar 
 sighing or panting character. When the patient was comatose, 
 generally a groaning noise accompanied expiration. Cheyne- 
 Stokes' respiration was never met with. 
 
 In nearly all of the twenty-six cases the pulse was very 
 rapid — frequently from 120 to 180 per minute. It often 
 became so rapid that it was impossible to count it. In several 
 cases, however, the rapidity of the pulse diminished shortly 
 before death to 90 or 100. In one case, in which I first saw 
 the patient about half an hour before death, the pulse was only 
 66 ; about an hour previously, I was informed that it had been 
 80 per minute. 
 
 The condition of the bowels (obstinate constipation in seven- 
 teen cases) lias already been referred to on p. 275. 
 
284 
 
 DIABETIC COMA. 
 
 In nearly all of the cases there was more or less cyanosis of 
 the limbs and face. 
 
 The temperature was usually subnormal ; it often sank 
 rapidly to 96° F., and remained low. In one case it sank to 
 
 F. a 
 102 
 
 101 
 
 100 
 
 99° 
 Nori 
 
 M 
 
 E 
 
 M 
 
 E 
 
 M 
 
 E 
 
 M 
 
 E 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 iJ I 
 
 
 
 
 
 
 
 98° 
 97° 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 I 
 
 V 
 
 -»- 
 
 
 -• 
 
 a 
 
 3 
 
 102 
 101° 
 
 100 _ 
 
 99° 
 
 Norn 
 
 M E 
 
 M 
 
 E 
 
 m;e 
 
 
 I 
 
 
 
 
 f 
 
 
 
 
 
 
 
 Jj 
 
 
 
 \ 
 
 98° 
 97° 
 
 
 
 1 
 
 .= i 
 
 / 
 
 
 
 1 
 
 Q 1 
 
 * 
 
 
 
 
 103 
 
 ro2° 
 
 101 
 
 100° 
 
 00 
 
 Nori 
 
 ME 
 
 M'E 
 
 M.E 
 
 M E 
 
 M' E 
 
 M E 
 
 M 
 
 L 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 Mil 
 
 
 
 
 
 
 
 
 98° 
 
 07 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 104 
 103° 
 
 102 
 
 lof 
 
 100° 
 
 00 
 
 Nori 
 
 m!e 
 
 M'E|MiE 
 
 M'E 
 
 MiE 
 
 M'E 
 
 M.E 
 
 
 
 
 ; 
 
 
 
 
 
 
 
 
 \\ 
 
 
 
 
 
 
 
 
 I 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 J 
 
 ,. 
 
 
 
 
 
 
 
 98° 
 97 
 
 \/ 
 
 /^ 
 
 
 
 
 
 
 
 / 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 Fig. 17. — Temperature charts in diabetic coma. 1, chart showing usual 
 fall of temperature ; 2, 3, and 4, charts showing unusual high tem- 
 peratures in three cases. 
 
 95°, but a few hours later it rose to 99° (just before death). In 
 two cases the temperature was practically normal, and in six out 
 of the twenty-six cases it was above normal. In one of these 
 six cases it rose suddenly and reached 1 04°, in another it reached 
 103°, in a third to 101 o, 6, in two other cases the temperature 
 
ANALYSIS OF SYMPTOMS. 285 
 
 was only slightly above normal. In one case it fell to 96°, 
 but next day rose to 101° F. Post-mortem examination made 
 in five of these cases failed to reveal any cause for the pyrexia. 
 No autopsy was obtained in the sixth case, but there was no 
 evidence of any complication which would account for the high 
 temperature (see Charts, .Fig. 17). 
 
 The peculiar so-called acetone smell of the breath was a 
 prominent feature in nearly all of the cases. 
 
 The quantity of urine and sugar excreted generally diminished 
 during coma ; but in none of the twenty-six cases did the sugar 
 disappear. 
 
 The urine was tested with perchloride of iron in twenty-four 
 cases ; a deep brown-red coloration was obtained in all before 
 the onset of coma. During the comatose condition a very 
 marked reaction was obtained in twenty. In two cases a deep 
 coloration was obtained with perchloride of iron just before the 
 onset of coma, but during the comatose condition the reaction 
 was much less. In two of the cases a deep coloration was 
 obtained just before the onset of coma, but the urine was not 
 tested with perchloride of iron during the comatose condition. 
 
 In sixteen cases, Legal's test for acetone was applied to the 
 urine during the comatose condition ; in fifteen the reaction was 
 obtained ; in one there was no reaction (in this case the per- 
 chloride of iron reaction was also slight). 
 
 The urine in nineteen of the cases was examined for albumin 
 during the coma ; it was present in all, but the quantity was 
 generally only very small. In some of the cases a trace of 
 albumin had been present for several days or weeks before the 
 onset of coma, and in these cases the amount of albumin in- 
 creased during the comatose stage. In other cases albumin had 
 been absent until the commencement of coma. 
 
 In sixteen cases the urine was examined for casts — they 
 were always present. 
 
 In fifteen cases enormous numbers of finely granular casts 
 were present during coma ; in one case only a few were seen. 
 In many of the cases the urine had been frequently examined 
 for casts, with negative results, just before the onset of coma. In 
 all of the fifteen cases there was a small, opaque, white or 
 yellowish white deposit exactly at the bottom of the urine glass. 
 Microscopically this deposit consisted of enormous numbers of 
 casts and a few epithelial cells. In most of the cases the urine 
 
236 
 
 DIABETIC COMA. 
 
 was carefully watched for a deposit for days before the onset of 
 coma, but it was always absent until just before the symptoms 
 of coma commenced ; it generally first appeared about twenty- 
 four hours before these symptoms were noted. In severe cases 
 of diabetes, I believe the appearance of this deposit (if micro- 
 scopical examination show it to be composed of casts) is a very 
 important sign of the approach of coma (see page 278). 
 
 In one case the patient lost 11 lb. in weight in the course 
 of the week preceding the coma. 
 
 Fig. 18. — Granular casts in the urine in diabetic coma. 
 
 Convulsions did not occur in any of the twenty-six cases. 
 
 The knee-jerks were examined in twenty-one cases during 
 coma ; they were absent in eighteen, present in three ; in one of 
 these three cases, however, they became exceedingly feeble about 
 two hours before death ; but in one they were obtained distinctly 
 half an hour before death. 
 
 The degree of coma varied. In some of the cases the patients 
 were quite unconscious, and could not be roused for many hours 
 before death. In other cases they were drowsy or semiconscious 
 
PATHOLOGY. 287 
 
 for many hours, but only became totally unconscious for a short 
 time before death. In some cases, though the patients passed 
 into a semiconscious state, they could be roused to answer 
 questions or to take medicine right up to the last hour of life. 
 In one case the patient, though she was semiconscious and did 
 not recognise her mother, was yet able to take her medicine 
 ten minutes before death. 
 
 The duration of definite symptoms of diabetic coma was 
 generally twenty-four to forty-eight hours ; but in one case, 
 five days elapsed between the onset of marked dyspnoea and 
 death. 
 
 Pathology of Diabetic Coma. — The pathological changes met 
 with in cases terminating in diabetic coma are not characteristic. 
 The most constant are the changes in the epithelium of the 
 kidney, described on p. 220, namely, the hyaline degeneration 
 of Cantani, fatty degeneration, necrosis of epithelium (Ebstein), 
 and the glycogenic degeneration of the epithelium of Henle's 
 loop. The changes found in other organs are varied — they are 
 generally clue to complications — and cannot be regarded as 
 important. 
 
 The literature relating to the pathology of diabetic coma is 
 very extensive, but it will only be possible to refer to the more 
 important views. 
 
 Fat emboli. — In some cases of diabetic coma there is an 
 excess of fat in the blood (see p. 188); and Sanders and 
 Hamilton ( 8 ) have detected fat emboli in the minute arteries 
 and capillaries of the lungs, to a less extent in the kidney and 
 other organs. They put forward the view that diabetic coma is 
 the result of fat embolism. But this view is now rejected by 
 almost all authorities. 
 
 As Dreschfeld ( 9 ) has pointed out, (1) in the majority of 
 cases of diabetic coma, fat emboli are not found ; (2) fat emboli 
 have been found in the kidneys and lungs, in the most varied 
 diseased conditions, and yet during life no symptoms thereof 
 have been observed ; (3) experiments on animals have shown 
 that, unless very large quantities of fat be injected, no bad 
 results follow, since the fat is gradually eliminated by the 
 kidneys. Hence we may certainly conclude that in most cases 
 of diabetic coma the symptoms are not due to fat embolism. 
 
 Cardiac failure. — In one variety of coma — the third form 
 escribed as diabetic collapse — the symptoms are in all prob- 
 
288 DIABETIC COMA. 
 
 ability the result of cardiac failure, as suggested by Frerichs, 
 Dreschfeld, and others. 
 
 Schmitz ( 10 ) points out that in advanced cases, just as the 
 muscles of the extremities and trunk become greatly enfeebled, 
 so also the cardiac muscles become weak and degenerated, and 
 death may occur from cardiac failure, with symptoms of collapse. 
 
 But in the variety of diabetic coma first described (the most 
 common form), some other explanation is probable. 
 
 The symptoms in this variety so strongly resemble those of 
 some forms of intoxication, that there can be little doubt that 
 in most cases the condition is the result of the action of some 
 poison or poisons developed in the organism of the diabetic 
 subject. 
 
 The facts that the blood and urine of diabetic patients 
 contain a large quantity of grape sugar, that during coma the 
 urine contains acetone, diacetic acid, and /3-oxybutyric acid, and 
 that the breath and the urine in diabetic coma have a peculiar 
 chloroform-like smell, all seem to point to a poisoning of the 
 organism. 
 
 Urcemia. — In diabetic coma, albumin in small or moderate 
 amount is probably always present in the urine ; also casts are 
 probably always present, even though but a few days previously 
 both were absent. Changes are frequently found in the renal 
 epithelium, as above described. Nevertheless the symptoms of 
 diabetic coma differ very much from those of uremia, and it is 
 not probable that there is the same toxic condition in the two 
 affections. 
 
 Acetoncemia. — In many severe cases of diabetes and in 
 diabetic coma the breath of the patient has a peculiar smell like 
 chloroform, and has been shown to contain acetone. The urine 
 has the same smell, and it also contains acetone (probably 
 constantly). Acetone has been obtained from the blood in 
 diabetic coma by some observers, but others have failed to 
 obtain any evidence thereof. Hence the symptoms of diabetic 
 coma have been attributed to the toxic effects of acetone in the 
 system. 
 
 But Frerichs and Dreschfeld have shown that acetone can be 
 taken in large doses by healthy men, without any symptoms of 
 importance being produced. The expired air contained acetone, 
 but only traces of this body were found in the urine. The 
 late Dr. Linderman of Manchester tried the effect of sub- 
 
PATHOLOGY. 289 
 
 cutaneous injection on himself. He found that the injection of 
 50 minims produced no bad effect. But acetone was present in 
 the urine for twelve hours after the last injection. 
 
 Dreschfeld found that large quantities of sugar together 
 with acetone could be taken without producing any toxic 
 symptoms. When given to diabetic patients also, no definite 
 symptoms were produced. 
 
 Subcutaneous injections of 5 to 10 minims of acetone, at 
 long intervals, produced no symptoms in rabbits, beyond slight 
 drowsiness ; but if five or six doses of 1 minims each were 
 given at short intervals, the rabbits became drowsy and comatose. 
 
 Acetone is found in the urine of diabetic patients, often for 
 weeks or months before any comatose symptoms occur ; also 
 acetone appears in the urine of healthy persons when carbo- 
 hydrates are cut off from the food entirely ; and it is found 
 in the urine in other diseases in which no symptoms of coma 
 occur. 
 
 Aceto-acetic acid — diacetic acid. — Much the same results have 
 been obtained in experiments with aceto-acetic acid, which has 
 been thought to be the toxic agent in diabetic coma, and which 
 is generally regarded as the substance giving the perchloride 
 of iron reaction in the urine. Many observers — Dreschfeld, 
 Frerichs, and others — have found that aceto-acetic acid can be 
 taken in large doses by healthy persons without any toxic 
 symptoms being produced ; also the acetone odour of the breath 
 has been detected, and the dark brown-red coloration with 
 perchloride of iron has been obtained in the urine, in many 
 diseased conditions besides diabetic coma (see p. 181). 
 
 The perchloride of iron reaction in the urine is often obtained 
 for weeks or months, in severe cases of diabetes, without any 
 comatose symptoms occurring. 
 
 Aceto-acetic ether given to healthy persons and to diabetic 
 patients in large doses has also produced no results. 
 
 Acid intoxication. — Walter ( n ) has shown that large doses of 
 dilute acids (hydrochloric and phosphoric) given to rabbits pro- 
 duce dyspnoea — the separate respiration being deep and laboured. 
 At this stage there is no evidence of cardiac failure, and the 
 blood pressure is not diminished. The dyspnoea is therefore not 
 due to cardiac changes. Walter believes that these acids give 
 rise first to a stimulation and then to a paralysis of the 
 respiratory centre, followed by fatal collapse. He attributes 
 <9 
 
2 9 o DIABETIC COMA. 
 
 the symptoms to a diminished alkalinity of the blood, since 
 they subside when sodium carbonate is injected into the blood. 
 
 These symptoms, produced in animals by acids, have some 
 resemblance to those of diabetic coma in man, and Stadel- 
 mann ( 12 ) believes that the latter condition is really due to an 
 acid intoxication. He attributes diabetic coma to a diminished 
 alkalinity of the blood, owing to the presence of an organic 
 acid. Minkowski ( 13 ) and others have supported this view. The 
 symptoms have been referred to the presence of crotonic acid or 
 to /3-oxybutyric acid in the blood. 
 
 Vaughan Harley ( u ) has found that by injecting grape sugar 
 into the jugular vein of dogs, symptoms resembling diabetic coma 
 were produced, providing the ureters were ligatured so as to 
 prevent any of the sugar or its products being eliminated by the 
 kidneys. The result was : (1) a stage of nervous irritation, 
 followed by (2) a comatose stage, varying from a few hours' 
 drowsiness, up to deep coma, ending in death. Harley believes 
 the coma to be, in part, the result of poisoning by substances 
 produced by the decomposition of sugar. He thinks, however, 
 that coma is principally due to a diminution of the alkalinity 
 of the blood, owing to the production of acids by the splitting 
 up of the sugar molecule. 
 
 Intestinal auto-intoxication. — Schmitz regards diabetic coma 
 as the result of an auto-intoxication, due to the formation of 
 toxic substances, by the putrefactive decomposition of albumins in 
 the intestine. A great amount of nitrogenous food is taken, and, 
 owing to the marked constipation, it remains long in the intestinal 
 tract ; decomposition occurs, toxic substances are formed, and, 
 being absorbed, they give rise to diabetic coma. Schmitz states 
 that by the use of purgatives (as castor-oil) he has caused the 
 symptoms of diabetic coma to disappear, when the treatment has 
 been commenced early. 
 
 It is very probable that constipation predisposes to coma (as 
 pointed out on p. 273); but, on the other hand, constipation 
 occurs to a marked degree in many affections without giving rise 
 to any symptoms of coma ; further, the bowels are often con- 
 stipated in diabetes for long periods without any symptoms of 
 coma developing ; and, finally, in some cases of diabetic coma 
 (though rarely) diarrhoea is present. Hence it appears more 
 probable that constipation is merely an exciting cause of diabetic 
 coma, and not the essential factor. 
 
PATHOLOGY. 291 
 
 Eogue, Devic, and Hugounenq ( 15 ) have made some interesting 
 observations on the toxic condition of the blood in a case of 
 diabetic coma. They found that the alkalinity of the blood 
 (estimated by sulphuric acid) was about half that of normal 
 blood. To 30 c.c. of blood serum from a diabetic patient, a 
 dilute solution of sodium bicarbonate was added, until the 
 alkalinity of the mixture was the same as that of normal blood 
 serum. In this way these observers were able to obtain two 
 samples of serum from the same diabetic blood, of similar com- 
 position, except that one was twice as alkaline as the other. In 
 order to determine whether the toxic action of the blood was the 
 result of its low alkalinity, experiments were made on rabbits. 
 It was found that 8 c.c. of serum of blood from the diabetic 
 patient were sufficient to cause death ; but, on using the diabetic 
 serum, the alkalinity of which had been raised to the normal 
 standard by adding sodium bicarbonate solution, it was found 
 that 23 c.c. were required to give a fatal result. Hence these 
 observers conclude that the toxic properties of diabetic blood are 
 greatly reduced by the addition of alkalies. 
 
 Klemperer ( 16 ) and v. Noorden ( 17 ) are of opinion that in 
 diabetic coma some toxic substances are produced in the organism, 
 which, acting on the brain, produce coma, and which also cause 
 destruction of protoplasm of the organism, and as a result thereof 
 oxybutyria acid is produced. Coma on the one hand, and the 
 formation of oxybutyria acid and the diminution of the alkalinity 
 of the blood on the other, are the result of the action of some 
 toxic substance formed in the diabetic organism. 
 
 The presence in the urine of acetone (in quantity), and of the 
 substance giving rise to the brownish red coloration with per- 
 chloride of iron, are indications of a severe and serious nutritional 
 change in the organism of the diabetic patient; and the con- 
 tinued excretion of /3-oxybutyric acid in considerable quantity in 
 the urine renders the onset of coma very probable. 
 
 In spite of the numerous researches and observations with 
 regard to the pathology of diabetic coma, the exact cause still 
 remains to be determined. Evidently some toxic substance or 
 substances are formed in the system, which by their action pro- 
 duce the symptoms that have been described above. 
 
 Whatever the exact poison may lie, as a rule, acetone, 
 aceto-aeetic acid, and /3-oxybutyric acid are found in the urine 
 coincident with this intoxication. It has not been proved, 
 
292 DIABETIC COMA. 
 
 however, that the presence of these substances in the blood cause 
 diabetic coma ; and in fact it appears more probable that they 
 are produced in the system by the action of some unknown 
 poison. 
 
 As already pointed out, in diabetic coma a small amount of 
 albumin and numerous casts are generally present in the urine — 
 probably they are always present — and very frequently changes 
 are found in the epithelium of the uriniferous tubules. Hence 
 it seems probable that failure of the renal function occurs, that 
 some poison ceases to be eliminated, and as a result the 
 symptoms of diabetic coma follow. Now, Yaughan Harley has 
 shown (see p. 290) that similar symptoms can be produced 
 in dogs by injecting grape sugar into the jugular vein, and 
 preventing elimination by ligaturing the ureters. The two facts 
 appear to me to throw much light on the pathology of diabetic 
 coma. 
 
 Diagnosis. — The diagnosis of diabetic coma is generally easy, 
 when the patient has been under observation for some time and 
 is known to be suffering from diabetes. 
 
 Important indications of commencing coma are: (1) 
 Epigastric pain and nausea; (2) rapidity of the pulse; (3) 
 dyspnoea (see case, p. 282); (4) drowsy mental condition, often 
 with restlessness ; (5) the appearance at the bottom of the urine 
 glass of a small, opaque, greyish white deposit, which on micro- 
 scopical examination is found to consist of enormous numbers 
 of casts. The appearance of this deposit is generally, but not 
 quite invariably, followed by coma, terminating fatally. 
 
 These indications are especially important if the patient is 
 young and wasted, if the bowels are markedly constipated, and 
 if the urine contains a small amount of albumin, and gives 
 G-erhardt's reaction with perchloric! e of iron and the tests for 
 acetone. 
 
 When the patient is seen for the first time in a comatose 
 condition, the diagnosis is, of course, much more difficult. The 
 differential diagnosis from other diseases may then generally be 
 settled by an examination of the urine, which in the case of 
 diabetic coma will have the characters mentioned above — 
 namely, it will contain sugar in considerable quantity, albumin 
 in small quantity, and almost invariably granular casts in 
 enormous numbers ; it will nearly always give a dark brownish 
 red coloration with perchloride of iron and the chemical test for 
 
DIAGNOSIS. 
 
 ■93 
 
 acetone. It is stated that sugar occasionally disappears from 
 the urine in diabetic coma ; certainly, if this is ever the case, 
 it happens exceedingly rarely. I have never met with such a 
 case myself. In a few rare cases the perchloride of iron reaction 
 is slight or absent. 
 
 If no urine is passed by the patient, it can generally be 
 withdrawn from the bladder by the catheter ; but if the bladder 
 should be empty, then the blood examination for the methylene 
 reaction described on p. 191, would be found of service in 
 diagnosis. 
 
 Uraemia, opium poisoning, alcoholic intoxication, and coma 
 from cerebral lesions of various kinds, are all liable to be mis- 
 taken for diabetic coma. 
 
 Uraemia. — Apart from the history, which of course cannot 
 always be obtained from the friends of comatose patients, the 
 following points in the symptomatology, during the stage of un- 
 consciousness, would be of importance : — 
 
 In Favour of Uraemia. 
 (Edema of face, legs, or genitals. 
 
 Anaemia. 
 
 Pulse slow ; when convulsions 
 1 occur, pulse small and frequent. 
 
 Cardiac hypertrophy present in 
 many cases. 
 
 Convulsions common. 
 
 Urine : Sugar absent, albumin 
 present, often in large quantities. 
 
 In Favour of Diabetic Coma. 
 Patient generally much wasted. 
 
 Slight cyanosis of face and limbs. 
 
 Pulse small and rapid. 
 
 Cardiac hypertrophy very rare in 
 severe forms of diabetes associ- 
 ated with great wasting. 
 
 Convulsions very rare indeed; knee- 
 jerks generally absent ; chloro- 
 form smell of breath. 
 
 Urine: Sugar present in large or 
 considerable amount ; albumin 
 present, but only in small 
 quantities. 
 
 Methylene blue reaction with blood. 
 
 Blood examination. The methylene blue reaction (see p. 171) would 
 r]i liii'jui-li diabetic coma from other forms of coma. 
 
294 DIABETIC COMA. 
 
 Opium 'poisoning. — In favour of opium poisoning would be the 
 smell of opium in the breath, the contracted pupils, and the absence 
 of wasting. In favour of diabetic coma, the " acetone " smell 
 of the breath, the above mentioned condition of the urine, and 
 the methylene blue reaction of the blood. 
 
 Alcoholic intoxication. — As mentioned on p. 281, in diabetic 
 coma sometimes the mental excitement is so great at the early 
 stage, that cases have been mistaken for alcoholic intoxication. 
 
 In alcoholic intoxication there is the alcoholic smell of the 
 breath, though this is not a sign of much importance, the peculiar 
 dyspnoea of diabetic coma is absent, and the urine and blood do 
 not present the signs characteristic of diabetic coma. 
 
 In cases of coma from cerebral hemorrhage and other cerebral 
 affections, it has been stated that occasionally sugar is found in 
 the urine ; but certainly this is very rare, and when sugar is 
 present, only a trace is usually found, and diacetic acid and 
 acetone are absent. Any evidence of unilateral paralysis or 
 paresis of the limbs, such as indications of the limbs being more 
 flaccid on one side than the other, or any unilateral affection of 
 the cranial nerves, would be in favour of a gross cerebral lesion. 
 In diabetic coma there are no indications of unilateral paralysis 
 or paresis, and the urine and blood have the characters already 
 described. In the coma of cerebral haemorrhage, and sometimes 
 in other cerebral lesions, the pulse is slow and full, whereas in 
 diabetic coma it is quick and feeble, often almost too rapid to be 
 counted. The presence of optic neuritis would exclude diabetes 
 and point to tumour, abscess, or meningitis. If the urine cannot 
 be obtained, or if its reactions are doubtful, I believe the blood 
 test with methylene blue will be of great service in diagnosis. 
 The difficulty in diagnosis is well illustrated by the case of 
 a patient admitted into the Manchester Eoyal Infirmary a short 
 time ago, under the care of Dr. Steell. A man, ait. about 28, 
 had been found unconscious in the hold of a ship (on the Ship 
 Canal) early on the morning of 9th January 1897. He had 
 done his work all right on 8th January, and no history of 
 accident or previous illness could be obtained. When first 
 examined he was quite comatose. The limbs were flaccid, but 
 there were no signs of unilateral paralysis. The pupils were 
 equal. The urine was drawn away by a catheter, and found to 
 give an abundant reduction of Fehling's solution. This roused 
 suspicions of diabetic coma, since the reduction' of Fehling's 
 
DIAGNOSIS AND PROGNOSIS. 295 
 
 solution indicated more than a trace or small quantity of sugar. 
 Nevertheless one could definitely reject the diagnosis of diabetic 
 coma, since the following symptoms and signs contra-indicated 
 that condition. The specific gravity of the urine was low, 1011. 
 There was no acetone smell of the urine or breath, and the urine 
 gave no brown-red coloration with perchloride of iron. The 
 pulse was not increased in frequency (70 per minute), and the 
 tension was high. The peculiar dyspnoea of diabetic coma was 
 absent, and the breathing was of the Cheyne-Stokes character. 
 There was no wasting, and no cyanosis. Though no indications 
 of external injury could be detected, still post-mortem examina- 
 tion showed that the case was one of fracture of the skull, the 
 fracture involving the squamous part of the temporal bone, and 
 running into the middle fossa. Between the dura mater and 
 the skull was a large clot of blood, compressing the anterior 
 part of the left cerebral hemisphere. 
 
 Prognosis. — The prognosis in diabetic coma is exceedingly 
 grave ; death usually occurs in from twenty-four to forty- eight 
 hours. When once the patient has become unconscious; the 
 case may be regarded as hopeless ; but sometimes the early 
 symptoms, such as dyspnoea, drowsiness, epigastric pain, and 
 nausea appear for a day or two, and then subside without the 
 patient passing into a comatose state. According to Hirschfeld, 
 such abortive symptoms are not rare. 
 
 In an advanced case of diabetes seen in private practice,- 
 these symptoms — dyspnoea, epigastric pain, nausea, and drowsi- 
 ness — became well marked ; the urine gave a deep brownish red 
 coloration with perchloride of iron, a trace of albumin was 
 present, and there was an abundant deposit of casts. The 
 patient was a man set. 24 ; he was much wasted, and the 
 symptoms pointed to the onset of diabetic coma. But large 
 doses of alkalies were prescribed, the symptoms of commencing 
 coma subsided, and the casts disappeared from the urine. The 
 patient lived for six weeks longer without any comatose symptoms 
 returning. 
 
 In another patient, set. 29, who had been under my care on 
 several occasions for two and a half years, the general symptoms in- 
 creased rapidly ; he became very weak and wasted ; the urine pre- 
 sented ;i white deposit, which, microscopically, was found to consist 
 of hyaline casts ; drowsiness developed, and the patient appeared 
 to be passing into a state of diabetic coma; but improvement 
 
296 
 
 DIABETIC COMA. 
 
 occurred in the general condition, the drowsiness passed off, and 
 casts disappeared from the urine. Death occurred a few weeks 
 later, without the symptoms of diabetic coma reappearing. In a 
 third case (which has come under my observation at the 
 Manchester Royal Infirmary), marked dyspnoea, epigastric pain 
 and drowsiness had developed ; but by treatment with very large 
 doses of potassium citrate, prescribed by Dr. Reynolds, who was 
 then the resident medical officer, recovery occurred, and the 
 patient lived for twelve months, and finally died of phthisis. 
 The case was under the care of Dr. Leech, and has been reported 
 by Dr. E. S. Reynolds ( 1S ). 
 
 REFERENCES. 
 
 Brit. Med. Journ., London, 21st August 
 
 1886. 
 '• Leber die Lebensweise der Zuckerkranken," 
 
 Wiesbaden, 1894. 
 Deutsche vied. Wchnschr., Leipzig, 1893, 
 
 No. 38. 
 Ibid., 1893, No. 27. 
 Lancet, London, 30th December 1893. 
 Arch. f. exper. Path. u. Pharmakol., Leipzig, 
 
 1883, Ed. xvii. S. 419. 
 "Leber den diabetes," Berlin, 1884, S. 165, 
 
 and 80-105. 
 8. Sanders and Edin. Med. Journ., 1879, vol. xxv. 
 Hamilton 
 
 Loc. cit. 
 
 '"Prognose rind Therapie der Zuckernkrank- 
 
 heit," Bonn, 1892, S. 50. 
 Arch. f. exper. Path. u. Pharmacol., Leipzig, 
 
 Bd. vii. S. 148. 
 Ibid., Bd. xvii. S. 418. 
 Ibid., Bd. xviii. 
 Brit. Med. Journ., London, 23rd September 
 
 1893. 
 f 5. Roque, Devic, and Rev. de med., Paris, December 1892. 
 Hugounenq 
 
 16. Klemperer . . . Berl. hlin. Wchnschr., 1889, No. 40. 
 
 17. v. Noorden . . . Loc. cit, S. 85. 
 
 18. Reynolds, E. S. . Med. Chron., Manchester, 1891, vol. xiv. p. 
 
 338. 
 
 1. Dreschfeld, J. 
 
 2. Ebstein, "W. . 
 
 3. HlRSCHFELD, E. 
 
 4. Schmitz, R. . 
 
 5. Grube, K. 
 
 6. Stadelmaxn . 
 
 7. Frerichs, F. T. 
 
 9. Dreschfeld, J. , 
 
 10. Schmitz, R. . 
 
 11. "Walter, F. . 
 
 12. Stadelmann, E. 
 
 13. Minkowski . . 
 
 14. Harley-, V. . . 
 
CHAPTER XII. 
 
 PATHOLOGICAL ANATOMY. 
 
 jSTumekous and varied pathological changes have been described 
 in diabetes mellitus, yet, strictly speaking, the disease has no 
 pathological anatomy. In various cases pathological changes 
 have been found in every organ of the body ; but they are 
 general secondary, or due to complications. On the other hand, 
 there is no organ which has not been found frequently normal 
 on pathological examination. In the absence of any information 
 as to the clinical history, or the examination of the urine or 
 blood, a pathologist would be unable to diagnose diabetes 
 mellitus by the macroscopical or microscopical changes found on 
 post-mortem examination. 
 
 It is somewhat remarkable that the negative or indefinite 
 nature of the pathological changes has not given rise to difficulties 
 medico-legally. It is quite possible that a few of the numerous 
 cases of death from unknown causes, in which post-mortem 
 examination has failed to furnish any satisfactory explanation, have 
 been due to diabetes mellitus — cases, for example, which are 
 occasionally attributed to " serous " apoplexy, or to sunstroke, 
 if the weather happens to have been warm. 
 
 Unless a specimen of urine be obtained from the bladder 
 post-mortem, and sugar detected, or unless the blood be proved 
 to contain an excess of sugar, the pathologist would not be able 
 to decide as to the cause of death, providing nothing was known 
 about the patient's symptoms during life. Also it is important 
 to remember that the sugar in the blood diminishes markedly 
 after death. 
 
 The pathological changes in the various organs are neither 
 constant nor characteristic ; they are chiefly secondary in nature, 
 and have already been described in connection with the various com- 
 plications, and in the chapter on etiology and etiological relation. 
 
 1. In the medulla, vagus nerves, and other parts of the base 
 
298 PATHOLOGICAL ANATOMY. 
 
 of the brain, definite macroscopical or microscopical changes 
 are occasionally met with, which are probably the cause of the 
 disease, but such cases are rare, and often the brain is normal 
 or only presents unimportant changes. The changes in the 
 brain have been described and discussed on pp. 125—40. Those 
 found in the spinal cord, the sympathetic and peripheral nerves, 
 are described on pp. 129, 242, 131, 264. 
 
 2. The pancreas in many cases presents changes which 
 are probably the cause of the disease, but in many cases this 
 gland is normal or only presents slight and unimportant 
 abnormalities. The pathological changes in the pancreas and 
 their relation to diabetes have been discussed in Chapter VIII. 
 
 3. The liver presents no characteristic changes : its condition 
 has been described on pp. 116—21. 
 
 4. The changes in the other organs are all secondary in 
 nature. Those met with in the heart and vessels have been 
 described on p. 216 ; those in the lungs on pp. 207—14; those 
 in the alimentary canal on pp. 205, 206. 
 
 It is said that the most constant pathological change in 
 diabetes is the glycogenic degeneration of the epithelial cells of 
 Henle's loop of the tubules of the kidneys ; but this change is of 
 course secondary in nature. The abnormalities met with in the 
 kidneys have been described on p. 219. The spleen is frequently 
 reported to be normal, and when changes have been found they 
 have been unimportant. The condition of the blood has been 
 described in Chapter IX. 
 
 Pathogenesis. 
 
 We have now surveyed the chief facts which are known 
 respecting the etiology, etiological relations, symptoms, and 
 pathological anatomy of diabetes, and some of the more import- 
 ant experimental work in connection with the disease has 
 been referred to. The question remains, What is the exact cause 
 of diabetes mellitus in man ? why do certain individuals become 
 diabetic ? The true nature of the disease still remains very 
 obscure, and it is quite impossible to refer to all of the numerous 
 theories that have been advanced. None of these are satisfactory, 
 however, and in spite of the very extensive literature respecting 
 diabetes, and the great amount of experimental and pathological 
 research in connection with the disease, the riddle still remains 
 unsolved, and will probably long remain so. 
 
PA THO GENE SIS. 299 
 
 As stated above in Chapter II., in health the urine is either 
 free from sugar, or it only contains minute traces, which cannot 
 be detected by the usual clinical tests. 
 
 Why does the urine contain a large quantity of sugar in 
 diabetes mellitus ? In this disease, do the kidneys simply allow 
 the sugar normally present in the blood to pass away into the 
 urine ? i.e. does the normal glucose simply separate itself from 
 the other constituents of the blood, and leak through the urinary 
 tubules into the urine ? Or is the large amount of sugar in the 
 urine clue to an excess of sugar in the blood ? The former 
 supposition is certainly not correct, since the blood would then 
 contain no sugar, or less sugar than in health, but chemical 
 examination shows the reverse ; the blood sugar is usually greatly 
 increased in amount. 
 
 Hence we may conclude that sugar appears in the urine, 
 because there is an excess of sugar in the blood, and the symp- 
 toms of diabetes are also owing to the same cause. This state- 
 ment is certainly true with reference to the majority of cases of 
 diabetes ; and Bernard, Pavy, and most observers believe that 
 glycosuria is always the result of an excess of sugar in the 
 blood (glycsemia). But Seegen ( 1 ), whilst admitting that the 
 above explanation holds good in the majority of cases, believes 
 that in a few of the milder forms, glycosuria is not simply 
 the result of an excess of sugar in the blood. In some of the 
 latter cases he has found that the blood sugar has scarcely 
 exceeded the normal limits. Putting aside these rare and dis- 
 puted cases, there can be no doubt that in a very large majority, 
 it' not in all cases of diabetes, the glycosuria is due to the excess 
 of sugar in the blood. The cause of this excess is the question 
 which requires to be answered. Some authors have attributed 
 it to (1) an excessive formation of sugar in the system ; some 
 have attributed it to (2) a diminished sugar destruction ; whilst 
 others believe (3) that in certain forms there is an excessive 
 sugar formation, but in other forms a diminished sugar destruction. 
 Now it is quite impossible, in a work devoted chiefly to the 
 clinical aspect of diabetes, to mention all the more important 
 arguments which have been advanced in support of each of these 
 views, and only a few can be briefly referred to. 
 
 With respect to the theory of excessive sugar formation, 
 Bunge ( 2 ) has pointed out the following objections: the great 
 excess of sugar could not be derived from the carbohydrates 
 
3oo PATHOLOGICAL ANATOMY. 
 
 of the food, since these are normally converted into sugar, 
 and then transformed into other substances. If the excess 
 of sugar in diabetes had its origin in these carbohydrates, 
 it would be owing to a diminished destruction and not to an 
 increased formation. Neither has the excess of sugar its origin 
 in fats, since diabetic patients can digest and assimilate them 
 in large quantities. As to the proteids, assuming that a diabetic 
 patient could consume 300 grms. in a day, this amount of 
 albumin would not form more than about 200 grms. of sugar. 
 And even if 200 grms. of sugar reached the blood daily, it would 
 not cause diabetes, so long as the decomposition of sugar remained 
 normal. A person whose power of sugar destruction was not 
 impaired, would be able to decompose daily 600 to 1000 grms. 
 of sugar (derived from starchy food), without glycosuria occurring. 
 Hence Bunge thinks that increased sugar formation cannot 
 account for the excess of blood sugar. 
 
 As regards excessive formation, in the strict sense of the 
 term, Bunge's objections undoubtedly hold good with reference 
 to the origin of sugar from foodstuffs, but do not apply to the 
 possible formation of sugar from the albumins of the tissues. 
 
 On the other hand, Kaufmann ( 3 ) brings forward evidence in 
 favour of the view that the excess of sugar in the blood in 
 diabetes is always due to an increase in the sugar formation, and 
 not to a diminished sugar destruction in the capillaries. He 
 believes that in the normal condition the destruction of sugar 
 in the blood is always compensated by an equivalent sugar 
 formation by the liver, and that any modification of the destruc- 
 tion of sugar in the tissues reacts at once on the liver by the 
 nervous paths or by the blood, and impresses on the sugar 
 formation an activity in direct relation to the destruction. 
 
 Kaufmann has isolated the liver by tying all its vessels in a 
 healthy dog, and also in another dog which had been rendered 
 diabetic by extirpation of the pancreas. The sugar in the blood 
 gradually became diminished during its circulation in other 
 parts of the body. But an hour after the isolation of the liver, 
 the blood had lost the same proportion of sugar both in the 
 healthy and in the diabetic animals. After having determined 
 the diminution of sugar, Kaufmann re-established the circulation 
 in the liver by removing the ligatures placed on the vessels. 
 He found that the sugar in the blood soon regained its previous 
 percentage. Hence he concludes that sugar destruction pro- 
 
PA THO GENESIS. 30 1 
 
 ceeds with the same activity in dogs rendered diabetic by 
 pancreas extirpation, as in the normal condition, and that the 
 excess of sugar in the blood is due to increased sugar formation 
 by the liver. Kaufmann believes that increased sugar formation, 
 and not diminished destruction in the capillaries, is the cause of 
 the excess of sugar in the blood in the diabetes produced by 
 pancreas extirpation, by lesions of the medulla, and in all other 
 ways. 
 
 Many authors, however, probably the majority, believe that 
 diabetes mellitus in man is usually due to diminished sugar 
 destruction. 
 
 As already mentioned, in the mild form of diabetes the 
 removal of carbohydrates from the diet causes the glycosuria to 
 cease. Now, in such cases the sugar appears to be derived in 
 some way from the carbohydrates of the food. It may be that 
 the sugar produced by the digestion of carbohydrates is absorbed 
 into the portal system, and then conveyed into the general 
 circulation, without being transformed into glycogen or other 
 substances. Or it may be, that the sugar is converted into 
 glycogen, but the hepatic glycogen is transformed into sugar 
 again in an abnormal manner. 
 
 Seegen ( 4 ) believes that the cause of the mild form, which 
 he terms the " hepatogenic," is the inability of the liver cells 
 to assimilate the carbohydrates in a normal manner. ' But it 
 has been shown by Ehrlich, Kiilz, and v. Mering that glycogen 
 may be present in the liver in considerable quantity in diabetes 
 mellitus. Also, pathological anatomy has failed to reveal any 
 macroscopic changes in the liver associated with diabetes. 
 Further, serious diseases giving rise to destruction of liver 
 tissue — cancer, acute yellow atrophy, cirrhosis, etc. — are not 
 associated with glycosuria. 
 
 According to Pavy (see p. 67), the two lines of defence 
 — intestinal villi and the liver — are inadequate to accomplish 
 their function of synthesising the carbohydrates. Hence the 
 latter reach the general circulation in excessive quantity, and 
 appear as sugar in the urine. 
 
 In the severe forms of diabetes, the sugar in the urine is 
 evidently not dependent simply on the carbohydrates of the food, 
 since the glycosuria persists when the diet consists only of fat 
 and nitrogenous substances, and even during starvation. 
 
 Pavy thinks that in these severe forms the sugar eliminated 
 
3 o2 PATHOLOGICAL ANATOMY. 
 
 is derived not only from the food, but also from the tissues. He 
 believes that the proteids of the body have a glucoside constitu- 
 tion, and that in diabetes of the severe form a carbohydrate is 
 cleaved off from these proteids by the action of some ferment 
 which he supposes to be present in the system. 
 
 According to Seegen, in the severe forms the cells and 
 tissues of the organism have lost their function of destroying 
 the sugar in the blood. 
 
 If we accept the views that the cells of the liver are at 
 fault in diabetes, and in severe cases also the tissues of the 
 organism, the question arises : Why do these tissues behave 
 abnormally ? 
 
 Pavy believes that, as regards the liver, there is a vasomotor 
 paralysis and dilatation of the small arterioles. As a result 
 thereof, the blood passes through the capillaries too readily, and 
 the contents of the small veins are not sufficiently de-arterial- 
 ised. He has shown that a hyperoxygenated state of the blood 
 favours the passage of carbohydrate matter into glucose. The 
 vasomotor paralysis Pavy refers to changes in the nervous 
 system. 
 
 Seegen points out that Pavy was able to produce diabetes 
 experimentally, by injury to the nervous system after ligature of 
 the hepatic artery, and he asks how vasomotor paralysis can be 
 regarded as the explanation, if this be true. 
 
 According to Ebstein ( 5 ), the amount of carbonic acid formed 
 in the tissues in diabetes is diminished. He believes that, owing 
 to some abnormal state of the protoplasm, there is an altered con- 
 dition of the internal respiration, with insufficient formation of 
 carbonic acid. Glycogen is present in the tissues in greater 
 quantity in diabetes than in health, and it is found in unusual 
 positions. Ebstein also believes that in the tissues there is a 
 diastatic ferment, and that the action of this ferment on glycogen 
 is hindered by carbonic acid. He believes that in diabetes the 
 diastatic ferment is present in normal quantity, but, owing to the 
 deficient amount of carbonic acid the inhibitory action of the 
 latter is less than normal. Hence an excess of sugar is pro- 
 duced ; the sugar in the blood is increased, and glycosuria 
 follows. Ebstein's views are in part based on the experi- 
 ments of Pettenkofer and Voit, which appeared to show that 
 the excretion of carbonic acid was diminished in diabetes ; 
 but these results have been disputed by Leo ( 6 ) and others. 
 
PATHOGENESIS. 303 
 
 According to Leo, the excretion of carbonic acid in diabetes is 
 quite normal ; and Ebstein's views have not been accepted by 
 most pathologists and physicians. 
 
 As pointed out on in Chapter VIII., section D, there is strong 
 evidence that some change in the nervous system is the starting- 
 point of the disease in many cases. 
 
 Also, pathological observations and the results of experiments 
 on animals render it very probable that in some cases of diabetes 
 pancreatic changes are the cause of the disease (see Chapter 
 VIII, section E). 
 
 It was pointed out on p. 15 6 that atheroma might act as 
 the starting-point of diabetes, by producing changes in the 
 nervous system or in the pancreas, and it is probable that a few 
 cases of diabetes originate in this way. 
 
 Diabetes has been attributed by some authors to disease of 
 the muscles. Bunge states that " as the bulk of the sugar is 
 normally decomposed in the muscles, it seems probable that 
 diabetes may fundamentally be due to a disturbance' of the 
 chemical processes in muscle." In stout diabetics a change 
 from a sedentary to an active life has occasionally caused the 
 glycosuria to cease, and a few cases of diabetes have been success- 
 fully treated by systematic muscular exercise. Bunge points 
 out that the chemical processes in muscles are subject to the in- 
 fluence of the nervous system, and that numerous observations 
 tend to show that the symptoms in diabetes are sometimes 
 caused by disturbances which originate in the central nervous 
 system. 
 
 The opinion appears to be gradually gaining ground, that 
 diabetes is not a pathological entity, but rather a group of 
 symptoms which may be produced by various morbid changes in 
 the system. Just as fever is the result of many diseased pro- 
 cesses, so it is quite possible, even probable, that diabetes is not 
 always due to the same primary pathological condition, and that 
 sometimes the starting-point of the disease is in the nervous 
 system, sometimes in the pancreas, occasionally in disease of the 
 arteries (arterio-sclerosis), possibly in the muscles sometimes, and 
 possibly it is due to various other causes and to endogenous 
 or inherited morbid conditions. 
 
3°4 
 
 PATHOLOGICAL ANATOMY. 
 KEFEKENCES. 
 
 1. Seegen, J. . . . " Der Diabetes mellitus," Berlin, 1893, S. 
 
 89-90. 
 
 2. Bunge, G. . . . " Text-Book of Physiological and Pathological 
 
 Chemistry," translated by L. C. Wool- 
 dridge, London, 1890, pp. 419, 420. 
 
 3. Kaufmann . . . Semaine med., Paris, 16th January 1895. 
 
 4. Seegen, J. Loc. cit., S. 91. 
 
 5. Ebstein . . . . " Die Zuckerharnruhr, ihre Theorie u. Praxis,"" 
 
 Wiesbaden, 1887. 
 
 6. Leo, H Centralbl. f. Min. Med., Bonn, 1892, No. 25. 
 
CHAPTER XIII; 
 
 FORMS OF DIABETES — TERMINATION — PROGNOSIS. 
 
 Eokms of Diabetes Mellitus and G-lycosukia. 
 
 Theke are two chief forms of the disease, severe and mild, 
 besides several other subvarieties. The main features of the 
 two more important forms of the disease may be here re- 
 peated. 
 
 1. In the severe form, the symptoms — thirst, hunger, and 
 diuresis — are well marked. The amount of sugar excreted is 
 large, and does not cease, though it may be diminished con- 
 siderably, when carbohydrates are rigidly excluded from the 
 diet. There is often rapid loss of strength and marked wasting 
 (hence the French term diabete maigre). This form of diabetes is 
 met with chiefly in young persons or in persons under middle 
 life. Often the patients are poor hard-working people, and a 
 large proportion of the cases of diabetes seen in hospital prac- 
 tice belong to this class. The urine often contains acetone, and 
 gives a dark reddish brown coloration with perchloride of iron 
 (diacetic acid). According to Schmitz, the urine excreted during 
 the night may contain a greater amount of sugar than that 
 excreted during the clay, even though carbohydrates be taken in 
 the food (see p. 307). The sugar excreted is therefore not 
 simply derived from the carbohydrates of the food. These cases 
 frequently run a comparatively rapid course, and are often 
 fatal in one to three years, death generally occurring from coma 
 or phthisis. 
 
 But occasionally cases are met with which run a very acute 
 course. The patients are often children or young people, under 
 the age of 30, though acute cases may occur later in life. The 
 disease may terminate fatally in a few weeks after the symptoms 
 are first noticed. In one patient at the Manchester Eoyal 
 Infirmary, who was IS years of age, the symptoms were first 
 
3 o6 FORMS OF DIABETES. 
 
 noticed only two weeks before death occurred ; in two cases the 
 symptoms were first noted only four weeks before the fatal 
 termination ; and in two other cases death occurred at the end 
 of six weeks and eight weeks respectively after the symptoms 
 first attracted attention. Death occurred from diabetic coma 
 in all of the cases. Of course it is possible that in some of 
 these cases the disease may have been present for a short time 
 before the symptoms attracted the patient's attention ; but 
 others are on record, in which the urine had been examined 
 and found free from sugar shortly before the diabetic symptoms 
 appeared. Thus in Wallach's case, referred to on p. 166, the 
 duration of the disease was under two weeks, as shown by the 
 urine examinations. 
 
 2. In the mild form of the disease the sugar disappears from 
 the urine on withdrawing carbohydrates from the food ; generally 
 thirst and diuresis are not so marked as in the severe forms ; 
 and dryness of the skin is often absent or slight. The patients 
 are often old or middle-aged persons, and frequently stout. (To 
 this form the term cliabUe gras is applied by French writers, 
 and the term chronic glycosuria by certain English authors.) 
 Not infrequently the patient also suffers from gout ; and some- 
 times the urine contains a great excess of uric acid, but usually 
 acetone and diacetic acid are absent. At an advanced stage, 
 especially if the patient has been kept on an injudicious diet, 
 marked wasting may occur. In this mild form of diabetes the 
 night urine contains less sugar than that excreted during the 
 day ; in fact, the night urine may be quite free from sugar. 
 The sugar in the urine appears to be derived from the carbo- 
 hydrates of the food. 
 
 The onset of the disease is usually very gradual ; fre- 
 quently the patient seeks advice on account of some complica- 
 tion, such as carbuncles, gangrene, etc., and on testing the urine 
 in a routine examination sugar is detected. 
 
 The mild form of diabetes generally runs a chronic course. 
 Sometimes the disease is associated with kidney mischief — 
 chronic parenchymatous or interstitial nephritis. 
 
 Some authors regard the mild and severe forms of diabetes 
 as two different diseased conditions, but others believe that the 
 mild form is simply an early stage of the severe variety. There 
 can be no doubt, however, that many of the severe forms of 
 diabetes have the characters of the severe variety from the onset, 
 
ONSET OF THE DISEASE. 307 
 
 and many of the cases of the mild form do not develop into 
 the severe form. On the other hand, cases are sometimes met 
 with in which at first the sugar excretion is very small, the 
 diuresis slight, and the symptoms are those of the mild variety, 
 but at a later date, months- or years afterwards, the sugar 
 excretion and diuresis increase, and all the symptoms of the 
 severe form develop. Several cases of this kind have come 
 under my own observation (see p. 166). There are, therefore, 
 besides the mild and severe forms, transitional and intermediate 
 varieties. 
 
 The following are examples of the differences in the 
 quantity of sugar excreted in the day urine and in the night 
 urine, in the mild and the severe forms of diabetes. 
 
 In a case of diabetes of the mild form in a very stout patient, 
 jet. 49 (who came under treatment for eczema of the vulva), 
 the specific gravity of the urine was 1029 ; and the total 
 quantity of sugar excreted in twenty-four hours was 270 grs., 
 the diet being unrestricted. When the diet was limited to flesh 
 meat, fish, milk (2 pints daily), beef -tea, green vegetables, and 
 a little bread, the sugar excretion diminished to 78 grs. daily. 
 On three occasions the night urine, passed between 1 a.m. and 
 8 a.m., was collected and kept separate from the rest of the 
 urine. It was found quite free from sugar. In other cases of 
 the mild form of diabetes I have found sugar absent from the 
 night urine, but present in the day urine. 
 
 In severe forms of diabetes I have never found the night 
 urine free from sugar. Sometimes I have found that more 
 sugar was excreted during the night than during the day ; at 
 other times the quantities have been about the same ; some- 
 times the sugar excreted during the night has been a little 
 less than that excreted during the day. 
 
 Thus in the case of a boy, set. 19, who was suffering from 
 a severe form of diabetes, the sugar excreted during the night 
 was either more than that excreted during the day, or it was 
 only a little less in amount. 
 
 The amount of urine from 8 a.m. to 9 p.m. was 102 oz. 
 Total quantity of sugar excreted during the day was 3769 grs. 
 The night urine from 9 p.m. to 8 a.m. was 106 oz. Total 
 quantity of sugar excreted during the night, 4223 grs. 
 
 The patient had breakfast at 8 a.m., dinner at 1, tea at 4.30, 
 and a little milk at 7 p.m. No food was taken afterwards. 
 
3 o8 FORMS OF DIABETES. 
 
 In some of the mild forms of diabetes no sugar appears in 
 the urine, even when a limited quantity of carbohydrate food is 
 taken ; but when the amount of carbohydrates in the food 
 passes beyond a certain quantity, glycosuria occurs. There is 
 thus a tolerance of a certain amount of carbohydrate food ; but 
 if this limit be exceeded, sugar is found in the urine. In other 
 cases the urine only remains free from sugar when all carbo- 
 hydrate food is withdrawn. 
 
 In some of the mildest cases the quantity of urine excreted 
 is normal, or almost normal, whilst the specific gravity is high ; 
 thirst and other diabetic symptoms may be absent. Such cases 
 have been described by the term diabetes decipiens. 
 
 In a case of the kind which I have followed for some time, 
 the urine has a specific gravity of 1040—42 ; the amount is 
 about 50 oz. ; the sugar varies from 5 to 7 per cent. There 
 is no thirst, no loss of flesh, and the patient (a young adult) 
 appears and feels in perfect health. 
 
 In another variety of the mildest form, the symptoms may 
 disappear from time to time ; this is known as intermittent 
 diabetes. Cases are sometimes met with in which the sugar 
 excretion ceases, and for a time the patient is quite well ; then 
 some indiscretion as regards mode of life, etc., causes the glyco- 
 suria and other symptoms to return. The sugar may disappear 
 on several occasions and return again, and finally the disease 
 may terminate fatally. 
 
 Thus, for example, a gentleman xb. 56 consulted me on 
 account of diabetes. He complained chiefly of dryness in the 
 mouth; his urine had a specific gravity of 1035, and was loaded 
 with sugar. The patient was stout, and for five or six years 
 sugar had been found in his urine by his medical attendant ; 
 but the glycosuria had been intermittent, and sometimes had 
 been absent for four to six weeks at a time. 
 
 Symptomatic glycosuria, in which the sugar excretion can 
 be referred to some primary affection of the nervous system, or 
 to other disease, has already been mentioned on p. 91. The 
 symptoms are often slight, and frequently temporary. 
 
 Diabetes with 'pigmentation of the skin — Diabtte bronze 
 (Hanot 1 ). — In 1882, Hanot and Chauffard ( 2 ) described a disease 
 characterised by the association of symptoms of diabetes mellitus 
 with hypertrophic cirrhosis of the liver and pigmentation of the 
 skin. To this disease Hanot gave the name of diabtte bronze. 
 
SYMPTOMATIC GLYCOSURIA. 309 
 
 In his second series of cases the liver was of normal size, but 
 the other symptoms were present. 
 
 The disease has since been described by Letulle ( 4 ), 
 Brault and Gaillard ( 5 ), Earth, Gonzalez Hernandez ( 6 ), Palma, 
 Mosse ( 7 ), de Massary and Potier ( 8 ), and Marie ( 9 ). 
 
 The following is Hanot's clinical description of the disease : — 
 
 " The symptoms consist of general signs of diabetes, with 
 abundant glycosuria ; of melanoderma, most marked upon the 
 face, the limbs, and the genital organs ; general symptoms of 
 hypertrophic cirrhosis, hypertrophy of the liver and of the spleen, 
 with slight ascites and distension of the abdominal cutaneous 
 veins. 
 
 " The onset is frequently sudden. Wasting, diarrhoea, oedema 
 of the lower limbs, soon appear ; loss of strength rapidly ensues, 
 and death supervenes, either during coma, or from septic pneu- 
 monia following diabetic gangrene. Although the pathogenesis is 
 not yet definitely proved, it appears to me that diaMte 'bronze is 
 a distinct morbid entity. 
 
 "In different cases the liver weighed from 1800 to 3500 
 grins. 
 
 " The density of the tissue is considerable, and its reddish 
 brown coloration recalls that of rust, or of old untanned leather. 
 The surface is smooth, and the biliary ducts are, as a rule, 
 unaffected. 
 
 " The intestinal glands are slaty-black in colour. The 
 ascites is generally moderate in amount, from 2 to 9 litres. 
 The spleen is increased in volume, indurated, and presenting a 
 reddish brown colour, somewhat resembling that of the liver. 
 The mesenteric and mediastinal glands have somewhat the same 
 rusty colour as the liver. The ochre-coloured pigment, which 
 infiltrates most of the organs and the Malpighian layer of the 
 skin, is an iron -containing pigment, as is proved by its reaction 
 with ammonic-hydric sulphate and potassium ferrocyanide. 
 Brault has insisted upon the properties which distinguish the 
 yellow pigment from the other ones, the melanin pigment and 
 the pigment of Addison's disease, which have no reaction for 
 iron. The pigment contained in melanotic sarcomas is soluble 
 in potash; and if one places a fragment of a melanotic sarcoma 
 in a tube containing potash, the pigment is entirely dissolved 
 out, and the liquid in the tube takes the colour of port wine. 
 The ochre-coloured pigment does not, on the other hand, dissolve. 
 
3 to FORMS OF DIABETES. 
 
 " In the liver, the pigment, which in Dr. Lapicque's experi- 
 ence appears to be a ' salt of iron,' is found between the con- 
 nective tissue fibres, in the fibres themselves, and in the hepatic 
 cells. The cells sometimes contain a few granules only, and 
 sometimes appear to be loaded with pigment, having lost their 
 nuclei, and becoming transformed into yellowish masses." 
 
 Hanot and Chauffard regarded the pigmentary cirrhosis as 
 the result of the diabetic condition. They believed that the 
 pigment was developed in the liver, and then became diffused in 
 the form of minute emboli through the whole organism. 
 
 Letulle does not think the pigment found in the various 
 organs is formed originally in the liver, but believes it arises 
 in various parts from the haemoglobin of blood or muscle. 
 
 Gonzalez and Eeiner have found that the pigment contains 
 iron. Mosse believes that the evidence is in favour of its origin 
 from altered haemoglobin, and Massary and Potier think that the 
 pigment is separated from the blood in the organs themselves. 
 
 According to Marie, diabete bronz6 is neither a pigmentary 
 cirrhosis in diabetes mellitus, nor a cachexie bronze', but a true 
 clinical and pathological entity ; in fact, a new disease. Hanot's 
 observations lead him to agree with this view. 
 
 Phosphatic diabetes ( 10 ) is the name which has been given to 
 a somewhat indefinite group of cases, in which the excretion of 
 phosphate of lime is excessive. There is great nervous irrita- 
 bility, derangement of digestion, great emaciation, and severe 
 aching pain in the back and loins. As the disease advances, 
 the amount of urine increases, and symptoms appear which are 
 somewhat analogous to those of diabetes, especially the insipid 
 form. Hence the name phosphate diabetes. The urine is 
 usually free from sugar, but Osier ( n ) states that in some 
 instances sugar has been present, and in others it has subse- 
 quently appeared. 
 
 Termination. 
 
 (a) Recovery. — According to Cantani ( 12 ), by a rigid diet, a 
 cure is not infrequently obtained in mild forms. But many 
 physicians, who have had a large experience of the disease, state 
 that they have never met with any case of true diabetes in 
 which permanent recovery has occurred. I have seen a number 
 of cases of symptomatic glycosuria, associated with cerebral 
 lesions, etc., in which the sugar excretion has ceased, but I have 
 
TERMINA TION. 3 1 i 
 
 never seen true recovery from a severe form of diabetes. In 
 the milder forms, by restricted diet, the sugar may disappear 
 from the urine entirely or partially ; but when a less rigid diet 
 is taken, the glycosuria returns. Patients suffering from the 
 mild form may, by following a restricted diet, live for years 
 without a return of the symptoms, or may live for years with- 
 out any symptoms beyond the presence of a small amount 
 of sugar in the urine. The urine of one of my patients — an old 
 man of 72, who suffers from a mild form of diabetes — has 
 contained sugar for seventeen years, but the other symptoms 
 have been slight. In some of these cases, after the diet has 
 been restricted for some time, the system is able to tolerate a 
 certain amount of carbohydrates. A small quantity of starchy 
 food may then be added without the appearance of sugar in 
 the urine. Occasionally the symptoms of diabetes in its mildest 
 form (or chronic glycosuria) disappear temporarily, or, very 
 rarely, permanently. In mild forms of diabetes in women, 
 occurring about the climacteric period, recovery sometimes takes 
 place. 
 
 In a case which has recently come under my care, a female, 
 set. 20, began to suffer from thirst and diuresis. Six to 
 eight pints of water were drunk daily, in addition to the usual 
 quantity of fluids at meals. The thirst and diuresis continued 
 two or three weeks, and then rapidly diminished. The urine 
 was examined, and found to reduce Fehling's solution. The 
 specific gravity was at one time as high as 1042. The diet 
 was restricted, and the thirst and diuresis completely dis- 
 appeared, and the reducing substance in the urine became less. 
 Three months after the symptoms had first commenced, I found 
 that the urine contained a very small quantity of sugar when 
 the patient was on a restricted diet ; but by allowing bread in 
 considerable amount, the sugar reached 3 per cent. The quan- 
 tity of urine was not increased, 36, 40, 44, 45 oz. being the 
 amounts of urine for four consecutive days. The specific gravity 
 varied from 1018 to 1032. That the reducing substance was 
 sugar was shown by the fermentation test, and the reaction with 
 phenyl-hydrazin. There was no thirst at the time, and the 
 patient felt quite well. One week later every trace of sugar 
 disappeared from the urine, and all diabetic symptoms were 
 absent, but one month afterwards the glycosuria returned. 
 
 A case, has come under my observation in which the patient, 
 
3 1 2 FORMS OF DIABE TES. 
 
 a woman set. 52, had suffered from very great thirst and very 
 great diuresis for six months during pregnancy. The urine was 
 at that time examined by her medical attendant, and diabetes 
 diagnosed. After the birth of the baby, the thirst and diuresis 
 disappeared at once, and did not return for fifteen years. She 
 then began to suffer from slight thirst about once a week; this 
 increased, and three years later she presented the symptoms of 
 a severe form of diabetes, which terminated fatally. For nine 
 years she had had much mental anxiety, worry, and overwork, 
 owing to her husband having become a helpless paralytic. 
 
 In another case which has come under my observation, a 
 female set. 60 had suffered from great thirst, and had passed a 
 large quantity of urine. She had lost flesh, and had suffered 
 from pain in the limbs. After these symptoms had been present 
 for nine months, the urine had been examined by a medical 
 man, and diabetes diagnosed. The sugar gradually diminished, 
 and all the diabetic symptoms disappeared, except pain in the 
 legs. Twelve months after the onset of thirst, the urine had a 
 specific gravity of 1014. Albumin was absent. A small quan- 
 tity of sugar was present, as indicated by Fehling's solution, by 
 the phenyl- hydrazin test, and by Seegen's test, after filtration 
 through animal charcoal. Trie amount of urine varied from 39 
 to 43 oz. One week later the urine was free from the slightest 
 trace of sugar, and remained so during the three weeks the 
 patient remained under observation. 
 
 When diabetes is associated with phthisis or other serious 
 complications, shortly before death the glycosuria and other 
 diabetic symptoms have occasionally disappeared, and only those 
 of phthisis or the complications have remained. 
 
 (b) Termination in other diseases. — 1. Albumin may appear 
 in the urine, and occasionally symptoms of chronic parenchymatous 
 or interstitial nephritis develop. In these cases sometimes 
 the kidney symptoms gradually advance, whilst the glycosuria 
 diminishes ; and, finally, the sugar excretion ceases, and only the 
 symptoms of nephritis remain. 
 
 2. Occasionally the symptoms of diabetes mellitus give place 
 to those of diabetes insipidus, and the sugar then disappears 
 from the urine. A few of such cases are on record in which, 
 finally, the polyuria has disappeared, and complete recovery 
 occurred. The relation of the two diseases has recently been 
 carefully discussed by Senator ( 13 ) (see p. 115). 
 
TERMINA TION AND D URA TION. 313 
 
 (c) Fatal termination. — Diabetes may terminate fatally in 
 the following ways : — 
 
 1. By diabetic coma. This is the most frequent termination. 
 
 2. By tubercular phthisis. This is a termination met with 
 chiefly in young patients suffering from the severe form of the 
 disease. As already stated, if phthisis becomes advanced, death 
 does not generally occur from coma (though there are exceptions 
 to the rule). 
 
 3. By other lung complications, such as pneumonia, pul- 
 monary gangrene, non-tubercular phthisis, etc. ; but these ter- 
 minations are rare. 
 
 4. By exhaustion and marasmus. 
 
 5. By nephritis. 
 
 6. By cerebral haemorrhage or softening. 
 
 7. By carbuncle or gangrene. 
 
 The following is the mode of termination in the last forty-two 
 fatal cases of diabetes which have come under my observation : — 
 
 Cases. 
 
 Diabetic coma . . . . . . . .29 
 
 Asthenia from phthisis and lung disease (six tubercular, 
 
 one non-tubercular) ...... 7 
 
 Cardiac failure ...... 
 
 Asthenia from cirrhosis of the liver 
 
 Pyeemic condition (multiple abscesses of the liver) 
 
 Acute pneumonia . . ... 
 
 Gangrene . . . . . . 
 
 42 
 
 Duration. 
 
 In some severe cases the duration of the disease may be 
 very short — two, four, or six weeks ; but these cases are rare, 
 and a duration of six months to three years or longer is much 
 more frequent ; whilst in the mildest forms the disease may run 
 a very chronic course of five, ten, fifteen, or even twenty years. 
 In one case, an old man of 72, the disease had been detected 
 over seventeen years before he came under my treatment. 
 
 PROGNOSIS. 
 
 Before giving a prognosis, it is important to place the 
 patient on a restricted diet (taking the precautions mentioned 
 
3i4 FORMS OF DIABETES 
 
 on p. 322), in order to decide to which form of the disease the 
 case belongs. The prognosis depends chiefly on the form of the 
 disease, and on the age of the patient. 
 
 The following are unfavourable prognostic indications : — 
 
 1. A severe form of the disease, when the most rigid exclu- 
 sion of carbohydrates from the diet fails to remove sugar from 
 the urine. 
 
 2. Early age. The younger the patient the more unfavour- 
 able the prognosis as a general rule, the prognosis being especi- 
 ally bad in children, and patients under 30. (In twenty-eight 
 cases of diabetes in children, collected by Wegeli,the duration of 
 the disease was very short, often some months only. The 
 longest duration was four and a half years.) 
 
 3. Marked wasting, great loss of weight and loss of strength, 
 especially when the disease is of short duration. 
 
 4. The occurrence of pulmonary tuberculosis as a com- 
 plication. 
 
 5. Severe general symptoms and marked digestive disturb- 
 ance. 
 
 6. The occurrence of gangrene. 
 
 7. A history of other members of the family having suffered 
 from a severe form of diabetes. 
 
 8. Unfavourable conditions of life. Amongst poor people 
 often the disease is not recognised until a late stage ; also the 
 treatment cannot be carried out so carefully. Mental anxiety 
 and worry often hasten the fatal termination. 
 
 9. The presence of a well-marked brownish red coloration 
 of the urine with perchloride of iron. 
 
 10. The sudden occurrence of a greyish white urinary deposit 
 consisting of casts. 
 
 11. A high degree of acetonuria or a marked increase in the 
 amount thereof (9, 10, and 11 indicate the approach of diabetic 
 coma). 
 
 12. The appearance of other symptoms indicating the onset 
 of coma. 
 
 13. The occurrence of symptoms of nephritis, especially if 
 the glycosuria gradually disappears and is replaced by albu- 
 minuria. 
 
 It is to be remembered that in severe cases diminution in 
 the amount of sugar in the urine is not always a favourable 
 sign. Sometimes, as the glycosuria diminishes, the wasting 
 
PROGNOSIS— DIA GNOSIS. 3 1 5 
 
 increases, the patient loses weight rapidly, and the general 
 condition becomes worse ; and this is especially liable to occur 
 when phthisis is a complication. In such cases, sugar may dis- 
 appear finally, and death occurs from the asthenia of phthisis. 
 
 Favourable indications. — 1. A mild form of the disease; the 
 sugar being removed from the urine by the withdrawal of carbo- 
 hydrates from the diet. 
 
 2. Advanced age. The older the patient the better the 
 prognosis as a rule. 
 
 3. The association of diabetes with obesity, gout, or the uric 
 acid diathesis. 
 
 4. A long duration of the disease without the occurrence of 
 complications, of much wasting, or of loss of weight. 
 
 5. A history of a mild form of diabetes in other members of 
 the family. 
 
 6. Favourable conditions of life ; absence of mental worry 
 and anxiety ; good social position and greater likelihood of early 
 recognition and careful treatment of the disease. 
 
 7. The onset of the disease about the climacteric period in 
 females. 
 
 8. The transition of diabetes mellitus into diabetes insipidus. 
 
 Diagnosis. 
 
 Usually the diagnosis of diabetes is very easy, but (1) the 
 disease may be overlooked and the patient treated simply for 
 one of the complications; or (2) various reducing bodies in 
 the urine may be mistaken for sugar; or (3) a temporary glycosuria 
 may be mistaken for a true diabetes. 
 
 When the patient complains of thirst and diuresis, when 
 the urine is found to be pale and straw-coloured, of high specific 
 gravity 1024 to 1040 or more, and when a well-marked reaction 
 is obtained with Fehling's solution, there can be no doubt about 
 the diagnosis. But difficulty arises chiefly, if the patient should 
 complain only of wasting or weakness, or of some of the com- 
 plications, and should make no mention of thirst or diuresis. 
 The emaciation often rouses suspicions of phthisis, and if the 
 urine bo not examined, diabetes is perhaps never thought of. 
 It is important, of course, to examine the urine in all cases, as 
 h mutter of routine, before making an exact diagnosis; but this 
 1 especially necessary in all cases of wasting. The urine ought 
 
3 1 6 FORMS OF DIABE TES. 
 
 also to be examined for sugar in all cases of pruritus or eczema 
 of the genitals, boils, carbuncles, gangrene, cataract. 
 
 It is important to remember, as pointed out on p. 10, 
 that Fehling's solution is reduced by other bodies besides glucose 
 — -bj r glucuronic acid, lactose, alkapton, etc. Especially when 
 the quantity of oxide of copper thrown down is small, or when 
 other diabetic symptoms are absent, there is often considerable 
 difficulty in diagnosis. It is then necessary to decide definitely 
 whether the reducing substance is or is not grape sugar. (The 
 detection of small quantities of grape sugar with certainty has 
 been considered in the chapter on "Urine Testing," p. 35.) 
 
 Having determined that the urine does contain grape sugar, 
 it is necessary to ascertain the quantity roughly if not exactly. 
 Then the question arises, Does the patient suffer from diabetes ? 
 It is important to remember that the presence of grape sugar in 
 a specimen of urine does not necessarily, per se, indicate diabetes 
 mellitus. The patient may be suffering from — 
 
 1. Temporary glycosuria: 
 
 (a) physiological — alimentary glycosuria (rare), or 
 
 (b) pathological, and secondary to some other disease. 
 
 In these cases (a and b) the sugar is generally small in quantity; 
 it soon disappears, and the urine is not increased, or only slightly 
 increased, in amount. 
 
 2. The glycosuria may be permanent, and the patient may be 
 suffering from 
 
 (a) the mild form of diabetes or chronic glycosuria, in which 
 
 sugar is removed by a restricted diet ; or 
 
 (b) the severe form of the disease, in which the glycosuria is 
 
 not removed by a restricted diet. 
 It is important, therefore, to watch the case for a little 
 time, to see if the glycosuria is persistent, and to determine 
 the quantity of urine and the amount of sugar. The duration 
 and the amount of the sugar are the points of greatest import- 
 ance in the diagnosis, and are not simply t)f theoretical interest. 
 Thus, for example, it is important not to pronounce the patient 
 to be suffering from diabetes simply because the urine contains 
 sugar. It is necessary to decide to which of the above groups 
 the case belongs. Great harm may be done to a patient who 
 suffers from a temporary glycosuria (physiological or patho- 
 logical) by a mistaken diagnosis of diabetes. The mental shock 
 might be great, especially if any relative of the patient had 
 
DIAGNOSIS. 317 
 
 died of diabetes. Thus a case has been recorded, in which the 
 mental shock produced by the discovery of a small amount of 
 sugar in the urine gave rise to Graves' disease, which termin- 
 ated fatally (Grube). The case should be therefore carefully 
 investigated, so that, if possible, the patient may be reassured, 
 at the same time that he is informed of the presence of sugar, 
 that he is suffering from a temporary or mild glycosuria and 
 not a true diabetes. Then, again, if it be found that the patient 
 suffers from true diabetes, the variety should be determined, so 
 that, in case it is very mild, a favourable prognosis may be 
 given. When the form is severe, the medical man ought to use 
 the greatest discretion in informing the patient so as to cause 
 as little mental shock as possible, and the patient ought also 
 to be cheered by pointing out any favourable features of his case, 
 if such exist. 
 
 In the severe form of the disease, an abrupt and hasty 
 statement of an unfavourable prognosis might do much to hasten 
 the fatal termination. 
 
 Then it is also important, for another reason, that the medical 
 man should decide to which of the above classes the case of 
 glycosuria belongs. His reputation may suffer considerably from 
 the hasty diagnosis of diabetes, when the case is really one 
 of mild or temporary glycosuria, from which the patient may 
 recover in a short time. Even when all the above precautions 
 are taken, there are still cases in which it is occasionally difficult 
 to say whether the patient is suffering from temporary glycosuria 
 or from the earliest stage of diabetes. Occasionally, only a 
 guarded diagnosis can be given, and the case watched. 
 
 There are various mild forms of the disease of long duration, 
 in which the quantity of urine is normal or only slightly in- 
 creased, that are better described as chronic glycosuria than 
 true diabetes. 
 
 Having decided that the case is one of diabetes or chronic 
 glycosuria, it is important to try and determine if it is secondary 
 to any other disease, as, for example, lesions of the brain, or 
 pancreas, atheroma, gout, obesity, etc., but at present this is 
 possible in a few cases only. 
 
 Occasionally, though very rarely, locomotor ataxia is com- 
 plicated with glycosuria or diabetes. The knee-jerks are 
 frequently lost in diabetes; often there are pains in the legs, 
 tenderness of the calf muscles, and other slight symptoms of 
 
3 i8 FORMS OF DIABETES. 
 
 peripheral neuritis. Hence a difficulty occasionally, though very 
 rarely, arises in diagnosing between an early stage of locomotor 
 ataxia with secondary glycosuria, and diabetes with loss of knee- 
 jerks and early symptoms of a secondary peripheral neuritis, 
 particularly if the latter be of the ataxic form. Again, a 
 patient may come under treatment for a perforating ulcer of 
 the foot ; on examination, the knee-jerks may be absent, and 
 pains in the legs may be troublesome. Such cases are also 
 liable to be mistaken for locomotor ataxia, especially if the 
 diabetic symptoms be slight. 
 
 A careful examination of the urine, and a consideration of 
 the symptoms and general history of the case, are usually sufficient 
 to enable a decision to be arrived at easily. Also the Argyll 
 Robertson pupil, the sharp shooting nature of the pains, and a 
 girdle sensation would be in favour of locomotor ataxia. Grube ( u ) 
 thinks, however, that even the failure of the pupillary reaction is 
 not always diagnostic, but believes that the presence of bladder 
 symptoms is important. Sudden and involuntary discharge of 
 urine or other slight bladder symptoms are frequently present in 
 locomotor ataxia, but do not occur in diabetes with peripheral 
 neuritis. 
 
 REFERENCES. 
 
 1. Hanot Brit. Med. Journ., London, 25th January 
 
 1896. 
 
 2. Hanot et Chauf- Rev. de vied., Paris, 1882, p. 385, avec 
 
 fard planches. 
 
 3. Hanot et Schach- Arch, dephysiol. norm, et path., Paris, 1886, 
 
 mann p. 90, avec planches. 
 
 4. Letulle .... Semaine med., Paris, 1885, p. 408. 
 
 5. Brault et Gail- Arch. gen. de med.. Paris, January 1888. 
 
 laro 
 
 6. Gonzalez Hern an- These de Montpellier, 1892. 
 
 dez 
 
 7. Mosse Congres de med. et d'hyg., 1894, p. 777. 
 
 8. M a s s a r y et Bull. Soc. anat. de Paris, April 1895. 
 
 Potier 
 
 9. Marie Semaine med., Paris, 22nd May 1895. 
 
 (The above references are taken from Hanofs paper.) 
 
 10. Ralfe, C. H. . . " Diseases of the Kidney," London, 1885, 
 
 p. 491. 
 
REFERENCES. 
 
 3 1 9 
 
 11. OSLER, W. 
 
 12. Cantani . 
 
 13. Senator, H. 
 
 14. Grube, K. 
 
 " The Principles and Practice of Medicine," 
 
 Edinburgh and London, 1895, second 
 
 edition, p. 776. 
 " Der Diabetes mellitus," German trans, by 
 
 Dr. Hahn, Berlin, 1880, S. 52-107. 
 Deutsche med. Wchnschr., Leipzig, 10th June 
 
 1897. 
 Neurol. Gentralbl., Leipzig, 1895, No. 1. 
 
CHAPTER XIV. 
 
 THE TREATMENT OF DIABETES. 
 
 Until more light has been thrown upon the exact pathogenesis 
 of diabetes mellitus, the treatment of many cases will remain 
 unsatisfactory. In the milder cases much can be done, and often 
 excellent results are obtained ; but in the more severe forms the 
 indications for treatment are less clear, and the results unsatis- 
 factory. In the latter cases treatment can only be empirical, 
 but it need not necessarily be unscientific. 
 
 Prophylaxis of Diabetes. 
 
 It has been shown that in healthy individuals the sugar- 
 destroying power of the organism is not unlimited. According 
 to v. Noorden, when more than 180 to 250 grms. of grape sugar 
 are taken, even in health, temporary glycosuria is produced. The 
 figures indicate the average quantities only, and in different 
 healthy individuals the limit varies somewhat, v. Noorden ( x ) 
 found that in four out of fifteen obese persons temporary glyco- 
 suria was produced by the administration of only 100 grms. of 
 pure grape sugar. Of these four persons, two became diabetic 
 some years later, and in the other two the lapse of time has 
 been too short to exclude the possibility of diabetes developing, 
 v. Noorden suggests that a large quantity of grape sugar should 
 be given, by way of experiment, in the case of persons who 
 have a family history of diabetes, or who are obese or gouty, 
 in order to ascertain how much sugar they can assimilate. If 
 a temporary glycosuria be produced by an amount of sugar which 
 is much below the average above stated, then these individuals 
 may be regarded as suspicious persons with regard to the possi- 
 bility of the development of diabetes, and they ought to observe 
 great caution with respect to amount of carbohydrates (especially 
 sugar) in their diet. v. Noorden would limit the amount of beer 
 
MANAGEMENT OF A CASE. 321 
 
 also in these cases. In this way it is possible that the develop- 
 ment of the disease may be prevented in some obese and gouty 
 persons, and in persons who have a family history of diabetes. 
 
 Management of a Case of Diabetes. 
 
 After a diagnosis of diabetes has been made, it is important 
 to note carefully a number of points before commencing the 
 treatment. The general condition of the patient should be care- 
 fully observed — whether he is well nourished or badly nourished. 
 The patient's weight should be ascertained, and whilst he is 
 under treatment he should be weighed once a week. It is 
 important to keep a sharp outlook on these two points. Even 
 though the sugar excretion and other symptoms should be 
 diminished by any form of treatment, this must be regarded as 
 unsatisfactory if the patient at the same time loses weight, and if 
 the general condition deteriorates. Hence a weekly or monthly 
 record of the patient's weight is perhaps more important in 
 severe cases of diabetes than the daily record of the amount of 
 sugar excreted. 
 
 Inquiries ought to be made with respect to the condition of 
 the bowels. This is especially necessary in the case of patients 
 just admitted to the wards of a hospital, since the confinement 
 to bed and the altered condition of life are alone sufficient in 
 many cases to produce obstinate constipation ; and if the bowels 
 have been previously constipated for a long period, the onset of 
 diabetic coma will be thereby favoured. Hence, means should 
 be taken to relieve severe constipation, if this symptom be 
 present. 
 
 A routine examination of the patient should be made, and 
 complications should be noted. The lungs should be carefully 
 examined, especially in wasted patients, for signs of phthisis, 
 since this is one of the two most serious complications which 
 may arise. Diabetics, especially hospital patients, are very 
 liable to develop tubercular disease ; hence precautions should 
 be taken, particularly if wasting be present, to avoid the risks of 
 tubercular infection, and to prevent the development of this 
 complication. 
 
 It is important to ascertain whether the urine gives the 
 dark brownish red coloration with perchloride of iron (Gerhardt's 
 reaction), .since this reaction points to a severe form of the 
 
3 2 2 THE TREA TMENT OF DIABE TES. 
 
 disease, and is an indication that a rigid diet should not be 
 prescribed. 
 
 In hospital patients the amount of sugar in the urine should 
 be estimated daily, and this is done most readily by the fermenta- 
 tion test. A chart should be kept, showing in a tabular form 
 the daily quantity of fluid taken, the amount of urine, the 
 specific gravity of non-fermented and of fermented urine, the 
 total daily excretion of sugar, the weight of the patient, and 
 the treatment. In private practice, the amount of sugar can 
 only be estimated periodically, as a rule. It is also of advan- 
 tage to estimate from time to time the amount of urea excreted, 
 in order to obtain an indication of the nitrogenous waste. 
 
 After admission to the hospital, the patient should be kept 
 on a diet which is not much restricted for a few days before 
 any observations are commenced with respect to the effect 
 of any special diet or treatment. Very frequently there is 
 a considerable diminution of the sugar excretion during the 
 first three or four days after admission to the hospital, apart 
 from any special form of treatment or restriction of diet. Hence 
 it is well to wait until this equilibrium has been attained, before 
 commencing any observations as to the effect of treatment. To 
 this general rule there are exceptions, however. The patient's 
 condition may make it necessary to restrict the diet at once, 
 in certain cases, when the treatment has been greatly neglected 
 before admission to the hospital, or when complications have 
 arisen which demand prompt anti-diabetic diet. Then, again, if 
 the patient has been on a very rigid diet at home, if an ordinary 
 diet be allowed suddenly, the rapid change might lead to serious 
 symptoms in advanced cases of diabetes. Diabetic coma has 
 been known to follow such a sudden change in diet. 
 
 The figures respecting the sugar excretion for the first two 
 or three days ought not to be taken into consideration in draw- 
 ing any conclusions as to the effects of drugs. 
 
 The diet should, then, be restricted, but it is important to 
 make the change gradually. Caution is especially necessary 
 when the urine gives a deep red coloration with perchloride of 
 iron, and in these cases it is also important that the diet should 
 not be rigid for a long period. 
 
 Potatoes should be excluded from the diet first, then bread, 
 and gradually all carbohydrates should be cut off. In the 
 place of bread, the aleuronat and cocoa-nut cakes mentioned 
 
GENERAL PRINCIPLES OF TREATMENT. 323 
 
 on p. 357 may be given. It is then noted whether the sugar 
 has disappeared from the urine ; if so, the patient is suffering 
 from a mild form of the disease. If sugar be still present, the 
 case belongs to the severe form of diabetes. 
 
 If glycosuria disappears when the patient is on a rigid diet, 
 then in course of time a little bread should be allowed, and the 
 amount gradually increased. It should be carefully observed 
 how much carbohydrate can be allowed before sugar reappears 
 in the urine. This amount is an indication of the quantity 
 of carbohydrates the patient can tolerate. 
 
 If exclusion of carbohydrates does not cause the glycosuria 
 to disappear, a very rigid diet must not be maintained, especially 
 if the urine give the dark brownish red coloration with per- 
 chloride of iron. 
 
 The effects of various drugs or other forms of treatment may 
 be tried with or without restriction of diet. 
 
 The urine ought to be examined frequently for albumin and 
 casts, in advanced cases of the severe form of the disease, as the 
 appearance of the latter is often, but not always, a forerunner of 
 diabetic coma. 
 
 GrENERAL PRINCIPLES OF TREATMENT. 
 
 There is no routine treatment for diabetes. Each case must 
 be separately considered, and the treatment varied according to 
 the form of the disease. 
 
 The chief objects to be aimed at in the treatment are — 
 
 1. To diminish the amount of sugar in the urine and 
 blood ; to diminish the amount of urine, and to relieve thirst. 
 
 2. To increase the weight, or at least to maintain it (except 
 in those mild cases in which the disease is accompanied by 
 obesity), and to improve the general condition. 
 
 3. To treat the various complications. 
 
 In mild cases of diabetes, the first object may be attained 
 by removal of the carbohydrates from the diet. 
 
 Leo ( 2 ) has pointed out that in the majority of severe cases of 
 diabetes there is a great increase of the nitrogenous metabolism, 
 and, as a result, a great loss of strength. The aim of treatment 
 in these cases ought to be to diminish not only the excretion of 
 sugar, but also the nitrogenous metabolism. 
 
 Careful observations have shown that sometimes improve- 
 ment occurs, and the patient g;iins in weight, in severe eases of 
 
324 THE TREATMENT OF DIABETES. 
 
 diabetes, when a little carbohydrate food is allowed, even though 
 the amount of sugar may be increased in the urine. If the 
 patient be losing weight, and the general condition be deteriorat- 
 ing;, then a diminution of sugar is of no avail. 
 
 In the slight cases, and in cases of medium severity, the 
 diminution or disappearance of sugar is a most favourable sign ; 
 but in the advanced and severe cases, accompanied by great 
 wasting, I have not infrequently found that a diminution of 
 sugar has been associated with an unfavourable course of the 
 disease, and has been accompanied by loss of flesh and deteriora- 
 tion of the general condition. The sugar has become less and 
 less, and has occasionally disappeared shortly before the fatal issue. 
 
 It ought to be our aim to make thin diabetics stout, as in 
 patients who are well nourished the prognosis is always much 
 better. 
 
 Those who are engaged in general practice amongst people 
 in good circumstances are apt to look upon the treatment 
 of diabetes as often fairly satisfactory ; whilst those who see 
 much of the disease in the hospitals of our large towns, or of 
 diabetes amongst the poor, will have been struck by the want of 
 success in treatment. 
 
 Much depends on the form of the disease. The results of 
 treatment in the mild form are very good, whilst in the 
 severe form they are often very unsatisfactory. In the former, 
 by treatment, the sugar may be removed from the urine, and 
 the symptoms disappear, so long as the patient follows the 
 treatment prescribed ; or, if the sugar does not entirely dis- 
 appear, it is diminished greatly, general improvement occurs, and 
 very frequently treatment enables the patient to live with a fair 
 amount of comfort and security. Life may be sometimes pro- 
 longed for ten or twenty years, or up to the natural limit. 
 
 The earlier the treatment is commenced the better ; it is 
 often very unsuccessful amongst the poor, because no attention 
 is paid to the disease until a late stage. Moreover, there are 
 very acute cases which sink rapidly, in spite of all treatment. 
 All these points ought to be borne in mind, in considering 
 the effects of treatment in any particular instance. 
 
 Dietetic Treatment. 
 It is well known that in certain parts of the world man 
 
DIETETIC TREATMENT 325 
 
 lives in perfect health on a diet from which carbohydrates 
 are almost absent. A mixed diet is no doubt most suited 
 for sustaining human life under the conditions of modern 
 civilisation, but a review of the dietetic customs in different 
 parts of the world (such as that given by Dr. Pavy in his 
 " Treatise on Food and Dietetics," London, 1874, p. 427), 
 shows that many races live almost entirely on food derived 
 from the animal kingdom. Thus in the Arctic regions the 
 Esquimaux live chiefly on animal food — on flesh and large 
 quantities of fat. The Pampas Indians, who pass the greater 
 part of their lives on horseback, are said to live entirely on 
 animal food — " the flesh of their mares " — and to have neither 
 bread, fruit, nor vegetables. The Guachos, the inhabitants of 
 the Pampas in the Argentine Ptepublic, who also spend the 
 greater part of their time on horseback, are said " to live 
 entirely on roast beef, with a little salt, scarcely ever tasting 
 farinaceous or other vegetable food," and their sole beverage 
 is stated to be mate taken without sugar. 
 
 Ever since Eollo published his book on diabetes in 1797, 
 and pointed out the value of restriction of the carbohydrates 
 in the food, it has been acknowledged that of all forms and 
 methods of treatment this dietetic one is the most important. 
 
 In undertaking the dietetic treatment of any case of 
 diabetes, it is necessary to give direction (1) as to the nature 
 of the food, (2) as to the quantity. But there is no hard-and- 
 fast rule on these points, and each case must be treated accord- 
 ing to the form of the disease to which it belongs. 
 
 What has already been pointed out with respect to the 
 varieties of the disease may be repeated, since different dietetic 
 treatment is required in each case. Diabetic cases may be 
 divided into two chief forms, mild and severe ; of each of 
 these forms Minkowski ( 3 ) recognises two varieties. 
 
 1. The mild form occurs chiefly in persons past middle 
 life — especially in obese or gouty persons. The removal of 
 carbohydrates from the diet is followed by the disappearance of 
 glycosuria. Food increases the sugar excretion, which is there- 
 fore less during the night than during the day. 
 
 Varieties — {a) Slight cases in which limitation of the 
 carbohydrates (without complete withdrawal) is sufficient to 
 remove the -glycosuria. 
 
 (b) More severe cases, in which complete withdrawal of 
 
326 THE TREATMENT OF DIABETES. 
 
 carbohydrates from the diet is necessary before the glycosuria 
 disappears. 
 
 2. The severe form is met with more frequently in persons 
 under the age of 45, and often amongst the poor and 
 hard-working people, and amongst hospital patients in large 
 towns. The withdrawal of carbohydrates from the food does not 
 cause the glycosuria to disappear. Wasting is a marked symptom, 
 and the urine often gives a dark reddish brown coloration 
 with perchloride of iron. As pointed out by Schmitz, the 
 urine passed during the day may contain less sugar than that 
 passed during the night. 
 
 Varieties — (c) In the milder cases of the severe form, sugar 
 is still excreted in the urine when the diet is free from 
 carbohydrates, but contains a very large quantity of albumin ; 
 if, however, the albumin in the food be restricted, the carbo- 
 hydrates being withdrawn, the glycosuria sometimes disappears. 
 
 (d) In the most severe form, neither withdrawal of carbo- 
 hydrates from the diet, nor restriction of the amount of albumin, 
 suffices to remove the sugar from the urine. Hence, in these 
 cases, sugar must be formed from the albumins of the 
 organism. 
 
 Many cases, even when first seen, belong to the severe form 
 of the disease ; sometimes, however, cases which at first present 
 the characteristic features of the mild form, at a later date pass 
 into the severe form. By good dietetic treatment, occasionally 
 a severe case may be transformed into a mild case ; frequently, 
 however, the case remains mild or severe from first to last. 
 
 The following are the indications for dietetic treatment in 
 the various forms of the disease ( 4 ) : — 
 
 1. In the mild, forms it is generally acknowledged that by 
 a restricted diet. i.e. by the withdrawal of carbohydrates from 
 the food, the greatest benefit is derived. 
 
 In the mildest variety of the mild form (form a, in which 
 limitation in the amount of carbohydrates in the food is 
 sufficient to remove sugar from the urine), the carbohydrates 
 should be withdrawn from the diet for a few weeks. Then 
 a little bread may be allowed, and if no sugar is excreted in 
 the urine a little more carbohydrate food may be given after 
 two weeks. It is often found that in time the patient will 
 be able to take a fair amount of carbohydrates without 
 glycosuria occurring, but even in these mild cases it is best 
 
DIETETIC TREATMENT. 327 
 
 to avoid excess of carbohydrates. Sugar and articles containing 
 sugar ought to be permanently excluded from the diet ; and 
 potatoes, bread, and beer ought to be limited. 
 
 In the more severe cases of the mild form (group b), 
 carbohydrates should be excluded from the diet for six or 
 eight weeks. Sometimes it is then found that a little carbo- 
 hydrate food can be tolerated, i.e. that the case has been 
 converted into one of the variety (a). But in very many cases 
 the urine can only be kept free from sugar by a diet from 
 which carbohydrates have been completely withdrawn. Such 
 a rigid diet cannot as a rule be continued for a very long 
 period, and hence, after a time, a small amount of carbohydrates 
 must be allowed, and we must be content if by a fairly restricted 
 diet we can limit the sugar excretion to 500 or 600 grs. 
 daily. 
 
 Whilst some authors hold that the diet ought to be as 
 rigid as possible, others believe that even in the mild form 
 patients do better on a moderately restricted but not absolutely 
 rigid diet. A rigid or fairly restricted diet may be allowed 
 in mild forms of the disease in stout persons, so long as the 
 patient feels well on such a diet, aud so long as definite wasting, 
 albuminuria, acetonuria, and diacetonuria (as indicated by the 
 ferric perchloricle reaction) are absent. 
 
 In the mild form of the disease (as in all forms), fatty 
 food is of great value as a substitute for bread, when the patient 
 is placed temporarily on a rigid diet for diagnostic purposes, 
 or when a rigid diet is prescribed as a means of treatment. 
 Cakes of aleuronat and cocoa-nut (see p. 357) will be found 
 very useful. 
 
 In the mild forms of diabetes, if the patient be stout, a 
 reduction of the total quantity of food, and especially a 
 reduction of the amount of nitrogenous food, is often of great 
 service ; but if the patient be wasted, no reduction of the 
 amount of food should be attempted. 
 
 Life can be prolonged in many mild cases of diabetes 
 to an advanced age, by careful dieting. Cantani states that 
 lie has cured many patients (in Italy) by absolute flesh and 
 fatty diet; but in England and Germany a cure of diabetes 
 in the true sense is certainly very rare, for if the glycosuria 
 disappears when the patient is on ;i rigid diet, it returns when 
 carbohydrates are taken (see p. 308). Occasionally glycosuria 
 
3 2 8 THE THE A TMENT OF DIABE TES. 
 
 which has persisted for some time disappears, but frequently 
 it returns at a later date. 
 
 2. In the severe forms of the disease the opinion has 
 gradually been gaining ground for some years, that a very 
 strict diet {i.e. complete withdrawal of the carbohydrates) is 
 injurious. In fact, it might almost be said that the milder 
 the form of the disease, the more rigid the diet indicated, 
 as regards withdrawal of carbohydrates ; and the more severe 
 the form, the more injurious a rigid diet will be. 
 
 As already mentioned, Leo ( 2 ) has pointed out that in the 
 majority of severe cases of diabetes there is a great increase 
 of the nitrogenous metabolism, and, as a result, great loss of 
 strength.- Marked wasting is a prominent symptom, and the 
 aim of treatment ought .to be to diminish not only the excretion 
 of sugar, but also the nitrogenous waste. 
 
 Fatty food is said to diminish the nitrogenous metabolism 
 (though this has been recently disputed by Dunlop), and Leo 
 has shown that carbohydrates have a similar action {i.e. albumin- 
 sparing influence). Leo has shown that the nitrogenous 
 excretion in the urine and faeces was less in two cases of 
 diabetes when the patients were taking a diet rich in albumin, 
 together with a certain amount of carbohydrate food, than when 
 they were taking the same diet without the carbohydrates. 
 Careful observations have shown that in severe cases of diabetes 
 improvement may occur, and the patient gain weight, when a 
 moderate quantity of carbohydrates is allowed, even though 
 the amount of sugar in the urine may be increased. 
 
 Hence, in these cases, not only should the sugar excretion 
 be estimated daily, but also the general condition of the patient, 
 and the weight, and the urea excretion should be carefully 
 watched. A diet rich in albumin and deficient in carbohydrates 
 is prescribed frequently to reduce obesity. A similar diet would 
 therefore appear little suited for a weak, emaciated, diabetic 
 patient. Then again, it is important to remember that in 
 many of these severe cases phthisis develops, and death may 
 occur from this complication. Hence it is necessary to prescribe 
 a diet on which the patient can maintain his weight, even if the 
 sugar excretion be somewhat increased thereby. On this account, 
 also, a rigid diet is unsuitable for these severe cases. 
 
 When the urine of a diabetic patient gives the dark 
 brownish red coloration (Gerhardt's reaction) with perchloride 
 
DIETETIC TREATMENT 329 
 
 of iron, the withdrawal of all carbohydrates from the diet is 
 especially dangerous. This has been pointed out long ago by 
 Ebstein ( 4 ) and many other writers. 
 
 When this reaction is obtained, there is even great danger 
 in suddenly placing the patient on a rigid diet temporarily, 
 for the purpose of ascertaining whether he is suffering, from 
 a mild or a severe form of the disease. Such a sudden change 
 might lead to diabetic coma. Hence, if it is desired to ascertain 
 the effect of a rigid diet on the amount of sugar excreted, the 
 restriction should be effected very gradually, and only continued 
 a very short time. 
 
 Hirschfeld ( 5 ) and others have shown (as previously men- 
 tioned) that acetonuria can be produced in healthy persons 
 by a diet consisting of albumin and fat only ; but the addition 
 of carbohydrates to the food causes it to disappear. If carbo- 
 hydrates are no longer burnt up in the system (as occurs in 
 diabetes), acetone is often found in the urine, in spite of the 
 carbohydrates in the food ; and when acetone is present in large 
 quantities, diacetic acid is also found. But, in the severe forms, 
 when the diet has been rigid for a long period, if the carbo- 
 hydrates of the food are increased, or if glycerin is added, the 
 acetone and diacetic acid in the urine diminish. 
 
 When the urine gives a marked reaction with perchloride 
 of iron, there appears to be good evidence that the rigid ex- 
 clusion of carbohydrates from the food has often a tendency to 
 bring on diabetic coma (Ebstein, 4 Hirschfeld, 5 Grube, 6 Schmitz 7 ). 
 Considerable evidence has been furnished that these cases often 
 improve, and the threatening diabetic coma is arrested, by 
 adding starchy food, potatoes, and bread in small quantities to 
 the diet. Schmitz has also drawn attention to the value of a 
 moderate amount of starch-containing foods in the most severe 
 form of diabetes ; he regards their complete withdrawal as 
 injurious. The quantity he would allow is never great, however, 
 and he regards saccharine carbohydrates as always decidedly 
 injurious. 
 
 Ebstein lias pointed out that the appearance of acetone 
 and diacetic acid in the urine is an indication for diminishing 
 the albumin, as well as for increasing the carbohydrates in the 
 diet. 
 
 Minkowski states that a rigid diet increases the formation 
 of organic acids in the system, and thus predisposes to coma. 
 
S3o THE TREATMENT OF DIABETES. 
 
 Schmitz ( 7 ) has drawn attention to the danger of a very large 
 amount of animal food in diabetes, since it passes into the 
 intestine undigested ; putrefaction occurs ; toxic substances are 
 formed and absorbed into the blood; and he believes that the 
 absorption of these toxic substances leads to the development of 
 the coma. But Seegen ( s ) and others oppose this view. 
 
 All authors agree as to the great value of fat in the severe 
 forms of diabetes ; and, in addition to fatty food, cod-liver oil, 
 lipanin, and other fats may be given. 
 
 In the severe forms of diabetes no attempt should be made 
 to restrict the total amount of food. 
 
 In certain cases of diabetes, for diagnostic purposes, the 
 effect of temporary and gradual withdrawal of carbohydrates from 
 the food, together with the reduction of the amount of nitrogenous 
 food, may be tried. If a diet restricted in loth these respects 
 causes the sugar to disappear, then the cases belong to the 
 variety (c), p. 326. But even in these cases a diet so restricted 
 is not advisable for any length of time. Any signs of increasing 
 weakness or loss of weight would be an indication to increase 
 both carbohydrates and nitrogenous food at once. 
 
 Briefly, then, in diabetes of the severe form, fatty food should 
 be given in large quantities, and a little alcohol taken to aid the 
 digestion thereof; nitrogenous food should only be given in 
 moderate quantities. Of the carbohydrates, saccharine food 
 should be excluded ; a certain quantity of starchy food, however, 
 is of advantage in the form of bread, but of course a great 
 quantity ought not to be allowed. Cream is of great service, and 
 milk in moderate quantities may be allowed. 
 
 Aeticles of Diet. 
 
 1. animal and nitrogenous food. 
 
 Most articles of food from the animal kingdom may be taken 
 freely by diabetic patients. Butchers' meat and flesh meat of 
 various kinds, poultry, game, and fish, may be taken in any form. 
 But in the cooking thereof flour or bread crumbs should not be 
 used, if a very strict diet is indicated ; aleuronat may be used 
 in place of flour, however (see p. 355), and butter and fats may 
 be freely employed. 
 
 The following articles may also be allowed freely : — Tongue, 
 
A NIMAL AND NITR O GENO US FO OD. 331 
 
 ham, bacon, potted beef, and chicken, preserved meats of various 
 kinds, sardines, tinned and preserved fish, beef-extracts, beef-tea, 
 meat juices, broth, soups, and jellies (when prepared without the 
 addition of any saccharine or starchy materials). 
 
 Eggs in various forms are most useful articles of diet. An 
 egg weighing about 600 grs. contains about 90 grs. of albumin 
 and 75 grs. of fatty material. When hard boiled they can 
 be taken with comparatively large quantities of butter. The 
 " buttered " egg or omelette is a useful form. 
 
 Custard made of eggs and milk, and sweetened with 
 saccharine or saxin, may be taken by diabetics ; it is a useful 
 substitute for rice and farinaceous puddings. 
 
 A kind of light fluid custard (commonly called " Samson " in 
 some parts of the north of England) is very suitable for diabetic 
 persons : — 
 
 Take one egg, beat up well ; make a mixture of cream and water, and 
 boil ; gradually add the boiled cream and water (whilst hot) to the egg, 
 stirring the mixture with a. spoon. Then place the mixture in a pan 
 over the fire, and continue to stir with a spoon until it becomes thick ; 
 then pour into a glass. It is important that the mixture should not 
 be heated too much (i.e. should not be boiled) as the albumin would 
 be coagulated thereby. Flavour with cinnamon, and sweeten with sac- 
 charine or saxin if desired. 
 
 Eggs are also useful in the preparation of bread, pancakes, 
 and various articles of diet for diabetic patients, as will be 
 pointed out in another section. 
 
 Liver, which contains a considerable amount of sugar, and 
 of substances capable of conversion into sugar in the alimentary 
 canal, ought to be avoided, but other organs, such as brain, 
 kidney, pancreas, thymus, lung, are harmless. 
 
 Oysters, cockles, mussels, and other mollnsca are to be 
 avoided, on account of the great amount of glycogen contained 
 in their large livers. The interior of crabs and lobsters are also 
 unsuitable articles of diet. Honey, of course, ought never to 
 be taken. Butter, cheese, and cream may be allowed in large 
 quantities. 
 
 With respect to milk, considerable caution is necessary. 
 This article of diet, so useful in health and in disease, contains 
 4-4'82 per cent, of sugar in the form of lactose, which for the 
 diabetic, of course, is an objectionable constituent. But milk 
 
33 2 
 
 THE TREATMENT OF DIABETES. 
 
 . 3*5 per 
 
 cent 
 
 . 3-98 
 
 >) 
 
 . 0-77 
 
 )j 
 
 . 4-00 
 
 ii 
 
 . 86-87 
 
 ii 
 
 also contains large quantities of fatty and albuminous materials, 
 which are most desirable. 
 
 According to Blyth ( 9 ), the following is the composition of 
 cow's milk : — 
 
 Milk fat . 
 \ Casein 
 (Albumin 
 Milk sugar 
 Water . 
 
 Experiments made by various observers have shown that, in 
 some cases of diabetes, when milk is added to the diet in con- 
 siderable quantity, the amount of sugar in the urine is not 
 increased. The lactose of the milk is therefore utilised in 
 the system. In other cases, however, the glycosuria has been 
 increased by this addition of milk to the diet ; but in these 
 cases the patient has nevertheless sometimes gained weight and 
 improved in general condition. 
 
 In the milder forms of diabetes, if the addition of milk to 
 the diet does not increase the glycosuria, it may be allowed ; 
 but when the glycosuria is ■ increased thereby, milk is better 
 avoided. When large quantities of milk have been employed in 
 the treatment, as in the so-called " milk cure," often bad results 
 have followed. If, however, the digestion is feeble, and other 
 foods are digested with difficulty, milk may be generally allowed. 
 In the most severe forms of diabetes, when a very rigid diet 
 is not indicated, milk may also be allowed in moderate quantities ; 
 indeed, some physicians recommend large quantities (3h pints) 
 daily in these cases. 
 
 Skimmed milk contains less albumin and fat, and a greater 
 percentage of carbohydrates, than ordinary milk. It has been 
 recommended, however, by Dr. Donkin. In his method of 
 treating diabetes he prescribes six to eight pints of skimmed 
 milk daily ; but most physicians who have tried this method 
 of treatment record bad results. 
 
 If a simple and easy method could be found for removing 
 the lactose, milk would then be a most valuable article of diet 
 for all diabetic patients, on account of the great percentage 
 of albumin and fat which it contains. 
 
 Wright ( 10 ) has suggested a method for making an artificial 
 milk for diabetic patients. A quantity of milk is diluted with 
 
Magnesium citrate 
 
 4-4 
 
 Dicalcium phosphate 
 
 8-0 
 
 Tricalcium phosphate 
 
 9-6 
 
 Calcium citrate 
 
 25-5 
 
 Calcium oxide 
 
 5-5 
 
 Sodium carbonate 
 
 40-0 
 
 ANIMAL AND NITROGENOUS FOOD. 333 
 
 three or four volumes of water, to which 1 to 2 parts per 
 1000 of acetic acid have been added (3iss— 3iii of acid acet. 
 fort, of the B.P. to a pint of water). This produces a precipi- 
 tation of all the casein and fat of the milk. The precipitate 
 is allowed to settle for a few minutes, and then strained through 
 a piece of calico. It is washed and redissolved in a 1 per cent, 
 solution of the following mixture : — 
 
 Sodium chloride . 11*5 parts, j Dimagnesium phosphate 4*0 parts. 
 Potassium chloride . 9 - 9 „ 
 Monopotassium phos- 
 phate . . . 13 - 8 ,, 
 Dipotassium phosphate 10 - ,, 
 Citrate of potassium . 5 - 9 ,, 
 
 A trace of saccharine may be used to sweeten the milk. 
 The salt solution is best used at about blood temperature, and 
 the casein and fat precipitate is to be mixed up with it, as in 
 making cocoa, to the desired thickness. By this method a fairly 
 palatable and entirely sugarless milk is obtained. 
 
 Einger ( xl ) has also described a method of preparing milk for 
 diabetic patients : — Add to a pint and a half of milk about 90 
 c.c. of a 10 per cent, solution of acetic acid. This precipitates 
 a curd caseinogen. It should be allowed to settle, and the clear 
 fluid siphoned off and distilled water added. After standing for 
 a time, the water should be decantered or siphoned off, and the 
 curd should be filtered and well washed with distilled water. 
 It is then rubbed up in a mortar with some calcium carbonate, 
 and water added ; all the caseinogen becomes dissolved, the 
 calcium carbonate soon settles, and the milky fluid can be 
 decanted. The dissolved caseinogen forms a substitute for milk. 
 
 (Martindale and Lee advise that the milk be diluted with 
 an equal quantity of water, and that the filtration and washing 
 of the caseinogen should be performed on a calico filter instead 
 of filter paper.) On the addition of 2 per cent of glycerin 
 to the mixture of caseinogen, a not unpalatable form of milk 
 is produced. Another plan proposed by Einger is the 
 following : — 
 
 " Milk is heated in a hot-water bath, then poured into 
 tubular parchment dialysing paper, and the whole placed in a 
 ves el filled with tap water; the water being kept running by 
 
334 THE TREATMENT OF DIABETES. 
 
 means of a tube connected with a tap, the tube ending in a piece 
 of glass tubing, long enough to reach the bottom of the vessel, so 
 that it is supplied from the bottom." In three days the milk 
 sugar is reduced to one-sixth of the original quantity. 
 
 Both these methods can be carried out easily in the labora- 
 tory, but for the average private patient, and especially for the 
 poorer classes of patients, they appear to be a little too com- 
 plicated. I have found that a very palatable drink, closely 
 resembling milk, can, however, be prepared from cream in the 
 simplest manner possible. Now cream contains milk sugar 
 (3 - 54 per cent., Konig; 2 - 8 per cent., Letheby) ; and though it 
 is allowed to diabetics, still, if taken in very large quantities, this 
 percentage of sugar would theoretically be objectionable, especially 
 when a strict diet is indicated. But cream contains also 25*72 
 per cent, of fat (Konig), which is of the greatest value to 
 the diabetic patient, and by the following exceedingly simple 
 method a palatable drink can be easily made. To about a pint 
 of water, placed in a large drinking-pot or tall vessel, three or 
 four tablespoonfuls of fresh cream are added and well mixed. 
 The mixture is allowed to stand for twelve to twenty-four 
 hours, when most of the fatty matter of the cream floats to the 
 top ; it can be skimmed off with a teaspoon easily, and on 
 examination it will be found practically free from sugar. This 
 fatty matter thus separated is placed in a glass and mixed with 
 water. Then the white of an egg is added, and the mixture well 
 stirred. The water and white of the egg are added in sufficient 
 quantities to make a mixture which has the exact colour and 
 consistence of ordinary milk. If a little salt and a trace of 
 saccharine be added, a palatable drink, practically free from 
 milk sugar, is produced, which has almost the same taste as 
 milk, and which contains a large amount of fatty material. 
 With very little practice the right proportions can be easily 
 guessed, and of course much larger quantities can be employed 
 than those mentioned above. A drink may thus be prepared, 
 which contains a large amount of fat, and which can be taken 
 ad libitum by diabetic patients, even if the most rigid diet be 
 deemed necessary. Devonshire cream, which has the consistence 
 of a soft paste, contains only 1'72 per cent, of milk sugar, and 
 hence may be used with advantage, in the place of ordinary 
 cream, in preparing milk as above described. 
 
 Cream contains much less lactose than milk, but seven times 
 
ANIMAL AND NITROGENOUS FOOD. 
 
 335 
 
 the amount of fat. Owing to this large percentage of fat in a 
 form easily digested, cream is a most valuable article of diet for 
 diabetic patients. It may be taken in considerable quantities with 
 coffee or tea, or in other ways. With a little care, three-fourths 
 of a pint of cream can be taken daily by a diabetic patient, and 
 in the severe forms of the disease it is most useful. The follow- 
 ing table shows the composition of cream and milk : — 
 
 
 Milk. 
 
 Cream. 
 
 Kbnig. 
 
 Blyth. 
 
 Konig. 
 
 Pavy. 
 
 Albumin .... 
 Fat 
 
 Milk sugar • . 
 
 87-40 
 3-40 
 3-66 
 4-82 
 
 86-87 
 4-75 
 3-50 
 4-00 
 
 66-40 
 3-70 
 
 25-72 
 3-54 
 
 66-0 
 
 2-7 
 
 26-7 
 
 2-8 
 
 Devonshire cream is a white, soft paste, and " is produced by 
 keeping the milk in large pans at a gentle heat for many hours. 
 The temperature is always far under the boiling point. This 
 application of a moderate heat during a lengthened time causes 
 the fat to coalesce and rise more rapidly than the ordinary 
 method." 
 
 Blyth ( 12 ) gives the following composition of Devonshire 
 cream : — Per cent 
 
 Milk fat 
 
 Casein 
 
 U O \J 1 1 
 
 . 3-530 
 
 Albumin . 
 
 0-521 
 
 Peptones . 
 
 0-050 
 
 Milk sugar . 
 
 . 1-723 
 
 Water 
 
 . 28-675 
 
 Ash .... 
 
 . . . 0-490 
 
 Owing to the small percentage of lactose, Devonshire cream 
 is particularly suitable for diabetic patients. 
 
 Batter-milk contains less sugar than ordinary milk, part of 
 the lactose having been decomposed. It still contains 3 -38 per- 
 cent, of lactose, however, along with lactic acid. It is therefore 
 a little more suitable than ordinary milk, but the taste is objec- 
 tionable to many patients. 
 
 Koumiss. — Lactose does not ferment under the influence of 
 the ordinary yeast fungus, but certain of the schizomycetes have 
 
'id 
 
 THE TREATMENT OF DIABETES. 
 
 the power of decomposing it, alcohol and lactic acid being 
 formed. The Tartar drink, koumiss, is fermented mare's milk ; 
 it can also be manufactured from cow's milk. Koumiss contains 
 less lactose than ordinary milk, and is in this respect more 
 suited for the diabetic patient. The Aylesbury Dairy Co. have 
 kindly supplied me with samples of their diabetic koumiss, and 
 a rough examination shows clearly that it contains very much 
 less lactose than ordinary milk. The following is the analysis- 
 furnished by the Aylesbury Co. : — 
 
 
 Koumiss. 
 
 One Day old. 
 
 Eight Days old. 
 
 Twenty-two 
 Days old. 
 
 
 92-24 
 
 92-38 
 
 92-55 
 
 
 •28 
 
 •35 
 
 •57 
 
 Fat 
 
 •51 
 
 •52 
 
 •51 
 
 
 2-19 
 
 2-13 
 
 2-05 
 
 
 •30 
 
 •25 
 
 •18 
 
 
 •36 
 
 •48 
 
 •65 
 
 Lactic acid ..... 
 
 •75 
 
 •86 
 
 1*22 
 
 
 2-7S 
 
 2-42 
 
 1-64 
 
 Ash 
 
 •59 
 
 •61 
 
 •63 
 
 In their " diabetic koumiss with glycerin," 10 per cent, of 
 glycerin is added. 
 
 The great objection to koumiss is the taste, which many 
 patients find disagreeable. 
 
 Koumiss has been recommended on theoretical grounds by 
 several writers, but I am not aware that it has been much used 
 in the treatment of diabetes in this country ; it appears worthy 
 of trial, however. 
 
 Stange ( 13 ) gives the table on p. 337, showing the composition 
 of koumiss after fermentation for varying periods. 
 
 Kephir, a Caucasian beverage, is really milk, in which the 
 lactose has been partly converted into alcohol, carbonic acid, and 
 lactic acid, by the action of a ferment ; the fatty elements are 
 almost unaltered, and the fluid is said to contain only about 
 1*5 per cent, of lactose. 
 
FATTY FOOD 
 
 337 
 
 
 Mare's 
 Milk. 
 
 Koumiss, after 
 
 Six Hours' 
 Fermentation. 
 
 Koumiss, after 
 Eighteen Hours' 
 Fermentation. 
 
 Koumiss, after 
 Thirty Hours' 
 Fermentation. 
 
 Koumiss, after 
 
 Four Days' 
 Fermentation. 
 
 Carbonic acid 
 
 
 3-8 
 
 6-0 
 
 7-0 
 
 11-0 
 
 Alcohol 
 
 
 18-5 
 
 19-5 
 
 30-0 
 
 30-0 
 
 Lactic acid . 
 
 
 3-9 
 
 5-6 
 
 6-4 
 
 6-4 
 
 Milk sugar . 
 
 51 
 
 18-8 
 
 16-3 
 
 
 
 Albumin 
 
 23 
 
 22-5 
 
 22-6 
 
 20-0 
 
 16-0 
 
 Fat . . 
 
 19 
 
 18-9 
 
 20-0 
 
 19-0 
 
 19-0 
 
 Salts . 
 
 5 
 
 4-5 
 
 4-0 
 
 4-0 
 
 4'0 
 
 The Caucasian ferment has been found to consist of a mix- 
 ture of yeast (Saccharomyces cerevisice) and a short bacillus, which 
 has been named Dispora Gaucasica. From this Caucasian ferment 
 Dr. Lehmann has prepared a more reliable kephir ferment. By 
 its action on milk, it is stated that home made kephir of an 
 agreeable taste can be easily prepared. 
 
 (The ferment can be obtained in small quantities, with 
 directions for use, from Dr. M. Lehmann, Berlin, C. Heiligegeist- 
 strasse, 43-44.) 
 
 2. FATTY FOOD. 
 
 Fats, both animal and vegetable, are the most valuable articles 
 of diet for diabetic patients, especially for those who suffer from 
 the severe forms of the disease, and may be allowed in large 
 quantities. 
 
 The more important articles of diet, containing a large 
 quantity of fat, are butter, cream, bacon, cheese, eggs, suet. 
 
 Butter contains 82 to 87 per cent, of fat; ordinary cream, 
 25 to 26 per cent. ; and Devonshire cream, 65 per cent. Cheese 
 contains from 10 to 30 per cent, of fat, and is also rich in 
 albumin. 
 
 Analysis of hitter. — Konig (quoted by Blyth) gives the 
 following as the mean of eighty-nine analyses : — 
 
 Mean. 
 Fat 83-11 
 
 Curd . 
 Ash . 
 Water 
 Milk sugar 
 
 •86 
 
 1-19 
 
 14-14 
 
 •70 
 
 22 
 
->8 
 
 THE TREATMENT OF DIABETES. 
 
 Analysis of several Varieties of Cheese (Blyth 14 ). 
 
 
 Water. 
 
 Ash. 
 
 Fat. 
 
 Nitrogen. 
 
 
 28-1 
 
 3-34 
 
 22-5 
 
 7-3 
 
 
 44-6 
 
 4-61 
 
 30-7 
 
 4-6 
 
 
 63-1 
 
 1-4 
 
 6-5 
 
 2-76 
 
 
 39-6 
 
 6-4 
 
 11-5 
 
 4-8 
 
 The yolks of eggs contain a large amount of fat. An egg 
 of ordinary size (600 grs.) contains 75 grs. of fat and 90 of 
 albumin. 
 
 Cod-liver oil is of considerable service, especially in the 
 severe forms of diabetes, and deserves to be more frequently 
 used when there is much wasting or loss of strength. Schmitz 
 speaks strongly in its favour. It is well to begin with a small 
 dose, 1 drachm daily, and to increase it up to | oz. once or 
 twice a-day after food. Lipanin may be employed, if the patient 
 objects very much to the taste of cod-liver oil. 
 
 Vegetable oils, salad oil, and olive oil may also be taken 
 freely with the food. Olive oil is largely used in the preparation 
 of food in Italy, but is not suited to the taste of the inhabitants 
 of Northern Europe. 
 
 Peach kernel oil (free from prussic acid) may be used in place 
 of cod-liver oil. Mr. Kirkby (of Messrs. Jewsbury & Brown, 
 Manchester) has prepared for me a very palatable emulsion of 
 peach kernel oil (50 per cent, by volume) with the yolk of egg, 
 saccharine elixir, spirits of chloroform, and essences of almond and 
 cinnamon. 
 
 Petroleum emulsion and pancreatic emulsion are also worthy 
 of a trial in cases accompanied by severe wasting. 
 
 Of all the fatty articles of food, butter and cream, in spite 
 of the small amount of milk sugar which both contain, are the 
 most suitable forms in which large quantities of fat can be 
 taken. 
 
 Ebstein records the case of a young diabetic lady, who took 
 225 grms. of butter daily, along with other fats and nitro- 
 genous food. The patient had become greatly wasted, but on 
 this diet the wasting disappeared, and she gained in weight ; the 
 sugar excretion was less, acetone and ace to-acetic acid dis- 
 
FATTY AND VEGETABLE FOODS. 339 
 
 appeared from the urine, and the general condition was 
 good. 
 
 Suet may be largely used in the preparation of puddings and 
 various articles of diet. 
 
 Fatty food is generally digested well by diabetics, and often 
 large quantities can be taken (half-a-pouncl or more) without 
 digestive disturbances occurring. Sometimes excess of fatty food 
 causes dyspepsia, but if a little alcohol (in the form of brandy or 
 non-saccharine wine) be taken after a diet rich in fat, digestion 
 is greatly aided ; and in this way the diabetic is enabled to take 
 large quantities of fatty food. The patient may be allowed to 
 take an unlimited amount of fat, so long as digestive troubles 
 are not produced thereby. 
 
 In the few cases, however, in which there is steatorrhoea 
 (probably owing to severe disease of the pancreas), large quantities 
 of fat cannot be digested. 
 
 If diarrhcea should be produced by a diet rich in fat, opium, 
 bismuth, and calcium carbonate will be found useful in the 
 treatment thereof. 
 
 d. VEGETABLE FOOD CARBOHYDRATES. 
 
 As already mentioned, the withdrawal or limitation of the 
 carbohydrates in the diet is the most important method of treat- 
 ment, and this is especially the case in the milder forms, The 
 caution necessary in the severe forms has been pointed out on 
 p. 328. Articles containing a large quantity of sugar must 
 always be avoided, and starchy food must be withdrawn, or 
 restricted, according to the nature of the case. The carbo- 
 hydrates are not equally injurious, however. 
 
 Starch is less injurious than sugar ; according to Schmitz ( 15 ) 
 sugar is ten times more injurious than starch ; and the with- 
 drawal of sugar is much better borne than the withdrawal of 
 starch, especially in the severe forms of the disease. 
 
 Inulin is an amylose which replaces starch in the roots of 
 many species of the composite. It occurs in dahlia tubers, in 
 the Jerusalem artichoke, and in several other roots ; and it has 
 been often recommended as a carbohydrate for diabetic patients. 
 Kiilz found that when inulin was taken by diabetic patients the 
 sugar excretion was not increased, and the inulin was there- 
 fore utilised in the system. He gives directions for the pre- 
 
34Q THE TREATMENT OF DIABETES. 
 
 paration of biscuit from inulin (see p. 360). Hale White 
 concludes, from his recent observations, that in moderate 
 amount, in the form of dahlia tubers, it does no harm to diabetic 
 patients. 
 
 Lichenin, a carbohydrate contained in Iceland moss, has been 
 recommended as a suitable carbohydrate in diabetes. In Ice- 
 land, bread is made from this moss, after the bitter principle has 
 been removed. 
 
 Of the sugar group of carbohydrates, glucose is the most 
 injurious, milk sugar and cane sugar rank next. Zcevulose is least 
 injurious. 
 
 in many cases milk sugar causes an increase in the sugar 
 excretion, but not to the same extent as does grape sugar. It 
 has been pointed out by Kulz, Voit, Minkowski, and others, 
 that Icevulosc and levorotatory carbohydrates may be utilised in 
 the system. Much, however, depends on the nature of the case 
 and the quantity of lsevulose taken. 
 
 Minkowski found that in the diabetes following pancreas 
 extirpation, the levorotatory carbohydrates were still burnt up in 
 the organism ; but he points out that after the use of lsevulose 
 in large quantities an increased excretion of dextrose may occur. 
 In human diabetes a small dose of lsevulose is almost completely 
 used up in the organism, and only a very slight quantity of this 
 sugar appears in the urine ; but a large quantity of lsevulose 
 causes a most decided increase of the glucose eliminated, the 
 greater portion of lsevulose at the same time passes off unchanged 
 by the urine. 
 
 Minkowski ( 16 ) has shown that — (1) Levorotatory carbo- 
 hydrates are partially destroyed in the system ; (2) partially 
 converted into dextrose and eliminated as such by the urine ; 
 (3) but when large quantities of lsevulose are given, a portion 
 passes off unchanged. 
 
 Haycraft ( 17 ), by experiments on animals and by observations 
 on three diabetic patients, has shown that — (1) In a case of 
 chronic diabetes in an old person, all the lsevulose taken (50 
 grms. daily) was utilised in the organism, and there was no 
 increase of the glucose excreted. (2) In some acute cases, a 
 part of the lsevulose taken with the food was excreted as such, a 
 part was utilised in the body, and a part was transformed into 
 glucose. (3) In rabbits, glycogen is formed from the lsevulose 
 taken, and is stored up in the liver. 
 
VEGETABLE FOOD. 341 
 
 Hale White ( 1S ) has made a number of observations on the 
 action of lsevulose. He found that if large amounts of lsevulose 
 were taken, some appeared in the urine. In none of the cases 
 did it have the pernicious effect (often produced by ordinary 
 carbohydrates) of increasing the output of sugar beyond the 
 extra quantity given. The excretion of sugar was usually 
 increased when lsevulose was given, but it was sometimes 
 diminished. In most of the cases, after giving lsevulose, much 
 less sugar was passed in the urine than would have been 
 excreted if the previous excretion of sugar had remained 
 stationary, and all the lsevulose had appeared in the urine. This 
 result seems to indicate that, in these cases, a portion of the 
 lsevulose taken was retained and destroyed in the body. 
 
 Some of Hale White's cases show that lsevulose can be 
 utilised better than dextrose. None of the patients felt worse 
 after taking lsevulose ; some felt better, and gained in weight. 
 
 Grube ( 19 ) concludes that lsevulose can be given in moderate 
 quantities, in the mild forms of diabetes, without any injurious 
 results as regards sugar excretion, state of urine, etc. These 
 patients appear to be able to utilise the lsevulose in the system, 
 though they are unable to utilise dextrose and cane sugar, which 
 are excreted in the urine. 
 
 Bohland ( 20 ) has obtained similar results. In mild cases he 
 found that lsevulose was utilised in the system ; in severe cases, 
 the sugar excretion in the urine was increased when lsevulose 
 was given. 
 
 Pure lsevulose is very expensive, too expensive for the 
 majority of diabetic patients, but most of the sugar in many 
 kinds of fruit consists chiefly of lsevulose. 
 
 Mannite or manna sugar is not a true sugar, but a hexad- 
 alcohol. It is the chief constituent of manna, and some fungi 
 contain mannite, but little carbohydrate material. It does not 
 increase the sugar excretion, and 30 grms. can be taken daily 
 by diabetics without producing any digestive disturbances, 
 but it has a slight purgative effect. It may be given with 
 coffee. 
 
 Pentaglucose, xylose, arbinosc, produce no bad effects in 
 diabetes, but they pass out of the body undecomposed 
 (Ebstein 2] ). 
 
 Lindemann and May ( 22 ) found that rhamnose, both in 
 healthy and in diabetic persons, was partly utilised in the 
 
342 THE TREATMENT OF DIABETES. 
 
 system, whilst a portion passed off in the urine. In health a 
 small percentage of rhamnose was excreted ; in diabetes they 
 found also a small proportion of rhamnose in the urine, along 
 with an increase of the grape sugar excretion. Ehamnose was 
 to some extent utilised in the system in diabetes, but not to 
 the same extent as in health. 
 
 Inositc (or muscle sugar) is really a benzol product, and 
 not a sugar. It is found in muscle tissue, is widely distributed 
 in the animal kingdom, and is also present in young French 
 beans. It is completely destroyed in the system in diabetes 
 (Hirschfeld). 
 
 Saccharine, a coal tar product, having a very sweet taste, may 
 be used in place of sugar to sweeten various articles of diet for 
 diabetic patients. 
 
 Saxin, 1 another coal tar product, has recently been recom- 
 mended for the same purpose. This substance is said to be six 
 hundred times as sweet as sugar. 
 
 Tablets of both substances are prepared ; a saxin tablet 
 the size of a hypodermic morphia tablet is equivalent to a lump 
 of sugar. The flavour of saxin is better than that of saccharine, 
 and is almost exactly the same as sugar. 
 
 The great problem with respect to the dietetic treatment 
 of diabetes is, to find a carbohydrate which is completely burnt 
 up in the system, and which can be readily procured at a 
 small cost. 
 
 Articles of Diet from the Vegetable Kingdom. 
 
 Fruits. — The most injurious constituent of fruit is sugar, 
 but in many kinds of fruit a large percentage of the sugar is 
 hevulose, which is utilised in small or moderate quantities in 
 certain cases of diabetes. 
 
 There is a considerable difference of opinion with respect to 
 the advisability of allowing a small quantity of fruit to diabetics. 
 Many physicians advise fruit to be withdrawn from the diet 
 entirely ; others would allow a small quantity in certain cases. 
 In any case of diabetes, if it has been shown that lasvulose can 
 be utilised in the system, i.e. that the glycosuria is not increased 
 by its use, then a small quantity of a fruit in which the sugar 
 is in the form of laevulose may be allowed. Otherwise, the 
 complete withdrawal of fruit would appear the more reasonable, 
 
FJR UJTS— VE GE TABLES. 3 43 
 
 especially in England, where fruit forms so small a portion of 
 the diet of most people. 
 
 Peaches and apricots are regarded as the least objectionable 
 kinds of fruit by Ebstein. Melons only contain 0*27 per cent, 
 of sugar. 
 
 The patient may be allowed stewed green gooseberries and 
 cream, sweetened with saxin, and a little soda-water may be 
 taken afterwards to neutralise the acidity of the gooseberries. 
 
 Dates, figs, currants, raisins, dried prunes and plums, and 
 other dried fruits which are rich in sugar, ought to be forbidden. 
 
 Nuts may be allowed, as a rule. Walnuts, almonds, filberts, 
 hazel nuts, and Brazil nuts may be freely allowed. Chestnuts 
 ought to be avoided, since they contain a large quantity of 
 carbohydrates, including 2 per cent, of sugar. 
 
 Pickled walnuts may be allowed freely. Walnuts may be 
 also boiled and used in the place of vegetables. 
 
 Place the walnuts (the shells having been removed) in boiling water, 
 and continue to boil for thirty minutes ; then drain away the water 
 carefully, place on a plate and sprinkle well with aleuronat flour (see 
 p. 355). Add salt, a little pepper, and butter also, if preferred. The 
 pan used should be enamelled or well lined, as an iron pan turns the 
 nuts black. 
 
 Cocoa-nuts contain a small quantity of sugar, but this may 
 be removed easily, and cocoa-nut flour may be used for the 
 preparation of biscuits (see p. 354). Almond flour is also 
 largely used for the preparation of cakes and biscuits. A bread 
 formed from pea-nut flour has recently been recommended. 
 These preparations from nuts will be referred to later. 
 
 Vegetables. — As a general rule, it may be stated that green 
 vegetables may be allowed, but white vegetables and root vege- 
 tables should be avoided. 
 
 The following may be allowed freely : — Salad (mustard and 
 cress), water-cress, the green parts of cabbage and lettuce, 
 Brussels sprouts, endives, broccoli, spinach, cucumber, mush- 
 rooms. Pickles — pickled cucumber, walnuts, and onions — may 
 also be allowed. A considerable amount of carbohydrates is 
 present in celery, onions, leeks, green French beans ; by some 
 they are forbidden, by others sanctioned, but when a very rigid 
 diet is indicated they are better avoided. 
 
 The following ought to be avoided in a rigid diet : — Potatoes, 
 
344 
 
 THE TREATMENT OF DIABETES. 
 
 rice, sago, tapioca, groats, arrow 7 root, macaroni, peas, beans, 
 turnips, carrots, parsnips, and beet-root. 
 
 Potatoes contain a less percentage of starch than the other 
 articles just mentioned (only 15 "5 per cent.). Sometimes the 
 craving for potatoes is great, and in certain cases they may be 
 allowed in small quantities, cooked in the form of potato chips. 
 A single potato may then " be made to fill almost a whole dish." 
 L. Brunton ( 23 ) gives the following directions : — 
 
 A large deep pan (not a frying-pan) 6 in. deep should be filled with 
 oil, or grease or dripping. This is placed on the fire until it appears to 
 boil. It is allowed to go on boiling, and all the water that is mixed 
 with the dripping boils away; and finally, in place of a boiling liquid, 
 there is a liquid with a perfectly smooth surface which is not boiling 
 at all. This is the time for cooking the potatoes. They are cut into 
 very thin shavings and thrown into the hot fat. They are then quickly 
 boiled and kept in the fat until slightly brown. Under the influence 
 of the great heat they have become firm and crisp upon the surface, 
 and the fluid that they contain is boiled within these crisp surfaces by 
 the fat, so that the chips are blown out, each little shaving of potato, 
 which was originally as thick as a bit of cardboard, is now three- 
 quarters of an inch thick. 
 
 Composition of various Articles of Diet from the Vegetable Kingdom. 1 
 
 Cekealia. 
 
 
 Water. 
 
 Nitro- 
 genous 
 
 Substances. 
 
 Fat. 
 
 Starch, 
 
 Sugar, 
 
 Gum, etc. 
 
 Cellulose. 
 
 Ash. 
 
 Wheat .... 
 
 13-56 
 
 12-42 
 
 1-70 
 
 67-89 
 
 2-66 
 
 1-79 
 
 Barley .... 
 
 13-78 
 
 11-16 
 
 2-12 
 
 65-51 
 
 4-80 
 
 2-63 
 
 Oats .... 
 
 12-92 
 
 11-73 
 
 6-04 
 
 55-43 
 
 10-83 
 
 3-05 
 
 Rice .... 
 
 13-23 
 
 7-81 
 
 0-69 
 
 76-40 
 
 0-7S 
 
 1-09 
 
 
 
 Leguminos 
 
 
 
 
 
 Garden beans 
 
 13-60 
 
 23-12 
 
 2-28 
 
 53-63 
 
 3-84 
 
 3-53 
 
 Peas .... 
 
 14-31 
 
 22-63 
 
 1-72 
 
 53-24 
 
 5-45 
 
 2-65 
 
 Lentils .... 
 
 12-51 
 
 24-81 
 
 1-85 
 
 54-78 
 
 3-58 
 
 2-47 
 
 1 v. Ziemssen's "Handbook of General Therapeutics," vol. i., Art. " On the Dietary of 
 the Sick and Dietetic Methods of Treatment," by Bauer, Trans., London, 1885. 
 
FR UITS— VE GE TABLES. 
 
 345 
 
 The moan composition of the potato '(Konig) is as follows : — 
 
 Water 
 
 Nitrogenous matters 
 
 Fat 
 
 Starch 
 
 Cellulose 
 
 Ash 
 
 Roots (Konig). 
 
 Field carrots 
 Turnips . 
 Beet-root . 
 Celery 
 Horse radish 
 Radish 
 Onions 
 
 Water. 
 
 87-05 
 91-24 
 87-07 
 84-09 
 70-72 
 86-92 
 85-99 
 
 Nitro- 
 genous 
 Substances. 
 
 1-04 
 0-96 
 1-37 
 1-48 
 2-73 
 1-92 
 1-68 
 
 Fat. 
 
 0-21 
 0-16 
 0-03 
 0-39 
 0-35 
 0-11 
 0-10 
 
 Sugar. 
 
 6-74 
 
 4-08 
 0-54 
 0-77 
 
 1-53 
 
 2-78 
 
 Non- 
 nitrogenous 
 Extracts 
 other than 
 Sugar. 
 
 2-60 
 1-90 
 9-02 
 11-03 
 15-89 
 6-90 
 8-04 
 
 Cellulose. 
 
 1-40 
 0-91 
 1-05 
 1-40 
 2-78 
 1-55 
 0-71 
 
 Ash. 
 
 0-90 
 0-75 
 0-92 
 0-84 
 1-53 
 1-07 
 0-70 
 
 Vegetables (Konig). 
 
 
 Water. 
 
 Nitro- 
 genous 
 Substances. 
 
 Fat. 
 
 Sugar. 
 
 Other Non- 
 nitrogenous 
 Substances. 
 
 Cellulose. 
 
 Ash. 
 
 Cabbage . 
 
 89-97 
 
 1-89 
 
 0-20 
 
 2-29 
 
 2-58 
 
 1-84 
 
 1-23 
 
 Cauliflower 
 
 90-39 
 
 2-53 
 
 0-38 
 
 1-27 
 
 3-74 
 
 0-87 
 
 0-82 
 
 Brussels sprouts 
 
 87-63 
 
 4-83 
 
 0-46 
 
 
 6-22 
 
 1-57 
 
 1-29 
 
 Spinach 
 
 90-26 
 
 3-15 
 
 0-54 
 
 0-08 
 
 3-26 
 
 0-77 
 
 1-94 
 
 Endive 
 
 94-13 
 
 1-76 
 
 0-13 
 
 0-76 
 
 1-82 
 
 0-62 
 
 0-78 
 
 Lettuce 
 
 94-33 
 
 1-41 
 
 0-31 
 
 
 2-19 
 
 0-73 
 
 1-03 
 
 Asparagus 
 
 93-32 
 
 1-98 
 
 0-28 
 
 0-40 
 
 2-34 
 
 1-14 
 
 0-54 
 
 French beans . 
 
 88-36 
 
 2-77 
 
 0-14 
 
 1-20 
 
 6-82 
 
 1-14 
 
 0-57 
 
 fhtnlen peasfnii 
 eeds) 
 
 80-49 
 
 5-75 
 
 0-50 
 
 
 10-86 
 
 1-60 
 
 0-80 
 
346 
 
 THE TREATMENT OF DIABETES. 
 
 Fruits. 
 
 
 Water. 
 
 Nitro- 
 genous 
 Matters. 
 
 Free 
 
 Acids. 
 
 Sugar. 
 
 Other Non- 
 nitrogenous 
 Matters. 
 
 Cellulose 
 
 + 
 Kernel. 
 
 Ash. 
 
 Apple 
 
 83-58 
 
 0-39 
 
 0-84 
 
 7-73 
 
 5-17 
 
 1-98 
 
 0-31 
 
 Pear . 
 
 83-03 
 
 0-36 
 
 0-20 
 
 8-26 
 
 3-54 
 
 4-30 
 
 0-31 
 
 Plum 
 
 84-86 
 
 0-40 
 
 1-50 
 
 3-56 
 
 4-68 
 
 4-34 
 
 0-66 
 
 Greengage 
 
 80-28 
 
 0-41 
 
 0-91 
 
 3-16 
 
 11-46 
 
 3-39 
 
 0-39 
 
 Peach 
 
 80-03 
 
 0-65 
 
 0-92 
 
 4-48 
 
 7-17 
 
 6-06 
 
 0-69 
 
 Apricot 
 
 81-22 
 
 0-49 
 
 1-16 
 
 4-69 
 
 6-35 
 
 5-27 
 
 0-82 
 
 Grape 
 
 78-18 
 
 0-59 
 
 0-79 
 
 24-36 
 
 1-96 
 
 3-60 
 
 0-53 
 
 Cherry 
 
 80-26 
 
 0-62 
 
 0-91 
 
 10-24 
 
 1-17 
 
 6-07 
 
 0-73 
 
 Strawberry 
 
 87-66 
 
 1-07 
 
 0-93 
 
 6-2S 
 
 0-48 
 
 2-32 
 
 0-81 
 
 Gooseberry 
 
 85-74 
 
 0-47 
 
 1-42 
 
 7-03 
 
 1-40 
 
 3-52 
 
 0-42 
 
 Orange (without 
 rind or pips) 
 
 89-01 
 
 0-73 
 
 2-44 
 
 4-59 
 
 0-95 
 
 1-79 
 
 0-49 
 
 Cucumber 
 
 95-60 
 
 1-02 
 
 
 0-95 
 
 1-33 
 
 0-62 
 
 0-39 
 
 Melon 
 
 95-21 
 
 1-06 
 
 
 0-27 
 
 1-16 
 
 1-07 
 
 0-63 
 
 Dried Fruits. 
 
 
 Raisins. 
 
 Figs. 
 
 Water .... 
 
 32-02 
 
 31-20 
 
 Nitrogenous matter 
 
 
 2-42 
 
 4-01 
 
 Fat 
 
 
 0-59 
 
 1-41 
 
 Free acid .... 
 
 
 
 1-21 
 
 Sugar 
 
 
 54-16 
 
 49-79 
 
 Other non -nitrogenous matters 
 
 
 7-48 
 
 4-51 
 
 Cellulose and seeds 
 
 
 1-72 
 
 4-98 
 
 Ash . . . 
 
 
 1-21 
 
 2-86 
 
 Nuts (Konig). 
 
 
 Water. 
 
 Nitro- 
 genous 
 Matter. 
 
 Fat. 
 
 Carbo- 
 hydrate. 
 
 Cellulose. 
 
 Ash. 
 
 Sweet almond 
 
 Walnut 
 
 Chestnut 
 
 5-39 
 
 4-68 
 
 51-48 
 
 24-18 
 16-37 
 
 5-48 
 
 53-68 
 
 62-86 
 
 1-37 
 
 7-23 
 
 7-89 
 
 38-34 
 
 6-56 
 6-17 
 1-61 
 
 2-96 
 2-03 
 1-72 
 
BREAD AND ITS SUBSTITUTES. 
 
 347 
 
 4. BEEA.D AND ITS SUBSTITUTES. 
 
 Bread is the article of diet with respect to which there is 
 the greatest difficulty in the dietetic treatment of diabetes, since 
 wh eaten bread contains a large percentage of starch. The mean 
 composition from a large number of analyses collected by 
 Koing( 24 ) is as follows: — 
 
 
 Mean for Fine 
 Bread. 
 
 Mean for 
 Coarse Bread. 
 
 Water .... 
 
 38-51 
 
 41-02 
 
 Nitrogenous substances . 
 
 6-82 
 
 6-23 
 
 Fat ... . 
 
 •77 
 
 •22 
 
 Sugar .... 
 
 2-37 
 
 2-13 
 
 Carbohydrates 
 
 49-97 
 
 48-69 
 
 "Woody fibre . 
 
 •38 
 
 •62 
 
 Ash .... 
 
 1-18 
 
 1-09 
 
 It is therefore unsuitable, at least in the ordinary quantities, 
 when a rigid diet is indicated. 
 
 Most patients, however, object to the withdrawal of bread 
 for any length of time from the diet, since it is so difficult to 
 find a satisfactory substitute. 
 
 In the most severe forms of diabetes, as previously men- 
 tioned, bread ought to be allowed in limited quantity, since a 
 rigid diet is dangerous. In the mildest cases of the mild form 
 of the diseases it is sometimes found, after keeping the patient on 
 a rigid diet for a short time, that a small quantity of bread may 
 be allowed, without any sugar reappearing in the urine, i.e. the 
 T)atient is able to tolerate a small quantity of carbohydrate food. 
 In these cases, of course, a little bread ought to be added to the 
 diet, so long as it does not cause glycosuria. But in other cases, 
 the more severe cases of the milder form (variety b, p. 325), the 
 addition of bread to a rigid diet always causes sugar to appear 
 in the urine. In such cases it is well to entirely replace bread 
 by some other substitute, temporarily at least. In course of time, 
 however, the patient's desire for bread often becomes so strong, 
 that this article of diet can only be in part replaced by some of 
 
348 THE TREATMENT OF DIABETES. 
 
 the bread substitutes, and a small amount of ordinary bread must 
 be allowed. 
 
 As already mentioned, for diagnostic purposes it is always 
 well to place a patient on a rigid diet — from which bread has 
 been withdrawn — for a short time, and then some bread sub- 
 stitute is necessary. 
 
 A number of bread substitutes have been from time to time 
 advocated, but unfortunately a large proportion of these are 
 useless. Many of them are exceedingly expensive, many are 
 very unpalatable, and many contain a very large percentage of 
 starch, as shown by analysis. 
 
 A large number of the specimens of diabetic bread and 
 biscuits which I have examined, have been coloured deep blue- 
 black by a drop of iodine and potassium-iodide solution, indi- 
 cating the presence of a large percentage of starch. Hence it 
 is not surprising that Schmitz, in writing of diabetic bread 
 substitutes, should state that they are of service chiefly to the 
 baker ; and that Saundby should state that diabetic breads are 
 often " neither more nor less than frauds." Some physicians, 
 especially on the Continent, prefer always to allow the patient a 
 small limited amount of ordinary white bread daily, rather than 
 trust to expensive, unpalatable, and unreliable substitutes. 
 
 For my own part, I fail to see the advantage of diabetic 
 foods, containing half as much starch as ordinary bread. Thus, 
 for example, some preparations of diabetic bread contain nearly 
 25 per cent, of carbohydrates. Now, ordinary bread only con- 
 tains about 5 per cent, of carbohydrates. Hence a patient 
 taking 12 oz. of this diabetic bread daily would consume 3 oz. 
 of carbohydrates, but he would only consume the same quantity 
 of carbohydrates if he took 6 oz. of ordinary bread, and the 
 latter he would find much less expensive, and much more palatable 
 and satisfactory, than. 12 oz. of certain diabetic breads. 
 
 All diabetic food preparations have not the above-mentioned 
 defects, however, and a number of these articles, when properly 
 prepared, are of great service in the cases indicated. I have 
 had many diabetic foods prepared at my own house ; as a 
 rule, they can be made easily; and diabetic bread and cakes 
 prepared at home are generally more satisfactory — more reliable, 
 more palatable, and much less expensive — than those obtained 
 from many firms. 
 
 Before recommending anv bread substitute, it is well to test 
 
BREAD AND ITS SUBSTITUTES. 349 
 
 a sample with a drop of a watery solution of iodine and potassium 
 iodide. Any preparation which gives a deep blue-black coloration 
 with this solution contains a large quantity of starch, and ought 
 to be rejected ; also samples of cocoa-nut or almond cakes ought 
 to be tested for sugar (by the fermeutation test). 
 
 The following are the most important substitutes for bread 
 and farinaceous articles of diet, which have been employed in 
 the treatment of diabetes : — 
 
 (1) " Torrified" bread, prepared by toasting thin slices of 
 bread until they are dark brown or almost blackened, is often 
 recommended for diabetic patients, in place of ordinary bread. 
 It is supposed that the starch and gluten of the bread are in part 
 decomposed by the heat. The patient is certainly not likely to 
 eat any large quantity of this burnt bread, and this is probably 
 its only advantage. 
 
 (2) Gluten bread is one of the oldest substitutes, and has been 
 perhaps more largely used than any other. It was first intro- 
 duced by Bouchardat, over fifty years ago, and has enjoyed 
 great popularity in France. It is prepared from gluten flour, a 
 substance obtained by washing away the starch from wheat flour. 
 A large number of specimens of gluten bread which I have tested, 
 have contained so much starch, however (indicated by the intense 
 blue colour produced with a drop of iodine solution), as to be 
 practically useless as bread substitutes. But all gluten breads 
 are not of this kind. Some preparations of gluten flour contain 
 only a small amount of starch and cellulose, 2 to 2 -5 per cent., 
 and are of great service to the diabetic patient. 
 
 Mr. E. 0. Bischof (28 Hanging Ditch, Manchester, and 35 
 Brooke Street, Holborn, London, E.C.) supplies a gluten flour which 
 has the following composition : — 
 
 Gluten, by direct extraction and drying . . . 82 "70 
 
 Loss in washing, principally soluble albuminous matter 
 Moisture ........ 
 
 Ash 
 
 Starch and cellulose • . ... 
 
 3-69 
 
 10-34 
 
 1-05 
 
 2-22 
 
 100 00 
 
 From this gluten Hour, bread can be easily prepared daily 
 at the patient's bouse, and for a time it may answer as a bread 
 substitute completely or partially. 
 
35 o THE TREATMENT OF DIABETES. 
 
 Recipe for making gluten bread. 
 
 Mix one pound of gluten flour (Bischof ) with three-quarters of a pint 
 or one pint of water, at 85° F. (JS T o yeast is required.) As soon as 
 mixed, put the dough into the tins, and immediately place them in the 
 oven, which should he ahout 430° F. Or the dough may be made into small 
 dinner rolls, and baked on flat tins. The loaves in the tins take about an 
 hour and a half to bake, and the rolls three-quarters of an hour. There 
 is not the slightest difficulty in the making of either. The addition of a 
 little salt improves the bread. 
 
 Gluten bread is light, dry, and crispy, and is improved by 
 being well buttered. The gluten bread bought from various 
 firms ought always to be tested with iodine and potassium iodide 
 solution, in order to form a rough estimate of the amount of 
 starch present. 
 
 By the use of a good gluten bread in the place of ordinary 
 bread, of course a marked diminution of the sugar- excretion may 
 be produced. For example, a patient suffering from a mild form 
 of diabetes, whose diet had been restricted for some time, was 
 allowed ordinary white bread for two weeks, but no other kind of 
 carbohydrate food. The average excretion of urine was 83 oz. 
 per diem, and the sugar excretion averaged 35 grs. per oz. of 
 urine. Then for two months the ordinary bread was replaced 
 by gluten bread, the diet otherwise remaining the same. The 
 quantity of urine averaged 54 oz. per diem, and the sugar 16 
 grs. per oz'. 
 
 Gluten flour may also be used for the preparation of pan- 
 cakes, puddings, etc. 
 
 Gluten padding may be prepared as follows : — 
 
 A batter of eggs, cream, and gluten flour is prepared. This is 
 flavoured with lemon or other essences, and baked. 
 
 Gluten pancakes are made by adding gluten floiir to one or two eggs, 
 and beating into a batter. The pancakes may be sweetened with 
 saccharine or eaten with a little glycerin. 
 
 (3) Bran cake. — The husk or bran of wheat consists of leguin 
 and an albuminoid substance ; after washing, to free it from 
 starch as much as possible, it may be made into bread with 
 butter and eggs. 
 
 Camplin ( 25 ) long ago strongly recommended bran cake as a 
 substitute for bread in diabetes. He found it necessary to have 
 the bran very finely ground, otherwise it was hurried through the 
 
BREAD AND ITS SUBSTITUTES. 351 
 
 intestines undigested, and diarrhoea was produced. When very 
 finely ground this disadvantage was overcome. 
 
 The following are his directions for the preparation of bran cake : — 
 "Take a sufficient quantity (say a quart) of wheat bran, boil it in two 
 successive waters for a quarter of an hour, each time straining it through 
 a sieve, then wash it well with cold water (on the sieve), until the water 
 runs off perfectly clear ; squeeze the bran in a cloth as dry as you can, 
 then spread it thinly on a dish, and place it in a slow oven ; if put in at 
 night let it remain until the morning, when, if perfectly dry and crisp, 
 it will be fit for grinding. The bran thus prepared must be ground in a 
 line mill and sifted through a wire sieve of such fineness as to require 
 the use of a brush to pass it through ; that which remains in the sieve 
 must be ground again until it becomes quite soft and fine. Take of this 
 bran powder 3 oz. (some patients use 4 oz.), the other ingredients as 
 follows — three newdaid eggs, H (or 2 oz. if desired) of butter, and about 
 half a pint of milk. 
 
 " Mix the eggs with a little of the milk, and warm the butter with the 
 other portion ; then stir the whole well together, adding a little nutmeg 
 and ginger, or any other agreeable spice. Bake in small tins (pattipans), 
 which must be well buttered, in a rather quick oven for about half an 
 hour. The cakes, when baked, should be a little thicker than a captain's 
 biscuit ; they may be eaten with meat or cheese for breakfast, dinner, 
 and supper; at tea they require rather a free allowance of butter, or 
 may be eaten with curd or any of the soft cheeses. It is important that 
 the above directions as to washing and drying the bran should be exactly 
 followed, in order that it may be freed from starch, and rendered more 
 friable. In some seasons of the year, or if the cake has not been well 
 prepared, it changes more speedily than is convenient — this may be 
 prevented by placing the cake before the fire for five or ten minutes 
 every day. Mr. Blatchley of 352 Oxford Street, London, makes the bran 
 biscuits, and prepares the powder for those at a distance who have no mill." 
 
 Bran flour can be obtained from Mr. E. Blatchley, 167 Oxford Street, 
 London, but Camplin recommends that the patient, if he does not pre- 
 pare the flour, should at least have it baked at his own house, rather 
 than purchase the dry hard biscuit of commerce. 
 
 (4) Soya biscuits or bread. — During the last seven years soya 
 beans have been used in the preparation of bread and biscuits for 
 diabetic patients. The soya or soja bean is used in Japan for 
 preparing various articles of food. The Japanese give the name 
 of daidsu to the plant from which the beans are derived, but 
 various botanical names have been given to it, Soja hispida, 
 Glycine hispida, etc. (IT. White 2C ). 
 
352 THE TREATMENT OF DIABETES. 
 
 In China an emulsion in water is made with the oil expressed 
 from the beans, and is drunk in those districts in which the 
 people are too poor to buy milk. 
 
 Dujardin-Beaumetz ( 27 ) recommended bread prepared from 
 the soja beans for diabetic patients, at the Berlin International 
 Congress in 1890, and since then it has been frequently 
 employed. 
 
 It is stated that the soja beans contain a large quantity of 
 nitrogenous matter and fatty material, and only a small per- 
 centage of starch ; and for this reason they are recommended as 
 a suitable article of diet for diabetic patients. But the various 
 analyses published show great discrepancies. According to 
 Dujardin-Beaumetz, soja bread contains less than 3 per cent, of 
 starch or sugar, 9 per cent, of fat, 20 per cent, of proteids, and 
 45 per cent, of water. The Lancet for August 16, 189 0, gives 
 2 - 72 as the percentage of starch in soja bread. Guy s Hospital 
 Gazette (28th February 1891) gives 6 '4 as the percentage of 
 starch; Saundby, 23 per cent, in the bread, 46 per cent, in the 
 biscuit, 45 in the flour ( 28 ). According to Professor Attfield, 
 soja flour contains 30 per cent, of cellulose, starch, and sugar. 
 
 Soja bread does not keep well, but the biscuits can be kept 
 for a long time. Most of the biscuits which I have tried have 
 had an unpleasant taste — some of them a very unpleasant taste. 
 One sample gave only a faint blue coloration with a drop of 
 iodine solution ; but several others, obtained at different times 
 from the same makers, contained a large amount of starch, 
 and gave a deep blue-black coloration with iodine solution. 
 Hale White ( 26 ) and others have obtained good results from the 
 use of soja preparations. 
 
 (5) Almond cakes were recommended long ago by Dr. Pavy 
 as a substitute for bread. Ground almonds, in the form of a 
 powder, can be obtained from many provision shops. The sweet 
 almonds contain 3 to 5 per cent, of sugar, but this may be removed 
 by pouring over the pulverised almonds boiling water acidified 
 with tartaric acid (Pavy), or acidified with a few drops of acetic 
 acid (Seegen). Almond flour washed in this manner is almost 
 free from sugar.] 
 
 Composition of the Sweet Almond (Konig). 
 
 Water 5-39 
 
 Nitrogenous matters ...... 24'18 
 
BREAD AND ITS SUBSTITUTES. 353 
 
 Fat 
 
 Carbohydrates 
 Cellulose 
 Ash 
 
 53-68 
 7-23 
 6*56 
 2-96 
 
 The following are the directions given by Seegen ( 29 ) for the 
 preparation of almond cakes : — ■ 
 
 Break up about a quarter of a pound of sweet almonds as fine as 
 possible in a stone mortar (or the almond flour may be used). Put the 
 flour into a linen bag, which is then immersed for a quarter of an hour 
 in boiling water, acidulated with a few drops of vinegar. This removes 
 the small amount of sugar from the almonds. Mix intimately with 
 3 oz. of butter and two eggs. Then the yolks of three eggs and a 
 little salt are added, and the whole stirred briskly for a long time. A 
 fine froth is made by beating the whites of three eggs, and then added 
 to the above mixture. The paste is made into biscuits, smeared with 
 melted butter, and baked with a gentle fire. 
 
 Almond cakes have long been used with a certain amount of 
 success, but, on account of the large quantity of fatty matter 
 which they contain, they are somewhat difficult to digest, and 
 are apt to give rise to dyspepsia. Their digestion is greatly 
 aided, however, by a small quantity of wine, brandy, or some 
 form of alcohol, taken with the biscuits, or directly afterwards. 
 
 Saundby gives the following direction for the preparation of 
 almond calces : — 
 
 One lb. of ground almonds, four eggs, two tablespoonfuls of 
 milk, a pinch of salt. Beat up the eggs, and stir in the almond flour ; 
 divide in twelve flat tins, and bake in a moderate oven for about fifteen 
 minutes. 
 
 I have found cakes prepared according to these directions 
 very palatable. They can be easily made at home. 
 
 I have also found almond pudding, prepared according to the 
 following directions, very palatable and useful : — 
 
 Two eggs, I lb. of almond flour, \ lb. of butter. Three tabloids of 
 saccharine dissolved in an ounce of brandy, or a little saccharine solution, 
 may be used to sweeten the mixture. 
 
 Warm the butter, beat in the yolks of the eggs and the almond 
 flour ; add the saccharine. 
 
 Whisk the whites of the eggs into a froth, then mix all together, 
 place in a mould, and bake in a quick oven. — Mrs. Hart ( :J0 ). 
 
354 THE TREATMENT OF DIABETES. 
 
 (6) Cocoa-nut. — The edible part of cocoa-nut forms a useful 
 addition to the diet of the diabetic. It may be purchased in 
 desiccated powder, which can be easily made into biscuits. 
 Cocoa-nut powder is very cheap, and costs only 4|d. to 6d. 
 per lb. Cocoa-nut contains a small amount of sugar, but a 
 large amount of fat (70 per cent.). 
 
 Almost all the samples of cocoa-nut powder which I have 
 examined have contained a very small quantity of sugar, and 
 sometimes it is adulterated with sugar ; hence it is well to test 
 the powder (by the fermentation test), to ascertain if there is any 
 quantity of sugar present. It may also be tested for starch with 
 iodine. A good sample contains, however, very little sugar ; and 
 specially prepared cocoa-nut powder can be obtained, which is 
 stated to be free from sugar, but most of the samples which I have 
 examined have contained a little, and it is somewhat expensive. 
 
 The trace of sugar can easily be removed from ordinary 
 cocoa-nut powder, however, by adding a little yeast and water, 
 and allowing the mixture to ferment before use. 
 
 Biscuits and cakes made of cocoa-nut may be prepared for 
 diabetic patients. The following are directions for home-made 
 cocoa-nut and cream cakes, which I have found to be very 
 palatable ! — 
 
 Three tablespoonfuls of cocoa-nut powder are mixed into a paste, 
 •with a little German yeast and water. The mixture is allowed to 
 remain by the fire, or in a warm place, for about twenty minutes, until 
 fermentation occurs, and it becomes " puffy." Then a little of a watery 
 solution of saccharine is added. 
 
 One egg is beaten up, and this with two teaspoonfuls of cream and 
 a little water are added to the cocoa-nut paste. The whole is well 
 mixed, and dropped into small tins, and baked in an oven for about 
 thirty minutes. 
 
 These cakes are excellent, but contain so much fatty material 
 that in many persons they cause slight dyspepsia. This may 
 easily be prevented by taking a little wine, or alcohol in some 
 form, soon after eating the cakes. 
 
 Saundby gives the following directions for the preparation of 
 cocoa-nut and almond cakes : — 
 
 Three-quarters of a lb. of the finest desiccated cocoa-nut powder, \ lb. 
 of ground almonds, six eggs, half a teacupful of milk. 
 
 Beat up the eggs, and stir in the cocoa-nut and almond flour. Divide 
 into sixteen flat tins, and bake twenty-five minutes in a moderate oven. 
 
BREAD AND ITS SUBSTITUTES. 355 
 
 Cocoa-nut can also be made into puddings, which will supply 
 the place of rice pudding. I have found the following useful :— 
 
 Take three tablespoonfuls of cocoa-nut powder, mix with a little 
 German yeast and water, and keep for twenty minutes in a warm place, 
 so as to decompose the small quantity of sugar present; add four 
 tablespoonfuls of cream, one egg, a little salt, half a pint of water 
 sweetened with saccharine. 
 
 Mix into a paste. Place in a dish greased with butter. Cook like 
 rice pudding, in a slow oven, thirty minutes. 
 
 Cocoa-nut pancakes can also be prepared. I have had these 
 made as follows : — ■ 
 
 Take one egg ; beat up in two tablespoonfuls of milk, or better, in a 
 little cream and water. Add a pinch of salt. 
 
 Then add two tablespoonfuls of cocoa-nut powder (freed from 
 sugar). Allow to stand five to ten minutes. 
 
 Add a little more cream and water. 
 
 Mix well, until it is a little thicker than ordinary pancake batter. 
 Place a little lard in a frying-pan ; heat until the lard is just melted ; 
 then drop in half of the above mixture. 
 
 Allow to remain over a moderate fire for a few minutes (five), until 
 the under surface is brown ; then turn the cake over, and heat for 
 another five minutes. 
 
 The other half of the mixture may be used for a second pancake. 
 
 (7) Aleuronat. — Professor Ebstein ( 31 ) has drawn attention 
 to the value of an albuminous substance, named aleuronat, in 
 Germany (dkevpov = flour). This is a vegetable albumin, 
 which Dr. Hundhausen prepares by a special process from wheat. 
 It is a light yellowish brown powder, and contains 80 to 90 
 per cent, of albumin in the dry substance, and only 7 per cent, 
 of carbohydrates. (It can be obtained from E. Hundhausen, 
 Hamm, Westphalia, Germany, in parcels of 4| kilos., for about 
 seven shillings.) Aleuronat powder is thus a cheap form of 
 albumin, and can be used as a substitute for ordinary flour 
 in various ways, in cooking, in the preparation of soups, sauces, 
 etc., and it can be baked into bread. But in the preparation of 
 bread it is necessary to add a considerable amount of ordinary 
 Hour. Ebstein gives directions for the preparation of aleuronat 
 breads containing 27'5 per cent, and 50 per cent, of albumin 
 in the dry substance. The following are his directions for the 
 preparation of bread containing 50 per cent, of albumin: — 
 
356 THE TREATMENT OE DIABETES. 
 
 About 6 or 7 oz. of ordinary white flour. 
 „ 6 or 7 oz. of aleuronat powder. 
 „ 5 oz. of butter (of the best quality). 
 
 One teaspoonful of salt. 
 
 Three-quarters of an ounce of baking powder. 
 
 The flour and aleuronat are mixed in a warm dish, and the melted 
 butter and milk (made lukewarm) are gradually added, then the salt, 
 and finally the baking powder (one part of sodium carbonate and two 
 parts of cream of tartar). The dough is well mixed, then formed into 
 loaves, and baked at a good heat. 
 
 I have had aleuronat bread prepared at my own house, and 
 have prescribed it for diabetic patients. It forms a useful and 
 fairly palatable brown bread ; but, as will be seen from its com- 
 position, it contains a considerable amount of starch. A patient 
 taking 16 oz. of this bread per day would consume as much 
 starch as a patient taking 8 oz. of ordinary white bread, but 
 the amount of albumin would be much greater in the former 
 case. After a time most patients would much prefer the 
 8 oz. of ordinary bread to the 1 6 oz. of aleuronat bread. It has 
 always appeared to me that half measures with regard to diabetic 
 food preparations are somewhat unsatisfactory, and that it is 
 better either to replace bread entirely by some substitute almost 
 free from carbohydrates, or to allow a definite small amount of 
 ordinary bread, with the addition of some starchless diabetic 
 bread substitute. 
 
 Aleuronat and cocoa-nut cakes. — After a number of experi- 
 ments, which I have had made at my own house, I have found 
 that very satisfactory cakes and buns can be formed by a com- 
 bination of aleuronat and cocoa-nut ( 32 ). Home-made prepara- 
 tions of these cakes are preferable ; and at the Manchester 
 Infirmary, for several years, excellent biscuits have been 
 made by the cook, for diabetic patients, according to my direc- 
 tions. Of course the patient regards all bread substitutes as 
 more or less objectionable after prolonged use. But I have 
 found that most of the numerous private and hospital patients 
 who have used these aleuronat and cocoa-nut cakes prefer them to 
 any other kind of diabetic bread, and their composition is reliable. 
 
 Cocoa-nut powder contains a large quantity of fatty material, 
 but a small quantity of sugar is also present ; as already 
 mentioned, the latter can be almost entirely removed by the 
 action of yeast. 
 
BREAD AND ITS SUBSTITUTES. 357 
 
 For the preparation of these cakes, 2 oz. of desiccated 
 cocoa-nub powder are mixed with a little water containing a 
 small quantity of German yeast. The mass is then formed 
 into a kind of paste, and this is kept for half an hour or longer 
 in a warm place. The small amount of sugar contained in the 
 cocoa-nut is almost entirely decomposed by the fermentation 
 produced by the yeast, and the cocoa-nut paste becomes spongy. 
 Two oz. of aleuronat, one egg beaten up, and a small quantity 
 of water, in which a little saccharine or saxin has been dissolved, 
 are now added to the cocoa-nut, and the whole well mixed until 
 a paste is formed. This is spread out on a tin and divided into 
 cakes, which are baked in a hot oven for twenty or thirty 
 minutes. 
 
 Messrs. Callard & Co. of Eegent Street have kindly 
 supplied me with a desiccated cocoa-nut powder, from which the 
 small amount of sugar has been almost entirely removed. If 
 this be employed, then the addition of the yeast is not necessary ; 
 2 oz. of aleuronat, 2 oz. of cocoa-nut, 1 egg (beaten up), 
 and a little water sweetened with saccharine, are simply mixed 
 together, divided into cakes, and baked. 
 
 The cakes are most palatable when newly made. If they 
 have been prepared more than twenty-four hours, the taste is 
 greatly improved if they are slightly warmed before the fire. 
 They are also improved by being buttered. 
 
 I have had excellent buns, about 1 inch in thickness, pre- 
 pared from aleuronat and cocoa-nut powder. Two oz. of desic- 
 cated cocoa-nut powder are mixed with a little yeast and water, 
 and kept in a warm place (by the fire) for fifteen to twenty 
 minutes. Then 2 oz. of aleuronat are mixed well with one tea- 
 spoonful of baking powder and a little salt. After the action 
 of the yeast, the cocoa-nut powder is added to the aleuronat, 
 together witli one egg, beaten up, and water (sweetened with 
 saccharine or saxin, if preferred). The mixture is worked into 
 a thin batter, and this is then placed in deep tins or tart dishes, 
 which are put at once into a hot oven for twenty to thirty 
 minutes. When half baked, the buns or rolls may be glazed 
 with a little egg albumin, and then placed in the oven again 
 until browned. 
 
 The above directions indicate the proportions of the sub- 
 stances, but of course large quantities can be employed. 
 
 The advantages of these cakes, buns, and rolls are — (1) A 
 
358 THE TREATMENT OF DIABETES 
 
 diabetic patient can have them easily prepared at his own home. 
 
 (2) If so prepared, they are cheap in comparison with most 
 diabetic foods. (Aleuronat can be obtained from Germany, 
 in parcels containing 4| kilos, for 7s., that is, about 9-|d. 
 per lb. ; fine desiccated cocoa-nut powder at 4§d. per lb.). 
 
 (3) They are much more palatable than most diabetic cakes ; 
 to many persons they are very palatable. (4) They contain 
 a large quantity of vegetable albumin and fatty matter, and 
 only a very small amount of carbohydrate. A solution of 
 iodine gives no blue coloration with the cakes, and hence 
 the starchy matter must be very small. When given to 
 diabetic patients in place of ordinary bread, of course a great 
 diminution of the quantity of sugar excreted in the urine is 
 observed. 
 
 I have found these biscuits and buns useful as a bread 
 substitute, when for diagnostic purposes a patient has been 
 placed on a rigid diet. They are also of great service in the 
 dietetic treatment of diabetes, to replace ordinary bread totally 
 .or in part when a rigid diet is indicated. 
 
 The table on p. 3 5 9 shows the diminution in the sugar excretion 
 and diuresis produced in a diabetic patient by replacing ordinary 
 bread by the same quantity of aleuronat and cocoa-nut biscuits 
 — the diet otherwise being the same. The patient was put on 
 a diet of meat, green vegetables, and bread, and 1 drm. of citrate 
 of potassium given three times a day. The sugar was carefully 
 estimated. As is usually the case, there was a diminution of 
 the sugar excretion during the first three days, but an equi- 
 librium was soon reached. Omitting the figures for the first 
 week from consideration, for the next six days, when the patient 
 was taking 14 oz. of ordinary bread daily, the sugar excretion 
 was from 3520 to 4800 grs. per day; the amount of urine 
 from 200 to 240 oz. daily. The urine gave no brown-red 
 reaction with perchloride of iron, and no reaction for acetone. 
 At the end of this time 14 oz. of aleuronat and cocoa-nut cakes 
 were given in place of the 14 oz. of bread for five days. Other- 
 wise the diet was kept exactly the same. The thirst became 
 less, the diuresis and glycosuria diminished ; the daily amount of 
 urine being 112 to 159 oz., and the amount of sugar 1792 to 
 2850 grs. To make the observation complete, at the end of five 
 days, 14 oz. of white bread were given in place of the aleuronat 
 and cocoa-nut cakes, the diet otherwise remaining the same. 
 
BREAD AND ITS SUBSTITUTES. 
 
 359 
 
 The thirst increased, and the diuresis and sugar excretion rose 
 to their previous level. In this case the bread was carefully 
 weighed daily. By replacing bread by the same quantity of 
 aleuronat and cocoa-nut biscuit, the average quantity of urine 
 excreted was thus reduced by 80 oz. daily, and the average 
 amount of sugar excreted was reduced practically by 2000 
 grs. per diem. 
 
 
 Ounces 
 
 of 
 Urine. 
 
 Sp. Gr. of 
 
 Unfermented 
 
 Urine. 
 
 Sp. Gr. of 
 
 Fermented 
 
 Urine. 
 
 Grains of Sugar 
 
 excreted 
 
 daily. 
 
 Diet. 
 
 , f 
 
 (6 days) 1 
 
 230 
 220 
 225 
 230 
 240 
 200 
 
 1025 
 
 1023 
 1023 
 1023 
 1024 
 1024 
 
 1007 
 1007 
 1007 
 1003 
 1004 
 1003 
 
 4140 
 3520 
 3600 
 4600 
 4800 
 4200 
 
 Chop ; steak or fish ; 
 green vegetables ; 
 milk, 2 pints ; tea ; 
 ' water ; 14 oz. of 
 bread ; citrate of 
 potash, 1 grm., 
 three times a day. 
 
 , f 
 
 (5 days) j 
 
 150 
 112 
 150 
 150 
 159 
 
 1023 
 1022 
 1021 
 1022 
 1023 
 
 1004 
 1006 
 1009 
 1006 
 1011 
 
 2850 
 1792 
 1800 
 2400 
 1908 
 
 "\ Medicine and diet 
 the same, except 
 that the 14 oz. of 
 
 V white bread were 
 replaced by 14 oz. 
 of aleuronat and 
 
 J cocoa-nut cakes. 
 
 in. / 
 
 (4 days) 1 
 
 188 
 258 
 220 
 220 
 
 1024 
 
 1024 
 1026 
 1026 
 
 1006 
 1002 
 1004 
 1006 
 
 3384 
 5676 
 4840 
 4400 
 
 | The aleuronat and 
 cocoa-nut cakes re- 
 V placed by 14 oz. of 
 j white bread ; diet 
 1 otherwise the same. 
 
 
 Average amount of 
 Urine daity. 
 
 Average amount of 
 Sugar daily. 
 
 First Stage. 
 
 
 
 14 oz. white bread . 
 
 224 oz. 
 
 4143 grs. 
 
 Second Stage. 
 
 
 
 14 oz. of aleuronat and cocoa- 
 nut biscuits in place of bread 
 
 144 ,, 
 
 2150 „ 
 
 Third Stage. 
 
 
 
 14 oz. of bread in place of 
 biscuits 
 
 221 ,, 
 
 4585 „ 
 
 I have also had cakes prepared from aleuronat and almond 
 flour as follows : — 
 
 Three oz. of aleuronat ; 3 oz. of almond flour ; one egg beaten up ; 
 about two teaspoonfuls of cream, and a little water. 
 
 Moisten the aleuronat with a little water containing saccharine, for a 
 few minutes ; then add the almond flour, the egg and cream, and water 
 
360 THE TREATMENT OF DIABETES. 
 
 as required, just to make a light paste. Spread on to a tin. Cut into 
 squares and bake in a moderate oven for twenty minutes. 
 
 Aleuronat may also be used in the following preparations : — 
 
 Aleuronat Pancake. 
 
 Take one egg ; beat up in a little water and cream : take two table- 
 spoonfuls of aleuronat powder ; add half a teaspoonful of baking powder 
 and a pinch of salt ; mix well, then add gradually to the egg and 
 cream, and beat into a batter ; allow to stand five minutes. If too 
 thick, add a little more cream and water. 
 
 Fry as an ordinary pancake in a frying-pan with a little lard. At 
 the end of about eight minutes, when the under surface is browned, turn 
 it over and continue to heat for about five minutes longer. 
 
 I have found that the following forms a useful, palatable, and 
 cheap combination for diabetic patients : — 
 
 Aleuronat and Suet Padding. 
 
 Take 2 oz. aleuronat flour, 2 oz. suet, one egg, a pinch of salt, half 
 a teaspoonful of baking powder. Sprinkle a little aleuronat flour on a 
 chopping-board. Chop tbe suet on this part of the board. Then mix 
 all the aleuronat with the chopped suet in a basin. Add the salt and 
 baking powder. Beat up the egg in about three tablespoonf uls of water, 
 to which a little saccharine has been added. Then add the egg 
 gradually to the mixed aleuronat and suet, stirring the whole mass well 
 into a paste. The addition of a little more water may be necessary. 
 Drop into a tin pudding mould, smeared with butter or lard, and float 
 it in a pan of water, and boil for two hours, taking care that the water 
 does not flow over into the pudding-mould ; or, better still, the pudding 
 may be baked in the oven. The addition of almonds (h oz.) improves 
 the taste. It can be eaten with a little red wine as a sauce. 
 
 (8) Inulin biscuits. — Inulin is assimilated by diabetic 
 patients, and Ktilz ( 33 ) recommends inulin biscuits prepared 
 according to the following directions : — 
 
 Fifty grms. of inulin are placed in a large porcelain basin, and, 
 while standing over a water bath, are rubbed up with 30 c.c. 
 of milk, and as much hot water as may be necessary, into a uniform 
 dough, with which the yolk of four eggs and a little salt are mixed. 
 To this the whites of the four eggs are added, having first been beaten 
 to a foam, and carefully worked in. The dough is finally baked in tin 
 moulds, previously smeared with butter. The taste of the biscuits may 
 be improved by the addition of vanilla or other spices. 
 
BREAD AND ITS SUBSTITUTES. 361 
 
 Inulin is so expensive, however, that this bread is not 
 likely to be of much practical use. 
 
 (9) Bread prepared from pea-nut flour has been recommended 
 by Stern. The kernel of the pea-nut (Arachis hypogce) contains 
 29 per cent, of proteids, 49 per cent, of fat, and 14 per cent, 
 of carbohydrates in the dry material. The oil is extracted and 
 used for various purposes in trade. The meal contains 5 2 per 
 cent, of proteids, 8 per cent, of fat, and 27 per cent, of 
 carbohydrates. The pea-nut is one of the cheapest foodstuffs 
 known. 
 
 The following are Stern's ( 3i ) directions for the preparation 
 of a " diabetic pea-nut flour " : — 
 
 The pea-nut kernels are boiled in water for half-an-hour, to extract 
 a portion of the oil which they contain. They are then dried and 
 pounded into fine particles by a rolling-pin. The pounded kernels are 
 then placed in boiling water, acidulated with tartaric acid or vinegar 
 in order (1) to extract saccharine elements, (2) to overcome the smell 
 and taste of the pea-nut, (3) to prevent emulsification of the remaining 
 oil. After having undergone a thorough boiling with acidulated water, 
 the grouud kernels are subjected to dry heat, and then rolled into fine 
 flour. The flour can be used in the form of porridge with milk ; bread 
 and biscuits can be baked from it ; but it may be utilised in the form of 
 the German pancake. As will be seen from the figures given above, it 
 contains a considerable amount of carbohydrates, however. 
 
 (10) O'Donnell recommends the following as a home-made 
 substitute for bread ( 35 ) : — 
 
 Six eggs are thoroughly beaten, then a teaspoonful of baking- 
 powder or its chemical equivalent and a quarter of a teaspoonful of salt 
 are added, and again the eggs are beaten. This mixture poured into hot 
 waffle irons, smeared with butter, is baked in a very hot oven. For 
 variety pulverised nuts (almonds) may be added. They may be eaten 
 hot with butter and cheese. 
 
 (11) Iceland moss is a lichen which is much used by the 
 inhabitants of Iceland, Lapland, and the Arctic regions as an 
 article of food. It contains a soluble gelatinous, carbohydrate 
 substance, known as lichenin, and two bitter acids — cetraric 
 and lichen -stearic. After the removal of these bitter substances 
 by repeated washing (or washing with a solution of bicarbonate 
 of potash), a bread may be formed from the moss, which is 
 Largely used in Iceland. 
 
362 THE TREATMENT OF DIABETES. 
 
 (12) Meat bread and biscuits. — A bread substitute composed 
 chiefly of flesh meat has been recommended by Baron Luhdorf ( 36 ). 
 Also meat biscuit and cheese biscuit are prepared by some 
 makers of diabetic foods, but naturally they have never been 
 much used. 
 
 Bevekages. 
 
 It is now generally acknowledged that it is unnecessary to 
 restrict the quantity of fluid taken by diabetic patients, unless 
 the thirst be exceedingly great. Restriction of the amount of 
 fluids has been shown to be distinctly harmful ; the volume 
 of urine and the amount of sugar excreted daily may be 
 temporarily reduced thereby, but the general distress increases. 
 A diabetic patient ought not to be allowed either to hunger 
 or thirst ; but of course food and liquids must be taken in 
 moderation. The greater the quantity of sugar excreted by any 
 diabetic patient, the greater the thirst as a general rule, and 
 vice versd. If, by treatment, the sugar excretion is diminished, 
 the thirst and diuresis are also diminished. 
 
 Prout has recommended that all liquids should be taken 
 tepid, since the thirst is thereby relieved much better than 
 when the liquids are cold. Of beverages, good drinking water 
 is the best. Tea and coffee both contain a small quantity 
 of carbohydrates, but the percentage is so very small (especially 
 in the case of tea) that they may be allowed freely ; of course 
 they must not be sweetened with sugar ; if a sweet taste is 
 preferred, saccharine or saxin may be added. Mate or kola 
 may also be allowed. Cocoa contains a considerable amount 
 of carbohydrates, and hence ought to be forbidden. A prepara- 
 tion of cocoa, from which the carbohydrates have been removed, 
 is sold in Neuenahr, and to this of course there is no objection. 
 Chocolate should be forbidden, on account of the considerable 
 percentage of carbohydrates which it contains. It has been 
 already pointed out that milk contains 4 per cent, of lactose; 
 whilst some patients take it without any bad effects, in others 
 the sugar excretion is increased by its use. Unless, therefore, 
 it has been proved that the case belongs to the former class, 
 milk ought to be forbidden, or allowed only in limited quantities 
 when a rigid diet is indicated. Cream is one of the most useful 
 articles of diet, and the artificial milk prepared from cream 
 
BEVERAGES. 363 
 
 (see p. 334) may be taken freely. Diabetic koumiss (see 
 p. 336) and kephir may also be allowed. Bouillon, beef- 
 tea, meat-broths (to which no carbohydrate material has been 
 added), eggs beaten up with cream and water, or with a little 
 milk, and sweetened with saccharine or saxin, may be taken 
 without restriction. Soda, potash, seltzer and Apollinaris waters, 
 or the alkaline waters of Neuenahr, Carlsbad, or Vichy, may 
 be allowed freely. But these carbonated alkaline waters should 
 not be given at meals, as they impair digestion, especially the 
 digestion of fats. 
 
 Alcohol does not increase the sugar excretion ; and when 
 taken in moderation and in a form unmixed with carbohydrates, 
 it is of great service in some cases. It may be given in the 
 form of brandy or cognac, or some of the wines which contain 
 very little sugar. In the milder forms of diabetes, alcohol is 
 unnecessary, but in the more severe forms it is very useful. 
 Alcohol is of value chiefly as an aid to the digestion of fatty 
 food. Fatty articles of food are of the greatest importance 
 in the diet of diabetics, especially in the severe forms of the 
 disease, but frequently these articles cause dyspepsia when taken 
 in considerable amount. The patient is, however, often enabled 
 to eat large quantities of fatty food, without nausea or 
 indigestion, if he take a small amount of alcohol in the form 
 of brandy or non-saccharine wine, with or just after the meal. 
 It is also of some service in aiding the digestion of nitrogenous 
 food. By its use, therefore, in the severe forms of the disease, 
 a much larger quantity of fatty food can be taken, and this 
 is a point of importance when there is much wasting, or when 
 tubercular complications are present. 
 
 Hirschfeld attaches considerable importance to the use of 
 alcohol in cases of commencing diabetic coma. 
 
 Wines and alcoholic beverages. — Of course, great caution is 
 necessary with respect to alcoholic beverages, as the thirsty 
 diabetic is very liable to take these to excess ; certainly they ought 
 only to be allowed with the meals. Some beverages are suitable, 
 whilst others are very unsuitable. Wines which contain a 
 large quantity of sugar must, of course, be avoided, but if the 
 percentage of sugar be very small they may be permitted. 
 Many old wines contain only a trace of sugar, and, as a rule, 
 the older the wine the more suitable it is for diabetic patients ; 
 new wines should generally be avoided. The name of the 
 
364 THE TREATMENT OF DIABETES. 
 
 wine is not always a reliable guide as to its composition, and, 
 before advising any wine or wines for frequent consumption, 
 it is important to obtain information respecting the results of 
 the chemical analysis. It is also to be remembered that wine 
 is a drink and not a food ; and large quantities ought not to be 
 taken by diabetic patients. But in addition to their value in 
 quenching thirst, and in aiding the digestion of fat (owing to 
 the alcohol they contain), wines are probably of value, as Schmitz 
 and others have pointed out, on account of the vegetable acid 
 and mineral elements present, especially since the diabetic 
 patient is to some extent deprived of these, through abstaining 
 from fruit. 
 
 Bordeaux wines contain only about 0*2 per cent, of sugar; 
 Bhiiie wines only 0*3 to 0'4 per cent. (Konig) ; Hock, Moselle, 
 and Ahr wines contain very little ; whilst some dry sherries are 
 free from sugar. 
 
 Austrian and Hungarian table wines also contain only a 
 very small quantity of sugar; and Hungarian Carlowitz is 
 practically free from sugar. All these wines may therefore 
 be permitted in moderation. A pleasant way of taking the 
 very acid table wines is to dilute them with a little soda 
 water. In this way there is less risk of " heart-burn " and colic 
 from the strong acidity of these wines. 
 
 Wines which contain sugar in large or considerable 
 quantities must be avoided. Must, Malaga, Port, Madeira, 
 Tokai and the other sweet Hungarian wines, Malmsey, cham- 
 pagne, and many Spanish, Sicilian, and Greek wines, contain 
 large quantities of sugar, and must therefore be avoided. 
 
 A champagne sans sucre can be obtained, however, by 
 those diabetic patients with whom this wine is a favourite 
 beverage. 
 
 Sweet ales, porter, rum, sweetened gin, and beer must also 
 be forbidden. Brandy, cognac, or old whisky may be taken in 
 small quantities. 
 
 In order to relieve thirst, it is important, of course, to try to 
 diminish the sugar excretion by restricting the diet and by the 
 use of drugs. As above stated, a reduction in the sugar excretion 
 diminishes the thirst. But, in addition, there are other means 
 of considerable service in relieving this troublesome symptom, 
 such as acid drinks of various kinds. Sir William Eoberts 
 recommends bitartrate of potash water, of which the following 
 
MODE OF LIFE, HYGIENIC CONSIDERATIONS. 365 
 
 is a pleasant form. A drachm or a drachm and a half of cream 
 of tartar is dissolved in a pint of boiling water, and flavoured 
 with lemon peel and saccharine ; when cold it may be taken 
 in small quantities throughout the day. 
 
 The following forms a useful lemonade, which is often 
 prescribed for the relief of thirst : — Citric acid, 1 grs. ; gly- 
 cerin, 4 drms. ; water, one pint. This may be taken in small 
 quantities during the twenty- four hours. Or a lemonade may 
 be made from fresh lemons, and sweetened with saccharine or 
 saxin. 
 
 Water containing a little dilute phosphoric acid or other 
 dilute acids is also often prescribed. 
 
 Then there are several other means of relieving thirst. 
 The patient may be allowed to suck ice, broken up into small 
 fragments. Washing out the mouth frequently with cold water, 
 with iced water, with acidulated water, or with weak brandy and 
 water, sometimes gives considerable relief. Bouchardat recom- 
 mends that the patient should be allowed to chew roasted coffee 
 beans, and diabetics tell me that this is successful in relieving; 
 thirst to some extent. 
 
 Large quantities of fluid should not be taken immediately 
 after meals ; the chief potations should precede the meals. 
 
 Mode of Life, Hygienic Considerations, etc. 
 
 The mental condition. — In the section devoted to etiology 
 it has been shown that not infrequently diabetes has followed 
 some severe mental shock. It is also well known that the 
 condition of a diabetic patient is often markedly affected by the 
 mental state. Severe mental shock or nervous excitement is 
 frequently followed by a decided advance of the symptoms, and 
 sometimes by diabetic coma. 
 
 Hence it is important to relieve the patients from mental 
 anxieties and worry as much as possible. Everything should be 
 done to cheer them and to make them forget their business, 
 professional, or family cares and anxieties. This is especially 
 desirable when the patients suffer, as is so frequently the case, 
 from great mental depression. A change of climate or residence, 
 the removal from old associations, in some cases retirement 
 from business or professional duties, may all be of great service 
 by their good effect upon the mental condition of the patient. 
 
366 THE TREATMENT OF DIABETES. 
 
 No doubt a considerable share of the good results of a visit 
 to Carlsbad or Neuenahr may be accounted for by the cheering 
 effect on the mind, produced by the absence from the worries 
 and anxieties of daily life. In fact, a visit to any health resort 
 may be of service, through its beneficial effect on the mental 
 condition, providing the patient is suffering from a mild form 
 of the disease only, and that he is not in an advanced stage, 
 and that there is no likelihood of coma being brought on by 
 the journey. 
 
 A few years ago we heard much of the supposed value of 
 music in therapeutics. Of course, nothing but slight improve- 
 ment could ever be expected from the influence of music in any 
 disease. But when we consider the undoubted influence of the 
 mental state on the symptoms of diabetes, it would certainly 
 appear that the effects of music are worthy of a trial, especially 
 if the patient be suffering from much mental worry and anxiety. 
 Many intelligent diabetic patients keep a sharp look-out 
 on the amount of sugar excreted daily, and are greatly 
 alarmed if the number of grains increases even to a slight 
 extent. Hence it is better, when the patient shows any anxiety 
 about the amount of sugar excreted, to prevent him obtaining 
 detailed information on this point. The instructive case re- 
 corded by Karl G-rube is well worth bearing in mind. A patient, 
 in whose urine a trifling amount of sugar had been discovered, 
 was so alarmed on hearing thereof, that symptoms of acute 
 Graves' disease developed, and rapidly proved fatal. This patient 
 had lost a relative from diabetes, and at once concluded that the 
 trace of sugar discovered in her own urine indicated an incurable 
 and severe disease. 
 
 Marriage. — Women who suffer from diabetes, especially in 
 its more severe forms, ought not to marry, as pregnancy is 
 dangerous. Abortion frequently occurs, and either confinement 
 or abortion has an injurious effect. 
 
 Cases of diabetes in males are sometimes met with, in which 
 the symptoms have developed not long after marriage, and it 
 is possible that sexual excess may have played some part as 
 an exciting cause. It is advisable that males suffering from 
 diabetes (especially in the more severe forms) should not marry ; 
 also all sexual excess should be avoided. 
 
 Clothing. — It is important that diabetic patients should be 
 well protected from cold, and that they should wear warm 
 
MODE OF LIFE, HYGIENIC CONSIDERATIONS. 367 
 
 clothing ; in winter, woollen clothing should be worn next to the 
 skin. Camplin long ago pointed out that diabetics were worse 
 in cold weather, and insisted on the importance of warm clothing. 
 
 Action of the shin. — Warm baths followed by cutaneous 
 friction are of service in promoting the action of the skin, 
 especially if the latter be excessively dry, but cold plunges or sea 
 baths ought not to be taken. 
 
 Exercise. — Kiilz has shown that in dogs, by vigorous exercise 
 the glycogen is made to disappear from the liver. He also 
 made experiments on five diabetic patients in order to determine 
 the effect of exercise. He found that exercise diminished the 
 sugar excretion in two powerful muscular patients, whilst in the 
 other three cases (patients who were badly nourished) the sugar 
 excretion was unchanged in two, increased in one. Zimmer 
 found that in well-nourished patients exercise diminished the 
 sugar excretion, but in patients suffering from a very severe 
 form of the disease, much exercise was useless or even injurious. 
 Seegen has obtained good results from moderate exercise in 
 the open air ; and when the disease is not too advanced, he 
 recommends the patient to spend the winter in a mild climate, 
 such as that of the Eiviera, where exercise in the open air 
 can be taken frequently. Schmitz refers to the cases of eight 
 soldiers who suffered from diabetes, and whose urine contained 
 from 1 to 4 per cent, of sugar; in each case the exercise of 
 a manoeuvre caused the sugar to disappear. Frerichs mentions 
 the case of an English medical man, whose diabetic symptoms 
 ceased when he took vigorous exercise in the country, hunting 
 with his brother. He also mentions the case of a landlord, 
 whose glycosuria disappeared through free exercise in the open 
 air. Brunton has drawn attention to the value of exercise in 
 the glycosuria of gouty persons. The value of exercise appears 
 to vary, therefore, according to the form of the disease. We 
 have already seen that the diet which is most suitable for mild 
 forms of diabetes is unsuitable for severe forms, and in like 
 manner a difference ought to be made with respect to the amount 
 of exercise recommended in the two varieties of the disease. 
 
 A considerable amount of exercise in the open air appears to 
 be of great service in the mild forms of diabetes, when the 
 patient's general condition is good, and especially in the case 
 of obese or gouty patients ; but over-exertion or undue strain 
 must be avoided. In other forms of diabetes only gentle 
 
368 THE TREATMENT OE DIABETES. 
 
 exercise can be recommended ; in severe cases it is especially 
 important to avoid much exercise, since any unusual strain 
 or vigorous muscular exercise would be liable to bring on 
 diabetic coma. 
 
 Massage is worthy of a more careful trial in diabetes, 
 especially when the patient is too stout, or too thin and weak, 
 to take much exercise. In these cases it is recommended 
 by Lauder Brunton, who points out that the flow of blood 
 through the muscles takes place three times as quickly under the 
 influence of massage. Brunton and Tunnicliffe ( 37 ) have also 
 shown that the total blood pressure is diminished. Grube ( 3S ) 
 has seen the best results from massage in cases of diabetes, 
 associated with arterio-sclerosis, and in such cases he thinks it is 
 specially indicated on account of its power of diminishing the 
 blood pressure. Balfe speaks favourably of massage, both 
 general and over the abdominal viscera ; he believes it improves 
 assimilation and promotes metabolism ; and when thoroughly 
 carried out be has " seen it effect a wonderful improvement 
 in the patient's general condition." But the most careful obser- 
 vations appear to have been made by Finkler eleven years 
 ago ( 39 ). His results show that massage alone, without any 
 restriction of diet or any medical treatment, is able to cause a 
 great diminution in the sugar excretion and diabetic symptoms. 
 Zander of Stockholm has employed regular Swedish gymnastics 
 in the treatment of patients, and though no cure has been 
 effected, considerable benefit is reported. 
 
 Teeatment by Alkali Mineeal Watees. 
 
 The mineral waters of certain continental spas have a great 
 reputation in the treatment of diabetes mellitus, and every year 
 these spas are visited by large numbers of diabetic patients from 
 all parts of Europe. The spas most frequented are Carlsbad, 
 Neuenahr, Vichy, Marienbad. The waters of Carlsbad and 
 Marienbad may be briefly described as alkaline sulphate water ; 
 those of Vichy and Neuenahr as alkaline bicarbonate waters. 
 
 As to the value of a visit to these spas, and as to the 
 value of the waters, there is much difference of opinion, and 
 the statements found in the writings of various authors are 
 somewhat conflicting. Whilst some writers state that it would 
 be much better and infinitely cheaper for patients to stay at 
 
ALKALI MINERAL WATERS. 369 
 
 home and have the water sent to them, others attribute any 
 good effects to the visit to the spa and not to the waters. 
 Whatever the explanation may be, there can be no doubt that a 
 large number of diabetic patients are greatly benefited by a visit 
 to Carlsbad and Neuenahr. Excluding the somewhat glowing 
 statements of many of the resident practitioners, we find that 
 many German and Austrian physicians, who have specially 
 devoted their attention to the clinical study of diabetes, and 
 who do not practise at any of these spas, testify as to the 
 undoubted benefit which many diabetic patients derive from a 
 visit to Carlsbad or Neuenahr. 
 
 Thus Frerichs ( 40 ) of Berlin stated that he had no doubt 
 that through the use of these waters the diabetic symptoms 
 diminish markedly, and sometimes disappear temporarily. Of 
 course no permanent cure is obtained ; but Frerichs stated that 
 he had often observed that every trace of sugar had been absent 
 for months, and then gradually returned. The beneficial influ- 
 ence he had seen mostly after the first visit. Then gradually 
 the good results diminished, and finally ceased. 
 
 Minkowski ( 41 ) of Strassburg states that in the milder forms 
 of the disease, frequently the sugar disappears from the urine 
 of patients at these spas, by the use of a less rigid diet than 
 would be necessary at home ; that the patients bear a restricted 
 diet better at the spas ; and that after returning home there is 
 frequently a greater tolerance for carbohydrates. Naunyn ( 42 ) 
 of Strassburg also bears similar testimony to the value of a visit 
 to Carlsbad in cases of slight or moderate severity, but admits 
 that in severe cases little benefit is derived. Seegen of Vienna 
 also speaks very highly of the value of Carlsbad waters. 
 
 Physicians such as Seegen and Kallay, who have had much 
 experience of the treatment of diabetes both in Carlsbad 
 and elsewhere, have obtained more successful results in 
 Carlsbad. 
 
 Mode of action. — There are other factors besides the drinking 
 of the waters which tend to improve the condition of the 
 diabetic patient during a visit to these spas. 
 
 (1) The patient is removed from his usual routine of life, 
 from the anxiety of business or professional cares ; he has com- 
 plete rest of mind as a rule ; and he lives a quiet, peaceful, and 
 regular life. it is well known, as has been pointed out on 
 j>. 365, that flu; disease is influenced by the mental condition. 
 24 
 
37o TBE TREATMENT OF DIABETES. 
 
 (2) He is in the open air a large portion of the clay. (3) He 
 takes suitable bodily exercise. (4) His diet is carefully regu- 
 lated, more carefully, as a rule, than when at home. (5) Further, 
 as Minkowski points out, the waters may act, not on the diabetic 
 state, but on the pathological condition which is at the bottom 
 of it, such as disease of the liver or pancreas. 
 
 Many physicians and physiologists attribute the improve- 
 ment in patients visiting Carlsbad and other spas entirely to 
 the above mentioned factors, and not to the influence of the 
 waters. Ktilz, Eiess, and Griesinger, Senator, and others 
 have tried the effect of Carlsbad water on patients in hospital, 
 and have not been able to detect any improvement by their 
 use ; also Carlsbad salts have not been of service to diabetic 
 patients at home. 
 
 Eiess ( 43 ), from observations which he has made respecting 
 the influence of Carlsbad waters, concludes that they are unable 
 to diminish the sugar excretion, apart from a nitrogenous diet, 
 in either slight or severe cases, and that in some cases they are 
 directly injurious. 
 
 Leichtenstern ( 44 ) states, however, that from his own experi- 
 ments he cannot agree with the conclusions of Eiess. 
 
 On the other hand, Seegen ( 45 ) asserts that he has observed 
 improvement in the mild forms of diabetes by the use of 
 Carlsbad waters, even when drunk at a distance; and in the 
 case of patients who cannot visit the continental spas, owing to 
 the expense, or for other reasons, he advises that two or three 
 times a year, for three or four weeks, the patient should take a 
 bottle of warmed Carlsbad water daily. 
 
 Whilst there is, therefore, considerable clinical evidence that 
 mild cases of diabetes do derive benefit from a visit to Carlsbad 
 and other spas, it is disputed how much of the benefit is to be 
 attributed to the waters, and how much to the altered conditions 
 of life and diet, etc. 
 
 The 'patients that derive benefit from a visit to these spas are 
 those who suffer from the milder form of the disease, especially 
 the obese and gouty. The severe cases with wasting derive 
 little or no benefit, but such cases do not derive much 
 benefit from any kind of treatment. Also a long journey is often 
 most injurious in the serious forms, and has frequently been the 
 exciting cause of diabetic coma. It is well to remember this 
 fact, and on no account to permit a patient suffering from 
 
TREA TMENT A T SPA S. 371 
 
 severe or advanced diabetes to take a long journey to Carlsbad, or 
 Neuenahr, or other distant spa. 
 
 Carlsbad. — The chief salt in the Carlsbad waters is sodium 
 sulphate, but the following are also present in smaller quan- 
 tities : sodium bicarbonate, sodium chloride, carbonate of calcium 
 and magnesium, etc. Seegen ( 45 ), who has had great experience 
 of the use of these waters, points out the following results which 
 he has observed in mild cases : — 
 
 1. Symptomatic improvement is soon obtained. The thirst 
 and dryness of the mouth diminish, the urine becomes less, and 
 therefore micturition during the night is less frequent, and the 
 patient sleeps better and feels stronger. These results have 
 been observed even when the sugar has not diminished. 
 
 2. In the majority of cases there is a real diminution of the 
 sugar excreted. This is most marked in the mild cases, but 
 even in many of the severe forms of the disease a diminution of 
 the sugar excretion occurs. 
 
 3. In many cases there is an increase in the body weight ; 
 but in the severe cases it remains the same. Seegen has 
 observed a diminution of body weight only in slight cases with 
 marked obesity. 
 
 4. The simple symptomatic improvement, without sugar 
 reduction, is never lasting, it disappears rapidly after the 
 termination of the treatment with the Carlsbad waters ; but in 
 the majority of the cases in which the sugar excretion is 
 diminished this improvement is more or less permanent. 
 
 5. Not infrequently, as a result of the Carlsbad treatment, 
 a greater tolerance of carbohydrates is established. 
 
 6. The improvement occurs without any regard to apparent 
 cause ; it is noted when the symptoms point to a brain lesion, 
 and also in cases in which there is a marked hereditary 
 tendency. 
 
 Indications and contra-indications of Carlsbad ivaters. — The 
 milder forms of the disease, especially in the obese and gouty, 
 are most benefited by a visit to Carlsbad. The waters are 
 unsuitable for the severe forms accompanied by great weakness 
 and wasting, and in such cases there is great danger of pro- 
 ducing diabetic coma by a long railway journey to Carlsbad. 
 According to some authorities, cases associated with nephritis 
 are also unsuitable. Seegen has seen improvement in cases com- 
 plicated with tuberculosis. 
 
372 THE TREATMENT OF DIABETES. 
 
 Carlsbad salts, natural and artificial, can be obtained and 
 taken at home very conveniently ; but, apart from their useful 
 purgative action, it has not been shown definitely that they are 
 of service in diabetes. 
 
 Carlsbad is situated in a beautiful part of the north of 
 Bohemia, and a visit may be made very enjoyable. But the 
 expenses are somewhat high, and the journey from England is 
 a long one. During the season, the little town is crowded with 
 patients from all parts of Europe, most of whom show their 
 faith in the waters by the great regularity with which they 
 follow out the treatment, and there can be no doubt that many 
 patients suffering from the milder forms of diabetes do return 
 home much improved. After a recent visit to Carlsbad, I do 
 not think that such improvement can be entirely attributed to 
 the climate and the more favourable condition of life, etc. 
 Carlsbad is situated in a valley, and is closely surrounded by 
 high hills. In summer the climate is frequently so hot and 
 relaxing, that most healthy persons are glad to leave after a short 
 visit.' Apart from the extra amount of time spent in the open 
 air, certainly there does not appear to be anything specially 
 favourable in the climate and conditions of life, and one can 
 hardly help coming to the conclusion that the improvement in 
 the patient's condition is partly due to the Carlsbad waters. 
 
 Neucnahr. — The waters of this spa are warm, and have the 
 great advantage of tasting pleasant. There can be no doubt 
 that they are of great value in relieving thirst. The chief solid 
 constituent is bicarbonate of soda; they also contain a very 
 small amount of magnesium and calcium carbonate, sulphate of 
 soda, sodium chloride, minute quantities of oxide of iron, silica, 
 aluminia, potassium, and lithium salts ; they contain, in addition, 
 free carbonic acid. 
 
 Indications and contra-indications. — B. Schmitz ( 46 ), who 
 practised at Neuenahr for many years, points out that the 
 waters of Neuenahr are of service chiefly in cases of glycosuria 
 complicating gout, the alkaline waters acting on the gouty 
 condition, to which it is probable that the glycosuria is secondary. 
 In other cases of diabetes they are also of service, and often 
 good results are obtained even when a strict diet is not pre- 
 scribed. Schmitz points that contra-indications to treatment 
 are considerable cardiac weakness, arterio - sclerosis with a 
 tendency to haemorrhages, which is not infrequently complicated 
 
TREA TMENT A T SPA S. 373 
 
 with glycosuria, and also great general weakness. He thinks 
 that Carlsbad and Vichy waters, on account of the large amount 
 of sulphate of soda contained in the former, and the large 
 amount of alkaline bicarbonate in the latter, are less suitable 
 than those of Neuenahr for weak patients. 
 
 Neuenahr is a beautiful little village situated in the valley 
 of the Ahr, to the west of the Rhine, not very far from Bonn. 
 It is more accessible than Carlsbad to patients from England 
 and the north of Europe. Apart from the intiuence of the 
 waters, the mode of life and surroundings are eminently 
 calculated to improve the general health. The place is very 
 quiet — for some patients it is too quiet ; but diabetics whose 
 illness has been brought on or aggravated by mental irritation 
 and anxiety, will here be able to lead a regular life free from 
 all excitement, in a healthy country village, with pleasant sur- 
 roundings and good hygienic conditions. 
 
 With military punctuality, the patients visit the Kurgarten, 
 and drink the water two or three times a day ; and from the 
 quantity which I saw drunk by the patients, during a visit to 
 Neuenahr a few years ago, it was difficult to believe that such 
 a large amount of bicarbonate of soda solution could be taken so 
 regularly without some effect, either for good or evil. 
 
 Vichy is much frequented by gouty and diabetic patients. 
 The water contains chiefly bicarbonate of soda, in addition to 
 smaller quantities of sodium chloride, calcium carbonate, 
 bicarbonate of potash, carbonate of magnesium, and sulphate 
 of soda. 
 
 Vichy water may be taken at home (3 to 6 oz.), half an 
 hour before each meal (Yeo). 
 
 The evidence in favour of the waters of the other spas is 
 even less decided. 
 
 The arsenical waters of La Bourboule (Puy de Dome, Trance) 
 are said to be of service, and are recommended in the severe 
 forms of diabetes. They can also be obtained at home, and a 
 minute quantity of arseniate of sodium is their most important 
 constituent ; they contain also free carbonic acid gas, chlorides 
 of sodium, potassium, lithium, and magnesium, with bicarbonates 
 of calcium and sodium, and sulphate of soda. 
 
 Seegen recommends the use of the arsenical mineral water 
 of Roncegus and Levico in the case of diabetes in children and 
 young people. 
 
374 THE TREATMENT OF DIABETES. 
 
 Medical Treatment of Diabetes. 
 
 The drugs which have been employed in the treatment of 
 diabetes mellitus are almost too numerous to mention, but only 
 a few have been proved to have any real beneficial influence. 
 Even in the cases in which the results have been most favour- 
 able, the action of the drugs has not been curative ; marked 
 improvement only has been produced. Nevertheless, to be able 
 to produce any improvement in the pa.tient's condition by drug 
 treatment is of great importance. 
 
 When we consider that diabetes is generally a chronic 
 disease, that apparently the symptoms are produced by an excess 
 of sugar in the blood or by some toxic substance, and that the 
 coma which so often terminates the disease is of the nature of 
 an intoxication — in other words, that in diabetes there is a 
 poisoning of the organism with chemical substances — treatment 
 by drugs appears particularly suitable ; yet the records of the 
 past have been disappointing. Nevertheless it does not seem 
 at all improbable that future research will lead to the discovery 
 of some internal remedy of great service. 
 
 The records of medical literature present the greatest 
 discrepancies as to the value of the various drugs which have 
 been used in the treatment of diabetes. A few drugs are 
 acknowledged by most writers to be of service ; a large number 
 are advocated by certain writers, but are stated to be useless by 
 the majority of observers. 
 
 It has been already pointed out that success in the dietetic 
 treatment of diabetes depends largely on the form of the disease, 
 the age of the patient, etc., and this is equally true with respect 
 to drug treatment. Too frequently these points have not been 
 considered in drawing conclusions as to the value of various 
 drugs. The records published in medical literature very 
 often omit to state details as to the form and duration of 
 the disease ; too often the results are those produced by re- 
 stricted diet plus the drug; and it is then, of course, impossible 
 to say how much of the improvement is due to diet, and how 
 much to the drug. To prove whether a certain drug will or 
 will not cure diabetes in any form and at any stage, is an easy 
 matter ; there is generally no difficulty in obtaining negative 
 evidence ; but to prove whether this drug has or has not any 
 beneficial influence is, as a rule, a most difficult task. Though 
 
MEDICAL TREATMENT 375 
 
 at first sight this might seem very simple, yet to anyone who 
 takes the trouble to study the natural course of the disease, and 
 the numerous sources of fallacy in drawing a conclusion, it will 
 be evident how useless is most of the literature with respect to 
 the drug treatment of diabetes. 
 
 The conclusions with respect to the action of many drugs 
 are simply examples of the post hoc propter hoc fallacy. A 
 number of drugs are stated by certain writers to have produced 
 marked improvement, whilst others state that they are useless 
 or injurious. Now, in many cases the change in the patient's 
 condition, or in the sugar excretion, has been purely accidental, 
 and due to variation in the diet, to altered surroundings, to the 
 course of the disease or other circumstances, and not to the 
 action of the drug. 
 
 It is important to remember, with respect to hospital 
 patients, that there is generally a marked diminution of sugar 
 excretion during the first three or four clays after admission, 
 even if no drugs be given and no special diet ordered. 
 
 Hence, in making scientific observations as to the action of 
 any drug on hospital patients, it is necessary to wait until this 
 equilibrium of sugar excretion has been produced before prescrib- 
 ing the drug, and to exclude the sugar excretion of the first four 
 days from consideration. 
 
 Probably many of the discrepancies with regard to the 
 action of various drugs may be owing to the fact that the 
 observations have been made by some practitioners in mild 
 forms or at early stages of the disease, and by others in 
 severe forms or in late stages. Then again, it is probable that 
 the pathological conditions at the foundation of the disease are 
 not always the same. In one case it may be that diabetes is 
 due to some change in the nervous system ; in another, to a 
 pancreas lesion, etc. If this be so, then these discrepancies in 
 the results of treatment are only natural. 
 
 Hirschfeld ( 47 ) states that in a number of careful observations 
 he has found that most of the drugs hitherto recommended cause 
 a temporary diminution of the sugar excretion, but this diminu- 
 tion is not important, and seldom amounts to more than 25 per 
 cent. Sometimes the diminution is followed by an increased 
 excretion, and the same drugs which cause a diminution of the 
 sugar in some patients cause an increase in others, whilst in a 
 third group they have no influence. 
 
3 7 6 THE TEE A TMENT OF EH ABE TES. 
 
 The records of the daily sugar excretion in a large number 
 of diabetic persons, who have been patients at the Manchester 
 Royal Infirmary, bear out the above statements with respect to 
 many of the drugs which have been employed in the treatment 
 of the disease in that hospital. These patients, however, have 
 generally been suffering from the most severe form of the disease, 
 often at an advanced stage. 
 
 The following is a summary of the evidence with reference 
 to the drugs which have been most frequently used in the treat- 
 ment of diabetes. 
 
 Opium and its alkaloids. — Opium has been long employed. 
 Dobson, in a paper published in 1779, mentioned amongst the 
 drugs used in the treatment of the disease, Dover's powder and 
 tincture thebaica. According to Dickinson, opium was first 
 recommended in 1812 by Warren, who gave large doses of 
 the drug with benefit. During the last fifty years it has been 
 extensively used ; and M'Gregor, Pavy, Seegen, Frerichs, and 
 numerous other writers speak of the great value of this drug 
 and its alkaloids. Years ago, Pavy ( 4S ) showed clearly how 
 useful these drugs were in diminishing the sugar excretion. 
 jSTo one, of course, regards opium or its alkaloids as direct 
 curative agents, except perhaps in the mildest form of the 
 disease ; but the evidence in their favour is greater than that in 
 favour of any other drug. Sometimes, in the milder form of 
 the disease, by this treatment, together with restricted diet, the 
 glycosuria disappears temporarily, though diet alone is not 
 sufficient to remove the sugar from the urine. But in most 
 cases only a diminution of the symptoms is obtained. 
 
 The effects of opium and its alkaloids are — 
 
 1. To diminish thirst. 
 
 2. To diminish the appetite. 
 
 3. To reduce the quantity of urine. 
 
 4. To diminish the amount of sugar excreted in the urine. 
 
 5. Sometimes to increase the weight, and to improve the 
 general condition of the patient. 
 
 6. To diminish nervous irritability. 
 
 Owing to the diminished excretion of urine, produced by the 
 drug, sleep is undisturbed by frequent micturition. 
 
 In some cases, opium, like every other remedy, is useless ; 
 but in other cases benefit is obtained, even when the diet is not 
 restricted. It is important to remember, however, that often 
 
0P1 UM AND ITS ALKALOIDS. 3 7 7 
 
 large doses of the drugs are necessary to produce good results. 
 Diabetic patients are very tolerant of opium or its alkaloids, 
 and can take large quantities without narcotism or any bad 
 effects being produced ; after a time, often 2, 4, or 5 grs. of 
 opium can be taken thrice daily. 
 
 It is best to begin with half a grain of opium, or of the 
 extract, three times a day, and then gradually to increase the 
 quantity; or the compound soap pill of the pharmacopoeia 
 (grs. 5 = 1 gr. of opium) may be given twice a day, and gradually 
 increased. 
 
 If the sugar can be removed from the urine completely by 
 diet, it is, of course, not necessary to give opium ; but when diet 
 alone is insufficient to do this, then opium or its alkaloids ought 
 to be tried. 
 
 Ealfe ( 49 ) has paid special attention to the treatment of 
 diabetes by opium and its alkaloids. From his own observations 
 he concludes that the most decided results are obtained when 
 opium is administered by the mouth. With regard to the best 
 time for administration — whether immediately before meals or 
 during digestion — he has found that when taken about an hour 
 after a meal it has a greater effect in restraining diuresis than 
 when taken on an empty stomach ; that not much difference is 
 effected on the sugar secretion ; but that the dose taken shortly 
 after food has the advantage of not deranging the stomach or 
 causing dyspeptic symptoms. 
 
 Ealfe prefers morphia to codeia, but states that the best 
 results are obtained when some preparation of crude opium is 
 added to either alkaloid. Thus liquor opii may be combined 
 with acetate of morphia. Diabetic patients often exhibit 
 iuclividual peculiarities as regards the different preparations of 
 opium. Ealfe mentions a case where codeine and morphine 
 had to be given up on account of the headache and giddiness 
 which they produced, whilst solid opium, in the form of com- 
 pound soap pill, was taken without discomfort. In fixing the 
 dose, it is important to remember that each patient has his own 
 capacity for the drug. Ealfe thinks that as regards dose we err 
 on the side of too much caution, and that as soon as the glycosuria 
 cea ies to be controlled by diet, opium should be given in doses 
 that sensibly affect the excretion of sugar, and should be 
 increased, until either it entirely controls the glycosuria, or 
 until no further reduction in the amount of sugar is obtained on 
 
373 THE TREATMENT OF DIABETES. ' 
 
 increasing the close. When this latter poiDt is reached, then it 
 is not wise to attempt to further increase the amount of opium. 
 Ealfe points out the danger of suddenly reducing the dose, 
 especially when opium has been administered for some time, in 
 advanced cases. After opium has been administered for a long 
 period, its good effects cease. 
 
 It is not known how opium acts in diminishing the diabetic 
 symptoms. Minkowski thinks that perhaps it inhibits the 
 formation of sugar from albumin. Sir William Eoberts attributes 
 the good influence of opium to its action in diminishing the 
 appetite, and as a consequence less sugar is excreted in the 
 urine. Also, it may produce good effects by inducing sleep and 
 allaying painful sensations and irritability. 
 
 The effects of opium ought to be carefully watched, how- 
 ever. It is liable to cause obstinate constipation and dyspeptic 
 troubles, the tongue may become thickly coated, and the 
 patient may suffer from epigastric pains. It is sometimes 
 necessary to discontinue its use on account of these symptoms. 
 In cases in which coma is threatening, it is well to avoid the 
 use of opium. In these cases, as a rule, there is constipation, 
 which probably has some indirect connection with the develop- 
 ment of coma, and by the use of opium the constipation is only 
 made worse. Also, in cases complicated with nephritis, opium 
 should be avoided or only given with caution. 
 
 As regards the form of the opium preparation, Frerichs, Sir 
 William Eoberts, Ealfe, Striimpell, and many others prefer crude 
 opium ; Bruce, Osier, and many observers prefer morphia, whilst 
 Pavy is in favour of codeine. 
 
 Morphine. — Kratschmer ( 50 ), Mitchell Bruce ( 51 ), and others 
 have clearly shown the value of morphine in reducing the 
 amount of sugar in the urine in cases of diabetes. Bruce found 
 that it acted much better when given by the mouth, than when 
 injected hypodermically. In two cases he compared the action 
 of acetate of morphine and phosphate of codeine, and found that 
 the former was distinctly more powerful in reducing the 
 glycosuria. Morphine is also much less expensive than codeine. 
 In the cases reported by Bruce, the patient was put on a rigid 
 diet, so that the results are due to the diet plus the drug. 
 
 If morphine be employed, it is well to begin with a small 
 dose, one-sixth of a grain three times a day, and gradually to 
 increase the amount. In course of time, diabetic patients are 
 
OPIUM AND ITS ALKALOIDS. 379 
 
 often able to take 1 gr. three times a clay, or even larger closes, 
 without any bad effects. 
 
 Codeine. — Pavy and other observers prefer codeine to opium 
 or morphine ; but many think that it is inferior to both. 
 Codeine should be given at first in small doses, half a grain 
 three times a day, and gradually increased. In this way 4 
 or 5 grs. are often taken three times a day, without any 
 bad effects. The advantages of codeine over morphine and 
 opium are, that it causes less constipation, that it is less liable 
 to derange digestion, and that it does not cause so much clrowsi- 
 ness. But against these advantages it is stated by Lauder 
 Brunton, Bruce, and others, that morphia is more powerful in 
 diminishing the amount of sugar excreted. Pavy, however, 
 thinks that it is quite as efficacious as opium or morphia. 
 
 In a number of cases which have come under my observation, 
 codeia has given fairly good results. I have obtained better 
 results with opium or cocleia than with any other drug. 
 
 Narcotine and narceine were found to be useless by Pavy. 
 
 Alkalies have long been used in the treatment of diabetes, 
 and numerous cases have been recorded in which such treatment 
 has appeared to be of service. Pavy, writing in 1862, stated 
 that an alkaline treatment was that which had perhaps received 
 most favour. 
 
 With reference to the alkaline treatment, it is interesting to 
 note that the urine is generally markedly acid in diabetes, and 
 that diabetic coma has been attributed to an intoxication with 
 organic acids. 
 
 Alkalies may be given in the form of the mineral waters, 
 which have been already referred to, or they may be given as 
 medicine in the usual way. The salts which have been chiefly 
 used are the bicarbonate, carbonate, or acetate of sodium, 
 the bicarbonate, carbonate, tartrate, citrate, or acetate of potash, 
 carbonate of ammonia, and carbonate of lithium. Good effects 
 can only be expected when very large closes are given. The 
 sodium salts are to be preferred, owing to the toxic action of 
 potash salts in large doses, and bicarbonate of soda is the salt 
 which has been most employed. If the patient be gouty, the 
 carbonate or citrate of lithium will be more suitable. The dose 
 of bicarbonate of soda ought to be large. Eichardiere ( r ' 2 ) 
 recommends oO to 1.50 grs. in the twenty-four hours, but 
 some writers recommend very much larger doses. 
 
3 So THE TREATMENT OF DIABETES. 
 
 In severe cases the alkaline treatment is useless. I have 
 never seen improvement in this form of the disease which could 
 be fairly attributed to alkaline treatment, except in cases of 
 commencing coma (these cases will be referred to later). In 
 mild forms of the disease there is a considerable difference of 
 opinion as to the value of alkalies. Some writers believe that 
 in these cases they have a distinct influence in diminishing the 
 sugar excretion. 
 
 Eichardiere has recently written strongly in favour of the 
 alkaline treatment (sodium bicarbonate) in mild cases. He 
 thinks it is better to give an alkaline course for two or three 
 weeks every three or four months instead of continuous alkaline 
 treatment. He believes that alkalies act by altering the general 
 nutrition. Among contra-inclications of the sodium bicarbonate 
 treatment, he mentions pulmonary tuberculosis, cachexia, and an 
 advanced stage of the disease. 
 
 Lime salts were employed long ago in the treatment of 
 diabetes. Willis, writing in 1679, mentions a diabetic. patient 
 who recovered from the disease, and who was treated with lime- 
 water, among other drugs. Quite recently, Kail Grube ( 53 ) has 
 pointed out the value of calcium salts. 
 
 Robin also recommends glycerophosphates of lime and 
 magnesium to be given with the meals, in order to counteract 
 the great loss of the phosphates of magnesium and calcium 
 which occurs in diabetes (see p. 386). 
 
 Arsenic has frequently been given in diabetes during the last 
 twenty years, and from time to time papers have been published 
 drawing attention to its value. Experiments on animals have 
 shown that the administration of arsenic, in sufficient doses and 
 for an adequate time, will cause the glycogen to disappear from 
 the liver, and puncture of the floor of the fourth vertricle is 
 then no longer followed by diabetes. Hence there appears to be 
 some theoretical evidence in favour of its use in diabetes ; but 
 whilst a number of observations have been published, recording 
 improvement, a number have also been published showing 
 negative results. Murray ( 54 ), a short time ago, drew attention 
 to the value of arsenic. He believes that after the sugar has 
 been reduced by diet and codeia, arsenic will often effect a cure.' 
 Murray's patients appear to have suffered from a mild form of 
 the disease, however. Arsenic may be given in the form of 
 liquor arsenicalis, and the dose gradually increased. Murray 
 
LIME SALTS— ARSENIC— JAMB UL. 381 
 
 recommends the doses to be increased until 1 minims of liquor 
 arsenicalis are given three times a day. 
 
 Frerichs has tried hypodermic injection of arsenic, but has 
 found it useless. 
 
 Clemens solution, arsenite of bromine, has been largely 
 employed in the treatment of diabetes, in doses of 3 to 5 
 minims once or twice a day after meals, but the evidence in its 
 favour is not convincing. By many observers it has been 
 found useless. 
 
 A combination of lithium and sodium arseniate, dissolved 
 in aerated water, has been much used in France, especially 
 in gouty cases. Dujarclin-Beaumetz recommends 5 grs. of 
 lithium carbonate, and 2 minims of Fowler's solution of arsenic 
 in a glass of Vichy or other alkaline water. But extensive trial 
 by others has failed to produce much evidence in favour of this 
 treatment. 
 
 I have given arsenic a fair trial both in severe and mild 
 forms of the disease, and have seen it tried in numerous cases 
 at the Manchester Eoyal Infirmary, but cannot say much in its 
 favour. Probably it is useless in the severe forms of the disease, 
 but in the milder forms it is worthy of trial, after a restricted 
 diet and opium preparations have been employed. 
 
 Janibul. — The seeds of Syzygium jamoolanum have been 
 recommended for diminishing the sugar excretion in diabetes, 
 and a few years ago the drug was pretty extensively tried, but 
 recently it appears to have been little used. It may be 
 administered in powders, cachets, pills, or as a liquid extract. 
 The dose of the latter is given as a half to 2 clrms. ; that of the 
 powder, 5 to 30 grs. 
 
 In numerous cases the use of the drug has not been followed 
 by any good results. In experimental pancreatic diabetes, 
 Minkowski found that its administration did not produce any 
 effect on the sugar excretion. 
 
 Lewaschew ( 57 ) states that, after giving a sufficient dose 
 of 20-0 to 40-0 grms. daily (about 300 to 600 grs.), he has 
 always found a decrease in the amount of urine and sugar, 
 of the thirst and other diabetic symptoms. He thinks that 
 the frequent failure of the drug is owing either to the dose 
 being too small, or to the impure condition of the drug. 
 
 In a case, under the care of Dr. Steell, which I had the 
 opportunity of watching for a long time in the Manchester Koyal 
 
382 THE TREATMENT OF DIABETES. 
 
 Infirmary, the sugar excretion diminished markedly under the use 
 of jambul, but the patient's general condition became gradually 
 worse as the sugar diminished, and the case terminated fatally. 
 
 Antipyrin has been frequently employed in the treatment of 
 diabetes during the last seven years. In closes of 30 to 60 grs. 
 daily, according to Dujardin-Beaumetz ( 5S ), it diminishes the 
 quantity of urine without increasing the percentage of sugar per 
 litre. Eobin ( 59 ) also employs antipyrin in his " alternating " 
 treatment, which will be subsequently described. If the gly- 
 cosuria be not modified in a few days, it is useless to continue 
 the drug. The action of antipyrin is only temporary ; it ought 
 not to be administered for a long period, as it is liable to give 
 rise to digestive troubles and temporary albuminuria. Many 
 observers, however, have found antipyrin useless. I have found 
 it apparently of service in relieving the pains in the limbs in 
 diabetes, but have not been able to detect real improvement 
 otherwise. 
 
 The evidence in favour of antipyrin is not therefore very 
 conclusive. Still, it is one of the drugs which are worthy of a fair 
 trial when diet and opium fail. 
 
 Sodium salicylate has been recommended, especially by 
 Ebstein ( 60 ), and during the last ten years numerous instances 
 of improvement, and diminution of sugar excretion under the use 
 of this drug have been reported, whilst many cases have also 
 been recorded in which it has appeared to be useless. 
 
 Ebstein states that its influence is greatest in recent cases. 
 Often, however, as Fiirbringer ( 61 ) has shown, it has no influence 
 on the glycosuria, but it reduces the nitrogenous excretion, and 
 in this way may be beneficial. 
 
 Ebstein recommends large doses, 75 to 150 grs. in the 
 twenty-four hours. I have never given it or seen it given in 
 such large doses, but in the usual doses, 10 to 20 grs. three 
 times a clay, I have not found any benefit in severe cases. 
 Patients sometimes prefer the drug to other remedies, however, 
 and state that they feel better when taking it. 
 
 Brunton and Ealfe think it is useful chiefly in the glycosuria 
 of gouty persons. 
 
 It is important to remember that the urine of patients 
 taking sodium salicylate gives a purplish brown coloration 
 with perchloricle of iron, which is liable to be mistaken for 
 Gerhardt's so-called diacetic reaction (see p. 181). 
 
SODIUM SALICYLATE, ETC 3S3 
 
 Salicylate of bismuth. — Sclimitz ( 62 ) of Neuenahr, who had 
 great experience in the treatment of diabetes, has recorded 
 four cases which show the value of salicylate of bismuth in 
 causing the sugar to disappear from the urine in mild forms of 
 the disease. He gives 5 decigrammes (about 7h grs.) twice 
 a day in powder. It is the only drug from which he has 
 obtained marked results. I have given it a fair trial in a very 
 mild form of diabetes, but without any good effect. 
 
 Potassium bromide is thought to be of service by v. Noorden, 
 Osier, Saundby, and other physicians, in cases in which there is 
 great nervous irritability or excitement. Owing to the depress- 
 ing action of the potassium salt, in many cases it would be 
 better to give sodium bromide, 15 to 20 grs. well diluted, or 
 lithium bromide in 5-gr.- closes. The bromides may be given 
 alone or in combination with opium, in the class of cases 
 mentioned. 
 
 Uranium nitrate. — Leconte long ago (in 1857) discovered 
 that by the prolonged administration of this substance glyco- 
 suria was produced in animals. Hughes found that when given 
 to diabetic patients the symptoms were diminished, but the drug 
 does not appear to have had a fair trial, and its use was dis- 
 continued until 1895, when West (° 3 ) recorded three cases, and 
 referred to others in which uranium nitrate apparently caused a 
 marked diminution of the sugar excretion. He commences the 
 treatment with a small dose, 1 to 2 grs. freely diluted with 
 water, twice a day after meals. He increases the amount at 
 intervals of a few days until its effects are produced. When 
 given in this way no digestive disturbances are produced, and no 
 albuminuria occurs, and the close may be increased in some 
 cases up to 15 or 20 grs. three times a clay. 
 
 According to West, the effects of the drug are (1) to 
 diminish thirst ; (2) to reduce the amount of urine ; (3) to reduce 
 the percentage of sugar. He points out that in giving the drug 
 it is well to discontinue it every three or four weeks, for a 
 few days, and after that time to resume it. In 1896, West ( 63 ) 
 reported five other cases in which favourable results were 
 obtained. He concludes that we have in uranium nitrate a drug; 
 of considerable value in diabetes, though it cannot be relied upon 
 to produce equally good results in all cases indiscriminately. 
 
 Quinine sulphates has often been given as a general tonic, but 
 it has not been shown to have any other action. It is worthy of 
 
384 THE TREATMENT OF DIABETES. 
 
 careful trial when the symptoms have followed malaria, or when 
 the patient has lived in a malarial district. 
 
 Iodide of 'potassium was carefully tried many years ago by 
 Dickinson ( 64 ). He found that a remarkable diminution of the 
 sugar excretion occurred when the drug caused loss of appetite 
 and general depression, but when the appetite was scarcely 
 affected there was very little change in the excretion of sugar 
 or urea. He therefore concludes that the diminution of the 
 sugar excretion which is sometimes produced by potassium 
 iodide " is not the result of any mitigation of the disease, but of 
 loss of appetite and general depression of function." He "has 
 never found the saccharine diminution thus caused to be per- 
 manent, nor been able to trace to it any amelioration in the 
 general condition of the patient." 
 
 Anti-syphilitic treatment. — It has been already pointed out 
 that a history of acquired syphilis is not common in diabetes. 
 Certainly, in the greater proportion of cases, acquired syphilis 
 is not the cause of the disease, but it is probable that in a 
 few rare instances syphilis may produce diabetes or glycosuria, 
 either by giving rise to changes in the nervous system, by 
 causing disease of the pancreas (see p. 113), or in some other 
 way. When, therefore, a history of syphilis is obtained, and 
 when there are indications of a syphilitic affection of the 
 nervous system or of other organs, anti-syphilitic treatment 
 ought to be prescribed. Iodide of potassium may be given in 
 10 gr. doses with aromatic spirits of ammonia, and mercurial 
 inunctions should also be employed. A few cases are on record 
 of the mild forms of diabetes, accompanied by syphilitic lesion 
 of the nervous system, in which anti-syphilitic treatment 
 caused great improvement. Feinburg ( 65 ) has reported such 
 cases ; and v. Noorden refers to twelve cases of diabetes preceded 
 by syphilis, in two of which decided and permanent improve- 
 ment was obtained by the use of mercury and potassium iodide ; 
 but in no case did complete recovery follow, v. Noorden ( C6 ) 
 points out, however, that in several cases fatal complications 
 occurred during the mercurial course — in one case, gangrene 
 of the foot ; in two, haemoptysis and rapid progress of pulmonary 
 tuberculosis. Hence the patient ought to be seen daily 
 and carefully watched, as mercurial stomatitis and dysenteric 
 intestinal catarrh are very liable to develop. 
 
 Iodoform has been strongly recommended by Moleschott, 
 
ANTISYPHILITIC TREATMENT. 385 
 
 and a number of cases have been reported which have improved 
 under the use of this drug. Frerichs confirms the favourable 
 reports ; he has found a moderate temporary improvement. In 
 mild cases the sugar disappeared, in severe cases it was di- 
 minished, but no actual cure was obtained, and the improvement 
 was not permanent. On the other hand, it has often been 
 found useless. 
 
 Oxygen inhalation has been highly spoken of, and in 
 mild cases is said to have caused the sugar to disappear when 
 the diet has been restricted, though diet alone and ordinary 
 treatment have caused no improvement ( 67 ). 
 
 Purdy ( 68 ) recommends 3 to 5 gallons of oxygen to be 
 inhaled twice daily, morning and afternoon, and has obtained 
 " the best results " thereby — what the exact results were is 
 not stated. 
 
 Lactic acial has been highly recommended by Cantani ( 69 ), 
 but others have not found it of service. Cantani prescribed 
 it in the form of lemonade ; 5, 15, or 20 grms. of lactic acid 
 are dissolved in 1 litre of water, and a little aromatic water, 
 such as peppermint or anise, is added. The patient is allowed 
 half a wine glassful of this mixture with \ grm. of bicarbonate 
 of soda every hour, or every two hours. 
 
 According to Cantani, pure lactic acid, given in .water 
 directly after a meal, aids the digestion of nitrogenous food, and 
 enables the patient to take a rigid nitrogenous diet for a long 
 period without the occurrence of the diarrhoea or gastro-intestinal 
 disturbances which are so liable to be produced. 
 
 Benzosol has been highly spoken of, and appears worthy of 
 further trial. 
 
 Methylene Hue. — Eecently Marie and Le Goff ( n and 72 ) have 
 employed this substance in the treatment of three cases of 
 diabetes. In two of the cases there was a gradual diminution 
 of the sugar, and an improvement in the general condition. 
 
 Pilocarpine injections have apparently caused a diminution 
 of urine and of the sugar excreted in some cases, but in other 
 cases they have been found useless. 
 
 Numerous other drugs, in addition to those referred to, have 
 been recommended from time to time by various physicians, but 
 the evidence in favour of their use has often been simply of the 
 post hoc propter hoc kind, and, on careful trial by others, generally 
 no good results have been obtained. Amongst the drugs of this 
 2 5 
 
386 THE TREATMENT OF DIABETES. 
 
 class may be mentioned chloral, carbolic acid, sulphonal, tincture 
 of iodine, phosphoric acid, benzoic acid, sodium benzoate, salol, 
 belladonna, phosphorus, strychnine, valerian, Calabar bean, picric 
 acid, sodium phosphate, peroxide of hydrogen, guaiacol, calcium 
 sulphide, creasote, camphor, nitro-glycerine, salicin, sulpho-car- 
 bolate of soda, cocaine, ouabain, cannabis indica, ergot, etc. 
 
 Kobin ( 73 ) has recommended an "alternatiug method" of 
 treatment. (1) He first prescribes antipyrin in 15 gr. doses along 
 with 7-| grs. of bicarbonate of soda in the form of a powder, to 
 be taken twice a day — one hour before breakfast and dinner. 
 He also prescribes cod-liver oil, and orders the bowels to be kept 
 regular by the use of saline aperients. (2) On the fourth or 
 fifth day he changes the medicine, and prescribes about 6 grs. of 
 quinine sulphate at the mid-day meal. This is taken for six days, 
 then discontinued for four days, and afterwards again taken for 
 six days. Twice a day he also gives a cachet containing arseniate 
 of soda, carbonate of lithium, and codeia. (3) After fifteen 
 days the above treatment is discontinued, and valerian, opium, 
 and belladonna are prescribed together in the form of a pill, or 
 potassium bromide is given. Cod-liver oil is discontinued, and 
 the patient is allowed a weak alkaline solution. To counteract 
 the loss of phosphates of magnesium and calcium, he recommends 
 the glycero-phosphates of lime and magnesium in cachets at each 
 meal. 
 
 If sugar is still present in the urine, the course is commenced 
 again, but, after the second course, diet only is trusted to. 
 
 Eobin has treated 100 cases by this alternating method, in 
 each of which the daily quantity of sugar excreted was 100 
 grms. or more. In twenty-four of these recovery has occurred, 
 i.e. for six months at least after treatment there has been no 
 reappearance of sugar in the urine. In twenty-five cases 
 recovery is still doubtful. Sugar has disappeared under treat- 
 ment, but has returned under the influence of unsuitable diet, 
 mental excitement, etc. In thirty-three cases there has been 
 considerable and permanent improvement. In eighteen cases 
 the results have been negative. 
 
 Glycerine was recommended years ago by Schultzen, but it 
 has not been found of any service by others. 
 
 Cocl-liver oil is of service, especially in those cases in which 
 the patient is wasted. Whilst it is often prescribed by some 
 medical men, it is somewhat remarkable that its use has 
 
SACCHARINE TREATMENT. 387 
 
 not become more general in this class of cases. Lipanin may 
 be given in place of cod-liver oil to those who object to the 
 taste of the latter. Peach kernel oil (free from prussic acid) 
 may also be used. It can be made into a pleasant emulsion 
 with eggs, sweetened with saccharine, and flavoured with cinna- 
 mon. Petroleum emulsion is also worthy of a trial (see p. 338). 
 
 Pancreatic emulsion, 1 to 3 drms. in a little water and 
 spirits, may be given once or twice a day, one or two hours after 
 food, in cases in which there is much wasting. 
 
 Pepsin and rennet were fairly tried years ago, but were 
 found to be useless by nearly all observers. 
 
 The action of electricity has been tried in diabetes, but with- 
 out success. 
 
 Saccharine treatment. — Many years ago, it was believed by 
 Piorry and others that the symptoms of diabetes were due to 
 the loss of sugar by the kidneys, and it was thought that 
 possibly relief might be obtained by increasing the amount of 
 sugar in the diet in order to compensate for the loss. This 
 method of treatment received a fair trial, and was found by nearly 
 all observers to be injurious. The theory on which this treat- 
 ment is based is of course erroneous, still nothing is more to be 
 desired in the treatment of diabetes than the discovery of some 
 form of sugar or other carbohydrate which the organism is 
 capable of utilising, and which may therefore be given in place 
 of the carbohydrates of food. 
 
 Lasvulose and lactose in certain cases can be burnt up in 
 the system, and may thus be used in moderation, in the diet of 
 such diabetic patients, to replace the injurious carbohydrates 
 (see p. 340). 
 
 Lepine's glycolytic ferment. — Lepine ( 74 ) has recorded the 
 result of the treatment of four cases of diabetes mellitus with a 
 glycolytic ferment prepared from the diastase of malt. In all 
 four cases there was a distinct improvement. As regards the 
 sugar excretion, in the first case, before treatment the mean 
 quantity for the twenty-four hours was 140 grms., under the 
 influence of the ferment it was 70 grms.; in the second case the 
 sugar was reduced from 41 to 11 grms. ; in the third case, from 
 116 to 80 grms.; in the fourth case, from 257 to 124 grms., 
 but later only to 163 grms. The ferment is not diuretic, it has 
 no injurious effects, but the improvement is only temporary. 
 
 Yeast in a very old remedy for diabetes, which was re- 
 
388 THE TREATMENT OF DIABETES. 
 
 commended many years ago, but it has not met with any 
 definite success. Eecently, I have seen a marked case with 
 wasting in a young man, under the care of Dr. Steell, in which 
 improvement occurred in the general condition when the usual 
 opium treatment was discontinued and yeast taken. The dose 
 was two dessert-spoonfuls of fresh barm, mixed with water to 
 such a consistency that it could be easily drunk. The improve- 
 ment was not permanent, however, and he finally died of coma. 
 
 I have heard of several other patients who have improved 
 under the yeast treatment. The above dose can be readily 
 taken, and is not particularly unpleasant. It is important to 
 obtain fresh yeast, and to take only the frothy part ; any fluid 
 settling to the bottom of the vessel ought to be rejected. If 
 the barm be not fresh, and if the fluid part be taken, severe 
 diarrhoea may result. 
 
 Cassaet ( 75 ) a short time ago reported good results from the 
 use of brewer's yeast. 
 
 Treatment by pancreatic preparations. — After the publication 
 of the brilliant results of the experiments of Minkowski 
 and v. Mering, de Dominicis, and others, on the relation of 
 the pancreas to diabetes (see p. 73), it was only natural 
 that those interested in the treatment of the disease should 
 have had the best hopes that the knowledge acquired would 
 be of great importance in therapeutics. The excellent results 
 which have attended the use of thyroid extract and other 
 thyroid preparation in the treatment of myxcedema, led to 
 the trial of various pancreatic preparations in diabetes. Un- 
 fortunately the expectations have not been realised. In the 
 first place, it was shown by the experiments of de Dominicis, 
 Minkowski, Thiroloix, and others, that in the diabetes produced 
 in dogs by the total extirpation of the pancreas, the sugar 
 excretion in the urine was not diminished by injections of 
 pancreatic infusion, whether the injections were subcutaneous, 
 intravenous, or intraperitoneal. Also by feeding the animal on 
 pancreas no benefit was derived. 
 
 Clinically, it has been found in most cases that pancreatic 
 preparations have been quite useless ; in a few cases some 
 improvement has followed their administration, but whether it 
 has been really due to the pancreatic preparations or to some 
 other cause, remains to be decided by future observations. 
 
 Pancreatic juice has been given by Mansell Jones ( 76 ) and H. 
 
PANCREATIC PREPARATIONS. 389 
 
 Mackenzie ( 77 ), but there is no evidence that it caused any real 
 improvement. Pancreatic extract has been tried with similar 
 results by H. Mackenzie and P. Watson Williams ( 78 ). 
 
 Patients have been fed on raw pancreas, but the records 
 published by Hale White ( 79 ) do not furnish any conclusive 
 evidence in favour of this treatment. Liquor pancreaticus has 
 been injected subcutaneously by Hale White, but no real benefit 
 has followed. 
 
 Cerenville ( 80 ) has reported the results of the pancreatic 
 treatment in five cases of diabetes. Four were fed on chopped 
 sheep's pancreas in doses of 4 to 10 to 30 grms. daily. In three 
 of these cases no improvement was detected ; in one there was 
 slight temporary improvement. The fifth case was treated with 
 pancreas extract, injected hypodermically, but no improvement 
 followed. Lightly cooked calf's pancreas was given by Ausset ( 81 ) ; 
 marked improvement followed, but only one case is recorded. 
 Bormenn ( 82 ) has given pancreatic preparations per rectum ; the 
 patient improved greatly, but again only one case is recorded. 
 Various pancreatic preparations have been injected subcutaneously, 
 and given in other ways, on the Continent, but, as a rule, without 
 benefit. Experiments have shown that it is not the absence of 
 pancreatic juice in the intestinal canal which causes diabetes in 
 dogs after extirpation of the pancreas, hence it is not surprising 
 that in most of the cases in which the pancreatic treatment has 
 been tried clinically, the results have been negative. The 
 prospects of success by injections of pancreatic extract appeared 
 more promising theoretically, but clinical experience has shown 
 them to be of no value. 
 
 Minkowski, Hedon, and others have shown that a graft of 
 pancreatic tissue under the skin of the abdominal wall is able 
 to prevent the occurrence of diabetes in dogs, when the whole 
 of the pancreas has been removed from the abdominal cavity 
 (see p. 76). In 1893 I suggested the possibility of treating 
 diabetes successfully by grafting a piece of the pancreas of 
 one of the lower animals under the skin of the abdominal 
 wall in pancreatic diabetes in man ( 83 ). About that time 
 I was treating a case of diabetes in a young man, and I was 
 strongly inclined to try the effect of grafting a piece of 
 sheep's pancreas under the patient's skin. A surgical friend 
 promised to perforin the operation, but I hesitated to advise it, 
 (1) because I could not be certain that the diabetes had a 
 
39° THE TREATMENT OF DIABETES. 
 
 pancreatic origin in this case, and (2) I was afraid that the 
 operation might lead to diabetic coma. The patient improved 
 somewhat ; he returned to his work, and after a few months 
 discontinued all treatment, except that he avoided certain 
 articles of diet. Eighteen months later he again came under 
 my care, and finally died of asthenia and phthisis. I obtained 
 permission to make a post-mortem examination at his home, and 
 removed the pancreas. On examination, I found it normal 
 microscopically and microscopically. 
 
 Of course it is possible, but not probable, that a graft of 
 pancreatic tissue under the skin might have been of service in 
 this case, though no pathological changes were detected in the 
 gland post-mortem ; but when the pancreas is normal, the opera- 
 tion does not seem to be indicated. 
 
 The operation of grafting pancreatic tissue in man has since 
 been performed, however. Williams ( 84 ) of Bristol had pieces of 
 sheep's pancreas grafted under the skin of the breast and 
 abdomen in a case of diabetes. Unfortunately, however, diabetic 
 coma developed on the third day after the operation, and termin- 
 ated fatally. 
 
 The two difficulties with respect to the treatment by sub- 
 cutaneous pancreatic grafts are therefore (1) the absence of 
 definite clinical indications which will enable us to diagnose 
 with certainty when diabetes is associated with pancreatic 
 disease ; (2) the risk of producing diabetic coma by the operation. 
 
 The grafting of pieces of pancreas subcutaneously appears, 
 however, to be the only form of pancreatic treatment from which 
 success can be expected (judging from the results of experiments 
 on animals) ; and it is possible that, in spite of the two diffi- 
 culties above mentioned, the treatment may yet be found to be 
 of service in certain selected cases of diabetes in man. 
 
 Hepatic extract. — Gilbert and Carnot ( 85 ) have recently tried 
 the effect of hepatic extract prepared from pig's liver, and given 
 per rectum as an enema. The results varied somewhat, but 
 these observers conclude that it caused a diminution of the 
 excretion of glucose. 
 
 The drugs, therapeutic agents, and methods of treatment 
 which have been referred to, are those which have been chiefly 
 used, or which have been found to be of most service. Numerous 
 others have been recommended, but it is quite impossible to 
 mention all. Neither is it possible to give more than a few 
 
PANCREA TIC PRE PA RA T10NS. 3 9 1 
 
 references to the various papers on the methods of treatment 
 described. But since so many of these methods have not been 
 attended with any benefit, the omission is scarcely to be regretted. 
 
 In the treatment of diabetes, attention to diet is the most 
 important point, especially the careful regulation of the diet 
 according to the form of the disease and condition of the 
 patient. Attention to the general hygienic conditions is also 
 of importance. If the patient be wealthy, if he be suffering 
 from a mild form of the disease, and if he be in a condition 
 suitable for travelling, a visit to Carlsbad or Neuenahr may be 
 attended with some benefit, whatever may be the exact cause 
 of the improvement. But the greatest caution should be 
 observed in recommending a visit to these spas, and the cases 
 should be very carefully selected. 
 
 As regards drugs, those which appear to be most deserving 
 of trial are opium, codeia, morphia, arsenic, antipyrin, sodium 
 salicylate, bismuth salicylate, jambul, uranium nitrate, cod-liver 
 oil. Brewer's yeast also appears worthy of a trial. 
 
 From a review of the treatment of diabetes, it will be 
 evident that what is required to enable us to combat the disease 
 successfully is either — 
 
 1. Some new form of carbohydrate which diabetic patients 
 can assimilate, and which can take the place of sugar and starch 
 in the diet ; or 
 
 2. Some internal remedy which will prevent the accumula- 
 tion of sugar in the blood — a remedy which will enable the 
 system to dispose of the carbohydrates in the normal manner. 
 
 Treatment of Complications and Troublesome Symptoms. 
 
 Thirst. — The relief of this troublesome symptom has been 
 already considered on p. 362. 
 
 Carious teeth and inflammation of the gums. — For these 
 common complications the use of a mouth wash several times 
 a day will be found serviceable. A weak solution of boracic 
 acid, or of borax in camphor water, may be used ; or the follow- 
 ing may be prescribed as a mouth wash : — 
 
 B: Borax, 3ij ; boric acid, 7A ; potassium chlorate, 3i ; camphor 
 water, 3xx (Yeo). 
 
 A 3 per cent, solution of sodium bicarbonate is also a useful 
 mouth wash. The teeth and gums should be kept clean by 
 
392 THE TREATMENT OF DIABETES. 
 
 the use of a tooth-brush ; but it is very important that the 
 brush should not be too hard, or mischief may be done by its use. 
 Dyspepsia is to be treated on general principles. A mixture 
 of alkalies and hydrocyanic acid may be prescribed. I have also 
 found frequent doses of bicarbonate of soda (10 grs.) in a 
 teaspoonful of milk, of service. Sir William Eoberts states that, 
 when craving for food and a sense of sinking at the epigastrium 
 are troublesome, 2 or 3 grs. of asafcetida in a pill two or three 
 times a day often gives relief. 
 
 For gastric crises Grube recommends that the bowels should 
 be kept regular, and an alcoholic extract of pancreas taken after 
 the three principal meals. 
 
 Constipation. — For this troublesome symptom, Carlsbad salts, 
 given in a tumblerful of lukewarm water before breakfast, are of 
 service. Artificial effervescing Carlsbad salts act quite as well, 
 and are much more pleasant to take. Other purgatives, such as 
 castor-oil, confection of senna, aloes, or Friedrichshall or Hunyadi 
 waters, may be employed. 
 
 For flatulence and intestinal catarrh, a pill containing creasote 
 or thymol is suitable. For diarrhoea, salicylate of bismuth, 
 8 grs. in powder twice a day has been found very useful. 
 
 Phthisis is such a common and serious complication, espe- 
 cially in the severe forms of the disease, that it is most important 
 that the diabetic patient should carefully avoid all risk of 
 tubercular infection ; and great attention should be paid to the 
 ventilation of the rooms in which he lives and to other hygienic 
 conditions. If phthisis should arise, then residence in a mild 
 climate is to be preferred rather than at a high altitude. A 
 large quantity of fatty food is especially to be recommended ; 
 and a small amount of alcohol should be given, in order to aid 
 digestion. Creasote, cod-liver oil, and the usual drugs prescribed 
 in phthisis may also be employed. 
 
 Itching of the shin. — Warm baths or frequent sponging with 
 tepid water are of service in relieving this symptom. It is 
 important also to keep the bowels regular. Potassium iodide 
 has been recommended internally. Seegen thinks that potassium 
 bromide is sometimes useful. When the symptom is very 
 troublesome, the calcium chloride treatment recommended by 
 Savill for general pruritus is worthy of trial — 
 
 R Calcii chloridi, 20 to 40 grs.; tinct. aurantii, 5i ; aqure 
 chloroformi, =i ; t.d.s. 
 
COMPLICATIONS— TROUBLESOME SYMPTOMS. 393 
 
 Pruritus and eczema of the vulva. — It has been already 
 mentioned on p. 223, that both these conditions are due to the 
 irritation of the saccharine urine, and often crusts containing 
 fungus mycelia and spores are found on the external genital 
 organs. To prevent the development of these troublesome 
 symptoms, the patient ought to be advised to dry the external 
 orifice of the urethra and surrounding parts with lint, or, 
 better still, with absorbent wool, directly after each act of 
 micturition. If eczema has actually developed, then the use 
 of remedies which prevent the fermentation of sugar is found 
 most beneficial. Boracic acid ointment is often of great service ; 
 or the parts may be washed with a warm, freshly made, 
 saturated solution of boracic acid, and, after careful drying, 
 zinc ointment should be applied (Saundby). A lotion con- 
 taining sodium hyposulphite, 1 oz. to 40 oz. of water, is 
 also a favourable remedy. Lawson Tait recommends an oint- 
 ment composed of 10 grs. of potassa sulphurata to 1 oz. of 
 benzoated lard. He has obtained the best results from this 
 ointment along with opium internally. 
 
 Hebra's unguentum diachylum has been found useful in 
 obstinate cases by Seegen. 
 
 Eczema of the 'prepuce. — To prevent this complication, the 
 patient ought to dry the end of the penis after each act of 
 micturition with lint or antiseptic wool. If eczema has actually 
 developed, boracic acid ointment or boracic acid lotion may be 
 employed. 
 
 Cystitis. — When this rare complication occurs, it may be 
 treated in the usual way, by washing out the bladder with 
 boracic acid solution (15 grs. to the oz.), or with a weak solution 
 of sodium salicylate (30 grs. to the oz.), in each case the solution 
 being made lukewarm by the addition of hot water before being 
 used. Internally, sodium salicylate, salol, or boric acid may be 
 given. 
 
 Boils and carhuncles. — In addition to the attempt to remove 
 sugar from the urine by the dietetic and usual medicinal treat- 
 ment, moist compresses of lint, soaked in a mild antiseptic such 
 as boracic acid, should be applied locally ; or an antiseptic poul- 
 tice may be applied. The latter is made by adding 2 drms. of 
 pure carbolic acid to 10 oz. of boiling water, and immediately 
 afterwards stirring in linseed meal. Quinine in 3 gr. doses four 
 times a day may be given internally. 
 
394 THE TREATMENT OF DIABETES. 
 
 Operative interference should be avoided, if possible, on 
 account of the danger of exciting coma thereby ; also, if 
 chloroform be administered, there is the danger of the narcosis 
 itself, owing to the depressing action on the heart. Hence, 
 ether narcosis is to be preferred, if an operation should be 
 necessary. 
 
 Gangrene. — The dietetic and medicinal treatment for diabetes 
 ought always to be enforced. As regards the local treatment, 
 Konig ( 86 ) and others point out that the chief indications are 
 to make the gangrenous parts dry and aseptic, if possible. All 
 moist dressings are to be avoided. The parts are to be dressed 
 with iodoform and wool, and access of air allowed. If cellulitis 
 be present, free incisions should be made. Konig points out 
 that amputation may be necessary in diabetic gangrene in two 
 conditions — (1) when, in spite of energetic antiseptic treatment, 
 the cellulitis increases, and fever continues ; (2) when the glyco- 
 suria persists in spite of anti-diabetic treatment, and coma is 
 threatening. 
 
 Godlee ( 87 ) points out that, when diabetic gangrene is asso- 
 ciated with extensive atheroma, the arterial changes will probably 
 extend as high as the knee. In these cases, if the gangrene 
 should be progressing rapidly, he advises amputation, but not 
 lower than the knee ; when the gangrene is due to neuritis 
 (see p. 226), he does not advise amputation, or, if amputation 
 be performed, it should be below the knee. 
 
 Corns. — Unless it is absolutely necessary, corns ought not to 
 be cut, since any skin irritation or wound produced might lead 
 to the formation of troublesome trophic ulcers, perforating 
 ulcers, or to superficial gangrene. 
 
 All operative treatment should, as a general rule, be avoided, 
 or should only be performed when absolutely necessary, since 
 wounds heal badly, gangrene is liable to occur, and there is 
 the danger of the operation being followed by diabetic coma. 
 The operation for diabetic cataract is an exception to the above 
 statement. 
 
 Tor sleeplessness, sulphonal, chloralamide, opium, etc., may be 
 given. When there is much nervous excitement, potassium 
 bromide is very useful. 
 
 In the rare cases in which oedema of the legs or other 
 regions occurs, apart from the presence of any kidney com- 
 plications, rest in bed is very successful in removing the 
 
COMPLICATIONS— TROUBLESOME SYMPTOMS. 395 
 
 anasarca. Internally, perchloride of iron may be prescribed 
 (Dickinson). 
 
 For the troublesome gnawing pains in the legs I have 
 found antipyrin useful (10 grs. in 1 oz. or \ oz. of peppermint 
 water, three times a day). 
 
 For neuralgia and sciatica the usual remedies may be 
 employed. 
 
 If diabetes is associated with very marked arterio-sclerosis, 
 iodide of sodium should be given internally, and the bowels 
 kept quite regular by frequent purgatives. 
 
 When nephritis (parenchymatous or interstitial) occurs as 
 a complication, then a rigid nitrogenous anti-diabetic diet must 
 not be prescribed. Only a medium quantity of nitrogenous 
 food should be allowed ; milk in large quantities is indicated ; 
 fatty food may be taken freely. All saccharine food must be 
 avoided, but a limited amount of carbohydrate in the form of 
 bread may be allowed. Ordinary drinking water and alkaline 
 mineral waters may be taken freely. Opium, morphia, or codeia 
 ought not to be prescribed. An alkaline treatment with 
 potassium or sodium citrate appears the most suitable. The 
 bowels should be kept freely open. 
 
 Treatment of Diabetic Coma. 
 
 The exciting causes of diabetic coma have already been pointed 
 out (p. 272), and by bearing them in mind we obtain several 
 hints as to the mode of life, etc., best calculated to prevent 
 the onset of this complication. 
 
 When the urine gives a marked reaction with perchloride 
 of iron, especially if the patient be wasted, and under the age 
 of thirty, there is a great risk of coma developing ; and, as 
 already pointed out on p. 273, in such cases the diet ought 
 not to be too rigid. A sudden complete withdrawal of carbo- 
 hydrates might lead to coma ; also a sudden change from a 
 rigid diet to an ordinary diet sometimes has the same effect. 
 Hence any change in either direction ought to be gradual. 
 
 The danger of a long railway journey and of over-exertion 
 has been pointed out (see p. 273). 
 
 Prolonged constipation ought also to be avoided, since it 
 is probable that it acts as an exciting cause. 
 
 When the earliest symptoms of diabetic coma have made 
 
396 THE TREATMENT OF DIABETES. 
 
 their appearance, the patient should, of course, be confined to 
 bed. It is advisable to increase the carbohydrates in the food 
 if the diet has been rigid previously. Schmitz ( ss ), Grube ( 98 ), 
 and others have pointed out that sometimes, in cases presenting 
 symptoms of early coma, an increase in the amount of carbo- 
 hydrate food — by the addition of a moderate amount of white 
 bread and a small quantity of potatoes to the diet — has been 
 followed by the disappearance of these symptoms. The quantity of 
 flesh meat, eggs, and other forms of nitrogenous food ought to 
 be diminished. Fatty food is especially indicated, and cream in 
 large quantities is the best article of diet. Milk may also be 
 allowed in moderate quantities. 
 
 A considerable amount of alcohol, given in small quantities 
 at a time, is advisable, not only as a stimulant, but also as an 
 aid to the digestion of fatty food. 
 
 Hirschfeld ( 90 ) recommends glycerine (100 to 150 grms.) daily, 
 in black coffee, in order to diminish the acetone formation. 
 
 v. Noorden thinks that, when the patient has been taking 
 large, quantities of carbohydrates before the onset of coma, 
 they ought to be diminished and replaced in part by albuminous 
 and fatty food. 
 
 If constipation is present, — and this is usually the case, — 
 it is important to relieve the bowels, but drastic purgatives 
 ought to be avoided (see p. 273). Mild purgatives and enemata 
 are best. Schmitz prefers castor-oil ; if this should prove 
 useless, calomel or compound jalap powder may be given. 
 Schmitz has recorded cases which appear to demonstrate the 
 importance of the relief of constipation (by castor-oil) in the 
 treatment of diabetic coma. In some of his cases, the early 
 symptoms disappeared after the constipation had been relieved. 
 
 Cardiac stimulants are indicated when the heart is failing ; 
 they are of most service in the third form of diabetic coma, which 
 appears to be due to degeneration of the cardiac muscle (see 
 p. 281). Ether, ammonia, digitalis, and alcohol are useful in 
 such cases. 
 
 The patient ought to be allowed to drink large quantities 
 of water, in order to wash out the tubules of the kidneys. 
 
 The inhalation of oxygen appears to be worthy of trial, 
 and several cases have been reported in which it has apparently 
 been of slight service. In one case it is stated that 
 consciousness was restored by the inhalation of oxygen for 
 
DIABETIC COMA. 397 
 
 five minutes, and the patient lived for two months after- 
 wards. 
 
 It has already been pointed out that many observers are 
 of opinion that diabetic coma is due to an acid intoxication, 
 and to counteract the effects thereof, alkalies have been often 
 recommended. There appears to be some evidence in their 
 favour, and as we know of no more satisfactory treatment, it is 
 advisable to give alkalies a fair trial in all cases. The alkalies 
 most suitable are the bicarbonate and citrate of soda and citrate 
 of potash. Of these the sodium salts are to be preferred, since 
 very large doses can be given without any risk of toxic effects. 
 
 In a case of diabetes of the most severe form, which was 
 under my care twelve months ago, alkalies appeared to be 
 of service. The patient was twenty-four years of age ; he 
 was very much wasted; and the urine was loaded with sugar 
 (3000 to 4200 grs. being excreted daily). Four months after the 
 onset of his symptoms he rapidly became much weaker, the 
 urine gave a dark brownish red coloration with perchloride of 
 iron, and a small quantity of albumin was present. He began to 
 suffer from nausea and vomiting, a deposit of casts appeared in 
 the urine, he became drowsy, and the breathing became deep 
 and laboured. The symptoms pointed to the onset of diabetic 
 coma. I prescribed 30 grs. of bicarbonate of soda every 
 three hours (to be taken in a little milk), and 100 grs. of 
 citrate of potash three times a day. The nausea and vomiting 
 ceased, and the drowsiness passed away ; the casts disappeared 
 from the urine ; the patient improved, and lived for one month 
 after the disappearance of these early symptoms of diabetic coma. 
 
 In a case w 7 hich I had the opportunity of observing at the 
 Manchester Infirmary under the care of Dr. Leech, large doses 
 of citrate of potash were prescribed by Dr. Eeynolds ( 91 ), who 
 was then resident medical officer, with the result that the 
 early symptoms of coma, drowsiness, and dyspnoea passed away, 
 and the patient lived for twelve months longer. 
 
 Huchard ( 92 ) records a case in which premonitory symptoms 
 of diabetic coma had appeared, but after the administration 
 of large doses of bicarbonate of soda (10 drms. daily) the 
 symptoms disappeared. 
 
 Cases similar to those just referred to have been recorded 
 from time to time. It is impossible, of course, to definitely 
 prove that fatal coma would have followed if the alkalies had 
 
398 THE TEE A TMENT OF DIA BE TES. 
 
 not been prescribed ; but such cases indicate the desirability of 
 giving alkalies a trial. Unfortunately, however, the comatose 
 symptoms usually advance to a fatal termination in spite of 
 similar alkaline treatment. 
 
 Huchard recommends sodium bicarbonate (2 to 10 drms. 
 daily), especially when hyperacidity of the stomach contents is 
 present, v. Noorden recommends 90 to 120 grs. of sodium 
 bicarbonate daily, to be taken in one or two bottles of Vichy 
 or Neuenahr mineral water. If there should not be any marked 
 hyperacidity of the stomach contents, probably the formula given 
 by Yeo is as satisfactory as any — 4 drms. of sodium bicarbonate, 
 2 drms. of citric acid, 5 oz. of water flavoured with a little 
 saccharine and essence of lemon. The whole of this mixture 
 should be taken once or twice daily in frequent doses. 
 
 Intravenous injections. — During the last ten years, intra- 
 venous injections of alkaline fluids have been often recommended 
 in diabetic coma, especially by those physicians who believe 
 that the condition is due to the presence of an abnormal acid 
 in the blood (acid intoxication). Others have recommended 
 the injection of a solution of sodium chloride and phosphate, 
 or even of sodium chloride alone. The treatment by intravenous 
 injections has now had a fair trial, and from the records in 
 medical literature, and from what I have seen of the method at 
 the Manchester Eoyal Infirmary, I think there is evidence 
 (1) that when properly performed, intravenous injections have 
 often a decided beneficial effect; but (2) that these results are 
 only temporary, and that the fatal termination is not prevented. 
 After an intravenous injection the pulse often improves, the 
 colour of the face is better, the slight cyanosis disappears, the 
 coma is less, and the patient can be more readily roused. Often, 
 when the patient is quite comatose and collapsed, and a speedily 
 fatal termination has appeared imminent, an intravenous injection 
 has caused decided improvement, and life has been prolonged 
 for twenty-four hours, though the patient has not recovered 
 consciousness. Occasionally a patient who has been quite 
 unconscious before, has recovered consciousness after the 
 injection, and remained conscious for several hours. But even 
 in the most successful cases, in the course of a few hours the 
 coma again becomes profound, and death occurs. Osier ( 93 ) states 
 with respect to the results of intravenous injections, that " of 
 seventeen cases collected by Chadbourne, in only one was it 
 
DIABETIC COMA. 399 
 
 successful ; in seven there was temporary improvement." It is 
 not stated whether the coma disappeared permanently in the 
 one case. I am not acquainted with the record of any case in 
 which there was permanent recovery from deep coma after 
 intravenous injections. Nevertheless it may be sometimes very 
 desirable, for various reasons, that a return of consciousness 
 should be produced for a few hours, if possible, and in such 
 cases intravenous injections are advisable. 
 
 In one case which I have seen, the patient had become 
 quite unconscious, but after an intravenous injection of 3 pints 
 of 0*75 per cent, solution of sodium chloride there was a 
 distinct improvement in the pulse and general condition. The 
 profound coma disappeared, and though the patient was in a 
 drowsy condition she was able to answer questions intelligently for 
 twelve hours, then she became quite comatose, and death occurred. 
 In another case there was improvement in the pulse and general 
 condition, but no recovery of consciousness, after an intravenous 
 injection of a similar salt solution. In a third case an injection 
 of a solution of sodium carbonate, chloride, and phosphate was 
 followed simply by slight improvement of the pulse. In a fourth 
 case there was no improvement after an injection of salt solution. 
 The solution injected consisted of 4 clrms. of sodium chloride 
 dissolved in two pints of water. This is mixed with an equal 
 quantity of warm water immediately before being injected. 
 
 Fagge ( 94 ) records a case in which an intravenous injection of 
 26 oz. of a solution of sodium phosphate and chloride, having 
 a specific gravity of 1020, was followed by return of conscious- 
 ness, but thirty-two hours later the patient again became drowsy, 
 and death occurred, v. Noorden ( 95 ) mentions a case in which 
 very marked improvement followed the intravenous injection of 
 1 litre of an - 6 per cent, sodium chloride solution (four times 
 250 c.c, at intervals of three hours); the patient regained 
 consciousness, and marked diuresis was produced. The improve- 
 ment was only temporary, however. Dickinson has observed a 
 return of consciousness after the intravenous injection of 456 
 oz. of a solution of potassium chloride, sodium carbonate, sodium 
 phosphate, and sodium sulphate. 
 
 The intravenous injection of a 3 per cent, sodium bicarbonate 
 solution has also been used. Hesse ( 9C ) has recorded a case of 
 diabetic coma in which decided improvement followed the intra- 
 venous injection of a 4 per cent, solution of sodium carbonate. 
 
4 oo THE TREATMENT OF DIABETES. 
 
 The patient relapsed, however, into a state of coma, but was 
 again relieved by a second injection ; after twenty-four hours he 
 became comatose and died. 
 
 Lepine ( 97 ) records the results of intravenous saline injections 
 in a case of diabetic coma. The patient regained consciousness 
 for several hours. Before the onset of coma the urine gave a 
 marked reaction with perchloride of iron ; no albumin was pre- 
 sent ; 9 to 1 litres of urine were passed daily, and the amount 
 of sugar was 48 to 55 grms. per litre. The first symptom of 
 coma was deep respiration. Sodium bicarbonate (25 grms. in 
 the twenty-four hours) was prescribed. Next day the patient felt 
 much worse ; coma developed, and became complete. The eyes 
 were closed, the pupils very contracted ; the respirations were 
 2 8 per minute, and very deep. The breath had the characteristic 
 odour of acetone. The pulse was very feeble, and the pulse- 
 beats 124 per minute. The temperature was subnormal. A 
 solution was prepared, consisting of 7 grms. of chloride of 
 sodium and 10 grms. of bicarbonate of soda, dissolved in 1 litre 
 of sterilised water. Two litres of this solution, at a temperature 
 of 38° 0., were injected into a vein of the arm in less than a 
 quarter of an hour. Whilst the fluid was being injected the 
 pulse became stronger, and the respiration less deep ; at last the 
 patient opened his eyes, and asked to drink ; 5 grms. of bicar- 
 bonate of soda dissolved hi water and wine were given to him 
 in the course of a few hours. Eour hours after the injection the 
 patient still remained conscious. The urine was very acid, pale, 
 and scanty. Three hours afterwards the patient became uncon- 
 scious. After the coma had continued eight hours, death 
 occurred. In the case recorded, the patient took altogether 
 90 grms. of bicarbonate of soda by the mouth, and 20 grms. were 
 given by intravenous injection. The urine remained very acid 
 in spite of the great quantity of alkali introduced into the system. 
 In another communication Lepine ( 98 ) states that he has never 
 known any case of diabetic coma recover. In order to have 
 any prospect of success, he thinks that the injection of alkaline 
 fluids ought to be made before the patient becomes comatose, 
 that is, when the prodromal dyspnoea first appears. He records 
 two cases in which early injections were made. The first case 
 was that of a boy, set. 16, who suffered from a very severe form 
 of diabetes. The urine gave an intense reaction with perchloride 
 of iron. Large doses of bicarbonate of soda internally (30 
 
DIABETIC COMA. 401 
 
 grms. daily) produced no effect. As the condition was becoming 
 worse, 2 litres of a solution containing 7 grms. of sodium chloride 
 and 10 grms. of bicarbonate of soda per litre, were injected sub- 
 cataneously — under the skin of the abdomen. The condition 
 of the urine remained practically the same; no improvement 
 occurred ; symptoms of commencing coma appeared, and the 
 patient was removed from the hospital by his parents. The 
 second case was that of a man, set. 37, who also suffered from a 
 severe form of the disease. Whilst under treatment, aceto-acetic 
 acid appeared in the urine, and the general condition became 
 much worse. He became restless, and a marked augmentation of 
 the respiratory movements was noticed. An intravenous injec- 
 tion of 2 litres of the above solution was made. After the injection 
 the quantity of urine increased slightly, and at least three times 
 the quantity of acetone, aceto-acetic acid, and oxybutyria acid 
 was excreted. Thus the injection caused abundant elimination 
 of these toxic substances. Next day there was improvement 
 of the general condition and diminution of the quantity of 
 urine excreted. Twenty-nine clays later the patient was 
 shown at the Lyon Medical Society, the improvement having 
 continued. 
 
 The small operation of intravenous injection can be easily 
 performed, just in the same way as transfusion of salt solution 
 in cases of hemorrhage, etc. All the apparatus necessary is a 
 funnel and piece of india-rubber tubing with clips and cannula, 
 One of the veins in front of the elbow is exposed, and the 
 cannula inserted. It is important that the fluid used should 
 not be cold. The solution of sodium chloride or sodium 
 bicarbonate is mixed with an equal quantity of warm water 
 just before being transfused. Lepine recommends that the 
 fluid injected should have a temperature of 38° C. The fluid 
 is apt to become cold if kept for a long time in the funnel 
 whilst the vein is being exposed. Hence it is better to expose 
 the vein first, before placing the warm fluid in the funnel and 
 tubing. In place of a funnel a vessel may be used which has 
 a cover, the fluid flowing from the lower part. [For details 
 respecting the operation of transfusion, see surgical works, 
 and article by Jennings, Brit. Med. Journ., London, February 8, 
 1896.] 
 
 26 
 
402 
 
 THE TREATMENT OF DIABETES. 
 
 1. V. NoORDEN . 
 
 2. Leo . 
 
 3. Minkowski . 
 
 4. Ebstein, W. . 
 
 5. Hirschfeld, F. 
 
 6. Geube, K. 
 
 7. Schmitz, E. . 
 
 8. Seegen 
 
 9. Blyth, A. W 
 
 10. Wright . 
 
 11. Einger, S. 
 
 12. Blyth . . 
 
 13. Stange 
 
 14. Blyth . . . 
 
 15. Schmitz . . 
 
 16. Minkowski . 
 
 17. Haycraft 
 
 18. Hale White 
 
 19. Grube, K. 
 
 20. Bohland . . 
 
 21. Ebstein . . 
 
 22. LlNDEMANN 
 
 May 
 
 AND 
 
 EEFEBENCES. 
 
 Beilage z. Centralbl. f. innere Med., Leipzig, 
 
 1895, No. 21, S. 36. 
 Deutsche med. Wchnsclir., Leipzig, 1892, No. 
 
 33. 
 "Encyklopsedie der Therapie," Liebreich, 
 
 Berlin, 1896, Bd. i. Abth. 3, S. 944. 
 u Ueber die Lebensweise der Znckerkranken," 
 
 Wiesbaden, 1892, and Med. Citron., Man- 
 chester, September 1892. 
 Deutsche med. Wchnschr., Leipzig, 1893, No. 
 
 38. 
 Lancet, London, 30th December 1893. 
 Deutsche med. Wchnschr., Leipzig, 1893, No. 
 
 27 ; " Prognose und Therapie der Zucker- 
 
 krankheit nach eigenen Erfahrungen," 
 
 Bonn, 1892. 
 "Der Diabetes mellitus," Berlin, 1893, S. 
 
 257, 258. 
 " Foods : their Composition and Analysis," 
 
 London, 1896, 4th edition, p. 250. 
 Brit. Med. Journ., London, 11th April 1891. 
 U)id., 7th December 1895, 14th December 
 
 1895, p. 1524. 
 Op. cit, p. 328. 
 
 Ziemssen's "Handbook of General Thera- 
 peutics," English trans., London, 1885, 
 
 vol. i. p. 395. 
 Op. cit., p. 341. 
 "Prognose und Therapie der Zuckerkrank- 
 
 heit," Bonn, 1892, S. 23. 
 Beilage z. Centralbl. f. 1dm. Med., Bonn, 
 
 1892, No. 25; Arch, f. exper. Path. u. 
 
 Pharmakdl., Leipzig, 1889, Bd. xxxi. S. 85. 
 Ztschr. f. physiol. Client., Strassburg, 1894, 
 
 Bd. xix, Heft 2. 
 Guy's Hosp. Rep., London, 1893, p. 133. 
 Ztschr. f. Idin, Med., Berlin, Bd. xxvi. 
 Jahresb. ii. d. Leistung. . . . d. ges. Med., 
 
 Berlin, 1894, Bd. ii. S. 52. 
 Virchoiv's Archiv, 1893, Bd. cxxxiv. S. 361. 
 Deutsches Arch. f. Idin. Med., Leipzig, Bd. 
 
 lvi. Hefte 3, 4. 
 
REFERENCES. 
 
 4°3 
 
 23. Brunton, Lauder 
 
 24. Blyth ........ 
 
 25. Camplin .... 
 
 26. Hale White . . 
 
 27. Dujardin-Beaumetz 
 
 28. Saundby .... 
 
 29. Seegen .... 
 
 30. Hart, Mrs. . . . 
 
 31. Ebstetn .... 
 
 32. Williamson, E. T. 
 
 33. Kulz .... 
 
 34. Stern .... 
 
 35. Saundby . . . 
 
 36. Ebstein . 
 
 37. Brunton and Tun 
 
 nicliffe 
 
 38. Grube, K. . 
 
 39. ElNKLER . . 
 
 40. Ererichs . . 
 
 41. Minkowski . 
 
 42. Naunyn . . 
 
 43. Kiess . . . 
 
 44. Leichtenstern 
 
 45. Seegen . . 
 
 46. Sciimitz, E. . 
 
 47. HmsciiFELU . 
 
 St. Barth. Hosp. Journ., London, February 
 1896, p. 68. 
 
 Op. cit., p. 201. 
 
 " On Diabetes and its Successful Treatment," 
 London, 1864. 
 
 Practitioner, London, May 1893. 
 
 Deutsche med. Wchnschr., Leipzig, 1890, No. 
 34. 
 
 Birmingham Med. Rev., 1893, vol. xxxiii. p. 
 275. 
 
 Loc. cit. 
 
 " Diet in Sickness and in Healtb," London, 
 1895. 
 
 Med. Chron., Manchester, September 1892 ; 
 Deutsche med. Wchnschr., Leipzig, 1892, 
 No. 19; ibid., 1893, No. 18; " Ueber die 
 Lebensweise der Zuckerkranken," Wies- 
 baden, 1892. 
 
 Brit. Med. Journ., London, 27th April 1895. 
 
 Ziemssen's "Handbook of General Thera- 
 peutics," London, 1885, vol. i. p. 296. 
 
 Med. Neivs, Phila., 8th June 1895. 
 
 "Lectures on Diabetes," Bristol, 1891, p. 187. 
 
 " Ueber die Lebensweise der Zuckerkranken," 
 Wiesbaden, 1892, S. 116. 
 
 Journ. Physiol., Cambridge and London; 
 1894, p. 364. 
 
 Ztschr. f. Idin. Med., Berlin, Bd. xxvii. Hefte 
 5, 6. 
 
 Verhandl. d. Cong. f. inner e Med., Wiesbaden, 
 1886, S. 190. ' 
 
 "Ueber den Diabetes," Berlin, 1884, S. 263- 
 265. 
 
 " Encyklopsedie der Therapie," Liebreich, 
 Berlin, 1896, Bd. i. Abth. 3, S. 944. 
 
 Samml. Idin. Vortr., Leipzig, No. 116. 
 
 Berl. Idin. Wchnschr., 1877, No. 39. 
 
 Ziemssen's "Handbook of General Thera- 
 peutics," London, 1885, vol. iv. p. 363. 
 
 "Der Diabetes mellitus," Berlin, 1893, S. 
 272. 
 
 " Prognose und Therapie der Zuckerkrank- 
 heit," Bonn, 1892, S. 38. 
 
 Berl. Kliniic, 1893. 
 
4°4 
 
 THE TREATMENT OF DIABETES. 
 
 48. Pavy 
 
 49. Ealfe 
 
 50. Kratschmer . . . 
 
 51. Bruce, Mitchell . 
 
 52. BlCHARDIERE . . 
 
 53. Grube 
 
 54. Murray .... 
 
 57. Lewaschew . . . 
 
 58. Dujardin-Beaumetz 
 
 59. Kobin . . 
 
 60. Ebstein . 
 
 61. FtJRBRINGEn 
 
 62. Schmitz . 
 
 63. "West . . 
 
 64. Dickinson 
 
 65. Feinburg . 
 
 66. v. noorden 
 
 67. Ascoli. . 
 
 68. PuRDY. . 
 
 69. Cantani, A. 
 
 70. V. ISTOORDEN . . . 
 
 71. Marie, P., and Le 
 
 GOFF, J. 
 
 72. Le Goff, J. . . . 
 
 73. Bobin 
 
 74. Lepine .... 
 
 75. Cassaet .... 
 
 76. Mansell-Jones . . 
 
 Guxfs Hosp. Rep., London, vol. xv. 
 Lancet, London, 30th April 1892. 
 Quoted by Seegen, loc. cit., S. 284. 
 Practitioner, London, January 1887 and 
 
 July 1888. 
 Union med., Paris, 1895, No. 7. 
 Milnchen. med. Wchnschr., 1895, No. 33. 
 Lancet, London, 25th February 1893. 
 Berl. Idin. Wchnschr., 1891, No. 8. 
 Deutsche med. Wchnschr., Leipzig, 1890, 
 
 No. 34. 
 Bull. Acad, de med., Paris, 1895, No. 23. 
 " Die Zuckerharnruhr," Wiesbaden, 1887, 
 
 S. 215. 
 Deutsches Arch. f. Idin. Med., Leipzig, 
 
 1878, Bd. xxi. S. 469. 
 " Prognose unci Therapie der Zucker- 
 
 krankheit," Bonn, 1892. 
 Brit. Med. Journ., London, 24th August 
 
 1895 and 19th September 1896. 
 " Diseases of the Kidney and Urinary 
 
 Derangements," pt. 1, " Diabetes," London, 
 
 1875, p. 141. 
 Jahresb. il. d. Leistung. . . . d. ges. Med., 
 
 Berlin, 1889, Bd. ii. S. 645. 
 Op. cit., S. 130. 
 Brit. Med. Journ., London, 19th October 
 
 1895. (Epitome.) 
 "Diabetes : its Causes, Symptoms, and 
 
 Treatment," 1890, p. 106. 
 " Specielle Pathologie u. Therapie der Stoff- 
 
 wechselkrankheiten — der Diabetes mel- 
 
 litus," Berlin, 1880, S. 380-390 (trans, by 
 
 S. Hahn). 
 Loc. cit., S. 130. 
 Bidl. et m6m. Soc. med. d. hop. de Paris, 1897, 
 
 p. 658. 
 "Sur certaines reactions chromatiques du Sang 
 dans le diabete sucre," Paris, 1897, p. 82. 
 Loc. cit. 
 
 Semaine med., Paris, 24th April 1895. 
 Ibid., 21st August 1895. 
 Brit. Med. Journ., London, 7th January 
 
 1893. 
 
REFERENCES. 
 
 405 
 
 77. Mackenzie, H. . . 
 
 78. Wood, Neville . . 
 
 79. White, W. Hale . 
 
 80. De Ce"renville . . 
 
 81. Ausset .... 
 
 82. BoRMANN .... 
 
 83. Williamson, E. T. . 
 
 84. Williams, P. W. . 
 
 85. Gilbert and Carnot 
 
 86. Konig, Fr. . . . 
 
 87. Godlee, E. T. . . 
 
 88. Schmitz, E. . . . 
 
 89. Grube, K. 
 
 90. HlRSCHFELD, F. 
 
 91. Eeynolds, E. S. 
 
 92. Huchard . . 
 
 93. Osler, W. . 
 
 94. Fagge, H. 
 
 95. 
 
 96. 
 97. 
 98. 
 
 V. NoORDEN, C. 
 
 Hesse, J. . . 
 
 Lepine . . 
 
 Do. . . . 
 
 Brit. Med. Journ., London, 14tli January 1893. 
 
 Ibid., 14th January 1893. 
 
 Ibid., 4th March 1893. 
 
 Rev. med. de la Suisse Rom., Geneve, tome 
 
 xii. p. 660. 
 Semaine med., Paris, 21st August 1895. 
 Wien. med. Bl., 17th October 1895. 
 Med. Chron., Manchester, April 1893, 
 
 pp. 54-58. 
 Brit. Med. Journ., London, 8th December 
 
 1894. 
 Semaine med., Paris, 10th May 1897. 
 Berl. Jdin. Wclmschr., 22nd June 1896. 
 Med.-Chir. Trans., London, vol. lxxvi. 
 " Prognose und Therapie der Zucker- 
 
 krankheit," Bonn, 1892; Deutsche med. 
 
 Wclmschr., Leipzig, 1893, No. 27. 
 Lancet, London, 30th December 1893. 
 Deutsche med. Wclmschr., Leipzig, 1893, 
 
 No. 38. 
 Med. Chron., Manchester, 1891, vol. xiv. 
 
 p. 338. 
 Rev. gen. de clin. et de therap., Paris, March 
 
 1893. 
 " Principles and Practice of Medicine," second 
 
 edition, Edinburgh and London, 1895, 
 
 p. 330. 
 '• Principles and Practice of Medicine," 
 
 London, 1888, vol. ii. p. 719. 
 "Die Zuckerkrankheit," Berlin, 1895. 
 Quoted by Purdy, "Diabetes," 1890, p. 110. 
 Semaine med., Paris, 1897, No. 10. 
 Lyon med., 1897, No. 15. 
 
APPENDIX. 
 
 In Chapter XIA 7 . the dietetic treatment of diabetes has been 
 discussed, and the class of cases pointed out in which a rigid 
 diet is contra-indicated. The following is a list of articles of diet 
 and beverages, arranged in a tabular form, which ought to be 
 sanctioned or forbidden, when for diagnostic or therapeutical 
 purposes a very strict diet is desirable : — 
 
 Articles of Food in Diabetes. 
 
 Sanctioned. 
 
 Butchers' meat of all kinds (except 
 liver) ; potted and preserved meats. 
 
 Ham, tongue, bacon. 
 
 Poultry, game. 
 
 Fish (fresh, dried, and preserved) ; 
 sardines, shrimps. 
 
 Broths, animal soups, and jellies 
 (prepared without the addition of 
 saccharine or starchy materials). 
 
 Eggs, cheese, cream. 
 
 Butter, suet, oils, and fats. 
 
 Custard (without sugar). 
 
 Reliable bread substitutes (gluten 
 bread, almond and aleuronat cakes). 
 
 Green vegetables — mustard and cress, 
 watercress, endive, lettuce, spinach, 
 turnip-tops, cabbage, broccoli, Brus- 
 sels sprouts, spring onions. 
 
 Cucumber. 
 
 Mushrooms. 
 
 Pickles (cucumber, walnuts, and 
 onions). 
 
 Nuts (walnuts, almonds, filberts, hazel 
 nuts, Brazil nuts), but not chest- 
 nuts. 
 
 Forbidden. 
 
 Sugar ; saccharine and farinaceous 
 
 articles of food. 
 Pastry and farinaceous puddings. 
 
 Rice, sago, arrowroot, tapioca, maca- 
 roni, vermicelli, semolina. 
 Potatoes. 
 
 Wheaten bread and biscuits. 
 
 Carrots, turnips, parsnips, beetroot, 
 beans, peas, large onions. 
 
 Liver. 
 
 Oysters, cockles, mussels, the " pud- 
 dings " of crabs and lobsters. 
 Honey. 
 
 All sweet fruit and dried fruits. 
 
APPENDIX. 
 
 407 
 
 Beverages. 
 
 Sanctioned. 
 
 Water, soda-water, and mineral waters. 
 
 Tea, coffee. 
 
 Dry sherry, claret, Burgundy, hock, 
 Moselle, Ahr wines, most Rhine 
 wines, Austrian and Hungarian 
 table wines (all in moderate quan- 
 tities, however). 
 
 Brandy in small quantities. 
 
 Forbidden. 
 
 Port, Tokay, champagne, and sweet 
 
 wines. 
 Must, fruit juices and syrups. 
 Sweet lemonade. 
 Liqueurs. 
 
 Beer, ale, porter, and stout. 
 Rum and sweetened gin. 
 Cocoa and chocolate. 
 Milk in large quantities. 
 
 The diabetic dietaries recommended by most authors agree 
 generally with that just given, but the following are minor points 
 of difference : — 
 
 Sir William Roberts ( 1 ) adds " torrified " bread and celery to the articles 
 sanctioned. 
 
 Pavy ( 2 ) sanctions also turnips, French beans, cauliflower, asparagus, and 
 vegetable marrow, but only in moderate quantity, and when boiled in a 
 large amount of water ; also radishes and celery. 
 
 Seegen ( 3 ) allows the following in moderate quantities : — Cauliflower, carrots, 
 turnips, white cabbage, green peas, berries (such as strawberries, rasp- 
 berries, currants) ; also oranges and almonds. But he forbids other 
 fruits, such as grapes, cherries, peaches, apricots, plums, and all kinds of 
 dried fruits. 
 
 In addition to the articles mentioned, v. Noorden ( 4 ) sanctions mussels, 
 oysters, lobster, cauliflower, spinach, onions, leeks, asparagus, sorrel, 
 French beans. He also allows the following (amongst other articles), 
 but only in very limited quantities : — Celery, green peas, beans, carrots, 
 mushrooms, radishes, two medium-sized tomatoes, a thin slice of cocoanut, 
 a thin slice of melon, one small acid apple, one or one and a half peach, 
 one tablespoonful of wild raspberries or strawberries, four spoonfuls of 
 currants, six greengages, twelve cherries, half a medium-sized pear. 
 
 Cantani ( 5 ) prescribes a very rigid diet of nitrogenous and fatty foods only. 
 He allows — 
 
 Bouillon of various kinds. 
 
 Beef, tongue, veal, mutton (but liver is forbidden), duck, goose, 
 
 chicken, pigeon, and game of all kinds. 
 Fish, crustacea, lobster, crabs, 
 in tli'- rooking aii'l preparation of the above he forbids the use of flour 
 sugar, wine, butter, vinegar, and lemon juice; but he sanctions the use 
 of olive oil and fats, am] recommends dilute acetic acid in place of 
 v inegar, and citric acid in place of lemon juice. 
 
4o8 
 
 APPENDIX. 
 
 Caxtani allows 500 to 600 grms. of nitrogenous food daily. In order to 
 improve the digestion, and to increase the strength of emaciated patients, 
 he recommends 60 to 200 grms. of pancreatic fat. The fresh pancreas of 
 a cow, calf, or lamb is cut into small pieces, and mixed with pig's suet. 
 After three hours it is lightly roasted. 
 
 The beverages which he allows are ordinary drinking water, or seltzer 
 water, with the addition of 10 to 30 grms. of rectified alcohol daily, and 
 small quantities of aq. ftenicul., aq. cinnam., aq. menth., etc. 
 
 1. Eoberts, Sir War. 
 
 2. Pavy, F. W. . . 
 
 3. Seegen, J. . . . 
 
 4. V. NoORDEN, C. 
 
 5. Cantani, A. . 
 
 On Urinary and Eenal Diseases," London, 
 
 1885, p. 289. 
 ' A Treatise on Food and Dietetics," London, 
 
 1874, p. 498. 
 ; Der Diabetes Mellitus," Berlin, 1893, dxitte 
 
 Aufl., S. 271. 
 : Die Zuckerkrankbeit," Berlin, 1895, S. 187. 
 ■ Specielle Pathol. Tberap. der Stoffwechsel- 
 
 krankbeiten," " Der Diabetes Mellitus," 
 
 aus dem Italienischen, von Dr. Hahn, 
 
 Berlin, 1880; S. 426. 
 
INDICES. 
 
INDEX OF SUBJECTS 
 
 PAGE 
 
 Abscess, ischiorectal .... 276 
 
 ,, of liver in diabetes . . . 117 
 
 Aceto-acetic acid in urine . . . 179 
 
 ,, as cause of coma . 289 
 
 Acetonemia 189, 288 
 
 Acetone in blood ..... 189 
 ,, in breath . . . .213 
 
 ,, in urine . . . . • . 181 
 
 test for 182 
 
 Acid intoxication .... 289 
 
 Acromegaly and diabetes . . . 137 
 
 Acute diabetes ..... 205 
 
 ,, infectious diseases and diabetes . Ill 
 
 Age of patients 95 
 
 ,, with reference to prognosis . . 314 
 Albuminuria .... 175, 218 
 
 Alcohol, value of . . . . 339, 363 
 
 Alcoholic beverages in diabetes . . 363 
 
 Alcoholism and diabetes . . . 110 
 
 Aleuronat ...... 355 
 
 ,, and cocoa-nut cakes and buns 356 
 
 ,, and suet pudding . . . 360 
 
 „ bread 355 
 
 ,, pancake .... 360 
 
 Alimentary canal 202 
 
 Alkali mineral waters .... 368 
 
 Alkalies 379 
 
 Alkaptonuria . . . . .11 
 
 Almond cakes ..... 352 
 
 ,, pudding ..... 353 
 
 Amblyopia 236 
 
 America, diabetes in United States of . 101 
 
 Ammonia excretion .... 174 
 
 Amylose group 5 
 
 Anasarca 227 
 
 Anti pyrin 382 
 
 Antisyphilitie treatment . . . 384 
 Appearance of patient . . . .162 
 
 Appetite 203 
 
 Arsenic 380 
 
 Arterio-sclerosis 217 
 
 ,, relation to diabetes . 156 
 
 Ascites 228 
 
 Atrophy of optic disc .... 234 
 
 Auto-intoxication, intestinal, in coma . 290 
 
 Bacillus, tubercle 210 
 
 Balanitis 224 
 
 li'-i-i -drinking and dialn-tc . . .110 
 
 Bev< rage 362 
 
 ' , all hi', uat and cocoa-nut . 356 
 
 ,, almond 352 
 
 ,, cocoa-nut and cream . . 354 
 
 TAGE 
 
 Biscuits, inulin 360 
 
 ,, soya 351 
 
 Bismuth salicylate .... 383 
 ,, tests for sugar . . .15 
 Elood, arterial, injection into portal 
 
 vein 66 
 
 ,, corpuscles in diabetes, number of 186 
 ., ,, ,, staining of 197 
 
 ., in diabetes . . . .185 
 ,, fat in . . .188 
 
 :,' ,, glycogen of . . 188 
 
 ,, ,, glycolytic ferment of 79,189 
 
 ,, ,, leucocytes of . . 188 
 
 , , , , methylene blue, reaction 
 
 of. 191 
 
 ,, ,, test for 
 
 amount of sugar 195, 196 
 
 186 
 
 186 
 
 186 
 
 190 
 
 197 
 
 191 
 
 186 
 
 54 
 
 190 
 
 81 
 
 71 
 
 225 
 
 15 
 
 ,, ,, reaction of 
 
 ,, ,, red corpuscles in 
 
 ., ,, specific gravity of 
 
 ,, ,, sugar in 
 
 ,, ,, staining 
 
 ,, ,, test for 
 
 ,, ,, water of 
 
 ,, sugar in normal 
 
 ,, ,, ,, diabetic 
 
 ,, ,, ,, pancreatic diabetes 
 
 ,, ,, ,, phloridzm diabetes 
 Boils 
 Bottger's test 
 
 Bourboule, La 
 Bowels, condition of 
 Brain complications 
 
 ,, in diabetes . 
 Bran cake 
 Bread and its substitutes 
 
 ,, aleuronat . 
 
 ,, gluten 
 
 ,, pea- nut 
 
 „ soya . 
 
 ,, torrified 
 Breath .... 
 Bromide of potassium . 
 Bronzed diabetes . 
 Butter .... 
 
 Cancer of pancreas 
 Cane sugar . 
 Carbohydrates 
 
 ( ' : i ) I m , j i i ' ■. acid I'oin 
 
 Carbuncles . 
 ( !ardiac failure 
 Carlsbad 
 
 ation in tissues 
 
 373 
 204 
 241 
 125 
 350 
 347 
 355 
 349 
 361 
 351 
 347 
 213 
 383 
 308 
 337 
 
 145 
 33 
 339 
 302 
 225 
 287 
 370 
 
412 
 
 INDEX OF SUBJECTS. 
 
 
 
 
 PAGE 
 
 
 TAGE 
 
 Casts in urine 
 
 
 
 177 
 
 Face, appearance of ... . 
 
 163 
 
 ,, ,, in coma . 
 
 
 
 286 
 
 Fat embolism . . . ' . 
 
 287 
 
 Cataract .... 
 
 
 
 230 
 
 ,, in blood ...... 
 
 188 
 
 Cerebellar cyst 
 
 
 
 134 
 
 Fatty food 
 
 337 
 
 Cerebellum, changes in 
 
 
 
 129 
 
 Fehling's solution .... 
 
 9 
 
 ,, injury of . 
 
 
 
 65 
 
 ,, ., Haines' modification . 
 
 14 
 
 Cerebrum, changes in . 
 
 
 
 129 
 
 ,, ,, Pavy's modification .. 
 
 15 
 
 Cheese ..... 
 
 
 
 338 
 
 ,, ,, quantitative estima- 
 
 
 Chemical note 
 
 
 
 4 
 
 tion of sugar by . 
 
 41 
 
 Children, diabetes in . 
 
 
 
 97 
 
 ,, test 
 
 8 
 
 Climate .... 
 
 
 
 101 
 
 ,, ,, after fermentation . 
 
 14 
 
 Climacteric diabetes 
 
 
 
 115 
 
 ,, ,, after filtration through 
 
 
 Clothing .... 
 
 
 
 366 
 
 charcoal (Seegen's modi- 
 
 
 Cocoa-nut .... 
 
 
 
 354 
 
 fication) 
 
 12 
 
 Codeine .... 
 
 
 
 379 
 
 ,, ,, in the cold 
 
 11 
 
 Cod-liver oil . 
 
 
 
 386 
 
 ,, ,, Worm-Muller's modifica- 
 
 
 Cceliac plexus, extirpation of 
 
 
 
 68 
 
 tion .... 
 
 12 
 
 Cold and wet, as cause of diabetes 
 
 
 112 
 
 Fermentation in bladder 
 
 173 
 
 , , fluids, drinking of 
 
 
 
 112 
 
 ,, test for sugar in the 
 
 
 Collapse, diabetic . 
 
 
 
 281 
 
 urine 
 
 19 
 
 Coma, diabetic 
 
 
 
 270 
 
 ,, ,, quantitative esti- 
 
 
 ,, ., diagnosis of . 
 
 
 
 292 
 
 mation of sugar 
 
 
 ,, ,, exciting causes 
 
 
 
 272 
 
 in the urine 
 
 39 
 
 ,, ,, ,, analysis 
 
 
 Food as a cause 
 
 113 
 
 
 of cases 
 
 274 
 
 ,, animal ..... 
 
 330 
 
 ,, ., forms of 
 
 
 
 281 
 
 ,, carbohydrate .... 
 
 339 
 
 ,, ,. pathology of . 
 
 
 
 287 
 
 ,, fatty ...... 
 
 337 
 
 ,, ,, prognosis 
 
 
 
 295 
 
 ,, nitrogenous 
 
 330 
 
 ,, ,, symptomatology . 
 
 
 276 
 
 ,, vegetable ..... 
 
 339 
 
 ,, ,, ,, analysis 
 
 of 
 
 
 Forms of diabetes .... 
 
 305 
 
 
 cases 
 
 
 282 
 
 Fourth ventricle, changes in 
 
 127 
 
 ,, treatment of 
 
 
 
 395 
 
 ,, puncture of 
 
 63 
 
 Complications, treatment of 
 
 
 
 391 
 
 Fruit 
 
 342 
 
 Constipation 
 
 
 
 204 
 
 
 
 Course 
 
 
 
 310 
 
 Gangrene . . . , . 
 
 225 
 
 Cream ..... 
 
 330, 
 
 335 
 
 337 
 
 Gastric catarrh ..... 
 
 204 
 
 Cystitis .... 
 
 
 
 222 
 
 ,, crises 
 
 ,, juice ..... 
 
 204 
 204 
 
 Definition of diabetes . 
 
 
 
 1 
 
 „ ulcer ..... 
 
 204 
 
 Derivation of name "diabetes 
 
 ' . 
 
 
 2 
 
 Geographical distribution 
 
 97 
 
 Diabete bronze 
 
 
 
 308 
 
 Gerhardt's reaction in urine . 
 
 179 
 
 ,, gras. 
 
 
 
 306 
 
 Glucoses 
 
 5 
 
 ,, maigre 
 
 
 
 305 
 
 Gluten bread ..... 
 
 351 
 
 Diabetes insipidus, relation to diabetes 
 
 
 Glycogen 
 
 52 
 
 mellitus 
 
 
 115 
 
 , 312 
 
 ,, in urine ..... 
 
 173 
 
 Diacetic acid 
 
 
 
 179 
 
 Glycogenic functions of liver 
 
 52 
 
 ,, as cause of coma 
 
 
 
 289 
 
 Glycolytic ferment .... 
 
 79 
 
 Diagnosis .... 
 
 
 
 315 
 
 Glycosuria, alimentary 
 
 85 
 
 Diarrhoea .... 
 
 
 
 205 
 
 ,, clinical .... 
 
 91 
 
 Diet, articles of . 
 
 
 
 330 
 
 ,, experimental 
 
 63 
 
 Dietetic treatment 
 
 
 
 324 
 
 ,, produced by poisons . 
 
 89 
 
 Digestive system . 
 
 
 
 202 
 
 ,, puerperal .... 
 
 88 
 
 Disaccharides 
 
 
 
 5 
 
 ,, symptomatic . . 91 
 
 , 308 
 
 Duration .... 
 
 
 
 313 
 
 Glycuronic acid 
 
 Gout . 
 
 10 
 109 
 
 Ears, the .... 
 
 
 
 238 
 
 Grape sugar in urine .... 
 
 5 
 
 Eczema .... 
 
 
 
 223 
 
 ,, detection of small quanti- 
 
 
 Eggs 
 
 
 
 331 
 
 ties of . 
 
 35 
 
 Emaciation .... 
 
 162, 
 
 163 
 
 305 
 
 ,, quantitative estimation of 
 
 39 
 
 Endogenous diabetes . 
 
 
 
 158 
 
 Gums ....... 
 
 202 
 
 Epilepsy and diabetes . 
 
 
 
 240 
 
 
 
 Erythema .... 
 
 
 
 223 
 
 Haemorrhages, retinal .... 
 
 232 
 
 Etiology .... 
 
 
 
 95 
 
 ,, vitreous 
 
 234 
 
 ,, of a hundred cases . 
 
 
 
 158 
 
 ,, in medulla . . 135 
 
 , 136 
 
 Exciting causes 
 
 
 
 95 
 
 Haines' fluid 
 
 14 
 
 Exercise 
 
 
 
 367 
 
 Head injuries .... 107 
 
 , 108 
 
 Exophth almic goitre 
 
 '87 
 
 , 92 
 
 , 132 
 
 Heart affections .....' 
 
 214 
 
 Experimental diabetes . 
 
 
 
 62 
 
 Hepatic extract ..... 
 
 390 
 
 Eyes, the ... 
 
 
 
 229 
 
 Heredity ...... 
 
 104 
 
INDEX OF SUBJECTS. 
 
 4'3 
 
 
 
 PAGE 
 
 
 
 PAGE 
 
 Historical note .... 
 
 
 2 
 
 Milk sugar ..... 
 
 
 32 
 
 Hoppe-Seyler's test . . " . 
 
 
 17 
 
 Monosaccharides .... 
 Moore's test ..... 
 
 
 5 
 7 
 
 Iceland moss .... 
 
 
 361 
 
 Morbid anatomy. See Pathological 
 
 Anatomy. 
 
 Increase of diabetes 
 
 
 102 
 
 Morphine ..... 
 
 
 378 
 
 India, diabetes in . 
 
 
 99 
 
 Mortality from diabetes 
 
 
 97 
 
 Indigo-carmine test 
 
 
 16 
 
 Mulder's test .... 
 
 
 16 
 
 Infection, possibility of 
 
 
 105 
 
 Muscle glycogen .... 
 
 
 53 
 
 Infectious diseases as a cause 
 
 
 111 
 
 Muscular changes and diabetes 
 
 
 303 
 
 Influenza as a cause 
 
 
 111 
 
 
 
 
 Injections, intravenous, in coma . 
 
 
 398 
 
 Nephritis 
 
 22C 
 
 , 312 
 
 Injury, external .... 
 
 
 107 
 
 Nerves, peripheral, affections of . 
 
 
 250 
 
 Inorganic salts in urine 
 
 
 175 
 
 Nervous diseases associated with 
 
 dia- 
 
 
 Insanity and diabetes . 
 
 92 
 
 , 241 
 
 betes .... 
 
 
 132 
 
 Intestinal auto-intoxication in coma 
 
 290 
 
 ,, system in diabetes . 
 
 
 239 
 
 Intestines, condition of 
 
 
 205 
 
 : , , , changes in . 
 
 
 125 
 
 Inulin ...... 
 
 
 339 
 
 ,, ,, diseases of, and 
 
 dia- 
 
 
 Iodoform 
 
 
 384 
 
 betes 
 Neuenahr 
 
 121 
 
 , 139 
 372 
 
 Jambul ..... 
 
 
 381 
 
 Neuralgia ..... 
 
 
 250 
 
 Johnson's test for sugar 
 
 
 15 
 
 Neuritis, peripheral 
 
 Nuts 
 
 
 259 
 343 
 
 Kephir ...... 
 
 
 336 
 
 Nylander's test .... 
 
 
 15 
 
 Kidneys, affections of . 
 
 
 217 
 
 
 
 
 ,, fatty degeneration 
 
 
 221 
 
 Obesity 
 
 
 109 
 
 ,, glycogenic ,, 
 
 
 221 
 
 OEdema ..... 
 
 
 227 
 
 ,, hyaline ,, 
 
 
 220 
 
 Oliver's test 
 
 
 16 
 
 ,, microscopical changes in 
 
 
 220 
 
 Onset 
 
 
 164 
 
 ,, necrosis of epithelium . 
 
 
 220 
 
 Opium ...... 
 
 
 376 
 
 Knee-jerks ..... 
 
 
 250 
 
 Optic disc, atrophy of . 
 
 
 234 
 
 Koumiss ..... 
 
 
 335 
 
 Orthonitro-phenylpropiolic acid test . 
 
 17 
 
 Kiissmaul's coma .... 
 
 
 280 
 
 Oxalate of lime in sputum . 
 ,, „ urine 
 
 
 214 
 175 
 
 Lactic acid 
 
 
 385 
 
 Oxybutyric acid in urine 
 
 
 185 
 
 Lactose 
 
 
 32 
 
 Oxygen 
 
 385 
 
 , 396 
 
 Lfevulose ..... 
 
 
 33 
 
 
 
 
 Lepine's malt ferment . 
 
 
 387 
 
 Pancreas, condition of, in a series of 
 
 
 Lichenin ..... 
 
 
 340 
 
 consecutive autopsies 
 
 
 152 
 
 Lightning stroke as a cause . 
 
 
 112 
 
 ,, extirpation of 
 
 
 74 
 
 Lime salts 
 
 
 380 
 
 ,, lesions of, and diabetes 
 
 
 140 
 
 Lipasmia ..... 
 
 
 188 
 
 ,, ,, ,, frequency 
 
 and 
 
 
 Liver, abscesses of ... 
 
 
 117 
 
 nature of . 
 
 
 141 
 
 ,, and sugar formation . 
 
 53, \ 
 
 , , symptoms of lesion of . 
 
 
 207 
 
 ,, condition during life . 
 
 
 206 
 
 „ transplantation of pieces 
 
 of . 
 
 76 
 
 ,, diseases and diabetes . 
 
 
 116 
 
 Pancreatic diabetes, experimental 
 
 
 73 
 
 ,, glycogen in 
 
 
 120 
 
 ,, disease, condition of urine in 
 
 153 
 
 ,, in diabetes .... 
 
 
 119 
 
 ,, emulsion 
 
 
 387 
 
 Lungs 
 
 
 207 
 
 ,, graft in treatment 
 
 
 389 
 
 ,, tuberculosis of . 
 
 
 207 
 
 ,, lesions in diabetes, cases of 
 ,, preparations in treatment . 
 
 142 
 388 
 
 Malaria as cause .... 
 
 
 112 
 
 Patella-tendon reflex . 
 
 
 250 
 
 Management of a case . 
 
 
 321 
 
 Pathogenesis .... 
 
 
 298 
 
 Marriage ..... 
 
 
 366 
 
 Pathological anatomy . 
 
 
 297 
 
 Massage ..... 
 
 
 368 
 
 ,, changes in blood 
 
 
 285 
 
 Medulla, changes in 
 
 127 
 
 133 
 
 ., ,, ,, brain 
 
 
 125 
 
 ,, microscopical changes in 
 
 130 
 
 134 
 
 ,, ,, ,, heart 
 
 
 214 
 
 „ section of 
 
 
 64 
 
 ,, ,, ,, intestines 
 
 
 206 
 
 Mental condition .... 
 
 239, 
 
 365 
 
 ,, ,, ,, kidneys . 
 
 
 219 
 
 „ emotion as a cause . 109, 
 
 121, 
 
 125 
 
 ,, ,, ,, liver 
 
 116, 
 
 119 
 
 Methylene blue in treatment 
 
 
 385 
 
 ,, ,, ,, lungs 
 
 
 207 
 
 ,, test for sugar in 
 
 the 
 
 
 ; , ,, ,, pancreas . 
 
 
 141 
 
 urine . 
 
 
 18 
 
 ,, ,, ,, spinal cord 
 
 129, 
 
 242 
 
 ,, ,, ,, diabetic blood 
 
 191 
 
 ,, ,, ,, stomach . 
 
 
 205 
 
 ,, ,, ,, estimation 
 
 of 
 
 
 Pavy's solution .... 
 
 
 42 
 
 sugar 
 
 in 
 
 
 Pentose 1 
 
 . 33, 
 
 341 
 
 blood 
 
 195, 
 
 196 
 
 Pepsin 
 
 
 387 
 
 ,. ., in diagnosis 
 
 of 
 
 
 Percliloride of iron reaction . 
 
 
 179 
 
 coma 
 
 293, 
 
 294 
 
 Perforating ulcer .... 
 
 
 226 
 
 Milk 
 
 
 331 
 
 Peripheral neuritis 
 
 
 259 
 
414 
 
 INDEX OF SUBJECTS. 
 
 
 PAGE 
 
 
 
 PAGE 
 
 Phenylhydrazin test 
 
 '. 23 
 
 Social position .... 
 
 
 104 
 
 ,, simple method 
 
 . 25 
 
 Soya bean ..... 
 
 
 351 
 
 Phloridzin diabetes 
 
 . 70 
 
 Spinal cord, division of 
 
 
 65 
 
 Phosphatic diabetes 
 
 . 310 
 
 ,, lesions of . 
 
 129 
 
 , 243 
 
 Phthisis .... 
 
 . 207 
 
 Stomach, condition of . 
 
 
 206 
 
 Physiological considerations 
 
 . ■ 51 
 
 Sugar excretion and exercise 
 
 
 171 
 
 Picric acid test . ... 
 
 . 15 
 
 ,, ,, ,, food 
 
 
 171 
 
 Pigmentation of skin in diabetes 
 
 . 308 
 
 ,. ,, diminished by inter- 
 
 
 Pnenmaturia 
 
 . 173 
 
 current affections 
 
 
 172 
 
 Pneumonia, acute croupous . 
 
 . 213 
 
 ,, in blood. See Blood, Sugar 
 
 in. 
 
 
 ,, broncho- . 
 
 . 213 
 
 ,, ,, normal urine . 
 
 
 46 
 
 ,, chronic 
 
 . 212 
 
 ,, ,, urine . . 
 
 
 170 
 
 Potassium iodide . 
 
 . 384 
 
 ,, intravenous injection of 
 
 
 56 
 
 Potatoes . . . 
 
 . 344 
 
 ,, tests ..... 
 
 
 7 
 
 Pregnancy and diabetes 
 
 . 114 
 
 Sympathetic nerve division of 
 
 
 65 
 
 Prognosis .... 
 
 . 313 
 
 ,, nerves and ganglia, changes 
 
 
 Prophylaxis .... 
 
 . 320 
 
 in . 
 
 
 131 
 
 Pruritus 
 
 . 223 
 
 Symptomatology .... 
 
 
 163 
 
 Pulse 
 
 . 217 
 
 Synonyms 
 
 Syphilis 
 
 
 2 
 
 113 
 
 Quantitative estimation of sugar ii 
 
 l urine 39 
 
 
 
 
 
 by Feh- 
 
 Teeth 
 
 
 202 
 
 
 ling's 
 
 Temperature . . . 202, 
 
 277 
 
 , 284 
 
 
 solu- 
 
 Termination 
 
 
 310 
 
 
 tion . 41 
 
 Tests for sugar in the urine . 
 
 
 7 
 
 >j 
 
 by fer- 
 
 Thirst 
 
 
 203 
 
 
 menta- 
 
 ,, relief of 
 
 
 364 
 
 
 tion . 39 
 
 Tongue 
 
 
 203 
 
 31 )> 
 
 byGer- 
 
 Treatment 
 
 
 320 
 
 
 rard's 
 
 ,, by medicines 
 
 
 374 
 
 
 cyano- 
 
 ,, ' dietetic 
 
 
 324 
 
 
 ciipric 
 
 ,, general principles 
 
 
 323 
 
 
 method 44 
 
 Trommer's test .... 
 
 
 8 
 
 >>• ii 
 
 by Pavy's 
 
 Tuberculosis of intestines 
 
 206 
 
 210 
 
 
 solu- 
 
 ,, ,, lungs . 
 
 
 207 
 
 
 tion . 42 
 
 
 
 
 11 11 11 
 
 by pic- 
 
 Ulcer, perforating 
 
 
 226 
 
 „• 
 
 ric acid 45 
 
 Uraemia . . 
 
 
 288 
 
 •> )) ii 
 
 by pola- • 
 
 ,, differential diagnosis 
 
 
 293 
 
 
 rimeter 46 
 
 • Uranium nitrate .... 
 
 
 383 
 
 ., ,, in 
 
 the 
 
 Urea excretion .... 
 
 
 173 
 
 I 
 
 lood 
 
 Uric acid . . . . . 
 
 
 174 
 
 
 191, 195 
 
 Urine, the 
 
 
 166 
 
 Quinine . . . . 
 
 .• 383 
 
 ,, colour of .... 
 ,, odour of 
 
 
 168 
 169 
 
 Racial influence . ■ . 
 
 . 99 
 
 ,, quantity of 
 
 
 166 
 
 Rarity of diabetes 
 
 - 97 
 
 ,, reaction of. 
 
 
 169 
 
 Reaction of blood . 
 
 ' . 186 
 
 ,, specific gravity of 
 
 
 169 
 
 ,, ,, urine . 
 
 . 169 
 
 ,, sugar in the 
 
 
 169 
 
 Reduction tests for sugar 
 
 8 
 
 ,, taste of ... 
 
 
 169 
 
 Reflexes .... 
 
 250, 258 
 
 
 
 
 Retinal changes . 
 
 . 23 
 
 Vagus nerve, irritation of 
 
 
 66 
 
 Robin's alternating treatment 
 
 . 386 
 
 ,, lesions of 
 
 
 130 
 
 Buhner's test 
 
 . 16 
 
 Vasomotor changes 
 
 ,, paralysis, in causation 
 
 of 
 
 227 
 
 Saccharin .... 
 
 . 342 
 
 diabetes . 
 
 
 67 
 
 Saccharine treatment . 
 
 . 387 
 
 Vegetable foods .... 
 
 
 343 
 
 Saccharoses .... 
 
 5 
 
 Vichy ...... 
 
 
 373 
 
 Saffranin test 
 
 . 18 
 
 Villi, intestinal . . . .6 
 
 , 67 
 
 301 
 
 Salicylate of bismuth . 
 
 . 383 
 
 
 
 
 ,, ,, sodium 
 
 . 382 
 
 Wasting. See Emaciation. 162, 
 
 305 
 
 163 
 
 Saliva ..... 
 
 . 202 
 
 Waters, mineral .... 
 
 
 368 
 
 Sex in diabetes 
 
 . 95 
 
 Weight, loss of . . . 162, 
 
 163, 
 
 305 
 
 Sexual excess . . . . 
 
 . 115 
 
 Wines . . . 
 
 
 364 
 
 ,, functions . . . . 
 
 . 237 
 
 Wrist-jerk 
 
 
 258 
 
 Skin ..... 
 
 . 222 
 
 
 
 
 , , action of 
 
 . 367 
 
 Xanthoma 
 
 
 228 
 
 ,, itching of 
 
 . 223 
 
 Yeast in treatment 
 
 
 387 
 
INDEX OF AUTHORS. 
 
 
 T 
 PAGE 
 
 
 
 
 
 PAGE 
 
 Alclehoff . 
 
 75 
 
 Cliauveau 
 
 Allen 
 
 . 18, 44 
 
 Chovstek . 
 
 ... 
 
 Althaus 
 
 263 
 
 Coolen 
 
 
 70 
 
 Aretteus 
 
 2 
 
 Cullen 
 
 
 . • .4 
 
 Armanni . 
 
 .220 
 
 
 
 
 Arthaud . 
 
 .06, 79 
 
 Devic 
 
 
 212, 291 
 
 Arthus 
 
 '80 
 
 Dickinson . 
 
 
 101, 123, 130, 384, 399 
 
 Ashdown . 
 
 . . 10 
 
 Dobson 
 
 
 3 
 
 Asher 
 
 .■ 108 
 
 Dominieis, de . 
 
 
 .... 74 
 
 Auche . . 
 
 264 
 
 Donkin 
 
 
 ... 332 
 
 Auerbach 
 
 251 
 
 Dreschfeld • 
 
 181, 211, 212, 213, 215, 
 
 A.usset 
 
 .'•'■: 389 
 
 240', 270', 271', 272; 277^ 
 281/282, 287, 288, 289 
 
 Bseyer 
 
 17 
 
 Dujardin-Beaumetz ... . . 352 
 
 Baldi 
 
 ... 81 
 
 Duncan, Matthews . . 88, 114, 382 
 
 Barral 
 
 . • . 79' 
 
 
 Baumel 
 
 72 
 
 Ebstein . .... 98, 108, 220, 240, 273, 
 
 Bernard, Claude 
 
 . 51, 54, 68 
 
 302, 329, 338, 355, 382 
 
 Bernardt . 
 
 126 
 
 Eckhard . .. . . . ' • . 65 
 
 Bettmann . 
 
 . . . 132 
 
 Ehrlich 
 
 
 
 
 120, 221 
 
 Biedl 
 
 ... 56 
 
 Eichhorst 
 
 
 
 
 251 
 
 Binet 
 
 . 26, 27 
 
 Einhorn 
 
 
 
 
 22 
 
 Blott 
 
 88 
 
 
 
 
 
 
 Blumenthal 
 
 59 
 
 Eagge " 
 
 
 
 
 399 
 
 Blyth 
 
 335 
 
 Fehling 
 
 
 
 
 8 
 
 Bohland . 
 
 . . 341 
 
 Feinberg » 
 
 
 
 
 113 
 
 Bond 
 
 92 
 
 Fichtner 
 
 
 
 
 221 
 
 Bormenn . 
 
 389 
 
 Fischer, E. 
 
 
 
 
 23 
 
 Bose . 
 
 •99 
 
 Fischer 
 
 
 
 
 107 
 
 Bouchard . 
 
 250 
 
 Finkler 
 
 
 
 
 172, 368 
 
 Bouchardat 
 
 . ■ . 349 
 
 Fleiner 
 
 
 
 
 157 
 
 Brault 
 
 309 
 
 Fliickiger 
 
 
 
 
 27 
 
 Bremer 
 
 197, 198 
 
 Foster, M. 
 
 
 
 
 53 
 
 Bruce 
 
 265, 378 
 
 Fraser 
 
 
 
 
 265 
 
 Briicke 
 
 47 
 
 Frerichs 
 
 89, 96, 116, 117, 224, 130, 202, 
 
 Bruns 
 
 261 
 
 206, 210, 221, 238, 369 
 
 Brunton, Lauder 
 
 344, 367, 368, 382 
 
 Fiirbringer .... 214, 222 
 
 Bunge 
 
 299, 303 
 
 
 Busseuius . 
 
 213 
 
 Gabrischewsky 188 
 
 Butte 
 
 . 66, 79 
 
 Gaillard 
 
 
 
 309 
 
 Buzzard 
 
 251,260 
 
 Gamgee 
 Gans . 
 
 
 
 
 89 
 204 
 
 Camplin . 
 
 350 
 
 Garrod 
 
 
 
 
 123 
 
 (Jantani 
 
 112, 113, 310, 385 
 
 Gaudard 
 
 
 
 
 238 
 
 Carnot 
 
 390 
 
 Gerhardt 
 
 
 
 
 179 
 
 ( !aroe 
 
 103 
 
 Gerrard 
 
 
 
 
 44 
 
 Cartier 
 
 90 
 
 Geyer 
 
 
 
 
 28 
 
 Cawley 
 
 71 
 
 Gilbert 
 
 
 
 
 390 
 
 1 !el ius 
 
 2 
 
 Glenard 
 
 
 
 
 120, 206 
 
 Cerenville . 
 
 389 
 
 Godlee 
 
 
 
 226', 394 
 
 Chadbourne 
 
 398 
 
 Goff, Le 
 
 
 
 195, 196, 198, 385 
 
 Cliarcot 
 
 . 2, 62 
 
 Gonzalez 
 
 
 
 310 
 
 1 li .mi Hard . 
 
 309, 310 
 
 Goolden 
 
 
 
 
 107 
 
416 
 
 INDEX OF A UTHORS. 
 
 
 
 PAGE 
 
 
 PAGE 
 
 Gowers 
 
 231 
 
 Lemoine . 
 
 141 
 
 Gregory . 
 
 4 
 
 Le Nobel . 
 
 182 
 
 Griesinger 
 
 222 
 
 Leo . 
 
 174, 323 
 
 Grube . . 96, 109, 
 
 156, 204, 251, 252, 
 
 Lepine 
 
 . 79, 103, 141, 156, 195, 
 
 318, 329, 
 
 341, 366, 368, 380, 
 
 
 387, 400 
 
 
 392, 396 
 
 Letulle 
 
 309 
 
 
 
 Leube 
 
 173 
 
 Haines 
 
 14 
 
 Levene 
 
 71 
 
 Hallervorden 
 
 174 
 
 Lewaschew 
 
 381 
 
 Hamilton . 
 
 287 
 
 Leyden 
 
 211, 248, 263 
 
 Hanot 
 
 308, 309, 310 
 
 Limbeck . 
 
 188 
 
 Hansemann 
 
 139, 150, 152 
 
 Lindemann 
 
 340 
 
 Harley, G. 
 
 91 
 
 Loeb . 
 
 166 
 
 Harley, Vaughan 
 
 . 58, 76, 290 
 
 Loewy 
 
 195 
 
 Haycraft . 
 
 186, 340 
 
 Ludwig 
 
 87 
 
 Hiklon 
 
 . 76, 81 
 
 Lustig 
 
 68 
 
 Henniker . 
 
 101 
 
 Lyonnet . 
 
 195 
 
 Hermanides 
 
 112 
 
 
 
 Hesse 
 
 399 
 
 MacCann . 
 
 88 
 
 Higgin 
 
 91 
 
 MacDonnell 
 
 55 
 
 Hirsch 
 
 2 
 
 Mackenzie, H. . 
 
 389 
 
 Hirschberg 
 
 . " 232 
 
 Mackenzie, S. . 
 
 220 
 
 Hirschfeld 
 
 183, 205, 273, 274, 
 
 MacMunn . 
 
 180 
 
 
 279, 329, 375 
 
 Maguire 
 
 177, 27S 
 
 Hoffmann . 
 
 . . 25 
 
 Marcus 
 
 81 
 
 Honigman . 
 
 204 
 
 Marie 
 
 195, 196, 310, 385 
 
 Hoppe-Seyler . 
 Huchard . 
 
 . 17, 157, 189 
 297 
 
 Marinesco 
 Maschka . 
 
 256 
 251 
 
 Hugoimenq 
 
 212, 291 
 
 Massary 
 
 310 
 
 
 
 May . 
 
 340 
 
 Imlach 
 
 115 
 
 Mayer 
 
 215 
 
 Israel 
 
 215 
 
 Mering, von 
 
 . 70, 74, 121 
 
 
 
 Minkowski 
 
 71, 74, 81, 325, 340, 369 
 
 Jaccoud 
 
 210 
 
 Minnich 
 
 249 
 
 Jackson, Hughlings . 
 
 256 
 
 Minor 
 
 248 
 
 Jacoby 
 
 . 90, 240 
 
 Moleschott 
 
 384 
 
 Jaeger 
 
 230 
 
 Moore 
 
 7 
 
 Jaksch, von 
 
 23, 87, ISO, 186 
 
 Moritz 
 
 228 
 
 James 
 
 186 
 
 Morris 
 
 228 
 
 Johnson, G. 
 
 . 15, 45, 47 
 
 Mosse 
 
 309 
 
 ,, G. S. . 
 
 47 
 
 Mosso 
 
 58 
 
 Jolles 
 
 22 
 
 Murray 
 
 380 
 
 Jones, Mansell . 
 
 388 
 
 
 
 
 
 Naunyn 
 
 92, 369 
 
 Kalmus 
 
 249 
 
 Nettleship 
 
 230, 236 
 
 Kaltenbach 
 
 88 
 
 Ney . . . 
 
 88 
 
 Kasaliara . 
 
 152 
 
 Nonne 
 
 256 
 
 Kaut'mann . 
 
 . 59, 81, 82 
 
 Noorden, von 
 
 . 85, 110, 182, 185, 186, 
 
 Kirkby . 
 
 300, 338 
 
 
 202, 207, 291, 384, 399 
 
 Klemperer 
 
 293 
 
 
 
 Kobrak 
 
 22 
 
 O'Donnell . 
 
 361 
 
 Koenig 
 
 226, 394 
 
 Ogden 
 
 . , . . 91 
 
 Kratschmer 
 
 378 
 
 Oliver 
 
 16 
 
 Kraus 
 
 . 56, 80, 87 
 
 Oppler 
 
 106 
 
 Kiilz .... 
 
 120, 167, 170, 174. 
 
 Ord . 
 
 113 
 
 
 178, 278, 360, 367 
 
 Osier 
 
 . 98, 310, 378, 398 
 
 Klissmaid . 
 
 270, 276 
 
 
 
 
 
 Paton 
 
 60 
 
 Lancereaux 
 
 156, 207 
 
 Pavy 
 
 32, 42, 47, 54, 59, 64, 
 
 Landenheimer . 
 
 241 
 
 
 65, 67, 91, 96, 190, 
 
 Lannois . 
 
 
 141 
 
 
 301, 352, 376, 379 
 
 Lanz . 
 
 
 87 
 
 Peiper 
 
 68, 112 
 
 Latham . 
 
 
 4 
 
 Pettenkofer 
 
 173 
 
 Leconte . 
 
 
 90 
 
 Poniklo 
 
 131 
 
 Lecorche . 
 
 
 261 
 
 Potier 
 
 310 
 
 Legal . , . 
 
 
 182 
 
 Prout 
 
 121, 362 
 
 Lehtnaun . . 
 
 
 337 
 
 Pryce 
 
 259 
 
 Leichtenstein 
 
 
 186 
 
 Purely 
 
 101, 124, 385 
 
 Leichtenstern . 
 
 
 370 
 
 Pusiuelli . 
 
 
INDEX OF A UTHORS. 
 
 4i7 
 
 
 
 PAGE 
 
 
 PAGE 
 
 
 272 
 
 Strauss 
 Striimpell . 
 
 221 
 '. 86, 110, 158 
 
 Ralfe 
 
 102, 310, 377 
 
 
 
 Rayer 
 
 
 125 
 
 Tait .... 
 
 115, 393 
 
 Reale 
 
 
 74 
 
 Tangl 
 
 58 
 
 Reiner 
 
 
 310 
 
 Tatbam 
 
 102 
 
 Renzi 
 
 
 74 
 
 Taylor, J. . 
 
 256 
 
 Reynolds . 
 
 
 296, 397 
 
 Teubaum . 
 
 175 
 
 Ricbardiere 
 
 
 380 
 
 Tbierfelder 
 
 27 
 
 Riess 
 
 
 370 
 
 Thiroloix . 
 
 81 
 
 Ringer 
 
 
 333 
 
 Tootb 
 
 249 
 
 Roberts, Sir We 
 
 1. 
 
 8, 13, 39, 125, 228, 
 
 Triboulet . 
 
 120 
 
 
 
 378, 392 
 
 Trommer . 
 
 8 
 
 Robin 
 
 
 382, 386 
 
 Tunnicliffe 
 
 368 
 
 Rokitansky 
 
 
 125, 141 
 
 Turner 
 
 88 
 
 Rollo 
 
 
 4 
 
 
 
 Roque 
 
 
 212,291 
 
 Ultzmann . 
 
 25 
 
 Rosenfeld . 
 
 
 . 72, 184 
 
 
 
 Rosenstein 
 
 
 204 
 
 Vogel 
 
 Voit .... 
 
 222 
 . 56, 173 
 
 Salkowski . 
 
 
 33, 180 
 
 
 
 Sanders 
 
 
 287 
 
 Wallacb . 
 
 100, 166 
 
 Sandmeyer 
 
 
 248 
 
 Walter 
 
 288 
 
 Saundby . 
 
 
 119, 131, 164, 234, 
 
 Weber 
 
 124 
 
 
 353, 354, 383, 393 
 
 Wedenski . 
 
 47 
 
 Schiff 
 
 65 
 
 Wegeli 
 
 97 
 
 Scbmitz 
 
 105, 171, 175, 222, 288, 290, 
 
 West 
 
 383 
 
 
 329, 339, 367, 372, 383, 396 
 
 Westpbal . 
 
 116 
 
 Seegen 
 
 . 12, 33, 58, 80, 96, 109, 
 
 Wbite, Hale . 
 
 131, 340, 352, 389 
 
 
 125, 174, 190, 206, 209, 
 
 Williams . 
 
 389, 390 
 
 
 214, 219, 299, 302, 371 
 
 Willis . 
 
 . 2, 121 
 
 Sen . 
 
 
 Windle . 
 
 131 
 
 Senator 
 
 107, 115 
 
 Worm-Midler . 
 
 12 
 
 
 88 
 
 Wrigbt . 
 
 332 
 
 Stadelmann 
 
 175, 279, 290 
 
 
 
 
 336 
 
 Zander 
 
 368 
 
 
 
 361 
 
 Ziemssen, von . 
 
 259 
 
 27 
 
CATALOGUE 
 
 OF 
 
 YOUNG J. PENTLAND'S 
 PUBLICATIONS. 
 
 EDINBURGH : 1 1 TEVIOT PLACE. 
 LONDON : 38 WEST SMITHFIELD, E.C. 
 
 0( TOBER 1897. 
 
CATALOGUE 
 
 OF 
 
 MEDICAL PUBLICATIONS 
 
 ISSUED BY 
 
 YOUNG J. PENTLAND, 
 
 EDINBURGH and LONDON. 
 
 Aitchison, Roberts., M.B., CM., F.R.C.P.Ed., 
 
 Visiting Physician, St. Cuthbert's Poorhouse Hospital, Edinburgh. 
 
 SYNOPSIS OF THERAPEUTICS. Arranged for the use of 
 Prescribers. With Posological Table and an arrangement of the 
 Poisons. i8mo, cloth, pp. xii., 120. Price 3s. 
 
 Alexander, William, M.D., F.R.C.S., 
 
 Honorary Surgeon, Royal Southern Hospital, Liverpool ; Visiting Surgeon, Liverpool 
 Workhouse Hospital ; Acting Honorary Consulting Surgeon, Epileptic Institution, 
 Manor House, Maghull. 
 
 THE TREATMENT OF EPILEPSY. 8vo, cloth, pp. x., 220, 
 with 9 illustrations. Price 7s. 6d. 
 
 Behrens, Dr. W. J. 
 
 TEXT-BOOK OF GENERAL BOTANY. Translated from the 
 Second German Edition. Revised by Patrick Geddes, F.R.S.E., 
 Professor of Botany in the University of Dundee. 8vo, cloth, pp. 
 viii., 374, with 408 illustrations, finely engraved on wood, and 4 
 analytical tables, new and cheaper edition. Price 5s. 
 
 Berry, George A., M.B., F.R.C.S.Ed., 
 
 Ophthalmic Surgeon, Edinburgh Royal Infirmary; Senior Surgeon, Edinburgh Eye 
 lensary ; Lecturer on Ophthalmology, Royal College of Surgeons, Edinburgh. 
 
 DISEASES OF THE EYE. A Practical Treatise for Students of 
 Ophthalmology. Second Edition, 8vo, cloth, pp. xvi., 730, 
 
4 CATALOGUE OF MEDICAL PUBLICATIONS 
 
 thoroughly revised and illustrated with many additional coloured 
 plates from original drawings. Price 25s. (Pentland's Medical 
 Series, Volume Second.) 
 
 THE ELEMENTS OF OPHTHALMOSCOPIC DIAGNOSIS. 
 For the use of Students attending Ophthalmic Practice. Crown 
 8vo, cloth, pp. xii., 83. Price 3s. 6d. 
 
 Billings, John S., A.M., M.D., LL.D., Harv. and 
 Edin., D.C.L., Oxon., 
 
 Member of the National Academy of Sciences, Surgeon, U.S.A., &c. 
 
 THE NATIONAL MEDICAL DICTIONARY. Including Eng- 
 lish, French, German, Italian, and Latin Technical Terms used in 
 Medicine and the Collateral Sciences, and a Series of Tables of 
 useful data. With the collaboration of W. O. Atwater, M.D., 
 James M. Flint, M.D., S. M. Burnett, M.D., H. C. Yarrow, 
 M.D., Frank Baker, M.D., R. Lorini, M.D., J. H. Kidder, 
 M.D., William Lee, M.D., C. S. Minot, M.D., Washington 
 Matthews, M.D., W. T. Councilman, M.D. In two very 
 handsome Imperial 8vo volumes, containing about 1600 pages. 
 Price 50s. nett. 
 
 Bramwell, Byrom, M.D., F.R.C.P.Ed., 
 
 Lecturer on the Principles and Practice of Medicine, and on Practical Medicine and 
 Medical Diagnosis, in the Extra-Academical School of Medicine, Edinburgh ; 
 Assistant Physician, Edinburgh Royal Infirmary. 
 
 DISEASES OF THE HEART AND THORACIC AORTA. 
 
 Large 8vo, cloth, pp. xvi., 783. Illustrated with 226 wood 
 engravings, and 68 pages of lithograph plates, exhibiting 91 
 figures — 317 illustrations in all. Price 25s. 
 
 INTRACRANIAL TUMOURS. 8vo, cloth, pp. xvi., 270, with 
 116 illustrations. Price 14s. 
 
 STUDIES IN CLINICAL MEDICINE. In one handsome 
 volume, crown 4to, cloth, bevelled boards, pp. xii., 344. Illustrated 
 with numerous wood engravings and full-page lithograph plates, 
 some coloured. Price 5s. 6d. nett. 
 
 Bruce, Alexander, M.D., F.R.C.P.Ed., 
 
 Lecturer on Pathology in the School of Medicine, Edinburgh ; Assistant Physician 
 (formerly Pathologist), Edinburgh Royal Infirmary; Pathologist to the Royal 
 Hospital for Sick Children. 
 
 ILLUSTRATIONS OF THE NERVE TRACTS IN THE MID 
 AND HIND BRAIN AND THE CRANIAL NERVES 
 ARISING THEREFROM. Royal 4 to. Illustrated with a series 
 of 27 coloured plates from original drawings, and numerous figures 
 throughout the text. Price 50s. nett. 
 
ISSUED BY YOUNG J. PENTLAND. 5 
 
 Bruce, J. Mitchell, M.D., F.R.C.P., 
 
 Physician to, and Lecturer on Medicine, Charing Cross Hospital; Physician, Con- 
 sumption Hospital, Brompton, London. 
 
 THE PRINCIPLES OF TREATMENT. In Press, 8vo, pp. xvi., 
 500. (Pentland's Medical Series, Volume Eighth.) 
 
 Bruen, E. T., M.D., 
 
 Assistant Professor of Physical Diagnosis, University of Pennsylvania ; one of the 
 Physicians to the Philadelphia and University Hospitals. 
 
 OUTLINES FOR THE MANAGEMENT OF DIET; or, The 
 Regulation of Food to the Requirements of Health and the Treat- 
 ment of Disease. Crown 8vo, cloth, pp. 138. Price 4s. 6d. 
 
 Burnett, Charles Henry, A.M., M.D., 
 
 Aural Surgeon to the Presbyterian Hospital ; one of the Consulting Aurists to the 
 Pennsylvania Institution for the Deaf and Dumb ; Lecturer on Otology, Women's 
 Medical College of Pennsylvania, Philadelphia. 
 
 DISEASES AND INJURIES OF THE EAR : their Prevention 
 and Cure. Crown 8vo, cloth, pp. 154, with 5 illustrations. Price 
 4s. 6d. 
 
 Carmichael, James, M.D., F.R.C.P.Ed., 
 
 Physician, Royal Hospital for Sick Children ; University Lecturer on Disease in 
 Children, Edinburgh. 
 
 DISEASE IN CHILDREN: a Manual for Students and Practi- 
 tioners. Crown 8vo, cloth, pp. xvi., 520. Illustrated with charts. 
 Price 10s. 6d. (Pentland's Sfttdenls' Manuals.) 
 
 Cheyne, W. Watson, F.R.S., F.R.C.S., 
 
 Professor of Surgery, King's College ; Surgeon to King's College Hospital, and 
 Paddington Green Children's Hospital, London. 
 
 TUBERCULOUS DISEASE OF BONES AND JOINTS: Its 
 
 Pathology, Symptoms, and Treatment. 8vo, cloth, pp. xvi., 374. 
 Illustrated with numerous wood engravings throughout the text. 
 Price 14s. nett. 
 
 THE TREATMENT OF WOUNDS, ABSCESSES, AND 
 ULCERS. Second Edition. Crown 8vo, cloth, pp. xii., 197. 
 Price 3s. 6d. 
 
 SUPPURATION AND SEPTIC DISEASES. Three Lectures 
 delivered at the Royal College of Surgeons of England. 8vo, 
 cloth, pp. xii., 102, with 4 illustrations. Price 5s. 
 
CATALOGUE OF MEDICAL PUBLICATIONS 
 
 Crocker, H. Radcliffe, M.D., F.R.C.P., 
 
 Physician to the Department for Diseases of the Skin, University College Hospital, 
 London. 
 
 ATLAS OF THE DISEASES OF THE SKIN. In a series of 
 Illustrations from Original Drawings with Descriptive Letterpress. 
 Complete in Sixteen Fasciculi, containing 96 plates, exhibiting 
 238 figures printed in colours, from original drawings specially 
 prepared for the work. Price 21s. each Fasciculus, and sold only 
 by subscription. Also in 2 vols., handsomely bound in half 
 morocco, gilt tops, the plates mounted on linen guards, — Price 
 18 Guineas, nett. 
 
 ' Compend " Series, The. 
 
 A Series of Handbooks to assist Students preparing for 
 Examinations. 
 
 COMPEND OF HUMAN ANATOMY, INCLUDING THE 
 ANATOMY OF THE VISCERA. By Samuel O. L. Potter, 
 M.D., M.R.C.P. (Lond.), Cooper Medical College, San Francisco. 
 Fifth Edition, revised and enlarged, crown 8vo, cloth, pp. 289, 
 with 117 engravings and 16 full-page plates. Price 5s. 
 
 COMPEND OF THE PRACTICE OF MEDICINE. By Daniel 
 E. Hughes, M.D., late Demonstrator of Clinical Medicine in the 
 Jefferson Medical College of Philadelphia. Fourth Edition, revised 
 and enlarged, crown 8vo, cloth, pp. 328. Price 7s. 6d. 
 
 COMPEND OF OBSTETRICS. By Henry G. Landis, A.M., 
 M.D., late Professor of Obstetrics and Diseases of Women in 
 Starling Medical College. Third Edition, thoroughly revised, en- 
 larged, and improved, crown 8vo, cloth, pp. 118, with 17 illustra- 
 tions. Price 4s. 6d. 
 
 COMPEND OF SURGERY. By Orville Horwitz, B.S., M.D., 
 Chief of the Outdoor Surgical Department, Jefferson Medical 
 College Hospital. Fourth Edition, revised, crown 8vo, cloth, pp. 
 272, with 136 illustrations. Price 5s. 
 
 COMPEND OF DISEASES OF CHILDREN. By Marcus P. 
 Hatfield, A.M., M.D., Professor of Diseases of Children, Chicago 
 Medical College. Crown 8vo, cloth, pp. 186, with coloured plate. 
 Price 4s. 6d. 
 
 COMPEND OF PATHOLOGY AND MORBID ANATOMY. 
 By H. Newbery Hall, Ph.G., M.D., Professor of Pathology and 
 Medical Chemistry, Post Graduate Medical School ; Surgeon to the 
 Emergency Hospital, &c, Chicago. Crown 8vo, cloth, pp. 204, 
 with 91 illustrations. Price 4s. 6d. 
 
ISSUED BY YOUNG J. PENTLAND. 7 
 
 COMPEND OF DENTAL PATHOLOGY AND DENTAL 
 MEDICINE. By Geo. W. Warren, D.D.S., Clinical Chief, 
 Pennsylvania College of Dental Surgery. Crown 8vo, cloth, pp. 
 109, illustrations. Price 4s. 6d. 
 
 Cunningham, D. J., M.D., F.R.S., 
 
 Professor of Anatomy and Chirurgery, Trinity College, Dublin. 
 
 MANUAL OF PRACTICAL ANATOMY. Second Edition. In 
 2 vols., crown 8vo, cloth, illustrated with 372 wood engravings, and 
 2 coloured plates. Vol. I. — Upper Limb, Lower Limb, Abdomen; 
 Vol. II. — Thorax, Head and Neck. Price per volume, 12s. 6d. 
 (Pentland's Students' Manuals.') 
 
 Davidson, Andrew, M.D., F.R.C.P.Ed., 
 
 Late Visiting and Superintending Surgeon, Civil Hospital; Professor of Chemistry, 
 Royal College, Mauritius. 
 
 GEOGRAPHICAL PATHOLOGY. An inquiry into the Geogra- 
 phical Distribution of Infective and Climatic Diseases. In 2 vols., 
 large 8vo, cloth, pp. xvi., 1008. Illustrated with maps and charts. 
 Price 31s. 6d. 
 
 HYGIENE AND DISEASES OF WARM CLIMATES. In a 
 Series of Articles by Eminent Authorities. Edited by Andrew 
 Davidson, M.D., F.R.C.P.Ed., Late Visiting and Superintending 
 Surgeon, Civil Hospital ; Professor of Chemistry, Royal College, 
 Mauritius ; Author of " Geographical Pathology." The Articles 
 are contributed by Sir Joseph Fayrer; Drs. Macnamara; 
 Patrick Manson ; Lane Notter ; E. A. Birch ; R. W. Cop- 
 pinger; David Bruce; G. M. Sternberg; Montague Lub- 
 bock ; Hv. Cayley ; Sonsino ; The Editor ; &c. &c. One 
 volume, royal 8vo, cloth, pp. xx., 1014. Illustrated with engravings 
 and full-page plates. Price 31s. 6d. 
 
 Davis, Edward P., A.M., M.D., 
 
 Professor of Obstetrics and Diseases of Infancy in the Philadelphia Polyclinic ; Clini" 
 cal Professor of Obstetrics, Jefferson Medical College, Philadelphia. 
 
 A TREATISE ON OBSTETRICS FOR STUDENTS AND 
 PRACTITIONERS. Large 8vo, cloth, pp. 553, illustrated with 
 217 engravings and 30 plates in colours and monochrome. Price 
 16s. nett. 
 
 Dentistry, The American System of. 
 
 In Treatises by Various Authors. Edited by Wilbur F. Litch, 
 M.D., D.D.S., Professor of Prosthetic Dentistry, Therapeutics, 
 
CATALOGUE OF MEDICAL PUBLICATIONS 
 
 and Materia Medica, in the Pennsylvania College of Dental 
 Surgery, Philadelphia. In three handsome volumes, royal 8vo, cloth, 
 containing about iooo pages each, with fully 1700 elaborate 
 illustrations. Price 30s. per volume, nett. For Sale by Subscrip- 
 tion only. 
 
 Dercum, Francis X., A.M., M.D., Ph.D., 
 
 Clinical Professor of Nervous Diseases, Jefferson Medical College, Philadelphia ; 
 President of the American Neurological Association. 
 
 A TEXT-BOOK ON NERVOUS DISEASES. By American 
 Authors. Royal 8vo, cloth, pp. xvi , 1056. Illustrated with 341 
 engravings and 7 coloured plates. Price 25s. nett. 
 
 *.£* The list of contributors includes the following well-known 
 names : — 
 
 N. E. Brill, M.D. James Hendrie Lloyd, M D. 
 
 Charles W. Burr, M.D. Charles K. Mills, M.D. 
 
 Joseph Collins, M.D. S. Weir Mitchell, M.D., LL.D. 
 
 Charles L. Dana, M.D. Charles A. Oliver, M.D. 
 
 F. X. Dercum, M.D. William Osler, M.D. 
 
 E. D. Fisher, M.D. Frederick Peterson, M.D. 
 
 Landon Carter Gray, M.D. Morton Prince, M.D. 
 
 C. A. Herter, M.D. G. E. de Schweinitz, M.D. 
 
 George W. Jacoby, M.D. Wharton Sinkler, M.D. 
 
 W. W. Keen, M.D., LL.D. M. Allen Starr, M.D. 
 
 Philip Coombs Knapp, M.D. James C. Wilson, M.D. 
 
 De Schweinitz, G. E., A.M., M.D., 
 
 Professor of Ophthalmology in the Philadelphia Polyclinic ; Clinical Professor of 
 Ophthalmology in the Jefferson Medical College, Philadelphia. 
 
 TOXIC AMBLYOPIAS; their Classification, History, Symptoms, 
 Pathology, and Treatment. Royal 8vo, cloth, pp. 238, with 46 
 figures in the text, and 9 full-page plates. Price 16s. nett. 
 
 Edinburgh Hospital Reports. 
 
 Edited by G. A. Gibson, M.D., D.Sc. ; C. W. Cathcart, M.A., 
 M.B. ; John Thomson, M.D. ; and D. Berry Hart, M.D. 
 Issued annually, 8vo, cloth, pp. xvi., 650 or thereby, handsomely 
 printed. Illustrated with full-page plates and engravings. Price per 
 volume, 12s. 6d. nett. Carriage free. Volumes I. II. III. IV. and 
 V. now ready. 
 
ISSUED BY YOUNG J. PEN TL AND. 
 
 Edinburgh Medical Journal. 
 
 Edited by G. A. Gibson, M.D., D.Sc, F.R.C.P.Ed., Assistant 
 Physician, Edinburgh Royal Infirmary. 
 
 Assisted in Special Depart7?ients by : — 
 
 A. Lockhart Gillespie, M.D., F.R.C.P.Ed., Medical Registrar, 
 Edinburgh Royal Infirmary {Medicine). 
 
 Alexis Thomson, M.D., F.R.C.S.Ed., Assistant Surgeon, Edinburgh 
 Royal Infirmary (Surgery). 
 
 Francis W. N. Haultain, M.D., F.R.C.P.Ed., Physician for Diseases 
 of Women, Deaconess Hospital, Edinburgh (Obstetrics and Gy?ice- 
 cology). 
 
 Ralph Stockman, M.D., F.R.C.P.Ed., Professor of Materia Medica 
 in the University of Glasgow (Therapeutics). 
 
 R. F. C. Leith, M.A., M.B., B.Sc, F.R.C.P.Ed., Assistant Physician, 
 Edinburgh Royal Infirmary (Pathology). 
 
 D. J. Cunningham, M.D., D.Sc, F.R.S., LL.D., D.C.L., Professor 
 of Anatomy in the University of Dublin (Applied Anatomy). 
 
 Thos. H. Milroy, M.D., B.Sc, Demonstrator of Physiology, Uni- 
 versity of Edinburgh (Physiology). 
 
 J. Batty Tuke, M.D., D.Sc, F.R.C.P.Ed., Physician Superintendent, 
 Saughton Hall Asylum; and J. MacPherson, M.B., F.R.C.P.Ed., 
 Medical Superintendent, Stirling District Asylum (Insanity). 
 
 W. Allan Jamieson, M.D., F.R.C.P.Ed., Physician for Diseases of 
 the Skin, Edinburgh Royal Infirmary (Diseases of the Skin). 
 
 P. M'Bride, M.D., F.R.C.P.Ed., Aural Surgeon and Laryngologist, 
 Edinburgh Royal Infirmary; and A. Logan Turner, M.D., 
 F.R.C.S.Ed., Surgeon for Diseases of Ear and Throat, Deaconess 
 Hospital, Edinburgh (Diseases of the Ear, Throat, and Nose). 
 
 George A. Berry, M.B., F.R.C.S.Ed., Ophthalmic Surgeon, Edin- 
 burgh Royal Infirmary; and W. G. Sym, M.D., F.R.C.S.Ed., 
 Senior Assistant Ophthalmic Surgeon, Edinburgh Royal Infirmary 
 (Diseases of the Eye). 
 
 G. H. Melville Dunlop, M.D., F.R.C.P.Ed., Extra Physician, 
 Royal Hospital for Sick Children, Edinburgh (Diseases of Children). 
 
 C. Hunter Stewart, M.B., D.Sc, F. R.S.E., Demonstator of Public 
 Health and Medical Jurisprudence, University of Edinburgh 
 (Public Health). 
 
 The Volumes begin with the Issues for January and July. 
 
 Subscription One Guinea per annum (in advance) post free. For the 
 Colonies and Abroad, Twenty-four Shilling's per annum (in 
 advance) post free. Single Numbers Two Shillings each. 
 
 Ewald, Dr. C. A., 
 
 Extraordinary Professor of Medicine at the University of Berlin; Director of the 
 Augusta Hospital. 
 
 DISEASES OF THE STOMACH. Authorised translation, with 
 special additions by the author, by Morris Manges, A.M., M.D., 
 
io CATALOGUE OF MEDICAL PUBLICATIONS 
 
 attending Physician, Mount Sinai Hospital, New York City. Large 
 8vo, cloth, pp. xvi., 498, with 30 illustrations. Price 16s. 
 
 Felkin, R. W., M.D., F.R.S.E., F.R.G.S., 
 
 Lecturer on Diseases of the Tropics and Climatology, School of Medicine, Edinburgh. 
 
 GEOGRAPHICAL DISTRIBUTION OF SOME TROPICAL 
 DISEASES AND THEIR RELATION TO PHYSICAL 
 PHENOMENA. Svo, cloth, pp. 54. Illustrated with 16 coloured 
 maps. Price 5s. 
 
 Frost, W. Adams, F.R.C.S., 
 
 Ophthalmic Surgeon and Lecturer on Ophthalmic Surgery, St. George's Hospital ; 
 Surgeon, Royal Westminster Ophthalmic Hospital, London. 
 
 THE FUNDUS OCULI, with an Ophthalmoscopic Atlas, illus- 
 trating its Physiological and Pathological Conditions. In one 
 handsome 4to volume, extra cloth, gilt top, with 47 plates, 
 exhibiting 107 beautifully coloured figures, and numerous engrav- 
 ings in the text. Price 63s. nett. 
 
 Fuller, Eugene, M.D., 
 
 Instructor in Genito-Urinary and Venereal Diseases, Post-Graduate Medical School, 
 New York. 
 
 DISORDERS OF THE MALE SEXUAL ORGANS. Large 8vo, 
 cloth, pp. 242. Illustrated with 8 plates and 25 figures in the text. 
 Price 9s. 
 
 Gibbes, Heneage, M.D., 
 
 Professor of Pathology in the University of Michigan ; formerly Lecturer on Histology 
 in the Medical School, Westminster Hospital. 
 
 PRACTICAL PATHOLOGY AND MORBID HISTOLOGY. 
 
 Svo, cloth, pp. 362. Illustrated with 60 photographic reproduc- 
 tions. Price 12s. 6d. 
 
 Gibson, G. A., IVLD., D.Sc, F.R.C.P.Ed., 
 
 Lecturer on the Principles and Practice of Medicine in the Edinburgh Medical 
 School ; Assistant Physician, Edinburgh Royal Infirmary. 
 
 DISEASES OF THE HEART AND CIRCULATION. Svo, 
 pp. xvi., 550, or thereby. Illustrated with 100 figures in the text. 
 In Press. {Pentland's Medical Series, Volume Sixth.) 
 
 TEXT BOOK OF MEDICINE. By British Teachers. Edited by 
 G. A. Gibson, M.D., D.Sc. In Press, royal Svo, pp. xxiv., 1250, 
 with illustrations. 
 
 . x ;\ The following, among others, will contribute : — Professors W. T. 
 Gairdner, Sir T. Grainger Stewart, Thomas Oliver; Sir 
 
ISSUED BY YOUNG J. PENTLAND. 
 
 William R. Gowers, Doctors J. Mitchell Bruce, R. Maguire, 
 F. W. Mott, Stephen Mackenzie, Hector Mackenzie, Hale 
 White, Lauder Brunton, Sidney Martin, W. Pasteur, A. A. 
 Kanthack, Patrick Manson, J. W. Moore, Wm. Hunter, W. 
 Allan Jamieson, J. O. Affleck, William Russell, Alex. 
 Bruce, R. W. Philip, the Editor, &c. &c. 
 
 Gibson, G. A., M.D., D.Sc, F.R.C.P.Ed., 
 
 Lecturer on the Principles and Practice of Medicine in the Edinburgh Medical 
 School ; Assistant Physician, Edinburgh Royal Infirmary ; 
 
 and 
 
 Russell, William, M.D., F.R.C.P.Ed., 
 
 Assistant Physician, Edinburgh Royal Infirmary ; Lecturer on Pathology 
 and Morbid Anatomy in the Edinburgh Medical School. 
 
 PHYSICAL DIAGNOSIS. A Guide to Methods of Clinical 
 Investigation. Second Edition, crown 8vo, cloth, pp. xvi., 376, with 
 109 illustrations, some coloured. Price, 10s. 6d. (Pent/and's 
 Studen is 1 Ma nuals.) 
 
 Graham, James, M.A., M.D., 
 
 Late Demonstrator of Anatomy, Sydney University ; Medical Superintendent, Prince 
 Alfred Hospital, Sydney. 
 
 HYDATID DISEASE IN ITS CLINICAL ASPECTS. Svo, 
 cloth, pp. xvi., 204, with 34 full-page coloured plates. Price 16s. 
 
 Hall, H. Newbery, Ph.G., M.D., 
 
 Professor of Pathology and Medical Chemistry, Post Graduate Medical School ; 
 Surgeon to the Emergency Hospital, &c, Chicago. 
 
 COMPEND OF PATHOLOGY AND MORBID ANATOMY, 
 
 Crown 8vo, cloth, pp. 204, with 91 illustrations. Price 4s. 6d. 
 
 Hare, Hobart Amory, M.D., B.Sc, 
 
 Clinical Professor of the Diseases of Children and Demonstrator of Therapeutics in 
 the University of Pennsylvania; Physician to St. Agnes's Hospital and to the 
 Medical Dispensary of the Children's Hospital, Philadelphia. 
 
 SYSTEM OF PRACTICAL THERAPEUTICS. By Various 
 Authors. Edited by Hobart Amory Hare, M.D. In 6 volumes, 
 royal Svo, cloth, of about 500 pages each. Uniform with the 
 "Cyclopaedia of Children's Diseases" and "Systems of Gynaecology 
 and Obstetrics." Price per volume, 12s. 6d. nett, carnage free. 
 
 Supplement, bringing the work up to date, in two volumes, 
 royal Svo, cloth. Price 25s. nett, carriage free. 
 
12 CATALOGUE OF MEDICAL PUBLICATIONS 
 
 Hatfield, Marcus P., A.M., M.D., 
 
 Professor of Diseases of Children, Chicago Medical College. 
 
 COMPEND OF DISEASES OF CHILDREN. Crown 8vo, cloth, 
 pp. 1 86, with coloured plate. Price 4s. 6d. 
 
 Haultain, F. W. N., M.D., F.R.C.P.Ed., 
 
 Physician to the Royal Dispensary ; late Clinical Assistant to Physician for Diseases 
 of Women, Royal Infirmary, Edinburgh ; and 
 
 Ferguson, J. Haig, M.D., F.R.C.P.Ed., 
 
 Physician to the New Town Dispensary; late Resident Physician, Royal 
 Maternity Hospital, Edinburgh. 
 
 HANDBOOK OF OBSTETRIC NURSING. Second Edition, 
 revised and enlarged, crown 8vo, cloth, pp. xvi., 244. Illustrated 
 with 2)Z wood engravings. Price 5s. 
 
 Hayem, Georges, M.D., 
 
 Professor of Clinical Medicine in the Faculty of Medicine of Paris. 
 
 PHYSICAL AND NATURAL THERAPEUTICS : the Remedial 
 Uses of Atmospheric Pressure, Climate, Heat and Cold, Hydro- 
 therapeutic Measures, Mineral Waters, and Electricity. Edited by 
 Hobart Amory Hare, M.D., Professor of Therapeutics and 
 Materia Medica, Jefferson Medical College, Philadelphia. Large 
 8vo, cloth, pp. 426, with 113 illustrations in the text. Price 14s. 
 
 Hirst, Barton Cooke, M.D., 
 
 Professor of Obstetrics in the University of Pennsylvania ; and 
 
 Piersol, George A., M.D., 
 
 Professor of Embryology and Histology in the University of Pennsylvania. 
 
 HUMAN MONSTROSITIES. In handsome folio, containing 
 about 230 pages of text, illustrated with engravings and 39 full- 
 page photographic plates from Nature. In four fasciculi. Price 
 25s. each nett, carriage free. The edition is limited, and is for 
 sale only by subscription. 
 
 Holland, J. W., M.D., 
 
 Professor of Medical Chemistry and Toxicology, Jefferson Medical College, Phila- 
 delphia. 
 
 THE URINE AND THE COMMON POISONS, MEMORANDA, 
 CHEMICAL AND MICROSCOPICAL, FOR LABORATORY 
 USE. Second Edition, revised and enlarged, oblong crown 8vo. 
 cloth, pp. 65, with 28 illustrations. Price, 4s. 
 
ISSUED BY YOUNG J. PENTLAND. 13 
 
 Horwitz, Orville, B.S., M.D., 
 
 Chief of the Outdoor Surgical Department, Jefferson Medical College Hospital, Phila- 
 delphia. 
 
 COMPEND OF SURGERY. Fourth Edition, revised, crown 8vo, 
 cloth, pp. 272, with 136 illustrations. Price 5s. 
 
 Hughes, Daniel E., M.D., 
 
 Late Demonstrator of Clinical Medicine in the Jefferson Medical College, Phila- 
 delphia. 
 
 COMPEND OF THE PRACTICE OF MEDICINE. Fourth 
 Edition, revised and enlarged, crown 8vo, cloth, pp. 328. 
 Price 7s. 6d. 
 
 Hygiene and Diseases of Warm Climates. 
 
 (See Davidson.) 
 
 James, Alex., M.D., F.R.C.P.Ed., 
 
 Lecturer on the Principles and Practice of Medicine in the School of Medicine, Edin- 
 burgh ; Assistant Physician, Edinburgh Royal Infirmary. 
 
 PULMONARY PHTHISIS. Its Etiology, Pathology, and Treat- 
 ment. Svo, cloth, pp. xii., 2S5. Price 9s. 
 
 Jamieson, W. Allan, M.D., F.R.C.P.Ed., 
 
 Extra Physician for Diseases of the Skin, Edinburgh Royal Infirmary ; Consulting 
 Physician, Edinburgh City Hospital ; Lecturer on the Diseases of the Skin, School 
 of Medicine, Edinburgh. 
 
 DISEASES OF THE SKIN. A Manual for Students and Prac- 
 titioners. Fourth Edition, revised and enlarged, 8vo, cloth, pp. 
 xvi., 676, with coloured illustrations. Price 21s. (Pentland's 
 Medical Series, Volume First.) 
 
 Johnstone, Alexander, F.Q.S., 
 
 Lecturer on Botany, School of Medicine, Edinburgh. 
 
 BOTANY. A Concise Manual for Students of Medicine and 
 Science. Crown Svo, cloth, pp. xvi., 260, with 164 illustrations 
 and a series of floral diagrams. Price 6s. (Pentland's Students' 
 Manuals.) 
 
 Journal of Pathology and Bacteriology. 
 
 Edited, with the Collaboration of Distinguished British and Foreign 
 Pathologists, by German Sims Woodhead, M.D., Director of the 
 Laboratories of the Royal Colleges of Physicians (London) and 
 
14 CATALOGUE OF MEDICAL PUBLLCATLONS 
 
 Surgeons (England), Assisted in Special Departments by Sidney 
 Martin, M.D. (Lond.) {Pathological Chemistry); S. G. Shattock, 
 F.R.C.S. (Morbid Anatomy and Histology); G. E. Cartwright 
 Wood, M.D. (Edin.) (Bacteriology); C. S. Sherrington, M.D., 
 (Cantab.) (Experimental Pathology). Subscription, One Guinea 
 per annum (in advance), post free. 
 
 It has been felt for some time that the want in this country of a 
 Journal dealing specially with General and Experimental Pathology 
 has militated most seriously against the free interchange of ideas, 
 not only between English-speaking pathologists, but also between 
 British and Foreign workers. 
 
 Although the Transactions of the Pathological Society deal with 
 communications which are brought before its members, there is no 
 medium in which longer articles, and especially those from workers 
 throughout the United Kingdom and abroad, can be brought 
 before a less limited circle of readers. It has been thought 
 desirable, therefore, to found a journal specially devoted to the 
 publication of original contributions to General Pathology, Patho- 
 logical Anatomy, and Experimental Pathology, including 
 Bacteriology. These articles will, of course, be mainly from 
 British Laboratories and Hospitals ; but the co-operation of many 
 distinguished Continental, American, and Colonial Pathologists has 
 been obtained, and papers written or edited by them will be placed 
 before our readers. 
 
 In order to increase the interest and extend the usefulness of the 
 Journal, it is intended that, in addition to original articles, critical 
 summaries of work done in special departments of Pathology and 
 Bacteriology shall from time to time be published. All articles 
 appearing in the Journal will be signed. 
 
 The Journal appears at least four times a year, but it will be 
 issued more frequently if necessary, in order to ensure publication 
 of all papers as early as possible after they are received. 
 
 The numbers issued throughout the year form a volume, royal 8vo, 
 of about 500 pages. They are printed on good paper, and freely 
 illustrated with Woodcuts and Chromolithographs. 
 
 Amongst those who collaborate are the following : — Sir Henry 
 Acland, Bart, Oxford ; J. G. Adami, Montreal ; S. Arloing, Lyons ; 
 B. Bang, Copenhagen ; Ch. Bouchard, Paris ; R. Boyce, Liver- 
 pool ; J. Rose Bradford, London ; H. Buchner, Munich ; Sir 
 Charles Cameron, Dublin ; Angelo Celli, Rome ; A. Chantemesse, 
 Paris ; A. B. Charrin, Paris ; A. Chauveau, Paris ; W. Watson 
 Cheyne, London ; H. Chiari, Prague ; Joseph Coats, Glasgow ; 
 W. T. Councilman, Boston ; D. D. Cunningham, Calcutta ; 
 S. Delepine, Manchester; J. Dreschfeld, Manchester; D. Drum- 
 mond, Newcastle; Von Esmarch, Koenigsberg; Ch. Firket, Liege; 
 R. H. Fitz, Boston ; P. Grawitz, Griefswald ; W. S. Greenfield, 
 Edinburgh ; E. H. Hankin, Agra ; H. Heiberg, Christiania ; 
 Victor Horsley, London ; F. Hueppe, Prague ; O. Israel, Berlin ; 
 
ISSUED BY YOUNG J. PENTLAND. 15 
 
 E. Klebs, Zurich ; Alfred Lingard, Poonah ; Lord Lister, London 
 O. Lubarsch, Rostock ; P. Marie, Paris ; E. Metchnikoff, Paris 
 
 F. W. Mott, London ; E. Nocard, Alfort ; T. Oliver, Newcastle 
 J. Orth, Gottingen; William Osier, Baltimore; J. F. Payne 
 London ; T. Mitchell Prudden, New York ; J. M. Purser, Dublin 
 J. C. Salomonsen, Copenhagen ; J. Burdon Sanderson, Oxford 
 A. M. Stalker, Dundee; J. Lindsay Steven, Glasgow; H. Stilling 
 Lausanne ; I. Straus, Paris ; T. P. Anderson Stuart, Sydney 
 R. Thoma, Dorpat ; J. Batty Tuke, Edinburgh ; L. Vaillard, Paris 
 Rudolf Virchow, Berlin; H. Marshall Ward, Cambridge; A. Weich- 
 selbaum, Vienna ; C. Weigert, Frankfort a/M ; W. H. Welch, 
 Baltimore ; Samuel Wilks, London ; A. E. Wright, Netley ; Von 
 Zenker, Erlangen ; E. Ziegler, Freiburg. 
 
 Keating, John M., M.D., LL.D., 
 
 Fellow of the College of Physicians of Philadelphia ; late Visiting Obstetrician to the 
 Philadelphia Hospital, and Lecturer on Diseases of Women and Children, Phila- 
 delphia Hospital. 
 
 CYCLOPAEDIA OF THE DISEASES OF CHILDREN, 
 MEDICAL AND SURGICAL. The Articles written especially 
 for the work by American, British, and Canadian Authors. Edited 
 by John M. Keating, M.D. In 8 vols., royal 8vo, of about 500 
 pages each. Illustrated with wood engravings in the text, and 
 numerous full-page plates. Price 12s. 6d. per volume nett, 
 carriage free. Detailed prospectus on application. 
 
 A NEW PRONOUNCING DICTIONARY OF MEDICAL 
 TERMS. Handsome 8vo, cloth, pp. Si 8. Price 18s. 
 
 Keating, John M., M.D., LL.D., 
 
 and 
 
 Coe, Henry C, M.D., M.R.C.S., 
 
 Professor of Gynaecology, New York Polyclinic. 
 
 CLINICAL GYNAECOLOGY, MEDICAL AND SURGICAL. 
 By American Teachers. 2 vols., royal 8vo, cloth extra, gilt tops, 
 pp. xviii., 994. Illustrated with 34 full-page plates and 377 
 engravings in the text. Price 25s. nett. 
 
 Keith, Skene, F.R.C.5., 
 
 Assisted by 
 
 Keith, George E., M.B., CM. 
 
 TEXT-BOOK OF ABDOMINAL SURGERY. A Clinical Manual 
 for Practitioners and Students. 8vo cloth, gilt top, pp. xvi., 508. 
 Price 16s. (Pentland's Medical Series, Volume Fourth.) 
 
1 6 CATALOGUE OF MEDICAL PUBLICATIONS 
 
 Kenwood, H. R., M.B., CM., L.R.C.P. (Lond.). 
 
 THE ESSENTIALS OF MEDICAL ANATOMY. i 2 mo, cloth, 
 pp. 52. Price 2s. 
 
 Landis, Henry G., A.M., M.D., 
 
 Late Professor of Obstetrics and Diseases of Women in Starling Medical College. 
 
 COMPEND OF OBSTETRICS. Third Edition, thoroughly revised, 
 enlarged, and improved, crown 8vo, cloth, pp. 118, with 17 
 illustrations. Price 4s. 6d. 
 
 Lawless, E. J., M.D., 
 
 Surgeon-Major, 4th V. B. East Surrey Regiment. 
 
 FIRST AID TO THE INJURED AND MANAGEMENT OF 
 THE SICK. An Ambulance Handbook and Elementary Manual 
 of Nursing for Volunteer Bearers and Others. Crown 8vo, cloth, 
 pp. xvi., 262, with plate and numerous wood engravings. 
 Price 3s. 6d. 
 
 Leuckart, Rudolf, 
 
 Professor of Zoology and Comparative Anatomy in the University of Leipsic. 
 
 THE PARASITES OF MAN AND THE DISEASES WHICH 
 PROCEED FROM THEM. A Text-Book for Students and 
 Practitioners. Translated from the German with the co-operation 
 of the Author by William E. Hoyle, M.A. (Oxon.), M.R.C.S., 
 F.R.S.E., Curator of the Museums, Owens College, Manchester. 
 Natural History of Parasites in General. — Systematic Account 
 of the Parasites Infesting Man. — Protozoa. — Cestoda. Large 
 Svo, cloth, pp. xxviii., 772. Illustrated with 404 engravings. Price 
 31s. 6d. 
 
 Liddell, John, M.D., 
 
 Honorary Physician, Harrogate Royal Bath Hospital. 
 
 THE MINERAL WATERS OF HARROGATE. Cr. Svo, limp 
 cloth, pp. 64. Price 2s. nett. 
 
 Lockwood, Charles Barrett, F.R.C.S., 
 
 Hunterian Professor, Royal College of Surgeons of England ; Assistant Surgeon to St. 
 Bartholomew's Hospital ; Surgeon to the Great Northern Central Hospital, London. 
 
 TRAUMATIC INFECTION. Hunterian Lectures delivered at the 
 Royal College of Surgeons of England. Crown Svo, cloth, pp. xii., 
 1 38. Illustrated with 27 wood engravings in the text. Price 3s. 
 
 ASEPTIC SURGERY. Crown Svo, cloth, pp. xvi., 233. Price, 4s. 
 
ISSUED BY YOUNG J. PENTLAND. 17 
 
 Longley, Elias. 
 
 STUDENTS' POCKET MEDICAL LEXICON. Giving the correct 
 Pronunciation and Definition of all Words and Terms in general 
 use in Medicine and the Collateral Sciences. New Edition, iSmo, 
 cloth, pp. 303. Price 4s. 
 
 McBride, P., M.D., F.R.C.P.Ed., 
 
 Lecturer on the Diseases of the Ear and Throat, Edinburgh School of Medicine ; 
 Aural Surgeon and Laryngologist, Royal Infirmary, Edinburgh ; Surgeon, Edin- 
 burgh Ear and Throat Dispensary. 
 
 DISEASES OF THE THROAT, NOSE, AND EAR. Second 
 Edition, revised and enlarged, Svo, cloth, pp. xvi., 682, with 
 coloured illustrations from original drawings. Price 25s. (Pent- 
 land 7 s Medical Series, Volume Third.) 
 
 McClellan, George, M.D., 
 
 Lecturer on Descriptive and Regional Anatomy at the Pennsylvania School of 
 Anatomy ; Professor of Anatomy at the Pennsylvania Academy of the Fine Arts ; 
 Member of the Academy of Natural Sciences, College of Physicians, &c, of 
 Philadelphia. 
 
 REGIONAL ANATOMY IN ITS RELATION TO MEDICINE 
 AND SURGERY. In 2 handsome volumes, large 4to of over 350 
 pages each. Illustrated with 100 full-page facsimile chromo-litho- 
 graphic plates, reproduced from photographs taken by the Author 
 of his own Dissections, expressly designed and prepared for this 
 Work, and coloured by him after Nature. Price per volume, 42s. 
 nett. Carriage paid. Sold only by subscription. 
 
 Maclaren, P. H., M.D., F.R.C.S.Ed., 
 
 Surgeon, Edinburgh Royal Infirmary; formerly Surgeon in charge of the Lock 
 Wards, Edinburgh Royal Infirmary ; Examiner in the Royal College of Surgeons, 
 Edinburgh. 
 
 ATLAS OF VENEREAL DISEASES. A Series of Illustrations 
 from Original Paintings, with Descriptions of the Varied Lesions, 
 their Differential Diagnosis and Treatment. In 10 fasciculi, price 
 6s. each ; or complete in one handsome royal quarto volume, extra 
 cloth, price 63s. nett. 
 
 Maguire, Robert, M.D., F.R.C.P., 
 
 Physician to, and Joint Lecturer on Pathology, St. Mary's Hospital ; Assistant 
 Physician, Consumption Hospital, Brampton, London. 
 
 DISEASES OF THE KIDNEYS. Svo, pp. xvi., 500, with illus- 
 trations in the text. In Press. (Pentland's Medical Series, 
 Volume Seventh.) 
 
1 8 CATALOGUE OF MEDICAL PUBLICATIONS 
 
 Martin, Sidney, M.D., F.R.S., F.R.C.P., 
 
 Assistant Physician and Assistant Professor of Clinical Medicine at University College 
 Hospital ; Assistant Physician to the Hospital for Consumption and Diseases of the 
 Chest, Brompton. 
 
 FUNCTIONAL AND ORGANIC DISEASES OF THE 
 STOMACH. 8vo, cloth, pp. xx., 506. Illustrated with numerous 
 engravings throughout the text. Price 16s. (Fentlands Medical 
 Series, Volume Fifth.) 
 
 Mills, Charles K., M.D., 
 
 Professor of Diseases of the Mind and Nervous System in the Philadelphia Polyclini 
 and College for Graduates in Medicine ; Lecturer on Mental Diseases in the 
 University of Pennsylvania. 
 
 THE NURSING AND CARE OF THE NERVOUS AND 
 THE INSANE. Crown 8vo, cloth, pp. 147. Price 4s. 6d. 
 
 Morrow, Prince A., M.D., 
 
 Clinical Professor of Genito-Urinary Diseases ; formerly Lecturer on Dermatology in 
 the University of the City of New York; Surgeon to Charity Hospital, New York. 
 
 SYSTEM OF GENITO-URINARY DISEASES, SYPHILOLOGY, 
 AND DERMATOLOGY. By Various Authors. Edited by 
 Prince A. Morrow, M.D. In 6 divisional volumes of about 550 
 pages each. Royal Svo, cloth extra, gilt tops, illustrated with 
 coloured plates and engravings throughout the text. Price per 
 volume, 14s. nett. Sold only by subscription. 
 
 Muir, Robert, M.A., M.D., F.R.C.P.Ed., 
 
 Lecturer on Pathological Bacteriology, and Senior Assistant to the Professor of 
 Pathology, University of Edinburgh ; Pathologist, Edinburgh Royal Infirmary ; 
 
 and 
 
 Ritchie, James, M.A., M.D., B.Sc, 
 
 Lecturer in Pathology, University of Oxford. 
 
 MANUAL OF BACTERIOLOGY. Crown Svo, cloth, pp. xvi., 519, 
 with 108 illustrations. Price 12s. 6d. (Fen Hand's Students' 
 Manuals.) 
 
 Murray, R. Milne, M.B., F.R.C.P.Ed., 
 
 Assistant Physician Royal Maternity and Simpson Memorial Hospital ; Lecturer on 
 Midwifery and Diseases of Women, School of Medicine, Edinburgh. 
 
 MANUAL OF MIDWIFERY. Crown Svo, with numerous illus- 
 trations, pp. 500, or thereby. Preparing. 
 
ISSUED BY YOUNG J. PENTLAND. 19 
 
 Muskett, Philip E., L.R.C.P. & S.Ed., 
 
 Late Surgeon to the Sydney Hospital ; formerly Senior Resident Medical Officer, 
 Sydney Hospital. 
 
 PRESCRIBING AND TREATMENT IN THE DISEASES OF 
 INFANTS AND CHILDREN. Third Edition, revised and 
 enlarged, i8mo, pp. xx., 336, in flexible leather binding for the 
 pocket. Price 6s. 6d. 
 
 Musser, John H., M.D., 
 
 Assistant Professor of Clinical Medicine in the University of Pennsylvania; Physician 
 to the Philadelphia and the Presbyterian Hospitals, &c. 
 
 PRACTICAL TREATISE ON MEDICAL DIAGNOSIS. For 
 
 Students and Physicians. Royal 8vo, cloth, pp. viii., 882. Illus- 
 trated with 162 engravings and 2 coloured plates. Price 24s. 
 
 Newman, David, M.D., 
 
 Laryngologist to the Glasgow Royal Infirmary ; Assistant Surgeon to the Western 
 Infirmary ; Examiner in Pathology in the University of Glasgow. 
 
 MALIGNANT DISEASE OF THE THROAT AND NOSE. 
 8vo, cloth, pp. xvi.j 212, with 3 illustrations. Price 8s. 6d. 
 
 Norris, W. F., A.M., M.D., 
 
 Professor of Ophthalmology, University of Pennsylvania ; 
 and 
 
 Oliver, C. A., A.M., M.D., 
 
 Surgeon to the Wills Eye Hospital, Philadelphia. 
 
 TEXT-BOOK OF OPHTHALMOLOGY. Royal 8vo, cloth, 
 pp. viii., 622. Illustrated with 5 coloured plates and 357 woodcuts. 
 Price 25s. 
 
 Oliver, Thomas, M.D., F.R.C.P., 
 
 Physician, Royal Infirmary, Newcastle-on-Tyne ; Professor of Physiology, University 
 of Durham ; Honorary Physician, Newcastle-on-Tyne Dispensary and Industrial 
 Schools. 
 
 LEAD POISONING, IN ITS ACUTE AND CHRONIC 
 FORMS. 8vo, cloth, pp. xii., 122, with 32 illustrations, mostly in 
 colours. Price 10s. 6d. 
 
 Osier, William, M.D., F.R.C.P., 
 
 Professor of Medicine in the Johns Hopkins University, and Physician-in-Chief to the 
 Johns Hopkins Hospital, Baltimore. 
 
 THE PRINCIPLES AND PRACTICE OF MEDICINE. 
 Second Edition, thoroughly revised and largely re-written. 8vo, 
 cloth, pp. xvi., 1 143, with charts and illustrations. Price 24s. 
 
20 CATALOGUE OF MEDICAL PUBLICATIONS 
 
 LECTURES ON ANGINA PECTORIS AND ALLIED 
 STATES. 8vo, cloth, pp. viii., 160. Price 6s. nett. 
 
 Park, Roswell, A.M., M.D., 
 
 Professor of the Principles and Practice of Surgery and of Clinical Surgery in the 
 Medical Department of the University of Buffalo, Buffalo, New York ; Surgeon to 
 the Buffalo General Hospital, etc. 
 
 A TREATISE ON SURGERY, by American Authors, for Students 
 and Practitioners of Surgery and Medicine. Edited by Roswell 
 Park, A.M., M.D. 2 volumes, large Svo, cloth, pp. 1603. 
 Illustrated with 807 engravings and 38 full-page plates in colours 
 and monochrome. Price 34s. nett. 
 
 Parvin, Theophilus, M.D., LL.D., 
 
 Professor of Obstetrics and Diseases of Women and Children in Jefferson Medical 
 College, Philadelphia, and one of the Obstetricians to the Philadelphia Hospital. 
 
 THE SCIENCE AND ART OF OBSTETRICS. Third 
 
 Edition, thoroughly revised, large Svo, cloth, pp. 701, with 269 
 wood engravings and 2 coloured plates. Price 18s. 
 
 Pekelharing, C. A., 
 
 Professor in the Faculty of Medicine, University of Utrecht ; 
 and 
 
 Winkler, C, 
 
 Lecturer in the University of Utrecht. 
 
 BERI-BERI : Researches concerning its Nature and Cause and the 
 Means of its Arrest. Translated by James Cantlie, M.A., M.B., 
 F.R.C.S. Svo, cloth, pp. xvi., 160, with coloured illustrations 
 from original drawings. Price 10s. 6d. nett. 
 
 Potter, Samuel O. L., M.D., M.R.C.P. (Lond.), 
 
 Cooper Medical College, San Francisco. 
 
 COMPEND OF HUMAN ANATOMY, Including the Anatomy 
 of the Viscera. Fifth Edition, revised and enlarged, crown Svo, 
 cloth, pp. 289, with 117 engravings and 16 full-page plates. 
 Price 5s. 
 
 Rentoul, Robert R., M.D., M.R.C.S., 
 
 Fellow of the Obstetrical Society, London. 
 
 THE CAUSES AND TREATMENT OF ABORTION, with an 
 Introduction by Lawson Tait, F.R.C.S. Svo, cloth, pp. xvi., 271, 
 with coloured plates and 35 engravings. Price 10s. 6d. 
 
ISSUED BY YOUNG J. PENTLAND. 
 
 Reports from the Laboratory of the Royal College 
 of Physicians, Edinburgh. 
 
 Edited by J. Batty Tuke, M.D., G. Sims Woodhead, M.D., 
 and D. Noel Paton, M.D. 
 
 Volume First, 8vo, cloth, pp. 212, with 23 full-page plates and 19 
 engravings. Price 7s. 6d. nett. 
 
 Volume Second, 8vo, cloth, pp. xvi., 280, with 43 full-page plates, 
 consisting of lithographs, chromo-lithographs, and micro-photo- 
 graphs. Price 10s. 6d. nett. 
 
 Volume Third, 8vo, cloth, pp. xii., 304, with 11 plates and folding 
 charts. Price 9s. nett. 
 
 Volume Fourth, 8vo, cloth, pp. xii., 254, with 25 plates and 
 folding charts. Price 10s. 6d. nett. 
 
 Rotch, Thomas Morgan, M.D., 
 
 Professor of Diseases in Children, Harvard University. 
 
 PEDIATRICS. The Hygienic and Medical Treatment of Diseases 
 in Children. For Students and Practitioners. 2 vols, royal 8vo, 
 pp. xvi., 1124. Iillustrated with 8 full-page coloured plates and 
 about 400 engravings in the text. Price 25s. nett. 
 
 Schafer, E. A., F.R.5., 
 
 Jodrell Professor of Physiology, University College, London. 
 
 TEXT -BOOK OF PHYSIOLOGY. By British Physiologists. 
 Edited by E. A. Schafer, F.R.S., Jodrell Professor of Physiology, 
 University College, London. In Press. 2 vols, royal 8vo, pp. 
 xxiv., 950 each or thereby, with numerous illustrations in the text. 
 
 * 4t * The following among others will contribute : — Professor 
 Burdon Sanderson (Oxford), Professor Gamgee, Dr. Gaskell 
 (Cambridge), Professor Gotch (Oxford), Professor Sherrington 
 (Liverpool), Professor M'Kendrick (Glasgow), Professor Halli- 
 burton (King's College), Professor Haycraft (Cardiff), Dr. 
 Pembrey (Charing Cross), Dr. Starling (Guy's), Dr. Rivers 
 (Cambridge), J. N. Langley (Cambridge), Professor Waymouth 
 Reid (Dundee), B. Moore (University College), F. G. Hopkins 
 (Guy's), Leonard Hill (London Hospital), W. H. R. Rivers 
 (Cambridge), the Editor, etc. etc. 
 
 Schech, Philip, M.D., 
 
 Lecturer in the University of Munich. 
 
 DISEASES OF THE MOUTH, THROAT, AND NOSE. 
 Including Rhinoscopy and Methods of Local Treatment. 
 Translated by R. H.Blaikie, M.D., F.R.S.E., formerly Surgeon, 
 Edinburgh Ear and Throat Dispensary ; late Clinical Assistant, Ear 
 and Throat Department, Royal Infirmary, Edinburgh. 8vo, cloth, 
 pp. xii., 302, with 5 wood engravings. Price 9s. 
 
22 CATALOGUE OF MEDICAL PUBLICATIONS 
 
 Schmiedeberg, Dr. Oswald, 
 
 Professor of Pharmacology, and Director of the Pharmacological Institute, University 
 of Strasburg. 
 
 ELEMENTS OF PHARMACOLOGY. Translated under the 
 Author's supervision, by Thomas Dixson, M.B., Lecturer on 
 Materia Medica in the University of Sydney, N.S.W. 8vo, cloth, 
 pp. xii., 223, with 7 illustrations. Price 9s. 
 
 Seifert, Dr. Otto, 
 
 Privat Docent in Wurzburg ; and 
 
 Muller, Dr. Friedrich, 
 
 Assistent der II. Med. Klinik in Berlin. 
 
 MANUAL OF CLINICAL DIAGNOSIS. Third Edition, revised 
 and corrected. Translated with the permission of the Authors 
 by William B. Canfield, A.M., M.D., Chief of Clinic for 
 Throat and Chest, University of Maryland. Crown 8vo, cloth, pp. 
 xii., 173, with 60 illustrations. Price 5s. 
 
 Sheild, A. Marmaduke, M.B. (Cantab.), 
 F.R.C.S., 
 
 Senior Assistant Surgeon, Aural Surgeon and Teacher of Operative Surgery, Charing 
 Cross Hospital. 
 
 SURGICAL ANATOMY: A Manual for Students. Crown 8vo, 
 cloth, pp. xii., 226. Price 6s. (Pentland's Students' Manuals.) 
 
 Skene, Alexander J. C, M.D., 
 
 Professor of Gynecology in the Long Island College Hospital, Brooklyn, New York. 
 
 MEDICAL GYNECOLOGY: A Treatise on the Diseases of 
 Women from the Standpoint of the Physician. 8vo, cloth, pp. 
 vi., 530, with illustrations in the text. Price 21s. 
 
 Smith, W. Ramsay, M.B., B.Sc, 
 
 Formerly Demonstrator of Anatomy, Edinburgh School of Medicine, Minto House ; 
 Late Senior Assistant to the Professor of Natural History, University of Edinburgh. 
 
 EXAMINATION QUESTIONS. Set for the Professional Exam- 
 inations in Edinburgh University during the past ten years, 
 selected from the Calendars. 
 
 NATURAL HISTORY, arranged and annotated. Price Is. nett. 
 
 BOTANY, arranged and annotated. Price Is. 6d. nett. 
 
 CHEMISTRY, answered and annotated. Price 2s. nett. 
 
 ANATOMY, answered and annotated. Price 2s. nett. 
 
 MATERIA MEDICA AND THERAPEUTICS, answered and 
 annotated. Price 2s. nett. 
 
 PHYSIOLOGY, answered and annotated. Price 2s. nett. 
 
ISSUED BY YOUNG J. PENTLAND. 23 
 
 MIDWIFERY AND GYNAECOLOGY, answered and annotated. 
 Price Is. 6d. nett. 
 
 PRACTICE OF PHYSIC, answered and annotated. Price Is. 6d. 
 nett. 
 
 SURGERY, answered and annotated. Price Is. 6d. nett. 
 
 MEDICAL JURISPRUDENCE AND PUBLIC HEALTH, 
 answered and annotated. Price 2s. nett. 
 
 *#* Other Volumes to follow. 
 
 Smith, W. Ramsay, M.B., B.Sc, 
 
 and 
 
 Norwell, J. Stewart, B.Sc. 
 
 ILLUSTRATIONS OF ZOOLOGY, INVERTEBRATES AND 
 
 VERTEBRATES. New and cheaper Edition, crown 4to, extra 
 cloth, gilt top, with 70 plates exhibiting over 400 figures. Price 
 7s. 6d. 
 
 Snell, Simeon, F.R.C.S.Ed., 
 
 Ophthalmic Surgeon, Sheffield General Infirmary. 
 
 EXAMINATION OF THE EYE. Crown 8vo, pp. xvi., 172, with 
 88 illustrations. Price 5s. 
 
 Starr, Louis, M.D., 
 
 Late Clinical Professor of Diseases of Children in the Hospital of the University of 
 Pennsylvania ; Physician to the Children's Hospital, Philadelphia. 
 
 DISEASES OF THE DIGESTIVE ORGANS IN INFANTS 
 AND CHILDREN. With Chapters on the Investigation of 
 Disease and on the General Management of Children. Second 
 Edition, post 8vo, cloth, pp. 396, with illustrations. Price 10s. 
 
 Stephenson, Sydney, M.B., F.R.C.S.Ed., 
 
 Surgeon to the Ophthalmic School, Hanwell, London. 
 
 EPIDEMIC OPHTHALMIA : Its Symptoms, Diagnosis, and 
 Management. 8vo, cloth, pp. xvi., 278. Illustrated with a coloured 
 plate and numerous figures in the text. Price 9s. nett. 
 
 System of Gynecology and Obstetrics. 
 
 By American Authors. Edited by Matthew D. Mann, A.M., 
 M.D., Professor of Obstetrics and Gynecology in the Medical 
 Department of the University of Buffalo, N.Y. ; and Barton 
 Cooke Hirst, M.D., Associate Professor of Obstetrics in the 
 University of Pennsylvania ; Obstetrician to the Philadelphia and 
 
24 CATALOGUE OF MEDICAL PUBLICATIONS 
 
 Maternity Hospitals ; Gynecologist to the Orthopaedic Hospital. 
 In 8 very handsome volumes, royal 8vo, cloth, of about 400 pages 
 each. Fully illustrated with engravings and coloured plates. Price 
 12s. 6d. each nett. For Sale by Subscription only. 
 
 Talamon, Ch., 
 
 Physician to the Tenon Hospital, Paris. 
 
 APPENDICITIS AND PERITYPHLITIS. Translated from the 
 French by Richard J. A. Berry, M.B., CM., Late President of 
 the Royal Medical Society, Edinburgh. Crown Svo, cloth, pp. 
 viii., 240. Price 6s. 
 
 Taylor, Robert W., M.D., 
 
 Clinical Professor of Venereal Diseases at the College of Physicians and Surgeons 
 (Columbia College), New York; Surgeon to Bellevue Hospital, and Consulting 
 Surgeon to City (Charity) Hospital, New York. 
 
 THE PATHOLOGY AND TREATMENT OF VENEREAL 
 DISEASES. Royal Svo, cloth, pp. 1002. Illustrated with 230 
 engravings in the text and 7 coloured plates. Price 22s. nett. 
 
 Text=Book on Nervous Diseases. By American 
 Authors. 
 
 See Dercum. 
 
 Text=Book of Physiology. By British Phy= 
 siologists. 
 
 See Schafer. 
 
 Text = Book of Medicine. By British Teachers. 
 
 See Gibson. 
 
 Thomson, J. Arthur, M.A., 
 
 Lecturer on Zoology, School of Medicine, Edinburgh. 
 
 OUTLINES OF ZOOLOGY. Second Edition, revised and enlarged, 
 crown Svo, cloth, pp. xx., 820, with 266 illustrations in the text. 
 Price 15s. (Pent land's Students' Manuals.) 
 
ISSUED BY YOUNG J. PENTLAND. 25 
 
 Vierordt, Oswald, M.D., 
 
 Professor of Medicine at the University of Heidelberg. 
 
 CLINICAL TEXT-BOOK OF MEDICAL DIAGNOSIS, for 
 Physicians and Students. Based on the most Recent Methods of 
 Examination. Translated, with additions from the second enlarged 
 German edition, with the Author's permission. By Francis H. 
 Stuart, M.D., Member of the Medical Society of the County of 
 Kings, New York. Large 8vo, cloth, pp. xvi., 700, with 178 fine 
 engravings, many in colour. Price 18s. 
 
 Walley, Thomas, M.R.C.V.S., 
 
 Late Principal of the Edinburgh Royal (Dick's) Veterinary College ; Professor of 
 Veterinary Medicine and Surgery. 
 
 A PRACTICAL GUIDE TO MEAT INSPECTION. Third 
 Edition, thoroughly revised and edited by J. M'Fadyean, Principal 
 and Professor of Comparative Anatomy, Royal Veterinary College, 
 London. Post 8vo, cloth, pp. xvi., 198, with coloured illustrations. 
 Price 10s. 6d. 
 
 Waring, H. J., M.S., B.Sc.Lond., F.R.C.S., 
 
 Demonstrator of Operative Surgery, and Senior Demonstrator of Anatomy, St. 
 Bartholomew's Hospital ; Senior Assistant Surgeon, Metropolitan Hospital; Surgeon 
 to the Belgrave Hospital for Children, London. 
 
 DISEASES OF THE LIVER, GALL BLADDER, AND 
 BILIARY SYSTEM; their Pathology, Diagnosis, and Surgical 
 Treatment. 8vo, cloth, pp. xvi., 385, with 58 engravings. Price 
 12s. 6d. nett. 
 
 MANUAL OF OPERATIVE SURGERY. In Press. Crown 8vo, 
 pp. 650 or thereby, with 400 illustrations in the text. 
 
 Warren, Geo. W., D.D.S., 
 
 Clinical Chief, Pennsylvannia College of Dental Surgery. 
 
 COMPEND OF DENTAL PATHOLOGY AND DENTAL 
 MEDICINE. Crown 8vo, cloth, pp. 109, illustrations. Price 
 4s. 6d. 
 
 Webster, J. Clarence, B.A., M.D., F.R.C.P.Ed., 
 
 Demonstrator of Gynaecology, M'Gill University; Assistant Gynaecologist, Royal 
 Victoria Hospital, Montreal; Late Assistant to the Professor of Midwifery and 
 Diseases of Women and Children in the University of Edinburgh. 
 
 ECTOPIC PREGNANCY : its Etiology, Classification, Embryology, 
 Diagnosis, and Treatment. 8vo, cloth, pp. xvi., 240, with 22 pages 
 of plates and figures throughout the text. Price 12s. 6d. nett. 
 
26 CATALOGUE OF MEDICAL PUBLICATIONS 
 
 RESEARCHES IN FEMALE PELVIC ANATOMY. Demy 4 to, 
 cloth, illustrated with 26 full-page coloured plates, from original 
 drawings. Price 30s. nett. 
 
 TUBO-PERITONEAL ECTOPIC GESTATION. Demy 4 to, 
 cloth, illustrated with n full-page plates, exhibiting numerous 
 figures. Price 16s. nett. 
 
 PRACTICAL AND OPERATIVE GYNECOLOGY. Crown 8vo, 
 cloth, pp. xvi., 295. Illustrated with 54 engravings. Price 7s. 6d. 
 
 MANUAL OF DISEASES OF WOMEN. In Press. Crown 8vo, 
 pp. 650 or thereby, with 300 illustrations. 
 
 Williamson, Richard T., M.D., M.R.C.P., 
 
 Medical Registrar, Manchester Royal Infirmary ; Assistant to Professor of Medicine, 
 Owens College, Manchester. 
 
 DIABETES MELLITUS, its Symptoms, Pathology, and Treatment. 
 8vo, pp. xvi., 400 or thereby. In Press. 
 
 Winckel, Dr. F., 
 
 Professor of Gynaecology and Director of the Royal Hospital for Women ; Member 
 of the Supreme Medical Council and of the Faculty of Medicine in the University 
 of Munich. 
 
 TEXT-BOOK OF OBSTETRICS, including the Pathology and 
 Therapeutics of the Puerperal State. Designed for Practitioners 
 and Students of Medicine. Translated from the German under 
 the supervision of J. Clifton Edgar, A.M., M.D., Adjunct Pro- 
 fessor of Obstetrics in the Medical Department of the University of 
 the City of New York. Royal 8vo, cloth, pp. 927. Illustrated with 
 190 engravings, mostly original. Price 28s. 
 
 Woodhead, Q. Sims, M.D., F.R.C.P.Ed., 
 
 Director of the Laboratories of the Royal College of Physicians (London) and 
 Surgeons (England). 
 
 PRACTICAL PATHOLOGY : A Manual for Students and Practi- 
 tioners. Third enlarged and thoroughly revised Edition, 8vo, cloth, 
 pp. xxiv., 652, with 195 coloured illustrations, mostly from original 
 drawings. Price 25s. 
 
 Woodhead, G. Sims, M.D., F.R.C.P.Ed., 
 
 and 
 
 Hare, Arthur W., M.B., CM. 
 
 PATHOLOGICAL MYCOLOGY : An Enquiry into the Etiology 
 of Infective Diseases. Section 1. — Methods. 8vo, cloth, pp. xii., 
 174, with 60 illustrations, mostly original (34 in colours). Price 
 8s. 6d. 
 
ISSUED BY YOUNG J. PENTLAND. 
 
 27 
 
 CLASSIFIED INDEX. 
 
 Abdominal Surgery, 
 Abortion (Causes and Treatment), 
 Abscesses (Treatment), 
 Ambulance Lectures, 
 Anatomy (Practical), 
 
 „ (Medical), 
 (Regional), 
 
 „ (Compend of) 
 
 ,; (Surgical), 
 
 ,, (Examination Questions), 
 
 ,, (Female Pelvic 
 Angina Pectoris, . 
 Appendicitis, 
 Aseptic Surgery, . 
 Atlases (Brain), 
 
 (Eye), . 
 (Skin), . 
 
 „ (Venereal Diseases) 
 
 Bacteriology, 
 
 „ and Pathology, Journal of. 
 
 Beri-Beri, . 
 
 Bones and Joints (Disease of), 
 Botany (Text-Book), 
 
 „ (Manual), 
 
 „ (Examination Questions), 
 Brain (Tumours), . 
 
 „ (Illustrations of Nerve Tracts), 
 
 Chemistry (Examination Questions), 
 
 Children (Diseases), 
 
 „ (Compend of Diseases), . 
 ,, (Cyclopaedia, Diseases of), 
 ,, (Treatment of Diseases), 
 
 Keith. 
 
 Rentoul. 
 
 Cheyne. 
 
 Lawless. 
 
 Cunningham. 
 
 Kenwood. 
 
 M'Clellan. 
 
 Potter. 
 
 Sheild. 
 
 Smith. 
 
 Webster. 
 
 Osler. . 
 
 Talamon. 
 
 Lockwood. 
 
 Bruce. 
 
 Frost. 
 
 Crocker. 
 
 Maclaren. 
 
 Muir & Ritchie. 
 Woodhead & Hare. 
 
 Pekelharing. 
 
 Cheyne. 
 
 Behrens. 
 
 Johnstone. 
 
 Smith. 
 
 Bramwell. 
 
 Bruce. 
 
 Smith. 
 Carmichael. 
 Hatfield. 
 Keating. 
 
 Muskett. 
 
2S 
 
 CATALOGUE OF MEDICAL PUBLICATLONS 
 
 Children (Diseases), 
 
 ,, (Digestive Organs in). 
 Clinical Diagnosis, 
 
 Gynaecology, 
 
 Medicine (Studies in), 
 
 Cyclopaedia of the Diseases of Children 
 
 Dental Pathology, 
 Dentistry (System of), 
 Dermatology (Atlas), 
 
 Diabetes Mellitus, 
 Diagnosis (Ophthalmoscopic), 
 
 ,, (Physical), 
 
 „ (Medical), 
 
 ,, (Clinical), 
 
 (Medical), 
 Dictionary of Terms (5 languages) 
 
 11 11 
 
 „ „ (Pocket), 
 
 Diet, 
 Digestive Organs in Children, 
 
 Ear (Diseases), 
 
 >) j) • • 
 
 Ectopic Gestation, 
 Edinburgh Hospital Keports. 
 Edinburgh Medical Journal. 
 Epidemic Ophthalmia, 
 Epilepsy, . 
 
 Examination Questions, 
 Eye (Diseases), 
 
 „ (Atlas), 
 
 ,, (Diseases), 
 
 ,, (Examination of), 
 
 ,, (Ophthalmia), 
 
 „ (Toxic Amblyopias), . 
 
 First Aid to Injured, 
 
 Genito-Urinary Diseases, . 
 Geographical Pathology, . 
 
 ROTCH. 
 
 Starr. 
 
 seifert & muller. 
 
 Keating & Coe. 
 
 Bramwell. 
 
 Gibson & Russell. 
 
 Musser. 
 
 Vierordt. 
 
 Keating. 
 
 Warren. 
 
 Litch. 
 
 Crocker. 
 
 Jamieson. 
 
 Morrow. 
 
 Williamson. 
 
 Berry. 
 
 Gibson & Russell. 
 
 Musser. 
 
 Seifert & Mui.ler. 
 
 Vierordt. 
 
 Billings. 
 
 Keating. 
 
 Longlev. 
 
 Bruen. 
 
 Starr. 
 
 Burnett. 
 M'Bride. 
 Webster. 
 
 Stephenson. 
 Alexander. 
 Smith. 
 Berry. 
 Frost. 
 
 Norris & Oliver. 
 Snell. 
 Stephenson. 
 De Schweinitz. 
 
 Lawless. 
 
 Morrow. 
 Davidson. 
 
ISSUED BY YOUNG J. PENTLAND. 
 
 29 
 
 Gestation (Ectopic), 
 Gynaecology (Text-Book), . 
 
 (Medical), 
 
 (Examination Questions), 
 
 (Practical), 
 
 (Manual), 
 
 and Obstetrics (System of), 
 
 Harrogate Waters, 
 Heart (Diseases), . 
 
 11 11 • • • 
 
 ,, (Angina Pectoris), . 
 Hospital Reports (Edinburgh). 
 Human Monstrosities, 
 Hydatid Disease, . 
 Hygiene and Diseases of Warm Climates, 
 
 Infective Diseases, 
 Intracranial Tumours, 
 
 Webster. 
 Keating & Coe. 
 Skene. 
 Smith. 
 Webster. 
 
 >> 
 Mann & Hirst. 
 
 Liddell. 
 br am well. 
 Gibson. 
 Osler. 
 
 Hirst & Piersol. 
 
 Graham. 
 
 Davidson. 
 
 Lockwood. 
 Bramwell. 
 
 Joints (Tuberculous Disease), . . Cheyne. 
 
 Journal of Pathology. 
 
 ,, (Edinburgh Medical). 
 Jurisprudence, Medical (Examination Questions), Smith. 
 
 Kidney (Diseases of), 
 
 Maguire. 
 
 Laboratory Reports (R. C. P., Edinburgh). 
 Lead Poisoning, . 
 Liver (Diseases of), 
 
 Malignant Disease of Throat and Nose, . 
 Materia Medica (Examination Questions), . 
 Meat Inspection, .... 
 Medical Anatomy, 
 Diagnosis, 
 
 Dictionary (5 languages), 
 
 Oliver. 
 
 Waring. 
 
 Newman. 
 
 Smith. 
 
 Walley. 
 
 Kenwood. 
 
 Musser. 
 
 VlERORDT. 
 
 Billings. 
 Keating. 
 
 Longley. 
 Skene. 
 
 „ (Pocket), 
 
 Gynaecology, 
 
 ,, Jurisprudence (Examination Questions), Smith. 
 Medicine (Text-Book of), . . . Gibson. 
 
 ,, (Compend of), . . . Hughes. 
 
 ,, (Text-Book of), . . . . Osler. 
 
3° 
 
 CATALOGUE OF MEDICAL PUBLICATIONS 
 
 Medicine (Examination Questions) 
 Midwifery (Treatise on), . 
 
 ,, (Compend of), . 
 
 „ (Manual of), 
 
 ,, (Text-Book), 
 
 ,, (Examination Questions), 
 
 ,, (System of), 
 
 „ (Text-Book of), 
 
 Mineral Waters (Harrogate), 
 Monstrosities (Human), . 
 Morbid Anatomy, 
 Mouth, Throat and Nose (Diseases), 
 
 Smith. 
 
 Davis. 
 
 Landis. 
 
 Murray. 
 
 Parvin. 
 
 Smith. 
 
 Mann & Hirst. 
 
 Winckel. 
 
 LlDDELL. 
 
 Hirst & Piersol. 
 
 Hall. 
 
 Schech. 
 
 Natural History (Examination Questions! 
 ., ,, (Illustrations), 
 
 ,, „ (Text-Book), 
 
 Nervous Diseases (Epilepsy), 
 „ ,, (Text-Book), 
 
 (Nursing), 
 Nose (Diseases), . 
 „ (Diseases), . 
 Nose and Throat (Malignant Disease of), 
 Nursing (Obstetric), 
 (First Aid), 
 ,, (Insane), . 
 
 Smith. 
 
 Smith & Norwell. 
 Thomson. 
 Alexander. 
 Dercum. 
 Mills. 
 McBride. 
 Schech. 
 Newman. 
 Haultain & Ferguson. 
 Lawless. 
 Mills. 
 
 Obstetric Nursing, . 
 Obstetrics (Text-Book of), 
 
 ,, (Compend of), . 
 
 „ (Manual of), 
 
 ,, (Science and Art of), 
 
 ,, (System of), 
 
 (Text-Book of), 
 Ophthalmia (Epidemic), . 
 Ophthalmology (Text-Book of), 
 Ophthalmoscopic Diagnosis, 
 Ophthalmoscopy (Atlas of), 
 Parasites of Man, . 
 Pathological Mycology, 
 Pathology (Geographical), 
 
 ,, (Practical), 
 
 „ (Compend of), 
 
 „ (Dental), 
 
 ,, (Practical), 
 
 Haultain & Ferguson. 
 Davis. 
 Landis. 
 Murray. 
 Parvin. 
 
 Mann & Hirst. 
 Winckel. 
 Stephenson. 
 Norris & Oliver. 
 Berry. 
 Frost. 
 Leuckart. 
 Woodhead & Hare. 
 Davidson. 
 Gibbes. 
 Hall. 
 Warren. 
 Woodhead. 
 
ISSUED BY YOUNG J. PENTLAND. 
 
 3i 
 
 Pathology and Bacteriology (Journal of). 
 Pediatrics, 
 
 Pharmacology, 
 
 Phthisis (Pulmonary), 
 
 Physical Diagnosis, 
 
 Physiology (Examination Questions), 
 ,, (Text-Book), 
 
 Poisoning (Lead), . 
 
 Practice of Medicine (Text-Book of), 
 ,, ,, (Compend of), 
 
 ,, ,, (Text-Book of), 
 
 „ ,, (Exam. Questions) 
 
 Pregnancy (Ectopic), 
 
 Prescribing and Treatment of Children's Diseases 
 
 Carmichael. 
 
 Hatfield. 
 
 Keating. 
 
 MUSKETT. 
 ROTCH. 
 
 Starr. 
 
 schmiedeberg. 
 
 James. 
 
 Gibson & Russell. 
 
 Smith. 
 
 Schafer. 
 
 Oliver. 
 
 Gibson. 
 
 Hughes. 
 
 Osler. 
 
 Smith. 
 
 Webster. 
 
 Muskett. 
 
 Regional Anatomy, 
 Reports, Edinburgh Hospital. 
 
 „ R. C. P. Laboratory, Edinburgh 
 
 Septic Diseases, 
 
 Sexual Organs (Diseases), . 
 
 Skin Diseases (Atlas of), . 
 
 Stomach (Diseases), 
 
 „ (Diseases), 
 Suppuration, 
 Surgery (Compend of), 
 
 „ (Abdominal), 
 (Aseptic), 
 
 ,, (Examination Questions), 
 
 ., (Operative), 
 Surgical Anatomy, 
 Sy philology, 
 
 System of Dentistry, 
 
 „ Genito-Urinary Diseases, 
 
 ,, Gynaecology and Obstetrics, 
 
 ,, Therapeutics, . 
 
 McClellan. 
 
 Cheyne. 
 
 Fuller. 
 
 Crocker. 
 
 Jamieson. 
 
 Morrow. 
 
 Ewald. 
 
 Martin. 
 
 Cheyne. 
 
 Horwitz. 
 
 Keith. 
 
 Lockwood. 
 
 Smith. 
 
 Waring. 
 
 Sheild. 
 
 Morrow. 
 
 Taylor. 
 
 Litch. 
 
 Morrow. 
 
 Mann & Hirsi 
 
 Hare. 
 
YOUNG J. PENTLANUS MEDICAL PUBLICATIONS. 
 
 Therapeutics (Synopsis), . 
 
 ,, (Principles of Treatment), 
 
 (System), 
 ,, (Physical and Natural), 
 
 „ (Examination Questions), 
 
 Throat (Diseases), 
 
 „ and Nose (Malignant Disease of), 
 ,, Nose and Ear (Diseases), 
 Toxic Amhlyopias, 
 Traumatic Infection, 
 Treatment (Principles of), 
 Tropical Diseases, . 
 
 ;j >) 
 
 Tumours (Intracranial), 
 
 Aitchison. 
 
 Bruce. 
 
 Hare. 
 
 Hayem & Hare. 
 
 Smith. 
 
 Schech. 
 
 Newman 
 
 McBride. 
 
 De Schweinitz. 
 
 lockwood. 
 
 Bruce. 
 
 Davidson. 
 
 Felkin. 
 
 Bramwell. 
 
 Ulcers (Treatment), 
 
 Urine, 
 
 Venereal Diseases, 
 
 Warm Climates (Diseases 
 Women (Diseases of), 
 ,, (Diseases), 
 „ (Diseases of), 
 Wounds (Treatment), 
 
 Cheyne. 
 
 Holland. 
 
 Maclaren. 
 
 Morrow. 
 
 Taylor. 
 
 Davidson. 
 
 Keating & Coe. 
 
 Skene. 
 
 Webster. 
 
 Cheyne. 
 
 Zoology (Examination Questions), 
 ,, (Illustrations of), . 
 ., (Outlines of), 
 
 Smith. 
 
 Smith & Norwell 
 
 Thomson. 
 
COLUMBIA UNIVERSITY LIBRARY 
 
 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 rules of the Library or by special ar- 
 rangement with the Librarian in charge. 
 
 DATE BORROWED 
 
 DATE DUE 
 
 DATE BORROWED 
 
 DATE DUE 
 
 MAY 2 1 1 
 
 94 5 
 
 
 
 AUG i 
 
 1945 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 it 
 
 
 
 
 m 
 
 T 
 
 
 
 5 
 
 3 
 
 
 
 § 
 
 
 
 
 ^^ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 C28(239)MI00 
 
 
 
 
Williamson 
 
 
 RC660 
 
 F672 
 
 1898 
 
 Diabetes 
 
 meliitus 
 
 and its 
 
 
 
 4d 666 
 
 /tig 
 
 ■ ■■