^'^ ^/ XTbe lUniversit^ ot Cbicaao FOUNDED BY JOHN D. ROCKEFELLER DIGESTIBILITY OF STARCH OF DIF- FERENT SORTS AS AFFECTED BY COOKING A DISSERTATION SUBMITTED TO THE FACULTY OF THE GRADUATE SCHOOL OF ARTS AND LITERATURE IN CANDIDACY FOR THE DEGREE OF DOCTOR OF PHILOSOPHY (department of SOCfOLOGV) BY EDNA D. DAY CHICAGO 1908 TLbc 'Clniversit^ ot Cbicaoo FOUNDED BY JOHN D ROCKEFELLER DIGESTIBILITY OF STARCH OF DIF^ FERENT SORTS AS AFFECTED BY COOKING A DISSERTATION SUBMITTED TO IHE FACULTY OF THE GRADUATE SCHOOL OF ARTS AND LITERATURE IN CANDIDACY FOR THE DEGREE OF DOCTOR OF PHILOSOPHY (department of sociology) BY EDNA D. DAY CHICAGO 1908 ^ T lie \J .-Ji vef&ity, F ? '09 -!> LETTER OF TRANSMITTAL. U. S. Department of Agriculture, Office of Experiment Stations, Washington, D. C, July 15, 1908. Sir: I have the honor to transmit herewith and recommend for pubhcation as Bulletin 202 of this Office a report of investigations on the digestibility of starch of different sorts as affected by cooking, conducted by Miss Edna D. Day, at the laboratory of physiological botany of the University of Michigan, and later at the University of Illinois and at the laboratories of physiological chemistry and plant physiology of the University of Chicago. The investigations were planned in such a way as to supplement the investigations on the changes brought about in food products by different cooking processes, which have been carried on as a part of the cooperative nutrition investigations of this Office, and during the work Miss Day was in consultation with the Director and his associates having charge of the details of the nutrition investigations. Of special interest and importance from the housekeeper's stand- point is Miss Day's conclusion that at least in the case of starch itself long continued cooking is not necessary, since it increases digestibility only slightly and out of proportion to the time and trouble involved, though with cornstarch an improvement m flavor was noticed after thirty to forty mmutes' cooking. In solving culi- nary ]iroblems the changes brought about in other constituents usu- ally associat(Ml with starch in foods must, however, be taken into account. Acknowdedgment should be made to Prof. F. C. Newcombe, of the University of Alichigan; to Dr. Charles Hottes, of the University of Illinois, and to Profs. Marion Talbot, A. P. Mathews, and Charles R. Barnes, of the University of Chicago, as well as to Dr. H. McGuigan and Dr. H. Ilasselbring, of the same institution. Respectfully, A. C. True, Director. Hon. James Wilson, Secretary of Agriculture. [Bull. 202) CONTENTS. Page. Introduction 7 Materials, apparatus, and methods used in experimental \v(irk 17 Experiments on tlie effect of cooking starch 22 Experiments with raw starch 22 Digestion experiment No. 1 23 Digestion expeiiment No. 2 24 Digestion experiment No. 3 24 Experiments with starch cooked below the hoiling temperatiu'e 25 Digestion experiment No. 4 26 Digestion experiment No. 5 28 Experiments with starch boiled five minutes 30 Digestion experiments Nos. 6 and 7 30 Experiments with starch boiled more than live minutes 30 Digestion experiment No. 8 31 Digestion experiment No. 9 31 Digestion experiment No. 10 '. 31 Digestion experiment No. 11 32 Digestion experiment No. 12 32 Digestion experiment No. 13 32 Digestion experiment No. 14 33 Experiments ^\dth starcli cooked in a water bath or double boiler 34 Digestion experiment No. 15 34 Experiments on the digestibility of starch by different diastases 34 Digestion experiment No. IG 34 Digestion experiment No. 17 35 Experiments on the digestibility of different preparations of the suiiic kind of starch 35 Digestion experiment No. 18 35 Digestion experiment No. 19 30 Digestion experiment No. 20 30 Additional attempts to identify rose amylose in ])otat() starch 37 Digestion experiment No. 21 37 Digestion experiment No. 22 37 Digestion experiment No. 23 37 The possibility of reversion in wheat and corn standics 38 Digestion experiment No. 24 38 Experiments on the digestibility of tapioca, sago, pearled l)aTley, and rice starch 38 Digestion experiment No. 25 39 Cornstarch v. wheat starch 39 Deductions and conclusions 39 Lliull. 2021 LLUSTRATIONS. Fic. 1. Dia.s^i'am representing Meyer's theory of a four-layered starcli grain . . A potato starch grain of hxrge size magnitied 385 times Raw starch grains magnified 385 times Raw starch grains tliat have been digested 38 hours Potato starch magnified 87 times, showing successive stages in paste fi )rmation (). Starc-h grains fn)ni unboiled pastes (6) tBull. 202] Page. 9 10 22 23 25 2fi THE DIGESTIBILITY OF STARCH AS AFFECTED BY COOKING. INTRODUCTION. The digestion of starch has been studied from various standpoints and by many investigators, but the subject still offers opportunities for investigation. Thus, it has been studied with reference to the characteristics of enzym activity; it has been studied as a means to knowledge of the composition of the starch molecule, and it has been studied from the standpoint of plant and of animal nutrition; but a search through the literature of the subject leads to the con- clusion that it has been very little studied for the purpose of helping solve the housekeeper's problem of how starcln^ foods may best be prepared for every day human digestion. As starch, next to water, is the substance most abundantly present in food, a knowledge of the effect of cooking on its digestibility is desirable, both from a practical and a scientific standpoint. The following comparative study of the relative digestibility of different forms of starch — uncooked, cooked below boiling tcnnperature, and boiled for short and long periods of time — is presented as a contri- bution to this large subject. A brief summary of current theories bearing on the subject may fitly precede a report of these investigations, and it is believed such a summary will be helpful to the student of starch who has found theories varied and conflicting. It will certainly be helpfid in under- standing the results of the experiments here recorded and in judging of the validity of the conclusions reached. The empirical knowledge, gained by generations of experience in cooking and eating starchy foods, first deserves attention. Judging from popular traditions among housekeepers, such experience teaches that (1) raw starch is indigestible or at least relatively so; (2) to be made easy of digestion starch should be cooked at boiling tempera- ture or higher; (3) after the starch has once been brought to the boiling point its cooking may be continued at a somewhat lower temperature, as, for instance, in the double boiler; (4) in general, the longer starch is cooked the better; (5) in cooking starch lumps (7) 5G551— Bull. 202—08 2 should be avoided; (6) cornstarch needs longer cooking than does wheat starch; (7) of the two starchy foods, rice and potatoes, rice is more easily digested than potato, and (S) French cooks believe that ])otato starch has better culinary properties than cornstarch or wheat starch (flour) in thickening sauces and gravies. Next to housekeepers, those most interested in this problem have been ]>hysicians, and statements from text-books on physiological chemistry follow: "Tlie individual grain [of starch] is inclosed in a capsule of so- called starch cellulose which is insoluble in water but which can be made to open by heating in the presence of mucli water. The con- tained starch granulose can thus be ()l)tained."" Another \\Titer'' states that according to ordinary opinion, starch grains consist of two dift'erent substances, starch granulose and starch cellulose, of which the fii'st only goes into solution on treat- ment with diastatic enzyms. Starch is practically insoluble in cold water. Tn warm water tlie grains swell up antl burst, yielding a paste. Boiled starch is quickly, and unboiled starch very slowly, con\'erted into sugai'. A'arious kinds of unboiled starch are con- verted witli (Hllereut degrees of rapidity. This theory of the outer, indigestible layer of the starch grain gives a reason for cooking starch, but it suggests no value in long cooking. It is the old theory of Nageli,<^ publislied in 185S, without e^ (Ml the modification inade by Brown and Tleron'^ in 1879. Brown and Heron taught that starch cellulose formed the outer la\(>r of the starch grain and that, unless it was broken meclianically, as by gi'indiug with glass, or dissolved ])y cytase, as in the plant cell, it |)revented entirely the digestion of raw starch. However, they found that, if the starch was made into a thin paste and treated with ])lenty of malt extract at a temperature above 45° C, the outer layer as well as the inner parts was entirely digested. But the ])otanists tell us that the theory of the outer layer of starch cellulose has been out of date with them for twelve years, and ])oint to Arthur Meyer's'' studies of starch grains as authority for this statement. As a result of his work he offers a theory of starch couiposition which is generally accepted by botanists to-day, which in brief is as follows: Starch grains are sphero-crystals, i. e., they are luore or less regular spheres composed of a mass of radiating needle- shaped crystals, which he calls trichites.^ « C. Simon. A Text-Book of Physiological Chemistry. Philadelphia, 1904, p. 73. '' 0. llammarsten. Lehrbiicli der Physiologischen Chemie. Wiesbaden, 1907, pp. :i2(iand 3-16. <' Die Stark ekorner. Zurich. 1858. 'Mdur. Chem. Soc. [London], 35 (1879), pp. 59(i-654. ' Uiitersuchungen ul)er die Stiirkekomer. Jena, 1895. /1M<1.. i.p. 100-129. [IJull. 202] Figure 1, which is redrawn from Meyer's report,'* represents dia- grammatically his theory of the structure of a starch grain having four concentric layers. Figure 2 shows the microscopic appearance of a normal potato starch grain showing 20 layers, magnified 385 times. A comparison of the two will give an idea of the very small size of tliese tricliites. The layers, he believes, are successively^ deposited one outside of the other, beginning with the hilum or point of origin. He suggests that the difference in appearance of the different layers is due to a differ- ence in the size of the trichites and to a difference in the amount of space between them, as shown in the figure. This, he believes, is due to alternating dift'erences in growth conditions in the plant cell in which the grain was formed, the loose layers being formed by partial solution, at the surface, during the night, of the starch deposited during the previous day.'' The outer layer he regards as no different from the otliers exce])t that in some cases it is moi-e dense, due to the fact that the starch is gathered after the end of the growth i)eriod in the plant, and the last growth is slower tlian the first.'' Some of these t richites, accord- ing to his theory, are composed of ^r-amylose and some of /^'-ani- ylose.'' They are both present in ever}' layer tliougli ])erhaps in different proportions.'^ In some grains part of the trichites are composed of amylodextrin. Meyer's /i-amylose corresponds in properties to tliest arch granulosc of earlier writers. It stains blue with iodin and is easily digestible. It is ins()lul)le in cold water./ In hot water (m° to 80°. C.) ]\Ieyer believes that the trichites absorb water and swell to a spherical form, and that the ordinary starch paste is an emulsion of these tiny sticky droplets.fi' At 138° C. he says that they go into solution in water, producing a clear licpiid. When a starch paste is exposed to dry air, a skin forms on the sur- face. Meyer says that this is due to the sticking together of these viscous swollen /i-amylose trichites, and that the}' can scarcely be separated again by shaking in hot water.* His n-amylose corresponds in a measure to the starch cellulose or amylocellulose of others. He says that it differs from ^-amylose in "Untersuchungf'ii uber , . , . . tiiiR's: per cent from the same sample of ])ot ato starch.' This is one I'cason he gives for beli(>v- ing that <^>'-aniylose is (>asily convertible into /j-amylose. and that the two ai'c essentially one substance. Amylodextrin is a term that has h(MMi used hy dilfei'ent wi'iters to designate dilferent substances. Meyer uses it to name the substance which is a tirst stage in the digestion of stai'ch in the cold with acid, and which stains red with iochn. It is slowly solul)le in cold water and readily in warm watei". It is easily changed hy further digestion to sugar.'' There are some starch gi'ains that slain \vi.\ instead of l)lu(> with iodin. From them Meyer claims to have extracted amylodextrin. vSometimes he luis seen a red color in starch grains which geneially stain blue and he considers 1 he color due to amylodextrin foi'med by some digestive pi'ocess.' (\)ntrary to the experiences of l>row ii and Heron, Meyei' i-ej)orts that unbroken raw starch grains can be (ligesi(>d by ])oth malt extract and saliva. This is (juite in accord with his theory of a ])()rous starch grain (see fig. 1), any |)art of which can be ))enetrated by enzyms without a j)relimin' y breaking of an outer wall. He ex|)lains the very slow digestior of raw starch by saying that since the j)()res are of less than microsco])ic size, (liffusion inward of enzyms and outward of digestion products takes place very slowly. He finds the grains that show cracks are much more (piickly digested raw than are solid grains. But with either the solid or more or less cracked grain the sui'face exposed is very small in comparison with that of an ecjual (|uantity of starch substance in the form of a paste.'^ According to Meyer, then, starch is very slowly digested in the raw form; quickly but not completely if it has been maxh^ into a paste " Untersuchungen iitxT die Starkcknrncr. Jena, 1895. p]>. t3, 14. ''I hill., p. •). '•11 lid., p. •"). [Bull. 202] ''Ibid.. ])]>. 27 'II. id., p. S2. /Ibid., p. -lA. 39. 11 mth water from 60° to 80° C. (140° to 176° F.), and quickly and completely if it has been cooked a long time or heated to lo8° C. There is notliing in liis theory to justif\^ the cook's insistence on the boiling temperature, nor is there warrant for cooking cornstarch longer than wheat starch, since he reports that the latter contains more o'-amylose than the cornstarch. Moreover, arrowroot starch, wliich physicians often recommend for invalids and young children as especially easy of digestion, he finds has more cr-amylose than either corn or wheat starch. In his book, Meyer reviews and criticises in the light of his own extensive observations and experiments the earlier investigations on the subject. Since its publication in 1895 several hundred articles have been published in scientific journals on the subject of starch, but only a few of these need be mentioned here. The theory of the spheroKuystalline form of the starch grain seems to be generally accepted by those who are familiar with it. Kra(Mnor" seems to be the only botanist who has recently insisted that the outer layer of lh(> starch grain is dilferent from the other layers. He suggests that it may be an anhydrid of starcli cellulose. Tn 1S97 BiUschli'' described some artificial starch grains pivcipi- tated from solutions of starch either by eva])oiiition oi- freezing. He said also that the skin that forms on staich ])aste on standing, which Meyer considers due to the sticking together of swollen «'-amylose tricliites, is of the same nature as tlie ])i-e('ipitat(Ml starch, and that they are botii harder of solution and digestion than raw starch. In 1899 Rodewald and Ivattein' reported the formation of artificial starch grains in starch solutions made })v heating starch with iodin atl30°*C. Recentl}' (1903-1 907) four associated scientists, A. Fernbach, E Roux, L. ]\Ia(|uenne, and J. Wollf, have given considerable atten- tion to this reversion of starch. They report that a clear ase])tic solution of starch |)re|)are(l by heating in an autoclave to I'Mf C. passes, on cooling, tlirough a series of retrogressive changes. It first becomes Yiko ordinary starch |)aste, in which form it is easily digested. From tliis form it passes by degrees into a final form, which they call amvlocellulose. Tiie characteristics of this form are tliat it drives no color with iodin and is difficult both of solution in boiling water and of digestion by diastase. Furtliermore, on treating with j)otassium hydroxid or by heating again with water to 180° it dissoh^es, forming "B()(. Guz., 34 (1902). pp. 341-354. ''Verhaudl. Naturhi.st. Mod. Ver. lloidellx-rg, 5 (1897), i)p. 457-472. ''Sitzuugsber. d. Berl. Acad., 33 and 34 (1899), ]). G2S. [Bull. 202] 12 a solution that colors blue with iodin but that does not gelatinize on coolinu: as does starch paste." They also report tliat tlie same reverted starch or amylocellulose is formed if starch pastes are liquefied by the first stages of digestion and then the digestive process checked/ This suggests the descri])tion given by Brown and Heron in 1879 of a preci])itate of starch cellulose from a starch paste digesting with malt extract in the cold. Brown and Heron reported two forms of starch cellulose, one more difficult of solution than the other. '' Maquenne and Roux'' conclude fi'om their work tliat amylose exists in a series of convertible foruis that may be regarded as different con- densation forms of the same fundamental nucleus. Amylocellulose stands at one end of the series and granulose at the other. Cold and concentration of solution tends to change tlie starch to the amylo- cellulose forui, and heat and dilution to the granular form, but the change to the amylocellulose form takes place oidy when tlie granu- lose has first been dissolveth In addition to tlie different forms of amylose tliere is, they believe, another substance ])resent in natural starch grains which they call auiylo|)ectin. It is the colloidal sub- stance which causes starch pastes to gelatinize. It does not stain with iodin and it is closely united with the amylose, and while in this combination prevents reversion. The two are separated by solution, whereupon the amylose reverts, and the reverted form containing no ainylopeclin does not gelatinize aft(>r nOieating with water. The isolation of amylop(M'tin has not yet heen reporteiL (See p. Ki.) This work on reverted starch suggests some new problems in starch cooking. It would seem that the more thoroughly starch is cooked the more (hmger there is of its reverting to its most indigestible, i. e., insoluble form, if it is not consumed at onc(\ It occurred to Roux '' that j)erhaps reversion took ])lace as bread ))ecanie stale, l)ut lie was unable to find any evidence of the fact, altiiougii two years before Lindet,' working apparently without this theory, had reported iinding the starch in stale bread harder to dis- solve than that in fresh bread. These four scientists seem to have overlooked an article by Syn- iewskif on the constitution of starch, which ap))eared in 1.S99, in which the author con(dudes that the amylocellulose of Brown and Heron (<^i'-amylose of Me^^er) does not exist in the natural starch grain, " ('(.inpt. Kend. Acad. Sci. [Paris], 137 (1903), pp. 12()6-12G8. 'M'ompt. Rend. Acad. Sci. [Pari.'^], 138 (1904), pp. 819-821 : 139 ( 1904 i, ])p. 1217-1219. <• Jour. Cheni. Soc. [London], 35 (1879), pp. 596-654. <« Compt. Rend. Acad. Sci. [ParL^], 140 (1905), p]). 1303-1308. f (\nn\A. Rend. Acad. Sci. [Paris], 138 (1904), pp. 1356-13.58. / Rul. Soc. Chim. Paris, 3. scr., 27 (1902), pp. 634-6.39. 9 Liebig's Ann. Chem., 309 (1899), pp. 282-315. [Bull. 202] 13 but is formed as a reversion product when for any cause starch goes into sohition and is not kept at a high temperature or quickly digestecL The reasons he gives for these conchisions are (1) that the methods they employed to isolate this substance were such as would allow of reversion; (2) they obtained varying quantities of it b}^ using different methods, more or less favorable to reversion; (3) they give no satis- factory explanation of the fact that under favorable conditions starch paste is as completely digested as it would be if it contained no indigesti})le material, and (4) he analyzed the substance prepared as directed for «'-amylose and found that it is a hydrated form of starch, instead of starch anhydrid or a condensation product of the same, as suggested by others. Its formula he gives as (C6lIio05),jn+nII.X). Since Syniewski wTote before the theory of amylopectin he natu- rally said nothing about it, but his theory easily explains without its help the fact that the reverted product does not gelatinize after reheating with water. The starch molecule has decomposed beyond the gelatin-forming stage before it reverts. Syniewski gives in this article and in one « that followed it a very interesting theory of the structural formula of tlie starch molecule and of the various changes that come to it on boiling with water and digestion with enzyms. Tiong boiling with water, he says, results in a series of hydration compounds unlike those given by digestion. Because of the place in the molecule where the changes take place, tliese compounds give the characteristic blue color with iodin, and do not reduce Fehling's solution, as do the prochicts of digestion. These compounds he calls, because of their blue color with iodin, amylodextrins (starch-like (h^xtrins). ITe has analyzed two of these. The hrst is the compound formed wh(>n l)y long boiling or by heating in an autoclave the starcli first forms a clear solution, lie gives to it tlie formula (CglliijOj)^,, + ouHoO. It is this compound, lie says, which easily reverts. There- fore he calls the product ''reverted amylodextrin " instead of re- verted starch. Its fornuda is (C,JIi„0-),jn-|-nir,0. The second of these auiylodextrin C()mi)ounds which he has analyzed he says is the final compound formed by very long boiling with water. He calls it amylogen, and gives its fornuda as (Collio0,r;)9 + .311,0 or CjJI,,/)^^- Ih' Ixdieves that the starch molecule is an anhydrid condensation product of four amylogen grou|)s, which the long ])oiling sei);irates into amylogen molecules. Syniewski agrees with other writers that by digestion with either saliva or malt extract the hrst stage is a change to a clear solu- tion, i. e., the ''soluble starch" of many writers and called by Syniewski "first amylodextrin." After this blue-staining com- pound, all agree that by continued digestion a mixture results that contains some maltose and a compound staining red with iodin. a Liebig's Ann. Chem., 324 (1902), pp. 212-268. (Bull. 202) 14 Syniewski calls this maltodextrin I. The hydration, he says, results in breaking one molecule of maltose from each amylogen group in the starch molecule. The next step results in the separation of more maltose, another molecule from each amylogen group, and a com- pound staining brown with iodin, his maltodextrin TI. The term erythrodextrin is often applied to both these red and brown dex- trins. The separation of a third molecule of maltose from each amylogen group, he says, results in a compound giving no color with iodin, formed of the resi(hie of these four groups. It is very slow of further digestion, l^ro vti and Heron called it ''stable dextrin." i\h)st writers now agree that it is fmally converted into maltose and glucose. If Syniewski is right in believing that there is no substance present in raw starch that is not changed to an easily digestible form by ordinary paste formation, there would seem to be no value in boiling starch a so-called "long time," for the more nearly the paste becomes a solution the more danger is there of reversion, if digestion can not follow al once. Moreover, this change to a .soluble form, which it takes hours of boiling to i)roduce, can be accompli.shed by the first minute of normal digestion which is followed at once by further changes tlnit carry the j)r()cess beycmd danger of reversion. Being a chemist :ind not a inicroscopist , Syniew.ski was apparently not interested in the j)hy>ical changes that occur in paste forma- tion — that is, whetiier outer layers burst or trichites swell. He worked with potato starch and has notiiing to say on the relative digestibility of dili'erent kinds of starch. As far back as 1S74 1). Levberg" reporte 1 that diU'erent ((uantities of saliva were necessary to i)ro luce the maxinnim (piantity of glucose in the same time of igestion. He oi)laine 1 from potato starch, with !) cubic centimeters of saliva, 00.8 per cent glucose: from arrowroot starch, with •> cubic centimeters of saliva, .')'.).(')_' p(>r cent glucose; from rice starch, with 10 cubic centimeters of saliva, 50.70 per cent glucose: and from wheat starch, with IC) cubic centimeters of saliva, ()"_'. S7 per cent glucose. By varying the experiment he found that it took S hours to produce the maxinuun result from arrowroot starch, hours from potato starch, 12 hours from wheat starch, and 14 hours from rice starch. In both cases it will be notice. 1 that the potato and arrowroot starches showed greater ease of digestion than the rice and wheat starches. In 1896 vStone'' com})ared the digestibility of diiferent kinds of starch paste when acted upon by various forms of diastase as shown l)y the time it took for the digestive mixtures to reach a stage where they could no longer color iotlin solution. He found that when treated with malt extract, potato starch required 12 to 15 minutes, «Inaug. Diss., St. Tetersburg, 1874. ^Expt. Sta. Record, vol. 8, p. 662. IBuU. 202] 15 wheat starch 60 to 90, and maize (corn) starch 90 to 120 mmiites. When sahva was used potato starch requu"ed 3 mmutes, maize starch 90 minutes, rice starch 165 minutes, and wheat starch 170 minutes. With commercial pancreatin, potato starch required 58 minutes, maize starch 337 minutes, while with both rice and wheat starch the digestion was not complete in 10 hours. As an explanation of the differences wdiich were observed in digestibility Stone suggests that starches, like sugar, possibly exist in different isomeric forms. In 1904 O'Sullivan" reported that "under similar conditions of hydrolysis with diastase * * * tl\e results obtained from potato starch show no (juantitative i-elationship with those obtained from the other starches." lie did not say what these other starches were, but he probably did not use arrowroot starch. U. Butyazin,'' in 1S87, reported that of rice, ])eas, millet, and buckwheat imder the same conditions of cooking and digestion, millet was the most easily digested, then buckwheat, rice, and peas. This author also concluded from his investigations that cooked starch was more digestible than raw starch, and that the digc^stihility increased with iho length of time of cooking. In this connection some Japanese work reported by S. Kano and S. lisliiina'^ is of interest. These investigators were themselves the subjects of tests undertaken to study the digestibility of katakuri starch made from KviitJironium clens-canis and kazu starch made from Pueraria thunhergldna. Aj)})arently tlie starch was cooked and was the sole article of (Het. The coellicient of (hgestibility of the katakuri starch was 95.7 per cent and of the kazu starch 96. G j^er cent, vahies which agree very closely. The digestii)ility of konnyaku, a f(-od product ])repared from the tuber of lli/drosme I'ivierl (ConophaUus konjak), wliicli contains man- nan as its princi{)al carbohydrate instead of starch, was also studied, the coellicient of digestibility of the carbohydrates being 82 per cent, or considerably lower than the starches. Juilging from tlie work of O'Sullivan and others there is a difference in tlie digest il)ility of (Ulfeivnt staiches, but there would seem to be more (HfTerence betwecMi ])otato and ''Other starches" than between coin and wheat starches. From the confusion and ct)ntradictions a few points stand out. Thus, all authorities agree that starch paste is very much more easily digested ihau raw starch. Investigators give no explanation of the cook's insistence on the boiling temperature in cooking starch. In fact, in j)rcparing their pastes foi- experiments they put no special emphasis on reaching the ffjour. Cheni. Soc. [London], 85 (1904), p. GIG. '^ Inaut!;. Diss., St. Potersburo;, 1887. <- Bui. Col. .Vgr., Tokyo Imp. l^niv., 2 (1891), No. 2— U. S. Dept. A.sjr., Oflico of Experiment Stations Bui. 159. 56551--Bull. 202—08 3 16 boiling point or on the length of time that the pastes are cooked. The length of time that a paste stands after it is made, or the tem- perature at which it stands, ov the presence of traces of impurities in the starch are all emphasized as affecting the digestibility, but no one seems to ha^'e comj)ared the digestibility of starch cooked below boiling temperature with that actually boiled. Some inyesti- gators ])ring their starch to a boil, others cook a few moments on the water bath after boiling, and others cook on the water bath appar- ently without any boiling. Their only effort in this direction seems to he to secure a homogeneous paste. Inyestigators as well as cooks use care to ayoid lumps in cooking starch. Many inyestigators use the autoclaye with a temperature of 130° to 140^ C. in pre])aring material for e.\periments; for since the pub- lication of Meyer's i-esults all agree that the entire starch grain goes into solution at that temperature. The exj)ression "long boiling," as used l>y inyestigators, is less dt'linite than the cook's "pinch of salt,'' but they all agree that it has the same effect as heating to 130° to 140° C. in changing the opal- escent ])aste to a clear solution. But since those who hn\o expcriniciUed with reyerted starch agree that starch solutions tend to reyert to forms nu)st didicult of diges- tion, long cooking would seem to be of (|uestionable yalue. There is a general agreement among investigators as to the pi'oj)- erties of the substance that foinis at least the chief mass o{ starch grains, but there is nuich disagrcMunent as to the presence in the stai'ch grain of one or moi'c other substances (such as rainiose, starch cellulose, ^»'-amylose, amylodextrin. and auiylo|)ectin^. and of the respect ivt> pi'o{)eities of each. Theie is an eyident (liscre})ancv between the beliefs of practical cooks and the yarious theories of different scientific investigators. In order to find, if possible, an explanation for this fact and to secure nioic definite knowledge as to the effects of cooking on the digesti- bility of starch, the experiments repoi'ted in tlie following pages were un(l(M-taken. Since tlu> investigations repotted in thi> hulletin were coin])leted and the report was prepari^l for })ul)lication seyeral articles by othei' investigators haye a|)i)e;i red whicli hav(> a bearing on the conchisions di-nwn. The following ])apers seem especially interesting in this connection. Mnie. Gratin-Gruzewska" claims to haye isolated the amylopectin of Miuiuenne and Roux (see [). 12) from })ot:ito starch l)y the use of alkali. She finds by the niicroscoi)e that the substance is composed of the outer layers of the starch grains. It stains red with iodin. "Conipt. Rend. Acad. Sci. [Paris]. 146 (1908), pp. 540-542. [Hull. -'112] 17 In wami water it forms a gelatinous paste wliile the solution from which it is separated does not gelatinize. Tliis so-called amylopectin is the same as the red amylose of this paper. That the blue amylose from which it is separated does not gelatinize is, perhaps, due to the fact that the alkali has hydrolyzed it beyond the gelatinizing stage. England" separates thdn, well-boiled pastes of corn and Inirky starches into three layers by using a centrifugal machine and then allowing them to stand for twelve hours. The first or upper layer, he finds, consists of a clear solution that gives the characteristic blue color with iodin and does not gelatinize on standing. He calls it soluble starch. The second layer consists of a white precipitate which stains a reddish color with a little iodin, a purple with more, and H deep blue if still more iodin is added. But if this precipitate is thoroughly washed with water to free it of soluble starch it gives the red(hsh color alone. On boiling with water it forms a gelatinous paste. lie calls it insoluble starch. The lowest layer, when well washed, is pearly white in ai)pearance and gives no color with iodin. He consiilers that it consists of the broken envelopes of starch grains and calls it cellulose. His insoluble starch probably consists of both rose and red amylose and his cellulose of reverted amylodextrin. Of more i)ractical interest is the work of Harcourt,'' who has recently tried some natural digestion experiments on the effect of cooking rolled oats and wheat farina. He reports that 98 per cent of the'^carbohydrate of the rolled oats was digested if cookc^l 20 minutes, while only 0.3 per cent more (98.3 per cent) was digested if the oats were cooked 8 hours. With the wheat farina 98.3 per cent of the carbohydrate was digested if the wheat was cooked 20 minutes, only 0.2 per cent more (98.2 per cent) if first cooked 8 hours. Thus, with natural as well as artificial digestion and with starch combined with the other constituents of the cereal as well as with isolated c(>real starch, there is found to be a slight, but only a slight, increase in digestibility secured by cooking for a long time. MATERIALS, APPARATUS, AND METHODS USED IN EXPERI- MENTAL WORK. The investigations here reported were carried on at the University of :Michigan and lat(>r at the University of Illinois and at the Univer- sity of Chicago 1903 to 1906. The materials used in these experiments were potato, wheat, corn, arrowroot, and several other forms of starch, sahva, malt diastase, and taka diastase. f The Cereal Modification of Cow's IMilk. [Philadelphia]. ( 1907). f* Ann. Ri)l. Ontario Atrr. Col. and Expt. Farm, 32 (1906), p. (Ki. [Bull. 202] 18 In the case of potato starch a commercial sample and several samples prepared in the laboratory were used. The commercial prejinration was neutral in its reaction and quite free from solid imj)urities, and therefore no attempt was made to purify it. For the laboratory preparations new and old potatoes were bought the lirst of May. The old ones were in very ^ood condition for the time of the year. The new ones were of s of cheese cloth. The iilterinu' removetl tlie cellulose. To rid the starch of the soluble im])iirities, it was washed many times 1)V decantatiou. first in tap water and then in distilled water. It was di'icd on un<2;la/.ed porcelain ))lates below ()()° C. Some of the old jiotatoes were saved and kept in the laboratory locker until the hrst week in July. wluMi they were found to be nuich sprouted and shriveled. Aftej' renioNiuL:; the s|irouts, starch was pr(»])are(l from these potatoes hy the same method. In the case of the cornstarch, and wheat starch also, connnercial sainpl(>s and laboratoiy |)reparations were used. It was found that there were dilVerences in the digestibility of the commercial pi'cpara- tions of cornstarch. Of i\w three sani])les tested and desionated "A," *'B." and "(\" respectively, "A" and "B" W(>re j)ut up for culinary use by dilferent firms and both contained solid inij)urities, lafii'ely cellular material. '"A" was so alkaline that it nnpiii-ed ().S6 cubic centimeters IK'l solution to neutralize 1 ijrani of it, using phenolplithalein as an indicator. "B" was wry sliuhtly acid in its i'(>a('tion. I)otli were washed with tap watei' until neutral, and then with distilled water a nund)er of times and dried on unglazed ])orce- lain |)lates. ■"C" was bought from a chemical supply house (under the name of wlu'at starch, though it was really coi-nstai-ch) and seemed pure enough to use as it was. Cornstarch was ])repai'ed in the laboratory by washing it from corn meal by the method used with the grated potatoes. The product was not pui'e. But it seemed much inoi'c important to have the starch as nearly as possible in its natural condition, although mixed with im|)urities. than to use any methods to get rid of these that might change the comj^osition (»f the stai'ch. Some diiliculty was experienced in obtaining a connnercial pre])ara- tion of wheat starch. Twice coi-nstarch was received from dealers in response to orders for wheat starch. The sample which was iinally obtained was used in the condition in which it was received. An impure starch was prepared in the laboratory from wheat Hour by the method used for cornstarch. In this case, too, the desire to kee|) the starch as nearly as ])ossible in its natural condition ])revented further ('(fort at ])urific-ation. [J5ull. ■J()2] 19 A commercial preparation of arrowroot starch was bought at a drug store in a tin-foil packag<\ It seemed to be very pure and was used as purchased. ^ The other forms of starch used were pearled l)arley, ordinary rice grains, tapioca in tlu-ee forms, and sago of two sorts. These were bought at a grocery store and used in c[ualitative experiments \\'ith- out any attempt at purification. As previously stated, the ferments used in the experimental work were malt diastase, taka diastase, and ptyalin of saliva. In the case of malt diastase, a sufficient quantity of a standard preparation from barley malt was purchased at the beginning for use in all the experi- ments here reported. A homogeneous mixture was made, wliich was kept closely corked in a dark-colored bottle. The taka diastase used was also the usual commercial preparation. The saliva was collected fresh for each set of experiments in which it was used. The apparatus used included a microscope and l)urettes, pipettes, flasks, etc., such as are needed for ordinary volumetric work. The hot-water stage for the microscope, mentioned on page 26, was made from a flat bottle of clear glass, which was fitted with a cork with two holes. Tlu'ough one passed a l)ent glass tube, the other end of wliich dipped in a hot-water bath, and througli tliis tube hot water was siphoned into the bottle. A glass tube through the other hole in the cork allowed for the exit of water. The water l)ath used for keeping the digestion mixtures at a con- stant temperature was made double. The outside was a largo galvanized-iron can, such as is used for garbage, holding 55 liters of water. Inside of this was a large deep can 10 inches less in diametci liolding about 13 liters. Tliis was supported by means of ten wires ])assing from the rim of the iimer can to that of the outer, as the spokes of a wheel from a hollow hub to the rim. Thermometers were kept in both the inner and outer parts. A small gas flame below supplied the heat. An attemjit was made at first to regulate the temperature by means of a mercury thermostat, but this ])roved unsatisfactory. It was found simpler and better to have the burner connected directly with the gas fLxture, with a long slcuider rod attached to the stopcock. In turning the gas up or down the end of this rod described an arc of a large circle. The point on tlus arc at which the rod rested, when under ordinary contlitions the temper- ature of the inside of the bath was 55° C, was marked on the table. With this normal position determined the gas coukl safely be turned up or down as needed, and the right temperature being obtained it was found easily possible to keep it constant within one-quarter of a degree for the important first half hour of digestion. In long experi- ments the t(>mperature often varied two or three degrees chu'ing the time when the changes could be noted. [Bull. 2021 20 Goiierallv speakinjr, the difrerent sorts of starcli tested were made into a j)aste witli distilled water before treatment with diastase. Tn niakinu' the paste the starch was dried overnight on wei*j;lied wateh crystals in a calcium chlorid desiccator, with approximately 1 ijram of starch on each watch crystal. This (juantity was then cpiickly adjusted to the exact weijj;ht of 1 o;ram, on a delicate balance, and then carefully transferred to an P^hrlenmeyer flask of 170 cid^ic centimeters capacity and shaken with ."> cubic centimeters of cold distilled water. To this was quickly added in three parts 45 cubic centimeters of boilinix distilled water, the mixture beins; shaken after each additic^n. The pre|)aration of all j^astes for quantitative work was hejxun in this way, since it involves no risk of losing any of the starch, as does the method ordinarily described of j^ourino; the mix- ture of starch and cold water into the boiling water. The further ])rej)aration of the pastes used in the work reported on pages 25-30 consisted in cooling them to 55° C, the temj)erature at which the}" were digested. Other ])astes were either boiled or cooked in the boiling water bath the indicated length of time, a return condenser being used in all cases to keep the concentration constant and to prevent skin formation. To prevent lumping, the pastes to be boiled were l)r()ught to the boiling temperature with constant shakings before attachment to the condensei-. After cooking the given length of time, the pastes were corked and cooled to 55° C, and by means of wires around the necks of the flasks suj)ported inside of the inner part of the constant temj^erature water bath. For digesting the starch, nnilt extracts were madi^ fresh evervdav. Three and a half grams of the commercial malt ])owder was mixed with 175 cid)ic centimeters of distilled water at n.'}° C. and kept at that temperature for an hour, being shaken every 15 minutes. The solution was then filtered with the aid of an as|)irator and heated again to 55° C. Twenty-live cubic centimeters of this extract meas- ured with a |)ipette was added to 50 cubic centimeters of {\w starch jiaste in each llask. Taka diastase extract, which was used for |)ur|)oses of comj)arison, was prepareic centimeters of Fehling's solution. From the calculated result was sul)tracted the 0.287 gram of sugar found in 25 cubic centimeters of the rualt extract used. This method was found to be much the most satisfactory of the methods tested involving the use of Fehling's solution. EXPERIMENTS ON THE EFFECT OF COOKING STARCH. EXPERIMENTS WITH RAW STARCH. vSonie preliminary experiments were made \\ illi law starch in whicli the microscopic apj)earance and its reaction \\itli ioihn wcr(> stu(nc(l and (hgestion experiments were ma(k\ As is well known, raw starch grains from (hll'creiU kinds of phints vary in tlicir microscopic appearance*. 'I'hc potato and arrowroot h Y\i:,. :i.— Uuw stan^li Kniins maguified ;iSo times: a, from potatu: b. from arrowroot; c anat and l)arlev starch grains are lens shajx'd, generally showing a hollo\V(>d space in the center also lens sha|)(Ml. Jlice and corn starch grains are generally i)olyhedral in sha[)e, with a more or less star-shaped crack in the center of each. As preliminary to the exj)erimental work the diflVrent starches were examiiKMl with a microscope and compared with standard samples as a means of establishing their identity. Ivaw starches of different origin show dillerences when treated with a .solution of iodin. When iodin solution is added to raw- starch grains a dark pur{)Iish blue results. The })otato and arrow- root grains showed but little of the ))m'ple, while the other starches which were used showetl on standing a decidedly red-jjurple color. [15iill.L'(l2J 9^ DIGESTION EXPERIMENT NO. 1 . In this experiment potato, arrowroot, corn, wheat, barley, and rice starches were used. One-half gram of each kind of starch was })laced in a flask and to each was added 25 cubic centimeters of cold distilled water, 3 cubic centimeters of fresh untiltered saliva, and a few drops of thymol solution. The flasks were stoppered and put in the constant temperature water bath at 40° C. The samples were tested \\'ith the naked eye and with the microscope at the end of 14 and 38 hour periods. With the naked eye no appreciable diminution in the quantity of solid starch in the bottom of the flask was noticeable at either time. At the end of 14 hours neither the potato nor arrowroot grains showed any change when examined with a microscope except a slight swelling. In most of the corn and rice starch grains the center was apparently all dis- solved, leaving be- hind a thick shell which showed long deep clefts along what were the edges of the solid polyhedral g r a i n (see fig. 4, a). In the case of wheat and l)arley starches many of the grains showed no change except corrosion at the surface, but many others were notliing l)iit hollow shells, tliinner than in the corn, with pitted surfaces and cracks around the edge nuich like the openings of oyster shells (see fig. 4, h). At the end of 3S hours many arrowroot grains showed surface cor- rosions (see fig. 4, c). Some potato starch grains showed the same kind of corrosions. Others showed a change at the small end (see fig. 4,(/). While the rest of the starch grain stained a dee}) l)hie with iodin, the small end did not stain at all, but api)eared to be of a colorless glistening substance less solid than the rest of the grain and showed the same laminated appearance. The corn, ric(\ wheat, and barley starch grains appeared j)ractically th(> same as the day before. They slowly stained a blue red with iodin. At some places at the surface the same glistening colorless substance noted in the potato grains was ai)|)arent. rBuU.-J02J Fig. 4.— Raw starch grains that have l)oon digested 38 hours: o, corn- starch showing clefts at what wore tlie edges of the grain; 6, wheat starch grain, side view, showing oyslcr-sliell-hke opening and pittings; c, arrowroot starcli grain sliowiiig corrosions; d. potato starch grain stainept that it is thinner at the small end in the potato starch grain l)ecause the hilum or point of origin is at this end. TT(Mice it is that in this grain the outer layer dissolves first at the snuill end. The colorless substance remaining at the end of the digestion period may be reverted amylodextrin, since digestion is very slow. -ludging by the results of the digestion tests, and from their normal condition and appearance without treatment, the six kinds of starch studied naturally grou}) themselves into two divisions, namely: Wheat, corn, barley, and rice starches and potato and arrowroot starches. In the case of the wheat, corn, barley, and rice starches the natural spherical form has become flattened. The starch grains show internal cracks and a red-purple stain with iodm on standing. The digestion is apparently more rapid than with the second group. In the case of the potato and arrowroot starches a rounded form is retained; normally no internal cracks are visible, and a blue-purple [Bull. 2021 25 stain is obtained with iodin -which does not become red purple on standing. The digestion is very slow. The starch grains of the first group are produced in the crowded cells of seeds and hence their flattened form. In these cells the loss of water on maturing is great and hence the cracks. The starch grains in the second group are produced in less crowded cells, and are not flattened. The drying at maturity is much less and so there are fewer cracks. EXPERIMENTS WITH STARCH COOKED BELOW THE BOILING TEMPERATURE. That starch may be cooked it is essential that eYerj starch grain be brought into contact with water of at least 60° to 80° C. (140° to 178° F.). If it is to be cooked below boiling temperature, this is best accomplished for quantitative work by simply adding boiling water to a mixture of starch and cold water and slinking. But a paste can x~§ a, 6 c Fig. o. — Potato starch magnified 87 times, showing successive stages in paste formation: Changes that come with the absorption of hot water— a, raw starch; x, hilum or point of origin; 6 and c, intermedi- ate stages, showing granulation beginning at the point of origin; d. condition in unboiled paste; y. outer layer; z, folds in outer layer. be made equally weU by the more common method of })ouring the mixture of starch and cold water mto boiling water while stirring. The same starches were used as in the previous tests. "\Mien the pastes were exammed with the naked eye, marked differences were noted between potato and arro%\Toot starch on the one hand and corn and wheat starch on the other. The potato and arrowroot pastes were colorless antl transparent, ^\■hile tlic others were whitish, more opaque, and appeared to be less thoroughly cooked. In all cases the pastes made without boiling separate on standing a short time, the starch sinking to the bottom. When examined with tlie microscope it was apparent that the starch grains were much swollen but were not broken. The transition from the raw to tlie swollen condition can be observed with the microscope if the paste is made by using all cold water and gradually raising the temperature. The absorption of water by the starch grain begins at the point of origin. It shows as a granulation which gradually works its way outward (cee fig. 5) until nothing solid but the [Bull. 2021 26 outer layer is left in the potato anil arrovrroot starch strains. Tn the corn and wheat some of the (z;ranular material also remains in solid form (see lig. ()). It is probably this which causes the o})acity in these pastes. The outer layer is ureatly stretched and, if the paste is cooled, collapses. Sometimes t hese collaj)sed forms arc hard to recoij^nize as such. Tn the wheat this is es- pecially true. The edo;es of the flat- tened form curve , ., , , , , over on themselves Fifi. 0. — Siaroh prains fromxinhouwl psistos: oand o, wheat staroh — n.col- liipscd swoUon };r;uii with cdKt's curved over and -sulid Kf^iiiil;!!' i»;»ti'- SO tliat I lu^ rCSUlt rial; h. the same j,'rain flattened out by heating on tlie liol-waler sta^'e; /)ft(»ii looks more c. collapseil swollen ^'rain of cornslareh showhiK' solid jiranular material. like a star than a ball (see i\<^. 0, a), but by usinu; a hot-wat(M' sta»j:e (see p. 10) on the micr()scoj)e these forms can be seen to unroll , b) ajid to stretch out to their ori corn and wheat starches are <:;rouped toi^ether. b(>cause in (>acli case these starches ^nnxe the same color I'eactions. T.\HLt; J. — Diiicstibility (is indicnttd hij indiii hsi of diffcrnit starchca cool'rd below 100° C. Titne of digestion of slarch. Anioiint of iodin usp(: Color proxtrins, called li<>re for the sake of simplicity according to the color they take with iodin, red dextrin and brown dextrin. The starch and amylodextrin have the greatest aflinity for the iodin, and only after more than enough iodin has been added to stain blue all the starch and amylodextrin present can any permanent red or puri)le be obtained. Alter this, if there is enough red dextrin present to dominate the blue, we see this color; otherwise we have the purple mixture. All the red and blue dextrin having united with iodin, the brown dextrin begins to show, anil if it is present in large enough (piantities the brown hue alone appears. As digestion proceeds, first the blue, then the red, and then the brown dextrin diminishes in volume and disai)pears. With the corn and wheat stardies there is the same series of color [Bull. 2021 28 changes, but in acklition tliere is another series which, especially with this method of cooking, often conceals the one corresponding to that of the potato and arrowToot starches. In order to get the gradation of color noted in columns 3 and 4 care must be taken to add enough iodin and to do it slowly. The "old rose" differs not at all in color from some mixtures of blue and red dextrin with iodin, and heretofore has evidently been taken for the same, but as shown here the compound stained old rose has a stronger allinity for iodin than the blue dextrin, while red dextrin has a less aflinity. The old rose always appears before the blue with these starches if iodin is added slowly. Moreover, this red is not quickly changed to browm by digestion, as is the red dextrin, but continues for days rather than for minutes. However, it must not l)e concluded, because it takes corn and wheat starches over 500 times as long as potato and arrowToot starches to reach the point in digestion where they no longer color iodm solution, that the relative tiigestibility of tli(» two pairs can be expressed by the ratio 1:500; for the part of the paste which is slow to digest represents only a small percentage of the whole, as is in- dicated by the shght color in the test tubes aft(>r 20 minutes' digestion. Moreover, there are probably factors, such as the accumulation of by-products, which make the (Hgestive j)ro('ess increasingly diflicidt as time progresses. When exaniin(Ml wilh the microscope it is seen that the swollen starch grains before digestion give to the })aste the blue color with iodin, but after a minut(>'s digesti(m this color is seen also in the solution outside of the gi'ains without any :i{)parent l)reak in the outer layer. ^^Tien most of the bhie substance has diffused out, this outer layer is seen to be stained red. It persists until the red dextrin is the domhiant compound in solution, and then (luite quickly disap- pears. After the solution of the outer layer the digesting i)otato and arrowroot mixtures are fr(>e from solid substance. The line granular substance of the corn and w heat starches, however, continues undis- solved. The particles refore, that it cau not l)e the amyhxiextrin, which Meyer thinks is present in those grains that stain red with iodin, for that is soluble in warm water and easily digestible. Nor can it be Syniewski's maltodextrin I (the erythrodextrin of others) which stains red with iodin, but is very easily soluble in warm water and very easily digestible, and which has less affinity for iodin than blue dextrin. It is evidently not a reversion product, for its color shows in the raw starch grains. It has some of the ])roj)erties of Meyer's rt'-amylose, in that it stains red with iodin, and is difficult of solution and digestion. However, this substance is not distributed in the starch grains as he believed <^r-amylose to be distributed. For the sake of simplicity this starch constituent may be called rose amy lose. Th(^re is j)resent in all the swollen starch grains tested a substance with the following properties: It stains blue with iodin, it digests rapidly even if the outer layer has not been broken, and it apparently constitutes the whole of the potato and arrowToot starcli grains except the outer layer and all of the inside of the corn and wheat starch grains except about 10 pcv cent which is rose amylose. It is evidently the /^-amylose of Meyer, the granulose of others. For the sake of uniformity this portion or constituent of the starch grain may be called blue a]iiylos(\ It was further (>videnl from the digestion experiments reported that the outer layer of the starch grain is not broken or dissolved in paste formation below lOO"" C. It stains red with iodin and is perhaps the cause of the slight i)urple hue seen in raw potato and arrowroot starches so stained: it is harder of digestion than blue amylose; it is easier of digestion than rose amylose, and it seems to [Biill.2li2] 30 differ from the other hiyers in more resj)ects than in density, since even in its stretched condition it stains red instead of bhie or rose with io(Un. Perhaps it is formed as a residt of changed contUtions in the cell sap that come after the end of the growth period in the plant, and wliich might affect the outer layer most because of its position even tliough the whole grain be porous. The substance of this layer may be called red amvlose. EXPERIMENTS WITH STARCH BOILED 5 MINUTES. The ])astes which are made by boiling the starcii for 5 minutes are more homogeneous and do not show the tendency to separate wliich is noted with uid>oiled pastes. The cereal starch pastes are still not clear wlien thus cooked though the others are. The microscope shows that large munbers of the starch grains have broken, corn and wheat starches showing a nudtitude of fragments of broken outer layers and more or less granulated inner ])arts. With tlie arrowroot and |)otato starclies the outer layers are more [)ersistent and, collapsed as they are, it is often hard to see tlie breaks, but with iodin it is easy to sec by their greater trans- ])arencv that they have lost their inner parts, wlii(di are more uni- forndy solid)l(' than in the other starches. DKiESTlOX KXlM'.lv'IMKXTS XOS. () .VXD 7. In thes(> experiments both methods and conditions were the same as in (Hgestion exj)eriments Nos. 4 and o e.xcept that the pastes were boiled o minutes, and the results obtained were on an average tile same as those given on j)ages 2()-o0 for tlie experiments with the unboiled })astes, all of ^\hich are included in Tables 2, 3, and 4 for purposes of comj)arison. Tlu> conclusion may therefore l)e drawn that brief boiling, while it breaks many starch grains and makes a better paste, does not ap|)recial)ly improve the digestil)ility of the starch. EXPERIMENTS WITH STARCH BOILED MORE THAN 5 MINUTES. To pre\('nt the formation of skin and the loss of water all the sani])les w(>re ])oiled in a iiask with a return condenser. With continued boiling the ))astes gradually change in ai)pearance and character. At the end of .'5 hours they are distinctly more homogeneous, less viscous, and more transparent, and they also acquire^ a slight yellowish tinge. The taste is changed from one more or less suggestive of the odor of boiling laundry starch to one very slightly suggestive of caramel. This slight change of color and taste is not due to any ordinary scorching, because there is n a]>i>arent sticking of the ])aste to tlie bottom or sides of the ilask. The microscope shows a gradual diminution in the solid nuiterial, i. e., the outer layers and granular substance ])resent. rniiii. I'd')! 31 DIGESTION EXPKKI.MEXT XO. 8. Potato and arrowroot starches wore used for this experiment, and the pastes were boiled for 3 liours and were treated with malt diastase in the usual way. The results obtained are summarized in the fol- lowing table, which shows the time required for the brown dextrin reaction to disappear from the digestion mixture. For purposes of comparison the results of similar tests with unboiled starch and with starch pastes boiled for o minutes are also included. Table 2. — Changes brought ahout in potato and (irrotrroot stnrchrs by long continutd cooking. Time required for disap- pearance of brown dex- trin reaction. Time required for disap- pearance of l)rown dex- trin reaction. Kind of starch. With '^^''^^ With pastes boiled 3 hours. Kind of starch. With unboiled pastes. With With pastes pastes boiled .") boiled 3 minutes, hours. ^^nuteli. Minutes. Minutes. I 10 -201 185-20.', \m-2ll 10i-205 j,^j^,,^_ Average. 20 20 Minute.t. Minutes. Minutes. 10-21 20-2.') 18-20 17-18 1 10-23 10-2U 20-2.') ' 10-24 23-25 20-25 Average. 21 211 20i DIGESTION EXPERIMEXT XO. 9. In this experiment arrowroot starch was used. The paste was boiled for 3 hours, when, after digestion in the usual way and tested with Fehling's solution, it was found that 71 to 73.5 per cent of starch had been converted into maltose — the same quantity which was notccl in the case of unboiled arrowroot jniste (see p. 28). It would therefore ap])ear from the results ol)tained in the diges- tion experiments Xos. 8 ami that ])otato and arrowroot starches are not mad(> more digestible by boiling. DIGESTIOX EXl'KKI.MEXT XO. 10. The cornslarch used was a commercial ])reparation designated '"A." The unboiled pastes and samples, boiled from ."> minutes to 3 hours, were prepared, digested in th(> usual way. and tested with iodin. The results which were obtained are as follows: Table 3. — Changci brought about in cornstarch by long continued cooking. Time required for disappearance of red dextrin color reaction with paste. Time required for disappearance of brown dextrin color reaction with paste. No. of test. Un- boiled. Boiled Boiled 5 min- 30 min- utes, utes. P,^j'!;f Boiled ^^;;i^"- 3hours. ,.„ 1 Boiled toiled. '^^-jy- Boiled 30 min- utes. Hrs. 2. 3-2. 5 2. 5-3. 3 Boiled 45 min- utes. Ilrs. 'i."2-i.'7' Boiled 3 hours. 1 Min. 7-16 5-14 Min. 0-12 8-11 Min. 7 -10 7. 5-12 Min. Min. 7- 9 8. 5-0. 5 6-13 ; 8 -12 IJrs. 23-24 24-25 Hrs. 28-29 26-27 Min. 51-50 ■{ 9-12 10-18 21-22 21-22 4 .\verage . . . N 11 10 9.25 11 0.5 23 1 27 2J '■1 ■'' [Bull. 202] 32 Comparison with Tiil)l(' 1 makes the results of this experiment more easily understood. The first part of Table 3 corresponds with the results given in Table 1 with samples cooked 10 minutes and shows the time which it takes for the disappearance of the red dextrin of the easily digested portion. As explained in connection with Table 1, it is impossible to see the transition at the end of 20 minutes with corn and wheat starches. The relatively rapid digestion of the rose amylose (23 hours instead of over 7 days, as noted in the experiment reported in Table 1) is doubtless due to some peculiarit}^ of the preparation of cornstarch used (see pp. IS, 27). From the experiuiental data it would appear that the l)hie amylose of corn, as of potato and arrowroot starches, is not made more easily (Hgestihle by long boiling. The rose aniylos(> of cornstarch is appar- ently nui(I(^ more quickly (Hgestihle by long ])oiling because some of it is changed to blue auiylos(>. DKJKSTIO.V EXPKin.MF.Xr X<). 11. In this, as in the preceding test, conuuercial cornstarch d(>signate(l ''A" was used. The pastes were boiled for o minutes, 30 mimites, 1 hour, aud .') hours, respect i\'ely, and dig(>sted for 30 mimites. When t(\sted with F(>hling's solution it was found that a sample l)oiled 5 mimites gave 5!) to 61 .2 })i'r cent maltose: the ])aste boiled ."!() mimites 00 to 02.1 per cent; the })aste boiled 1 hour OiJ.o to Oo.l ])er cent, and tlie ])aste boiled 3 hours 02.1 to ().'). 7 per cent, as compared with 5r lengths of tini(>. DICESTION' KXPHHIMFA'T XO. 12. Samples of pastes made from wlieat starcli wdo boilml for 1 and 3 hours, respectively, and digested with malt <>\tract in tlie j)roportion commonly used in the experiments, namely, 2.1 cubic centimeters of malt extract and 50 ciil)ic centimeters of water to 1 gram of starch. ^Mien tested with iodin all the samples re(|uired over 7 days for the (hsa])pearance of the characteristic d(>xtrin color, the only noticeable difference being that with the pastes boiled for the longer period there was less substance left undigested at the end of 20 minutes than with th(> other pastes. DKJESTIOX EXPERI.MEXT NO. 13. In this experiment wheat starch was used and the paste was boiled for (hfl"erent periods of time. The proportion of malt extract to starch used for digesting the paste was ten times greater than in the majority [Bull. 202] 33 of the tests, 50 cubic centimeters of malt extract and 20 cubic centi- meters of water beinij used to 0.2 (rram of starch. The resuhs obtained upon testin^:; with iodin are shown in the followino; table : Table 4.— Changes brought about in wheat starch by cooking for dljfc rent lengths of time. Kiiui of starch. Time required for disappearance of red dextrin color reaction with paste. Time required for disappearance of l)r(nvn dextrin color reaction with paste. Un- boiled. Boiled .5 minutes. Boiled 1 hour. Boiled .3 hours. Un- boiled. Boiled ,5 minutes. Boiled 1 ' Boiled 3 hour. hours. W.'ieat starch: First test Second test Average Mmules. Minutes. 3-.') 5-7 \nnutes. .5-6 .5-6 Minutes. :v>-.5 4-."> Hours. 4-18 4-18 Hours. Hours. 4-18 4-18 4-18 1 4-lS Hours. 1-li A\\ 5 oil 4a! ' Xo diflVrence is a])])arent from the data included for the dis- apj)earance of the brown dextrin color reaction, owini!; to the fact that the end points of the reaction occurred some time durintj the ni<;ht.\\hen it was not possible to make the necessary observations. There was, however, as in experiments Nos. 6 and 8, a noticeal)le difference between the })astes boiled not more than o minutes and those boiled 1 and W hours, as shown by the (piantity of im(liij;ested material left at the end of 20 minutes. It secMiis fair to con(dude, tluM'efore, that increasin*]; the [)r()portion of diastase to starch shortens the time re(piir(>d for both stat!;es of digestion. Lon<2; boilin*: makes the rose amylose of wiieat starcli more easily di<;estil)le, as is the case with cornstarch, and as previ- ouslv found has no etfect on the blue amvlose. DKJESTION' KXPEKl.MEXT XO. 14, In this test wheat starch ])astes boiled for dilferent periods of time were used. The pastes were digested with the ordinary proj)ortions of malt extract, namely, 25 cubic centimeters malt extract and 50 cubic centimeters water for 1 gram of starch. On testing with Fehling's solution it was found that the uuhoih'd pastes showed (54.8 to 68.3 per cent maltose in the first test antl ().3.8 to 66 per cent in the second test. In the case of the samples boiled 3 hours 64.8 to 67.3 per cent maltose was noted in the first test, and 67.5 to 70.5 per cent in the second test. The increase in (Ugestihility due to long boiling, as indicated by the production of maltose, is small, being even less than was noted with cornstarch. There seems to be a discrepancy between the results of the iodin test (ex))eriment No. 13) and the test with Fehling's solu- tion. The iodin test would indicate that wheat starch is more diffi- cult to digest than cornstarch, while the test with Fehling's solution [Bull. 202] 34 would iiulicalc the reverse. Further iiivestiijjatioii is recjuired, there- f()r(>, l)el'ore general couchisioiis eaii be drawn regarding the relative digestibility of wheat and eorn starches. In general, the conclusion seems warranted, from experiments Nos. 8 to 13, that pastes made from potato and arrowroot starches, which have been boiled even for 3 hours, do not digest more quickly than similar pastes made with hot water but not heated to the boiling point. Corn and wheat starches are rendered somewhat more digestible by long boiling owing to the gradual change of the rose aniylose which they contain to blue amylose. A long cooking j)(>riod is, however, necessary, as the change to l)lue amylose is not coni])lete after 3 hours' boiling. EXPERIMENTS WITH STARCH COOKED IN A WATER BATH OR DOUBLE BOILER. AVheii starch paste is cooki^l in a water bath or double boiler, in which the water is kej)t at boiling temperature, the starch grains do not break but settle to the bottom of the vessel in which the paste is cooked. DKilOSTION EXPERIMENT XO. 1"). Cornstarch was used in this experiment and was cooked in a boiling water bath for 1 hour, using a flask with a return condenser. The temperature of the starch mixture itself was found to be 95° C. When tlu> mixture, after digestion in the usual way, was tested with iodin it was found that the brown dextrin disappeared in 5 to 18 hoiu's as compared with 23 hours in the case of unboiled paste and '2\ hours in the case of starch boiled 30 minutes, the values which were obtained in previous tests. T\\v conclusion is drawn, therefore, that cooking cornstarch below 100° C., as in a. water bath, increases its digestil)ility, l)ut not so ra])idly as cooking at 100° C. EXPERIMENTS ON THE DIGESTIBILITY OF STARCH BY DIFFER- ENT DIASTASES. It- was felt that it was desirable to compare the digestibility of starch by malt extract with that by other diastases, and so experi- ments were undertakt^n in which ptyalin and taka diastase were used. DIGESTION EXPERIMENT NO. 1(). Wheat, corn, and ])otato starch pastes were prepared in the usual way and boiled for 1 minute. One sample of wheat starch was also boiled for 3 hours. The samples were digested with saliva (see p. 10) and then tested with iodin. In the case of potato starch [lUill.202] 35 boiled for 1 minute the brown dextrin reaction disappeared in 8 to 12 muiutes, and with corn and with wheat starch, cooked the same length of time, in 24 hours in each case, while with wheat starch cooked 3 hours 3 to 18 hours were required. The corn and wheat starches showed the slow disappearance of rose color, but the potato starch did not. The ptyalm of saliva, like malt extract, according to the data reported, digests potato starch quickly and corn and wheat starch slowly. When wheat starch was boiled for 3 hours it was digested more quickly than when boiled for 1 minute. DIGESTION' EXPERIMENT XO. 17. Wheat, corn, and potato starch ])astes were prei)arrepare(l from a commercial wheat starch, from a sample of wheat starch jH-epared in the laboratory, and from wheat starch in the form of wheat fiour, and were tested with iodin after digestion in th(> usual way. In all cases digestion occurred in two stages, the old-rose color appearing. [Bull. 202] 37 The conclusion was reached, therefore, that both bhie and rose amylose are normal constituents of wheat starch. In general, it seems fah to conclude that the results obtained in experiments Nos. 4 to 18 on the composition and digestibility of wheat, corn, and potato starches are normal and are not ascribable to the different methods used in isolating the starches. ADDITIONAL ATTEMPTS TO IDENTIFY ROSE AMYLOSE IN PO- TATO STARCH. Since recognizable quantities of rose amylose in wheat and corn starch are found to vary with experimental conditions, it was thought that it might possibly be detected in potato starch under conditions unfavorable for digestion and three experiments were made to study this point. DIGESTION EXPERIMENT NO. 21. A sample of unboileil paste was ])repared and digested at 61° C. instead of 55° C. On testing with iodin it was found that the time of digestion was slightly increased, but there was no indication of rose amylose. DIGESTION EXPERIMENT NO. 2-. Unboiled potato starch paste was treated with a fifth as much dias- tase as w^as used in the majority of the experiments, the quantities being 5 cubic centimeters of malt extract and 50 cubic centimeters of water to 1 gram of starch. The time recpiired to reach the achro- matic point was increased from 20 to SO minutes, but there was no indication of rose amylose. DIGESTION EXPEKIMENT NO. 2.S. Raw potato starch ground w iih glass fragments to break the outer layers of the starch grain was tested with iodin and it was apparent that digestion occurred much more ra])idly than with unbroken raw starch, but much more slowly than with starch grains which had been swollen with hot water. \Vhen examined with the microscope they showed that the outer layer was red with iodin whenever it was sepa- rated from the inner parts, but the digestion mixture tested with iodin in test tubes showed no old-rose color before the blue and no tendency of the blue to revert to the okl-rose color on standing. The general conclusion seems warranted, therefore, that rose amy- lose is not present in potato starch. The red amylose of the outer layer of the potato starch grain does not give the same reaction. [Bull. 202] THE POSSIBILITY OF REVERSION IN WHEAT AND CORN STARCHES. AccDidiiig to Maqiienne and Jloiix, reversion does not occur above 60° C. The following experiment was made to determine whether rose amylose is a reversion product: DIGESTION' HXPEKIMENT XO. 24. Unboiled pastes were made from wheat and corn starches and care was taken not to cool the pastes below 61° C, and digestion was con- tinued at this temperature. Treatment with iodin showed that di- gestion was retarded in both cases, but the results were otherwise the sauie as in previous experiments. Froui th(> results of this ex])eriuient and the fact that the rose color is also api)anMit in the raw graiu on treatuient with iodin, the deduc- tion would seem warranted that rose amylose is not a reversion ])r()(luct. Little work was done on the subject of the formation and proper- ties of the possible reversion product in connection with the investi- ii;ations reported in this bulletin, but enough was accom])lished to suiigest that such a reversion product i> formed when raw starch di- gests slowly, when starch i)astc> are fro/en, when a >kiin forms on stai'ch paste, and when a partially digested starch i)ast(> stands at ordinary laboratory temperature. It does not seem to be formed when dilute (2 i)er cent ) starch pastes are kept st(»])pere(l for 24 hours at ordinary laboratory temperature. As regards the properties of this reversion |)roduct, it does not stahi with iodin; it is more dillicult of solution and digestion than rose amylose, and it dissolves in ])ota:rsium hydroxid, forming a solution which after neutralization gives the characteristic blue color with iodin. This reversion product, which is probably the same as Syniewski's reverted amylodextrin, may be regarded as distinct from the red amylose and rose-amylose Ixxlics, which are apparently now for the first time definitely identified, but it has undoubtedly been ct)id"used with them l)y earlier investigators who have variously called it starch cellulose, amylocellulose, farinose, and rt-aniylose. Further investiga- tions are necessary before definite conclusions can be drawn ;is to the (liet(4ic significance of the reversion of starch. EXPERIMENTS ON THE DIGESTIBILITY OF TAPIOCA, SAGO, PEARLED BARLEY, AND RICE STARCH. In coimection with the investigations which have been reported it seemed desiral)le to test the digestibility of some other common sorts of food starches and accordingly tapioca, sago, pearled barley, and ric(> w(>re selected for study. [Bull. L'U'.'l 39 DIGESTION EXPERIMENT NO. 25. Three commercial preparations of tapioca, two of sago, one sample of pearled barley, and one of rice were selected for this experiment. No attempt was made to free the starch from cellulose or other sub- stance, the materials as purchased bein^; simply cooked on a water bath until soft. When the starch was dio;ested tlu> proportion of malt extract to starch was not definitely fixed, since no attempt was made to secure other than qualitative residts. When tested with iodin, it was found that the tapioca and sago digested in one stage as did potato and arrowToot starch (see Table 1). Rice and barley showed distmctly the two stages found in the digestion of corn and wheat starches (see Table 1). These results, therefore, confirm the conclusions from earlier experi- ments that the starches produced in the cereal grains difTer from those in roots and stems in that they contain rose aniylosc, and hence are more diliicult of complete digestion unlc^ss cooked for a long time. CORNSTARCH v. WHEAT STARCH. The experiments here reported luive shown that, in so far as there is a difrerence, cornstarch digests more rapidly than wheat starch and the digestibility of both is ahke increased by long-continued cooking. The housekeeper's rule that cornstarch should be cooked longer than wheat starch seems to be founded on th(> fact that cornstarch })astes, unless cooked .'50 to 40 minutes, have a |)ecnhar Ihivor, disagreeable to most persons, which is not noticed with wheat and other starches. Efforts were ukuU' to learn the cause of this flavor, but without suc- cess, and more work is (Uvsirabk^ on this point. It seems probable that the flavor is due to an impurity in the commercial starch which either evaporates or is deconipos(Ml on long-continued cooking. From the standpoint of Uavor. therefore, cornstarch shouhl be cooked 30 to 40 minutes. Moreover, if Pawlow" is right in concluding that foods which })lease the palate are more easily digested than those which do not, it may be true that th(> long-continued cooking increases the digestibility of the cornstarch under natural conditions more than would be hidicated by t heart ilicial digestion experiments Iumc reported. DEDUCTIONS AND CONCLUSIONS. In applying the results of these experiments it must be reinendx'red that in general the starches used in the experiments had been s(>parated from the other sid^stances with which they are ordinarily combined in foods. Hence, hi cooking starchy foods, such as potatoes, it is neces- sar}' to consider the time required for the heat to j)enetrate to the center of the mass, and to reineml)er that the effect of cooking })otatoes, for instance, includes changes in the cellulose ])resent as well as in the « The Work of the Digestive Glands. LoikUhi, 1902. [Bull. 202] 40 starch. Furthermore, the relative digestibihty of two starchy foods may (k^pend on other factors than the starch content. For instance, the fact that rice is considered easier of di.ij:estion or better tolerated than potatoes, an opinion which seems to be based veiy largely on experience gained in invalid dietetics, is, if true, probably due to the relative digestibility of other constituents than the starch in the foods, since rice starch, as shown by the experiments here reported, contains the slowly digesting rose amvlose not found in potato starch. It nnist l)e remembered, further, that the digestion experiments reported were made by artificial rather than natural methods, and that it is hardly possible in lal)oratory experiments to ikiplicate all the condi- tions i)resent in natural (Hgestion. Lee," among others, has given reasons against drawing conclusions from test-tube experiments with regard to the limits of digestion. He found that when a dialy/.er was used to nuuove the products of dia- stase action as th(\v wcvc formed, digestion was carried much further than unr usual test-tube conditions. However, such objections are not so applicable whei'e the results sought are oidy comparative, as was the case in the present work. As suggested earlier (see p. 2S), it is probal)l(>that ros(> amy lose would have been found to digest more rapidly had a dialyzer Ixhmi emj)loye(l, but such a modification would also aid the digestion of hlue amylose, and it is probable that the ratio between the rates of digestion of the two woidd not have been much alteretl. However, it is reasonable to suppose that there may be physiological factors in natural digestion which were not pivsent in these experi- ments, which would alter the relative r(>sults obtained from laboratory tests.'' Bearing in mind these limitations, it se(>msfair to draw the fol- lowing conclusions from the experimental evidence reported witk |)otato, arrowroot, taj)ioca, sago, wheat, corn, rice, and barley starches. Three substances arc found in raw starch grains, which are designated in this bulletin according to the color they give with iodin — blue amylose, red amylose, and rose amylose. \M\w amylose constitutes tke entire inside of the starch grains of potato, arrowroot, ta|)ioca, and sago and !>() per cent or more of the inside of the cereal starches studied, luimely, wheat, corn, rice, and barley, and is identical with the substance called granulose or //-amy- lose by other investigators. It gives a blue color with iodin and in th(> solid form found in raw starch it is digested slowly because of its dejisity. It takes up water at 60° to S()° C. and forms the sticky colloidal substance known as starch ])aste, in which form it is very easily digested, l^ong l)oiling, at least to th(> ext(Mit of 3 hours, does not make it more quickly digestible. ".Tour. Physiol., 11 (1890). p. 226. &Sce the rcsiill of Ilarcouri's work, ]). 17. [Bull. 202] 41 Red amylose constitutes the outer layer of the starch grains. It ffives a red color with iodin and is more difhcult of dio-estion or change in water than blue amylose. The ])resence and the density of this layer hinders, though it does not prevent, the digestion of raw starch grains. \Mien starch ])aste is made without l)oiling the red amylose layer stretches, though it does not break, and in this condition is easily permeable and does not interfere^ with the more rapid digestion of the inner portions of the grains. When starch paste is boiled this layer breaks and a more homogeneous, though not more digestible, paste results. Rose amylose forms about 10 per cent of the inside of cereal starches and is not found in i)otato, arnnvroot, tapioca, and sago starches. It gives an old-rose color with iodin and shows a greater affinity for the iodin than the blue amylose, takhig the color sooner and retaining it longer. It digests more slowly than (Mtherbhie or red amylose, and hence it is fair to conclude that the cereal starchcf^ are not as readily digested as the other starclu^s studied. Uose aitiy- lose is slowly changed by cooking to the blue amylos(» form, hcmce cereal-starch pastes are made somewhat more (>asily digestible by cook- ing them for several hours. There is a fourth substance not found in luitural starch grains which has been fre([uently confused with red and with rose amylose. It is formed as a reversion ])roduct from solutions of starch and has been designated reverted amylodextrin by Syniewski. It is found in th(> skin that forms on a starch paste exposed to dry air. It does not stain with iodin and is more diflicult of solution and of digestion than any of the amylose forms of stai'ch. Raw starch, as shown by the ex|)erinunits reported, digests very much more slowly than the starch in the forms of a ])aste. Starch i)aste made below th(> boiling t by long-continued cooking. On tile other hand, the cereal starches are made more easily digest- ible by long cooking, though the change occurs very slowly and perhaps the increased digestibility is not sullicieutly great to justify the trouble, under ordinary circumstances at least, for separated starch such as is used in cookery. However, in the case of starch still inclosed in cellulose cells, as in many starchy foods, the long- continued cooking may be necessary. The commercial j)re])arations of cornstarch recpiire 30 to 40 mimites' cooking because of the improvcMuent in flavor which results. [Bull. 2021 42 Skin formation as well as lumps should be avoided in cooking; starch — the latt<>r contain raw starch, the former reverted amylodex- trin, and both are very slow of digestion. The selection of potato starch instead of corn or wheat starch for thickening sauces, in accordance with the custom of French cooks, is rational, since it contains no rose amvlose and so forms a clearer and more digestible sauce, and since it does not require 40 minutes' boiling for improvenuMit in flavor, as is the case with cornstarch. Increasing th(> |)i'()poft ion of diastase to starch in artiiicial digestion experiments markedly hastens the (hgcvstion of both blue and red amylose. It would appear, therefore, that eating starchy foods slowly, which would of coui'se increase the jjroportion of saliva and ptvalin to a given (puuitity of starch, would he of more value from the standpoint of digc^stion than would any cooking of starch beyond the stage of ])ast(> formation. [Bull. 2