WHEAT STUDIES OF THE FOOD RESEARCH INSTITUTE

VOL. XI, NO. 6 (Price $.50) FEBRUARY 1935

STARCH AND FLOUR QUALITY

HE PART starch plays in determining the quality of flour T is often overlooked because it is less conspicuous than that of gluten. In this WHEAT STUDY, existing information bearing on the influence of starch on baking value has been collected, some of the gaps in our knowledge are pointed out, and some of the problems awaiting solution are presented. With fuller knowledge, it may appear that variations in the properties of the starches of different flours influence baking quality ma­ terially. Already it is clear that starch absorbs about as much water as all other flour constituents combined. Variation in the absorption of flours might therefore depend to some extent upon the water capacity of their starches. The susceptibility of starch to diastase is one of the factors affecting the diastatic power of flour and in consequence panary fermenta­ tion. This susceptibility to attack depends not only upon the size of the starch granules and upon their intrinsic ability to resist conversion into sugar by diastase, but also upon their location-whether surrounded by a protecting envelope of gluten or freed from it. It is because fine grinding separates large numbers of granules from the gluten envelope that fine grinding increases the apparent diastatic power of flour. But it also injures starch granules so that they swell in cool water, thereby increasing the absorption of the flour; and at the same time some starch goes into solution. In consequence, diastatic power is increased because injured granules and dissolved starch are readily attacked by the enzyme. Finally, changes that the modified starch of bread undergoes in time seem to be the major factor in causing bread to grow stale.

STANFORD UNIVERSITY, CALIFORNIA February 1935 WHEAT STUDIES OF THE FOOD RESEARCH INSTITUTE

Entered as second-class matter February 11, 1925, at the Post Office at Palo Alto, Stanford University Branch, California, under the Act of August 24, 1912. Published by Stanford University for the Food Research Institute. Copyright 1935, by the Board of Trustees of the Leland Stanford Junior University. STARCH AND FLOUR QUALITY

The leavening of bread, as everyone knows, other functions, though less important ones is the consequence of the formation of car­ than gluten. These functions arc clearer now bonic-acid gas in dough by the yeast plant. than formerly because more is known regard­ Buhbles of this gas, trapped in the dough, ing the physical and chemical nature of the lIIake it porous, and this porosity becomes starch granule. The time would seem to be more pronounced when the oven's heat first ripe for the reappraisal of the r(lle of starch in expands the bubbles and then sets their walls, Hour, dough, and bread.' It is to such an ap­ as the white df egg in an omelet souffie is coag­ praisal that this study is devoted. ulated and solidified in the oven. Thus dough THE ROLE OF STAHCH IN GENERAL a(~quires the solidity and texture characteris­ tic of bread. Wheat and rye are the only The structure of starch and {lour.-Viewed erop plants from which through a suitable micro­ bread leavened in this way scope, wheat starch is seen can be baked, for only CONTENTS to consist of small discrete wheat Hours, and to a granules which range from PAGE 2 lesser extent those from The Role of Starch in General 229 about two microns to a rye, form tough and elastic Consequences of Variability little over 40 microns in doughs. If mixtures of in the Nature of Starch . .. 21ft diameter. The larger ones !lour and water did not Summary...... 252 are lenticular, the small form doughs but merely ones spherical or, less halters in which bubbles oft en, pol y g 0 n a I. Co m­ would grow, expand, rise to the surface, and pound granules, i.e., aggregates of two to 25 burst, a porous loaf would not be formed, granules, also occur. The granules do not ap­ but rather a batter cake or a cracker. pear clear and homogeneous but consist of The primary role that the toughness and more or less concentric layers or rings of alter­ elasticity of wheat-flour doughs play in bread nating more or less diffracting material so that making has centered the attention of students they have a laminated appearance. upon them. Because these properties depend Starch is the only constituent of wheat that upon the constituent of flour known as gluten, consists of discrete particles or granules. and only indirectly upon other factors, an Gluten, the other principal constituent, is enormous amount of scientific study has been present as a more or less continuous matrix in devoted to gluten, while the other constituents which the starch granules are imbedded like of Hour have received far less attention. Al­ raisins in a cake.3 However, unlike raisins, though over 70 per cent by weight of flour is the starch granules are generally imbedded in starch, its role as a factor determining the the matrix not singly but in groups. In difl'er­ quality of Hour and the properties of bread ent kinds of wheat the starch bears different has been regarded by cereal chemists as minor. relations to the gluten and this, in a measure, They usually treat starch as if it were merely determines the gross appearance of the ker­ a filler and a food for the yeast, though an im­ nels. With a sufficiently high ratio of gluten portant contributor to the food value of bread. to starch, the starch and the gluten are ce- Their attitude is to be explained not merely hy the preponderant role played by gluten in 1 Cf. C. H. Bailey, TIle Chemistry of Wheat Flour determining the character of bread, but also (New York, Chemical Catalog Company, 1!/25), p. 287; hy the widespread but erroneous belief that all M. P. Neumann, Brolgetreide lind Brot (Berlin, Parey, 192:1), p. 210. slarches are alike, and that therefore the vari­ 2 One micron (the symhol is 110) = 111000 milli­ ability of flours does not depend in an impor­ meter; 25.4 millimeters = 1 inch. tant degree upon variability in the properties 3 E. VerschalTeIt and E. vun Teutel11, "Die Anderung of their starches. Yet starches do vary, and del' mikroskopischen Struktur des Brotes beim Alt­ backcnwcrdcn," Zeitschrift fur physiologische Chemie, starch, in addition to serving as a filler, has October 1915, XCV, 1:10-:15. WHEAT STUDIES, Vol. XI, No.6, February 1935 [ 229 ] 230 STARCH AND FLOUR QUALITY lllented togethcr solidly as the grain dries out different proportions of gluten particles of dif­ in ripening and the kernels appear vitreous or ferent sizes and that, if separated into frac­ 1linty; while in the absence of sufIicient gluten, tions by sifting, soft flours will yield larger air spaces appear rendering the grain soft and fractions passing through the finer-meshed also serving as light-refracting surfaces that sieves than hard-wheat flours. make the grain appear opaque. A grain of It is obvious, furthermore, that the propor­ hard 'wheat may be made to lose its vitreous tion of free unimbedded starch granules must appearance by soaking it in water till it begins depend upon the granulation or "dress" of thc to swell and then drying il,l This causes air flour. Flours from hard vitreous wheat tend spaces to appear within the berry. A true soft to be coarsely granular because of the nature wheat naturally has air spaces diffusely scat­ of their gluten unless the miller deliberately tered through the interior of the berry, i.e., grinds them fine. Flours from soft wheats tend through the endosperm.2 When grown under to be finely granular without special grinding. unfavorable conditions, normally vitreous By grinding flour fine, gluten clumps are wheat varieties may assume the appearance of broken up, the starch set free, and some of the soft wheat in variable degree. The opaque soft starch granules injured and even shattered. part may be localized in a part of the berry or The significance of the state of the starch in in spots. The air spaces to which this appear­ flour will appear as the role of starch in bread ance is due are then localized in these areas; making is unfolded. they are not diffusely distributed throughout The distribution of starch and diastase in the endosperm as in a genuine soft wheat. wheat and flour.-Starch is found in greatest It is obvious that a wheat with such air proportion in the interior of the wheat berry spaces, a soft wheat, is more friable and must and this proportion decreases from within break up into smaller particles in the milling outward. There is also a tendency for the process than a hard vitreous wheat with few starch to be more abundant in the immediate or no air spaces. In consequence, the starch is neighborhood of the germ or embryo.s It fol­ distributed differently in strong and in weak lows that the interior of the berry is poorer in flours. In a hard-wheat flour, the gluten is gluten and richer in starch than those portions present in relatively large clumps in which the of the berry that lie closer to the bran coat. greater part of the starch is imbedded. Rela­ Diastase4 is a chemical agent occurring tively few starch granules lie free. In soft­ naturally in the wheat berry. Under favorable wheat flour, the gluten clumps are smaller and conditions it attacks starch and converts it carry imbedded in them correspondingly less into gum-like substances known as dextrines starch; relatively more of the granules are free. or into fermentable sugar which is known as It is obvious that the two types of flour contain maltose or malt sugar. It is formed only by living things and belongs to the class of agents

1 Cf. W. Johannsen, "Devcloppement et constitution known as enzymes. Its chemical nature is not de l'endosperme de l'orge," Resume du comple-rendu as yet at all well known. It can be recognized des trauaux du laboratoire de Carlsberg, 1883, II, 60-77. only by its effects, of which the one already 2 The endosperm is the main part of the wheat berry from which flour is formed by separating it from the mentioned, the formation of sugar from starch, seed coats, aleurone layer, and the germ. is the most striking. It diffuses very slowly 3 The germ or embryo is that part of the berry from through solutions and cannot pass through which in germination the wheat plant develops. The such biological membranes as surround cells. rest of the seed furnishes the food upon which the embryo is nourished until the young plantlet develops Its commonest commercial source is barley functioning roots. The embryo is a small body com­ malt. posing about 1.5 to 2.0 per cent of the wheat seed Diastase is present in the several organs of located at the end of the berry opposite to the end which is armed with the tuft of hairs. In making white barley in different amounts. It is most abun­ flour hy the roller-milling process, most of the germ dant in the scutellum of the germ or embryo, is removed and appears in the offal, to the feed value and in the aleurone layer lying directly undcr of which it contributes markedly. the bran coat, and also occurs in minimal 4 Some modern investigators use the word "amylase" instead of "diastase." amounts in the endosperm which forms the THE ROLE OF STARCH IN GENERAL 231 bulk of the interior of the seed and contains Since modern roller milling is a process of most of the starch.1 While information con­ gradual reduction of the wheat berry to flour cerning the localization of diastase in the and offal, the flour separated at different stages wheat berry is far less complete, what there is of the roller process is derived preponderantly of it is in full accord with the observations on from different portions of the berry. That ob­ barley, so that it may safely be assumed, pend­ tained from the interior of the berry in the ing evidence to the contrary, that in this re­ earlier stages of reduction would naturally spect wheat is like barley. Martin2 found tend to contain somewhat more starch and that in the wheat berry diastatic activity in­ less diastase than that obtained at later stages, creases progressively from the interior to the and indeed this is what analyses of the dif­ exterior of the berry. In the inner endosperm, ferent mill streams indicate.4 The different that is the endosperm without any traces of streams are of different quality. Moreover, the aleurone layer, the amount of diastase is flours of different fineness obtained by sepa­ too small for any appreciable conversion of ration through sieves with different size the starch. Moreover, the diastase of the inner meshes differ in quality." To what extent endosperm, unlike that of the embryo and of varying starch and diastase content is respon­ the aleurone layer, does not form appreciable sible for varying quality will be discussed amounts of fermentable sugar.3 below. Absorption of moisture by starch.-Starch 1 F. Stoward, "A Hesearch into the Amyloclastic is very hygroscopic, that is to say, it absorbs Secretory Capacities of the Embryo and Aleurone moisture with the greatest avidity. Starch as Layer of Hordeum with Special Reference to the Ques­ tion of the Vitality and Auto-depletion of the Endo­ it comes on the market, although dry to the sperm," Annals of Botany, July 1911, XXV, 799-841. touch, nevertheless contains apprecial?le 2 F. J. Martin, "The Distribution of Enzymes and amounts of moisture-how much depends Proteins in the Endosperm of the Wheat Berry," Jour­ upon the species of plant from which the nal of the Society of Chemical Industry (London), De­ cember 15, 1920, XXXIX, 327-28T. starch is derived, upon the temperature, and 3 H. P. Wijsman, Jr., "La diastase considen\e comme upon the humidity of the atmosphere. Newton un melange de maltase et de dextrinase," Recueil and Cook6 have shown by direct measurement des trauaux chimiques des Pays-Bas, 1890, IX, 1-13; Stoward, op. cit. that wheat starch suspended in cool water 4 Bailey, op. cit.; H. Kalning and A. Schleimer, "Die binds about 30 per cent of it. Rodewald,? using chemische Zusammensetzung des Weizens und seiner a different method, found that wheat starch Mahlprodukte," Zeitschrift fiir das gesamte Getreide­ placed in an atmosphere saturated with water wesen, 1913, V, 199, cited in C. Oppenheimer, Die Fer­ mente und Ihre Wir!wngen (Leipzig, Thieme, fifth edi­ vapor at 21.5° C. binds about 36 per cent of tion, 1929), Vol. IV, Part 2, p. 304; M. P. Neuman, water. This water is in some manner loosely "Brot," in Ergiinzzzngswerle zu Muspratt's EnzylelO­ held by the starch. It can be removed again podischem Handbucl! der Technischen Ch em ie, edited by B. Neuman, A. Binz, and F. Hayduck (Braun­ by exposure to an atmosphere free from water schweig, Vieweg, 1915), Vol. IV, Part I, pp. 343-96. vapor. The last traces can be expelled only 5 H. Nuret and A. v. Ugrimoff, "Einige Versuche tiber with extreme difficulty. To dry the starch die Kornung oder Grosze der Mehlpartikelchen der completely, it must be heated in hot dry air, Weizenmehle," Zeitschrift fiir das gesamte Getreide­ Miihlen- und Biicleereiwesen, April 1933, XX, 95-98; for in hot undried air it retains some moisture ,J. H. Shollenberger and D. A. Coleman, Influence of most tenaciously.s Granulation on Chemical Composition and Baleing The absorption of water by starch at ordi­ Quality of Flour (U.S. Department of Agriculture, De­ partment Bulletin 1463), December 1926. nary temperatures is not accompanied by any 6 R. Newton and W. H. Cook, "The Bound Water of visible change in structure except a moderate Wheat-Flour Suspensions," Canadian Journal of Re­ degree of swelling (see below). None of the search, December 1930, III, 560-78. starch placed in cool water dissolves. It is 7 H. Rodewald, Untersuchungen iiber die Quellung der Starlee (Kiel and Leipzig, Lipsius und Tischer, otherwise when the granules are caused to 1896), p. 69. swell or gelatinize by heating them in water. 8 L. Maquenne, "Sur la dessication absolue des ma­ They then swell till ultimately they take on tieres vegetales," Comptes rendus hebdomadaires des seances de [,Academie des Sciences, OCtober 16, 1905, the appearance of tiny translucent globular GXLI, 609-12. sacks filled with liquid. For the most part, the 232 STARCH AND FLOUR QUALITY

swollen granules of wheat starch do not dis­ zation starch is capable of absorbing large integraLe and hurst. In the process of swell­ (JuanliLies of water-whether as water of ing, a considerable portion of thc starch suh­ hydration or as water less firmly held. stance di/ruscs out of thc granulc and dissolves A bsorption of water a criterion of flour in the surrounding water. In swclling, whcn quality.-The amount of water that can he heated in water, slarch absorhs morc water ineorporated in Hour without producing than it docs in water without heating. Wheat "slackness" of the dough, i.e., too great soft­ starch bchavcs in this way at a tcmperalurc ness, is one of lhe critcria the commercial well below that of hoiling water. haker uses to evaluate flour. Plainly the more If the quanlity of water in which the starch waLeI' he can use, the greater the number of is heated is not too great, the healed suspen­ loaves he obtains per harrcl and the greater sion sets to a more or less stifl" jelly or gel on his profits, for the amount of water remaining cooling. It is a solid, starch paste; the water in the loaf as it leaves the oven depends in part has disappcarcd as such. But it does not neces­ upon the amount of water in the dough. sarily follow that the gel contains more water Therefore, other things being equal, the baker of hydration than ungclatinizcd starch, for, if prefers that Hour which absorbs the greatest such a gel he ground in a pehhle mill or othcr­ all10unt of water. Indeed, Hour strength has wisc disintegrated, it hecomes much less still', sometimes been defined as capacity to absorb may indeed hc convertcd into a sirup.! "E-xami­ water.B nation with the microscope shows that the tiny This important trait of flour, absorption, is swollen slarch sad,s have heen ruptured and commonly thought to be related to the gluten lheir more fluid contents relcased. Ohviously alone without consideration of the role of starch pastc does not contain all of its watcr as starch. Yet, as we have seen, starch absorbs water of hydration; some of it is held mechan­ about one-third of its weight of water. Since ically hy the peculiar structure of the paste. starch constitutes nearly four-fifths by weight It contains water in at least two forms, namely, of nour, it is ohvious that in doughing nour water of hydration c10scly associatcd with or the starch plays a major role in absorbing and bound to the starch molecules, and water holding the water added. How does it com­ mercly mechanically held. Whether in the pare with the other constituents of flour? proccss of gelatinization the starch molecules The distribution of moisture in dough.­ acquire a greater capacity to bind water is not The only other flour constituent present in known. It is not improhahle, for Dumanski" large enough amount to be a factor in absorp­ has rcported that boiled potatoes have a water­ tion is gluten, a mixture of proteins, which, as binding capacity of ahout 1.5 grams per gram we havc seen, gives to wheat-flour doughs their of dry material. Since hoiled potatoes are prc­ main characteristics of elasticity and tough­ dominantly starch, this is suggestive of in­ ness. SIwvholl4 has based estimates of the dis­ crease in hydration capacity resulting from tribution of water in dough upon his own de­ hoiling. In any event in the process of gclatini- terminations of the total bound water in dough and upon the hydration determinations upon 1 c. L. Abhc.·g and E. E. Perry, "Thc Effect of Grind­ starch of Newton and Cook," assuming that ing upon Starch and Starch Pastes," Proceedings of llle their figure of 30 per cent may be applied with­ Socielll for Experimental Bio(o(1Il (lnd Medicine, Octo­ he.· 1 !J24, XXII, 60-61. out serious error to all the nonprotein sub­ 2 A. Dumansld, "Die Bestimmung del' Menge des stances of dough. The average of his rcsults gchundcncn Wass(,l"s in dispersen Systemen. I Methode obtained with three difl'erent Hours was as fol­ der Hefruktometrie und Polarimet.·ie," J(o{[oid-Zeif­ scbriff, Novcmbe.· 1!);J3, LXV, 178-84. lows: Of the water present in the doughs, 43.5 a F. B. Guthrie, "The Absorption of Water by the per cent on average was bound by the dough Gluten of Different Wheats," Af/riclI{[lIra( Gazelle of suhstances. Of this bound moisture or water New SrJlltll Wale.~, Septembcr 18!J6, VII, 583-89. of hydration, the fraction held by the nonpro­ 10. Skovholt, Effect of P(a.~licitll and Related Physi­ tein substances, predominantly starch, was cal I'r()pertie.~ of llle Substratum on Ellzllme Activit/} (Dissertation, Minnesota, 1 !J;J4). 23.3 per cent of all the water present. The 1\ Op. cil. fraction held by the proteins was 20.2 per cent TIlE ROLE OF STARCH IN GENERAL 233 of all the water in the dough. In other words A positivc answcr to this question cannot lhe sLarch and other nonprotein substances as yct hc givcn, but a number of facts arc held somewhat more water bound as water of known which point to an answcr consistcnt hydration than the proteins. Starch is there­ with prescnt knowledge. fore at least as important in determining Gluten is a solid Lough jclly or gcl. Locba hydration as gluten. According to the calcu­ has suggested that such substances have a lations of Skovholt the average hydration of definite struclure and that watcr is "en­ lhe proteins in the three flours studied was trapped" in thc intersliccs of this structurc. 19G per cent. Though gluten has far greater The structurc itself has been likened to a avidity for water than starch-close to 200 per "brush heap" made of interwoven aggrcgates eent as against 30 per cent-the proportion of of molecules betwecn which watcr is hcld en­ slarch in flour is so much greater than that of trapped. This hypothesis of thc naturc of gels gluten that the starch holds about as much is consistent with thc behavior of such jellies waleI' as the gluten or even a little more. when subjected to mechanical manipulation. The determinations by Newton and Cook1 McBain1 has shown that shearing jellies of upon suspensions of nour in water and by Lhis charactcr rcduces their viscosity, a phe­ Skovholt upon doughs of the amounts of water nomenon that is best explaincd by Lhe break­ hound by gluten are in fact inferences. These ing up of the aggrcgatcs and the resulting investigators actually determined the amounts freeing of "cntrappcd" water. of water lhat were made incapable of serving Morcover, Alsbcrg and Griffing" have shown as solvent for other substances, and this water, that severely overgrinding a Hour changes the removed as it were from the scene, they term character of its gluten. Thc gluten washed hound moisture or water of hydration. After from such a Hour swells lcss in dilutc acid subtraction of this amount there remains over than the control gluten preparcd from the !iO pCI' cent which is still available as solvent. same flour beforc it was subjected to over­ But this water is not free as is the water in the grinding. Thc difIerence was so dcfinite that pores of a coarsc spongc. Skovholtz was un­ it was possiblc by mcre inspection to dis­ ahle to express it by means of a hydraulic tinguish the two kinds of swollen gluten. The prcss under a pressure of 45 kilograms per control gluten, as it swelled in dilute acid, be­ square centimcier. Yet whcn dough is dried, came translucent, shiny, and runny; while a great part of the water evaporates with ease. gluten from the samc flour overground re~ Gradually, however, it evaporates less and less maincd morc opaquc, swelled less, and was easily as the moisturc content of the drying much firmer to the touch. This phenomenon dough becomes less, until the final fractions is best explaincd on the assumption that severe are removed only with thc greatest difficulty. and long-continued grinding breaks down thc In short, ovcr half the walcr in doughs behaves structure of gluten so that it is less capable of likc ordinary liquid water so far as its solvent holding walcr "entrapped" in its interstices. powcr and abiliLy to form steam or vapor arc The manner in which glutcn is laid down involved. Yet this fraction cannot rcadily bc during growth in the endosperm of the wheat scparated from dough by mcchanical means. berry and the manner in which dough is made, How is it held 1 kneading and mixing, presumably producc interstices of variable dimensions; in othcr 111lid. words, dough contains pores of variable sizc. ~ OfJ. cit. :t .1. Locb, Pl'oieins (lnd tile Theol'/! of Colloidal Be­ In these, what wc have termed the "entrapped" iWlJiol' (New Yori{, McGraw-Hill, first cdition, 1922). water is containcd. This, as we have seen, 4.1. W. McBain, "Thc Apparcnt Viscosity of Colloidal amounts to over 50 pCI' ccnt of all the water in Solutions lind a Theory of Neutral Colloids as Solvated Micelles Capable of Aggrcgation," .TollI'1l(l1 of Pllysica/ dough and, unlikc water of hydration, is avail­ Chcmisll'!/, February 1!)26, XXX, 239-47. able as a solvent for all those elements of nour r, C. L. Alsberg and E. P. Gl'ifling, "The Errect of Dry which, unlike gluten and starch, arc soluble Grinding UPOll Gels," I'l'oceeciin(1s of lIre Society for K'rperimenia/ Biolo(]lJ (lnd Medicine, N ovcmbcr 1925, in water. Though the total proportion of these XXIII, 142-43. is small, their numbcr is large, for Hour is cs- 234 STARCH AND FLOUR QUALITY sentially crushed modified vegetable cells and water, would naturally tend to hasten absorp­ therefore contains small amounts of a host of tion and therefore developmen,t of the gluten. substances usually found in the tissues of Large gluten particles would take more time plants. Sugars, organic and inorganic salts, to develop than small ones. We should there­ diastase, and non-gluten proteins are among fore expect flours of fine granulation to absorb those significant in the present connection. water more quickly than coarser ones,2 and In so far as these substances occur in mole­ such is in fact the case. Since hard-wheat cules or aggregates of molecules of small size flours as a rule are more coarsely granulated -for example, salts and sugars-they are able than soft-wheat flours, and since hard-wheat to dissolve in all the "entrapped" water and flours contain more gluten than those made diffuse into the smallest pores, for it is well from soft wheat, we should expect hard­ known that such substances diffuse through wheat-flour doughs to take more time to de­ jellies and gels like gelatin almost as fast as velop than doughs from soft-wheat flours, and through water. On the contrary, those flour this indeed is what is observed. constituents that exist in large molecules or It is obvious, therefore, that kneading large aggregates of molecules-for example, speeds up gluten development by mixing so as the non-gluten proteins and diastase-though to bring the gluten particles more rapidly into they are water soluble, probably cannot diffuse contact with water. If gluten has a "brush­ into the gluten, for it is well known that such heap" structure, then perhaps kneading has substances do not diffuse readily into gels. In another function, namely, to facilitate the consequence, there must be a very uneven dis­ drawing of water into the interstices of the tribution of the water-soluble substances in "brush heap." Water enters these interstices dough. Those with small molecules-salts, by capillarity; but the forces of capillarity are sugars, etc. - are fairly evenly distributed surface forces. Under the stress of these throughout the dough mass. Those with large forces, the interstices of the "brush heap" en­ molecules-proteins, diastase, etc.-are prob­ deavor to assume that shape which reduces ably limited to the coarser interstices, namely, the stresses arising from surface forces to a the surfaces of the gluten masses and the fis­ minimum. This shape is the sphere. Knead- sures and grosser spaces of the dough. They must be in fairly concentrated solution, and 1 It may be that the distribution of diastase in one of the effects of kneading is to distribute dough is to some extent independent of the water available for its solution. Diastase is adsorbed by their solution more widely through the dough starch, i.e., it combines loosely with it ecf. H. C. Sher­ mass.1 man, M. L. Caldwell and M. Adams, "Enzyme Purifica­ The relation of kneading to absorption by tion by Adsorption: An Investigation of Pancreatic Amylase," Journal of the American Chemical Society, flour.-The hypothesis that gluten has a November 1926, XLVIII, 2947-56; and "Further Experi­ "brush-heap" structure may explain the ef­ ments upon the Purification of Pancreatic Amylase," fects of kneading and overkneading. It is a Proceedings of the Society for Experimental Biology and Medicine, March 1926, XXIII, 413-15). It would well-known fact that a certain amount of time therefore tend to be removed from the "free" water in must elapse before a dough achieves its maxi­ which it is dissolved by adsorption on starch granules. mum of toughness and elasticity; the phe­ This would occur more in the case of a flour with many free starch granules than in a flour with few of them. nomenon is known as the "development" of Starch granules apparently have considerable adsorb­ the gluten. Kneading facilitates, and tends to ing power, for they adsorb salts readily (H. Lloyd, hasten, development. Many suggestions in ex­ "The Adsorption of Some Substances by Starches," Journal of the American Chemical Society, July 1911, planation of why and how gluten "develops" XXXIII, 1213-26). J. G. Malloch ("Studies on the Re­ have been offered; but the phenomenon is sistance of Wheat Starch to Diastatic Action," Cana­ probably quite simple. It takes time for water dian Journal of Research, July 1929, I, 110-47) found that starch prepared from flour by washing still retains to diffuse into the particles of gluten and give some diastase. to it the dough characters of toughness, elas­ 2 J. H. Shollenberger, W. K. Marshall, and J. F. ticity, and adhesiveness. Kneading, by bring­ Hayes, "Influence of the Size of Flour Particles on ing the gluten particles, as well as free starch Baking Quality," National Miller, December 1921, pp. 29-31 and 66; cited in Shollenberger and Coleman, granules, into more intimate contact with 0p. cit., p. 2. THE ROLE OF STARCH IN GENERAL 235 ing distorts this structure, the interstices are not the case when bakers scald a part of the J1attened and pulled out, and their surfaces are flour before working it into the dough, as they increased. Thereby the surface energy on the do, for example, in making barm for the internal surfaces of the interstices is in­ Scotch bread process. There is, then, an ap­ creased. Therefore, when they endeavor to preciable amount of starch dissolved in the approach again the spherical shape, force is "free" water, the water-binding power of the exerted with which water is drawn into the dough is increased, and the bread keeps fresh fine meshes of the dough. The gluten is not longer.4 The addition of boiled potato to dough merely hydrated, but water is also "en­ serves similar ends. trapped." The considerations above presented are Further support is given to the "brush­ based on the assumption that both the gluten heap" structure hypothesis by the fact that and the starch become fully hydrated and excessive kneading tends in the end to reduce "entrap" the maximum of water. There is no the toughness and elasticity of dough. It may experimental evidence that this assumption is be assumed that overkneading tends to break warranted. It may be that neither starch nor up the "brush-heap" structure of gluten, just gluten is fully saturated when kneading ceases as grinding breaks up the structure of starch or that only one of them is. Indeed, Mohs has pastes,1 and shearing breaks up the structure suggested that added water does not penetrate of soap jellies.2 Consistent with this hypothe­ the gluten particles of excessively strong sis is the well-known loss of viscosity which flours, but is absorbed on their surface. In flour-water suspensions undergo in time. ordinary flours, on the contrary, part of the Blair, Watts, and Denham3 have attributed water, Mohs assumes, penetrates the particles this time effect to syneresis. Syneresis is the of gluten causing them to swell, while when all sweating-out of liquid by a jelly after a time, the water penetrates the particles and none re­ for example, the separation of whey from milk mains on their surface the gluten is charac­ coagulated with rennet or of serum from a terized by being weak. 5 blood clot. Probably the effect of mechanical Since it takes time to develop gluten, it may treatment (kneading, shearing, shaking, well be that those starch granules which are grinding) and syneresis go back to the same imbedded within gluten masses are unable to phenomenon. In the case of the separation of become completely saturated-at least not fluid from a gel by mechanical treatment, we until after development is ended. Water must have release of "entrapped" liquid by mechan­ diffuse through a layer of gluten before it can ical rupture of the structure. In the case of reach such granules. It is therefore quite pos­ syneresis, we have the release of "entrapped" sible that in the ordinary process of kneading liquid because the particles aggregate into many of the starch granules of a strong flour coarser clumps, so that the structure becomes never become fully saturated with water or less fine-meshed and in consequence not cap­ else become so only gradually as fermentation able of holding so much water. progresses, whereas in weak-flour doughs In doughs made from normal flours, there is most of the starch grains become fully satu­ practically no starch dissolved in the "en­ rated almost immediately. Possibly one of the trapped" water; for whole, intact, unheated modes of action of the high-speed mixer is to starch granules are quite insoluble. This is hasten the saturation of the starch granules and thus to contribute to the more rapid ab­ 1 Alsberg and Perry, "The Effect of Grinding upon sorption effected by these machines. At the Starch and Starch Pastes." same time friction during kneading produces 2 McBain, op. cit. heat and warms up the dough. Raising the 3 G. W. S. Blair, G. Watts, and H. J. Denham, "Effect of Concentration on Viscosity of Flour Suspensions," temperature of gluten up to a certain point Cereal Chemistry, January 1927, IV, 63-67. increases its absorption power (see below). 4 Neumann, Brotgeireide und Brat, p. 432. Possibly it is in imperfect hydration and uK. Mohs, Neue Er1cenntnisse aut dem Gebiete der development that the explanation of lack of J}~iillerei und Biic1cerei (Dresden and Leipzig, 1922); cited in Bailey, op. cit. direct proportionality between gluten content 236 STARCH AND FLOUR QUALITY and absorption is to be sought. With rising the much greater hydration of soft-wheat gluten content of flour, one would expect ab­ flour is due to finer granulation and conse­ sorption to rise faster than gluten content, quent more perfect hydration, for Newton because each unit of gluten replacing a unit and Cook2 found the same hydration capacity of starch has greater water-absorption ca­ for the gluten of strong and of weak flours. pacity than the starch. The question does They experimented under conditions favor­ not seem to have been tested directly by exact able to perfect hydration. They concluded measurements relating gluten percentage to that the greater water-absorption capacity of absorption, but experiments on fermentation strong flours depends not on greater hydration rates and baking tests (see below) indicate of the gluten but rather upon the nature of that absorption rises less rapidly than gluten its "brush-heap" structure which causes it to percentage. "entrap" more water than the gluten of a It has generally and tacitly been assumed weak flour. that, so far as absorption is concerned, the We see, then, that physical conditions are of only variable is gluten, and lack of propor­ great importance in determining absorption; tionality between gluten content and absorp­ and it is not unreasonable to surmise that the tion has been attributed to some as yet un­ location of the starch granules, whether free or known change of the character of the gluten imbedded, may well be an important factor in as its proportion in flour increases. This is determining absorption. When a starch gran­ possible, but it is here suggested that the ule is imbedded in a gluten particle, it will starch may also playa role. As above pointed compete with the gluten of that particle for the out, in high-gluten flours, fewer starch gran­ water, as the water passes into the interior ules are free than in low-gluten flours. More of the gluten particle where the starch granule of their starch is but slowly accessible to lies. The tendency would be for the hydra­ hydrating water than in a soft low-gluten tion and consequent development of the glu­ flour. Moreover, free starch granules may con­ ten to be slowed up, and in fact, as we have ceivably serve as centers for the formation of seen, strong flours tend to develop more tiny pockets in the dough in which water is slowly than weak ones. In any event, it is enclosed mechanically. Perhaps in hard flours obvious that the quantity of starch in a flour, with few free granules but little water is in­ and probably its location, whether free or corporated mechanically in this way. imbedded, play a large role in determining The only estimations of the degree of hy­ absorption. dration of the gluten in doughs of different Fermentation.-Normal flour contains very kinds of flours are those of SkovhoW above little fermentable sugar, but sugar must be discussed. Two of the flours he studied were present in dough if it is to ferment, because hard-wheat bread flours and therefore, pre­ yeast is only able to ferment certain sugars, sumably, coarsely granular; one of them was i.e., to form carbonic-acid gas from them. a soft-wheat cake flour and therefore, pre­ Yeast is quite incapable of fermenting starch. sumably, finely granular. The cake-flour glu­ As soon as yeast is incorporated in dough, ten when doughed was hydrated 277 per cent, it begins to ferment the small amount of pre­ whereas the bread flour glutens were hy­ formed sugar present, and fermentation would drated 179 and 183 per cent, respectively. It cease as soon as this small amount of sugar may be that these differences are merely inci­ had been consumed, if fresh sugar were not dental, for the methods used involve an ap­ supplied. Leavening would remain incom­ preciable and undetermined margin of error. plete, unless in some way fresh supplies of If it should turn out on further study that fermentable sugar were offered to the yeast. such a difference between the two types of In most flours, such fresh supplies are of­ flour in fact is normal, it would suggest that fered through the action of diastase on starch. Hence the action of diastase upon the starch lOp. cit., Table XVI. in flour is very important; indeed, it is one 20p. cit. of the factors determining the baking quality THE ROLE OF STARCH IN GENERAL 237 of flour. The ability of a flour to form sugar enmeshed starch granules by mechanical in this way is known as its diastatic power. means, if it is to attack them. While, there­ In order that diastase may form sugar fore, there is some evidence that the location from starch, it must be able to get at the of the starch in flour and dough is a factor granules. As we have seen, diastase diffuses in sugar formation, so many other factors are slowly through a solution and probably en­ involved, as we shall see, that it is not justi­ ters the gluten masses slowly if at all.1 Such fiable, for the present, to say more than that starch granules as are imbedded in gluten the degree to which the starch is separated matrix presumably are still surrounded by from the gluten is one of the factors determin­ the protoplasmic layer within which they were ing the diastatic power of flours. formed, and diastase from without passes With the formation of gas in the dough by through this layer with difficulty if at all.2 fermentation, the dough mass is puffed up, Such granules are therefore inaccessible to its volume increases, the dough "rises." This diastase from without or accessible only with "rise" depends in part upon the rate of fer­ difficulty. Starch grains which have been mentation, which in turn depends upon the freed from the matrix, it is to be expected, diastatic power of the flour and upon the will have had these protoplasmic layers rup­ activity and quantity of yeast. How the dia­ tured or removed. One would therefore ex­ static power is related to the "freedom" of the pect flour with many free granules to show starch has been pointed out. But the rise greater diastatic action than flour with few. depends also upon the elasticity and tenacity Such evidence as there is points in this direc­ of the gluten which keep gas bubbles "en­ tion. Thus Malloch3 found that any treat­ trapped" and do not permit them to escape. ment which made a flour finer (fine grind­ As we have seen, absorption of flours in­ ing, extraction with ether) without injuring creases with their gluten content but less (see below) many of the granules increased than proportionately. This seems also to be diastatic power. true of the rise of the dough and consequently Furthermore, very thorough mixing tends of the volume of the baked loaf. As the gluten to increase the rate of sugar formation at the content becomes greater, each increment of beginning of doughing. Thus SkovhoW found gluten causes a smaller increment in loaf a considerable increase of sugar formation in volume than the increment preceding. Each doughs mixed for ten minutes over the same increment is progressively less effective in dough mixed but one minute. One is tempted increasing the size of the loaf. Why this to infer that this is because the mixing has should be has not as yet been explained sat­ brought more starch granules in contact with isfactorily. This lack of direct proportionality diastase, for, as we have seen, diastase having between gluten content and absorption has molecules of large dimension does not diffuse been attributed to some as yet unknown modi­ readily into the interior of gluten masses. It fication of the character of the gluten as its needs to be brought in contact with gluten- proportion increases. This is entirely probable; but it is also 1 Some investigators have suggested that diastase occurs in flour in a form that is not freely soluble, possible that the starch may playa role as well perhaps loosely combined with protein. J. S. Ford and just as it may playa role in absorption. The J. T. Guthrie ("The Amylolitic and Proteolytic Fer­ greater the gluten content, the less the tend­ ments of Wheaten Flours, and Their Relation to 'Bak­ ing Value,''' Journal of the Society of Chemical In­ ency of the flour to be very finely granulated, dustry [London], April 30, 1908, XXVII, 389-93; cf. also for, as has been pointed out, the endosperm Bailey, op. cit., pp. 233-34) found it difficult to extract of a high-protein wheat is less friable than all the diastase from flour unless they first partially digested the protein with the enzyme, papain. If in fact that of a low-protein wheat. In consequence, the diastase of flour is combined in some manner so it the flour from high-protein wheat, as also is not freely soluble, the starch imbedded in gluten pointed out, contains more of its starch locked matrix may well be quite inaccessible to attack. up in a gluten matrix and less of it free. More 2 H. Ziegenspeck, "Ueber Sparstarke," Botanisches Arc/til), August 1924, VII, 251-73. of its starch is therefore inaccessible to dia­

3 Op. cit. 4 Op. cit., Table XVII. stase as well as slower to hydrate. The con- 238 STARCH AND FLOUR QUALITY

sequences are probably less rapid sugar for­ dilution of the gluten by starch is so great in malion, slower initial rate of fcrmcnlation, such instances that the gluten particles are not and less rapid rise of the dough. afforded an opportunity to agglutinate into a Some investigators have studied the rela­ coherent mass in the washing process."2 Here tion of gluten percentage in flour to flour we see occurring naturally what may he strength by diluting flour with starch and not­ brought about artificially by the addition of ing the effecU The extensibility of the dough starch to a stronger flour. is reduced, its gas-retaining capacity is di­ One would expect the dough from the minished, and the quality of the loaf is im­ starch-Hour mixtures to be less strong me­ paired. Most students of these queslions have chanically, and perhaps this is one of the assumed that starch served merely as a dilu­ reasons why Johnson and Bailey8 found that ent. They have not considered that different doughs made from strong Houts to which wheat starches, in all probability, resist dia­ starch had been added exhibited a decrease in stase to variable degrees (see below). They gas-retaining capacity. have not taken into their calculations the But the results of Johnson and Bailey also fact that added starch is not present in the show that the addition of starch increased the flour-starch mixture in the same state as the total volume of carbon-dioxide gas formed and natural starch of flour. Much of the latter, that the rate of gas formation was more rapid especially in the stronger, high-gluten flours, at the beginning of fermentation than in is locked up in the gluten matrix and there­ doughs to which no starch had been added. fore but slowly accessible to water of hydra­ This is what one would expect from the fact tion and accessible with great difficulty, if that the starch added is free starch accessible at all, to the diastase. Merely for mechanical to immediate hydration and exposed to im­ reasons the effect of the added starch would mediate attack by diastase. be diITerent from that of starch naturally Furthermore, since starch absorbs as much present in the flour. Since the added starch as 30+ per cent of moisture, it follows that is all free, it would become saturated with if the starch-flour mixtures were doughed by water at once and it would interfere with the the same formula, using the same amount of tendency of the gluten particles to stick to­ water as for the unmixed flours, the two sets gether. Mechanically, it must act quite dif­ of doughs would have different distributions ferently from the natural starch of flour, be­ of the water. Since the water capacity of cause this is in great measure already sur­ gluten is greater than that of starch, in the rounded and protected by gluten. "When starch-11our mixtures there would tend to be patent or high grade flours containing an more water bound neither by gluten nor by unusually low percentage of gluten are mixed starch. Such doughs would tend to be into a paste or dough and washed with water, "slacker," and they would therefore tend to it proves impossible at times to recover an bake out somewhat differently. appreciable quantity of wet crude gluten. The Behavior of doughs in the oven.-In the oven, water vapor4 is formed in the dough 1 The literature has been reviewed by Bailey, op. cit., which evaporates into the carbonic-acid gas p. 261. bubbles, thereby aiding materially in the ex­ 2 Bailey, op. cit., p. 257. pansion of the loaf. At the same time both 3 A. H ..Johnson and C. H. Bailey, "Gluten of Flour and Gas Retention of Wheat Flour Doughs," Cereal gluten and starch alter their water-absorption Chemistry, March 1925, II, 95-106. capacity. When gluten, as it occurs in flour, 4 Dough contains small amounts of alcohol which is is moistened, it swells, becomes hydrated, and vaporized at a lower temperature than water. Its va­ in addition takes up water into its finer inter­ por no doubt contributes slightly to the expansion of the loaf in the oven. According to Neumann (Brotge­ stices. When it is warmed in water, it ab­ Ire ide und Brot, p. :l85) fresh bread still contains sorbs slightly more of it5 up to a temperature 0.25-0.4 per cent of alcohol. of 50 0 C. or slightly higher. Above 50° C., G F. W. Upson and J. W. Calvin, "On the Colloidal its behavior is reversed and its swelling power Swelling of Wheat Gluten," Journal of the American Chemical Socieill, May 1915, XXXVII, 1295-1304. becomes progressively less. Heat-coagulated TIIB ROLB OF STARCH IN GENBRAL 239 gluten! and gluten coagulated by treatment Alsberg and Griffingn found that swelling be­ with alcohol-ether followed by drying2 have gins to some extent at even lower tempera­ greatly reduced swelling power, are less hy­ tures. drated,S and hold water less firmly. It follows that, within the range of tem­ Starch also swells when moistened in cool perature over which glulen is losing water water and as it is healed continues to swell capacity, starch is gaining water capacity. at temperatures at which gluten is already Clearly the distribution of water in the baked losing swelling power. Nyman1 observed loaf must be quite difTerent from that in starch to begin to swell at a temperature as dough. low as 50 0 C., while LaWall and Graves" and When the fermented and proofed dough is brought into the oven, diastaLic action must continue for a time, at any rate until the 1 C. L. Alsberg and E. P. Griffing, "The Heat Coagu­ lation of Gluten," Cereal Chemistry, November 1927, temperature in lhe baking loaf reaches about IV, 411-23. 800 C. In a moderate oven a temperature 2 Newton and Cook, op. cit. of 80° C. is reached in five to eight minutes.1 3.J. R. Katz, "Uber das Altbackenwerdcn des Brotes However, the action of diastase cannot be lind die Moglichkeit, diese Veranderung hintanzu­ hulten," Zeitschrift fiir da.~ gesamte Getreide- Miihlen­ very extensive during this period, for very und BiicJcereiwesen, March 1934, XXI, 6:1-68. little of the products of its action is found 4 M. Nyman, "Untersuchungen tiber die Verkleister­ in bread. From bread crumb, Stone8 ex­ ungstemperatur bei Starkekornern," Zeitschrift fiir tracted comparatively little soluble material Unlersuchung der Nahrungs- und Genussmittel, Decem­ her 1, 1912, XXIV, 673-76. derived from starch, while Katz9 found some­ r, C. H. LaWall and S. S. Graves, "Studies in Carbo­ what more in Dutch bread. The starch gran­ hydrates. The Composition and Digestibility of Wheat ules cannot therefore be gelatinized by the Bread and Allied Foods. Gelatinization of Starches," Transactions of the Wagner Free Institute of Science, oven's heat in the same way as when heated 1913, Vol. VII, Part 2, pp. 37-45; cited in A Comprehen­ in an excess of water, for under the latter sive Survey of Starch Chemistry, compiled and edited conditions an appreciahle proportion of the hy R. P. Walton (New York, Chemical Catalog Com­ pany, 1928), Vol. I, Part 2, p. 133. starch granule becomes soluble in water. Cer­

(J "'The Heat Coagulation of Gluten." tainly there is not enough water in dough to 7 L. E. Stout and F. Drosten, "Heat Flow through permit anything like complete gelatinization Bakery Products. I. Time-Temperature Relationships of the starch even if all the water, water of Existing during the Baking of Bread," Industrial and Engineering Chemistry, April 1933, XXV, 428-30. hydration included, were available for that 8 W. E. Stone, The Carbohydrates of Wheat, Maize, purpose. This has been shown definitely by Flour, and Bread and The Action of Enzymic Ferments Jago and Jago,1o by Whymper,l1 and by others. upon Starches of Different Origin (U.S. Department of Some change, nevertheless, takes place in the Agriculture, Office of Experiment Stations Bulletin 34), 1896, p. 28. starch granules, even though there is not 9 J. R. Katz, Het Oudbakken Worden van het Brood enough water available to form genuine starch in Verband met het Vraagsluk van den Naclltarbeid der paste. No unchanged granules remain in or­ Bakkers Cs-Gravenhage, van Langenhuysen, 1917), dinary bread crumb.12 They have an altered Part 2, p. 23. appearance under the microscope, and all the 10 W .•Jago and W. C. Jago, The Technology of Bread­ Making (London, Simpkin, Marshall, Hamilton, Kent, granules may be stained to varying degrees 1911), pp. 81, 428. with Congo red, showing that some alteration II R. Whymper, "Colloid Problems in Bread-Mak­ has occurred.13 Normal intact granules do not ing," in "Colloid Chemistry and Its General and In­ dustrial Applications, Third Report of the Committee take up this dye. In the presence of insuffi­ .... ," British Association for the Advancement of cient moisture to completely gelatinize, heat­ Science, Report of the Eigbty-eighth Meeting, Cardiff- ing probably disorganizes the crystalline struc­ 1920, Auaust 211-28 (London, Murray, 1920), Appendix, pp.61-74. ture (see below) of the grains. They take up

12 VerschaffeIt and van Teutem, op. cit.; A. Maurizio as mueh water as is available and only swell (Die Nahrunasmittel aus Getreide [Berlin, Parey, sec­ to that extent. It is likely that in this condi­ ond edition, 1924], I, 350) observed that large and tion they have acquired greater water capacity small granules behave alike. than they had in the unaltered state. 13 C. L. AIsberg, "Starch in Flour," Cereal Chemistry, November 1927, IV, 485-92. In any event, the changes dough undergoes 240 STARCH AND FLOUR QUALITY

in the oven while being converted into bread Behavior of starch in bread.-This redistri­ cause changes in the water capacities of the bution, however, is probably not permanent, constituents and a fundamental change in for starch jellies change their state. In time, water distribution. Gluten, because coagu­ they show the phenomenon of syneresis (see lated, loses water capacity; starch gains it. above), and Katz! has made this the basis of Presumably this is accompanied by some a hypothesis regarding the manner in which transfer of water from gluten to starch, and bread goes stale. It has long been known that this transfer must depend to some extent upon drying out is not the reason why bread grows the spatial relations of the starch granules to stale; it shows the characteristics of staleness the gluten. Possibly an explanation is here to long before it has lost appreciable amounts of be found for the fact that some types of bread water. Indeed, bread placed in a hermetically grow stale more rapidly than others (see sealed container in which it cannot dry out below). becomes stale about as soon as in air. Probably the starch granules have been con­ The behavior of starch paste upon which verted into an extremely concentrated jelly. Katz bases his hypothesis is sometimes known But many dry jellies-for example, dry gela­ as "retrogradation." In time, the starch in tin-have great avidity for water; they swell pastes undergoes a change: the paste be­ while absorbing it. A concentrated starch comes harder and tougher and water separates jelly, of which the granules after baking are from it. In a system consisting of starch gelat­ probably composed, would be avid for water. inized in so much water that it is merely a In summary, we may say that starch shares cloudy viscous liquid or a very thin jelly, at least equally with gluten in determining much of the starch ultimately settles out leav­ the absorption of flours-possibly it may even ing supernatant liquid which contains some hold more water. In different types of flour, starch substance in solution. French investi­ the starch is in different relations to the glu­ gators interpret these phenomena as indicat­ ten. In hard-wheat flour, nearly all of the ing that a part of the starch which has been starch is surrounded by gluten; very little of changed by the gelatinizing process gradually it is free. Therefore, most of the starch of returns to its original state in the unaltered such flours is more or less protected from the granule. For reasons which it is unnecessary attack of such agents as diastase. In soft­ to detail here, a much more probable expla­ wheat flours, contrary conditions obtain. The nation is that the phenomenon is merely due location of the starch is undoubtedly a factor to change of the state of aggregation of the in determining the rate of fermentable-sugar starch whereby, as explained above, the mole­ formation by diastase in dough and therefore cules clump together in larger aggregates and the initial rate of panary fermentation, but it thereby lose power to hold water.2 If this is not possible to say what is the importance occurs in bread, there must in time again be of this factor relative to the numerous others a redistribution of water in the loaf. There involved in the diastatic powers of flour and is as yet no way to demonstrate the occur­ the fermentation rates of doughs. Finally, in rence of such a redistribution of water by ex­ the oven, starch gains in water-absorbing periment, but it is a fact that as bread goes power, while gluten loses, and a redistribution stale the amount of soluble starch that can of water in the loaf different from that in be extracted from it with water becomes less. dough follows. Furthermore, the change in gelatinized starch which is termed retrogradation takes place more rapidly at lower temperatures than 1 R. J. Katz, "Gelatinization and Retrogradation of Starch in the Bread Staling Process," in Walton, op. at higher ones; and Katz3 has shown that the cit., Part 1, pp. 100-117. rate at which bread grows stale passes through 2 For a fuller discussion of retrogradation the reader a maximum at -2° to -3° C. At the tem­ is referred to Oppenheimer, op. cit., Vol. I, Part 9, pp. perature of liquid air (-185° C.) and also be­ 661 and 681-82. tween +60° C. and +90° C., bread remains S "Gelatinization and Retrogradation of Starch in the Bread Staling Process." fresh for a long time. On this behavior of CONSEQUENCES OF VARIABILITY IN THE NATURE OF STARCH 241 bread, Katz! has based a commercial method than the crumb, and some of the starch is for controlling the rate of growing stale of changed to dextrines, quite hygroscopic sub­ bread. stances. In consequence. after the loaf has Katz has also proposed a method for de­ cooled, the dry crust attracts moisture from termining the degree of staleness of bread the moister crumb beneath it, whereby the based on the observation of Lehmann2 that crust becomes softer and more pliable. fresh bread imbibes more water and swells more than stale bread. The method is as fol­ CONSEQUENCES OF VARIABILITY IN THE NATURE lows: A weighed piece of bread crumb is dis­ OF STARCH integrated in water and the suspension forced Thus far we have assumed that starch is through a sieve. The resulting finer suspen­ the same in all flours. Only differences in the sion is made up to a known volume by adding spatial relation of the starch granules to the water, poured into a graduated cylinder, and gluten matrix have hitherto been considered. allowed to settle. The volume of the sedi­ We have next to determine whether the phys­ ment is noted and compared with that of ical and chemical character of the starch itself fresh bread crumb similarly treated. The is the same in all flours, and, if it is not, to staler the bread, the smaller the volume of what extent the variability of the starch may the sediment as compared with that of fresh affect the behavior of flour in doughing, fer­ bread. mentation, and baking. Whether or not the hypothesis of Katz The properties with respect to which starch proves to be the final explanation of why may be conceived to vary may be grouped into bread grows stale, it seems quite certain that the following categories: in the process starch plays the major role. Variability with respect to chemical com­ Katz's hypothesis applies only to bread position. crumb and not to crust. In fresh bread, the Variability with respect to size and size dis­ crust is crisp and brittle; as bread begins to tribution of granules. grow stale, the crust becomes soft and pliable. The state of the granules-whether intact This is due to an increase in the moisture or injured. content of the crust. In the oven, the crust, Variability with respect to water-holding being at the surface of the loaf, becomes drier capacity. Variability with respect to resistance to 1 Het Oudbakken Worden van het Brood in Verband diastase. met het Vraagstuk van den Nachtarbeid der Bakkers. Variability of starch in chemical composi­ 2 K. B. Lehmann, "Hygienische Studien iiber Mehl­ und Brot. Theil V: Beitrage zur physikalischen Ber­ tion.-Until recent decades the belief pre­ schaffenheit des Brotes," Archiv fur Hygiene, 1894, XXI, vailed that, chemically speaking. all starches 215-67, especially p. 238. are identical. In 1913 Reicherta brought for­ 3 E. T. Reichert, The Differentiation and Specificity of Starches in Relation to Genera, Species, etc. (Car­ ward evidence that different starches reacted negie Institution of Washington Publication 173), differently to various chemical reagents. In 1913. recent years it has been shown by a number 4 There is much confusion in the nomenclature. of investigators that starch is not a single Some authors term Meyer's a-amylose "amylopectin," a term that is objectionable because this substance has chemical individual but that the bulk of the none of the characteristic chemical properties of granule consists of two substances which A. pectins; nor does it serve similar functions. The p­ Meyer termed a-amylose and ~-amylose.4 Be­ amylose of Meyer is often termed simply "amylose." It has seemed best in this study to follow H. C. Sher­ sides these, small amounts of silica, of iron. man and his school in retaining the nomenclature and of hemicellulose have been reported. originally introduced by A. Meyer in 1895, in his Other mineral substances occur accidentally Untersllchungen uber die Starkekorner (Jena, Fischer, 1895) . because of the great adsorptive power of the

5 T. C. Taylor and J. M. Nelson, "Fat Associated with granules. Starch," Journal of the American Chemical Society, Alpha-amylose contains an acid. either August 1920, XLII. 1726-38; H. A. Iddles. Separation of the Amylases in Some Common Starches (Disserta­ phosphoric acid or a fatty acid,5 as a con­ tion, Columbia. 1925). stituent part of its molecule, whereas ~-amy- 242 STARCH AND FLOUR QUALITY lose contains none.1 The a-amylose of wheat position by weak hydrochloric acid, whereas starch is hardly, if at all, soluble in cool water ~-amylose breaks down readily into sugars.3 and contains phosphoric acid; the ~-amylose Furthermore, malt diastase attacks a-amylose is soluble and is free from phosphoric acid.2 more slowly than ~-amylose.4 Alpha-amylose is rather resistant to decom- It has been established that the percentages of a-amylose and of ~-amylose vary from 1 Compare, however, A. R. Ling and D. R. Nanji, species to species.G It may therefore be re­ "Studies on Starch. Part I. The Nature of Polymerised garded as definitely proved that the starches Amylose and Amylopectin," Journal of the Chemical Society (London), Transactions, 1923, CXXIII, 2666-88. of different species have different chemical 2 J. Field II, "Specific Rotation and Phosphate Con­ composition. Iddleso has even found differ­ tent of Cold-Water-Soluble Fractions of Ground Corn ences in the percentages of a-amylose and ~­ and Wheat Starches," Proceedings of the Society for Experimental Biology and Medicine, May 1928, XXV, amylose in different samples of starch of the 711-12. same species (corn, rice, potato). Beta-amylose should not be confused with the "sol­ As yet no direct determinations of the pro­ uble starch," so-called, of chemical literature. This is starch that has acquired solubility by chemical treat­ portions of a-amylose and of ~-amylose in the ment. It is altered starch that does not occur as such starch from different samples of wheat seem in the natural granules. Beta-amylose occurs as such to have been made. ReicherF reached the con­ in the granules, for it may be extracted from them by first shattering the granules and then dissolving in clusion that the starches of different wheats watel·. There is no reason to believe that merely crack­ react very similarly to the reagents he em­ ing granules mechanically produces any chemical ployed but that the starch of emmer, Triticum change. dicoccum, is more sensitive to these reagents 3 T. C. Taylor and H. A. Iddles, "Separation of the Amyloses in Some Common Starches," Industrial and than that of common wheat, T. vulgare. Man­ Engineering Chemistry, July 1926, XVIII, 713-17. gelsB has found that the starches from dif­ 4 H. C. Sherman and .J. C. Baker, "Experiments upon ferent kinds of wheat contain different per­ Starch as Substrate for Enzyme Action," Journal of centages of phosphoric acid and that the the American Chemical Socie!", September 1916, XXXVIlI,1885-1904. same variety may vary in this regard in dif­ G Ling and Nanji (op. cit.), however, have stated ferent seasons. Since the phosphoric acid is that a-amylose and ~-amylose practically always oc­ contained in the a-amylose and not in the ~­ cur in the ratio 2 : 1. Taylor and Iddles found rela­ amylose, it follows that percentage of phos­ tively much more ~-amylose. G Separation of tile Amylases in Some Common phoric acid in wheat starch is an index to the Starclles (Dissertation). percentage of a-amylose it contains.9 Mangels' 7 Op. cit. observations therefore indicate that there is 8 C. E. Mangels, "Varietal and Regional Variations probably more a-amylose (and correspond­ in Properties of Wheat Starches," Cereal Chemistry, October 1934, IX, 571-85. ingly less ~-amylose) in a starch sample with o However, there is some evidence that the a-amy­ a high phosphoric-acid percentage than in lose of different kinds of starch may vary in phos­ those with a low percentage. Wheat starches phoric-acid content, for Samec has reported that in may therefore differ in their chemical com­ different starches all the phosphoric acid is not bound with equal firmness (M. Samec and H. Haerdtl, "Stu­ position. dien tiber Pflanzenkolloide, IX. Zur Kenntnis ver­ This inference is supported by the observa­ schiedener SU'trkearten," Kolloidchemische BeilIefte, tion of Fernbach10 upon potato starch. He Heft 12, 1919, XI, 281-300). Moreover, several investi­ gators have found that extraction with dilute hydro­ washed potato starch with dilute hydrochloric chloric acid removes a variable fraction of the phos­ acid and then separated it into large and small phoric acid from natural starch grains. Furthermore, granules by fractional sedimentation. The there is some evidence that the ~-amyloses from dif­ ferent starches are not identical (cf. .J. J. Lynst Zwik­ small granules contained appreciably greater ker, "L'action des enzymes amylolytiques sur les percentages of phosphoric acid than the larger grains d'amidon naturels et la structure colloldale de ones. This would indicate that small potato­ l'amidon," Recueil des trauaux botaniques neerlandai.~, 1921, XVIII, 1-102). These are questions that await a starch granules contain more a-amylose than final solution. large ones. 10 A Fernbach, "Quelques observations sur la com­ It remains to be determined whether wheat­ position de l'amidon de pommes de terre," Comples starch granules of different sizes contain dif" rendus .hebdon:wdaires des seances de ['Academie des Sciences, February 15, 1904, CXXXVIII, 428-30. ferent proportions of phosphoric acid and, CONSEQUENCES OF VARIABlLI1'Y IN THE NATURE OF STARCH 243 therefore, different proportions of a-amylose. such great variability in the character of their If it turns out that they do, then this fact may starch grains, it is reasonable to expect other well explain some of the differences in the species, such as wheat, also to show analogous hehavior of different wheat starches to be variations. discussed below, for we know that they differ As we have seen, wheat-starch granules vary with respect to granule size and we also know in size; while all sizes occur over the size that a-amylose and ~-amylose possess differ­ range most of them are either large or small. ent chemical and physical properties. Armstrong1 found that the small granules Variability of starch with respect to size range in size from 3 to 5 11 and the largest and size distribution of granules.-Reichertt from 30 to 35 11. Naudain5 divides them into seems first to have claimed that the size, two classes: those less than 7 11 and those shape, and other properties of starch granules more than 7 11 in diameter. The size of the are characters differing from variety to va­ granules and the proportions of small and riety of the same species and heritable. Darbi­ large granules are somewhat different in the shirez and Gregory3 found that smooth-seeded several species of the wheat genus, Triticum, garden peas have simple starch grains, where­ but the differences are not very great except as wrinkled-seeded peas have compound ones, in the case of Triticum monococcum, or Ein­ the component granules of which are smaller korn, which has appreciably smaller granules in size than the simple granules of smooth­ than the common wheats." Buchanan and seeded peas. Intermediate varieties that are Naudain7 found the size of the starch gran­ pitted rather than wrinkled have simple ules different in different varieties of the same grains. If the different horticultural varieties species, and on the whole the strongest wheats of a single species like the garden pea show tended to have the smallest granules. This is quite in accord with the observations of Holz­ lOp. cit. nerS and of ThausingO that vitreous barley Z A. D. Darbishire, "On the Result of Crossing Round kernels contain a relatively smaller number of with Wrinkled Peas, with Especial Reference to Their Starch-Grains," Proceedings of tlle Royal Society (Lon­ large starch granules than soft ones, a fact don), 1908, Series B, LXXX, 122-35. definitely corroborated by Johannsen,lo who 3 R. P. Gregory, "The Seed Characters of Pisum by comparing corresponding portions of the sa/iuum," New PItytologist, December 22, 1903, II, 226- endosperm of hard and soft barley kernels 28. from the same harvest established that hard 4 E. F. Armstrong, "The Chemical Properties of Wheaten Flours," Journal of tIte Board of Agriculture, kernels contain a much larger number of 1910, XVII, Supplement 4, pp. 45-52. small granules than soft ones. One may en­ 5 G. G. Naudain, "Study of the Properties of Wheat counter analogous differences in a batch with Starch and the Baking Qualities of Flour," American only vitreous kernels. Food .Journal, May 1925, XX, 251. It has long been known that the larger fl L. Eynon and J. H. Lane, Starclt. Its Cltemistry, Tec1lno[ogy and Uses (Cambddge, Heffel', 1928), p. 79. granules of any given starch gelatinize at

7 J. H. Buchanan and G. G. Naudain, "Influence of lower temperatures than the smaller ones,ll Starch on Strength of Wheat Flour," Industrial and though some starches exhibit the opposite be­ Engineering Cltemistry, October 1923, XV, 1050-51. havior.12 Nageli long ago observed that the 8 Cited hy Johannsen, op. cit. larger granules of potato starch begin to ge­ ° Ibid. latinize at 55° C., whereas the smallest gran­ lOOp. cit. ules do not begin to gelatinize until a tem­ 11 R. Whymper, "Microscopical Study of Changes 13 Occurring in Starch Granules during Germination of perature of 65° C. is reached. Nyman made Wheat," in SeuentIt International Congress of Applied similar observations and suggested that this Chemistry, London, Mall 27 to June 2, 1909 (London, might be due to the fact reported by Lenz14 Partridge and Cooper, 1910), Section VI A, pp. 7-13. that the membranes of the larger grains are 12 Reichert, op. cit. thinner than those of the smaller ones. The 13 Op. cit. reason may, however, be merely that in swell­ 14 W. Lenz, in Arbeiten ails dem Pltarmazeutisclten lnstitut der Universitat Berlin, 1910, Vol. VII; cited by ing large grains necessarily develop greater NYman, op. cit. internal pressures per unit of granule surface 244 STARCH AND FLOUR QUALITY

than small ones.1 Naudain2 found that the firmed, it may well prove to be that differences larger granules of wheat starch swelled rela­ in the size distribution of the granules ac­ tively much more than the small ones, which count in part for the differences in the vis­ may indicate that internal pressure may have cosities of the pastes of starches from differ­ much to do with the difTerent behavior of ent horticultural varieties of the same species. small and large granules. Naudain also found Grewe and Bailey,8 contrary to Buchanan the small granules to resist swelling much and Naudain9 and to Naudain,lO were unable more than large ones, but his findings were to find any direct correlation between granule based on only two flours. They must there­ size and baking quality. They concluded that fore be regarded as merely suggestive until difTerences in the properties of starch that are confirmed with more samples. related to baking strength of flour or to dia­ Very probably variability in size distribu­ static activity must be attributable to proper­ tion accounts for the difTerences in gelatiniza­ ties of the starch other than those involved in tion temperatures of granules and in the vis­ the size of the granule alone. These conclu­ cosity of pastes that have been observed in sions were based on the measurement of the difTerent varieties of starch from the same granules from seventeen flour samples. species. Dox and Roark3 found that the Naudain,ll again using but two samples, gelatinization temperatures of thirteen va­ found that small granules imbibe relatively rieties of maize starch ranged from 64. 10 to less water than large ones, but Grewe and 71.10 C.4 Francis and Smith" made analogous Bailey,12 using seventeen samples, were un­ observations. Rask and AlsbergG have shown able to discover any correlation between that the viscosity of pastes made from the granule size and heat of imbibition such as starches of difTerent varieties of wheat varies one might expect if granules of different sizes greatly. One of the factors that determines possessed different powers of combining with the viscosity of pastes is the size to which the or imbibing water. However, the gelatiniza­ individual granules swelJ.7 If Naudain's state­ tion of starch is a complex phenomenon,13 and ment that small and large wheat-starch gran­ the results of Naudain and of Grewe and ules swell to difTerent sizes relatively is con- Bailey are not necessarily contradictory. In this connection, one further question needs consideration. As we have seen, small 1 C. L. Alsberg, "Studies upon Starch," Industrial and Engineering Chemistry, February 1926, XVIII, 190. granules tend to swell at higher temperatures 2 Op. cit. than large ones, and they also swell more 3 A. W. Dox and G. W. Roark, Jr., "The Determina­ slowly. It follows that in the oven large gran­ tion of Gelatinization Temperatures of Starches by ules will begin to swell first. In doing so they Means of an Electrically Heated Chamber on the Micro­ scope Stage," .Journal of the American Chemical So­ must absorb water and thereby afTect the ciety, April 1917, XXXIX, 742-45. plasticity of the baking dough. If the average 4 Eynon and Lane, op. cit., p. 48. granule size is small, this effect would come ,; C. K. Francis and O. C. Smith, "The Determination later than if the average size is large. The of the Gelatinizing Temperature of the Starches from the Grain Sorghums by Means of a Thermo-Slide," longer the dough remains plastic, the larger, .Journal of Industrial and Engineering Chemistry, June presumably, the spring in the oven, other 1916, VIII, 509-11. things being equal. And so it may be that in 6 O. S. Rask and C. L. Alsberg, "A Viscosimetric this way granule size is one of the numerous Study of Wheat Starches," Cereal Chemistry, January 1924, I, 7-26. factors that determine the volume of the 7 Alsberg, "Studies upon Starch." baked loaf. B E. Grewe and C. H. Bailey, "The Concentration of Diastase attacks intact granules at the sur­ .Glutenin and Other Proteins in Various Types of face. It is obvious that small granules expose Wheat Flour," Cereal Chemistry, May 1927, IV, 230-47. 90p. cit. a proportionately greater surface than large lOOp. cit. ones. Therefore, other things being equal, 11 Ibid. diastase should form sugar more quickly in 12 Op. cit. Cf., however, Jago and Jago, op. cit., p. 322. flours with small-grained starches than in 12 Alsberg, "Studies upon Starch." those with large-grained ones. Such flours CONSEQUENCES OF VARIABILITY IN THE NATURB OF STARCH 245 should exhibit greater diastatic power and general they grow larger from the periphery doughs made from them should begin to fer­ inward,7 and this is true also for garden peas.8 ment at a faster rate. Perhaps the size of the While similar ohservations do not seem to granules is the key to the observation of Col­ have heen put on record for wheat, it is to be latzl that the starch of strong flours appears assumed that the same condition obtains. As to be more easily hydrolyzed by diastatic fer­ we have seen, in the modern process of mill­ ments than that of weaker ones. However, ing, Hour is produced separately from dif­ Grewe and Bailey, as stated above, were not ferent parts of the berry; these fractions or able to demonstrate any relationship between "streams" are combined by the miller to pro­ granule size and diastatic power (see also duce the several grades of flour. As we have below). also seen, these streams differ in respect to But the area of surface exposed to diastase the location of the starch and in respect to action may not be the only factor that deter­ their diastase content. To these differences, mines susceptibility to diastatic attack. J 0- we may now add differences in the size of the hannsen2 found that the smaller granules of starch grains; those streams coming from the barley have thicker walls and are firmer-ob­ outer regions of the endosperm should tend servations in harmony with those of Lenz." to have larger proportions of small granules Furthermore, it is probable, as we have seen, as compared with those from the interior. that the chemical composition of granules of This may he a further factor contributing to different sizes is different and this, too, may differences in the quality of the different he a factor in determining the variability in streams. resistance to diastase, which has been estab­ Effect of injury on starch granules.-As al­ lished by Whymper,4 Collatz," and Rumsey." ready pointed out, the starch granule in its Obviously the resistance of starch granules is natural state is wholly insoluble in cold water. the resultant of a number of factors; to what If, however, it is injured mechanically-if, for extent susceptibility to diastatic attack is a example, a bit is chipped out of its circumfer­ function of granule size needs further inves­ ence-it swells at the site of injury and some tigation. of its substance dissolves. The more com­ In barley the starch in the outer layers of pletely the granule is shattered, the greater the endosperm and toward the tip of the ker­ the amount dissolved.o No very great force is nels consists of quite small granules, but there necessary, although the microscopic size of are small granules in the interior also. In the granules makes it technically difficult to apply what little force is required. 1 F. A. Collatz, Flour Strength as Influenced by the That the injured granule swells is shown Addition of Diastatic Ferments (American Institute of Baking Bulletin 9), 1922. by the fact that a 10 per cent suspension of 20p. cit. ground-potato starch in cold water is at first 30p. cit. merely sirupy and viscous. In time, it becomes 4 "Microscopical Study of Changes Occurring in much thicker and may ultimately form a me­ Starch Granules during Germination of Wheat." chanically fragile jelly.10 Therefore, a flour GOp. cit. with many injured granules absorbs more o L. A. Rumsey, The Diastalic Enzymes of Wheat Flour and Their Relation to Flour Strength (American than if it had few or noneY The dough of a Institute of Baking Bulletin 8), 1922. severely overground flour has a somewhat 7 Johannsen, op. cit. modified consistency and tends to exhibit a 8 Gregory, op. cit. shiny surface probably due to dissolved starch. ° Aisberg and Perry, "The Effect of Grinding upon Very few dyestuffs stain intact unheated Starch and Starch Pastes." starch granules, but injured ones stain readily 10 E. P. Griffing, cited in Alsberg, "Studies upon Starch." at the site of the injury and staining is con­

11 This phenomenon is not peculiar to starch. It is fined to this area. Granules swollen by heat­ well known that grinding cellulose (wood pulp) very ing in water also stain, but the entire granule fine increases its affinity for water (cf. J. R. Katz, "Die Quellung," Ergebnisse der Exakten Naturwissenschaf­ takes up the dye throughout. Hence, it is easy len, 1924, III, 347). to detect injury to, or swelling of, granules by 246 STARCH AND FLOUR QUALITY staining them, for example, with Congo red. granules give off more or less starch to the The presence of injured granules in flour may surrounding watery medium. Wanklyn and readily be detected in the following way: Cooper,4 Alsberg and Griffing,5 and also Shol­ Place a tiny bit of the flour on a glass slide, lenberger and Colemano have shown that over­ add a drop or two of water containing a little grinding increases the portion of flour sol­ of the dye, Congo red, and observe under the uble in cold water. A part of this is starch microscope. Most of the starch granules which in the dissolved state is more suscep­ will remain unstained, but wherever a gran­ tible to attack by diastase than intact starch ule has been injured it takes up dye at the granules. site of the injury and usually it can be seen It follows that if a flour contains appre­ to be swelling at this spot. With the swelling, ciable numbers of injured starch granules it some of its substance is going into solution. must exhibit greater diastatic power than if Another method is to stain with an iodine it contained few of them, for the dissolved solution so dilute it tinges intact granules but starch and the injured granules are very read­ faintly. Injured granules will be seen to be ily attacked by diastase. Such a flour would stained intensely at the site of injury.l Injury exhibit more rapid initial dough fermentation, may also be detected by examining with a for the diastase acting upon dissolved starch polarizing microscope. The injured granules and injured granules would make sugar rap­ have more or less lost their birefringence.2 idly available to the yeast. It should be pos­ Diastase acts but slowly upon ungelatinized sible to remedy deficient diastatic power of starch; but it has long been known that sugar some flours by overgrinding them slightly to is formed with great rapidity by diastase act­ increase the injured granules or by blending ing upon mechanically injured raw starch them with a small amount of heavily over­ grains.3 Indeed, there is little difference be­ ground flour. Alsberg and Griffing7 found that tween the rate of sugar formation from shat­ heavily overground flour when doughed be­ tered granules and from boiled ones. gins to ferment very much more vigorously Moreover, as already pointed out, injured than it did before it was overground, and this has been confirmed by Shollenberger, Mar­ 9 1 Scheffer, "Ueber den Nachweis von mechanischen shall, and Hayes,8 and by Malloch. Beschaedigu ngen der Staerkekoerner," Zeitschrift fur A number of investigators at the University das gesamte Getreidewesen, May-June 1919, XI, 41-43. of Minnesotalo have found that an appreciable 2 Ibid. amount of sugar is formed very early in 3 H. T. Brown and J. Heron, "Contributions to the History of Starch and Its Transformations," Journal dough. It may be that it originates from in­ of the Chemical Society (London), Transactions, 1879, jured granules, for these must always be XXXV, 596-654; 1. Maquenne, A. Fernbaeh and J. present though in ordinary flours probably in Wolff, "Retrogradation et coagulation de l'amidon," Comptes rendus hebdomadaires des seances de l'Acad­ small numbers. Perhaps the augmentation of emie des Sciences, .January 4, 1904, CXXXVIII, 49-51; early sugar formation by severe mixing is to 1. Maquenne, "Sur la nature de la fecule erue," ibid., be attributed to the speedy bringing of dia­ February 8, 1904, CXXXVIII, 375-77; E. P. Griffing, unpublished work cited in C. L. Alsberg, The R6le of stase in contact with such granules. It is of Starch in Bread Making, in Walton, op. cit., Part 1, course also conceivable, though it seems im­ p. 91. probable, that severe mixing injures granules. 4 J. A. Wanklyn and W. J. Cooper, Bread Analysis. In any event, in future comparisons of the A Practical Treatise on the Examination of Flour and Bread (London, Regan, Paul, Trench, Trtibner, 1881). diastatic power of different wheats, it will be 5 C. L. Alsberg and E. P. Griffing, "Effects of Fine necessary to make certain that in the grinding Grinding upon Flour," Cereal Chemistry, November of flour from them no undue numbers of 1925, II, 325-44. starch granules have been injured. 6 Op. cit. In interpreting the effects of overgrinding 7 "Effects of Fine Grinding upon Flour." upon flour, one must take into consideration 8 Op. cit. 9 Op. cit. that all of its diastase does not readily dis­ ll 10 Reported by Skovholt, op. cit., pp. 60 ff. solve. Ford and Guthrie have shown that to 11 Op.cit.' get the maximum of diastase into solution it CONSEQUENCES OF VARIABILITY IN THE NATURE OF STARCH 247 is necessary to digest the flour with proteo­ But flours with many injured starch gran­ lytic enzymes, for apparently the proteins of ules have much material, ~-amylose, dissolved Hour in some manner lock up a part of the out of the starch granules, as was explained diastase. If they do so by mechanically sur­ above. This starch rcquires watcr for its so­ rounding the diastase so it cannot be reached lution and therefore enters into competition by water, then fine grinding should set more for water with the gluten, the starch, and the of the diastase free and thus increase the other water-soluble constituents of flour. It diastatic power of flours. The effects, then, of is therefore only to be expected that over­ overgrinding are rather complex. ground flours absorb more water than the Shollenberger and Coleman1 believe that in same flour before it was overground, and this commercial practice overgrinding to an extent is actually the fact.4 great enough to be of practical significance Finally, over grinding is not the only way in does not occur. However, ordinary commer­ which starch granules may be injured. They cial milling presumably injures at least some are injured through diastatic attack when of the granules, for the cold-water-soluble car­ wheat sprouts, but the effects are best consid­ bohydrate material is greater in flour than in ered below in connection with water absorp­ grain. This is probably due to some injurious tion and diastatic power. effect upon starch grains during milling. It is Variability of starch in water-holding ca­ well known to millers that heavy grinding, or pacity.-As has been pointed out, starch is handling in horizontal conveyors, tends to responsible for at least half the absorption of impair certain properties of the flour produced flour. We have also seen that freedom of the from middlings. This is probably due to a individual starch granules and mechanical number of causes.2 Perhaps injury to starch injury influence absorption, and the possi­ is among them. At any rate Miller attributes bility that starch-granule size does so has been the "instability" and the poor keeping quality discussed. We have next to consider whether of flours milled from wheat stocks that are granules injured otherwise than by fracture too dry to the formation of an excessive and intact granules exhibit differences in their amount of soluble starch by the milling proc­ capacity to absorb water. ess. 3 Of course no "soluble starch" is formed When wheat sprouts, some of its granules (see p. 245) but ~-amylose is released. Such are attacked by diastase and this injures flours undoubtedly include unusually large them. According to Hesse,s diastase in at­ numbers of injured starch granules which tacking the surface of ungelatinized starch permit ~-amylose to diffuse and disperse, i.e., granules increases their water-binding power. to dissolve out at the points where the granule To this effect upon the starch granules Hesse is injured, into the free water of the dough. attributes the fact that doughs made with the Perhaps such flours are more hygroscopic addition of commercial diastase preparations than normal flours. If they are-and this re­ stiffen some time after the kneading is over mains to be determined-then their greater and may therefore be made up thinner, that "instability" may be accounted for, since is, with more water. In such doughs, Hesse moist flours tend to spoil. states, the water is more u-niformly distrib­ uted, more firmly bound, and less is given off lOp. cit. in baking. Bread yields are correspondingly 2 Bailey, op. cit., pp. 284-87. larger. 3 E. S. Miller, "The Significance of Moisture," North­ Differences in absorptive power of differ­ western Miller, September 30, 1925, CXLIII, 1428. ent uninjured starches might be due to (a) 4 AIsberg and Griffing, "Effects of Fine Grinding upon Flour." differences in the hydration capacity of the

5 A. Hesse, "Fermente in del' Nahrungsmittelindus­ starch substance itself, or (b) differences in trie," in Oppenheimer, op. cit., Vol. IV, Part 2, pp. 306, the physical structure of the granules, render­ 323. ing them capable of holding more "en­ 6 w. Nossian, "Ueber das hygroskopische Verhalten mehrer Stiirkearten," Journal fur praktische Chemie, trapped" water. 1861, LXXXIII, 41. NossianG has reported that the starches of 24~ STARCH AND FLOUR QUALITY different plants absorb different amounts of hot, dry seasons the wheat plant produces moisture from an atmosphere of the same hu­ wheat starches less susceptible to action of midity, but Grewe and Baileyl seem to be the swelling agents than when the starches are only investigators who have made comparative produced in seasons of lower temperature and studies on the hydration capacity of the more abundant rainfall." Hot, dry weather starches of different wheats. Their method tends to favor the production of hard vitreous was indirect: the determination of heat of kernels, and such kernels, as we have seen, imbibition. They found differences between tend to have more numerous small granules. different samples; these were not correlated Perhaps, small granules are less susceptible with granule size. In view of these findings to swelling, for, as we have seen, they tend and in view of the different behavior of dif­ to require a higher temperature for gelati­ ferent wheat starches in swelling, in gelati­ nization. nization, and in the viscosity of the pastes they It has long been known that dry granules form, it is quite probable that direct measure­ swell when brought into water, and the in­ ments may uncover differences in the capacity crease in the length of their axes may be as of different starches to bind water of hydra­ much as 15 per cent or more.4 Such shrinking tion. The significance of differences in gelat­ and swelling, unlike swelling in hot water, is inization and viscosity will be considered not accompanied by any change in structure below in connection with the analysis of con­ visible under the polarizing microscope, ex­ ditions in the loaf. Here we shall consider cept that in drying rifts may develop which other forms of swelling only. when granules are swelled again in water do Mangels and Bailey2 have shown that aque­ not again close completely. Moreover, drying ous suspensions of starches from different deprives granules of their power to reassume types of wheat will vary in viscosity when their former volume when wetted.5 treated with agents that swell the granules This shrinking and swelling without heat without heat. Later, Mangels3 extended these of starch granules may seem astonishing in observations and found further that not view of the fact that X-ray analysis has shown merely varietal but also seasonal and regional starch granules to possess crystalline struc­ factors affect the properties of wheat starches, ture,a if by this is meant an orderly arrange­ as indicated by the degree of their swelling in ment of the atoms and molecules. The bodies the cold under the influence of various swell­ we have been accustomed to regard as crystal­ ing agents, such as sodium hydroxide and line have geometrical shapes, do not swell, and urea. He reached the conclusion that varia­ could only do so with the sacrifice of their tion with regional environmental conditions is geometric characteristics. However, types of not due to a single chemical or morphological crystals are known that do swell, for example, factor. Seasonal variation indicated "that in protein crystals,1 and in doing so they change lOp. cit. their shape. Moreover, many vegetable fibers, 2 C. E. Mangels and C. H. Bailey, "Relative Viscosi­ now known through X-ray analysis to possess ties of Wheat Starches," Indus/rial and Enoineerino an orderly arrangement of their molecules, Chemistrll, April 193:!, XXV, 456-60. 3 Op. cit. behave very much like starch in that they also 4 Meyer, op. cit., p. 127. swell in cold water. o H. de Vries, "LeeI'boek del' Plantenphysiologie," But drying also affects the power of gran­ in C. A. J. A. Oudemans and H. de Vries, Leerboelc der Plantenlcunde (Nijmegen, Alfen, Noman, 1895), I, 147. ules to swell when heated in water. Potato 6 On the crystalline structure of wheat starch see starch dried to a moisture content of 8 per H. O. Herzog and W ..Janeke, "Uber den physikalisehen cent and then heated in water was foundS to Aufbau einiger hochmolelmlarer organiseher Verbin­ swell to such a size that the area of its largest dungen," Berichte del' Deutschen Chemischen Gesell­ schaft, November 13, 1920, LlII, 2162-64. optical cross section was only 57 per cent of 7 A. F. W. Schimper, "Uebel' die Krystallisation del' that of the undried starch when it was gelat­ eiweissartigen Substanzen," Zeitschrift fiir Krystallo­ inized. After allowing the dried starch to oraphie und Mineralooie, November 16, 1880, V, 130-68. remain in contact with the atmosphere and to S E. P. Griffing, unpublished work cited in Alsberg, "Studies upon Starch." absorb moisture from it for two days, swelling CONSEQUENCES OF VARIABILITY IN THE NATURE OF STARCH 249 increased 10 per cent. This observation is in hot, dry weather after flowering tends to raise accord with that of Mangels and Bailey,! who the protein content of the berry and make it found that a wheat starch prepared without vitreous;5 and Johannsen" has found that vit­ drying at a temperature higher than that of reous barley kernels have more small starch the laboratory changed its capacity to swell granules than soft kernels. It is stated by when it was wetted and redried at 80 0 C. certain investigators7 that small granules ex­ These observations of Griffing, moreover, hibit a higher temperature of gelatinization are consistent with those of Mangels, above than large ones, that is, they resist swelling cited, to the eITect that hot, dry weather ren­ in hot water more. All these observations ders starch less susceptible to swelling than hang together, so that one is tempted to as­ cool, moist weather. In the light of these facts, sume some sort of causal relationship be­ the statement of Whymper2 that the tempera­ tween growing conditions and starch proper­ ture of gelatinization varies with the state of ties. maturity of the starch granules, and the ob­ Further evidence that cultural conditions servations of Fernbach and W oltra that starch aITect the properties of starch is furnished from immature peas is different from that by the findings of van de Sande-Bakhuyzen,S of a mature sample of the same variety, gain who found that wheat grown with continuous significance. illumination and under uniform conditions In this connection, it is perhaps worth of temperature and humidity produced wheat while to bring to the reader's attention that berries with starch grains that did not show but little gluten is formed in the development the lamellation or rings that can be seen with of the wheat berry before its moisture con­ the microscope in the starch of ordinary tent has dropped to 40 per cent or less.4 With wheat seeds. further reduction of the water content, the The eITect of growing conditions which point is approached at which starch in the seems to be correlated with granule size and berry would have to give up water of hydration ability to swell may in the last analysis prove (36 per cent, according to Rodewald; 30 per to be the expression of chemical composition. cent, according to Newton and Cook; see As we have seen, starch granules are composed above). It is to be presumed that as this of a-amylose and ~-amylose, of which a-amy­ point is reached the nature of the starch lose alone contains phosphoric acid. As we formed must be affected, if indeed starch is have also seen, there is reason to believe that still formed so late in the life of the berry. small granules contain more phosphoric acid Very possibly the rate at which the berry dries than large ones. One is therefore justified in out and ripens determines some of the charac­ surmising that small granules contain rela­ teristics of its starch. Certainly we know that tively more a-amylose than large ones. Alpha­ amylose when dispersed in water is thick and lOp. cit. viscous, whereas ~-amylose yields rather thin 2 "Microscopical Study of Changes Occurring in and mobile solutions. This may well indicate Starch Granules during Germination of Wheat." that a-amylose binds more water than ~-amy­ 3 A. Fernbach and J. Wolff, "Analogie entre l'ami­ don coagule par l'amylocoagulase et l'amidon de pois," lose. If this be true, one would expect small­ Comples rendlls hebdomadaires des seances de l' Aca­ granule starches with a high phosphoric­ demic des Sciences, June 5, 1905, CXL, 1547-49. acid content and therefore a high a-amylose 4 For a fuller discussion, see C. L. Alsberg, "Envi­ ronment, Heredity, and Wheat Quality," WHEAT STUD­ content to bind more water than those with ms, March 1934, X, 229-49. low content. That this is likely ultimately G Ibid. to be proved to be the case is indicated by e Op. cit. o 7 Nyman, op. cit. the conclusion of Mangels: "An inverse S H. L. van de Sande-Bakhuyzen, "The Structure of relationship between phosphorus content and Starch Grains from Wheat Grown under Constant Con­ swelling capacity or viscosity is indicated." di.tions," Proceedings of the Society for Experimental BIOlogy and Medicine, January 1926, XXIII, 302-05. However, the experiments of Grewe and o Op. cit. Bailey,'° above discussed, point in the opposite lOOp. cit. direction. 250 STARCH AND FLOUR QUALITY

In short, while there is as yeL little direct moisLened. However, moisture can enter the evidence on record that difTerenL sLarches bind interior of the wheaL berry only at the germ. or absorb difTerent percentages of water, there Therefore, in tempering, the bran coaL is the is much indirect evidence that investigation portion of the wheat that is moistened and of this question is likely to show LhaL this is therehy toughened. The miller tal{es care indeed the facL. The few measurements of not to add so much water or to permit the hydration of wheat starch made hy Newton tempering process to lake so much time that and Cook J are suggestive, for their values much moisture enters the endosperm. If the range from 18 to 30 per cent. endosperm becomes moist, it is toughened, The conclusion seems jusLifiahle that much and when ground between the rolls it tends evidence points to the variability in the proper­ to nake rather than to crack. It is possible, tics of sLarch being a factor in determining therefore, that very dry wheats still remain the variability in respect to absorption of abnormally brittle after tempering, with cor­ difTerent llours. responding danger that in grinding starch Perhaps such a causal relationship ex­ granules may he injured. This risk would be plains the very high ahsorption and relatively greater in softer wheats, like the Persian and poor baking qualities of llours from certain Indian wheats, than in hard vitreous ones. To wheats grown in very hot, very dry localities overcome excessive briLtleness of very dry and harvested very dry (5-8 per cent mois­ wheats, it is often the custom to temper in two ture), for example, some Indian and Persian stages. A portion only of the tempering water wheats. Their great ahsorption is not an is added at the beginning and the wheat is al­ indication of great llour strength; it is prob­ lowed to stand for as much as two days to give ably due in the main to the dryness of the time for some of the moisture to penetrate sLarch which gives it very great moisture into the endosperm and render it less brittle. capacity. But it is possible also that a change The rest of the water is then added and the in the physical character of the starch is in­ wheat held for a few hours longer before volved, for, as pointed out elsewhere, sharply grinding.2 dried starch docs not swell to its original The capacity of starch to absorb water dur­ volume when moistened nor does it swell to ing gelatinization by heat is its most impor­ the same degree when gelatinized. Further­ tant property in the oven. The importance in more, it is probable that excessive drying this regard of size of granules, chemical com­ damages gluten, hut this is a question beyond position, and injury has already been pointed the scope of this study, out; but the past history of a starch seems also We must remember also thaL in milling to have a marked efI'ect upon its gelatinization such very dry wheats there may be danger that temperature. Whymper3 found that the tem­ large numbers of starch granules may be perature of gelatinization varies with the state injured. To be sure, before the wheat is sent of maLurity of the starch granule. LaWall and to the rolls it is tempered, that is, it is Graves,4 Nyman,n and Reicherto state that the gelatinization temperature of a starch freshly lOp. cil. isolated from living tissue is different from 2 Cf. Bailey, op. cil., p. 124. that of a starch which has been dried previous a "Microscopical Study of Changes Occurring in to gelatinization. Drying potato starch raises Starch Granules during Germination of Wheat." the gelatinization temperature from 59° to 4 Op. cil. 65° C. According to Samec,7 the length of r, Op. cit. time a starch has been wet is also a factor r, Op. cit. modifying its gelatinization temperature. 7 M. Samec, "Stl1dicn lihcl' Pflanzenlwlloide. 1. Die Losl1ngsql1cllung del' Stiirlw hei Gegenwart von Kris­ Furthermore, starch grains very readily talloidcn," J(olloidchem ische lJeihe[le, Decembcr 10, ahsorb ions of calcium, magnesium, potas­ 1 H12, III, 12B-fj(). sium, sodium, etc.,8 much as does the mineral ~ H. TryIler, "Beilriige zur Chemie del' Kartoffel­ sliil'kcfahl'ikation," Chemiker-Zeilllll(J, November 11, zeolite or the arLificial mineral "permutite," 1920, XLIV, 845-47; Lloyd, op. cil. which is used for water softening. The ab- CONSEQUENCES OF VARIABILITY IN THE NATURB OF STARCIl 251

sorption of different metals seems Lo alrect Lhe tion is raised a few degrees, the shift in 'vater physical propcrtics of granules. Fernbach and which occurs as the dough is converted inLo WolfP found that using watcr containing bread will be delayed rclative Lo the coagula­ limc in the preparation of starch afJ'ects thc tion of gluten. The spring in Lhe oven, the viscosity of the pastc. This would indicate porosity, and texLure of the loaf might well be that the presence of calcium in the granulc a/rected. However, no experimental evidcnce affects the tempcrature at which it gclatinizes is available on these points. and also thc degrce to which it ahsorbs water. Variability of tbe resistance of storch to There are many ohservations indicating that diaslase.-In preceding sections it has been the mineral-salt con Lent of thc water used to shown Lhat the nu mber of starch granulcs thaL makc dough is of some influence upon the hak­ lic free and are unsurrounded by gluten and ing quality of flour." These effects have heen thc number of granules that arc injured play assumed to he the result of action of thc saIt their part in contributing to the diastatic upon glutcn, for saIts contained in thc wash power of flour. Granules that have been in­ watcr, espcciaUy calcium, havc markcd effccts jured by diasLatic attack whilc still in the upon thc properties of glutcn as it is washed wheat before it was milled tend to increase the out of flour. 3 SLudents of thcse questions havc apparent diastatic power of Jlour because they not considercd that starch may ahsorb calcium arc corrodcd, pitted, and even perforatecl and and may be changed in its physical behavior therefore expose more surfacc to attack.4 Such thcrcby, thus contributing to the effect of the granules are especially numerous in /lour prcsence of mineral salts in dough and bread. made from sprouted wheat. For this rcason, Yct a part of these effects may well be starch and also because they contain larger amounts effects. of diastase than Jlour made from unsproutcd It is entirely within the range of possibilities wheat, such /lours exhibit great diastatic that variability in gelatinization tempcrature powcr. Flour deficienL in diastatic powcr may is a factor in causing variability of behavior be improved by addiLion of /lour made from in the oven. If the temperature of gelatiniza- slightly sprouted whcat." Or a little sproutcd wheat may be blended with wheat deficient il} lOp. cil. diastatic power before it is milled. " For a discussion sce Bailey, of!. cit., pp. 278ff. In years past much work was done upon "D. B. Dill and C. L. Alsberg, "Some Critical Con­ the resistance of gelatinized ~tarches to dia­ sidt'l"Utions of the Glutcn \Vashing Problem," Cereal Chemisil"/J, Scptemhcl' I H24, I, 222-,16. stase,a but these investigations are not signifi­ 1 Malloch, op. cit. cant for present purposes since wc are inter­ r, H. C. Sherwood and C. H. Bailey. "Control of Dia­ ested in the resistance of raw starch as it static Activity in Wheat Flour. r. Production of Dia­ occurs in flour. There is now considerahle static Floul' and Effect of Large Dosages," Cereal Chemi.~lrlJ, March 1926, III, 107-:i6; "II. Experiments evidencc that resistance to diastase varies with Flour Milled on a Commcrcial Scale," ibid., May from spccics to species and that in a si ngle 192G, III, 163-82. plant the starch may vary in Lhis respect from . fl This literature is discussed in Oppenheimer, op. 7 Cll., Vol. I, Pal-t 9, pp. 640 ff. organ to organ. 7 Ziegenspecl{, op. cit. Humphries and Simpson" brought forward 8 A. E. Humphries and A. G. Simpson, "Gas-Making evidence to show Lhat the rate at which the Capacity as a Factor in the Estimation of Strength in starch granules are attacked, and not the Wheaten Flour," in Sellcnil1 International Congress of Applied Cl1emistl"/J, London, Mall 27 10 .JlIne 2, 1909 activity of the enzyme, is the prime factor thai (London, Pal-tridge and Cooper, IHIO), Section VI A, determines the gas-making capacity of a fer­ pp. 27-:i8. menting dough. Rumsey," on thc basis of a fJ Op. cit. lOOp. cii. review of thc literature and his own experi­

II F. A. Collatz and O. C. Hacke, "Effects of Diastase ments, reached the conclusion that such dif­ and Malt Extract in Doughs," Cereal Chemisirl/, .July ferences do exist. CoIIatz10 and Collatz and I !l25, II, 213-27. Rackell found the starch of dilferent flours to 12 C. E. Mangels, "Factors Affecting the Diastatic Aetivity of Wheat Flour," ibid., September 1926 III possess different powers to resist diastase. ~16-il2. ' , Mangels12 found different flours to dill'er in the 252 STARCH AND FLOUR QUALITY

susceptibility of their starch to diastatic at­ the a-amylose molecule where the phosphoric tack. He also found durum-wheat starch more acid is attached.4 readily attacked than starch from hard-red­

spring wheat. Hermano and Rask! also noted SUMMARY that different wheat starches show significant variation in susceptibility to diastatic action. Starch contributes not merely to giving flour Malloch2 noted differences in different kinds and bread their food value and bulk but it also of wheat and in the same kind grown in dif­ serves the following purposes; ferent localities. His data also indicate that (1) Starch absorbs moisture, and when resistance increases with maturity. flour is doughed its starch takes up about as There can therefore be no doubt that dif­ much water as all the other flour components ferences in diastatic power of different flours combined. It is responsible for about half of may depend upon difference in the resistance the total "absorption" of flours. While it ab­ of their starches and that in this way starch sorbs only about one-third of its weight of properties may be an important factor in de­ water, it makes up nearly four-fifths of flour termining the baking quality of flours. -hence, its important part in absorption. To what can this variability of starches be It is impossible to say whether the starches due? It may well be due to granule size or to of different Wheats absorb different amounts chemical composition or to both. of water; this problem has as yet been insuffi­ In the discussion of the significance of ciently investigated. Evidence is presented granule size (see above), it was pointed out indicating that such differences probably exist. that, a priori, small-granule starches should If further study proves this to be true, then yield sugar more rapidly under the influence variability of wheat starches with reference to of a constant proportion of diastase, because their capacity to absorb water may be a factor they present a larger surface per unit of mass in determining the absorption of different than large granules. This suggestion does not flours. necessarily imply that the individual small In the wheat berry, the starch grains lie im­ granule is more easily attacked by diastase bedded in gluten. In milling, some of the than the large one. As a matter of fact, granules are loosened so that they are no Barnstein3 has found that large wheat-starch longer completely surrounded by gluten; they granules are much more rapidly attacked than then absorb water more rapidly and more per­ the small ones. We have, therefore, two fectly. The rate at which a flour absorbs mois­ factors working in opposite directions. Where ture when doughed and probably the total there are proportionally many small granules, amount absorbed depend to some extent upon there is more surface exposed to attack, but the location of its starch granules with respect more of the granules tend to be relatively re­ to the gluten. How many starch granules are sistant and vice versa. The relationships are freed from their gluten capsules depends upon too complex to warrant a positive statement. the kind of wheat that is milled and the fine­ Perhaps the reason why small granules are ness to which it is ground. Soft-wheat flours more resistant to diastase may be that they tend to have more of the granules separated probably contain more a-amylose (see above), from their gluten covering than hard-wheat which is more resistant than ~-amylose. More­ flours. Grinding a flour fine tends to free large over, diastase does not attack that portion of numbers of starch grains and to favor absorp­ tion of water. In the process of milling, some starch ! A. J. Hermano and O. S. Rask, "A Consideration of Certain Reactions of Starches with Special Reference granules are injured; in fine grinding a great to Enzyme Hydrolysis," ibid., November 1926, III, 361- many may be. Injured granules swell at the 92. site of injury and some of the starch substance 20p. cit. dissolves. Therefore, a flour containing many 3 Cited by Neumann, Brotgetreide und Brot, p. 242. 4 M. Samee, Kolloidchemie der Starke (Dresden and injured granules absorbs more water than it Leipzig, Steinkopf, 1927), p. 457. would if it contained none, and it also contains SUMMARY 253

more starch soluble in cold water. In good modified in various ways. Starch from commercial practice few granules are injured sprouted wheat is less resistant than it was and these phenomena are not important. How­ before sprouting began. Therefore if flour is ever, it is well known to millers that heavy made from parcels of wheat containing any grinding or excessive manipulation of mid­ material number of sprouted kernels, its dia­ dling stocks tends to reduce Hour quality; and static power is increased, in part because its this is perhaps due in part to injury to starch starch is less resistant, in part because granules. sprouted wheat contains much diastase. (2) Starch sustains fermentation. The Gelatinizing starch, as by heating with enzyme diastase, present in flour, attacks the water, lessens its resistance to diastase. There­ slarch of dough with the formation of sugar fore, mixing in some scalded flour in dough which the yeast ferments to alcohol and car­ favors early fermentation; this may also be bonic acid. The carbonic acid together with accomplished by adding boiled potatoes or water vapor formed during baking leaven the other gelatinized starchy material. bread. If there were no starch or no diastase Injury to granules renders them easily at­ in flour, it would be necessary in some manner tacked. Therefore not merely does fine grind­ to furnish small amounts of sugar continu­ ing increase diastatic power by rendering ously to the yeast; otherwise leavened bread, more starch grains accessible to diastase, but as we know it, could not be produced. if the grinding is severe enough to injure large The rate at which a flour can produce sugar numbers of granules these will be attacked from starch when doughed is known as its rapidly, as will also that portion of the sub­ diastatic power. This depends, obviously, both stance of injured granules which becomes dis­ upon how much diastase a flour contains and solved. The diastatic power of such flours is upon how resistant its starch is to diastatic much increased. attack. The resistance of starch depends both (3) Starch is probably the principal sub­ upon its location and upon its intrinsic powers stance involved in the growing stale of bread. of resistance. In the oven, gluten coagulates and loses ca­ If most of the starch is locked up inside pacity to hold water while starch is gelatinized clumps of gluten where diastase cannot easily to the extent that the water present permits reach it, because diastase diffuses with diffi­ and gains in water-holding capacity. In fresh culty if at all through gluten, such starch bread, as a result, starch holds more water and granules are not attacked. If, however, the gluten probably less water than in dough. granules be free and not surrounded by a pro­ It is known that the gelatinization tempera­ tecting envelope, there is nothing to prevent ture of different starches differs somewhat diastase from reaching them. Hence breaking and that in the same sample large granules up gluten clumps by grinding, or otherwise, gelatinize at lower temperatures than small tends to increase the diastatic power of flours. ones. It is also known that the size distribu­ There is considerable evidence, though not tion of the granules is different in different yet enough to be finally conclusive, that the wheats. It may be that the differences in gelat­ starches of different wheats exhibit differences inization temperatures are related to the size in their resistance to diastase. It is extremely distribution of the granules. In any event, dif­ probable, therefore, that the diastatic power ferences in the temperature of gelatinization of different flours may vary for this reason. will cause starch to swell and absorb water The reason for such variability among wheat earlier or later in the oven and thereby affect starches is unknown. It may be due to differ­ the plasticity of the dough. It is therefore pos­ ences in the size of the granules or it may be sible that the nature of the starch affects the due to differences in chemical composition. "spring" in the oven and the loaf volume. This There is some evidence that in different wheat is a matter for future investigation. starches the percentages of a-amylose and In time, the gelatinized starch in bread [:I-amylose are 110t the same. undergoes spontaneous changes, and to such Intrinsic power to resist diastase may be change the growing stale of bread has been 254 STARCH AND FLOUR QUALITY attributed. While more study is needed before Much of what is presented in this study is the mechanism by which bread grows stale hypothetical and speculative. This is deliber­ may be regarded as wholly solved, it seems ate. The aim is to attract the attention of in­ clear already that starch plays the major role vestigators to the study of starch and of its in that mechanism. The growing stale of role in the preparation of baked products and bread is certainly not due Lo loss of water, i.e., to indicate some of the problems that await drying out. solution.

This study is the work of Carl L. Alsberg WHEAT STUDIES of the FOOD RESEARCH INSTITUTE

VOLUME IX No. 1. Price Spreads and Restraint of United States Wheat Exports. October 1932. $.60 No.2. The Voluntary Domestic Allotment Plan for Wheat. November 1932. $.75 No.3. The World Wheat Situation, 1931-32: A Review of the Crop Year. December 1932. $1.00 No.4. Survey of the Wheat Situation, August to November 1932. January 1933. $.50 No.5. Estimation of End-Year World Wheat Stocks from 1922. February 1933. $.50 No.6. Price Relations between July and September Wheat Futures at Chicago since 1885. March 1933. $.75 No.7. World Wheat Crops, 1885-1932: New Series, with Areas and Yields, by Countries. April 1933. $.75 No.8. Survey of the Wheat Situation, December 1932 to April 1933. May 1933. $.50 No.9. Britain's New Wheat Policy in Perspective. July 1933. $.75 No. 10. Survey of the Wheat Situation, April to August 1933. September 1933. $.50

VOLUME X No. 1. British Preference for Empire Wheat. October 1933. $.50 No.2. Price Leadership and Interaction among Major Wheat Futures Markels. November 1933. $.75 No.3. The World Wheat Situation, 1932-33: A Review of the Crop Year. December 1933. $1.00 No.4. World Wheat Survey and Outlook, January 1934. January 1934. $.50 No.5. Price Relations between May and New-Crop Wheat Futures at Chicago since 1885. February 1934. $.75 No.6. Environment, Heredity, and Wheat Quality. March 1934. $.50 No.7. World Wheat Survey and Outlook, May 1934. May 1934. $.50 Nos. 8 and 9. Decline and Recovery of Wheat Prices in the 'Nineties. June and July 1934. $1.00 No. 10. Pacific Northwest Wheat Problems and the Export Subsidy. August 1934. $.75

VOLUME XI No. 1. World Wheat Survey and Outlook, September 1934. September 1934. $.50 No.2. Decline in Wheat-Flour Export during the Depression. October 1934. $.75 No.3. Prices of Cash Wheat and Futures at Chicago since 1883. November 1934. $1".00 No.4. The World Wheat Situation, 1933-34: A Review of tlle Crop Year. December 1934. $1.00 No.5. World Wheat Survey and Outlook, January 1935. January 1935. $.50 No.6. Starch and Flour Qllality. February 1935. $.50

RECENT CONTRIBUTIONS FROM THE FOOD RESEARCH INSTITUTE (Numbered reprints available free on request) G 65. "Wheat, Wheat Policies, and the Depression," Joseph S. Davis. Review of Economic Statistics, April 1934 G 66. "New Chinese Agricultural Statistics," M. K. Bennett. Joumal of Farm Economics, April 1934 G 67. "A Random-Difference Series for Use in the Analysis of Time Series," Holbrook Working. Jour­ llal of the American Statistical Association, March 1934 G .68. "Wheat Consumption during the Depression," Carl L. Alsberg. Proceedings of the World's Grain Exhibitioll and COllference, Regina, Canada, 1933 G 69. "Observations on Wheat Consumption," Carl L. Alsberg. Proceedillgs of the Fifth Pacific Sci­ ence Congress, Victoria and Vancouver, Canada, June-JUly, 1933 G 70. "What about Wheat Prices?" Holbrook Working. Proceedings of the Fifth Pacific Science Con­ gress, Victoria and Vancouver, Canada, June-July, 1933 E 44. "The Effect of Diet on the Distribution of Glycogen in the Skeletal Muscle of the Rat," Mel­ ville Sahyun, Ray Simmonds, and Holbrook Working. The American Jourllal of Physiology, June 1934 E 45. "The Bitter Glucoside of the Olive," W. V. Cruess and C. L. Alsberg. Journal of the American Cllemical Society, October 1934 FOOD RESEARCH INSTITUTE STANFORD UNIVERSITY, CALIFORNIA

A research department of Stanford University, established in 1921 jointly by Carnegie Corporation of New York and the Board of Trustees of the Leland Stanford Junior University, for research in the production, distribution, and consumption of food.

DlRECI'ORS CARL L. ALSBEllG JOSEPH S. DAVIS ALONZO E. TAYLOR

ECONOMISTS MEllnILL K. BENNETT HOLBROOK WORKING

PUBLICATIONS

WHEAT STUDIES FATS AND OILS STUDIES Ten issues yearly, published at approximately A series (books, issued irregularly) of studies monthly intervals; $6.00 per volume. Each in fats and oils of animal and vegetable origin, volume includes a comprehensive annual re­ dealing primarily with economic aspects-­ view of the world wheat situation, a survey production, trade, prices, and utilization-but and outlook at four-month intervals, and with due reference to technical knowledge. usually six special studies. For partial list, see inside back cover.

GRAIN ECONOMICS SERIES MISCELLANEOUS PUBLICATIONS A series (books, issued irregularly) covering A series (books, issued irregularly) embody­ topics in grain economics not suited to full ing the results of research in fields other than discussion in Wheat Studies. those covered by the series listed above.

CONTRIBUTIONS Reprints from periodicals of articles written by members of the Food Research Institute.

List of publications available free on request. Address all communications to

FOOD RESEARCH INSTITUTE STANFORD UNIVERSITY, CALIFORNIA

European Sales Agents: Great Britain: P. S. KING & SON, LTD., 14, Great Smith Street, Westminster, S. W. 1, Londol1 Continental Europe: MAllTINUS NIJHOFF, 9 Lange Voorhout, The Hague, Holland