THE DIFFERENTIATION AND SPECIFICITY OF STARCHES IN RELATION TO GENERA, SPECIES, ETC. STEREOCHEMISTRY APPLIED TO PROTOPLASMIC PROCESSES AND PRODUCTS, AND AS A STRICTLY SCIENTIFIC BASIS FOR THE CLASSIFICATION OF PLANTS AND ANDIALS BY EDWARD TYSON REICHERT, M.D. Professor of Physiology in the University of Pennsylvania Research Associate of the Carnegie Institution of Washington IN TWO PARTS PART I WASHINGTON, D. C. Published bt the Carnegie Institution of Washington 1913 CARNEGIE INSTITUTION OF WASHINGTON Publication No. 173, Part I 31 q'] PBESa OF J. B. LIPPINCOTT COMPANY PHILADELPHIA PREFACE. The present memoir, which is purely in the nature of a report of a preHminary investigation, is complementary and supplementary to Publication No. 116 of this Institution, entitled "The Differentiation and Specificity of Corresponding Pro- teins and other Vital Substances in Relation to Biological Classification and Organic Evolution: The Crystallography of Hemoglobins," in the preface of which the following statement was made of the hypothesis upon which the research was founded, and of the support of the hypothesis by the results of the inquiry: "The trend of modern biological science seems to be irresistibly toward the explana- tion of all vital phenomena on a physico-chemical basis, and this movement has already brought about the development of a physico-chemical physiologj', a physico-chemical pathology, and a physico-chemical therapeutics. The striking parallehsms that have been shown to exist in the properties and reactions of colloidal and crystalloidal matter in vitro and in the living organism lead to the assumption that protoplasm may be looked upon as consisting essentially of an extremely complex solution of interacting and interdependent colloids and crystalloids, and therefore that the phenomena of life are manifestations of colloidal and crystalloidal interactions in a peculiarly organized solution. We imagine this solution to consist mainly of proteins with various organic and inorganic substances. The constant presence of protein, fat, carbohydrate, and inorganic salts, together with the existence of protein-fat, protein-carbohydrate, and protein-inorganic salt combinations justifies the belief that not only such substances, but also such combinations, are absolutely essential to the existence of life. "The very important fact that the physical, nutritive, or toxic properties of given substances may be greatly altered by a very slight change in the arrangement of the atoms or groups of molecules may be assumed to be conclusive evidence that a trifling modifi- cation in the chemical constitution of a ^'ital substance may give rise to even a profound alteration in its physiological properties. This, coupled with the fact that differences in centesimal composition have proved very inadequate to explain the differences in the phenomena of living matter, implies that a much greater degree of importance is to be attached to peculiarities of chemical constitution than is universally recognized. "The possibilities of an inconceivable number of constitutional differences in any given protein are instanced in the fact that the serum albumin molecule may, as has been estimated, have as many as 1,000 million stereoisomers. If we assume that serum globulin, myoalbumin, and other of the highest proteins may have a similar nimiber, and that the simpler proteins and the fats and carbohydrate, and perhaps other complex organic sub- stances, may each have only a fraction of this number, it can readily be conceived how, primarily by differences in chemical constitution of vital substances, and secondarily by differences in chemical composition there might be brought about all of those differences which serve to characterize genera, species, and individuals. Furthermore, since the factors which gi\-e rise to constitutional changes in one vital substance would probably operate at the same time to cause related changes in certain others, the alterations in one may logically be assumed to serve as a common index of all. IV PREFACE. "In accordance with the foregoing statement, it can readily be understood how environ- ment, for instance, might so affect the individual's metaboHc processes as to give rise to modifications of the constitutions of certain corresponding proteins and other vital mole- cules which, even though they be of too subtle a character for the chemist to detect by his present methods, may nevertheless be sufficient to cause not only physiological and mor- phological differentiations in the individual, but also become manifested physiologically and morphologically in the offspring. "Furthermore, if the corresponding proteins and other complex organic structural units of the different forms of protoplasm are not identical in chemical constitution, it would seem to follow, as a corollary, that the homologous organic metabolites should have specific dependent differences. If this be so, it is obvious that such differences should constitute a preeminently important means of determining the structural and physio- logical peculiarities of protoplasm. "It was such germinal thoughts that led to the present research, which I began upon the hypothesis that if it should be found that corresponding vital substances are not identical, the alterations in one would doubtless be associated with related changes in others, and that if definite relationships could be shown to exist between these differences and peculiarities of the living organism, a fundamental principle of the utmost importance would be established in the explanation of heredity, mutations, the influences of food and environment, the differentiation of sex, and other great problems of biology, normal and pathological. "To what extent this hypothesis is well founded may be judged from this partial report of the results of our investigations: It has been conclusively shown not only that corresponding hemoglobins are not identical, but also that their peculiarities are of posi- tive generic specificity, and even much more sensitive in their differentiations than the ' zooprecipitin test.' Moreover, it has been found that one can with some certainty pre- dict by these peculiarities, without pre^•ious knowledge of the species from which the hemo- globins were derived, whether or not interbreeding is probable or possible, and also certain characteristics of habit, etc., as will lie seen by the context. The question of interbreeding has, for instance, seemed perfectly clear in the case of Canida and Muridce, and no difficulty was experienced in forecasting similarities and dissimilarities of habit in Sciuridce, MuridxB, Fdidcc, etc., not because hemoglobin is -per se the determining factor, but because, accord- ing to this hypothesis, it serves as an index (gross though it be, with our present very limited knowledge) of those physico-chemical properties which serve directly or indkectly to differentiate genera, species, and individuals. In other words, vital peculiarities may be resolved to a physico-chemical basis." Before and since the inception of the foregoing research, data have been slowly accumulating which point more and more strongly to the extremely important interrelationships that exist between the intramolecular configurations of various substances that play active roles in life's processes and the configurations of pro- toplasm. Hence, any progress in the application of stereochemistry to metabolic processes brings us closer to an understanding of those peculiar mechanisms of protoplasm which give rise to the phenomena which in the aggregate constitute life in its normal and abnormal manifestations. Hemoglobin, next to protoplasm, is unquestionably the most important organic substance of vertebrate life, and in conjunction with the stroma with which it is associated is an active functionating protein, the main function of which is the conveyance of oxygen from the external organs of respiration to the internal organs of respiration or the tissues generally. Starch is similarly an extremely important PREFACE. V constituent of a vast number of forms of plant, life, but its role in vital processes, while, on the whole, as essential to the continuance of life, is of an entirely different character. INIoreover, the general and special characters of these substances in relation to those of the bodies which originate them, and the mechanisms of their formation, are likewise strikingly different. Hemoglobin constitutes nearly the whole of the erythrocyte or red-blood corpuscle, and that portion of the ery- throcyte which is not this substance may properl}'^ be regarded as being in the nature of an adjunct, but nevertheless essential. In early embryonic life the erythrocytes are nucleated and probably derived directly from the mesoblastic elements, and they increase in number by mitosis. Later, proliferation occurs in all parts of the circulation, in certain capillary areas more than others, especially in those of the liver, spleen, and bone-marrow. During the progress of fetal devel- opment the erythrocytes, primarily spherical and nucleated, in time lose their nuclei, and become smaller, and take on the peculiar disk or cup-shaped form of postnatal life. After birth the red bone-marrow is the chief or sole seat of formation of erythro- cytes. It is the common conception that in this structure these corpuscles arise from nucleated red cells which exist at first as colorless, nucleated erythroblasts, and subsequently as smaller, denser, colored, nucleated normoblasts. The former, which are looked upon as the hereditary representatives of the embryonal erythro- cytes, are generally conceived to be converted into normoblasts by mitosis, and the latter in turn to become ordinary erythrocytes upon the disappearance of the nuclei by solution or extrusion. It is, however, more likely, as suggested in 1882 by Malassez, and very recently (1912) by the investigations of Emmel by means of plasma cultures, that the erythrocj^te of late fetal and post fetal life is formed from the cytoplasm of the erythroblast by a simj^le jjrocess of budding and detachment. According to either conception the erythrocyte is a separated portion of the mother substance that has been set free in a highly specialized life-sustaining medium, but in a distinctly modified form, inasmuch as it has a much higher hemoglobin content and is lacking in the amoeboid activities and power of reproduction of the parent substance, the latter differences being readily accounted for in the absence of nuclear matter. Starch, on the other hand, is a synthetic product of metabolic activity which bears no resemblance to the protoplasm that gave rise to it, and which is destined to serve an entirely different purpose from that of hemoglobin in the life-history of the organism. With hemoglohm as it exists associated with the stroma in the erythrocytes we are dealing with an active, living, functionating, highly specialized form of protoplasm; with starch, we deal with an absolutely inert, non-living, non-functionating, extremely complex carbohydrate in the nature of a stored-up pabulum, and a synthetic product of plastids which are specialized forms of protoplasm. In the hemoglobin research it was shown that the hemoglobin molecule is modified in specific relationship to genus, species, etc., which may be taken to mean that the form of protoplasm that is exj^ressed by the term erythrocjiie is correspondingly stereochemically modified; with starch it has been found, as will be seen by the context, that the molecule is likewise changed in specific relationship to genera, species, etc., which accordingly may also be VI PREFACE. taken to mean that during synthesis the products of activity are altered in their molecular peculiarities in specific relationship to the stereochemic modifications of the forms of protoplasm which produce them. In other words, one may lay down the dictum that each and every form of protoplasm existent in any organism is stereochemically peculiarly modified in specific relationship to that organism, and that, as a corollary, the products of synthesis will be modified in conformity with the molecular peculiarities of the protoplasm giving rise to them. It follows, therefore, that if the plastids of any given plant be of different stereochemic structure from those of others, the starch produced will show corresponding stereochemic variations, and hence he absolutely diagnostic in relation to the plant. Abundant evidence will be found in the pages which follow in justification of this statement. Moreover, if such differences are diagnostic, it is evident that they constitute a strictly scientific basis for the classification of plants. The author takes advantage of this opportunity to record his heartfelt obli- gation to the Carnegie Institution of Washington for the grants which made this investigation possible; and also to President R. S. Woodward and Dr. S. Weir Mitchell for invaluable assistance — assistance easy of mention, but difficult of adequate expression. Edward Tyson Reichert. From the S. Weir Mitchell Laboratory of Physiology, University of Pennsylvania, April, 1912. CONTENTS. Part I. PAQE Chapter I. Introduction 1 Objects of the Research 1 The Starch-Grain 1 Stereochemistry and some of its Applications 2 Differentiation of Stereoisomers 12 Conceptions, Methods, Plan, and Conduct of this Research 13 Assistants and Sources of Supply of Material 14 General Characters of the Investigation and Records 15 Chapter II. The Starch-Substance and the Structure, Form, and Mechanism op Formation op THE Starch-Grain 17 Various Views of the Nature of the Starch-substance, and of the Structure, Form, and Mechanism of Formation of the Starch-Grain 18 Occurrence of the Starch-Grain in Plant-life 60 Peculiar kinds of Starch, and Starch-hke Bodies 62 The Chief Forms and Classifications of Starch-Grains 64 Schleiden's Classification of Starch-Grains 64 Nageli's Classification of Starches from Different Sources 66 Meyer's Classification of Starch-Grains 67 Muter's Classification of Starches 71 Kraemer's Classification of Starch-Grains 71 \\'inton's Classification of Starches 72 Properties of the Starch-Grain in relation to Mendelism 72 Starch-Grain as a Spherocrystal 75 Conclusions relating to the Starch-Substance and the Structure and Mechanism of Formation of the Starch-Grain, based chiefly upon the foregoing Literature and in part upon Observations recorded in Subsequent Chapters 79 Chapter III. The Primary and the Reverted Decomposition Products of Starch 83 Synopsis of the more Important Literature up to the Investigations of Griessmayer, Briicke, and O'Sul- livan in 1872 84 The Basic Investigations of Griessmayer, Briicke, O'Sullivan, and Musculus and Gruber 93 Nature of the Chemical Processes involved in the Dextrinization and Saccharification of the Starch- Molecule 95 The Processes in the Conversion of Raw Starch into Starch-Paste, Pseudo-Solution, and True Solution 95 The Process in the Conversion of Starch into Dextrin 97 The Process in the Conversion of one form of Dextrin into another and into Sugar 100 Soluble Starch, Reverted Starch, Coagulated Starch, "Artificial Starch" and "Artificial Starch-Grains.". . . . 101 Soluble Starch 101 The Reversion of Starch-Paste, Soluble Starch, and Amylodextrin into Coagulated and Insol- uble forms of Starch. "Artificial Starch" and "Artificial Starch-Grains" Ill Amylodextrin and Maltodextrin 113 Erythrodextrin, Achroodextrin, Grenzdextrin, etc 120 Isomaltose, Maltose, Glucose, Saccharose, etc 138 Differences in the Products of Acid and Enzymic Action 149 Unusual Products of the Decomposition of Starch 154 Differences in the Decomposition Products of Different Starches 154 The Synthesis of Starch 156 Summary and Conclusions 160 Chapter IV. The Differentiation op the Starches from Different Genera, Species, etc 165 The Histological Method 166 Proximate Constituents and other Features of General Chemical Composition 166 The Proximate Principles 166 The Starch-Substance 166 The Water of Starch 167 The Ash 168 The Reaction of Starches 169 Miscellaneous Chemical Differences 170 Color Reactions 170 Iodine 170 vii Vrir CONTENTS. Chapter IV. The Differentiation of the Starches from Different Genera, Species, etc. — Continued, p^^^^ Color Reactions. — Continued. Iodine Vapor 171 Iodine-Chloral Hydrate 171 Iodine-Lactic Acid 171 Phenols with Sulphuric or Hydrochloric Acid 172 Reactions with AniUne Dyes 172 Reactions with Various Agents, with especial reference to the demonstration of the Structure and Com- position of Starches from Different Sources 172 Temperatures of Swelling and Gelatinization 174 Refractive Indexes of Starches 175 Reactions in Polarized Light 175 Characters of Starch-Paste and Pseudo-Solutions formed by Starches from Different Sources 170 Phenomena of Digestibility 177 Raw Starches 177 Boiled Starches 190 Summary and Conclusions 195 Chapter V. A Systematic Summary of the Gross Histological Properties of Starches from Various Sources 197 Nageli's Classification of Starches from Different Sources 197 Type 1 . Grains simple, centric, spherical 198 2. Grains simple, centric, lenticular 203 3. Grains simple, centric, oval 207 4. Grains simple, centric, spindle-shaped 212 5. Grains simple, centric, bone-shaped 212 6. Grains simple, eccentric, inverted-cone-shaped 213 7. Grains simple, eccentric, cone-shaped 214 8. Grains simple, eccentric, cuneiform, or flattened 221 9. Grains simple, eccentric, rod-shaped 229 10. Grains simple, structure obscure 232 1 1 . Grains semi-compound 251 12. Grains compound, with fused components 252 13. Grains compound, in one or two rows 254 14. Grains compound, of few components of equal size 255 15. Grains compound, components few and unequal 268 10. Grains compound, many components 273 17. Grains compound, hollow-spherical 293 Chapter VI. Methods Used in the Study op Starches in this Research 295 Histological Method 296 Iodine Reactions 297 Actions with Aniline Dyes 297 Reactions in Polarized Light, without and with Selenite 297 Temperature of Gelatinization 298 Actions of Swelling Reagents 298 Preparation of the Starches 299 Photomicrographic Records 300 Curves of the Reaction-Intensities of Different Starches 300 Charts of Comparative Reaction-Intensities of Starches with each Agent 301 Note in Regard to Pari II 301 Chapter VII. Differentiation and Specificity of Starches in Relation to Genera, Species, etc., AS Demonstr.\ted by the Methods employed in this Research 302 Starch-Substance a non-unit Substance 302 Methods employed in this Research, and their Values in the Differentiation of Starches 305 Differences in the Reaction-Intensities of Different Starches 313 Mean Temperatures of Gelatinization of Various Starches 334 Chapter VIII. General Applications of this Research 337 Specificity and Constancy of the Stereochemic Characters of Starches in Relation to Genera and Species. . 337 Applications of the Results of the Research to Pharmacognostics, Commerce, and Technical Pursuits 337 Applications of Principles to Pharmacodynamics 338 General Applications of this Research in Systematic Botany 340 Summary and Conclusions 342 CONTENTS. IX Part II. — The Starches. PAGE Starches of Graminace.« 343 Genus Zca 343 Starch of Zea mays var. everta (Golden Queen). Plate 1, Figs. 1 and 2. Chart 1 344 Starch of Zea mays var. cvcrta (White Rice). Plate 1, Pigs. 1 and 2. Chart 2 345 Starch of Zea m.ays var. indurata (North Dakota). Plate 1, Figs. 1 and 2. Chart 3 346 Starch of Zea mays var. indurata (Compton's Early). Plate 1, Figs. 1 and 2. Chart 4 347 Starch of Zea mays var. indentata (Early Learning). Plate 1, Figs. 1 and 2. Chart 5 348 Starch of Zea mays var. indentata (Hickory King). Plate 1, Figs. 1 and 2. Chart 6 349 Starch of Zea mays var. saccharata (StowcU's Evergreen, a dent sweet corn). Plate 1, Figs. 1 and 2. Chart 7 350 Starch of Zea mays var. saccharata (Black Mexican, a flint sweet corn). Plate 1, Figa. 1 and 2. Chart 8 351 Starch of Zea mays var. saccharata (Golden Bantam, a flint sweet com). Plate 1, Figs. 1 and 2. Chart 9 352 Differentiation of the Starches of the Genus Zea 354 Notes on the Starches of Zea 356 Genus Andropogon 357 Starch of Andropogon sorghum var. (White Kaffir Corn). Plate 1, Figs. 3 and 4. Chart 10 357 Starch of .Vndropogon sorghum var. (Yellow Branching Sorghum). Plate 1, Figs. 3 and 4. Chart 11. . . 359 Starch of Andropogon sorghum var. (Shallu). Plate 1, Figa. 3 and 4. Chart 12 360 Differentiation of the Starches of the Genua Andropogon 361 Notes on the Starches of Andropogon 362 Genus Panicum 362 StarchofPanicumcrus-galU var. (Japanese or Barnyard Millet). Plate 1, Figa. 5 and 6. Chart 13 362 Genus Oryza 303 Starch of Oryza sativa var. Plate 2, Figs. 7 and 8. Chart 14 363 Genus Triticum 364 Staich of Triticum sativum var. vulgare. Plate 2, Figs. 9 and 10. Chart 15 364 Starch of Triticum sativum var. dicoccum. Plate 2, Figs. 9 and 10. Chart 16 366 Differentiation of the Starches of the Genus Triticum 367 Notes on the Starches of Triticum 368 Genua Secale 368 Starch of Secale cereiile var. (Mammoth Winter). Plate 2, Figs. 11 and 12. Chart 17 368 Starch of Secale eereale var. (Spring). Plate 2, Figs. 11 and 12. Chart 18 370 Differentiation of the Starches of the Genus Secale 371 Notes on the Starches of Secale 372 Genus Hordeum 372 Starch of Hordeum sativum var. (Champion). Plate 3, Figs. 13 and 14. Chart 19 372 Genua Avena 374 Starch of Avena sativa var. (Clydesdale). Plate 3, Figs. 15 and 16. Chart 20 374 Genus Arrhenatherum 375 Starch of Arrhenatherum elatius var. Plate 3, Figs. 17 and 18. Chart 21 375 Notes on the Starches of Graminacea;. Charts 22 to 30 376 Starches of Leguminos.e 378 Genus Vicia 378 Starch of Vicia sativa. Plate 4, Figs. 19 and 20. Chart 31 378 Starch of Vicia villosa. Plate 4, Figs. 21 and 22. Chart 32 380 Starch of Vicia faba. Plate 4, Figs. 23 and 24. Chart 33 381 Starch of Vicia fulgena. Plate 5, Figs. 25 and 26. Chart 34 382 Starch of Vicia gerardi. Plate 5, Figs. 27 .and 28. Chart 35 384 Differentiation of Certain Starches of the Genus Vicia 385 Notes on the Starches of Vicia 386 Genua Phaseolus 386 Starch of Phaseolus vulgaris var. (Rod Kidney Bean). Plate 6, Figs. 31 and 32. Ch.art 36 386 Starch of Phaseolus lunatus var. (Henderson's Bush Lima Bean). Plate 6, Figs. 33 and 34. Chart 37. . 387 Differentiation of Certain Starches of the Genus Phaseolus 388 Notes on the Starches of Pha-seolus 389 Genus Dolichoa 389 Starch of Dohchos lablab. Plate 6, Figa. 35 and 36. Chart 38 389 Genus Mucuna 391 Starch of Mucuna pruriena. Plate 5, Figs. 29 and 30. Chart 39 391 Genus Lens 393 Starch of Lens eaculenta var. Plate 7, Figa. 37 and 38. Chart 40 393 Genua Lathyrua 395 Starch of Lathyrus odoratus var. shahzada. Plate 7, Figs. 39 and 40. Chart 41 395 Starch of Lathyrus sylvestris. Plate 7, Figa. 41 and 42. Ch.art 42 397 Starch of Lathyrus latifoliua var. albus. Plate 8, Figs. 43 and 44. Chart 43 398 Starch of Lathyrus magellanicus var. albus. Plate 8, Figs. 45 and 46. Chart 44 399 Differentiation of Certain Starches of the Genus Lathyrus 401 Notes on the Starches of Lath3'rus 402 Genus Pisum 402 Starch of Piaum sativimi var. (Eugenie, green). Plate 8, Figs. 47 and 48. Chart 45 402 Starch of Pisum sativum var. (Eugenie, yellow). Plate 8, Figs. 47 and 48. Chart 46 404 Starch of Pisum sativum var. (Thomas Laxton). Plate 9, Figs. 49 and 50. Chart 47 406 X CONTENTS. Starches of Lequminos.'E. — Continued. page Genus Pisum. — Continued. Starch of Pisum sativum var. (Electric Extra Early). Plate 9, Figs. 51 and 52. Chart 48 407 Starch of Pisum sativum var. (Mammoth Grey Seeded). Plate 9, Figs. 53 and 54. Chart 49 408 Starch of Pisum sativum var. (Large White Marrowfat). Plate 10, Figs. 55 and 50. Chart 50 409 Differentiation of Certain Starches of the Genus Pisum 410 Notes on the Starches of Pisum 412 Genus Wistaria 413 Starch of Wistaria chinensis. Chart 51 413 Genus Arachis 414 Starch of Arachis hypogoea (Jumbo Peanut). Plate 10, Figs. 57 and 58. Chart 52 414 Notes on the Starches of Legimiinosa; 415 Starches op Polygonace.b 418 Genus Polygonum 418 Starch of Polygonum fagopyrum var. (American). Plate 10, Figs. 59 and 00. Chart G2 418 Starch of Polygonum fagojiyrum var. (Japanese). Plate 10, Figs. 59 and 60. Chart 63 419 Differentiation of Certain Starches of the Genus Polygonum 420 Notes on the Starches of Polygonum 420 Starches op the Copulifeb.e 421 Genus Quercus 421 Starch of Quercus alba. Plate 11, Figs. 61 and 62. Chart 64 421 Starch of Quercus muehlenbergi. Plate 11, Figs. 63 and 64. Chart 65 423 Starch of Quercus prinus. Plate 11, Figs. 65 and 66. Chart 66 424 Starch of Quercus rubra. Plate 12, Fig. 67. Chart 67 426 Starch of Quercus texana. Plate 12, Fig. 68. Chart 68 427 Differentiation of Certain Starches of the Genus Quercus 428 Notes on the Starches of Quercus 430 Genus Castanea 430 Starch of Castanea americana. Plate 12, Figs. 69 and 70. Chart 69 430 Starch of Castanea sativa var. numbo. Plate 12, Figs. 71 and 72. Chart 70 432 Starch of Castanea sativa var. Chart 71 433 Starch of Castanea pumila. Plate 13, Figs. 73 and 74. Chart 72 434 Differentiation of Certain Starches of the Genus Castanea 435 Notes on the Starches of Castanea 437 Notes on the Starches of CupUiferae. Charts 73 and 74 437 Starch op Sapindace^e 438 Genus ^sculus 438 Starch of iEsculus hippocastanum. Plate 13, Figs. 75 and 76. Chart 75 438 Notes on the Starch of jEscuIus hippocastanum 439 Starches of Aroide.e 440 Genus Arum 440 Starch of Arum palastinum. Plate 13, Figs. 77 and 78. Chart 76 440 Starch of Arum cornutum. Plate 14, Figs. 79 and 80. Chart 77 441 Starch of Arum italicum. Plate 14, Figs. 81 and 82. Chart 78 443 Differentiation of Certain Starches of the Genus Arum 445 Notes on the Starches of Arum 445 Genus Arisaema 446 Starch of Arisajma triphyllum. Plate 14, Figs. 83 and 84. Chart 79 446 Genus Dracunculus 447 Starch of Dracunculus vulgaris. Plate 15, Figs. 85 and 86. Chart 80 447 Genus Richardia 449 Starch of Richardia elliotiana. Plate 16, Figs. 91 and 92. Chart 81 449 Starch of Richardia africana. Plate 16, Fig. 93. Chart 82 451 Starch of Richardia albo-maculata. Plate 16, Figs. 95 and 96. Chart 83 452 Differentiation of Certain Starches of the Genus Richardia 454 Notes on the Starches of Richardia 454 Genus Dieffenbachia 455 Starch of Pith of Dieffenbachia seguine var. nobilis. Plate 17, Figs. 97 and 98. Chart 84 455 Starch of Cortex of DiclTrnbai'hia seguine var. nobilis. Plate 17, Figs. 99 and 100. Chart 85 457 Starch of Pith of Dielfenbachia seguine var. maculata. Plate 17, Figs. 101 and 102. Chart 86 458 Starch of Cortex of Dieffenbachia seguine var. maculata. Plate 18, Figs. 103 and 104. Chart 87 460 Starch of Pith of Dieffenbachia seguine var. irrorata. Plate 18, Figs. 105 and 106. Chart 88 462 Starch of Cortex of Dieffenbachia seguine var. irrorata. Plate 18, Figs. 107 and 108. Chart 89 464 Starch of Pith of Dieffenbachia illustris. Plate 19, Figs. 109 and 1 10. Chart 90 466 Starch of Cortex of Dieffenbachia illustris. Plate 19, Figs. Ill and 112. Chart 91 467 Differentiation of Certain Starches of the Pith of the Genus Dieffenbachia 469 Differentiation of Certain Starches of the Cortex of the Genus Dieffenbachia 470 Notes on the Starches of Dieffenbachia 471 Notes on the Starches of Aroidoea. Charts 92 to 96 a 471 Starches of Liliace* 474 Genus Lilium 474 Starch of Lilium candidum. Plate 20, Figs. 115 and 116. Chart 97 474 Starch of Lilium longiflorum var. giganteum. Plate 20, Figs. 117 and 118. Chart 98 476 Starch of Lilium var. eximium. Plate 20, Figs. 119 and 120. Chart 99 478 Starch of Lilium parrvi. Plate 21, Figs. 121 and 122. Chart 100 479 Starch of Lilium rubellum. Plate 21, Figs. 123 and 124. Chart 101 480 CONTENTS. XI Starches of Liliace.k. — Continued. pagb Genus Liliuin. — Continueil. Starch of Liliuin ])hil;ulclpliicuin. Plate 21, Figs. 125 and 120. Chart 102 481 Starch of l.iliuni ligiinuni var. splendens. Phite 22, Figs. 127 and 12S. Chart 103 483 Starcli of Liliuni luMn-yi. Plate 22, Figs. 129 and 130. Chart 104 484 Stari-li of Liliuni ainatuni. Plato 22, Figs. 131 and 132. Chart 105 485 Starch of Lilium spcciosuni var. album. Plate 23, Figs. 133 and 134. Chart 106 487 Starch of Lilium niartagon. Plate 23, Figs. 135 and 136. Chart 107 488 Starch of Lilium superbum. Plate 23, Figs. 137 and 138. Chart 108 489 Starcli of Lilium tenuiiolium. Plate 24, Figs. 139 and 140. Chart 109 491 Starch of Lilium pardalinura. Plate 24, Figs. 141 and 142. Chart 110 492 Starch of LiUum puberulum. Plate 24, Figs. 143 and 144. Chart 111 493 Ditlerentiation of Certain Starches of the Genus Lilium 494 Notes on the Starches of Lihum 498 Genus Frilillaria 498 Starch of Fritillaria meleagris. Plate 25, Figs. 145 and 146. Chart 112 498 Starch of Fritillaria pyrenaica. Plate 25, Figs. 147 and 148. Chart 113 500 Starch of Fritillaria pudica. Plate 25, Figs. 149 and 150. Chart 114 501 Starch o{ Fritillaria aurea. Plate 26, Figs. 151 and 152. Chart 115 503 Starch of Fritillaria armena. Plate 26, Figs. 153 and 154. Chart 116 504 Starch of Fritillaria imperialis var. aurora. Plate 26, Figs. 155 and 156. Chart 117 505 Starch of Fritillaria liliacea. Plate 27, Figs. 157 and 158. Chart 118 506 Starch of Fritillaria recurva. Plate 27, Figs. 159 and 160. Chart 119 507 Differentiation of Certain Starches of the Genus Fritillaria 509 Notes on the Starches of Fritillaria 511 Genus Calochortus 511 Starch of Calochortus albus. Plate 28, Figs. 163 and 164. Chart 120 511 Starch of Calochortus maweanus var. major. Plate 28, Figs. 165 and 166. Chart 121 513 Starch of Calochortus benthami. Plate 28, Figs. 167 and 168. Chart 122 514 Starch of Calochortus lilaeinus. Plate 29, Figs. 169 and 170. Chart 123 515 Starch of Calochortus nitidus. Plate 29, Figs. 171 and 172. Chart 124 516 Starch of Calochortus howellii. Plate 29, Figs. 173 and 174. Chart 125 517 Starch of Calochortus leichtlinii. Plate 30, Figs. 175 and 176. Chart 126 518 Starch of Calochortus luteus var. oculatus. Plate 30, Figs. 177 and 178. Chart 127 519 Starch of Calochortus splendens. Plate 30, Figs. 179 and ISO. Chart 128 520 Differentiation of Certain Starches of the Genus Calochortus 522 Notes on the Starches of Calochortus 523 Genus Tulipa 524 Starch of Tulipa hageri. Plate 31, Figs. 181 and 182. Chart 129 524 Starch of Tulipa sylvestris. Plate 31, Figs. 183 and 184. Chart 130 526 Starch of TuUpa greigi. Plate 31, Figs. 185 and 186. Chart 131 527 Starch of Tulipa billietiana. Plate 32, Figs. 187 and 188. Chart 132 528 Starch of Tulipa didieri. Plate 32, Figs. 189 and 190. Chart 133 529 Starch of Tulipa didieri var. mauriana. Plate 32, Figs. 191 and 192. Chart 134 530 Starch of Tulipa didieri var. fransoniana. Plate 33, Figs. 193 and 194. Chart 135 531 Starch of Tulipa clusiana. Plate 33, Figs. 195 and 196. Chart 136 532 Starch of Tulipa clusiana var. persica. Plate 33, Figs. 197 and 198. Chart 137 533 Starch of Tulipa oculus-solis. Plate 34, Figs. 199 and 200. Chart 138 534 Starch of Tulipa pra;co.x. Plate 34, Figs. 201 and 202. Chart 139 535 Starch of Tulipa australis. Plate 34, Figs. 203 and 204. Chart 140 537 Differentiation of Certain Starches of the Genus Tulipa 538 Notes on the Starches of Tulipa 540 Genus Scilla 540 Starch of Scilla sibirica. Plate 35, Figs. 205 and 206. Chart 141 540 Starch of Scilla peruviana. Plate 35, Figs. 207 and 208. Chart 142 542 Starch of Scilla bifolia. Plate 35, Figs. 209 and 210. Chart 143 543 Differentiation of Certain Starches of the Genus Scilla 545 Notes on the Starches of Scilla 546 Genus Chionodoxa 546 Starch of Chionodoxa lucilia;. Plate 36, Figs. 21 1 and 212. Chart 144 546 Starch of Chionodo.xa tmolusi. Plate 36, Figs. 213 and 214. Chart 145 547 Starch of Chionodoxa .sardensis. Plate 36, Figs. 215 and 216. Chart 146 548 Differentiation of Certain Starches of the Genus Chionodoxa 550 Notes on the Starches of Chionodoxa 550 Genus Puschkinia 551 Starch of Puschkinia scilloides. Plate 37, Figs. 217 and 218. Chart 147 551 Starch of Puschkinia scilloides var. libanotica. Plate 37, Figs. 219 and 220. Chart 148 552 Differentiation of Certain Starches of the Genus Puschkinia 554 Notes on the Starches of Puschkinia 554 Genus Ornithogalum 555 Starch of Ornithogalum nutans. Plate 37, Figs. 221 and 222. Chart 149 555 Starch of Ornithogalum umbellatum. Plate 38, Figs. 223 and 224. Chart 150 556 Starch of Ornithogalum narbonense (pyramidale). Plate 38, Figs. 225 and 226. Chart 151 557 Starch of Ornithogalum thyrsoides var. aureum. Plate 38, Figs. 227 and 228. Chart 152 558 Differentiation of Certain Starches of the Genus Ornithogalum 559 Notes on the Starclies of Ornithogalum 560 XII CONTENTS. Starches of Liliace-e. — Continued. paqe Genus Erythronium 561 Starch of Erythronium dens-canis. Plate 39, Figs. 229 and 230. Chart 153 561 Starch of Erythronium dens-canis var. grandiflorum. Plate 39, Figs. 231 and 232. Chart 154 562 Starcix of Erythronium americanum. Plate 39, Figs. 233 and 234. Chart 155 563 Starch of Erythronium grandiflorum. Plate 40, Figa. 235 and 236. Chart 156 564 Starch of Erythronium citrinum. Plate 40, Figs. 237 and 238. Chart 157 565 Starch of Erytlu-onium californicum. Plate 40, Figs. 239 and 240. Chart 158 567 Differentiation of Certain Starches of the Genua Erythronium 568 Notes on the Starches of Erythronium 569 Genus Hyacinthus 570 Starch of Hyacinthus orientalis var. alba superbissima. Plate 41, Figs. 241 and 242. Chart 159 570 Starch of Hyacinthus orientalis var. albulus (white). Plate 41, Figs. 243 and 244. Chart 160 571 Starch of Hyacinthus orientalis var. albulua (Itahan). Plate 41, Figs. 245 and 246. Chart 161 573 Differentiation of Certain Starches of the Genus Hyacinthus 574 Notes on the Starches of Hyacinthus 575 Genus Galtonia 576 Starch of Galtonia candicans. Plate 42, Figa. 247 and 248. Chart 162 576 Genus IVIuscari 577 Starch of Museari botryoides. Plate 42, Figs. 249 and 2.50. Chart 163 577 Starch of Museari paradoxum. Plate 42, Figs. 251 and 252. Chart 164 579 Starch of Museari micranthum. Plate 43, Figs. 253 and 254. Chart 165 580 Starch of Museari conicum. Plate 43, Figs. 255 and 256. Chart 166 581 Starch of Museari commutatum. Plate 43, Figs. 257 and 258. Chart 167 582 Starch of Museari racemosum. Plate 44, Figs. 259 and 260. Chart 108 583 Starch of Museari compactum. Plate 44, Figs. 261 and 262. Chart 169 585 Starch of Museari comosum. Plate 44, Figs. 263 and 264. Chart 170 586 Differentiation of Certain Starches of the Genus Museari 587 Notes on the Starches of Museari 589 Genus Brodiaja 589 Starch of Brodiaa peduncularis. Plate 45, Figs. 265 and 266. Chart 171 589 Starch of Brodiiea ixoides var. splendens. Plate 45, Figs. 267 and 268. Chart 172 591 Starch of Brodiaja Candida. Plate 45, Figs. 269 and 270. Chart 173 592 Starch of Brodi»a lactea. Plate 46, Figs. 271 and 272. Chart 174 594 Starch of Brodia?a laxa. Plate 46, Figa. 273 and 274. Chart 175 595 Starch of Brodiiea coccinea. Plate 46, Figs. 275 and 276. Chart 176 596 Starch of Brodia-a grandiflora. Plate 47, Figs. 277 and 278. Chart 177 597 Starch of Brodia;a californica. Plate 47, Figs. 279 and 280. Chart 178 598 Starch of Brodia;a purdyi. Plate 47, Figa. 281 and 282. Chart 179 600 Starch of Brodia?a stellaris. Plate 48, Figs. 283 and 284. Chart 180 601 Starch of Brodi.Ta capitata. Plate 48, Figs. 285 and 286. Chart 181 602 Starch of BrodiiPa congesta. Plate 48, Figs. 287 and 288. Chart 182 603 Differentiation of Certain Starches of the Genus Brodiaea 605 Notes on the Starches of Brodiaea 607 Genus Triteleia 60S Starch of Triteleia uniflora. Plate 49, Figs. 289 and 290. Chart 183 608 Genus Lachenalia 609 Starch of Lachenalia ijendula. Plate 49, Figs. 291 and 292. Chart 184 610 Starch of Lachenalia tricolor var. luteola. Plate 49, Figs. 293 and 294. Chart 185 611 Differentiation of Certain Starches of the Genua Lachenalia 612 Notes on the Starches of Lachenaha 613 Notes on the Starches of Liliaoea. Charts 186 to 200 613 SxAnCHES OF CoNVALLARIACEiE 616 Genus Convallaria 616 Starch of Convallaria majalia. Plate 50, Figs. 295 and 296. Chart 201 616 Genus Trillium 618 Starch of Trillium grandiflorum. Plate 50, Figs. 297 and 298. Chart 202 618 Starch of Trillium ovatum. Plate 50, Fig. 299. Chart 203 619 Starch of Trillium sessile var. californicum. Plate 50, Fig. 300. Chart 204 620 Differentiation of Certain Starches of the Genus Trillium 621 Notes on the Starches of Trillium 622 Notes on the Starches of Convallariaceae. Chart 205 622 Starches op Colchicace^ 623 Genus Colchicum 623 Starch of Colchicum parkmsoni. Plate 51, Figs. 301 and 302. Chart 206 623 Starches of Amartllidace.e 625 Genus Amaryllis 625 Starch of Amaryllis belladonna major. Plate 51, Figs. 303 and 304. Chart 207 625 Genus I lippeastrurn 627 Starch of Ilipiieastrum vittatum. Plate 52, Figs. 307 and 30S. Chart 20S 627 Starch of Hippcastrum equestre. Plate 52, Figs. 309 and 310. Chart 209 629 Starch of Hiijpeastruni aulicum var. robustum. Plate 52, Figs. 311 and 312. Chart 210 630 Differentiation of Certain Starches of the Genus Hipiieastrum 631 Notes on the Starches of Hii)peastrum 632 Genus Vallota 633 Starch of Vallota purpurea. Plate 51, Figs. 305 and 306. Chart 211 633 CONTENTS. XTII Starches of Amartllidace.e. — Continued. faob Genus Crinum 634 Starch of Crinum fimbriaUilum. Plate 53, Figa. 313 and 314. Chart 212 634 Starch of Crinum amoricaniim. Plate j3, Figs. 315 and 31(5. Chart 213 636 DilTorentiation of Certain Starches of the Genua Crinum 638 Notes on the Starches of Crinum 639 Genus Zepliyrauthes 639 Starcli of ZephjTanthes Candida. Plate 54, Figs. 319 and 320. Chart 214 639 Starch of Zepliyranthes ro.sea. Plate 54, Figs. 321 and 322. Chart 215 640 Differentiation of Certain Starches of the Genus ZephjTanthes 642 Notes on the Starches of Zephyranthes 642 Genus Sprekelia 642 Starch of Sprekelia formosissima. Plate 53, Figs. 317 and 318. Chart 216 642 Genus Hamanthua 644 Starch of Hajmanthus katherinse. Plate 54, Figs. 323 and 324. Chart 217 644 Gentis Hymenocallis 646 Starch of Hymenocallis undulata. Plate 55, Figs. 325 and 326. Chart 218 646 Starch of Hymenocallis calathina. Plate 55, Figs. 327 and 328. Chart 219 648 Differentiation of Certain Starches of the Genus Hjinenocallia 650 Notes on the Starches of Hymenocallis 650 Genus Leucoium 651 Starch of Leucoium a;stivum. Plate 55, Figs. 329 and 330. Chart 220 651 Starch of Leucoium vernum. Plate 50, Figa. 331 and 332. Chart 221 653 Differentiation of Certain Starches of the Genus Leucoium 654 Notes on the Starches of Leucoium 654 Genus Galanthus 655 Starch of Galanthus nivalis. Plate 56, Figa. 333 .and 334. Chart 222 655 Starch of Galanthus elwesii. Plate 56, Figs. 335 and 336. Chart 223 656 Differentiation of Certain Starches of the GeniLs Galanthus 658 Notes on the Starches of Galanthus 658 Genus Alstrcemeria 658 Starch of Alstrcemeria ligtu. Plate 57, Figa. 337 and 338. Chart 224 659 Starch of Alstrcemeria brasilienais. Plate 57, Figs. 339 and 3 10. Chart 225 660 Starch of Alstrcemeria aurantiaca (aurea). Plate 57, Figs. 341 and 342. Chart 226 661 Differentiation of Certain Starches of the Genus Alstrcemeria 662 Notes on the Starches of Alstrcemeria 663 Genus Stembergia 664 Staroli of Sternbergia lutea. Plate 58, Figs. 343 and 344. Chart 227 664 Genus Narcissus 665 Starch of Narcissus horsfieldii. Plate 58, Figs. 345 and 346. Chart 228 0(35 Starch of Narcissus maximus. Plate 58, Figs. 347 and 348. Chart 229 667 Starch of Narcissus bulbocodium. Plate 59, Figs 349 and 350. Chart 230 668 Starch of Narcissus bulbocodium var. conspicua. Plate 59, Figs. 351 and 352. Chart 231 669 Starch of Narcissus bulbocodium var. monophyllus. Plate 50, Figs. 353 and 354. Chart 232 G70 Starch of Narcissus incomparabilis. Plate 60, Figs. 355 and 356. Chart 233 671 Starch of Narcissus odorus. Plate 00, Figs. 357 and 358. Chart 234 672 Starch of Narcissus poeticus. Plate 00, Figs. 359 and 360. Chart 235 673 Starch of Narcissus biflorus. Plate 01, Figs. 361 and 362. Chart 236 674 Starch of Narciscus jonquilla. Plate 01, Figs. 363 and 364. Chart 237 675 Starch of Narcissus jonquilla var. rugulosus. PKate 61, Figs. 305 and 366. Chart 238 677 Starch of Narcissus jonquilla var. campernelli rugulosus. Plate 62, Figs. 367 and 368. Chart 239 678 Starch of Narcissus tazetta var. orientalis. Plate 62, Figs. 309 and 370. Chart 240 679 DitTerentiation of Certain Starches of the Genus Narcissus 680 Notes on the Starches of Narcissus 683 Notes on the Starches of .\maryllidacea3. Charts 241 to 253 683 Starches of Taccace^ 686 Geiuis Tacca 686 Starch of Tacca pinnatifida. Plate 62, Figs. 371 and 372. Chart 254 686 Starches op Iridacb.e 688 Genus Iris 688 Starch of Iris florentina. Plate 63, Figs. 373 and 374. Chart 255 688 Starch of Iris pallida var. speciosa. Plate 03, Figs. 375 and 376. Chart 256 690 Starch of Iris pumila var. cyanea. Plate 63, Figs. 377 and 378. Chart 257 691 Starch of Iris bismarckiana. Plate 04, Figs. 379 and 380. Chart 258 693 Starch of Iris iberica. Plate 64, Figs. 381 and 382. Chart 259 694 Starch of Iris .xiphium var. Grand tresorier. Plate 64, Figs. 383 and 384. Chart 260 695 Starch of Iris xiphium var. WUhelmine. Plate 05, Figs. 385 and 386. Chart 261 696 Starch of Iris xiphium var. Lustitanica. Plate 05, Figs. 387 and 388. Chart 202 697 Starch of Iris tingitana. Plate 65, Figs. 3S9 aufl 390. Chart 203 698 Starch of Iris reticulata. Plate 00, Figs. 391 and 392. Chart 264 699 Starch of Iris histrio. Plate 00, Figs. 393 and 394. Chart 205 700 Starch of Iris alata. Plate 00, Figs. 39."> and :J90. Chart 200 702 Starch of Iris caueasica. Plate 07, Figs. 397 and 398. Chart 267 703 Differentiation of the St.arches of the Genus Iris 704 Notes on the Starches of Iris 708 Genus Morcea 708 Starch of Monsa tristis. Plate 67, Figs. 399 and 400. Chart 268 708 XIV CONTENTS. Stakches of Iridace.e. — Continued. paoe Genus Homeriii 709 Starch of Homeria collina. Plate 67, Figs. 401 and 402. Chart 269 710 Genus Tigiidia 711 Starch of Tigridia pavonia var. grandiflora alba. Plate 68, Figs. 403 and 404. Chart 270 711 Starch of Tigridia jiavonia var. conchiflora. Plate 68, Figs. 405 and 406. Chart 271 713 Differentiation of Certain Starches of the Genus Tigridia 714 Notes on the Starches of Tigridia 715 Genus Gladiolus 715 Starch of Gladiolus bvzantinus. Plate 68, Figs. 407 and 408. Chart 272 715 Starch of Gladiolus primulinus. Plate 69, Figs. 409 and 410. Chart 273 716 Starch of Gladiolus cardinalis var. (Blushing Hridc). Plate 09, Figs. 411 and 412. Chart 274 717 Starch of Gladiolus floribundus. Plate 69, Figs. 413 and 414. Chart 275 718 DitTerentiation of Certain Starches of the Genus Gladiolus 719 Notes on the Starches of Gladiolus 721 Genus Watsonia 721 Starch of Watsonia humilis. Plate 70, Figs. 415 and 416. Chart 276 721 Starch of Watsonia iridifolia var. o'brieni. Plate 70, Figs. 417 and 418. Chart 277 722 Starch of Watsonia meriana. Plate 70, Figs. 419 and 420. Chart 278 723 Differentiation of Certain Starches of the Genus Watsonia 725 Notes on the Starches of Watsonia 725 Genus Tritonia (Montbretia) 726 Starch of Tritonia crocata. Plate 71, Figs. 421 and 422. Chart 279 726 Starch of Tritonia crocata var. lilacina. Plate 71, Figs. 423 and 424. Chart 280 725 Starch of Tritonia crocata var. rosea. Plate 71, F'igs. 425 and 426. Chart 281 729 Starch of Tritonia sccurigera. Plate 72, Figs. 427 and 428. Chart 282 730 Starch of Tritonia pottsii. Plate 72, Figs. 429 and 430. Chart 283 732 Starch of Tritonia crocosma>flora. Plate 72, Figs. 431 and 432. Chart 284 733 Differentiation of Certain Starches of the Genus Tritonia 735 Notes on the Starches of Tritonia 736 Genus Freesia 736 Starch of Freesia refracta var. alba. Plate 73, Figs. 433 and 434. Chart 285 736 Starch of Freesia refracta var. leichtlinii. Plate 73, Figs. 435 and 436. Chart 286 738 Differentiation of Certain Starches of the Genus Freesia 739 Notes on the Starches of Freesia 740 Genus Antholyza 740 Starch of "Antholyza crocosmoides. Plate 73, Figs. 437 and 438. Chart 287 740 Starch of Antholyza paniculata. Plate 74, Figs. 439 and 440. Chart 288 742 Differentiation of Certain Starches of the Genus Antholyza 743 Notes on the Starches of Antholyza 743 Genus Crocus 743 Starch of Crocus susianus (Cloth-of-Gold). Plate 74, Figs. 441 and 442. Chart 289 744 Starch of Crocus versicolor (Cloth-of-Silver). Plate 74, Figs. 443 and 444. Chart 290 745 Starch of Crocus var. (Baron von Brunow). Plate 75, Figs. 445 and 446. Chart 291 747 Differentiation of Certain Starches of the Genus Crocus 748 Notes on the Starches of Crocus 750 Genus Romulea 750 Starch of Romulea rosea var. speciosa. Plate 75, Figs. 447 and 448. Chart 292 750 Genus Cypella 751 Starch of Cypella herberti. Plate 75, Figs. 449 and 450. Chart 293 751 Genus Marica 753 Starch of Marica gracilis. Plate 76, Figs. 451 and 452. Chart 294 753 Genus Gelasine 754 Starch of Gelasine azurea. Plate 76, Figs. 453 and 454. Chart 295 754 Genus Sparaxis 756 Starch of Sparaxis grandiflora alba. Plate 76, Figs. 455 and 456. Chart 296 756 Starch of Sparaxis var. (Albertine). Plate 77, Figs. 457 and 458. Chart 297 758 Differentiation of Certain Starches of the Genus Sparaxis 759 Notes on the Starches of Sparaxis 760 Genus Ixia 760 Starch of Ixia speciosa. Plate 77, Figs. 459 and 460. Chart 298 760 Starch of Ixia viridiflora. Plate 77, Figs. 461 and 462. Chart 299 762 Starch of Ixia var. (Emma). Plate 78, Figs. 463 and 464. Chart 300 763 Differentiation of Certain .Starches of the Genus Ixia 764 Notes on the Starches of Ixia 764 Genus Babiana 765 Starch of Babiana var. (Violacea). Plate 78, Fig. 465 and 466. Chart 301 765 Starch of Babiana var. (Athraction). Plate 78, Figs. 467 and 468. Chart 302 766 Differentiation of Certain Starches of the Genus Babiana 767 Notes on the Starches of Babiana 768 Notes on the Starches of Iridacea,-. Charts 303 to 319 768 Starches of Mdsace^ 771 Genus Musa 771 Starch of Musa eavendishii, obtained from the stalk. Plate 79, Figs. 469 and 470. Chart 320 771 Starch of Musa eavendishii, obtained from green fruit. Plate 79, Figs. 471 and 472. Chart 321 773 CONTENTS. XV Starches of Mdsace^. — Continued. page Gonus Musa. — ContinuiHl. Starch of Musa sapiontum, obtaiiicil from the stalk. Plate 79, Figs. 473 and 474. Chart 322 774 Starch of Musa ensote. Plate SO, Fig.s. 475 and 476. Chart 323 777 Differentiation of Certain Starches of the Genus Musa 776 Notes on the Starches of Musa 778 Starches of Zingiberace.e 779 Genus Zingiber 779 Starch of Zingiber otticiiiale. Plate SO, Figs. 477 and 478. Chart 324 779 Starch of Zingiber otlicinale var. Jamaica No. 1. Plate SO, Figs. 479 and 480. Chart 325 781 Starch of Zingiber otlicinale var. Jamaica No. 2. Plate 80, Figs. 479 and 480. Chart 326 782 Starch of Zingiber olficinale var. Cochin. Plate 81, Figs. 481 and 482. Chart 327 784 Differentiation of Certain Starches of the Genus Zingiber 785 Notes on the Starches of Zingiber 786 Genus Hedychium 787 Starch'of Hedychium coronarium. Plate 81, Figs. 483 and 484. Chart 328 787 Starch of Hedychium gardnerianum. Plate 81, Figs. 485 and 486. Chart 329 788 Different iation of Certain Starches of the Genus Hedychium 790 Notes on the Starches of Hedychium 790 Genus Curcuma 790 Starch of Curcuma longa. Plate 82, Figs. 487 and 488. Chart 330 791 Starch of Curcuma petiolata. Plate 82, Figs. 489 and 490. Chart 331 792 Differentiation of Certain Starches of the Genus Curcuma 794 Notes on the Starches of Curcuma 794 Notes on the Starches of Zingiberaccse. Charts 332 to 334 794 Starches of Cannace.e 796 Genus Canna 796 Starch of Canna warszewiczii. Plate 82, Figs. 491 and 492. Chart 335 796 Starch of Canna roscoeana. Plate 83, Figs. 493 and 494. Chart 336 798 Starch of Canna musafolia. Plate 83, Figs. 495 and 496. Chart 337 800 Starch of Canna eduHs. Plate 83, Figs. 497 and 498. Chart 33S 801 Starch of Canna var. (Konigen Charlotte). Plate 84, Figs. 499 and 500. Chart 339 802 Starch of Canna var. (President Carnot). Plate 84, Figs. 501 and 502. Chart 340 804 Starch of Canna var. (L. E. Bally). Plate 84, Figs. 503 and 504. Chart 341 805 Starch of Canna var. (Mrs. Kate Grev). Plate 85, Figs. 505 and 506. Chart 342 806 Starch of Canna var. (Jean Tissot). Plate 85, Figs. 507 and 508. Chart 343 807 Starch of Canna var. (J. D. Eisele). Plate 85, Figs. 509 and 510. Chart 344 808 Differentiation of Certain Starches of the Genus Canna 809 Notes on the Starches of Canna 812 Starches of Marantace^ 813 Genus Maranta 813 Starch of Maranta arundinacea. Plate 88, Figs. 523 and 524. Chart 345 815 Starch of Maranta arundinacea var. No. 1. Plate 88, Figs. 525 and 526. Chart 346 818 Starch of Maranta arundinacea; var. No. 2. Plate 88, Figs. 527 and 528. Chart 347 820 Starch of Maranta massangeana. Plate 89, Figs. 529 and 530. Chart 348 822 Starch of Maranta leuconeura. Plate 89, Fig. 531. Chart 349 824 Starch of Maranta musaica. Plate 89, Figs. 533 and 534. Chart 350 826 Differentiation of Certain Starches of the Genus Maranta 829 Notes on the Starches of Maranta S31 Genus Calathea 831 Starch of Calathea lietzei. Plate 90, Figs. 535 and 536. Chart 351 831 Starch of Calathea vittata. Plate 90, Figs. 537 and 538. Chart 352 832 Starch of Calathea wnotiana. Plate 90, Figs. 539 and 540. Chart 353 833 Starch of Calathea vandenheckei. Plate 91, Figs. 541 and 542. Chart 354 834 Differentiation of Certain Starches of the Genus Calathea 836 Notes on the Starches of Calathea : 837 Genus Stromanthe 837 Starch of Stromanthe sanguinea. Plate 91, Figs. 543 and 544. Chart 355 837 Notes on the Starches of Marantaceie. Charts 356 to 358 838 Starches of Ntmph«ace« 840 Genus Nymphxa 840 Starch of Nympha;a alba. Plate 91, Fig. 545. Chart 359 840 Starch of Nymphaa marliacea var. albida. Plate 92, Figs. 547 nnd 548. Chart 360 842 Starch of Nymphsea marliacea var. carnea. Plate 92, Fig.s. 549 and 550. Chart 361 843 Starch of Nymphaea gladstoniana. Plate 92, Figs. 551 and 552. Chart 362 844 Starch of Nymphsa odorata. Plate 93, Figs. 553 and 554. Chart 363 845 Starch of Nymphaea odorata var. rosea. Plate 93, Figs. 555 and 556. Chart 364 846 Differentiation of Certain Starches of the Genus Nymphaea 847 Notes on the Starches of Nymphaea 848 Genus Nelumbo 849 Starch of Nelumbo nucifera. Plate 93, Figs. 557 and 558. Chart 365 849 Starch of Nelumbo lutea. Plate 94, Figs. 559 and 560. Chart 366 850 Differentiation of Certain Starches of the Genua Nelumbo 851 Notes on the Starches of Nelumbo 852 Notes on the Starches of Nymphaeaceae. Charts 367 and 368 852 XVI CONTENTS. FAOB Starches of Anemonace^ 853 Genus Anemone 853 Starch of Anemone apennina. Plate 94, Figs. 561 and 562. Chart 369 853 Starcli of Anemone fulgons. Plate 94, Fig. 563. Chart 370 854 Starch of Anemone blanda. Plate 94, Fig. 504. Chart 371 855 Starch of Anemone japonica. Plate 95, Figs. 565 and 506. Chart 372 856 DilTerentiation of Certain Starches of the Genus Anemone 857 Notes on the Starches of Anemone 858 Starches of Delphinace^ 859 Genus Aconitum 859 Starch of Aconitum napellus. Plate 95, Figs. 567 and 568. Chart 373 859 Starches of Helleborace^ 861 Genus Actaea 861 Starch of Actsea alba. Plate 95, Fig. 569. Chart 374 861 Starch of Actaea spicata var. rubra. Plate 95, Fig. 570. Chart 375 862 Differentiation of Certain Starches of the Genus Aetata 863 Genus Cimicifuga 863 Starch of Cimicifuga racemosa. Plate 96, Figs. 571 and 572. Chart 376 863 Genus Eranthis 865 Starch of Eranthis hyemalis. Plate 96, Figs. 573 and 574. Chart 377 865 Notes on the Starches of Helleboraceaj. Charts 378 to 380 866 Starches op Ranunculace.s; 867 Genus Ranunculus 867 Starch of Ranunculus bulbosus. Plate 97, Figs. 577 and 578. Chart 381 867 Starch of Ranunculus ficaria. Plate 97, Figs. 579 and 580. Chart 382 868 Differentiation of Certain Starches of the Genus Ranunculus 870 Notes on the Starches of Ranunculus 870 Genus Adonis 870 Starch of Adonis amurcnsis. Plate 96, Figs. 575 and 576. Chart 383 870 Notes on the Starches of Ranunculacea; 871 Starches of Crdciferace.e 872 Genus Cochlearia 872 Starch of Cochlearia armoracia. Plate 97, Figs. 581 and 582. Chart 384 872 Starches of Euphorbiace.«! 874 Genus Jatropha 874 Starch of Jatropha curcas. Plate 98, Figs. 583 and 584. Chart 385 874 Genus Manihot 876 Starch of Manihot utilissima (commercial tapioca). Plate 98, Figs. 585 and 586. Chart 386 876 Notes on the Starches of Euphorbiacca; 877 Starches op PRrMULACE.E 878 Genus Cyclamen 878 Starch of Cyclamen repandum. Plate 99, Figs. 589 and 590. Chart 387 878 Starch of Cyclamen coum. Plate 99, Figs. 591 and 592. Chart 388 880 Differentiation of Certain Starches of the Genus Cyclamen 881 Notes on the Starches of Cyclamen 881 Starches op Solanacea 882 Genus Solanum 882 Starch of Solanum tuberosum. Plate 100, Figs. 595 and 590. Chart 389 882 Starches of Convolvdlace.e 884 Genus Batatas 884 Starch of Batatas edulis. Plate 100, Figs. 597 and 598. Chart 390 884 Starches of Gesnerace^; 886 Genus Gesneria 886 Starch of Gesneria tubiflora. Plate 100, Figs. 599 and 600. Chart 391 886 Genus Gloxinia (Sinningia) 888 Starch of Gloxinia var. Plate 101, Figs. 601 and 602. Chart 392 888 Notes on the Starches of Gesneracea; 890 Starches of CtrctiRBiTACE^ 890 Genus Trianosperma 890 Starch of Trianosperma ficifolia. Plate 98, Figs. 587 and 588. Chart 393 890 Starches of Cycadace.i; 892 Genus Cycas 892 Starch of Cycas revoluta. Plate 101, Figs. 603 and 604. Chart 394 892 Starch of Cycas circinalis. Plate 101, Figs. 605 and 606. Chart 395 894 Differentiation of Certain Starches of Genus Cycas 895 Notes on the Starches of Cycas 896 Genus Dioon 896 Starch of Dioon edule. Plate 102, Figs. 607 and 608. Chart 396 896 Genus Zamia 898 Starch of Zamia integrifolia. Plate 102, Figs. 609 and 610. Chart 397 898 Notes on the Starches of Cycadacea;. Charts 398 to 400 900 PART I. THE STARCH-SUBSTANCE AND STARCHGRAIN. THEIR FORMS, STRUCTURE, MECHANISMS OF FORMATION, CLASSIFICATION, PROPERTIES, COMPOSITION, DECOMPOSITION PRODUCTS, SYNTHESIS, RELATED BODIES, METHODS OF DIFFERENTIATION. AND DIFFERENTIATION AND SPECIFICITY IN RELATION TO GENERA, SPECIES, ETC. GENERAL APPLICATIONS OP THE RESULTS OF THIS RESEARCH. CHAPTER I. INTRODUCTION. ORjECTS OF THE RESEARCH. This research was undertaken with tliree primary objects in view: First, to deter- mine if the hypothesis underlying the hemoglobin in\'estigation would be supported by the stereochemic peculiarities of other complex synthetic metabolites; second, to add materially to our knowledge of one of the most important substances in the life-history of both plant and animal kingdoms; and third, to throw open fields of investigation which offer extraordinary promise, particularly in adding to our knowledge of the all-important properties of protoplasm. In the previous research it was clearly demonstrated that hemoglobin exists in isomeric forms which are specifically modified in relation to genera and species, and differing to so marked a degree that by means of such variations species and genera could be recognized. Starch was selected as the second subject of investigation, primarily, as stated in the Preface, because of its being so different from hemoglobin in its relations to the parent substance, in its chemical composition and con- stitution, and in its role in life's processes; and, moreover, because of its especial value in such an inquiry owing to its extremely large and complex molecule and high carbon content. THE STARCH-GRAIN. Starch exists in situ both as " soluble starch " and starch-grains or granules, almost solely as the latter. In the soluble form it has been found in the cell-sap of the epidermal cells of Saponaria officinalis and Arum ilalicum, and in a number of other plants. Starch- grains have long been recognized as occurring in a great variety of forms. Fritzsche, in 1834, recorded that not only are the starches from different plants not alike, but also that often the form was so characteristic as to determine the plant, or at least indicate the genus and family to which the plant belonged. This observation was confirmed by Schlei- den and other investigators. Nageli, the most noted authority on starches, stated in his elaborate memou- pubhshed in 1858, after examining over 1,200 different kinds of starches, that all genera of a natural order frequently contain starch-grains which are closely related, and that sometimes a distinction is shown between genera, and then usually also between species, so that plants may be classified into natural groups according to the structure of their starches. Nageli's records also show that starches of like gross histological characters may be found in entirely different genera, families, and orders; and, moreover, that even in the same family the starches of different genera may differ more or less markedly, or, indeed, be entirely unlike. Even though such differences in form exist, it is by no means necessarily implied that such variations are an expression of different kinds of starch, because they may be entirely accounted for upon the basis of attendant conditions. The starch-grain, excepting perhaps at its earliest stage of formation, is a sphero- crystal, and is produced by specialized protoplasmic bodies which are designated plastids. The forms of the starch-grains are dependent in part upon the molecular peculiarities of the plastids, the form of the plastid, and the position of the plastid in relation to the sur- face of the grain; and in part upon a number of incidental conditions, such as variations in the composition of the cell-sap, the presence of foreign bodies in the form of crystals of protein or other solid masses, mutual pressure of starch-grains, etc. Were the plastid 1 Z DIFFERENTIATION AND SPECIFICITY OF STARCHES. free to operate in the absence of disturbing external conditions, the histological char- acteristics of the grain would doubtless be specific in relation to the peculiarities of the plastid; but the conditions in the plant are ever-changing; and, as is well known, a given substance being deposited in a crystalline or a non-crystalline state under varying attendant conditions may take on different forms. The cell-sap is not only of varying composition in different plants, but also changing from time to time in the same plant or cell, which factor of itself may materially modify the shape of the grain. Such an influence is illustrated in the crystallization of calcium oxalate, which substance may be found in a variety of forms in different plants and differ- ent parts of the same plant, and modifiable by changes in conditions in the given part of a plant in which the crystals arise. It may be seen as needles, rods, crypto- crystalline forms, rosette aggregates, tetragonal pyramids, rhombic plates, combination of pyramid and prism, combination of rhombohedron and hemipyramid, and in various other forms which are stated to belong to the tetragonal or monocUnic systems. In many plants the form of these crystals is quite as distinctive of the plant as the form of the starch- grain, yet the differences are attributed to extraprotoplasmic conditions, and not to any inherent molecular peculiarities of the substance itself. There are many instances on record to show that such variations may be dependent solely upon the presence of certain substances in the mother-liquor, or upon other incidental conditions: Thus, for instance, in the case of sodium chloride, which under ordinary conditions crystallizes in cubes, but which in the urine or in a solution of urea appears as octohedrons; or in ammonium fluo- silicate which crystallizes from an aqueous solution in hexagonal forms at 6° C, in cubes at 13°, and in both forms at intermediate temperatures. Not only are starch-grains during their growth subjected to the influences of the pecuUarities of composition of the cell-sap, but also to changes in temperature which may amount to many degrees, and also to alterations in the plastids themselves, which are ever influenced in their operations by variations in both internal and external conditions. It seems, therefore, obvious that the histological peculiarities of different starches may be expressive merely of variations in the conditions attending starch formation, with- out in the least indicating differences in the starch molecules themselves any more than differences in the forms of geometric models indicate differences in the wood of which they are made. It follows, as a consequence, that if there are inlierent differences in the starches of different plants, as was found in hemoglobin in relation to animals, such variations must be determined by methods that are available for the differentiation of isomers. STEREOCHEMISTRY AND SOME OF ITS APPLICATIONS. Modern science has brought to Ught extraordinarily important disclosures in expla- nation of the differences in the properties of isomerides, and this development has resulted in the formation of a new department of chemistry that is known as stereochemistry, but which as yet seems to be but little understood among the rank and file of students. As the laws and principles of this new science underlie the present series of investigations, it seems desirable to set forth certain basic facts which wUl serve particularly to show: (1) that it is theoretically possible for a complex carbon compound, such as starch or hemoglobin, to exist in a countless number of stereoisomeric forms; (2) that the slightest alteration in the configuration or arrangement of the component units of a molecule may give rise to a change of properties that may be profound, and sometimes of a predictable specific character; (3) that stereochemistry is inseparably associated with protoplasmic processes, and hence with the problems of nutrition, species, disease, heredity and the innumerable manifestations of protoplasmic activity which in the aggregate constitute life. For many years the chemical composition of substances was expressed by their molec- ular forviulce, but in the course of time this was found to be inadequate to designate INTRODUCTION. 3 certain peculiarities of substances which experience determined to be owing to differences in cliemical constitution, or the manner of linkage of the molecular components, and hence sliould be expressed by stnidwal formula. Still later it was recognized that even the structural formuhr, which obviously indicate the relations of the atoms and groups of the molecule to one another in only two dimensions of space, are insufficient, and that a full conception of the causes underlying the differences in isomers rests in the recogni- tion of the arrangements of the molecular components or units in the three dimensions of space or, in other words, to be recorded by space formula}. Stereochemistry, therefore, treats of the physical and chemical properties of atoms or structural units in space of three dimensions, the arrangement of the component units of the molecule being expressed by the term configuration of the molecule. This branch of chemistry seems to have had its inception in an article by Wollaston in 1808, in which he states that in order to understand the mutual relations of atoms we must have a geo- metrical conception of their arrangements in all dimensions of space. But Uke many a gem, it lay by the wayside unnoticed. Many years later (1848) Gmelin wrote that no contro"\-ersies as to the mode of writing formulae could be settled without the support of a recognized conception of the arrangements of the atoms of the molecule. After another such long period (1872), Wislicenus forcefully pointed out the insufficiency of structural formulae, and suggested that the deficiencies in our conceptions of the relations of atoms could be accounted for in "the different positions of atoms in space." About a decade previous to this (1861), Pasteur made the first substantial step in la3'ing the foundation of stereochemistry by his investigations with tartaric acids, in which he found that under proper conditions there could be obtained three kinds of tartaric acid differing in crystalline form, and in optical, chemical, and physiological properties. Two forms of these crystals he determined to be enantiomorphous, that is, related to one another as an object is to its mirror-image, the hemihedral faces which appear on the right side of one of the forms appearing on the left side of the other. Crystals that were composed of equal quantities of these substances showed an absence of hemihedrism. TMien the enantiomorphous forms were examined in polarized light it was found that they were oppositely but equally optically active, one form rotating the ray of polarized light to the right (dextro-rotatory), and the other to the left (Isevo-rotatory) ; and that when they were present in equal quantity the acid was inactive (racemic). When solu- tions of ordinary tartaric acid (dextro-tartaric acid) and of racemic tartaric acid (in the form of tartrates, with a little albumin), were subjected to the influence of PeniciUium glaucum, it was found, upon testing with the polarimeter, that as fermentation proceeded, in the first solution optical activity decreased, and that the second or optically inactive solution became active, but herein the plane of polarization was rotated to the left, or in the opposite direction, and hence was laevo-rotatory. As fermentation progressed, optical activity was correspondingly increased, and reached its highest limit when fermentation ceased. Hence, in the racemic solution the acid was split, the dextro-tartaric acid being consumed by the PeniciUium as in the first solution, leaving the Isevo-acid. All three acids are in chemical composition and structural formula identical, yet, as will be observed, they differ markedly, not only in their crj'staUine forms and optical properties, but also to an extraordinary degree in their physiological pecuharities. This latter phenomenon, as will be pointed out, is in its applications one of the most important in the whole domain of protoplasmic processes. The principles underlying the work of Pasteur were developed in 1874 by van't Hoff and Le Bel, working independently, who found that every optically active carbon com- pound contained at least one asymmetric carbon atom, that is, an atom which has attached to it four dissimilar elements, groups, or masses; and, moreover, that in order to explain why this asjTiimetry should cause crystaUine and optical asjrmmetry and the accompany- 4 DIFFERENTIATION AND SPECIFICITY OF STARCHES. ing differences in chemical properties, it is necessary to represent the arrangement of the atoms in the three dimensions of space. Wlien substances have the same kintls of atoms and the same number of each of the kinds of atoms, they are isomers and have tlie same molecular formula. If they differ in any of their properties there is a difference in the linkage of their components which is expressed in differences in their structural formula; or, if they have the same structural formula but differ in their properties, while the com- ponents have the same linkage the units differ in their positions in the three dimensions of space, and are then distinguished as stereoisomers. Now, it is an extremely important fact, as pointed out by Fischer, that stereoiso- mers may show far greater differences in their properties than may be obser\'ed among related isomers. As shown by van't Hoff, substances which contain one or more asym- metric carbon atoms are optically active, and it has since been found that every optically active carbon compound contains at least three carbon atoms, one of which is asymmetric. If a compound containing one asymmetric carbon atom be so modified that one of its attached atoms, groups, or masses is substituted by another so as to give rise to a plane of symmetry, as in the case of isoamyl alcohol to form amylen hydrate, when the hydroxyl group is replaced by hydrogen, and the hydrogen by a hydroxyl group, optical activity disappears for the obvious reason that asymmetry no longer exists in the molecule: Isoamyl alcohol. Amylen hydrate. CH3 CH3 I I C2H5— C— H C0H5— C— OH I " I CHoOH CH3 Each stereoisomer, whether it contain one or more asymmetric carbon atoms, is enantiomorphous, or, in other words, there are two complementary forms which are char- acterized by having opposite but equal effects on polarized light, and also by other differ- ences, all of which are due to modifications in the space relations of the components of the molecules. There are therefore two types of optically active stereoisomers which are desig- nated the dextro and lajvo forms in respect to their optical effects. Moreover, inactive isomers may exist if the two enantiomorphous forms be mixed in equal proportion to form a mechanical mixture, as in racemic tartaric acid; and also when the units of the mole- cules of the two forms combine in the molecule in equal (luantity to form a true compound, as in meso-tartaric acid. In the first instance, the substance is inactive because of external compensation, the effect of the dextro-molecule being compensated for by that of the la;vo-molecule; in the second instance the compensation is internal and due to the units which constitute one half of the molecule neutralizing those of the other half. Obvi- ously, those of the first kind can be separated into the two enantiomorphous forms, while those of the second can not. The first are known as racemic substances, and the second as ?/teso-substances. There are, therefore, in the case of tartaric acid, four distinct stereoisomeric forms: dextro-rotatory, lajvo-rotatory, racemic, and meso-tartaric acids, respectively. The dextro-active, the inactive, and the te\'o-active forms may be likened physically to the differences in the three images seen when the head is viewed in the three-sided mirror. When a substance contains a number of asymmetric carbon atoms, the number of possible stereoisomers increases with each addition, and soon becomes considerable, so that there may be many of each of the dextro, Isevo, and inactive forms. When the number is relatively small, as in the members of the aldohexose group which have the formula CH20H.(CHOH)4.COH, in which there are only four such atoms, the theoretical INTRODUCTION. number is sixteen. These are in pairs, twelve of the sixteen are known, and three occur in nature. The relative configurations may be expressed by the following formulae: Unknown. Glucoses. Mannoses. Dextro. Lievo. Dextro. Lievo. CH2OH H.OH H.OH H.OH H.OH COH CH2OH HO.H HO.H HO.H HO.H COH CH2OH H.OH H.OH H.OH HO.H COH CHoOH HO.H " HO.H HO.H H.OH COH CHoOH H.OH H.OH HO.H H.OH COH CH2OH HO.H HO.H H.OH HO.H COH CHoOH H.OH H.OH HO.H HO.H COH CH2OH HO.H HO.H H.OH H.OH COH Guloses. Idoses. Galactoses. Taloses. Dextro. Lsevo. Dcitro. LiBVO. Dextro. Lfflvo. Dextro. Lffivo. CH2OH HO.H H.OH HO.H HO.H COH CHoOH H.Oli HO.H H.OH H.OH COH CH.>OH HO.H" H.OH HO.H H.OH COH CHoOH H.OH HO.H H.OH HO.H COH CH2OH H.OH HO.H HO.H H.OH COH CHoOH HO.H H.OH H.OH HO.H COH CHoOH H.OH HO.H HO.H HO.H COH CHoOn HO.H H.OH H.OH H.OH COH The relative arrangements of the component units of molecules are determined by an arbitrary standanl, because there is no known way of ascertaining the absolute con- figuration. If, therefore, we adopt in any pair a standard formula for the dextro form, the formula of the Ikvo form will, of course, be the mirror-image of that of the first; and the standards for the several pairs will, in their relations to one another, be established upon the same basis. In addition to this differentiation in accordance with the qualita- tive effects upon polarized light, the various dextro and la^vo forms differ in their quanti- tative effects; and, moreover, each form may exist in more than one form. Inasmuch as each asymmetric group has a rotatory value of its own, dextro or Isevo, it seems obvious that when a compound contains a number of such groups the rotatory power must corre- spondingly be influenced. This conception led van't Hoff to the assumption that when a compound contains a number of asymmetric groups the rotatory power is the algebraic sum of the rotational values of the several groups. Both Guye and Walden have con- firmed, but Rosanoff has disputed, this assumption; but the whole subject of the causes of the quantitative differences in optical activity is under discussion. For many years it was believed that the asymmetric carbon atom is specific to living matter, just as was held in respect to organic substances until Wohler in 1828 synthetized urea. He in common with Darwin, van't Hoff, and almost every discoverer or expositor of some great and revolutionary truth or doctrine was subjected to intemperate criticism, and so insistent were his critics that when all attempts failed to locate the mysterious vital entity in vitro, the proposition that the necessary living factor lay in the presence or manipulations of the chemist certainly carried destructive criticism to its ultra-micro- scopic limit. Since Woliler's time, hundreds and thousands of organic substances have been synthetized in vitro; and it has likewise been found that the asymmetric carbon atom is an atom not merely of life, but of nature. In fact, so impressed were chemists with the specificity of the carbon atom in respect to living matter that its very presence was regarded as sufficient of itself to determine whether or not a given substance was of vital origin. This is well-illustrated in the instances of petroleum, bitumen, and asphaltum, whose origin had been a matter of speculation for generations; but which, since Ragusin has shown that petroleum is optically active, has been considered settled. 6 DIFFERENTIATION AND SPECIFICITY OF STARCHES. Starch is given the molecular formula ?i(C(jHio05)/i, and were it known, or had we good reason for believing, that the molecule is a complex polymeride of preformed groups such as CeHjoOs, or of larger or smaller groups, as is commonly stated, the number of pos- sible stereoisomers could readily be determined were the molecular weight known. Assum- ing that the latter is as high as 25,000, as set forth upon the basis of apparent osmotic pressure, the number of stereoisomers would be a matter of hundreds rather than of thou- sands. But, as stated in the preface, there is abundant evidence to be convincing that starch is not a uniform substance and that it may exist in countless different forms, and hence that the word starch should be used to express not only an individual, but also a group of substances which have certain fundamental typical characteristics in conmion, but which differ from each other in individual ways. The existence of starch in a vast number of different chemical forms can be accounted for: (1) in forms of isomers having structural formulse which differ so little that all have the typical characteristics of a given prototype; (2) in forms of stereoisomers; and (3) in mechanical mixtures of two or more isomers or stereoisomers. Starch, like cellulose, is undoubtedly a complex or aggregate of molecular units of great molecular weight, high carbon content, and extreme complexity of molecular configuration. According to the very recent experiments of Fouard (page 83) with a non-colloidal solution of starch, the lowest possible molecular weight is 15,000. Figiues by former observers ranged usually between 20,000 and 30,000. Knaffe prepared a body from glycogen ("arnmal starch") by the agency of chlorine which had a molecular weight of 23,630, from which glycogen was regenerated having a molecular weight of 15,350. Geinsbergen recorded for cellulose 5,508, Bumcke and Wolfenstein 1,944, and Nastukoff 6,480. The very striking relationships between starch, ceUulose, and glycogen in their element- ary composition, molecular formula, color reactions with iodine, decomposition products in the form of dextrins and sugars, and in a number of other ways that need not be specific- ally stated, are very familiar to every biologist. Hence, a knowledge of the chemical peculiarities of one substance becomes of direct value in explaining or indicating peculiar- ities of another. Our information of the chemistry of the starch molecule is very limited, but the common conception that it is a complex polymeride of molecidar units which are in the nature of preformed hexose or other similar groups is not only apparently with- out justification, but is opposed by the logical conception, and by modern literature, that it is an extremely complex labile aggregate of unknown groups of ionic units. The investigations of Cross and Bevan with cellulose, extending during a period of over ten years, have brought to Ught many important facts from which deductions may be made in respect to starch. For instance, these authors have shown that there is not a single cellulose, but a group of celluloses that is divisible into subgroups in accordance with their chemical peculiarities. Their classification is based on: (1) the degrees of resistance to hydi'olytic and oxidative agents ; (2) the per cent yield of fm"f m'ol when decomposed by boUing HCl.Aq; and (3) the elementary composition in respect to the ratio C : 0. The most important findings of a fundamental character may be tabulated as follows: Cotton. Subgroup A. Bleached cotton. Wood cellulose. Subgroup B. Jute cellulose. Cereal cellulose. Subgroup C. Straw cellulose. Elementary com- J q"" position \^^ Furfurol 44.0 to 44.4 SO 1 : 1.13 0.1 to 0.4 No active CO groujw. 1 Pass througli tlie reac- tions involved in their .solutions as xanthate without hydrolysis to soluble derivatives. 43.0 to 43.5 51 1 : 1.18 3.0 to 6.0 Some free CO groups. In xanthate reactions they are partially resolved to alkali- soluble derivatives. 41.5 to 42.5 53 1 : 1.26 12.0 to 15.0 Considerable activity of CO groups. Still less resistant than the preceding subgroup, and more especially the furfurol yielding comjionents which are selectively attacked under cer- tain conditions. Other characters: - INTRODUCTION. 7 These groups were found by Cross and Bevan to "pass by imperceptible gradations into a Iieterogeneous class of natural products which, while possessing some of the char- acteristics of the celluloses proper, are so readily resolved by hydrolytic treatment that they must represent a very different constitutional type or types. To this group of com- plex carboliydrates the name hemicellulose has been assigned," which group is stated to be readily resoh-etl into crystalline monoses. From the foregoing data it is obvious that there are certain forms of cellulose that are not strictly speaking isomeric, and that there are also forms that are isomerides. Cellulose is optically active, and therefore contains one or more asynmietric carbon atoms, and, as a corollary, each isomer may have a number of stereoisomeric forms, the number varying with the number of asymmetric carbon atoms. Reasoning from this, starch may likewise exist as groups which are not strictly speaking isomeric, yet in which the molecules differ so little in their molecular and struc- tural formulae as to have the essential characteristics of a given prototj^je; secondly, each group may be made up of a mmiber of stereoisomers; and thirdly, the starches of nature, as observed in the starch-grains of a given plant, may, as it seems certain, be vari- able mechanical mixtures of two or more different chemical forms. The number of possible starch stereoisomers is entirely problematical. Miescher has estimated that the serum albumin molecule having 40 carbon atoms may have as many as a thousand million stereoisomeric forms. If we assume that the molecular weight of starch is as low as 15,000, and that the molecular formula is nCCoHjoOs)?!, the total number of carbon atoms in the molecule is at least 550. What proportion are asymmetric is unknown, but judging from the relatively high percentage in such compa^ati^•ely simple substances as the aldohexoses, and the striking tendency for protoplasm during the synthesis of organic substances to form bodies with asymmetric carbon and asymmetric nitrogen atoms, it is probable that nearly all are asymmetric. Moreover, if we conceive, as we should upon the present basis of our knowledge, that the molecule is not a poljoneride of preformed atomic groups, but an aggregate of labile groups of ionic units, the possible number of stereoisomers is absolutely inconceivable. It has already been pointed out that a trifling transposition of elements, groups, or masses attached to an as;ymmietric carbon atom may cause a marked change in crystalUne form, and in optical, chemical, and physiological properties; and also that in stereoisomers changes in the configuration of the molecule, however sUght, may give rise to greater dif- ferences than may be shown by isomerides which have entirely different structural formulie. This is a matter of the most profound fundamental importance in connection with proto- plasmic processes, and in it we seem to have the key to unlocking many baffling problems of physiology, toxicology, and pathology — not to speak of those of general biology; nor is it necessary to enter into speculation for illustrations of such applications, because many instances in literature, dating from Pasteur's experunents, are at one's disposal. In earlier pages, experunents of Pastevu* were referred to which showed the marked differences in the dextro-, laevo-, and racemic tartaric acids m relation to Penicillium glaucum, the dextro form being consimied, the racemic form being split into the dextro and laevo forms, and the dextro form being used, but the laevo form discarded and re- maining in solution. It has since been shown tliat the dextro forms of glucose, mannose, galactose, and fructose are fermentable, or, in other words, consumed by various kinds of micro-organisms as foodstuffs; while the laevo forms are not at all or but inappreciably affected. Again, when mandelic acid (racemic) is subjected to the action of Penicillium glaucum the compound Ls split, the Itevo form is consumed, but not the dextro form; whereas, in the presence of Saccharomyces elUpsoideus, the dextro form but not the laevo form disappears. With glyceric acid treated with Penicilliimi and Bacillus elhaceticus, respecti\'ely, the primary splitting process is the same as with mandelic acid, but the Penicillium uses the dextro form and the Bacillus the laevo form. 8 DIFFERENTIATION AND SPECIFICITY OF STARCHES. These astonishing manifestations of selectivity, which of course mean important differences in the configurations of the molecules and specific relationshii)s of these differ- ences to the protoplasmic structures and nutritive mechanisms of the organism, must be of the deepest fundamental significance. Not only do these organisms thus distinguish, as it were, between the dextro and Ltvo isomers, but also between different dextro forms. Thus, while 07ily the dextro varieties of hexoses are fermentable, 7iot all of them are fermentable; and those which are exhibit different degrees of fermentability that vary in specific relationship to their molecular configurations. As has been shown, dextro-glucose, dextro-mannose, and dextro-fructose are closely alike in their configurations, and all three are attacked with readiness; dextro-galactose and dextro-talose are likewise closely related, but the former is fermentable by means only of a specially prepared yeast-juice, while the latter is, so far as known, unfermentable. All of these differences are readily explainable upon the basis of the existence of slight differences in the relative positions of OH groups. The enzyme emulsin is without action on both a-methyl dextro-glucoside and a-methyl dextro-galactoside, but it breaks down /3-methyl dextro-glucoside and (3-methyl dextro-galactoside; whereas, the destruction by yeast is the reverse, the former substance being metabolized but not the latter. Studies of the properties of stereoisomers from the aspects of toxicity and general physiological actions have afforded interesting and even startling examples of the as yet little appreciated importance of molecular configuration of substances in their relations to protoplasmic processes. A few instances hastily gathered from the researches of various investigators are all that are required at this junctm-e. Brion, in experiments on dogs with the dextro-, laivo-, meso-, and racemic tartaric acids, found marked differences in the extent to wliich they were oxidized. He recorded that the lajvo and meso forms are destroyed in equal degree, the dextro form to a less degree, and the racemic form to a very slight degree. Pohl, two years before, had also found, in experiments on rabbits and dogs, that tartaric acid (dextro form) is only to a slight extent destroyed in the system. Neuberg and Wohlgemuth, in investigations on rabbits with dextro-, laevo-, and racemic arabinose, re- corded that when these substances were given by the mouth the percentages which disap- peared were in order, Isevo-, racemic, and dextro-arabinose ; and they also noted that only the Isevo form gave rise to glycogen formation. Beyerinck recorded that the nutritive values of the stereoisomeric tartaric, fumaric, and maleic acids may be markedly different. Nagano noted differences in the absorbability of certain kinds of sugars: Thus, a 2.5 per cent solution of dextro-mannose was distinctly less absorbable than the same concen- tration of dextro-galactose, dextro-glucose, and dextro-fructose, while Isevo-xylose was decidedly more absorbable than Isevo-arabinose. With stronger solutions differences were noted in the first three. Mayer found that dextro-mannose is more readily oxidized than Isevo-mannose, but that, unlike the difference noted in the enantiomorphous forms of arabinose, both (instead of only one) led to the formation of glycogen. Neuberg and Mayer, in experiments on rabbits with the three mannose stereoisomers, recorded not only that the dextro form is the best in respect to nutritive value, but also that both the dextro and racenoic forms are glycogen-builders, the latter being partially split, and all three yielding a derivative of the glucose series. McKenzie found that the dextro-^- oxybutyric acid is more reacUly broken down than the lajvo form, and that after giving racemic salts, racemic and lievo acids were excreted in the urine, the dextro acid disap- pearing in tlie body. Instances to show that the configuration of molecules is of great importance in rela- tion to the degree of toxicity and general physiological actions are numerous. Maleic and fiunaric acids are stereoisomers, and in the exijeriments of Isliizuka on dogs it was found that tlie former is distinctly the more toxic, and confirmatory results are recorded by Kahlenberg and True in experiments with Penicillium glaucum and Lupinus albus. INTRODUCTION. 9 Mayer noted, in in^'cstigat.ions with dextro and tevo forms of nicotine on guinea-pigs, tluit the former is only luilf as poisonous as the latter, and tliat the two differ markedly in certain respects in their physiological actions. Poulsson, in studies of the actions of the cocaine group, found that the application of a 5 per cent solution of racemic cocaine gives rise to a quicker and more intense but less lasting action than the same strength of solution of the la!V0 form. Albertoni, in his investigations of the actions of certain drugs on the brain, recorded differences in cinchonine and cinchonidine, which are stereo- isomers, the former being the dextro form and the latter the lajvo form. The latter is the less acti^'e, and its actions differed somewhat from those of the other. Ehrlich and Einborn found that the physiological actions of laivo-hyoscine (scopolamine) and racemic hyoscine (atroscine) differ. Piutti records that dextro-asparagine is sweet and Ltcvo- asparagine tasteless; and Menozzi and Appiani discovered the same peculiarity in gluta- minic acid. Werner and Com-ad, in examinations of odors, noted differences in the esters of the dextro- and Ijevo-trans-hexohydro-terephthalic acids; and Schmidt and Tiemann state that the optically active terpenes are more odorous than the racemic forms. Cushny, in experiments with atropine (racemic hyoscyamine) and its optical isomers dextro-hyo- scj'amine and hevo-hyoscyamine, found "that the two hyoscyamines differ to a marked extent in their pharmacological actions, the lajvo-rotatory natural base possessing a very powerful action on the terminations of the nerves of the salivary glands, heart, and iris, while the dextro-rotatory artificial base is almost devoid of effect on these organs, but exercises a stronger stimulating action on the central nervous system of .the frog. The action of atropine (racemic hyoscyamine) is the resultant of the actions of its compo- nents, la3vo- and dextro-hyoscyamine, and it thus affects the ner\'e terminations about half as strongly as lajvo-hyoscyamine, while possessing a more distinct stimulant action on the central nervous system." In another study in which the dextro-, tevo-, and racemic hyoscines were the subjects, it was recorded that Icevo-hyoscine acts twice as strongly as the racemic form on the secretory nerves of the salivary glands and on the inhibitory nerves of the heart; and that lajvo- and racemic hyoscines have the same effect on the central nervous system in man and animals and on the terminals of the motor nerves of the frog, in which animal, however, they seem to be without effect on the central nervous system. Such facts might be materially amplified, but anyone who knows them and has any real ap])reciation of the meanings of the differences in physiological, pharmacological, and toxicological actions, and of the relations of these differences to the configurations of the molecules, must, if he does not conceive of the stupendous importance of stereochemistry to protoplasmic processes, be absolutely devoid of imagination. Moreover, the fact that a given isomeride may he transmuted into another form under conditions which are identical with or similar to those which exist in plant and animal life is one of apparently the greatest and most widespread fundamental biological impor- tance. It has already been found that many stereoisomers show marked evidences of unstability, tending under certain conditions to be changed with the greatest readiness from one form into another; that the degree of unstability varies, the direction of the change usually being from the more unstable to the less unstable configuration; and that in some, and perhaps in all, instances, the transmutation may be more or less markedly accelerated or inhibited by very simple conditions, as has been shown in the case of cer- tain sugars and related bodies, ethylenes, cobalt-amines, diazo-compounds, chromium and silicon and platinum compounds, and other bodies. Such transformations may be brought about dircdhj, whereby one form is changed into another by a simple trans- position of molecular components; or indirectly, by the formation of an intermediate compound. Direct transmutation may be caused by spontaneous change, in accordance with the laws governing the establishment and maintenance of equiUbrium of solutions; 10 DIFFERENTIATION AND SPECIFICITY OF STARCHES. by mere solution in water; by a slight increase of temperature; by the action of sunlight; by crystallization; by repeated recrystallization; by the actions of certain chemical agents, such as traces of halogens, alkalies, certain acids, etc. It has been found in recent investigations that the transmutation of stereoisomers can be accelerated by bases, by salts generally, and by certain acids, etc.; and inhibited by alcohol, phenol, certain acids, and a number of other bodies. The readiness with which such transformations may occur is instanced in the following cases: Wlien a-glucose in aqueous solution is converted into i3-glucose; when maleic acid is heated slightly above the melting-point and the stereoisomeric fumaric acid appears; when the benzyl-/3-amido- crotonic ester is converted by sunlight into the a form; when the solution of a-methyl- cinnamic acid is crystallized, yielding the j3 form; when the aceto-acetic ester phenyl- hydrazone is by simple repeated recrystallization transformed into a stereoisomer having a nmch higher melting-point and other differences in properties; when in the reversal of the enzymic maltose-dextrose reaction isomaltose appears instead of the initial mal- tose; when maleic acid is converted into fumaric acid by the agency of the combined catalytic actions of sulphur dioxide and sulphureted hydrogen; and when a-glucose in solution is transmuted with great rapidity into /3-glucose by the agency of a mere trace of alkali, etc. Perhaps among all of the instances of the transmutation of stereoisomers that might be quoted, none is of broader and more fundamental importance in its bearings in biology than when under certain conditions there occurs a spontaneous conversion of one form of glucose into another when in aqueous solution. The solution of natural glucose is that of an equilibrated mixture of two stereoisomerides, a-glucose and /3-giucose, the former having the lower solubility, but the higher crystallizabihty and rotatory power. The rotatory power of the pure solution, as found by Dubrunfaut in 1846, lessens with marked slowness, which change has been shown in recent years to be due to the transformation of some of the a form into the fi form. If such a solution be concentrated until crystal- lization occurs, some of the a form crystallizes out, thus causing the solution to be tlirown out of equihbrium, which disturbance immediately tends to be compensated for by a con- version of some of the /3 form into the a form, to replace the moiety of the latter which has passed into sohd form. If, now, the preparation be warmed so as to cause a re-solution of the crystaUine a form there will be a conversion of some of the a form that is in solu- tion into the /3 form until equilibrium of solution is again established. If, now, we suppose that only one or the other of these forms is especially adapted to the requirements of protoplasm, as is to be inferred from the illustrations previously given, it is clear how through such transformations there may be maintained not only an equilibrium of solu- tion, which often is of such vast importance in vital processes, but also a continuous sujiply of pabulum in a form that is especially adapted to nutritive exactions. Such instances indicate not only how readily a given form of stereoisomer may, in order to meet nutritional needs, or because of casual conditions, be transmuted into another form wliich may have markedly different properties, but also that inasmuch as the conditions existent in living matter are of a highly favorable character to stereo- isomeric transformations, that such changes are continually going on, and inferentially of the greatest importance. Coupling the foregoing statements with the recognized extreme labihty or unstability of protoplasm, it requu-es no effort of thought to con- ceive of how the more or less marked and continual changes in conditions, internal and external, to which organisms are subjected may bring about with equal or greater facility transformations in the configurations of the slereochemic units of protoplasm, thus altering to a greater or less extent the physico-chemical mechanism, and in turn giving rise to physiological and morphological changes which may be temporary or permanent, and even heritable. INTRODUCTION. 1 1 Anyone who is familiar with modern biology, general and medical, need not give more than a few moments' thought to the foregoing facts and deductions without being led irresistibly to the application of stereochemistry to the explanation of the infinitude of enzymic actions winch constitute so essential a part of life's processes; of the mechanisms that are concerned in the determination of sex, species, and genera; of the mechanisms of fecundation and heredity; of the changes produced in the labile protoplasm of organisms by the alterations of environment, and their effects on structm'e and function; of the mechanisms which underlie tmnor formation, malformations, and reversions; of the specifi- cities of toxins, hemolysins, and agglutinins; of anaphylaxis; of symbiosis; of bodily idio- syncrasies; of asepsis, infection, and immunity; of the selectivities of different cells and tissues for different substances; of the restrictive activities of hormones in relation to the specific reactive organs; of the specificities of the actions of medicinal substances; of the oft predetermined physiological or pharmacological properties of substances synthetized m vitro; of the extraordinary tolerance of pigeons and dogs to morphine; of the differences in the susceptibility of different species and races of animals to disease; of the seemingly chemical identity of the component parts of different forms of lichens which arise from the same alga;; and of a host of other commonplace and special vital phenomena. These statements, as barren as they are of detail and accompanying facts and obser- vations which would add greatly to their interest, applications, and value, must be ade- quate to subserve the purposes for which they are intended: that is, (1) to show by the results of laboratory experience that even trivial changes in the configurations of the molecules of stereoisomers may cause more or less marked, or even profound, alterations in the properties of substances in theu* relations to protoplasm, and of protoplasm itself; (2) to show that different forms of organisms exhibit specific selectivities for substances in accordance with the configm'ation of the molecules, some organisms utihzing the racemic form and others the Isevo or dextro form, or both, but with different avidities, and also that in a given organism one form is complementary, as it were, to one kind of protoplasm, but not to another; and (3) to give evidence which leads to the inevitable deduction that if there be, as justifiably held, countless stereoisomeric forms of starch, each form differ- ing, however little, in its configuration from that of the others, each must have specific individual properties by which it can, by appropriate means, be absolutely distinguished from the others, whatsoever the number. Pasteur, in explanation of the selectivity of Penidllium glaucum for dextro-tartaric acid, likens it to the mechanical relations of "male and female screws." Fischer, to whom we owe more than to any other individual for our conceptions of the actions of enzymes, states that the explanation of the selectivities of these mysterious entities probably rests in the complementary configurations of the enzymes and the bodies acted upon, an enzyme not affecting any substance that has not a molecular configuration complementary to its own, or, figm-atively speaking, when the adjustments are not like those of "lock and key." Ehrlich would conceive of a similar complementary relationship between the configura- tion of the toxine and that of the protoplasm. And thus one might go on, but this would take us far beyond the necessary restrictions of this memoir. Yet a final word : The history of starch from the moment of the utilization of carbon dioxide and water to form an aldehyde, through the various steps of synthesis of mono- saccharoses, disacchai'oses, and polysaccharoses to the ultimate appearance of starch, and the reversal of these steps when the starch is consumed as food, is upon logical grounds con- ceived to be one continuous and consecutive enzyme action. The enzymes synthetize starch and its precursors, and no other substances, because they can build up only such chemical structures as have configurations complementary to themselves, each tending naturally to build those forms which have the closest configuration; and likewise each analyzes only such substances as have the same stereochemic relationships. If, as stated, protoplasm uses 12 DIFFERENTIATION AND SPECIFICITY OF STARCHES. only such organic substances as have a complementary stereochemic form, it follows, as a corollary, that starch has a corresponding configuration, and that if starch has such a con- formable structure so also must have the enzymes that produce it. In other words, every enzyme formed by any given kind of i)rotoplasm is specifically produced to carry out oper- ations which are directly or indirectly essential to the existence of the protoplasm itself, and must ipso facto bear a stereochemic relationship to its mother substance; therefore, proto- plasm, enzyme, and product have in common the same fundamental stereochemic peculiari- ties. In fact, as the results of these researches go to show, every synthetic organic sub- stance produced by any given kind of protoplasm through the agency of its enzymes has a configuration in agreement with that of the protoplasm. If, as must be admitted, corre- sponding kinds of protoplasm in different organisms differ, the corresponding synthetic metabolites will differ; and, conversely, if the latter differ, so must the former. Hemo- globins, which are corresponding substances, have been shown to differ in specific relation- ship to genera and species; and the same extraordinary phenomenon has been brought to light in respect to starches. Therefore, from specific stereochemic differences in corresponding substances we are led to corresponding specific differences in protoplasm; from these to differences in vital processes in general; and from these in turn to those which characterize life's processes in all of their enigmatic phases. Indeed, it is far from visionary to conceive that through the advances of science along the lines indicated it will be found that inasmuch as any given organism, for instance, gives rise to a number of enzymes which are the essential instruments by which the necessary vital processes are directly or indii'ectly carried out, the sum of the configurations of the enzymes represents a corresponding sum or composite of the configurations of the component units of protoplasm. Therefore, each configuration formula would be as specific of the enzyme and the form of protoplasm as is that of glucose ; and the sum or composite of these formula! would be as specific of the organism as the aldo- hexose group of fornmke is specific of the aldehexoses. Hence, we may logically assume that the time viay come lohen any given form of protoplasm or any given organism may be expressed by the physiological chemist as specifically in terms of configuration formulae as it now is de- scribed by the biologist in terms of morphology. DIFFERENTIATION OF STEREOISOMERS. Our methods for the differentiation of stereoisomers are, on the whole, inexact or even absolutely crude ; and in some instances, as in the preparation, separation, and identi- fication by Fischer and his pupils of members of a single group, such as the aldohexoses, the work is tedious and difficult. Obviously no aid is to be expected by centesimal analysis and, as a consequence, dependence rests upon such procedures as will elicit differences in optical reactions, crystalline form and crystallizability, solubility, melting or gelation point, color, color reactions, digestibility or fermentability, decomposability, toxicity, physio- logical and pharmacological actions, etc. Optically active substances owe their effects on light to either /nier-molecular or inira-molecular arrangements; that is, in the first place, to the arrangements of the molecules in relation to each other, and, in the second place, to the arrangements of the component molecular units within the molecules themselves. Sodium chlorate, for instance, is optically active when in ciystalline form, but inactive when in solution; saccharose is active when in crystalline form, and also when in solution. In the first instance, optical activity is attributable to intermolecular arrangement, the mole- cules being so disposed in relation to one another as to cause asymmetry; in the second instance, activity is due jirimarily to the configuration or asymmetry of the units of the molecule itself; and activity becomes more marked in both instances as the number of asymmetric molecules of the substance present is greater. INTRODUCTION. 13 Starch, like saccharose, is optically active in both solid state and solution. Starch grains are, as a rule, markedly ojitically active, which activity is due in part to the asymmetric molecules, and in part to the intcrmolecular arrangements of the starch in the solid form. If grains of potato starch, for instance, are examined in water with the i)t)larizing microscope, and the slide be gradually heated to the minimal temperature of gelatinization, it will be found that gelatinization and swelling will begin at a given point or points, and that the progress of the process can accurately be traced by the cor- responding disappearance of optical activity. This alteration is obviously due to a break- ing down of the intermolecular structure of the grain; but the gelatinized starch is never- theless still optically active. Brown, Morris, and Millar found the specific rotatory power of 2.5 to 4.5 per cent concentrations of soluble starch at 15.5 C. to be («)„ = -\- 202. In differentiating isomerides which are in crystalline form, the melting-point, or, in other words, the temperature at which the intermolecular organization is broken down, is one of the most important means of distinction. Hence, as the starch-grain is a spherocrystal, the temperature of gelatinization must be regarded as being equally important in the differentiation of starch stereoisomers. In various parts of this memou-, especially in Chapters IV and VI, the various methods employed in the differentiation of starches are given adequate consideration to meet the conditions of the investigation. CONCEPTIONS, METHODS, PLAN, AND CONDUCT OF THIS RESEARCH. Opinions may differ widely in regard to various features pertaining to the concep- tions, methods, plan, and conduct of this research. Thus, it might be held that such differences as have been noted in different starches might be attributable to mixtm'es of a unit substance which may exist in multiple forms having variable physical but not chemical properties, but such a view finds insignificant support in the literatm'e of starch, or in the results of this research; while, on the other hand, the evidence points positively to the existence of a vast number of stereoisomers, which, however, may be multiple in the starch of any species or even in any given starch-grain. As regards the methods selected, choice was based in part upon the results of recorded work, and in part upon deductions upon general principles. The histological method, in use for two centuries, has shown that starch-grains from different plants appear in a great variety of forms, and that peculiarities of form may be definitely related to the genus or species, etc.; the polariscope has demon- strated differences in the degrees of polarization, differences in the interference figure or "cross," and differences in reaction in the presence of the selenite plate; iodine and anilines have been found to vary in their reactions with different starches ; temperatures of gelatiniza- tion have been recorded by Lippmann and others as ranging witliin wide limits; the degrees of digestibility of both raw and boiled starches are stated by different authors to vary markedly; and a number of "swelling reagents" have been found to elicit differences in minute histological structure, as well also in the intensity of action. In the preliminary work of this research it was found that digestion experiments, whether upon raw or boiled starch, were of no value in detecting differences in the starch- substance per se. Apart from this method, all of the others enumerated were used. The selection of the particular anilines and swelling reagents was based to only a minor degree upon the results of experience. It is quite probable, indeed almost certain, that better results are to be obtained by other reagents; but this must be shown by experience. It has been taken for granted that the phenomenon of swelling, whether brought about by heat or chemical reagents, is a manifestation of adsorption affinity that is a specific prop- erty of the starch, and variable in relation to the molecular configuration and intermolec- ular structure of the different hypothetical stereoisomers; and, conversely, that quanti- tative and qualitative differences in the reactions to these agents are indicative of corre- 14 DIFFERENTIATION AND SPECIFICITY OF STARCHES. spondingly different stereoisomers, which is entirely consistent with the facts embodied in the literature of adsorption-affinities — or whatever term one may use to express the very conspicuous and important residual-affinilics that exist in substances in which, according to the laws of stoicliiometry, the affinities are satisfied; or, in other words, in which there exists chemical saturation. Neither heat nor any of the other gelatinizing and swelling agents used (chloral hydrate, chromic acid, pjTogallic acid, ferric chloride, and Purdy's solution), except probably chromic acid, causes any apparent notable degree of decomposition of the starch molecule during the periods of observation; but they do change the raw starch into a distinctly different constitutional form by altering the intermolecular arrangements. The individuality in the behavior of the chromic acid, coupled with its peculiarities, leads to the belief that the first effect of this reagent is a constitutional transmutation, as above stated, which is probably followed by a series of oxidation processes. No doubt more or less de- comjiosition is brought about sooner or later by pyrogallic acid and ferric chloride, both of which substances are in concentrated solution; laut with Purdy's solution the decomposi- tion, if any, that may be caused by the small amount of alkaline hydrate present need not be considered. It is manifest from this that these methods of gelatinization represent essen- tially a group of a number of different groups of chemical methods which are available for the study of the changes in the starch molecule and its derivatives tlirough all of their ramifications of changes of constitution and decomposition to the ultimate CO2, H2O, etc. Doubt may exist as to whether or not the primary objects of the research would not have been better accomplished by a study of a less number of specimens by a larger num- ber of methods, especially by methods akin to those employed by Fischer and his pupils in the separation and identification of stereoisomers, and by Cross and Bevan in the study of cellulose. But this alternative plan, for certain reasons, was practically impossible; and, as regards the one adopted, it was believed that a few methods would be sufficient to elicit such fundamental differences as might be necessary in a preliminary investigation, and that the larger the number and variety of starches the better the idea of what is to be expected from the really serious research that must follow. Then again, it might seem that a study of a large number of members of a single genus, or of representatives of quite a number of genera of only a single family, might be .sufficient as an index of what is to be anticipated in other corresponding groups; but the wisdom of the plan adopted of examining in a number of instances, as in Liliacecp, Amaryllidacea, and Iridacece, representatives of a dozen or more genera, and as in Lilium, Narcissus, and Iris quite a number of species and varieties, is clearly shown by the records. And it is also evident that the examination of representatives of scattered families and genera has brought fruitful results. ASSISTANTS AND SOURCES OF SUPPLY OF MATERIAL. The very exacting demands upon the head of one of the most important departments of a first-class medical school, and the enormous amount of labor required in an investi- gation of this character, made it necessary for the author to assign certain laborious and routine parts of the work to assistants. At the inception of the research it was believed by the writer, who is not a botanist, that the best results were to be obtained by the con- stant assistance of an expert botanist. To this end an arrangement was made with the late Dr. Louis Krautter, instructor in physiological botany in the University of Penn- sylvania, but whose tragic and lamentable death occurred before the investigation had been started. The author secured the material, laid down the methods and lines of inves- tigation, directed the work, and made the photomicrographs which accompany the histo- logical and polariscopical descriptions, but the routine laboratory studies of the properties of the starches were of necessity but regretfully assigned to assistants. Nearly all of this INTRODUCTION. 15 laboratory work was done by Dr. Elizabeth E. Clark, B. A. (Bryn Mawr), M.D. (Women's Modical College of Philadelphia), who was well fitted for such an investigation by her training in the laboratories of these colleges, and who devoted two years to the investi- gation. Dr. Clark made all of the studies recorded in Part II, with the exception of a group including the studies of Vicia, Lalhyrvs, Quercns, Castanea, Lilium, Tuli-pa, Convallaria, Amaryllis, Crinum, Sprekelia, Hocnianthus, Hymenocallis, Leucoium, Crocus, Sparaxis, Cur- cuma, Maranta, and Zamia, and some incidental investigation here and there. Practically all of the studies included in the latter group, and also the determinations of the tempera- tures of gclatinization, were made by Miss Martha Bunting, B. L. (Swarthmore), Ph.D. (Bryn Mawr), whose painstaking work in various laboratories is so well known to biolo- gists as to render an introduction needless. The author has also received help from Dr. Clark and Miss Bunting and other assistants in getting together the literature quoted, and in other ways; and for three and a half years he has devoted to this research all of the hours that could possibly be taken from the very exacting requirements of pro- fessorial work. It goes without saying that, inasmuch as the author is not a botanist, the botanical data given are based upon other and recognized authorities. Free use has been made of the voluminous work of Engler and Prantl, Die Natiii'lichen Pflanzenfamilien, begun in 1889, but unfortunately still incomplete; of the Index Kewensis; of Warming's Systematic Botany; and of many other standard authorities, especially of the admirable four- volume Cyclopedia of American Horticulture by Bailey, which often has been followed quite closely. The classification and nomenclature of plants are undergoing continual change, and it has been puzzling at times how best to identify plants and to state relationships, but the data of this character found at the heads of the various chapters have been based on the highest authorities. The brief descriptive introductory notes on the different genera, species, etc., will doubtless prove of value to many readers in refreshing their memories, as well as in other ways. Altogether over 300 starches were studied, a goodly number, but ridiculously small when one recalls to mind that the total known number of plant species is over 230,000 (over half of which are Monocotyledones and Dicolyledones) — not to speak of the thousands of varieties, hybrids, etc. Most of the material from which the specimens were prepared was obtained from E. H. Krelage & Son, Haarlem, Holland; James Veitch & Sons, London; Henry A. Dreer and Henry F. Michell & Co., Philadelphia; J. ]\I. Thorburn & Co. and Peter Henderson & Co., New York; Bobbink & Atkins, Ruther- ford, New Jersey; and Reasoner Brothers, Oneco, Florida. Some were secured from or through the Botanical Department of the University of Pennsylvania by the courtesy of Prof. J. M. Macfarlane (to whom the author is indebted for help in many ways), and quite a number from dealers and growers in various parts of the world. GENERAL CHARACTERS OF THE INVESTIGATION AND RECORDS. A glance through the pages of this memoir will be all that is necessary to convince the reader of the superficiality and general crudeness of the investigation; and that the records of the laboratory studies of the properties and differentiation of the various starches are of a purely tentative character and therefore, for comparative purposes, in the nature of merely temporary standards. It is obvious that if data are to be had that are to be accepted as constayits, certain absolutely essential conditions wliich have not been suffi- ciently recognized in this work must be satisfied. As, for instance, every plant from which starch is obtained must positively be identified botanically; the starch must be so pre- pared as to eliminate, as far as possible, without injury to the molecule, all contaminating substances; examinations should in every instance be made with two or more specimens obtained from different plants of the same species ; the influences of age, environment, and 16 DIFFERENTIATION AND SPECIFICITY OF STARCHES. other conditions must be determined; the time-reactions of the effects of each reagent should be repeatedly observed, as often as occasion demands, in the case of each specimen under given conditions of experiment in order to establish mean standards, and so on. However, if the tlii'ee fundamental objects stated in the beginning of this chapter have been fulfilled, and if by thi'owing open the shutters investigators may perceive roads and pathways leading to tlie elucidation of the extraordinarily important relations that exist between stereochemistry and protoplasmic processes, the labor has not been in vain. In the final arrangement of this report it has been found to be a mechanical necessity to carry the detailed laboratory records into a separate volume, and these accounts now compose Part 11 of this publication. This arrangement, it is hoped, will make this mat- ter easy of reference, although the first impression may be that they have been taken out of their more logical position. CHAPTER 11. THE STARCH-SUBSTANCE, AND THE STRUCTURE, FORM, AND MECHANISM OF FORMATION OF THE STARCH-GRAIN. The wide distribution of starch in plant-life, its great food-value to both plants and animals, and its extensive field of usefulness in domestic life and in many of the arts, sciences, and trades, have made this substance a subject of study for generations. In 1S36, Poggendorff, in re\'ievving the literature of starch, wrote: No substance has been more investigated and yet less known. It affords a striking proof of the diffuse manner in which a subject may be treated if it fall into improper hands. After ten years of investigation, in which the most various views have been set up on the nature of starch, during which all its characteristics as a proximate vegetable substance have been discussed, we are little or nothing in advance of the old jwint of view; and, although perhaps we may not be wholly with- out some extension of our knowlctlge on secondary points, we are still entirely without fundamental grounds in proof of our having arrived at the truth. Even as late as 1885 Brown and Heron wrote: There is i)robably no one sul)ject in the whole range of chemistry which has attracted more workers during the last 60 years than the transformation which starch undergoes when submitted to the action of diastase or dilute acids; and in no respect are the opinions of chemists, even at the present time, more at variance. As recently as 1895, Mej^er, in his elaborate memoir on starch, states that our knowledge of the chemical substances which compose the starch-grain and the products of decomposition is very meager. Notwithstanding the accumulation of an exceedingly voluminous literature and the many advances in our knowledge of the chemistry of starch since the time of Poggendorff, there are doubtless not a few biochemists who will hold that the statement of this investi- gator and critic is applicable in a very large measure to our information at the present day. The synthesis, proximate constituents, microscopical structiu-e, and molecular constitu- tion of starch, and the exact processes and products of the disintegration of starch thi'ough the actions of dilute acids, alkalis, heat, oxidizing agents, enzymes, etc., are but few of the very many instances of important problems which remain partially or wholly unsolved. However, there are sufficient data pertaining to the subject-matter of this chapter to show that starch is produced only in certain plants and in certain parts of plants, it being sub- stituted by some analogous substance or substances in non-starch-producing plants or plant parts; that the starch-substance from different sources is not identical, but exists in a number of forms ; that the starch-grain is produced by specific starch-forming structures which are more or less markedly differentiated, not only in different plants but also in cUfferent parts of the same plant ; that the forms and other structural characteristics of the starch-grain differ in certain ways in specific relationship to the peculiarities of the starch- producing structures and other specific intracellular conditions; that the starch-grains pro- duced by any given part of the plant may be very variable in size, form, and other structural features, yet as a whole exhibit a certain type; that the starch-grain is a sphero- crystal, and therefore has properties which render it available for study by certain crystallo- 17 18 DIFFERENTIATION AND SPECIFICITY OF STARCHES. graphic methods; and that the chief or essential primary proximate decomposition prod- ucts, and conversely probably the essential final synthetic substances, are erytlirodextrin, achroodextrin, maltose, and dextrose. VARIOUS VIEWS OF THE NATURE OF THE STARCH-SUBSTANCE, AND OF THE STRUCTURE, FORM, AND MECHANISM OF FORMATION OF THE STAKCH-GRAIN. Although starch was known to the ancients, as shown by the writings of Dioscorides, our knowledge of this substance had its origin practically in the microscopical examina- tion by Leeuwenhoek in 1716, by wliich he differentiated two fundamental structural components. He found that when the grains were heated in water the inner part dis- appeared, leaving nothing but the integuments (Hiillen); and that in the excrement of bu'ds that had fed on grain the same integuments could be found. From these observa- tions he concluded that starch consists of a kernel or nucelus which is fit for nourishment (nalirenden Substanz), and an outer, insoluble non-nutritive envelope. From tliis time until the early part of the last century no material addition was made to the literature of starch, at which period an era of research was initiated by Vaquelin (1811) and Kir- choff (1811), the former finding that when starch is subjected to torrefaction it is converted into a gummy substance soluble in water; and the latter discovering that weak acid changes starch into gum and sugar. The results of these investigations were confirmed by a num- ber of contemporaneous experimenters. During the following twenty-five years, however, very little was added to our knowl- edge apart from certain discoveries relating to the decomposition products and to the agents wliich give rise to them. In 1819, De Saussure (Ann. de chim., 1819, xi, 379) found that after setting raw starch aside for two years there were present a sort of paste and sugar; a substance insoluble in cold water, but soluble in hot water, and giving a blue reaction with iodine; and a body that was insoluble even in hot water, and which closely resembled cellulose. To the substance soluble in hot water he gave the name amidinc, and to the cellulose-like body the name ligneux amylacee. Some years later Raspail (Ann. de sciences natur., Oct. et Nov. 1825, et Mars, 1826; quoted by W. Nageli, Der Starkegruppe, Leipzig, 1874) reported that starch when heated on a plate is converted in part into a sort of gum that is soluble in cold water, leaving a residual insoluble sheath. The former stained blue with iodine, but the latter not. Ras- pail's statement was at once disputed by Caventou (Ann. de chim., 1826, xxxi, 358), who recorded that all parts of the grains stain blue with iodine; that raw starch does not con- tain a substance that is soluble in cold water; and that when starch is subjected to dry heat at 100° it is changed into a soluble substance resembling the amidine of De Saussure, which substance he looked upon as a modified starch (amidon modijie). It was then shown by Guibourt (Ann. de chim. et phys., 1829, xl, 183) that raw starch does contain a substance that is soluble in cold water. He found that when starch- grains have been comminuted in a mortar they are rendered partially soluble in cold water; that this solution yields a blue reaction with iodine, and that upon drying it yields a resi- due that is insoluble in cold water. The soluble substance he identified with the amidine of De Saussure and the gum of Raspail, which latter, however, he holds is not a gum, as stated by Raspail. He records that there is still another component (the integuments) which stains blue with iodine, but is insoluble in cold water. Following Guibourt's article there appeared four contributions by Guerin-Varry (Ann. de chim. et phys., 1834, lvi, 225, and lvii, 108; 1835, lx, 32; 1836, lxi, 66). In the fii'st of these Gu6rin-Varry states that starch consists of tliree substances: One (amr- dine) that is soluble in cold water; another {aniidin soluble) that is insoluble in cold water by itself, but which is held in solution by amidine; and an insoluble substance (amidine tegumenlaire) , the integuments or capsules of the grains. The amidin tegumentaire was STARCH-SUBSTANCE, AND THE STRUCTURE, ETC., OF THE STARCH-GRAIN. 19 fouiul to constitute 2.96 per cent of the starch, while the remaining 97.04 per cent was recorded as being made up of 60.45 parts of amidine and 39.55 parts of amidin soluble. Payen and Persoz (Ann. de chim. et phys., 1834, lvi, 337) regarded the starch-sub- stance as a single body, wliich they refer to as amidone. Gu(5rin-Varry in his second article states that the amidone of Payen and Persoz is a mixture of two substances, one soluble and the other insoluble in cold water; and in his third article he studies the actions of diastase on starch, and the characteristics of starch- paste and the sugar formed, and he refers to the great resistance of raw starch to the action of malt extract. He placed a preparation of raw potato starch and malt extract in a sealed tube, keeping it at a temperature of 20° to 26° C. for 63 days, and found upon microscopic examination at the end of tliis period that the grains showed no alteration. In his fourth article he fm'ther considers the three components of starch, as held by him (see also Chapter III, pages 86 and 177). Fritzsche (Ann. d. Physik u. Chemie, 1834, xxii, 129) seems to have been the first to study the form and structure, and the mechanism of formation of the starch-grain. He, in the first place, takes exception to the assertion of Raspail that every grain is com- posed of two substances, one an enveloping membrane or integument (Hiille) which is insoluble in water, and another part wliich is intraintegumentary and soluble in water. The author's studies were mostly made with potato starch, but confirmed by observations on other starches. He states that the grains of potato starch have a variety of forms and sizes, and that the normal shape is that of a somewhat compressed, oval structure. A uni\-ersal and constant characteristic, he records, is the presence of concentric rings which vary in distinctness, regularity, arrangement, and number. These rings proceed from a spherical point which, on account of its peculiar chemical relation, he terms the kernel or nucleus (Kern) of the gi'ain (in later years and at present known as the hilum) . The hilum was not always found to be located at the center, but might be anywhere in the long axis of the grain. From the transparency of the hilum he was led to believe that it is a funnel- like hole, which he thought proved by the fact that when a grain is compressed between glass plates the hilum retains its position and canals are never observed to run from this point peripherally. About this spherical hilum rings are arranged, the inner ones usually uniformly, and the outer ones spread out more on one side of the grain than the other, producing an oval shape. Every grain he states is composed of as many concentric layers as there are rings. Fritsche then considers the question as to whether the outermost layer is produced first in the form of a skin (Haut), and then the inner part by infiltration; or whether the hilum is the first part formed and the layers deposited on it. This he found most easily answered by examining compound grains which are found among the simple grains, and which he states may be regarded as deviations or monstrosities. Observations of these grains led him to conclude that all grains are formed by the deposition of the outer layers upon the mner. The compound grains he believes have arisen tlirough the union of two grains in apposition by means of the deposition of a conmion combining layer, or by a smaller grain being inclosed by a layer of a larger grain during the process of growth of the latter, such grains always having a small, distinct space present between the fused grains. The outermost layer of the starch-grain was found to have a special density, by which it can resist external influences better than the inner layers and he gives this as the reason why unbroken grains are insoluble in water, and why crushed grains are soluble. The reason for the greater density of the outermost layer he attributes to its being in con- tact for a long time with the cell-sap which contains a large amount of vegetable albumin; or that the grains may be covered with a precipitate that could easily be formed from an albuminous body. He believes that differences in the densities of the layers account for their visibility, and that the variations in density are perhaps owing to the influence of STARCH-SUBSTANCE, A'D THE STRUCTURE, ETC., OF THE STARCH-GRAIN. 21 if starch be rubbed for a pend of half an hour in a mortar with double the volume of water, there is formed a viscidstiff "salve," capable of being drawn into threads. A large number of the granules when sen under the microscope appeared to be crushed in various ways, and partly ground into aiall flakes. The inner layers are combined with more water by friction, exliibiting a finel^floccular or granular but connected mass, which is colored blue with iodine, v,-hile all tfe water surrounding the mass remains wholly uncolored. On heating the grains upon a mall plate he observed that one can trace gradual changes brought about by heating, an that thus may be found the best explanation of the struc- ture of the grains. The fii-stiction is one of drying, by wliich the so-called nucleus or hilum is converted into an ai-bubble. The several layers separate simultaneously, and in consequence of inspissation he lines of separation become sharper, darker, and broader, and also even recognizable biader and narrower layers of air, the layers hanging closer together at some places thanit others. By degrees the separate layers peel away from each other like the scales of abulb. If the grains are boiled in water their outlines grow more and more distinct, but te particles cling together in the form of a paste-like mass. Under the microscope, by mens of iodine, he found that we may recognize the separate and swollen granules, while th water added is never colored blue. He believes from these phenomena that while starch my take up a large quantity of water, it can never be properly dissolved. The layers of the^tarch-grain he regards as being more aqueous as they lie further inward. By testing vth iodine he noted that all parts of the grain are stained equally, and that while there nay be slight differences in the external layers in relation to solvents, differences which le believes arise from the adhesion or infiltration of traces of albumin, fat, or wax, such ifferences merely delay the action of the solvents. Schleiden studied the pheomena of erosion and formation of the grains in the potato. During erosion the grains retaied their solidity to the last moment, and were only gradually attacked from the exterior in-ard, the extremities of the longitudinal sections offering the greatest resistance, on wLh account the grains after a time resembled knotty twigs. In young growing potatoes h found exceedingly minute granules and large grains, the former being the more numeras. He states that if we regard the minute granules as the rudiments of structure, and tke the different sizes as indexes of their age, the younger the granules the more truly aherical they appear, the ovoid or irregular outline being subsequently acquired, and te deviations from the original form being caused by the unequal thickness of the outr layers. The innermost layers continue to exliibit the original spherical form which he youngest granules present. Starch therefore grows, he concludes, by the deposition f outer new layers upon the inner older layers. This he found confirmed when starche from the other plants were compared with grains from the potato, as, for instance, grainirom Diejffenbachia seguine. Schleiden also noted that the forms of starch-grains are excedingly various, and he made a tabular list of grains from various sources based upon hi:ological peculiarities (see page 64), The view that starch-grais grow by the external apposition of layers was asserted by Walpers (Botanische Zeitug, 1851, ix, 329) to be disproved by the fact that in twin grains, which occur so abundatly, the liila are not located near the line of fusion of the component grains, but alway near the circumference; and that the formation of " be- tween-lay ers " which takes pise at the line of contact is not explained by deposition of layers from without. The thory of external apposition he states is contradicted by the fact or assumption that the lost outer layer is the oldest, and the layers within the youngest, in succession from priphery to center. The outer layers he states are usually the densest, yet they must be ain to be able to stretch to attain the maximum size of the grain. Walpers holds that t adjust such an indisputable contradiction, two possible answers present themselves: 1) the outer older layers grow simultaneously with the growth of the interior, which 'an not be the case, because the lamellation indicates an 22 DIFFERENTIATION AND SPECIFICITY OF STARCHES. intermittent growth; or (2) in many kinds of starcli both external and internal lamella- tion go on independently. The latter a.ssumption he holds explains the peculiar structure of starch-grains often found in Chile arrowroot (Alsirceineria), but rarely found in potato starch, in which several common dense layers surround a fused group of starch-granules. Maschke (Jour. f. prakt. Chemie, 1854, lvi, 400), from studies of the structure of wheat starch, concluded that the grains have the form of membranous vesicles, the periphe- ral layer consisting of cellulose, which in starch-paste does not stain blue with iodine. He heated starch in water to various temperatures. At 40° C. the grains or vesicles showed large numbers of well-defined rings which are alternately hght and dark; at 60°, instead of intact rings, outlines of smaller grains were noticed in the center of the large grain; at 70° breaks and cracks were found, due to the swelling of the vesicles; and at 100° the grains were rather irregular in shape, resembling collapsed bags. On the addition of iodine he noted in the blue mass, granular brown-colored lumps. These phenomena can be explained, he states, if one assumes that every grain is composed of from 3 to 5 vesicles placed concentrically one within the other, and between them the granular amylon or starch. The appearance and disappearance of so many rings he attributes to the swelling and the separation of the granular amylon, ultimately leaving the 3 to 5 circles which are the envelopes of as many concentrically arranged vesicles that compose the grain. He believes that he has proved the existence of such an arrangement by means of the action of iodine and sulphuric acid. The internal part of the starch he conceives to consist of a soluble and an insoluble substance, and that by evaporation the former is converted into the latter, whereas by solvents, such as lye or hot water, the reverse occurs. These two modifications of starch, therefore, are believed by him to preexist in the grains. The explanation of the light and dark rings, he writes, is to be found in these two modifications of starch, the light rings representing the insoluble modification and the dark rings the soluble modification. The hilum he describes as a central hollow of the innermost vesicle, which hollow may, owing to drying, be without contents, or it may be filled with amylon in solution. The insoluble starch, he assumes, is present in granules around which the soluble form of starch is present in liquid form. The view that starch-grains grow by external accretion, and that the outer layers are the youngest and last formed, received further support in the investigations of Crliger (Botanische Zeitung, 1854, xii, 41) of the starch-grains in the plant-cells of Philodendron grandifolkim, Dieffenbachia seguine, Batatas edulis, and other plants, the starch of Batatas edulis being especially suitable for the study of compound grains. He endeavored par- ticularly to determine where and how the outer layers are formed. In a number of the plants examined the younger cells contained only round grains, whereas in older plants there were large grains of various irregular forms, from which facts he believes that the forms found in the older cells were round originally and developed into larger grains which depart during growth from the spherical form because of the deposition of layers on the outside. He makes the observation that all starch-grains are formed upon the protoplasmic layer which lines the inner surface of the cell. The starch-grain is formed here, he holds, as long as it is capable of further growth and as long as protoplasm exists in the cell. In all kinds of starches with definite layers and definite eccentric hila it was seen that the liilum is always located in the part of the grain farthest removed from the point of attachment of the grain. When the part of the starch resting upon the protoplasm or chlorophyl was examined it was found that the outside layer at this point behaved differently with iodine from other parts of the grain, for while the mass became blue this outer layer tm-ned yellow or dark brown, the same color as the protoplasm or chlorophyl assumes, but it did not take the stain readily. This outer layer is described as being of varying thickness in different starches, in some hardly perceptible ; and he regards it as a substance which is in the proc- ess of becoming starch, and that it probably contains nitrogen or albuminous material. STARCH-SUBSTANCE, AND THE STRUCTURE, ETC., OF THE STARCH-GRAIN. 23 Small grains treated with iodine were colored yellow or gave no color reaction; and young grains reacted very slowly, and the blue color was not so pronounced as in the older grains. The parenchjana cells of Batatas edulis were found to contain large numbers of single starch-grains. In later stages of development the grains by enlargement approach each other, and ultimately form groups with the flattened surfaces of the grains in contact with each other. At this stage the transition substance was still perceptible, but in older cells it was absent. A compound grain he therefore regards as being nothing more than two or more single grains whicli develop singly, and that later, owing to the disappearance of the transition substance between them and by the development of a common outer envelope, become one compound grain. The di\-ergent views as to whether the starch-grain grows by a stalactite-like deposi- tion upon the outer surface or whether the grain, like the wood- or bast-fibers, increases in size by a deposition watliin, led Harting (Botanische Zeit., 1855, xiii, 905) to an investi- gation, from which he found that in the early stages of development of the starch-grain an integument is present, this being especially prominent in the starch of the potato tuber. This was colored brown with iodine, not blue like the starch-layers beneath it. According to his view the integument belongs to the "epigon cell" which produces the starch-grain. In the mature grain an integument was no longer discernible, all layers now being colored blue with iodine. He believes that it must be this integument, or the integument of the epigonen or mother cell, which holds the groups of grains together in such plants as Smilax syphilitica. The similarity after absorption of water of the "cambial layer," or outer layer of the starch-grain, to the cross-section of the cambial layer which produces wood- and bast-fibers, is so striking that he beUeves one is warranted in the supposition that it is the outermost layer of the starch winch is the cambial wall of the cell, and that the deeper layers are formed later, and that while new layers are being formed on the inner surface of the pre^'iously formed structure, the latter and the cambial wall increase in size through intussusception. Reinsch (Neue Jalirbuch. f. Pharm., 1855, iii, 65) states that the granules of potato starch contain dextrin and sugar already formed, which can be dissolved in water from the pulverized grains. Melsens (Institut, Ire sect., 1857, xxv, 161; quoted by W. Nageli, Der Starkegruppe, Leipzig, 1874) found that when starch-grains were treated with weak acids, pepsin, or dias- tase they become so changed that they no longer yield a blue reaction with iodine, yet retain their form, therefore having a non-starch skeleton or framework. In 1858 there appeared the elaborate monograph of Carl NageU (Die Starkekorner. Morphologische, physiologische, chemisch-physicahsche und systematisch-botanische Mon- ographic, Ziirich, 1858, 25 Tafellen, 625 S.) which covers a wide field of inquiry. To this author chiefly is due the conception, which even to the present day receives almost uni- versal acceptance, that the starch-grain consists fundamentally not only of two substances, granulose and cellulose, but also that they have markedly different properties. The differ- entiation of these components was brought about by subjecting raw starch to the action of saliva for several months at 45° to 55° C. The larger portion of the grains was slowly dissolved, while the remaining part was found to retain the form and structure of the original grains. To the former he gave the name granulose and to the latter cellulose, which latter he thought identical with the substance of the same name of plant structures, and which he regarded, when obtained in this way, as being in its purest form. He looked upon the grains as being mixtures of granulose and cellulose in which these substances are combined in the form of a sort of a diffusion. The proportions of the two substances, he records, differed in different kinds of starch and in the different layers of the same grain, but the quantity of granulose was considerably greater than that of cellulose, the latter being present often in very small quantities, probably representing only one-eighth 24 DIFFERENTIATION AND SPECIFICITY OF STARCHES. of the entire mass. The quantity of celhilose was found to be in direct proportion to the resistance of the grains to swelling and solvent media. For this reason, he concludes, the dense layers contain relatively more cellulose than the soft layers, and the external more than the internal parts. The cellulose, he states, is not insoluble, but difficultly soluble. Besides granulose and cellulose the grains were found to contain water, both in the fresh and air-dried condition, and also "condensed gases;" but usually no other substances were present in perceptible quantities, although sugar, dextrin, and soluble starch may exist in small amount. He thought that large quantities of gases are condensed in the grains, which idea was doubtless suggested by the evolution of gas during decomposition processes. Differences in the reactions of starch granulose and cellulose with iodine were recorded : The granulose was found to take up iodine from a weak solution and to become blue before the cellulose becomes colored, and this occurred even though the iodine solu- tion had to penetrate the cellulose to reach the amylose. While the amylose was colored a bright red to a blue or a black, according to the quantity of starch present, pure cellulose was colored a dull-red, or a brownish-red, which differences formed a means of distinguish- ing one from the other. Studies of the mechanism of development of starch-grains led Nageli to the view of growth by intussusception, and hence to oppose the theory of growth by external accre- tion that was held by many of his predecessors. It must be remembered, he writes, that all grains in every stage of development are solid, and that only in abnormal cases, owing to solution, are they hollow. If growth takes place by deposition on the outside, then the young grain and the inner layers of the large growing grain must be identical in form, structure, and substance. Wliile the forms of the two are very similar they nevertheless differ very markedly, there being present in the interior of large growing grains "layer complexes" which are never found in the small mature grains. In all kinds of starch, without exception, the substances of the young and of the large growing grains are differ- ent. In the large grains there are present, from the periphery to the center, alternate dense and less dense layers. During no stage of the development is there a soft layer on the outer surface, but always in the growing grain a dense peripheral layer. Swelling solvent mecha, such as hot water, acids, and alkalis, act on the inner substance of large grains, as well as on the "part-grains" of "half-compound grains" (see page 60), and dissolve this substance, while the small mature grains are not acted upon. The external layer in aU stages of development of the grain is the same, so that solvents disorganize and dissolve the entu'e inner substance in both large and small grains in a similar manner, but leave the outer layer in the form of a membrane. Young grains up to a certain stage are entirely homogeneous, but after a time layers become evident in the interior of the grains. In other grains, which in earlier stages show no structural differentiation, inclosed part-grains appear whereby such grains become half-compound grains. The phenomena of form and structure of compound grains, Nageli holds, are also in opposition to the apposition theory of growth. The part-grains increase in size by growth along the inner and not the outer radius. If growth occurred from outside there must be pressure exerted on the outside of the grains, but the compound grains show no trace of such external pressure, since the part-grains have sharp edges and corners and flat surfaces. Not only, he states, is growth not by apposition, but the half-compound and many of the compound grains do not originate by the fusion of simple grains, but both are produced by internal processes, the originally homogeneous substance becoming lamellated and divided into part-grains. He regards it unlikely that there may also be a deposition on the exterior of the grain, because the peripheral layer of large and small and young and old grains is identical in its resistance to the action of solvents, and because one would have to assume that the extei-nal layer, as it is covered by the newly deposited substance, changes its nature, and that as the layers are deposited each in turn takes on STARCH-SUBSTANCE, AND THE STRUCTURE, ETC., OF THE STARCH-GRAIN. 25 the properties of the external layer. He believes that the conditions which regulate the growth of the grain by intussusception bear a very close relationship to the physical and chemical jiroportics of the grains. This he sums up in the following law: The more rapid the growtli in any part of the grain, the softer the substance at that point, and the more readilj^ it is acted upon by swelling solvents, and since the chemical composition runs l^arallel with growth, the larger will be the amount of granulose in relation to cellulose. The least growth, Niigeli holds, takes place in tlie exterior part of the grain, as can be seen in both simple and half-compound grains, which grains from the earliest stages do not increase in thickness but grow in the same plane. While the outer layer under all circum- stances is the densest and richest in cellulose content, the small young grains, unlike the mature grains, consist of a dense substance throughout. In these grains there appears at the center a soft hilum which is a very small point, and which increases in size and density, while the surrounding substance does not change much in density. In a later contribution (Botanische Zeit., 1881, xl, 633) NiigeU again discusses the mechanism of formation of starch-grains, and adheres to the theory of growth by intus- susception. He assumes that the differentiation of the substance into lamella) is present in all starch-grains, and that the homogeneous appearance of some grains is due to the optical appliances not being sufficiently powerful to perceive a lamellation which may be exceedingly indefinite and indistinct. That the grains are solid can be demonstrated, he states, in the sprouting of starch-containing plant parts, in which the grains, like inorganic crystals, dissolve from the surface inward until they disappear completely, during all stages of which solution the grains are solid. The variations in the densities of the lamellie he attributes to differences in the water-content. The lamellis are not of uniform thickness throughout, and the inequalities he states may be distributed regularly or irregularly over the entire layer. Very often two or more layers fuse into one, or one layer may split into two or more. Such splitting takes place jjrincipally in the dense layers, but it may occur in the more watery loose layers. A tliick lamella which on one side is simple may on the other side be split into several parts of ec^ual density, between which there may appear parts of less density. Complete and incomplete lamellte are arranged about a common central point which sometimes is the mathematical center of the grain, but usually eccentric for the entire grain and concentric only for the inner layers. The form of the individual layers and the arrangement of the lamellae from the center to the periphery, in other words, the structure of the grain, he states are due to the form of the hilum. In grains with a lenticular hilum the short axis of the grain is in a line with the shortest diameter of the hilum, which is centrally located, and the layers are usually of uniform thickness. In grains with an elongated hilum, the axis of the greatest diameter of the grain is in line with the greatest diameter of the liilum, wliich is central, and the layers are rather uniform. In spherical grains the hilum lies at the center of the grain, and the laj^ers are always circular, and have a uniform tliickiiess. All radii of such grains are equal. Another type of grain is one in which the hilum is spherical and eccentrically located, and in which the axis passes tlii'ough the center, on one side of which it extends through the longer radius and on the other the shortest radius of the grain. Such a type is the potato starch. The principal forms of grains with eccentric nuclei are: (1) spherical with the hilum between the center and the periphery; (2) elongated (lanceolate to conical), with uniform ends or one end more attenuated, the liilum being located at either the narrower or wider end; (3) wedge-shaped, elongated (1 to 3 times as long as wide), compressed on one side, narrow and not compressed on the other side, where the hilum is located; (4) expanded (0.66 to 2.5 as wide as long), compressed, the hilum located near the narrow end; (5) irregular forms with projecting edges and angles. Nageli notes that in some instances lamellated grains have several hila, with as many systems of lamella;. Such a grain appears as an aggregate of se\'eral lamellated part- 26 DIFFERENTIATION AND SPECIFICITY OP STARCHES, grains wliich are surrounded by a common lamellated envelope. A grain of this type may be said to be "half-compound." These part-grains number 2 to 10, rarely 25 to 40, in one grain. If they become relatively large, and the common enveloping layers are thin, a transition type between tlie simple and the compound grain is formed. If the lamella- tion becomes rather indistinct, the half-compound grain resembles a single grain, but it can easily be distinguished from the latter by the presence of several hila. Half-compoimd grains occur usually in grains with eccentric liila. The structure of the part-grains is similar to that of the simple grains. The part-grains of a half-compound grain are usually simple, but rarely are half-compound, each with several liila (see Chapter V). In another part of the memoir Nageli studies the formation of starch-grains in "chlo- rophyl vesicles." Starch-grains, he writes, occur not infrequently in the granular or vesic- ular protoplasmic structure of the cell-contents, but they are constant constituents of the chlorophyl vesicles. The formation of the grains within the chlorophyl, he found, can often be traced. The grain is described as appearing in a homogeneous green mucus in the form of small points that increase in size and often attain sufficient dimensions to be recognized as starch-grains. In some instances these particles remain rather small, and during their entire development are inclosed entirely by the chlorophyl; in other cases they steadily increase in size, gradually push out through the surrounding chlorophyl, and finally lie free. Grains lying together in the same vesicle may become flattened on their contact surfaces and be united as a compound grain. Such stages of development were observed in the leaf of Begonia dichotoma. In the oval chlorophyl vesicles of Nitella syncarpa 3 to 10 white points were seen and recognized as white starch-granules, very minute in size and not becoming larger. In young cells of Chara hispida the chlorophyl vesicles are arranged in series and are polygonal in shape. They contained 1, 2, or 3, rarely 4, starch-grains, which color blue with iodine. In older cells the chlorophyl grains are considerably larger, with their margins more rounded. In such cells the chlorophyl vesicles are entirely filled with starch-gi-ains, containing 1 to 4, rarely 7, starch-grains, which are in contact with each other, and at the surface still covered with a thin layer of chlorophyl. In still older cells the chlorophyl had disappeared entirely. The chlorophyl vesicles are arranged along the ceU-wall, and they are compressed, with the flat surfaces in contact with the cell-waU. The starch-grains at first, without coming in contact with each other, lie side by side in the same plane, and they grow chiefly in the plane parallel with the cell-wall, and be- come more or less tabular. Later they touch each other, and their ends become superim- posed upon each other. The young grains are perfectly homogeneous, but later one or more points are noticed which are probably areas where solution has taken place. Cliloro- phyl vesicles in the leaf parenchyma of Begonia sp. are perfectly homogeneous at fii'st, and consist only of green-colored protoplasm. After they have increased in size one observes from 2 to 7 shining dots. In still larger chlorophyl vesicles there are 1 to 3, very rarely G, starch-grains which are recognized by means of iodine. With the growth of the starch-grains the green protoplasm associated with the formation of the grains becomes more and more replaced by starch, and finally instead of protoplasm one observes only starch-grains. In the younger stages the chlorophyl vesicles have 3 to 7 small dots, while among the mature grains none consists of more than 3 part-grains. The explanation of this phenomenon is that either an unequal number of starch-grains originates in a chloro- phyl vesicle at different periods of growth, or, as in Phipsalis, some of the points which are visible in the early stages are oil droplets; or it may be that the chlorophyl vesicles divide, as in Chara and Nitella. Studies of the solubility of raw starch, and also of the structure of the starch-grain, were made by Jessen (Ann. d. Physik u. Chemie, 1859, cvi, 497), who states that one can easily con\'ince himself of the solubility of raw starch by crusliing the grains of potato STARCH-SUBSTANCE, AND THE STRUCTURE, ETC., OF THE STARCH-GRAIN. 27 starch, maranta starch, etc., in a little water, when the beginning of solution is soon noted, for the entire mass becomes viscous and mucilaginous. On the addition of more water a clear solution is obtained, on whose surface float torn and broken "integuments" of the starch-grains, while the unbroken grains sink to the bottom. The filtrate shows no precipitate, but becomes blue with iodine. Jessen agrees with Harting (page 23) that the starch-grain has a structure like that of cells, and that the concentric lines on the grains are cell-walls or condensed layers between wloich the soluble starch is deposited. In a later research (page 29) Jessen records that starch consists of three substances : the integ- uments, starch soluble in cold water, and starch insoluble in cold water. Wicke (Ann. d. Physik u. Chem. 1859, cviii, 359) holds, in opposition to many previ- ous observers, that pulverized starch-granules are almost entirely insoluble in cold water, only very small portions going into solution. The statement of Nageli that the cellulose of the starch-grain remains as a residue after the solution of the granulose seemed to von Mohl (Botanische Zeit., 1859, xvii, 225) to be not well-founded. Von Mohl therefore undertook a series of experiments to prove or disprove this point. He used chiefly the starch from the rhizome of Carina indica, and for the extraction of the substance (granulose) which was colored blue with iodine, he utilized saliva. At a temperature of 35° to 40° C. the extraction of the .soluble substance began slowly, and proceeded regularly from the periphery of the grain to the center. At a temperature of 50° to 55° the extraction was completed in several hours. An exainination of the grains from which the soluble constitutents had been thus removed showed that they had lost in weight; that they float more readily in water; that they are less refractive to light; and that they are smaller, but just how much was difficult to determine. The 1am- ellation underwent no change, except that in many cases, as in sprouting wheat, it became more evident during the process of solution. The grains from which the soluble matter had been extracted by saliva behaved exactly towards polarized light as do unchanged grains. Von Molil opposes Nageli's statement that granulose and cellulose can be differentiated by iodine. He states that whether iodine produces a red or a blue color depends neither on the fact that the object stained may be starch (granulose) or cellulose, nor on the amount of iodine present, but essentially on the behavior of the organic substance toward water. If a small amount of water is absorbed a red coloration ensues; if a larger amount, there is a blue coloration. One can produce, he states, a beautiful blue in cellulose, and a red or violet in starch (granulose) without bringing about a chemical change in the objects colored, the coloring being regulated by the amount of water absorbed. It is therefore clear, he holds, that the blue coloration of cellulose with iodine is in no wise proof that the cellulose is changed wholly or partly into starch; and, moreover, that iodine furnishes no means whatever of differentiating granulose from cellulose. Furthermore, von Molil holds it can be demonstrated, from both physical and chemical standpoints, that the substance (cellulose) remaining in starch-grains after treating the grains with saliva is not identical with the cellulose of plants. The so-called cellulose of the starch-grain, he states, is very brittle, while plant cellulose is tenacious to a remarkable degree. The effects of polarized light on the two are opposite. Caustic potash dissolves starch cellulose instantly, but swells up plant cellulose, and dissolves it only after a num- ber of hours. In nitric acid and muriatic acid starch-grains are dissolved at once, but plant cellulose is dissolved only by boiling. Other chemical reagents showed similar con- trasts in their effects on the two substances. Von Mohl proposes the name farinose as a substitute for the term cellulose of Nageli. The nature of the substance of raw starch that is soluble in cold water was also examined by Delffs (Aim. d. Physik u. Chemie, 1860, civ, 648), who macerated com- minuted starch-grains in water for 24 hours, after which a clear liquid was filtered off which gave a blue reaction with iodine. The soluble constituent he regards as being a 2S DIFFERENTIATION AND SPECIFICITY OF STARCHES. form of dextrin, which he thinks differs from the three forms derived by the action of malt extract, sulphuric acid, and torrefaction, respectively. He views this substance as being in the nature of a starch-building material, and he discusses in this connection the mechan- ism of the formation of the starch-grain. He writes that if one assumes the starch-grain to be an organized body which grows by intussusception there must be present a substance which is soluble and which can enter the grains by osmosis. To serve the function of starch- building no substance is more suitable for the purpose than the various kinds of gums and dextrins, because they on the one hand have the proper solubility, and on the other belong to a group that is isomerous with the insoluble starch-substance, to which grouj) belongs the soluble constituent of starch. This substance which may be built up into starch he states may be termed amylogen, and he holds that, if these views are correct, the question as to the age of the various layers of the starch-grain is solved, for the outer layers must be regarded as the oldest, the greater age and density of the outer layers protecting the soluble contents of the grain. Objection was made by Knop (Chem. Centrablatt, 18G0, v, 367) to the statement of Jessen and of Delffs that by crusliing the starch-grain a gelatinous substance is dissolved out by cold water. Knop believes that the heat generated while rubbing the grain is sufficient to cause gelatinization ; but Jessen (Ann. d. Physik u. Chemie, 1864, cxxii, 482) showed later that the ground pulp has a temperature of only 22°, and therefore, in opposition to Knop, not sufficiently liigh to cause gelatinization. Shortly after the investigations of Jessen, Delffs, and Ivnop, Fliickiger (Zeit. f. Chemie, 1861, IV, 104), in a brief article, showed that pulverized starches from various sources are partially dissolved, especially in the presence of calcium chloride. According to Dragendorff (Jour. f. Landwirthsch, 1862, vii, 211), starch consists of 3 or 4 components: (1) a base which remains after starch has been heated to 60° in a concentrated solution of sodium chloride in a 1 per cent muriatic acid; (2) true starch which is insoluble in cold water; (3) Schultze's amidulin (Delff's amylogen); and (4) occasionally dextrin. The influence of light on the formation of starch in chlorophyl granules was studied by Sachs (Botanische Zeit., 1862, xx, 365), who concludes that starch is to be regarded as a product of the assimilative activity of the chlorophyl substance, this activity, being due to the agency of strong light, the starch being built up from CO2 and water in the presence of mineral salts, and being distributed to growing buds and storage centers. He believes that starch formation occurs through a series of transformations in the chlorophyl bodies, and that the starch of non-green parts of plants has wandered from the green parts. In a later article {ibid., 1864, xxii, 289) Sachs reports the results of experiments in connection with the formation and disappearance of starch in a number of plants. He found that starch which originates in chlorophyl granules under the influence of light disappears when the plant is withdrawn from the light ; that starch formed during the day is partly dissolved during the night; and that the entire starch-content disappears in the dark in 48 hours. In accordance with his results, he assumes that during a summer night of 8 hours one-sixth of the starch is dissolved. It would seem, he states, that these reversed processes should throw light on the mechanism of the formation of starch-grains. From a study of the chemical properties of starch in relation to various solvents, especially haloid salts, glycerol, and saliva, Kabsch (Zeit. f. analyt. Chemie, 1863, 11, 216) held that the assumption is not justified that starch consists of two different substances, one of which is real starch (granulose) and soluble in saliva (Nilgeli and von Mohl), in dilute acids (Melsens), and in cold water after crushing (Reinscli, Jessen, Delffs, and others); and the other substance (cellulose according to Niigeli, and farinose according to von Mohl) insoluble in these media. The grains he found appear at first very small, solid, and gran- ular, and during growth become more or less condensed, and arranged in layers which STARCH-SUBSTANCE, AND THE STRUCTURE, ETC., OF THE STARCH-GRAIN. 29 vary in density, the outermost and densest part resisting all action of chemical substances longer than the less dense layers by whicli the latter are surrounded and impregnated. In further investigations of the constitutcnts and decomposition products of starch, Jessen (Jour. f. prakt. Chemie, 1868, cv, 65) concludes that starch-grains are composed mainly of three constitutcnts: (1) the envelopes, or cell-membranes, or integuments, which are insoluble in hot or cold water (the amidine tegumentaire of Gu^rin-Varry) ; (2) the starch-substance, or amylogcn, which is soluble in cold water (the amidone of Payen andPersoz; the amylogcn of Delffs ; and the amidine of Guerin-Varry etc.) ; (3) starch which is insoluble in cold water, but soluble in water at 55° to 80° C., which may be called amijlin (the amidin-soluble of Guerin-Varry, and the amylin of IVIaschke, Delffs, Melsens, and Fr. Schultze). Other constituents were found to be present in very small quantities, and in ordinary analyses some of them may be entirely overlooked. Some of the minor constitu- ents, he states, are dextrin (in wheat, pointed out by JVIaschke); chlorophyl and wax (in potato, according to Guerin-Varry); nitrogen (0.1 to 0.25 per cent); fat, soluble in alcohol (0.001 per cent in potato and 0.0005 to 0.006 per cent in wheat, according to Rousseau). The envelope, Jessen states, forms only a small part of the grain (2.96 per cent accord- ing to Guerin-Varry; 5.7 per cent in potato, 3.1 per cent in maranta, and 2.3 per cent in wheat according to Fr. Schultze; and 4.8 per cent according to Payen and Pensoz). By continued boiling the envelopes pass more or less into solution, and they behave generally like other cell membranes and give the reactions for cellulose. Often they became blue in the presence of iodine, owing, he assumes, to adherent particles of starch. He states that when these are removed the envelopes are colored with iodine only upon the addition of chloride of zinc, dilute sulphuric acid, etc. The proportions of amylogen and amylin Jessen gives as 58.68 per cent and 38.38 per cent respectively. He accepts the figures of Guerin-Varry (60.45 and 39.55 respec- tively), although he asserts that the proportion of amylogen is probably less. Both sub- stances yield a blue reaction with iodine. The amylogen exists chiefly in the inner parts of the grains, and it can be obtained by placing the crushed grains in cold water. The inner layers he believes to be the youngest. Starch-paste he holds is neither a simple substance nor a chemical compound, but a mechanical mixture of all of the constituents of the grains. Amylogen, he found, passes into dextrin very readily, far more readily than amylin; and the transformation occurs in a pure solution at room temperatures after 2 to 3 days. Whether or not amylogen and amylin of different kinds of starches are differ- ent, he does not know; and he states that very little is known about the composition and properties of amylogen, and that in chemical composition amjdogen and amylin may be the same. Sachs (Text-Book of Botany, 1875) adopts in modified form Nageli's view of the two proximate constituents of starch. He states that every grain of starch consists of starch, water, and inorganic substances. The starch has the same percentage composi- tion as cellulose, to which it bears the greatest similarity of all known substances in both chemical and morphological properties. It occurs in the grain in two forms: one, granu- lose, which is the more easily soluble and which gives a blue reaction with iodine ; and the other, starch-cellulose, which shows less solubility and wliich comes nearer to cellulose. Both occur in every part of the grain. If the granulose is extracted, the celluolse remains behind as a skeleton which shows the internal organization of the whole grain, but is less dense, and its weight represents only about 2 to 6 per cent of the weight of the whole grain. The internal organization of the grain he holds is not homogeneous, and that it differs in relation to the varying proportions of water in the several parts. Every part of the grain contains water, the amount usually increasing from without inwards, and attaining a maximum at a fixed point in the interior. With the increase in the proportion of water there is a decrease in cohesion and density, and also in refractivity, on which partl^^ the 30 DIFFERENTIATION AND SPECIFICITY OF STARCHES. power of perceiving these differences depends. The outermost, most dense, and least watery layer is succeeded by a more watery layer beneath, and this in turn is followed by a stUl more watery layer, and so on until the innermost and least dense and most watery layer surrounds a very watery part, the nucleus. Sachs believes that the hypothesis of the growth of the starch-grain by intussuscep- tion alone affords an explanation of all of the phenomena arising during the growth of the grain. New particles of formative material become, he tliinks, intercalated between those already existing, both in radial and tangential directions, by which means the pro- portion of water at the points of deposition is lessened. Did the formation of layers occur by external deposition, grains would be formed in which the outermost layer is the most watery; but this never occurs, because the outermost layer is always the least watery and the most dense. According to the view of growth by apposition, he contends that the nucleus would possess the properties of young grains, which are dense, whereas in mature grains the nucleus is always soft. The theory of growth by apposition he believes could be accepted to explain only the formation of partially compound grains if we were to assume that the common layers which inclose the single simple grains had been deposited about them, in wliich case the common layers would have a different form and the fissures in the interior of such grains would remain unexplained. In 1874 a monograph appeared by W. Nageli (Beitriige zur naheren Kenntnis der Starkegruppe in chemischer und physilogischer Beziehung, Leipzig, 1874, S. 106) in which he treats especially of the effects of dilute acids on starch, and of the means of preparation and the properties of amylodextrin. (See Chapter III, p. 114.) He notes that while starch in a natural condition is insoluble in water, it becomes soluble after long soaking, or when the grains are broken; and that there are several modified forms of starch which are char- acterized by their different i^owers of resisting solution and by their reactions with iodine. In the solid state these forms are said to take color in the order of blue, \aolet, red, orange, and yellow, as their power of resistance to solution and their affinity for iodine increase. Iodine added to a starch solution will always turn it blue, because, he states, on boiling with water the other modifications gradually go over into the blue-reacting form. Tliis latter form, being the more readily soluble, possesses the power of absorbing the less soluble modifications; the disappearance of the former precipitates the latter. The various kinds of starches, he states, may be distinguished by the different proportions of these modified forms. The difference in the substances composing the starch groups may, he believes, be a chemical one, but that it more likely rests on the physical condition of greater or less distribution. Musculus (Botanische Zeit., 1879, xxxvii, 345) accepts the view of C. Nageli that starch consists fundamentally of two substances (granulose and cellulose), but he goes on to show that the cellulose (referred to by him as amylo-cellulose) is nothing else than an insoluble modification of granulose. Granulose can by drying, he states, be transformed into cellulose, and cellulose can be transformed by sodium hydroxide into granidose. If alcohol be added to the sodium-hydroxide solution of cellulose, a gelatinous precipitate results which, after repeated washing, shows all of the properties of granulose. This precipitate is completely soluble in hot water, and the solution is colored blue with iodine. If the precipitate is chied, it undergoes a partial transformation into cellulose. The first of an important series of contributions wliich have had a marked influence on our views, even up to the present, was published by Schimper (Botanische Zeit., 1880, xxxviii, 881; 1881, xxxix, 185, 201; 1883, xli, 121), in which he studies the structure and mechanism of formation and other properties of the starch-grains, and also the specific starch-forming structures of various plants. He records in the first article that if the cldorophyl-grains are spherical, the starch-grains may appear at all points of the chlorophyl- grains; but if the form is plate-like, the starch-building is limited to the equatorial zone. STARCH-SUBSTANCE, AND THE STRUCTURE, ETC., OF THE STARCH-GRATN. 31 and that such chlorophyl-grains may produce six or more starch-grains. Starch-grains which arc produced within the chlorophyl-grain, and remain surrounded by it, attain a concentric structure, and most grains originating in this way remain very small and without definite structure. In Vanilla plamfolia the mature starch-grains consist of hundreds of colorless, polyhedral granules of equal size, which originate as very small points in the clilorophyl, and become larger, and at the same time polyhedral through mutual pressure. Starch- grains which develop at the periphery of the chlorophyl-grain show a differentiation into hila and lamelliT'. Such grains are always eccentric, and the growing side of each grain is that to which the chlorophyl-grain is attached. It is obvious, Schimper states, that the unequal growth on opposite sides of the hilum is the result of unequal "nourishment." Such a conclusion, he finds, is supported by the fact that where starch-grains are partly in contact with chlorophyl-grains they become gibbous at the points of contact. Starch-grains developing in flat chlorophyl-grains are wedge-shaped at first, being flattened in the same manner as the chlorophyl-grains. W^iere starch is produced very actively, the chlorophyl-grain assumes an isodiametric form, decreases in density and later in size, until finally it is present as a mere remnant, while at the same time the starch-grain becomes denser and assumes an oval shape. With the dis- appearance of the chlorophyl the growth of the starch-grain ceases. In chlorojihyl-grains capable of producing starch-grains within their entire mass, the starch-grains may appear near the surface of the chlorophyl, and later on break through. Such grains are eccentric. Scliimper's examination of non-green starch-bearing plants showed that starch-grains are not surrounded by ordinary protoplasm, but are inclosed by or attached to refractive, spherical, or spindle-shaped bodies which are but little affected by alcohol and are rather unstable. A study of the youngest stages of these bodies (later designated leucoplasts) demonstrated their presence before starch-grains are formed and that the starch-grains, at their appearance and during their development, indicate similar relations to these bodies as are borne by other starch-grains to chlorophyl granules. The starch formed in these albuiuinous-hke bodies has an eccentric structure like the eccentric grains produced by chlorophyl-grains, and the entire behavior of these bodies is like that of chlorophyl-grains, and therefore they are starch-builders. Schimper then gives the mechanisms of formation of starch by starch-builders in non-chlorophyllous plants and refers to the formation of starch in roots as an instance of starch-building by these colorless bodies, or leucoplasts. He also noted that in most cases the colorless starch-builders may, under the influence of light, be transformed into chlorophyllous bodies. In the second and third communications Schimper (Botanische Zcit., 1881, xxxix, 185, 201) goes more in detail into the processes and stages of development of starch-grains in different plants. He found that the grains growing in many chlorophyllous plants show certain constant peculiarities. Tabular grains have irregular lobes and are laterally de- pressed, and at times are porous; their broad surfaces are uneven and present a spotted appearance, owing in part to surface structure and in part, in many cases, to internal vacuoles. These phenomena he ascribes to partial solution, which occurs when starch is used for the growth of the organ in which the starch-grains are formed, as in germinating seeds of Zea mays. Even after the starch-bearing organs cease to grow, or slacken growth, starch is formed. There may originate spherical structureless grains; or corroded grains already present in the cells may take on new growth, which occurs in the form of a layer around the corroded grain, thin at first, but gradually becoming thicker and more refrac- tive. This layer shows the projections and depressions of the corroded grain. Layers subsequently formed go through the same process, but the roughness of the surface becomes less pronounced as new layers are deposited, so that the mature grain appears smooth; but iDy suitable illumination the original corroded grain may be seen in the center of the mature grain. 32 DIFFERENTIATION AND SPECIFICITY OF STARCHES. Schimper states that the processes of formation of starch-grains were very similar in the several plants studied. The most important stages in the development he gives as follows: (1) the appearance of starch-grains in the form of strongly refractive bodies; (2) the differentiation of the originally homogeneous grain into a central watery hilum and a dense peripheral layer; (3) in later stages the surrounding of the hilum by three layers, the middle always being watery (a watery layer never occurs at the outside, and such a laj'er probably originates by an alteration of an original dense layer) ; (4) the number of layers increases, and the outer is always the most dense; (5) the watery content of the inner parts of the grain increases with the increase in the volume of the grain. Schimper found that pressure on starch-grains causes the formation of numerous cracks, which in simple grains are perpendicular, but never parallel to the lamella?. The cohesion of the starch-grains varies in a very striking manner with the line of direction in wliich the pressure or strain is exerted — in a tangential direction it is very small, but in a radial direction it is great. In the tangential direction he holds that there is no elas- ticity of the substance, whereas in the radial direction it is marked. Schimper refers to the observation of previous investigators that in the swelling of starch-grains in water the water is stored much more abundantly in a direction parallel than perpendicular to the lamellae, and he states that one reason for this is that the cracks, originating during the drying of starch, always run at right angles to the lamellae, and that cracks would also appear in other directions if the water were distributed equally in the grains. The fact that the water is stored in larger quantities in a tangential than in a radial direction causes tension in the grains. The grain is assumed to consist at first of a homogeneous substance of uniform density. If the ever-increasing tension resulting from the unequal storing of water reaches a point where the elasticity can no longer withstand the tension, the substance of the center of the grain must expand and assume a condition of greater swelling, and become less refractive. Observation shows, he states, that when the grain has attained a certain size during swelling, the feebly refractive but much swollen hilum appears at the center. The formation of the hilum is therefore brought about by the action of tension, this tension being caused by unequal deposition of water molecules, and not (as was assumed by Nageli) by the tleposition of starch molecules. The formation of the hilum decreases the tension of the grain, but with the storing up of new substance in the dense layer around the hilum the tension increases until it finally is sufficient to over- come the elasticity. The result is that the dense layer is differentiated into three layers — a soft middle layer and dense inner and outer layers. The dense outer layer behaves just as the first layer, and when its tension reaches a certain intensity it also undergoes division into a soft inner layer and a dense outer layer. As a result of the storing of new substance in the grain the inner parts are expanded, the soft layers increase in density, and the dense layers increase their water-content as they become softened, and, consequently the inner parts of the grain are less resistant to swelling and solvent media than the outer parts. Schimper holds that the form of the starch-grain is determined by the manner of "nourishing." Concentric grains result, he states, when they are completely surrounded by the starch-producing substance or chloroplast, wliile eccentric grains are found when the plastid is in contact with a part of the surface of the growing grain, the most rapid growth taking place at the point or points of contact of the starch-forming substance. Flat grains -with a central hilum originate in lens-shaped chlorophyl-grains, and their flat sides are parallel to the flat sides of the chlorophyl-grains. Elongated starch-grains are formed in spindle-shaped chlorophyl-grains. The different shapes of the starch-grains can be explained only by unequal "nourishment." If we try, he states, to conceive the manner in which the grain is nourished by its mother substance, we imagine the latter to be a form of solution which impregnates the chlorophyl-grain. The starch-grain and the chlorophyl- grain which forms it do not lie in the cell-sap, but are embedded in the protoplasm. STARCH-SUBSTANCE, AND THE STRUCTURE, ETC., OF THE STARCH-GRAIN. 33 In later communications by Schimper (Botanische Zeit., 1883, xli, 121; Jahrbiicher f. vviss. Bot. 1SS5, XVI, 7) studies are made especially of the origin and functions of chlo- roi)hyl-grains anil homologous structures. Leucoplasts, he writes, originate in the vege- tating cells of the plant, and but rarely from chloroplasts. Certain fruits which are green in the young stages and white in the mature condition contain leucoplasts which develop from the chloroplasts, as, for instance, in the white fruit of the snow-berry, Symphoricar- pus racemosus. Leucoplasts are widely distributed and serve as starch-builders; rarely are they functionless, as in the roots of Dahlia. Many leucoplasts may be transformed into chloroplasts, especially when they are exposed to light of sufficient intensity, but some leucoplasts are incapable of this change. In non-chlorophyllous starch-producing plants, or parts of plants, the starch is formed by leucoplasts. Referring to chromatophores, Schimper records that they assume very different forms, and that it is difficult to discover an external relation between them. The green tabular chloroplasts, the colorless (usually very delicate) spherical leucoplasts, and the chromoplasts with manifold forms differ widely from each other. Nowhere is the capacity shown by liv- ing substance to assume such diverse forms and properties as is exhibited by these simple protoplasmic structures. One and the same chromatophore, as a leucoplast, can form and store up starch from assimilated materials, or as a chloroplast can decompose carbonic acid and produce organic substance from carbon and water, or as a yellow or red chromoplast can fill the passive role of attraction for animals. Chromatophores originate exclusively from preexisting chromatophores. In the simpler plants, as in Chlorophycece and Diato- macece, leucoplast formation is a subsequent process, that is, chromatophores which con- tain pigment are transformed into colorless chromatophores, or leucoplasts; but in higher plants the opposite process takes place. This condition in the simplest plants leads to the belief that leucoplasts are to be regarded as metamorphic forms of pigmented, assimilating chromatophores. The chromatophores of a large number of Angiosperms form protein crys- tals, and the shape of the chromatophore is often due to the presence of these crystals, and thus the shape and structure of the starch-grain are correspondingly influenced. In Phaius (orchid), starch-grains are produced which have an eccentric structure and some are tri- angular in shape. These are developed on an elongated chromatophore, the elongation being due to the presence of protein crystals. That these crystals thus indirectly influence the shape of starch-grains is also shown in the rhizomes of some Zingiheracew. In such plants the chromatophores of the rhizomes always produce elongated, spindle-shaped, or needle-like crystals, and these cause the odd structure of the starch-grains. In the rhizomes of Canna, large, flattened, triangular, mark- edly eccentric starch-grains develop. These grow on leucoplasts which form two kinds of crystals — flattened octohedra and needle-like. The grains lie both on and within the chro- matophores, and in the latter location attain an eccentric structure which is influenced by the presence of crystals. Colorless bodies, pyrenoides, are found in the chromatophores of many Alga^, and appear to act as nuclei of the chromatophores. They bear a close relation to starch-formation. In these plants, the starch-grains which originate within the chromato- phores appear first and in the largest numbers about the pyrenoides, which they surround in the form of a shell. Observations indicate that the pyrenoid substance is in the nature of a reserve material that is used by the chromatophore in starch-building, and that they are probably crystals, since they behave like protein crystals of the higher plants. Schimper also believes that in young cells of certain of the lower plants, cell nuclei furnish material for starch-formation before the chromatophores are developed; but starch subsequently appears everywhere in the cells about the chromatophores, by which time the nuclei apparently have ceased to yield starch-forming material. Chloroplasts and pyrenoids in Algce, chloroplasts and paramyluni bodies of Eugknoe, and pjTenoids in bacilli were reported by Schmitz (Die Chromatophoren der Algen, 1882; Jahr. f. wiss. Bot. 1884, xv, 1). 3 34 DIFFERENTIATION AND SPECIFICITY OF STARCHES. Vines (The Physiology of Plants, 1886) assumes a difference in the functions of chlo- rophyl corpuscles and leucoplasts in that in the former the synthetic processes begin with such simple substances as CO2, water, and salts, and are effected under the influence of light, whereas in the latter they begin with tolerably complex substances, such as asparagin and glucose, light not being essential. The contention of Schimper that starch-grains are formed by apposition brought forth another contribution by C. Nageli (Botanische Zeit., 1882, xl, 633) in defense of the intussusception theory. Nageli WTites that he was led to the adoption of the intussuscep- tion theory because in the originally dense grain there appears a soft hilum, and that later, when the dense layers have attained a certain thickness, there is inserted a soft layer. He assumes that the starch-grain is composed of very small invisible particles (which he here refers to as micellce) which have a crystalline character and grow as crystals do, and which attract water to their surfaces. If unilateral pressure is applied to the grains, tangential as well as radial cracks form. The cohesion of the grain he holds is less in the radial than in the tangential direction, on account of the alternate dense and soft layers; and the cohesion of the single layers is also greater in the radial direction. For various reasons it seemed to Nageli that the soft layers possess a gelatinous, brittle consistency, and are not semifluid, and therefore that cracks might occur in these layers, which would not be the case were they semifluid. Stains and solvents in small quantities, he states, penetrate the starch-substance and are deposited in varying proportions. Such deposition is dependent on two causes : on the relation of the stain and the solvent to the starch and the peculiar micellar constitution of the starch-substance, and on the dynamic action which is conditioned by the constitution of the starch-substance. The molecules of the stain he conceives to remain dissolved in the imbibitional liquid, or extracted from it and deposited in the starch micellae. That different stains behave differently is shown by the dissimilar relations of starch-grains toward the same stain in different solvents, and by their relation towards different stains in the same solvent. Starch-grains in their natural condition, he found, may not take stain, while the swollen grains take an intense stain. This he explains upon the assumption that unchanged starch substance offers a greater resist- ance to the deposition of stains because of the regular and compact arrangement of the micellse. In the swollen, disorganized starch substance the micellae are disarranged, and therefore lack the power of resistance against foreign molecules. Nageli believes that the softer parts of starch can not be regarded as a paste-like substance disorganized by swelling, because the cohesiveness of the grain prevents such a disorganization. Tension in the starch-grain which surpasses the limit of elasticity brings about a rendering of the substance, but not a swelling. In grains diied and again moistened, cracks were found evident only in the unchanged starch-substance. By slow action of artificial swelling agents, cracks appear, and disorganization takes place in pro- portion to the strength of the disorganizing medium. He holds that if the negative tensions which result from the growth-processes within the grains are to be explained satisfactorily by the storing of water (as Schimper believes), and not by a storing of starch-substance according to the intussusception theory, the formation of the soft center or hilum and the softer layers can not be accounted for, and by such an explanation the interior of the grain must be disrupted by cracks. NiigeU also attacks Schimper's view that water is stored more abundantly parallel than perpendicular to the lamellae, and that on this account different conditions of tension are caused. He contends that the starch-grain at every stage of its growth is surrounded by a watery liquid, by an intermicellar water-containing system, whose tensions are always at an equilibrium. When the grain is dried cracks are formed, which shows that the equi- librium is disturbed by drying; and the cracks have a radial direction, crossing the layers at right angles, which is proof that more water is lost in a tangential rather than in the STARCH-SUBSTANCE, AND THE STRUCTURE, ETC., OF THE STARCH-GRAIN. 35 radial direction. He found that wlien artificial swelling media act upon the starch-grain slowly, the vohune of the grain is increased and radial cracks are formed which indicate that during this process more water is stored radially than tangentially. The existence of tensions Nageli deduced from the theory of growth by intussuscep- tion. Starch-grains in which radial cracks appear on drying do not lose the cracks on moistening, but on the other hand the cracks become larger and hence more evident. The lamellie of the grain in its natural state have a tendency to store more water tangentially, and therefore grow chiefly in length and breadth, and this tendency is more pronounced in the external layers, so that these in contrast with the inner lamella; have a positive ten- sion. He states that if this were not the case, on moistening the dried grains the cracks would become imperceptible. Tensions could not occur, he holds, in growth by apposition. If we imagine a starch-grain to lie in the cell-protoplasm or in the cell-liquid, it is in con- tact with water, and it has absorbed as much water as it can, because the micellje attract as much water as their molecular force can hold. The particles of micellae are formed in such a manner that there is an equilibrium between the particles of starch and water. The tensions determine the separation of the soft center or hilum and the softer layers of the grain, and this is brought about by the simultaneous growth by intussusception. Another fact that Nageli assumes aids in disproving the apposition theory is the presence of a thin peripheral layer which possesses different properties from the rest of the grain. He states that when this outer layer is very thin it does not react with iodine, and that when somewhat thicker it is colored a reddish-violet, while the part inclosed by this layer always becomes a deep blue. This peripheral layer likewise resists the action of solvents (acids, etc.), while the entire inner mass is dissolved. Since both large and small grains have this peripheral layer or "membrane," growth by intussusception, according to Nageli, is the only possible explanation of the mechanism of formation of the starch-grain. The hypothesis proposed by Schimper that the starch-grain is a spherocrystal of a carbohydrate was supported by Meyer (Botanische Zeit., 1881, xxxix, 841), who states that it furnishes the simplest explanation for the lamellation, the apposition-growth of these lamellse, the later deposition of the outer layers by external accretion, and the rela- ti\'ely low density of the inner parts of the older starch-grains. He compares the growth, the formation of lamella?, and the variations in density of the different parts of the grain with similar phenomena of spherocrystals generally, and shows analogies in causes and explanations. He reasons that if starch-grains are spherocrystals, it may be assumed with certainty that they grow in a manner analogous to that of other spherocrystals. If sphero- crystals of a carbohydrate, such as sugar, are caused to form, and if conditions attending crystallization are altered periodically, as when crystallization takes place at a window where the sun warms the preparation periodically, spherocrystals are deposited in the form of a concentric lamellated structiu-e; but if crystallization occurs under constant conditions of temperature, the crystals show no lamellse. It is therefore to be inferred, Meyer holds, that the formation of the layers of the starch-grain is due to fluctuations in external conchtions. The hHa of the spherocrystals of sugar are usually less dense than other parts of the crystals, and they inclose a mother substance, rarely air. Very seldom is a dense hHum observed. Applying these facts to the growth of the starch spherocrystal, Meyer states that they show that most starch-grains, even in young stages, must ha\'e a small, relatively soft center; and that the starch-grains, since the assimilation of the plants is subject to periodic fluctuations, must be built up of layers of alternating density. The youngest, outermost layer is always the most dense, and the successive, deeper-lying layers become less dense accorchng to their age. The varying density, he assumes, is due to the action of ferments. Thus, if one assumes that a layer just formed is exposed to ferment action 36 DIFFERENTIATION AND SPECIFICITY OF STARCHES. for n hours, it will decrease in density, and it wUl lose, for example, m grams of starch- substance which is dissolved by the ferment. If a new layer is now formed, it will be the most dense. If the starch-grain is again exposed n hours to ferment action, the first layer loses 2vi grams of substance, and the last-formed layer, that is, the second layer, loses m grams by solution. Meyer states that there is present a second kind of solution, a corrosion of starch, which occurs in the living cell. In and on the chlorophyl-grains solution of the starch-grains takes place if the assimilation of the plants becomes sufficiently low. Starch-grains in non-green cells are also dissolved on the exterior while they are still attached to the starch- builder or leucoplast. Grains which are slowly dissolved from without take on a corroded appearance, owing to a loss of some of their substance, and during the action of the fer- ment they are surrounded by a layer of less dense substance. The thickness of this layer depends upon the length of time the ferment acts and upon the degree of activity of the ferment. This outer corroded layer remains visible even after new starch is deposited. Such a deposition on corroded grains takes place in both chloroplasts and leucoplasts, and relatively thick layers originate which may be designated secondary layers in contrast with the primary layers produced by the previously described change of conditions of crystallization. In another article which appeared at about the same time (Botanische Zeit., 1881, XXXIX, 856) Meyer reports the results of studies of starch-formation in the rliizomes of 7ns gcrmanica and Iris pallida. The mature grains are described as oval, cylinchical, or spherical, and as showing great diversity of forms and with an eccentric hilum. Usually the grains in the cells nearest the tip of the rhizome are faintly lamellated, but very often closed layers are entu-ely wanting. In parts of the one-year-old rhizome back from the tips, grains are found, of approxunately equal size, in which the closed layers are more prominent than in the tips. The increase in the distinctness of lamellation he explains upon the assumption that there is a solvent action on the starch in the cells. The grains show conical lamellation, and the outUnes of the layers which become visible through ferment action are similar to those of the young stages of the grains of the rhizome tips. Meyer confesses his inability to determine whether the relatively low density of the bases of the successive closed layers has arisen by the contact with the starch-builder and the rapid supply of material, or whether it is a form of lamellation determined by the ferment, or whether it is due to both. It seems certain, however, he states, that the ferment action emphasizes conical lamellation. In older parts of the rhizomes cells are sometimes found filled with starch, but in others they are entirely starch-free. In going from such old (mostly non-starch-bearing) parts toward the tip of the rliizome, starch-builders are found which bear starch-grains; but these grains are mostly attached obhquely to the builders, the grains having moved somewhat, so that now the original bases do not come in contact with the builders. The grains are usually more or less corroded, and conunonly the side of the grain in contact with the builder is "eaten out." In parts of the rhizome where the most active solution of the starch takes place, it is found in the autumn that there are grains which have secondary lamellation. The presence of such lamelte in grains that are usually attached diagonally to the builder supports the theory that in such parts of the rliizome periodic solution and formation of starch-substance take place in the same starch-builder. Since starch-builders have until their death the capacity to produce starch, new grains may be formed or, if corroded grains are present, such grains may take on sec- ondary lamellation; but the new grains are distinguished by their almost regular round form and by their prominent radial striations, which are developed when the grains are swollen in potassium hydroxide. The theory of the growth of starch-grains by apposition furnishes, he states, a simple explanation of all of the phenomena observed in the devel- opment of the starch-grains of the Iris rhizome. STARCH-SUBSTANCE, AND THE STRUCTURE, ETC., OF THE STARCH-GRAIN. 37 Meyer, in an elaborate research on starch-grains published in 1895, again takes up the study of the formation and structure of the starch-grain, reference to which will be found in later pages (page 47). Our understanding of the structure and mechanism of formation of the starch-grain was added to materially by the careful observations of Strasburger (Ueber den Bau und das Wachstum der Zellhflute, Jena, 1882). He states that since there is a great analogy of structure between cell-membranes and starch-grains, it is natural to go from the former to the latter in the study of the structure and growth, and that the large, flat, eccentric starch-grains of Phaius grandiflorus {P. wallichii) are very suitable for such investigations. The structure of the grains is such that the layers, except the innermost, are not complete, that is, they end blindly on the side of the grain (see plate 102, fig. Gil). In general, all starch-grains from the same bulb agree fully in structure and in other prop- erties. When magnified, these grains show definite lamellations which are indicated by broader light and narrower dark lines. The lines or lameUse may vary greatly in width. Upon this alternation of light and dark lines is based the view of alternation of dense and less dense layers, and of layers that are poor or rich in water. As a matter of fact, he states, there are in the starch-grain, as in the cell-membrane, consecutive lamelte wliich resemble each other rather closely. The darker lines are the adhesion surfaces of the consecutive lamellae. Wliere these dark lines follow each other very closely they indicate directly the boundaries of the lamelte; usually, however, they indicate only single adhesion surfaces, and thus separate layer-complexes from each other. The more sharply marked the separation-surfaces appear, the longer are the intervals between the periods of forma- tion of the lamellae. The dividing lines become more pronounced by individual layers becoming more dense on their outer surfaces. This accentuation, he states, may also be brought about by an interruption in the growth. Small variations in the constitution of the protoplasm which produce the lamellae, as a difference in the water-content, may also possibly cause the variation. Strasburger found that by a slow action of potassium hydroxide on starch-grains the grains swell. The surface was found to be the most resistant, and the outermost layer was seen to be not continuous around the entire grain. The anterior and the acute end of the grain was the most resistant and is the oldest part of the grain. With the beginning of the action numerous fine cracks appeared on the surface, and ran with approximate regularity perpendicularly to the direction of the lamellae; immediately the substance between the cracks was dissolved at a number of points. The entire surface of the grain was now dotted regularly, and the dots were very small and not continuous. At the same time there was seen a rather large crack, which, passing out from the hilum, mantle-like, inclosed the inner part of the anterior half of the grain ; it separated the denser outer part from the less dense inner part; the latter escaped from the grain through openings made by the rapidly dissolving posterior lamellae; the inner mass separated into two parts. The dense layer and the layer-complexes do not swell as much as the less compact layers and radial cracks form in them. These cracks are usually very irregular and anastomose with each other. During the swelling the lines of separation between the layers and the layer- complexes do not become wider, yet they become more evident. The formation of cracks and their direction in the swelhng process indicate undoubtedly that a radial structure is present in the grains, and leads to the conclusion that the elements of the grain are minute rods arranged radially. The regularity of the cracks, Strasburger holds, opposes the view that the cracks have arisen through tension in the grains. In the fresh starch-grains the hilum was recognized with difficulty, but in the dried grains it was quite evident. In the latter case it appeared usually to be hollow, and the fissures that crossed the lamellae at right angles were directed toward the hilum. Such fissures in the dried grains extended only to the concentric lamella?, and they usually terminated where the lamelte became incomplete. 38 DIFFERENTIATION AND SPECIFICITY OF STARCHES. Studies were also made by Strasburger of the starch of Cycas drcinalis. In the rela- tively large part-grains, each grain consisting of 2 to 8 parts, which constitute a single grain, a radial structure was also observed. Likewise in the centric, oval grains of Phase- olus vulgaris a radial structure of the lamellae was observed. In Phaius the starch-builders are rod-like and easily seen, and from them the starch-grains project. The entire starch- grain or only the base was observed to be covered by a delicate membrane separate from the main mass of the starch-builder, and the lamellse extended only to this membrane. Between this membrane and the mass of the starch-builder was a less dense substance, described by Schimper as being a delicate and more or less swollen layer of the starch- builder which borders on the grain. Some of the grains, Strasburger found, do not originate in the chlorophjd bodies and starch-buUders directly, but in the cell-plasma. Such grains were observed in the macro- spores of Marsilia (water fern), and the grains were remarkable for the fact that each was covered with a network. The formation of the starch in these macrospores only begins after the macrospore has been developed. The grains are embedded in the proto- plasm, and no speciaUzed starch-buUders are present. These grains, he states, grow by apposition. Starch-grains in the cells of the medullary rays of Conifers are also produced without the agency of starch-buUders. In Pinus sylvestris the formation of starch begins 4 or 5 cell-lengths from the cambium. In the cells of the rays at this point very small starch-grains appear in the strands of the plasma-network, and appear to come directly from microsomes. When the grains become larger they lie in the meshes of the network, and have microsomes attached to their surfaces. In principle, the process of starch- formation in Marsilia, Pinus, etc., was not found to differ essentially from the formation in starch-builders, since in both cases protoplasm and mircosomes are utilized in the process. Where starch-builders are present there is merely a more extended division of labor in the protoplasmic cell-body. Strasbm-ger states that it is evident that all eccentric grains originate from differ- entiated starch-buUders, and that the eccentric structure is a result of unilateral deposi- tion on the starch-grain, which explains at once the presence of the starch-builder at only one end of the grain. Not all centric grains, however, originate without the agency of starch-builders, because a uniform growth on the entire outside of the grain is possible only so long as the grain is entirely surrounded by the substance of the starch-builder. Finally, that compound starch-grains originate by the union or fusion of grains originally isolated can be established, Strasburger states, in grains of Marsilia diffusa. In further studies of the structure of the starch-grain, Strasburger records that up to that time (1882) the lesser density of the interior of the starch-grain had not been satis- factorily explained. He compares the lamellse of cell-walls and starch-grain, and notes that in cell-walls the inner constantly growing lamellse wliich border the cell-contents behave differently from the lamellse of the membranes which are removed from the cell- contents. Every lamella, as it becomes covered by later-formed lamellae, decreases in its refractive capacity, this decrease being due to an increase in the water-content. The same relation holds good for the starch-grain. The lamellae deposited on the outside of the grains becomes less refractive but richer in water as they become farther separated from the outer surface. The parts of the grain which cease to increase, on wliich no new starch is deposited, retain their original density, and their density even increases with age. The anterior or acute end of the very pronounced eccentric Phaius grains is dis- tinguished by its density, and it rejiresents the oldest part of the grain. On account of the increased water-content as the lamellae lie farther from the surface layer, a })ulling or state of tension is created in every lamella, and the tendency is to distribute this strain and thus increase the size of the entire starch-grain, and hence on account of the strain in the inner lamellae, the storage of water in them is favored. This water-storage he con- STARCH-SUBSTANCE, AND THE STRITCTITRE, ETC., OF THE STARCH-GRAIN. 39 ceives takes place in a tangential direction only. The tendency of the grain is to increase constantly on its periphery, but this is hindered by the inner expanding lamellae. Thus there is a positive tension between every lamella and the next inner lamella, and a negative tension between it and the next outer lamella. Strictly speaking, Strasburger states, only in centric starch-grains does the water- content of the lamella; increase from without inward. In eccentric grains the increasing water-content of the lamellae appUes only to the oldest central parts of the grain. The greater density of the surface is due to the lower water-content. The influence of the sur- rounding medium lowers the swelling capacity of the surface, and it also gives rise to the following phenomenon noted by Harting: By contact with water there very often form den- ticular projections w^hich are caused by a more rapid swelling of the layers closely packed below the surface, whereby the surface-membrane becomes turned over at some places. Strasburger notes that by cutting or crushing starch-grains sweUing is increased because of the greater or less disorganization. Cohesion in starch-grains is feeblest in a tangential direction, and tension relations are therefore equalized by means of radial cracks. In concentric grains the innermost and most watery layer and the hilum lose most water in drying, and hence the cracks center at the hilum. The adhesion of the lamellte is stated to be very great, so that the layers can not be separated by means of pressure, wliile the formation of radial cracks is easily accomplished by pressure. Strasburger writes that the radial structure of starch-grains suggests the conception that the rod-like elements of the grains are crystal needles arranged radially. The component of raw starch that is soluble in cold water was studied by Bruckner (Monatschefte f. Chemie, 1883, iv, 889) and identified with the amidulin described by Nasse (luaug. Dissert. Halle, 1866) and with the granulose of C. NageH. Bruckner dried the grains and crushed them between glass plates, causing many cracks. After macera- tion in water, and repeated filtering, a fUtrate was obtained which became blue upon the addition of iodine. In another experiment, the whole grains (Zea) were placed with three volumes of water in a cylinder and set aside for three weeks, during which period the mixture was shaken daily. At the end of this time the preparation was filtered, and the filtrate was evaporated to one-fifth of its original volume. No trace of a blue reaction with iodine was obtained. From these results, Bruckner concludes that the outer firm layers or coats serve as a protective membrane to the inner layers, and that the soluble starch can be extracted by cold water only after cutting, breaking, or otherwise injuring the outer layers, and thus exposing the inner soluble portion. Shubert (Monatschefte f. Chemie, 1884, 472; Jour. Soc. Chem. Ind., 1885, iv, 236) investigated the behavior of starch-granules on being heated, and concluded that changes in form and structure, especially the lamination, are not solely determined by the amount of moisture in the air-dried granules, but depend also on the different physical and chemical properties of each layer. The effect of heat was found sufficient to make these differences more prominent. The starch-granule, under the influence of high temperature, is altered in such a manner that the layers that are rich in granulose are at once converted into solu- ble starch and dextrin, while the principal portion of the layers that are rich in cellulose undergoes this transformation only after a time. When starch which has been heated is treated with water of the ordinary temperature, the soluble starch and dextrin are removed and an organized residuum is left, which resembles the form and structure of the original granule and contains small quantities of unchanged granulose. This granulose can be further removed by extracting with water, and it appears to be in such a state as to be readily changed in its chemical properties. The grains give up the greater portion of the granulose to the water, thereby losing in mass but not in volume, retaining their structure, the residue consisting chiefly of cellulose. Grains extracted in this way become colored blue, or, at least, bluish-violet, on the addition of sulphuric acid and iodine, and the individual 40 DIFFERENTIATION AND SPECIFICiri' OF STARCHES. laj'ers swell up and separate from one another. When the granules are heated more strongly with a drop of water on the object glass, a deep-blue coloration ensues. The residue is not a uniform body, but contains, in addition to cellulose and granulose, a transformation product of starch, similar to dextrin, which reduces Fehling's solution, is colored red by iodine, and undergoes decomposition on treatment with water. It could not be decided whether or not this body was erythrodextrin. If a large quantity of the grains is triturated with powdered glass, which presumably would not produce decomposition or alteration, the substance yielding the red color can not be removed by repeated treatment with cold water; on the contrary, by the addition of iodine the residue becomes more intensely colored blue, while the rotatory power remains practically unaltered. The development of the starch-grain in the lactiferous cells of Eiiphorhiaccce, first observed by Meyer in 1836, was studied by Potter (Jour. Linnean Society, 1884, xx, 446), who describes phenomena of much interest. These grains are formed in the interior of rod- or spindle-shaped corpuscles which lie in the parietal protoplasm of the cell. The starch-grain is at first visible as a tliin streak in the interior of the corpuscle. Tliis streak, through the deposition of starch, assumes a rod- or spindle-shape, and both grain and corpuscle increase in size. When the grain has attained nearly its maximum dimensions in length and breadth, the starch-forming corpuscle collects at both ends of the rod-sha]3ed grain and forms masses at the ends which cause the former to assume a remarkable shape, resembling a long bone, such as the tibia. The lactiferous cells are polynuclear, and since when very young their diameter does not much exceed that of the nuclei, it follows that the starch-forming corpuscles which are always formed near the nucleus must be developed at the sides of it. The smallness of the diameter of the lactiferous cell necessitates the starch-forming cell being much longer than broad, and hence it comes about that the prunitive shape of the grain should be that of a rod. Later, however, when the cell has increased in diameter, the rod can also increase in diameter. The hilum of these grains is seen in the form of a line in the middle of the grain in the tlu-ection of its long axis. Tlie lines of stratification inclose the hilum and are roughly parallel to the outline of the grain. The grains are doubly refractive, and in all respects agree with starch-grains from other sources, since they are developed in the interior of starch-forming corpuscles and are strati- fied and rendered doubly refractive through the agency of the lamella?. The view that the starch-grain is composed, according to C. Niigeli, of granulose and cellulose was opposed by De Vries (Botanische Jahresberichte, 1885, i, 122), who holds that the so-called starch-skeleton does not consist of cellulose, because most or all of such skeletons are colored blue by boiling in Lugol's solution. He believes that only one carbo- hydrate (amylum or granulose) is present in the grains, that it may exist in different degrees of density, and hence that the starch cellulose is merely a dense form of granulose. In the same year Mikosch (Botanische Jahresberichte, 1885, i, 122) reported the results of his inquiries into the seat of origin of starch-grains. He placed leaves on a sugar solution, and observed that starch-grains arise not only in chlorophyl grains, but also in any part of the protoplasm. He found the same to take place in the potato tuber. In the cotyle- donary leaves of Zea mays and in the young tissue of Elodca canadensis the grains were observed to originate in the plasma of leucoplasts. At the same time an article by Belzung (Botanische Jahresberichte, 1885, i, 122) appeared on the mechanism of the development of the starch-grain. He fomid in sprouts develoji- ing in the dark that leucoplasts arise which soon after their appearance form several small starch-grains. After a short period of growth the grains fill the leucoplasts, and finally the grains in one leucoplast may fuse mto a single grain, or by atrophy of the leucoplast they may lie free as small grains in the protoplasm. In general, he states, there are three methods of development of starch-grains: (1) The formation of grains within the chromatoi:)hore by "resorption," but without marked subsequent growth (seen in sprouts, leaves); (2) the for- STARCH-SUBSTANCE, AND THE STRUCTUnE, ETC., OF THE STARCH-GRAIN. 41 mation in chromatophores, but with marked subsequent growth (in many Cotyledones) ; (3) the formation at tlie surface of the chromatophores (Phaius). The first method of growth gives rise to a type of small, lamellated starch-grains; the second and third methods to types of large lamellated grains. In the development of types of grains 1 and 2, the growth of the grains takes place presumably by a chemical metamorphosis of the chromatophore- substance, while in the growth of type 3 the role of the chromatophore is problematical. The so-called starch cellulose or skeletons prepared by C. Nageli by subjecting starch to the prolonged action of saliva, were obtained by Meyer (Botanische Zeit., 1886, xlix, 097, 713) by the agencies of dilute acids, pepsin, diastase, and saliva. Meyer found that after digestion had proceeded sufficiently long the skeletons are colored yellowish or reddish- brown with iodine, and that they consist of pure amylodextrin. This substance, he notes, crystallizes very readUy, usually in the form of spherical spherocrystals (rarely plate-like), resembling starch-grains which have a concentric structure. These spherocrystals, he states, behave similarly to starch-grains in polarized light, except that the dark cross is not orthogonal but diagonal, a phenomenon which e\'idently is dependent, he states, upon the orientation of the needle-crystals. The saliva-skeletons, as well as the acid- skeletons, behave exactly like the starch-grains towards polarized light. The microchemical similarity of the spherocrystals of amylodextrin and the saliva and acid skeletons proves, he contends, the identity of the substance of the three structures. This statement he sup- ported by the results of experiments with acids, alkali, and chloride of zinc and iodine, by which it was found that all were affected in the same way. The starch cellulose or skeleton-like substance described by Nageli and others was prepared by Griessmayer (AUgem. Bauer- u. Hopfenzeit, 1887, xxvi, 147) by subjecting 1000 granxs of potato starch to the action of G Hters of a 12 per cent hydrochloric acid for 100 days in the cold. From this preparation, after washing free of acid, sugar, etc., and drying, a quantity of starch cellulose was obtained that weighed 300 grams. This sub- stance was almost completely soluble in boiling water; and by freezing such a solution amylodextrin crystallized out in the form of spherocrystals composed of minute radial needles. Brown and Heron (Ann. d. Chem. u. Pharm. 1879, cxcix, 189) question whether cellulose is an original constituent of the grain or an after-formation. The conception of the rod-like crystalline arrangement of the particles of the starch- grain received further support in the investigations of Buscalioni (Botanische Jahresbe- richte, 1891, xix, 489) with Zea starch. The seeds of corn, not quite ripe, were broken up and boiled for 30 seconds in 1 c.c. of chloroform containing several drops of a solution of chromic acid. The pieces of the seeds were spread out and examined under the micro- scope, and numerous grains of starch swollen to various degrees were noted. In the slightly swollen grains radial streaks were seen which passed from the center of the grain to the periphery, and wliich were placed in two directions so as to cross each other at right angles, thus giving the grain the appearance of being composed of numerous rhombic pieces regularly arranged. In the more swollen grains the lines were replaced by small points arranged concentrically and radially. If the action of the chromic acid goes on for a longer time every trace of a definite structure vanishes and the whole grain appears to be a hyaline mass. In a later article (ibid., 1899, xxvii, 282), Buscalioni found in the cortex of the root of J uncus tenuis, especially within the endoderm, an accumulation of simple and com- pound starch-grains. The grains which appear in cells poor in starch were sometimes inclosed in a space of varying form and consistency, and in the nature of a dense mucus substance, so as to constitute a capsule. These capsules lie in a corner of the cell or on the cell-wall, while others are attached to the wall by a little stalk. Aggregations of grains are sometimes inclosed in a common capsule. Capsules were also observed which did not contain a trace of starch. Sometimes pectin, and also cellulose, were present. The author 42 DIFFERENTIATION AND SPECIFICITY OF STARCHES. states that the envelope of the starch-grain undoubtedly originates in the cytoplasm, as is shown in the different stages of development observed in the various i^reparations. The inclosed grains are stated to have an extraordinary similarity to the grains of the seeds of Vicia narhonensis, with the difference that in Juncus the capsules usually have stalks, while those of Vicia never have. The cause of the lamellation of the starch-grain is due, according to Zimmermann (Beitriige z. Morph. u. Physiol, d. Pflanzenzelle, 1890; Botanical Microtechnique, Trans, by Humphrey, 1893, 226), to varying water-content, which, according to the author, may be shown by examining moist and cLry starch-grains in Canada balsam. He notes that the complete removal of water can only be accomplished by drying at a temperature of 50° to 100° C, and that the use of dehydrating media, such as absolute alcohol, does not give demonstrative results. In an investigation of the mechanism of the growth of starch-grains, Eberdt (Jahr- biicher f. wissensch. Botanik, 1891, xxiii, 293) attempts the support of the then practically abandoned theory of growth by intussusception, and among other things he holds very different views from Schimper regarding the role of the colorless starch-builders, or leuco- plasts. Eberdt contends that the starch-grains which originate in the interior of chlorophyl granules become larger after they are freed from the granules, which to him indicates a formation by concentric lamellation, a process that can only be explained by growth by in- tussusception. The appearance of lamellation in starch-grains, he states, be it concentric or eccentric, might be explained by a leaching-out process brought about by plant acids, but the tensions present in the starch-grains would remain to be explained, and hence one must at the same time assume that simultaneously with this process swelling takes place. Eberdt agrees with Schimper that in non-asshnilating plant parts the formation of starch is due to albuminous bodies. Schimper ascribed an active role to these bodies, looking upon them as starch-builders, but Eberdt believes that they are passive, and that the plasma is the active agent in the transformation of assimilation products into starch. Eberdt holds that during the life of the plant, when no reserve material is produced, these albumin bodies perform no apparent function, and that one might in many cases regard them as waste products ; but when reserve material is being formed the role played by them is passive, and they become transformed into starch. Tliis transformation may be brought about, he states, in two ways, either from within outward, as when the form of the starch grain is analogous to the form of the body; or, the transformation takes place gradually from without inward, in which case the body is completely dissolved, and the form of the resulting starch-grain shows no resemblance to its form. In every case the transformation is assumed to take place only in the presence and by the agency of protoplasmic substance, which is the active factor, and which may become green under the influence of light. Eberdt gives a very lengthy discussion of the subject, but his chief conclusions may be summarized as follows: (1) The origin of the small albuminous bodies (Stiirkekorner of Schimper), which on account of their behavior he terms the "Stiirkegrundsubstanz," result from a differentiation of the plasma of the cell. (2) These bodies may be attracted to the cell nucleus, and later be arranged in groups or deposited singly about the nucleus. In every instance they are surrounded by a layer of protoplasm which is connected with the peripheral plasma of the cell by plasmic threads. (3) After the indi\'idual grains of the groups have been transformed into starch, the plasmic layer completely surrounds every group, or the layer breaks and the groups separate. In the former case the groups are surrounded by the plasma until the starch-grains are mature, and the individual starch- grains show no lamellation, or the grains separate before maturity, and in such cases con- centric lamellae appear, and in some instances eccentric layers. Grains of the last two types may increase in size after lying free in the cell-space. (4) In instances where the bodies of the Stiirkegrundsubstanz are not grouped around the cell-nucleus, the plasmic STARCH-SUBSTANCE, AND THE STRUCTURE, ETC., OF THE STARCH-GRAIN. 43 layer separates, and a part of it is in contact with and completely surrounds every one of these bodies. This plasmic layer brings about the transformation of the body into starch in such a manner that the albumin molecule is split by the action of the plasma. After the solution of the entire body, and after the newly formed starch-grain has broken through the plasmic layer, a part of this plasmic layer remains attached to the grain in the form of a hood, which usually has attachment to the rest of plasm. (5) Such grains always show eccentric lamellation which appears after the grains have broken tlirough the plasmic layer. (6) No growth occurs after the disappearance of the plasmic hoods. (7) The plas- mic layer or plasmic hoods might with reason be called starch-builders. The starch-building processes of non-chlorophyllous structures were also studied by Konigsberger (Botanische Centralblatt, 1892, xliv, 47). This author gives the results of his investigations of starch-building in Angiosperms. He does not agree with Schimper that in many plants the starch-grain originates only in the peripheral parts of the chloro- phyl-grains and soon breaks through the chloroplast, and on account of unilateral accre- tion becomes eccentric. He observed starch-grains in the parenchyma cells of Pelargonium which were eccentric although entirely inclosed by the chlorophyl. The following is a very brief summary of Konigsberger's results: (1) The formation of reserve starch in Angiosperms takes place through the agency of the leucoplasts and also by the direct activity of the protoplasm itself. The first process, which is observed in many Monocotyledones and in only a few Dicotyledones, must be viewed as the earlier, from which the second has been developed and predominates in many Dicotyledones. (2) In the Dicotyledones the leucoplasts, having performed their function, have in many plants entirely disappeared. (3) The beginning of the starch-grain is probably in the form of a deposition of amylodextrin. (4) The capacity to polymerize carbohydrates of less molecular weight into those of greater molecular weight, which latter are in the definite form of reserve material, was originally peculiar to the leucoplasts, but in many of the higher plants was later transferred to the protoplasm. (5) The starch-grains probably originate from amylodextrin, and when the starch is rendered into a soluble form for pur- poses of transportation it is reverted into amylodextrin. The form and method of formation of starch-grains were studied coincidently by Dodel (Flora oder Allegemeine Botanische Zeitung, 1892, lxxv, 266) and Binz {ibidum, 1892, Lxxvi, 34). Both made their studies mostly with the starch from Pellionia daveauana. Dodel makes rather general statements, while Binz goes somewhat into details. Dodel writes that after careful study he reached the conclusion that this plant fur- nishes the most suitable material for the study of both the morphology of starch-grains and of the starch-builders; that both grains and builders of Pellionia are well adapted to answer the question as to whether or not the grains grow by intussusception or apposition ; and that he is convinced that the growth is exclusively by apposition. He records that a section of the stem at a young internode shows in the parenchyma region many starch- grains of various sizes and forms, most of which have chloroplasts attached to them. In the region of the fibrovascular bundles, and in the pith, the starch-grains have attained most perfect and uniform development, while towards the periphery of the section the grains and the chloroplasts decrease in size. The form of the grains was found to vary considerably. All of the grains are stated to originate as small spheres which are found in the middle or toward the periphery of the spherical or ovoid chloroplast. As soon as the grains ha\'e attained sufficient size to protrude through the green starch-builders, the spherical form is changed, and the grains begin to grow unilaterally, that is, the main point of growth is where they are in contact with the chloroplasts, while the part of the grain that is not in contact with the chloroplast not only grows slowly but finally ceases to grow. The grains, at first spherical, soon become oval, wedge-shaped, reniform, cylindrical, or irregular in shape. All of this diversity of form, he states, is due to growth by apposition. 44 DIFFERENTIATION AND SPECIFICITY OF STARCHES. The form and size of the grain are determined by the form and number of the chloroplasts attached to the grain. Frequently two or more grains were noted to appear simultaneously in different parts of the chloroplast, when a compound grain results. Dodel then takes up the question as to whether or not the grains are formed by the attached chloroplast, and to answer this he studied the half-grown grains in a section of a stem of Pellionia that had been in absolute alcohol for several months, and in which the starch-builders had become decolorized. A section was washed in distilled water, and stained with a weak solution of methyl-violet, when it was found that the half-grown grains, besides being partially surrounded with the hood-shaped chloroplast, were inclosed by a very thin, colorless plasmic layer. In the matm-e grains this layer was no longer evident. Frequently several chloroplasts were in contact with one starch-grain, this being due to the division of the original chloroplast. The starch-builders increased in mass as long as the starch-grains continued to grow. The lamellae of the grains of Pellionia were not perceptible until a comparatively late stage of development. The young grains, as long as they were spherical, appeared to be homogeneous. It is very striking, he states, that the part of the grain about the hilum, which is the oldest, is not lamellated. The solution of the starch-grains, he found, takes place on the entire surface of the grain, even on the part covered with the choloroplast. During the process of solution the chloroplasts modify their form as the shape of the grains is changed. After a considerable i:)art of the starch- grain has been dissolved the chloroplast may resume its activity, and thus there may be an intermittent activity of the starch-builders. The formation of irregular grains is prob- ably due, he holds, to such secondary activity of the chloroplasts. According to Binz, the majority of the starch-grains of Pellionia are simple, and there are both half-compound and compound grains; but the compound grains do not origi- nate by division of the hilum of the simple grain, as Nageli assumed. The simple grains, he states, fall into two sharply defined categories: In the young state the grains are regular (spherical) in form, and inclosed by the green starch-builders. After they break through the chloroplasts they assume an oval form and become eccentric. Binz assumed that it is not the inner constitution of the starch-grain which determines the form of the grain, but the position of the grain in relation to the starch-builder. After the grains protrude from the chloroplasts, the latter are in contact with the grain in the form of a hood, and the growth of the grain takes place at the point of contact. Such grains take on a more or less regular form. The second category of simple grains includes grains of irreg- ular form, representing advanced stages of the regular forms. When the grains are sub- jected to a 4 per cent solution of potassium hydroxide the lamellse become evident. Two kinds of lamella} are noted, one kind being complete, and the other being incomplete and wedge-shaped. The complete or closed layers are few in number and pass entirely around the hilum. These layers are evidently formed when the grain is entirely inclosetl by the chloroplast. In every grain the innermost part is stated to be soft and watery, then follows a complete dense layer, then a watery layer, and so on, until the outermost or last complete layer is very dense, and the boundaries between the complete and incomplete layers are very marked. The inner complete layers are not complete in the early stages of the growth of the grain, and are only perceptible after the starch-builder has become ruptured by the extrusion of the grain and has assumed a hood-like form, and perhaps not visible before the first of the incomplete layers is deposited. This is a point, states Binz, that has always been used to oppose the apposition theory of growth, although nothing can be proved by it in such application. It is possible that the layers as such are deposited as a homogeneous mass from the very beginning, and that by subsequent changes, as by the varying absorp- tion of water, are rendered apparent. Binz agrees entirely with Schimper in that the growth of starch-grains occurs by appo- sition, yet he admits that there are no direct unimpeachable proofs in support of the theory. STARCH-SUBSTANCE, AND THE STRUCTURE, ETC., OF THE STARCH-GRAIN. 45 The softer layers, he observes, are present in niucli larger numbers than the denser layers. Very often a grain shows only 4 or 5 dense layers, while between them are numerous soft layers. The lamella' are separated from each other by dark lines which Binz holds must be regarded as very thin layers, for it is evident, he states, that when two substances of tlitfcrent rcfractivity are in contact a boundary line must result. Binz holds that Nageli's view that layers arise by the splitting of layers already present, is not justified, and that the correct view is that formulated by Schimper, which is that the starch-builder deposits the layers. Whether the lamella; are deposited directly or whether they are secondary phenomena Binz could not determine; nor was lamellation found to be attributable to the action of ferments, as Meyer assumed. Binz discusses various other topics, such as the relations between the structure of the chloroplast and the form of the starch-grain; irregular starch-grains, compound and half- compound grains; solution phenomena of starch-grains; and the structure of chloroplasts. The main conclusions of Binz's researches, briefly stated, are: (1) Nageli's theory that layers arise by the splitting of other layers does not hold good for the starch-grains of PeUioma. (2) The outer layer of the starch-grain is the youngest, and the innermost the oldest. (3) The spherical part of the grain forms when the grain is within the chlo- roplast, complete layers being formed so long as the grain is entirely within the chloro- plast and the wedge-shaped incomplete layers being formed after the grain has broken tlirough the builder, so that the form of the starch-grain is directly related to the form of the chloroplast. (4) The extruded grain grows at the point at which it is attached to the chloroplast. (5) Outgrowths and secondary growths are sharply defined from the original starch-grain. (6) If the chloroplast becomes detached from a certain part of the grain, the gro\\i.h of the grain ceases at that part. (7) Compound and half-compound grains do not originate by division of the original simple grain, but by the formation of several starch-forming centers in a chloroplast. The compound grains may also originate by the grouping together of several starch-builders, as seen in the pith of Philodendron and Convallaria stanhopea. All these points, he writes, favor the apposition theory of growth, but the theory can not be proved definitely until the origin of the lamella; is better under- stood. (8) The structure of the starch-grain has no influence upon the manner of corro- sion, but upon the intensity of it, for the softer parts are more easily corroded than the denser, and the several layers are more easily corroded in a radial than in a tangential direction. (9) The cliloroplast consists of a homogeneous ground-mass, or stroma, with embedded pigment spheres, the grana. (10) Starch-builders in the form of leucoplasts are present in the growing tips and they are structures homologous to the cliloroplasts, since under the influence of light they are transformed into chloroplasts. The microscopic structure of the starch-grain was studied by Biitschli (Botanisches Centralblat, 1893, lvi, 150), who found that starch-paste of medium density, on drying, first assumed a honey-comb structure and subsequently a fibrous form. In gelatinous paste, and in a dilute solution of starch that is frozen, the same characteristic honey- combed appearance was noted. From these results he concluded that the lameUation of starch-grains bears a close relationship to a honey-combed structure. Examinations of starch-grains in water were found by Biitschli to indicate that there are traces of a honey-comb structure in the inner layers and also in the hUum, but that such observations did not furnish definite results. When grains of a form of commercial arrowroot (canna starch) were heated in water at 60° to 70° C. until they began to form a paste, it was seen that every layer of the grains is radially striated, and that every layer is composed of "honey-combs" whose waUs consist of firm starch-substance, and whose contents are water or dissolved starch-substance. These radial structures, BiitschU states, do not extend through the entire grain, but alternate with each other in successive layers. This structure was seen in slightly swollen starch, but not when much swollen, because in 46 DIFFERENTIATION AND SPECIFICITY OF STARCHES. the latter there is brought about an increase in vohime in the outer layers, and this causes a disintegration of the internal organization. The honey-combed structure was seen even more clearly in the so-called artificial starch-grains which were prepared by condensation of a starch-solution, when under the skin that covers the solution a layer of grains consisting of starch was found. These grains, Biitschli recorded, show a characteristic honey-combed structure, and in polarized light behave exactly like natural starch-grains. Such artificial starch-grains, having a beautiful finely honey-combed appearance, are produced by freezing a dilute starch solution. Biitschli holds that there can be no doubt about the chemical nature of these artificially produced grains, and he also states that the structure of the starch-grain seems incompatible with the theory of growth by intussusception, but entirely in harmony with the theory of growth by apposition. On evaporation of a thin starch- paste there arose at the outer edge of a droplet layers having lamellated structural char- acteristics bearing a marked resemblance to the lameUse of the starch-grain. The origin of the honey-combed structure Biitschli conceives to be due to a separation of water during the concentration of the starch-solution, and to further loss of water by freezing. This investigation was supplemented by Biitschli (Botanisches Centralblat, 1896, Lxviii, 213) with the conclusion that the structure of the spherocrystal of inulin and that of the starch-grain is identical, and that both kinds of spherocrystals have arisen tlirough a honey-combed method of building. He obtained spherocrystals of starch in this way: A watery starch solution was prepared by boiling starch in water for 3 to 4 hours, when the solution filtered until entirely clear. To the filtrate was added an equal volume of a 5 per cent solution of gelatine, and the preparation was then evaporated almost to dry- ness, when spherocrystals appeared. These crystals had a diameter of 0.05 mm. In polar- ized light such crystals behaved exactly like the natural starch-grains, and the behavior of the two towards iodine was similar, but solutions of chloride of calcium and of chloral hydrate acted differently on the natural and artificial starches. Biitschli agrees with Meyer (see page 47) as to the presence of a-amylose and 0- amylose in natural starch, but he obtained only one of them in the form of artificial starch- grains. Artificial starch-grains have also been prepared by Rodenwald and Kattein (Sitzungs- ber. Kgl. pr. Akad. Wiss., 1899, xxiv, 628), Roux (Compt. rend., 1905, cxl, 440, 943, 1259), Maquenne and Roux {ibid., 1303), and St. Jentys (page 58). The former heated starch in a solution of iodine and iodide of potassium in a sealed tube for 15 minutes at 130°. A greenish liquid was formed which, they state, consisted essentially of a solution of starch- iodide and some sugar. The clear filtrate, they record, became decolorized upon heating, owing to the liberation of the vaporous iodine, which could be removed by a current of steam. Upon slowly cooling the solution, a white substance separated in the form of starch granules. These granules gave the characteristic blue reaction Avith iodine; they were insoluble in water; they formed a thick paste on boiling; and they swelled and formed a paste with potassium hydroxide. Roux found that by an incomplete degradation of amylocellulose an artificial starch can be produced which has a cellular structure similar to that of natural starch. These granules, while giving the blue reaction with iodine, do not gelatinize in hot water, and they therefore differ somewhat from both natural starch and the artificial starch-grains of Rodenwald and Kattein (see pages 59 and 111). The honey-comb theory of Biitschli was opposed by Puriewitsch (Berichte d. deutsch. Botan. Gesellsch., 1897, xv, 246). This experimenter used arrowroot, potato, wheat, and canna starches, whereas Biitschli used only canna starch. Puriewitsch records that starch- grains of arrowroot in the fresh condition show no honey-comb structure, nor is such a structure apparent in starch-grains that have been in water for 24 hours. On slightly swelling the grains, either by heating or by reagents, there is developed a quite definite honey-comb appearance, but this Puriewitsch does not regard as an actual honey-comb STARCH-SUBSTANCE, AND THE STRUCTURE, ETC., OF THE STARCH-GRAIN. 47 structure. The lines of laniolkitidn of such swollen grains are not regular but zig-zag, and the presence of these small projections gives rise to the illu-sory honey-combed appear- ance. Starch-grains from the potato likewise did not show a honey-comb structure in the fresh condition, and in only very few of the grains could such a formation be seen after swelling. Here also the appearance of such a texture is illusory, and merely an optical effect due to the zig-zag lamella?. In a few instances traces of the honey-comb structure could be seen in Cmma indica. This form of starch is very suitable for such an investiga- tion on account of the large and prominent lamellse. Grains from wheat did not show a honey-comb peculiarity, either in the fresh or swollen condition. Puriewitsch therefore declines to accept Biitschli's view of the peculiar honey-combed structure of starch. An important contribution to the literature of starch, covering quite a broad field of investigation, was published in 1895 by Meyer (Untersuchungen uber die Starkekorner. Wesen und Lebensgeschichte der Starkekorner der hoheren Pflanzen, Jena, 1895, mit 9 Tafeln und 99 in den Text gedruckten Abbildungen, S. 318). Meyer notes that even at this time our knowledge of the chemical substances which compose the starch-grain, and of the products of decomposition, is very meager in spite of the enormous amount of work that has been done. Meyer concluded from his investigations that in ordinary starch- grains only amylose and small amounts of amjdodextrin are present. The former he states is present in two forms, one of which dissolves in water at 100°, but the other not. The difference, he assumes, is due to the existence in starch-grains of anhydrous crystals, of crystals which contain water, of crystals that are soluble with difficulty in water, and of crystals easily soluble in water. The easily soluble modification of amylose he terms )3-amylose, and the difficultly soluble he terms a-amylose. a-Amylose. — Just as there is a dextrose anhydride, Meyer writes, which in cold water does not take up water of crystallization directly, there also appears to be amylose anhy- dride which upon boiling in water passes into a hydrate very slowly, and this fact has partly given rise to the conception of the so-called starch-cellulose. The term starch-cellulose, he notes, has been applied to very different substances — to mixtures of amylodextrin with a-aniylose, of dissolved /3-amylose and a-amylose, of dissolved /3-amylose and nitrogenous and fatty impurities, and to amylodextrin in an almost pure state. Meyer states that he concluded in 1886 that the skeletons of starch formed during acid and sali\'ary digestion do not consist of a substance that is contained originally in the starch-grain, but of amylo- dextrin, a transformation product of amylose. Later researches, however, showed him that this conclusion was not strictly correct, and that acid skeletons are of variable composition, and consist, depending upon the length of time the acid acts, of a mixture of j8-amylose, a-amylose, and amylodextrin ; or of a-amylose and amylodextrin ; or of amylo- dextrin alone. That amylodextrin alone is left after acid action for a sufficient time was proved, he records, by an experiment in which a 12 per cent hydrochloric acid solution was allowed to act on potato starch for eight and one-half years, at the end of which time the skeletons contained only amylodextrin. It is different, he records, with the saliva- skeletons, which, if the amylose hydrate has been dissolved out of the grains, consist of a mixture of amylodextrin and a-amylose. a-amylose he obtained by treating starch-paste with malt extract, or by the action of hot dilute hydrochloric acid on whole starch-grains. He found that when starch-grains are treated for a short time with saliva or with a cold dilute acid, and the skeletons extracted with hot water, there remained a residue which differed in its properties from other substances of the grains in that it was insoluble in hot water, and was colored red instead of blue with Lugol's solution. Meyer beUeves it probable that a-amylose is to some extent present in starch-grains from the beginning in crystallized forms which offer great resistance to boiling water, and to the penetration of iodine into the small crystals. That a-amylose is already con- tained in the intact grains is shown, he states, in the following experiment: If arrowroot 48 DIFFERENTIATION AND SPECIFICITY OF STARCHES. starch is allowed to soak at 70°, and the soaked grains are treated for several minutes with cold saliva, skeletons are obtained which no longer color blue with iodine, and which, at least in part, must consist of a-amylose, because the a-amylose could hardly be formed in so short a time from the solid crystals by the action of the saliva. He believes that it must therefore be assumed that a-amylose is present from the beginning in varying quantities in many gi'ains. Meyer goes on to state that the results of the experiments made up to that time do not show clearly the relationsliip of a-amylose to /3-amylose; that it seems to him that the differences between the substances are insignificant ; and that the future may show that they are merely hydrous and anhydrous forms of the same body. fi-Aviylose. — Meyer found that when starch-grains are subjected in water to a temper- ature of 138° there is obtained an apparently uniform solution of amylose, since at this temperature a-amylose goes into solution as j8-amylose. On cooling this solution, amylose separates from its solution in very small microscopic, viscous droplets, which do not mix with water below 138°, and wliich can only be regarded as a solution of water in amylose. This watei'-amylose solution can be produced at the swelling temperature of starch-grains, that is, at 60° to 70°, at which temperature the trichites of /3-amylose are changed into drop- lets of viscous solution, while a-amylose apparently remains unchanged. The trichites of a-amylose are so minute, he states, that they can not be seen singly, but they may become visible by fusion of a number of individuals in the form of a droplet. /3-amylose is insoluble in water at a temperature of 30° or below. At 60° the crystallized /3-amylose in small quantity forms a thick solution. At 100° it requires more than half its weight of water to pass from the crystallized to the liquid state. Amylodextrin. — Meyer states that this substance is of great interest because it is present in grains of starch which are colored red with iodine. Grains which color blue with iodine can readily be transformed into pseudomorphs consisting of amylodextrin. Amylodextrin was prepared by Meyer by the action of acids upon starch, 200 grams of starch in 2 liters of water and 8 grams of oxalic acid being heated to 100°. The unchanged amylose is allowed to settle by setting the preparation aside for 12 hours. The filtrate which is rich in amylodextrin is then concentrated and frozen to remove the amylose, and finally the amylodextrin solution is greatly concentrated, from which, upon standing, spherocrystals of amylodextrin separate. By recrystallization and washing with alcohol, a preparation is obtained which colors a pure red with iodine. The quantity of amylo- dextrin produced by this method was not very great, being only about 2 per cent of the amount of starch. Amylodextrin he found to be difficultly soluble in water (at 8°, 0.13 per cent; at 30°, 1.58; at 6°, 3.98; at 70°, 4.66; at 80°, 9.33; at 90° a dense solution that could not be filtered) . It is soluble in 50 per cent boiling alcohol, and more soluble in acid and saline solution than in water. Meyer also studied the growth of the starch-grains and the relations of the grain to the chromatophore. He states from the results of his researches that he is convinced that the substance of the chromatophore, whether it be chloroplast, cliromoplast, or leuco- plast, incloses the starch-grain as long as it lies witliin the lining cell, and that during no period of its existence does the starch-grain come in direct contact with the cytoplasm. The complete inclosure of the starch-grain by the substance of the starch-builder permits a direct influence of the latter on the grain, because every part of the grain is in contact with the mother substance. The cliloroplast appeared to be a more or less viscous droplet of a colorless or light-yellow substance {stro?na), in which was a viscous droplet (grana) of a substance colored dark-green with chlorophyl. In the stroma there may also be present well-formed crystals of protein and also starch-grains. In his earlier experiments, Meyer found difficulty in determining whether the stroma is colorless or a very light green, but he now expresses his conviction that the stroma contains no chlorophyl, that it is usually colorless, and seldom yellowish. In the leucoplasts of Pellionia, and in the "turned- STARCH-SUBSTANCE, AND THE STRUCTURE, ETC., OF THE STARCH-GRAIN. 49 green" leucoplasts of potatoes, he observed colorless grana lying in the stroma, and these he designatetl albigrana, but since that time he has not studied them. Inasmuch as only the chromatophores which contain grana; separate oxygen, it seemed to him likely that the grana» constitute the ajiparatus of assimilation. The colorless as well as the green parts of the chloroplasts have, he states, the capacity of forming starcli-substance. The starch-grain, Meyer holds, grows entirely within the chloroplast, and the approxi- mate proportion between the thickness of the chromatophore layer and the starch-layers growing below it makes it probable that the chromatophore controls the production of starch-substance. The stroma probably produces the starch, because the grain is most actively added to where the layer of stroma is thickest. Chromatophores embedded in a relatively fine-grained emulsion of cytoplasm show the simplest form — a rather viscous drop. But the form varies and is alterable in relation to contents which originate in the chromatophore. The various forms of these bodies have a close relationship to the forms of the starch-grains growing in them, provided that the shape of the chromatophore is retained for a sufficient length of time. Protein crystals growang in the chromatophore produce a change in form which may be constant for a long time, and they are very important in relation to the growth of the starch-grain. Pigment crystals act in a similar manner, and occasionally influence the form of the starch-grains. But the most far-reaching changes in the forms of the chroma- tophore are due to the starch-grains themselves. Not only does the growing starch-grain act upon the shape of the chromatophore, but the varying form of the chromatophore reacts upon the growing grain. There are cases where there are simultaneous growths of eccentric starch-grains and chromatophores, when after a certain stage in development both grain and chromatophore increase in size yet constantly retain the same form. The young chloroplasts of Dieffenbachia are mostly spherical and contain mostly centric starch- grains, a few eccentric. On further growth all of these grains were found to become elon- gated and eccentrically layered, so that the chloroplast bulges out on one side into a thin membrane. A grain wliich is centric at first may develop into an eccentric or concentric grain, according to the nature of the cliromatophore, which is influenced by the surrounding cytoplasm. The pressure of a relatively thin but viscid layer of cytoplasm upon the cliro- matophore exerts a marked influence upon the shape of the chromatophores, and also therefore upon the starch-grains growing within them. The pressure of the cytoplasm, together with the viscosity of the cliloroplast and the pressure of the growing starch-grain, gives rise to an accumulation of chloroplast substance at two diametrically opposite points of the starch-grain, so that the chloroplast assumes the form of a thin-walled sac with thick- walled ends. Several grains may grow in one chromatophore. Pellionia serves as a good example, Meyer states, of the change of form of such clilo- roplasts as contain two starch-grains. In well-nourished types of this plant the growth of the starch-grains normally exceeds that of the chloroplasts. The margins of the grains push diagonally through the chloroplast, and as the grains grow the tendency is to push the chloroplast out into a layer over the contact surfaces of two grains. In Dieffenbachia the chloroplasts are of rather thick consistency, so that two growing starch-grains press out the layer of the chloroplast substance between them more slowly, and owing to the greater mutual pressure of the grains the grains assume an irregular shape. A change in the form of the chromatophore that is far-reaching and very significant in its influence on the ultimate form of the starch-grain is brought about by an active solution of the contained starch-grains. Meyer states that if one observes in Dieffenbachia the chloro- plasts which surround the normal starch-grains during the various stages of solution, it is seen that the main mass of the chromatophore becomes rounded and that the thin layer of the chromatoplast which covers the largest part of the grain is closely joined to all of the transverse furrows that originate in the grain. At times he observed that the thin 4 50 DIFFERENTIATION AND SPECIFICITY OF STARCHES. stroma layer is decidedly denser in grains that are undergoing solution than in grains in process of growth, and here and there was noticed a pecuUar thickening of the layers in the furrows which results from the flow of the stroma layer to the furrows. In further studies of form and lamellation, Meyer confirms the observation of Schimper tliat the unequal growth of the starch-grain on different sides of the hilum is owing to the unequal deposition due to the relations of the grain with the starch-builder. Meyer points out that the starch-grains of Iris germanica grow most rapidly upon the side in closest con- tact with the mass of the leucoplasts, and also that the closed layers are widest on the side lying in closest proximity to the leucoplasts. The form of the last starch-layer deposited is approximately similar to the form of the chromatophore if the chromatophore does not undergo a change of form during the deposition. The final thickness of the layers, Meyer states, is dependent upon two factors: first, the thickness of the first layers of the grain, and second, the solution in situ which removes some of the starch before the occurrence of the following period of starch-growth. Like many spherocrystals of other carbohydrates, starch-grains are composed of alternate loose and rather compact layers, but entirely apart from this difference the layers can be differ- entiated from each other by the varying proportions of a-amylose, /3-amylose, and amylo- dextrin. In spherocrystals forming from a single substance in a homogeneous mother sub- stance the deposition of the layers occurs in such a manner that the saturation relations or other relations of the mother substance change periodically, so that a complete cessation of growth, or a partial solution, does not necessarily take place. If partial solution occurs, the lamellations become very prominent. Similarly, in the formation of starch-grains there are fluctuations in the state of the mother substance which affect the formation of the lamella;. In Meyer's experiments with Pellionia, twigs were starved so that the starch-grains were dissolved to such a degree as to leave only open lameUse. The twigs were now nourished until the deposition of starch began, when it was found that for every day of growth there was formed a thick, dense layer, and for every night a thin, soft layer. Dm-ing the day the twigs were assimilating actively and the growth of the layer proceeded steacUly, but at night there is a small supply of material, so that there may be deposited a thin, loosely compact layer, or there may occur a partial solution of the outer layer. Experiments were also made by Meyer with the starch of Adoxa moschatellina and Hyacinthus. The results in general showed that a relatively dense layer is deposited on the grain when the clu'omatophore produces starch-substance actively and steadily, and that when there is a slow and irregular production of starch-substance a thin, incompact layer is formed. The thickness of the layer is always proportional to the length of the period during wliich starch is continually produced. An examination of the periphery of a grain that had been exposed in the chromatophore for a long time to the action of diastase showed, Meyer states, that this part of the grain is never surrounded by a loose laj^er, as some have supposed. The diastase penetrates every part of the grain and brings about internal solution, but forms no sharply separated incompact peripheral layer. Even if the internal solution plays no part in the formation of lamellae it brings about a porous structure of the layers of the grain exposed to prolonged solvent action. This is demonstrated, he states, by the fact that potato starch-grains, which lie for some time in the sprouting tubers, are less refractive than the grains of the tubers which have sprouted. This factor is, he states, hardly sufficient to account for the presence of loose, porous layers in the central parts of many grains. Such a phenomenon is, he admits, explained by the fact that many starch-grains in their earlier stages of growth ai'e subjected to conchtions which cause the formation of open, porous layers. In Adoxa moschatellina the lamellation was studied by following the stages of development. In the growing grains of young shoots eccentric layers are deposited corresponding to the form of the chromato- phore. As long as the layers are not affected by solvent action, all of the layers, both STARCH-SUBSTANCE, AND THE STRUCTURE, ETC., OF THE STARCH-GRAIN. 51 dense and porous, are entirely closed. At the apex of the grains the layers are very soft, but entirely closed, and the layers widen toward the base of the grain. Later, during short periods of solution, approximately uniform layers of substance are dissolved from the jieriphery of the grain, but the solvent action is more marked at the base, with the result of opening some of the lamelliE laterally. If such open layers are later surrounded by newly formed closed layers, it is seen that in the periphery single open layers are located between numerous closed layers. When the grain has attained its normal mature size, its base is often subjected to active solution, whereby the grain becomes attenuated and the layers of the base loosened. Meyer suggests that the dense layers are relatively rich in a-amylose, and that amylo- dextrin occurs in largest quantity in the less dense, porous layers. It is, he states, possible that there are starch-grains which contain only amylodextrin, or amylodextrin and /3- amylose, since not every chromatophore possesses an equally high capacity of condensation. The less dense layers may be distinguished, according to Meyer, by placing the grains in a solution of methyl violet and adding calcium nitrate, which causes a granular precipita- tion of these layers, which becomes stained. Such precipitation-staining experiments were carried out in exienso by Fischer (page 55). Referring to the causes of the variations in the forms of starch-grains in the same plant, Meyer writes that the differences are due not only to the fact that the grains were not formed at the same time, but also to the fact that every cell has its own biology, and that even every chromatophore has its own individual properties. In cUfferent organs of the same plant starch-grains can take on a great variety of forms. In the tubers of potato, for example, the leucoplasts form mostly solitary, monarch, eccentric, conical, or oval grains, with a length of 2Q0^i, and having definite irregular layers. The polytone starch-grains of a gi\'en plant part differ more from one another than the monotone grains, because in them the original differences are magnified. (Further reference to Meyer's investigations and liis descriptions of starch-grains will be found on page 67.) In opposition to the view that the alternation of light and darkness is a cause, or the cause, of the lamellation of the starch-grain, Fischer (Beihefte z. Botan. Centralbl., 1902, XII, 227) records that when cuttings of Pellionia daveauana were kept in the dark for two weeks, and then put for one week where the liglit of an incandescent lamp would fall on them continuously, a few lamellated grains were found like those described by Meyer as owing theu- lamellation to alterations of illumination. (See also St. Jentys, page 58.) Some of ]\Ieyer's conceptions and conclusions were criticized by Robert (Ber. d. deutsch. botan. Gesellsch., 1897, xv, 231). According to Meyer, /3-amylose remains unchanged during the conversion of raw starch into starch-paste, but Robert states that, if this were true, when the temperature falls below the minimum required for forming paste, the form of amylose that is insoluble in water would again assume its original con- dition, which is not the case. At the temperature of gelatinization, Robert states, amylose undergoes a permanent change, being converted mto a substance capable of swelling to a much greater degree, the nature of the change being apparently that of a hydrolytic sphtting of the molecule of /3-amylose into smaller molecules of sunilar composition, and probably a further spUtting into still smaller molecules at 138°, at which temperature the paste is soluble in water. Robert contends that the terms a-amylose and /3-amylose are not acceptable, and he proposes as substitutes farinose and gi-anulose, respectively, and he suggests that the term amylose be retained for the substance which results from the gelatinization of granulose (/S-amylose). He disagrees with Meyer's statement that every grain is completely and constantly surrounded by the substance of the chromato- phore, without, however, mentioning the nature of his observations. Salter (Jahrbiicher f. wissensch. Botanik, 1898, xxxii, 117) also followed up Meyer's work, and some of Meyer's statements he confirms, but others he opposes. The following 52 DIFFERENTIATION AND SPECIFICITY OF STARCHES. is a brief summary of Salter's results: (1) The starch-grain, in all stages of its develop- ment, is sharply differentiated from the plastid in which it originates. In no instance are transition lamellse noted. The substance of the grain is separated out by protoplasm, but not produced by the gradual transformation of the successive layers of protoplasm. (2) Meyer's observation that the chromatophore forms a complete and constant but very delicate integument of the starch-grain was confirmed in many instances. (3) The stain- ing reactions prove the accuracy of the view that the lamellated appearance of starch- grains is primarily to be ascribed to the differences in density and also to the varying absorption capacity of the different layers; and that the coloration by stains is due merely to the imbibition of coloring matter between particles, the layers which stain dark being comparatively loose and watery, the darker the staining the less the density. (4) The conclusion reached by Nageli regarding the uniform density of the young grains, and the manner of formation of the hilum and the first soft lameUse, was confirmed. (5) All grow- ing grains appear to possess a dense peripheral layer, which gives no indication of a lamel- lated structure. The lamellae attain their distinguishing characteristics when covered by the lamellae formed subsequently. (6) A progressive but not uniform decrease in density is noted in passing from the periphery toward the hilum, or structural center of the grain. (7) Lamellation bears a close relationship to the diverse conditions wliich arise during the formation of the layers. It is noted that layers soft at first become dense after a time. (8) Changes of the surface of the grains may be owing to ferment action. (9) Every lamella consists of structm-al elements arranged radially. The substances designated a-amylose and ;8-amylose by Meyer were studied by Syniewski (Annal. d. Chcm. u. Phar., 1899, cccix, 282), who found that when starch-paste is treated for several minutes with malt extract all of the paste is dissolved except a small portion that remains as a flocculent mass, which is the amylocellulose of Meyer. According to Meyer this substance, when boiled in water and again treated with malt extract, is for the most part dissolved, leaving a residue that is a-amylose. Syniewski ascertained in these experiments that from starch-paste amylocellulose could be obtained in quantities ranging from 0.7, 2.4, 3.6, and even to 13 per cent, according to the concentration of the paste; and also under different conditions variable percentages from paste of equal concentration. These facts led to the belief that amylocellulose is formed from the starch-substance that originally is dissolved or swollen. He notes that when a 5 per cent starch-paste is heated in a closed vessel under a pressure of 3 to 4 atmospheres, the starch-substance is entirely dissolved. From tliis solution on cooling there separated a gelatinous mass which is insolu- ble in cold water and not acted upon by diastase. On heating this mass it went into solu- tion, and upon coohng it again became gelatinous, but upon frequent repetition a stage was reached when not all of the gelatinous matter was dissolved upon heating. This residue is not soluble in hot water, and is, he states, the a-amylose of Meyer. Syniewski foimd that, from a solution which according to Meyer contained only ;3- amylose he could obtain amylocellulose, and also that by continued boiling under pressure he could obtain the a-amylose. Amylocellulose and a-amylose therefore, he writes, originate subsequently from the substance of the starch-grains that first went into solution. When malt extract was added to starch-grains suspended in water, and the mixture heated to 70°, the starch was found to be dissolved completely without leaving any a-amylose, which would necessarily remain if it preexisted in the starch-grain. There is, therefore, states Syniewski, no evidence to uphold the theory that starch is a compound of two substances. To the contrary, he holds, since the starch-substance is homogeneous at 138°, a temperature not favorable for the origin of derived products, even if these by transformation go into solution, that it is safe to assume that the starch-substance is a single body. By the action of sodium peroxide on starch a form of soluble starch occurs which Syniewski regards as the simplest structural element of the complex starch-molecule, and wliich he terms STARCH-SUBSTANCE, AND THE STRUCTURE, ETC., OF THE STARCH-GRAIN. 53 amylogen. He believes that st arch (as well as the products obtained from it, such as Meyer's a-aniylose, Lintner and Diill's araylodextrin, the granulose of various investigators, etc.) consists of many amylogen molecules joined together in the form of carbinolanliydrid combinations. (See Chapter III, p. 130.) Strasburger (Text-Book of Botany, 1903) supports the view of the existence of a single fundamental substance, which, however, he admits exists in two forms in the grains, one of which (the inner part) is soluble in water at 100° C, and the other (the outer part) not soluble. He also notes that amylodextrin may be present, and that grains of certain plants, such as Oryza saliva var. and Glutinosa may consist principally of amylodextrin, and therefore color red with iodine. (See Alocasia, Chelidonium, and Macis.) Bloemendal (Woch. f. Brau., 1909, xxxiii, 436, 449) is also among the opponents of Meyer's conception of the nature of the substances constituting the starch-grain. Bloe- mendal studied the chemical composition of various starches, and came to the conclusion that the o-amylose and |3-amylose of Meyer correspond to the cellulose and amylose of Niigeli, and if not identical differ only in water-content; and that the one form can be readily transformed into the other. He states that no amylodextrin is found in normal starch if sufficient care has been exercised in the course of preparation. (See page 170.) Pfeffer (The Physiology of Plants, 1900-6, 2d edition, 3 vols., Oxford, Trans, by Ewart) records that starch-grains remain within the chloroplastids or leucoplastids in which they were produced until they are dissolved or removed, so that under normal conditions they are never found lying free in the protoplasm or cell-sap. The grains are described as being composed mainly or entirely of amylose, usually turning blue with iodine; but in certain cases, as in the seed-coat of Chelidonium, Oryza, etc., they are recorded as being mainly composed of amylodextrin, and other dextrins as well, so that a red coloration is produced with iodine. These dextrinous substances are stated to be formed as intermediate products of diastatic action, so that the starch-grains which redden with iodine may be regarded as having undergone partial conversion into sugar. Starch-grains, whether formed by chloroplastids or leucoplastids, are able, he WTites, by virtue of their power of imbibition and swelling, to take up dissolved substances, and hence to interpolate new particles be- tween the older ones. They might therefore, states Pfeffer, grow by intussusception, although the researches of Schimper and Meyer have shown that starch-grains usually grow by apposition. The structure and lamellation of the starch-grain, he holds, are mainly the result of its growth by the apposition of successive layers, but, as in the cell, secondary modification is possible by means of solvent and other agencies acting on the surface of the starch-grain. Starch, like reserve cellulose, may be partially or entirely dissolved when required for food, and hence at any time there may occur a solution or a renewed deposition. A starch-grain does not dissolve only from the outer surface, but also from within, so that frequently a skeleton of the grain is produced. Typical sphero- crystals often dissolve in an equally peculiar manner, and changes in the condition during their formation may result in the production of denser layers than those first formed. Pfeffer records that the shape and growth of the starch-grain depend upon a variety of factors, such as the specific character and activity of the amyloplastid (leucoplastid or chloroplastid), the position of the starch-grain in it, and also upon a number of condi- tions which influence these and other relationships. Hence, starch-grains in the same cell are not always precisely similar, while in diversely differentiated cells of the same plant they may assume widely different shapes, as, for example, those in the lactiferous cells of Euphorbia when compared with those in other cells of the same plant. Usually the grain continues to grow only so long as it is in contact with the plastid; and when the latter is attached to one side only, growth takes place in this direction, and as a consequence an eccentric lamellation results. The enlarging starch-grain, he states, not only regulates its own growth bj^ causing the distension and shifting of the plastid, but also, as in case 54 DIFFERENTIATION AND SPECIFICITY' OF STARCHES. of a growing crystal, the part already deposited influences the shape of the subsequent additions. The internal strains observed in starch-grains, he holds, could easily be pro- duced by apposition combined wdth subsequent internal changes, and no arguments as to the mode of growth can be deduced from the supposed molecular structure, which is itself a mere hypothetical abstraction. The power of forming starch is, according to Pfeffer, possessed by etiolated chloroplasts as well as by many non-chlorophyllous chromatophores, but all chromatophores have not this power, and certain chloroplastids never contain starch, perhaps because solvent enzymes may be present which dissolve the starch as fast as it is formed. The sequence of events which occur in the formation of the starch-grain was studied by Timberlake (Annals of Botany, 1901, xv, 619). His inquiries were made with Hydro- dictyon. He notes that the pyrenoid (first described by Schmitz) is a spherical protein body that forms a part of the chromatophore, and that it bears a morphological relation to it similar to that of the nucleolus to the nucleus. In cells of this plant chlorophj'l is distributed in the whole peripheral protoplasmic layer of the cell. No distinct chromato- phores are observable in the cells. In the cells that contain an abundance of starch, prac- tically the whole layer of the protoplasm, from the plasma membrane on the outside to the vacuolar membrane on the inside, is filled with starch-grains, all of which originate in pyrenoids and later are transferred bodily to other parts of the cell. The whole process of starch formation could be traced from certain structural changes occurring in the body of the pyrenoid. The first indication of the changes leading to the formation of starch consists, Timber- lake found, in a difTerentiation of the body of the pyrenoid into two portions, one of which is destined to become transformed into a starch-grain and the other to remain unchanged. The part that is to form the starch-grain stains less densely, and instead of red becomes a neutral gray or a faint orange, in safranin-gentian-orange stain. The dense homogeneous structure becomes spongy, with regions of varying density. Very often the dense regions are so distributed as to give an alveolar appearance. The denser regions become more prominent and take up the blue stain. Between the fully formed starch-grains and the unchanged remainder of the pyrenoid a thin zone of slightly stained material appears. When the grain is fully formed it is seen to he in a vesicle or vacuole in the cytoplasm, but without being surrounded by a differentiated membrane. The mature grain was observed to have practically the shape of the pyrenoid from which it was formed. This point, states Timberlake, aids in establishing that all starch is formed from the pyrenoid. When starch is produced rapidly a second grain will be built at once before the pyrenoid regains its original form. The long axis of the second grain is at right angles to that of the first. As the process precedes rapidly, the grains as they are formed are continually crowded outward by the last-formed grains, until finally they are densely packed through- out nearly the whole of the protoplast. He therefore looks upon the pyrenoid as being directly the seat of a process which results in the formation of starch, which process, he states, is an exceedingly complicated one, as is shown by the structural and microchemical changes ehcited by differences in staining. Since the pyrenoid seems to be of a protein nature, and since a part of it (at least in Hydrodictyon) seems to be converted into starch, Timberlake suggests that the process involves the breaking down of a protein into carbohydrate. He notes that Boubier sug- gested the hypothesis that the pyrenoid is comparable to the leucoplast of the higher plants, and that the method of starch formation in it is similar to that in the leucoplast. The most serious objection to the comparison suggested by Boubier seems to lie, Timber- lake believes, m the fact of the difference in structure between the two. Timberlake states that it is difficult to differentiate the leucoplast from the rest of the protoplasm, and that when it is differentiated it has a granular or reticulate appearance; but the pyrenoid STARCH-SUBSTANCE, AND THE STRUCTURE, ETC., OP THE STARCH-GRAIN. 55 appears homogeneous, dense, and sharply Iiounded. Hence the relations of the methods of starch formation in Hijdrodiclijon and in chromatophores without pyrenoids must be regarded at present as uncertain. In a series of precipitation-staining experiments, Fischer (Beihefte z. hot. Centralbl. 1902, XII, 226) tried various aniline and other dyes, using picric acid as the precipitant instead of calcium nitrate, which was employed by Meyer (page 51). Such grains were found (1) to be unaffected by nigrosin, Hessian purple, diamond red, Kongo red, carmine, aniline blue, and cyanin; (2) to be uniformly colored by acid fuchsin, corrallin, eosin, crocein, tropseolin, Martin's yellow, and hemotoxylin ; (3) to be in the form of fine pre- cipitates by fuchsin, safranin, indigo carmine, methyl blue, methylen blue, and indulin; (4) to give large crystalUne grains by methyl violet and gentian violet; (5) and exhibit radial needles by Bismarck brown, chrysodin, malachite green, brilliant green and thionin. These precipitation-staining reactions were confined to the less dense layers. The outer surface or layer was unaffected. Ivi-aemer (Botanical Gazette, 1902, xxxiv, 341) conceives the starch-grain to be a direct product of the polymerization of soluble carbohydrates of either the glucose or cane-sugar group, and that during the processes of formation the products consist of three substances, two crystalloidal in the form of starch-cellulose and granulose, and one colloidal. These are assumed to occur as follows: (1) In the point of origin of growth (the hilum) the colloidal substance is associated with a small proportion of cellulose, as also in the alternate lamellae. (2) In the other layers occurs the granulose associated with a small amount of colloidal substance, and possibly also some cellulose. (3) The peripheral layer of the grain is not readily acted upon by reagents, and is quite elastic and more or less porous, probably consisting of an anliydride of cellulose. (4) In some cases some of the dextrins, or some of the non-coUoidal or crystalline carbohydrates, such as maltose, dex- trose, levulose, etc., may be present, probably formed as results of alterations taking place in the grain. Certain aniline dyes were used by Kraemer to differentiate the lamelltc of the starch- grain. Freshly prepared starch-grains, or commercial starches, were treated with weak aniline solutions of safranin and gentian violet and allowed to dry at ordinaiy temperature, when it was observed that certain parts of the grain took up the stains more readily than others. He also used iodine, water, and various other reagents for the same purpose, and he makes the important observation that the grains from different plant som'ces do not react the same. The gentian-violet stain was found to be more pronounced in its effect upon potato starch than upon the starch of wheat and corn, the stain being held by the point of origin of growth (the hilum) and by the lamellse alternating with it. On the other hand, he ascertained that safranin is a better differential stain for wheat starch, being held in certain of the lamelke (usually not more than three or four of them being affected) and also in numerous radial clefts and channels. Corn starch did not appear to take up these stains as readily as either wheat starch or potato starch, and there was no dilTerentia- tion of lamellae, which Ivraemer thinks as being probably due to the peripheral layers being denser and less permeable. In referring to the previous work of Salter (Jahrbiicher f. wissensch. Bot., 1898, xxxii, 117), &aemer takes exception to his statement that the dye has not a selective or spe- cific action on the layers, and he states that Salter's figures show that certain parts of the grain stain more than others, and that he believes Salter's work, as well as that of Meyer {loc. cit.), with methyl violet, correspond with his own in that the layers that are colloidal in character take up the stain. He notes that the lamellae which are not affected by the aniline stains become blue with iodine, the alternate laj'ers and the point of origin of growth remaining unaffected. The layers thus affected by iodine, he states, are the ones that are rich in granulose and more clearly defined in the grains of potato and wheat starch 56 DIFFERENTIATION AND SPECIFICITY OF STARCHES. than in corn starch. These lamellte, Kraemer found, become crystalloidal in character on treatment with water at 60° or 65° for about an hour, and also with chromic acid, calcium nitrate, saliva, and other reagents. The effects of these reagents upon different starches were found to be not identical. Upon potato starch the first effect is to make the lamellae more distinct; this is followed by the development of the crystalloidal character of the lamellae, which is most pronounced in those lamella; which are colored blue with iodme; this in turn is followed by the pro- duction of small tracts or channels which connect contiguous lamella?; then larger channels form which are plume-like in appearance, the grain in the meanwhile sweUing quite per- ceptibly, the middle portion becoming clearer and assuming a zig-zag outline, between which and the periphery a number of crystalloidal lamellse arise; the grain now becomes spherical and marked by a number of concentric lamellse near the periphery, and the lamellse finally rupture, followed by a gradual solution of the grain (see page 172). In wheat starch the development of the crystalloidal character of the lamellfe is fol- lowed by the formation of narrow, interrupted or continuous, radial channels near the periphery of the grain, which are sometimes connected with lamellse located near the middle of the grain; the grain meanwhile swells perceptibly, the center becomes clearer, and the contents are crowded into crescent-shaped halves which are still connected at the poles; the contents of each of the halves consist of crystalloidal lamellse in wliich are then produced small tracts or channels connecting the contiguous lamellse, the halves in some instances finally separating and slowly dissolving. The first effect of reagents upon corn starch is to bring out the point of origin of growth, which becomes larger and in some cases more or less zig-zag in outline; between this and the periphery of the grain there arise more or less interrupted or continuous radial channels, usually the latter; the crystalloidal structure of the grain develops slowly and is most pronounced when the grain has swollen to two or three times its normal size. At this stage the center of the grain has become clear and the point of origin of growth has become obliterated in some cases, and between it and the periphery occur numerous crystalloidal lamella; similar to those observed in potato starch. Finally, the peripheral layer ruptures and there is a gradual disintegration of the grain. Sometimes it was noted by Kraemer that the grain appears to separate into as many parts as there were arms to the point of origin of growth, particularly when acted upon by saliva or diastase. Kraemer, in summing up liis observations, concludes: "The starch-grain consists of colloidal and crystalloidal substances, these being arranged for the most part in distinct and separate lamellse, that is, at the point of origin of growth, and in the alternate lamella; the colloidal substance preponderates, associated with the crystalloid cellulose; whereas, in the other layers the crystalloid substance, consisting for the most part of granulose, occurs in greater proportion. As further evidence of the presence of these crystalloidal and colloidal areas we may say that the peculiar behavior of the colloidal layers toward aniline stains is analogous to the behavior of a section containing nnicilage cells towards these dyes, the latter being taken up by the mucilage cells alone." Kraemer holds that differences in the starch-grains show that starch, instead of being a uniform substance, is in fact composed of several substances in varying proportions, but more or less definitely arranged. Denniston (Trans. Wisconsin Acad. Science, Arts and Letters, 1904, xv, 664), in his studies of the growth and organization of the starch-grain, also made use of aniline dyes to differentiate the different layers, and he furthermore noted that the differences of the various layers of the same grain vary when the grain is mounted in water, iodine solution, and a solution of gentian violet and orange G, respectively. Layers which, for instance, appear single in water may appear double or multiple in a color reagent, and a single layer brought out by one stain may appear as two by means of another. The results of the STARCH-SUBSTANCE, AND THE STRUCTURE, ETC., OF THE STARCH-GRAIN. staining reactions with canna starch are summarized by Denniston in tables 1 and 2, the order of the parts being from the margin inward. Table 1. A (in water). B (in iodine solution). C (in gentian violet and orange G). a. Refractive layer. 1. Crevice. 6. Refractive layer. 2. Dark line. c. Dark, slightly refractive layer. f a' . Light blue. I a". Dark blue. L Dark line. 6. Pale-blue layer. 2. Dark line. c. Dark-blue layer. a' . Orange. a". Light blue. 1. Narrow blue layer. ( b' . Dark-blue layer. \ h". Light-blue layer. 2. Narrow light-blue layer, c. Daik-blue layer. Table 2. A (in water). B (in iodine). a. Highly refractive region, a'. Layer which has not so fully taken on the nature of starch, hence is faintly blue in color, a". A blue starch layer. 1. A dark line, probably a ' This layer is broader and crack filled with watery colloidal mass. h. Highly refractive layer. 2. A dark line similar to 1. c. Slightly refractive layer. paler in color. and Contracted slightly stains blue, c and layers anterior to c have contracted, leaving space at 2. In iodine this layer stains uniformly with those next to it on inside. It is pale blue in color. C (in alcohol). This layer is now pale blue in color, the color becom- ing lighter from inside toward periphery. There is no sharp line separat- ing two parts. In alcohol this layer is still broader. Contracted a little more and blue partly removed. Contraction goes on with consequent broadening of 2. This iodine is easily remov- ed, leaving layer jjale blue in color. D (in gentian violet and orange G). The layer a' is of different composition and takes orange; a" is starch and takes gentian violet like rest of grain. This layer is about the same width as in alcohol. It stains pale blue. Characteristic blue with gentian violet. Stains pale blue, contains relatively small amount of starch. This layer stains less deep- ly than a or h. The layers which take the deepest color with iodine and gentian violet Denniston regards as being the more dense, but, in case of such precipitants as were used by Meyer {loc. cit., p. 51) and Fischer {he. cit., p. 55), the less dense. The limitation of the orange layer, he holds, can not be due to hindrance to the penetration of the dye, because the layer does not become thicker in time and because in the case of crushed grains in which the dye has access at once to all of the layers the parts adjacent to the outer layer do not become yellow. He also noted that the central part of Canna grains stain yellow, and that frequently young grains stain entirely orange with the exception of one or two dots, thus in agreement with the view expressed by others that the young grain is of different composition from the later superposed starch. As further evidence of a differentiation of the outer layer, he found that weak iodine may penetrate to the inner part of the grain, coloring it blue without in the least coloring the outer part (as had been found by Nageli). The outer layer he believes is in the natm-e of a transition substance undergoing erosion or deposit. Further evidence that starch is not a uniform substance was found by Harz (Beiheft. z. botan. Centralbl., 1905; Woch. f. Brau., 1905, xxii, 721) in experiments with solutions of chromic acid, and cliromic and sulphuric acids, in which the starch-grains were macerated for 24 hours and then washed with cold water. Not only did the various kinds of starch, but also different grains of the same starch, differ widely in their behavior; from which Harz asserts that starch can not be a physically uniform substance which consists of granules differing merely from each other according to a denser or looser constitution of their ultimate complexes. He states that amylodextrin also did not behave like a uniform substance, but seemed to be made up of a number of molecular groups which differ in com- 58 DIFFERENTIATION AND SPECIFICITY OF STARCHES. plexity and density of internal structure, and tliat in examining the decomposition products of starch it was not until the achroodextrin stage of the degradation of the starch molecule was reached that the products exhibited an apparently uniform molecular condition. Aniline dyes were used by Gastine (Compt. rend., 1906, cxlii, 1207; Jour. Soc. Chem. Ind., 1906, XXV, 655) for the detection of rice flour in wheat flour. A sample is treated on an object glass with a solution of a suitable dyestuff, drying the preparation at about 30° C, and then completing the desiccation by heating for a few minutes at 110° to 130° C. The preparation, he states, should be mounted in cedar-wood oil or Canada balsam and examined under the microscope. For staining, a solution of aniline blue or green may by employed of a strength of 0.05 grm. in 100 c.c. of 33 per cent alcohol. This treatment has the effect of showing the hilum of the minute rice starch-grains, while wheat starch rarely exhibits a visible hilum. In rice flour isolated starch-grains are rare, the grains generally occurring in clusters in starch-bearing cells. These clusters, according to the above method, have a very characteristic appearance, since the hilum of each starch-granule appears as a reddish-colored point, these red points being grouped quite uniformly in symmetrical arrangement resembling a mulberry under a high magnifi- cation. Wheat and rice starches do not take up the dyestuff; only the nitrogeneous mat- ters are dyed. The fragments of rice flour therefore appear colored; the medium-size and large granules of wheat are practically uncolored, but the small granules of the wheat, in which the interstitial nitrogeneous matter is more abundant, are distinctly colored. The grains of corn and buckwheat starches behave like rice. Potato, arrowroot, and sweet potato starches, unlike the cereal and leguminous starches, absorb the dyestuff directly. According to Maquenne and Roux (Compt. rend., 1905, cxl, 1303; 1906, cxliii, 124) starch consists of about 90 per cent of amylocellulose and about 10 per cent of amylo- pectin. The former they describe as being devoid of gelatinizing power, but the latter as gelatinizable. (See Chapter III, page 112.) Day (U. S. Dept. Agriculture, Office Expt. Sta. Bull. 202, 1908) records three substances in starch-grains, which are designated blue amylose, red amylose, and rose amylose, in accordance with the color reaction with iodine. (See Chapter IV, page 166.) Some extremely interesting and original views of the chemical nature, and also of the cause of the peculiar structure, of the starch-grain were published by St. Jentys (Bull. d. I'academie d. sc. d. Cracovie, 1907; Jalir. ti. d. Fort. d. Theirchemie, 1907, xxxvii, 99), who records that the presence of tannin in the cell-sap, and the fact that a solution of starch yields with this substance a preparation which is insoluble in cold water, led him to the supposition that tannins enter into the composition of the starch-grain. Doubts as to the components of the starch-grain were first aroused in St. Jentys by the results of a series of experiments in connection primarily with the peculiar reactions of the starch- grain and starch-paste with iodine. St. Jentys observes that starch-grains, as is well known, are rarely colored a pure blue with iodine, but usually violet, and sometimes even black. A still greater variety in color may be observed in dried starch-grains and starch- paste which have been treated with iodine, the violet parts gradually going over into a cherry-red, copper-red, or orange, and finally into a brownish-yellow. These various colorings (which Nageli had already noticed) are due, St. Jentys states, to the presence in the starch-grain of substances which differ in their behavior toward iodine. For exam- ple, the starch-grains of potato pulp, from which the tannin had been removed by means of a methyl-blue solution instead of water, were colored an indigo-blue with iodine after the removal of the methyl-blue; the outer surface of the intact grains was colored violet, while the inner part became more of a blue color. Other reagents also acted differently on the outer and the inner parts of the grains. St. Jentys could extract nothing by cold water from intact grains, but some particles passed into solution from the pulverized grains. Boiling with water, he states, evidently STARCH-SUBSTANCE, AND THE STRUCTURE, ETC., OF THE STARCH-GRAIN. 59 causes the outer layer to swell or to dissolve, the whole grain producing starch-paste; but the starch does not go into complete solution, because undissolved flakes always remain behind and take on a violet-color with iodine, while the solution turns blue. The process of gelatinization also varied in starches of different origin. For instance, in acorns the starch gelatinized only after being heated in water to 77.5° to 87.5° C, while the other starches did so at a much lower temperature, the difference being due, he states, to the richness of the starch in tannin. The presence of tannins also explains, he believes, the failure to obtain starch-paste from raw boiled potatoes and from acorns. In fact, he found that starch-paste will not be formed if the grains are boiled in a weak solution of tannic acid. The starch in such a mixture is in flakes and the preparation can not be used as a paste. The presence of tannin as a constituent, especially of the outer layer of the starch- grain, is further indicated by the behavior of starch to iodine in solutions of calcium, magnesium, and zinc chloride, of potassium bromide and potassium iodide, and of con- centrated Ij^e. The action of these reagents depends, as revealed by the microscope, upon the solution of the outer layer of the grain by these compounds. A certain power of resist- ance to solutions of diastase is one of the peculiar properties of the outer layer. Starch- grains were found to be affected by diastase only after the outer layer had here and there been eroded by enzymes. St. Jentys attempted to prove by direct experiments the hypothesis that tannins enter into the composition of the starch-grain. Potato starch was digested with concen- trated sodium hydroxide, and from the solution substances were extracted by means of concentrated alcohol which separated spontaneously from the solution, one of which was obtained in crystalline form. These substances gave either a yellow or a copper-red reac- tion with iodine, and also the characteristic color reaction for tannin with ferric chloride. The alcoholic mother-liquor was colored greenish-yellow, and it also gave the tannin color reaction with ferric chloride. The mass wliich was insoluble in concentrated lye and alcohol, and which may be compared with Nageli's granulose, is dissolved readily in water, and it is the substance in which a blue color with iodine predominated. Nor was this part of the starch found to be of uniform composition, as was proved by the addition of a surplus of iodine. The solution which at first was blue turned black, and after successive shakings with chloroform and ether this black solution became violet and then blue, while bodies staining a dark-red or yellow went into solutions in the chloroform and ether. Now that the part which tannin plays in the iodine reaction, as well as in other reactions of starch, was recognized, it was natural, St. Jentys writes, to suppose that the lamellated structure of the grain is due to the presence of tannin. He states that when changes in concen- tration occur, the formation of concentric layers must be caused by crystallized bodies like tannins. In fact, from solutions of granulose, to which tannin had been added, and allowed to evaporate, there separated lamellated structures resembling the most beautiful starch-grains, such as are found in only a few plants like Dioscorea or Canna. Adding gallic acid instead of tannin to a granulose solution caused a radial structure to appear with con- centric lamellse, and the grains which separated from the solution resembled the starch-grains of wheat, buckwheat, or Chinese sokyes, which is only the more remarkable inasmuch as the grains with a more radially striated structure gave a more violet color, sometimes a red color, with iodine, also a color smiilar to that of gaUic acid. The lamellated grains, as St. Jentys found, could now be studied not only in granulose, but also with methyl-blue solu- tions, these solutions being allowed to evaporate after the addition of tannic acid. St. Jentys states that in view of these phenomena it is clear that the starch-grain in the plant is formed neither by apposition nor by intussusception, but by solidification of concentrated solutions; and, furthermore, since tannins also possess the property of being colored with iodine, not every part of the grain which shows a lamellated structure and which is colored by iotline need necessarily be considered a carbohydrate. Since in the 60 DIFFERENTIATION AND SPECIFICITY OF STARCHES. process of the saccharification of starches with mineral acids or with diastase the iodine reaction shows all the changes of color which may be observed in the intact grain, as well as of starch-paste under certain conditions, it may be assumed that tliis process (contrary to the universally accepted view of the progressive formation of the dextrins) depends upon the splitting up of the compounds of sugar and various tamiins in the form of glu- cosides, and to the successive decomposition of separated or free tannins. St. Jentys believes that numerous observations seem to support this theory. For instance, starch- paste which had been colored blue with iodine on being treated with powdered leather, first turned violet and then became colorless, and it had the property of reducing Fehling's solution. On distilling acidified starch-paste with sulphuric acid in order to test for sugar, an aromatic volatile body with a peculiar, unpleasant odor went into the distillate, which combined with iodine without a color reaction. OCCURRENCE OF THE STARCH-GRAIN IN PLANT LIFE. C. Nageli (loc. cit.) gave careful study to this subject, and this section is a free trans- lation from his memoir. He states that starch has a very general distribution throughout the vegetable kingdom, but that it is absent from the Fungi, Diatomacece, ChroococcacecE, Nostocaceoe, and many other cellular plants, and also apparently from some vascular Pteri- dophyta. Little or no starch is found in colorless parts of plants of one year's growth from which no new structures arise, but on the other hand more or less starch will always be found in tissue containing chlorophyl. The vegetative parts which develop organs often store up considerable amounts of starch in their colorless portions when they are not too near the surface and are of the right age, for example, the miderground parts of perennial herbaceous plants, all of which contain a great deal of nutritive material, and from which starch is seldom wholly absent. Furthermore, the stems and branches, also portions of the roots of trees and shrubs, contain starch, which is present in the pith, in the medullary rays, and in the wood-cells up to a certain age of these organs, generally in small quan- tities, or occasionally in considerable quantities, in the region near the leaves. Finally, it is observed in the pith of this year's sprouts, in the receptacle, and even in the placenta. In seeds the presence or the absence of starch-grains is more sharply defined than in other portions of the plant which contain nutrient material. Generally all of the genera of one order correspond in this respect, this holding good for eleven-twelfths of the natm-al families; and the genera of the same order are seldom dissimilar in this particular, and still more rarely the species of a genus. No starch-grains occur in the seeds of about four-fifths of the natiu-al families and about nine-tenths of all the genera of Phanerogams. Starch is present in the seeds of about half of the families and genera of Gymnospermce and Mono- colyledones, while it is absent from the other half. Starch is found in the seeds of about one-sLxth of the Dicotyledones, and in only one-fourteenth of the famiUes and in a still smaller fraction of the genera of Gamopetalce. When starch-grains are present in seeds, other reserve foods are almost entirely excluded; if seeds which are rich in starch have perisperm, then the embryo with few exceptions contains oil and no starch-grains ; but if perisperm is absent from the seeds, the starch is found in the cotyledons, while only oil is usually present in the cells of the canicle and plumules. Whenever the various genera of an order differ in this respect, it will occur usually among those which have seeds of different sizes, starch being found in the large seeds and lacking in the small seeds. The resting spores of Cryptogams usually have no starch-grains, but if two kinds of resting spores are present they usually agree in that starch is in both either present or entirely absent. Starch is absent in the majority of the pollen grains of Phanerogams. If we consider the development of the entire plant, the formation of starch always occurs in the tissues at certain stages of their growth and disappears from them at a definite OCCURRENCE OF THE STARCH-GRAIN IN PLANT LIFE. 61 age. Tliis aji;recs in general with similar transitions found in vegetable life. Before the decay of the lateral parts the starch is again dissolved and conducted to the main organs, for example, from the periphery and the growing vegetative ends. Sometimes starch formation is found in all of the plant parts in consecutive order (organs, tissues, cells), while now and then it is omitted from one of them. A noteworthy suspensioia in starch formation occurs during the transition from the active to the latent period of ^•egetation, and this may be due to the disappearance of the starch and its re- placement by another reserve material, such as fatty oil, or it may be that starch formation has ceased before the end of the period of rest. Thus, in the development of Phanerogams the formation of starch may extend to the ovary, the placenta, the testa, the outer and inner coats of the ovule, and to the embryo; and the starch may remain stored up where it is formed in the embryo, in the endosperm, etc., or it may disappear from these parts. The underground parts of plants usually contain both simple and compound starch- grains, and usually in equal proportions, although it may appear that there are more of one kind than of the other. The compound grains consist of a larger or smaller number of components, depending on whether there are the same or a greater number of these grains present; if they are in the majority they are composed of from 10 to 12 components; when they are almost exclusively present 20 components; and in exceptional cases as many as 200 to 500 components may enter into the composition of the grain. The components of the compound grains are sometimes of the same size, but more frequently differ in size. The simple grains probably never belong to the centric-lamellated type, but an eccentric structure can in most cases readily be distinguished. In closely allied species, or in the same species, the simple grains are larger in proportion as they outnumber the compound grains, while they become smaller and show a less distinctly developed structure as they decrease in number. Although the parts of the plant above the ground (bark, pith, wood, leaves, recep- tacle, and pericarp) are essentially similar in structure to the underground parts, they contain starch-grains which are smaller and less well developed. Starch formation is less general in the vegetative organs, among which Nageli includes all organs except the seeds, and often shows considerable variations, even in different parts of the same plant. In this respect nearly related plants may differ so completely that they can easily be distinguished by the place of formation of the starch-grains, while usu- ally no characteristic difference is found in this particular in an entire order of plants. All forms of starch-grains may occur in seeds, but generally the simple grains with cen- tric lamellae and the compound grains of many equal components considerably exceed all others in number. The compound grains with many components (usually many thousand) are more hkely to be found in the perisperm, and the simple grains with centric lamellse are less numerous, wliile those of an eccentric structure very rarely occur. On the other hand, in the cotyledons the simple grains are more numerous, sometimes with centric, sometimes with eccentric structiu-e, the former usually predominating, while the compound grains usually consist of 2 to 8 or in exceptional cases of 24 to 40 components. If starch is found in both the perisperm and the cotyledons the grains show a similar structure. The starch in seeds, especially in the spores and endosperm, show much likeness in form, variations occurring only within narrow limits. Very often only simple or only compound grains will be found in these parts, although it is not unusual to find both kinds in the coty- ledons. All genera of one natural order frequently contain closely related starch-grains, but sometimes a distinction is shown between genera, and usually then between species, so that these may be classified into natural groups according to the structure of their starch. The starch-grain is found almost exclusively in the plant cells. It is absent from the vascular bundles and also from the latex, except in Euphorhiacece. It is found in the ceUs in large number, even to 100, and then of small size; or a few or even one single grain of 62 DIFFERENTIATION AND SPECIFICITY OF STARCHES. considerable size. In the cells they sometimes constitute the only non-nitrogenous nutrient reser\'e substance; sometimes other equally important compounds, such as fat, cellulose, pectin bodies, etc., are present. In most cases the grains lie immediately within the cell- cavity, and when present in large numbers they fill up the whole sj^ace and are sometimes so crowded that they become flattened by mutual pressure into polyhedrous forms. They may be in direct contact with each other, or a thin layer of protoplasm may intervene, so that each grain appears to be lodged in its own compartment. When they are present in smaller masses they frequently form a lining to the wall of the cell, and may if they lie in a single layer become polygonal, owing to their crowded position; if still less numerous they sometimes cover merely the cell nucleus or circumscribed places of the cell. The starch-grains during their entire existence, or at least in earUer stages, are inclosed within plastids. In the nuclei and in the mucilage cells there may also occur oil-drops, mucus granules, and chlorophyl granules. Starch usually occurs alone in the chlorophyl bodies, but occasionally oil-drops are present. Owing to the crowded position of the grains they are sometimes flattened by mutual pressure and may remain coalesced as a pseudo-compound grain if set free by the disappearance of the surrounding protoplasm. In the clilorophyl of the Desmidacece and the Zygnemacecc, as well as in several other lower forms of Algce, the starch appears first in the form of homogeneous rings (globular shells) inclosed in protoplasm, and which later, through radial division, are converted into a compound, spherical, hollow grain. PECULIAR KINDS OF STARCH, AND STARCH-LIKE BODIES. A substance termed soluble starch was found by Dufoiu- (Bull, de la Soc. vaud. de Sci. nat., XXI, Nr. 93; Zimmermann, Botanical Microtechnique, 1893, 229) in solution in the cell-sap of the epidermal cells of a few plants, notably Saponaria officinalis. The chemical composition was not determined, but it agrees with the soluble starch made in vitro from starch-grains in forming a solution in water, and in yielding a blue, violet, or red reaction with iodine. Ewart (Pfeffer's Physiology of Plants, trans, by Ewart, 1900-6, i, 473), in referring to the soluble starch in Saponaria, etc., states that — Similarly, the cell-sap in the epidermal cells of Arum italicuni turns violet when treated with iodine, the color disappearing on heating and returning on cooling. The substance giving this reaction escapes from the cells as soon as they are killed, and the watery extract yields on evapora- tion a transparent, slightly gummy residue, which turns violet or blue with a watery solution of iodine, but reddish-b^o^vn when alcoholic iodine is added, turning blue in the presence of water. After prolonged boiling a more reddish reaction is given, and also partial digestion with diastase or ptyalin, while ultimately the color reaction disappears, a reducing sugar being formed. This "sol- uble starch" has a very much feebler osmotic value than cane sugar or dextrose, and its molecule is presumably large and complex. Its peculiar distribution points rather to its possessing some biological function (hindrance to transpiration, protection, etc.) than of its having any special value in nutritive metabolism. It may occur in small quantity in the cell-sap of the guard-cells of the stomata, though it seems always to be more abundant in the surrounding epidermal cells, and it may be still present in almost undiminished abundance after a prolonged sojourn in darkness (ten days), although no starch is then present in the mesophyll. The soluble starch soon escapes from the epidermal cells when placed in 50 per cent alcohol, and the same also occiu-s in absolute alcohol, though more slowly. Incidentally in this connection might properly be mentioned the discovery of Weder- hake (Centralbl. f. allg. Path. u. path. Anat., 1905, xvi, 517) of the occurrence of what he terms genuine starch-grains in the human secretions and excretions. The so-called starch was found in the fresh spermatic juice and in the testis, and also in a gonorrheal discharge, sputum, tuberculous sputum, pus, and both normal and abnormal urine. The grains yielded a deep-blue reaction with the tincture of iodine, and lost their color upon heating and recovered PECULIAR KINDS OF STARCH, AND STARCH-LIKE BODIES. 63 it on cooling. But they did not give the starch reactions with iodine-sulphuric acid, methyl violet, and gentian \i()lct. (This might be a form of glycogen, since Claude Bernard found a glycogen in paralyzed muscles that gave a blue reaction with iodine, and because glycogen is an important constituent of the human organism, while starch heretofore has been found onl}' as a transient food-stuff that is confined to the alimentary tract.) A starch-like substance known as "Floridean or Rhodophycean starch" has been observed in a number of Floridece. This body, which seemingly is not identical in different plants, has been examined by Van Tieghem (Compt. rend., 1865, lxi, 804), Belzung (Ann. d. Sci. nat. Bot., Ser. vii, v, 179; quoted by Zimmermann, he. cit.), Hansen (Mitth. a. d. Zool. Station, z. Neapel, 1893, xi, 276, 283), Golenkin (Alogologische Notizen, 1894, 4) and Burns (Flora, 1894, Erg. bd., 159). Van Tieghem found that the grains of this body agree in most of their chemical properties with ordinary starch, and that in the polarizing microscope they showed a similar cross or interference figure. With iodine, however, a jellow-brown or brownish-red reaction was obtained. Belzung states that the starch of many Floridece, especially the young grains, yields a blue reaction. The " Phseophj'cean starch" described by Schmitz (Die Chromatophoren der Algen, Bonn, 1882, 154; Jahrbiicher f. wissensch. Botanik, xv, 1) is in the form of colorless bodies that are found in the cytoplasm. They do not yield a color reaction with iodine, and it is held by Berthold (Jalirbiicher f. wissensch. Botanik, xiii, 569) that such bodies do not exist. Paramylon or paramj'lum grains have been examined by Klebs (Untersuch. a. d. bot. Institut z. Tiibingen, 1883, i, 233), Schmitz (Jahrbuch. f. wissensch. Botanik, 1884, xv, 111), Schimper {ihidum, 1885, xvi, 199), and Zopf (Schenk's Handbuch, Bd. iii, Hefte 2, 1). These grains have been observed in several Euglence and other low organisms, but there is doubt as to their actual character. They have been seen as disk-shaped, rod-shaped, and ring- shaped forms, and in some instances they have been found to be lamellated after being sub- jected to the action of certain swelling media, but the lamellation differed from that of ordi- nary starch-grains, inasmuch as it was in the form of complete concentric rings or plates without a conmion center or liilum. They do not yield a color reaction with iodine, and they differ in their behavior towards certain swelling and solvent reagents, and certain stains. Glycogen, which occurs in both plant and animal life, chiefly in the latter, is so closely related to starch as to be called "animal starch." In the saccharification of glycogen, dextrins appear as intermediary bodies, as in the saccharification of starch (see Tebb, page 153). With iodine it yields an orange to a reddish-brown or wine-color reaction, according to the strength of solution and form of the glycogen, the color disappearing on heating and reappearing on cooling, as with starch. It is soluble in water, forming an opalescent solution, and it exists in the li\ing tissues usually as colorless, refractive bodies. Errera (Bot. Zeit., 1886, 316; Zeit. f. w. Mikrosk, in, 277) found that glycogen is widely distributed in fungi, in which it seemingly replaces starch, inasmuch as chromatophores are absent from these organisms and therefore probably no starch formed, and since it seems to serve the purpose of a reserve food. It is also formed, sometimes very abun- dantly, in yeast cells, as shown by Errera (Recueil de ITnstitut botanique, Bruxelles, 1906; Compt. rend., 1885, ci, 253), Laurent (Jahresbr. ii. Gahrungsorg., 1900, i, 54), Meiss- ner (Centralbl. f. Bakt., 1900, ii, 6), Cremer (Ber. d. d. chem. Gesellsch., 1899, xxxii, 2062) and Pavy and Bywaters (Jour. Physiology, 1907, xxxvi, 149). In large fungi, as in Phallus, Clautriau (Jahresb. ii. Gahrungsorg., 1895, vi, 51) has shown that glycogen disappears rapidly during growth, it being altered in the same manner as starch under similar conditions in phanerogams. Other substances that bear a more or less close relationship to starch are cellulin bodies discovered by Pringsheim (Ber. d. d. botan. Gesellsch., 1883, 288) in the hyphse of Saprolegniacece. These bodies occur in the form of spherical, circular, or polyhedral granules which are occasionally lamellated. They do not yield a color reaction with iodine. Noll 64 DIFFERENTIATION AND SPECIFICITY OF STARCHES. found cellulin bodies in Siphonecp, and Denniston (Trans. Wisconsin Acad. Sciences, Arts and Letters, 1904, xv, 664) studied them in Saprolegnia. Denniston found that they, like the outer layer of starch-grains, show a strong affinity for orange-G aniline, in contra- distinction to the inner part of the grain, or the granulosa, which shows a correspondingly strong affinity for methyl violet. He regards these bodies as being made up of a substance that is intermediate in character in the synthesis and analysis of starch. (See Denniston, pages 56 and 57.) Related to these bodies are spherical granules found by Weber van Bosse (Ann. Jard. bot. de. Buitenzorg., 1892, viii, 165) in Phylophysa treubii. These granules do not yield a reaction with the iodine-potassium iodide solution, but they become blue in the presence of iodine-zinc chloride, and violet on the addition of iodine-sulphuric acid. Bodies in the form of U or V shaped cups, hollow cones, and hollow cylinders, known as fibrosin bodies, were discovered by Zopf (Ber. d. d. botan. Gesellsch., 1887, 275) in several Erysiphece. They do not yield a color reaction with either the iodine-potassium iodide or iodine-zinc cUoride solutions. Some forms of cellulose are in the nature of a food reserve, and bear a close relationship to starch in this respect, and also chemically. Like starch, they yield sugars upon decompo- sition in the presence of dilute acid or appropriate enzymes, the cellulose-reducing enzymes being designated cytases. (See Brown and Morris (Jour. Chem. Soc. Trans., 1890, lvii, 458), Schulze (Zeit. f. physiol. Chemie, 1890, xiv, 227, and 1892, xvi, 387), Newcombe (Annals of Botany, 1899, xiii, 49), and Heruaaey (Revue d. bot., 1903, xv, 345).) Phajtophycean starch, paramylum bodies, cellulin bodies, and fibrosin bodies are doubtless closer relatives of cellu- lose than of starch. Some forms of plant mucus and cellulose give a blue reaction with iodine. THE CHIEF FORMS AND CLASSIFICATIONS OF STARCH-GRAINS. It must have been recognized by Leeuwenhoek, and by many of the investigators of the earliest part of the last century, that starches from different sources are not mor- phologically identical, but if so it does not seem to have attracted any particular atten- tion until the investigations of Fritzsche (Ann. d. Physik. u. Chemie, 1834, xxxii, 129), although Payen and Meyen and others examined a number of different starches. Fritzsche desciibed the starches obtained from a variety of plants, including Solanum tuberosum, Cos- tus speciosus, Tulipa gesneriana, Fritillaria meleagris, Lilium bulbiferum, Amaryllis formosis- sima, Bromelia sp., Hyacinthus orientalis, Iris florenlina, Ixia crocata, Narcissus poeticus, Crocus vernus, Colchicum autuvmale, Bulbocodium vernum, Gladiolus communis, Arum dracunculus, Pisum sativum, Canna edulis, Hedychium flavescens, and H. hirsutum. He noted not only that the starches from different sources were different, but also that often the form was so characteristic as to determine the plant, or, at least, indicate the genus and family from which the specimen was obtained. This statement was confirmed some years later by Schleiden (Principles of Botany, 1849, 14), who examined a number of starches, mostly not described by Fritzsche. From the differences observed he was enabled to tabulate the various forms, and he published a classification that has continued to be quoted in various standard works, even at the present time. Schleiden 's Classification of Starch-Grains. I. Amorphous Starch. Amorphous starch was found in only two plianerogamous plants, it occur- ring paste-like in the cells, as in the seeds of Cardamomum minus and in the bark of Smilax ornata (Jamaica sarsaparilla) . In the ease of the latter it is not improbable that the method of drying by the fire, common in the prep- aration of sarsaparilla, may change the character of the starch. The paste is most frequently found in abnormally red roots, and less frequently in the yellow roots, neither of which have hitherto been esteemed in commerce as varieties of the Jamaica sarsaparilla. CHIEF FORMS AND CLASSIFICATIONS OF STARCH-GRAINS. 65 II. Simple Grains. Tlu; majority of i)I;uits exhibit perfectly simple iiuliviiliia,! grains, among which doublets and triplets only occur as exceptions. The following groups may be distinguished: 1. Roundish Bodies. A. With the ciidnd cuvilij or hiliun ajjjmrcntly absent. 1. Quito small, almost spherical gramdes, occurring almo.st evcrj-where from time to time in the vegctaljle kingdom as cellular contents, as for instance, in carrots, in the cambium in the winter; in leaves as the bearers of chlorophyl, etc. 2. Large, irregular, knoljby, often truncated midtiangular grains, as, for instance, in the bulbous buds of Saxifraga granulata and in the pseudo-tubers of Ficaria verna. B. With small roundish central cavities or hila. (a) With a perceptible laminated formation. 3. Very large, rough grains, deformed as it were. Found in the pith of the Cijca- dacea. There are somewhat similar grains in the imdcrgrouuil leaves of Lathrwa squamaria, in which the iimer layers form an ovoid gi-ain almost similar to those of potato starch; the few grains formed in external layers, on the contrary, are so irregular, and generally so disproportionally thick- ened at one or two sides, that the whole grain assumes a broadish triangular figure. 4. Ovoid granules. In the potato. 5. Mussel-like granules. In the bulbs of the larger Liliacece, as in Fritillaria and Lilium. 6. Almost triangular. In Tulipa. (Jj) With an indislinct or deficient lamellatcd formation. 7. Rounded-off polyhedric grains. In the albumen (perisperm) of Zea mays. 8. Sharp-edged, polyhedric, very small grains. In the albumen of Oryza sativa. (c) W^ith an elongated central camty. 9. Roundish or oval grains, in a dry condition, generally showing a star-like cleft in the iimer layers. In the Leguminosce, as in the seeds of Pisum and Pha- seolus. (d) Perfectly hollow, apparently cup-like grains. 10. Very marked in the rhizome of Iris Jlorentina and in kincbed species. 2. Flatly compressed lenticular granules. 11. Sometimes with, sometimes without, a decided lamellated formation; some- times with a central, or eccentric, or less rounded, or more elongated, or radiated torn-up cavity or bilum. In the albumen of Triticum, Hordeum, and Secale. 3. Perfectly flat discs. 12. With more distinct lamella, in which it is, however, at times doubtful whether they pass entirely around or are only menisci laid over one another. The former appeared probable owing to analogy and the phenomena presented in roasting and on dissolving in sulphuric acid. We do not find it in the rhizomes of all the Scitaminew, as Meyer asserts, but exclusively in the Zin- giberaceoe Lindl; and neither in the Cannacece, nor in the Marantaceoe. 4. Elongated grains. 13. With an elongated central cavity in the milk-juice of the indigenous and a few of the tropical Euphorhiaceos. 5. Very irregular grains. 14. In the milky juice of many tropical Euphorbiaccoe. 5 66 DIFFERENTIATION AND SPECIFICITY OF STARCHES. III. Compound Granules. Here we find a simple grain as an exception in the plant or part of the plant. / . The separate grains in the compound grains without evident central cavity or hilum. 15. Compounded according to the simplest types in 2, 3, or 4, as in the rhizomes of Marantacece (West Indian arrowroot). Also in tubers of Aponogdon, in the thickened vagination of the leaves of Marattia, and in roots of Bryonia. 16. Generally arranged regularly, seldom irregularly, and composed of from 2 to 6. In the bark of the roots of the various sorts of sarsaparilla. 2. The separate grains in the compound grain having a distinct central cavity or hilum. (a) All the parts of the grains of nearly the same size. 17. United according to simple types from 2 to 4. The central cavity or hilum small and roundish, as in the tubers of Jatropha manihot. 18. Combined according to simple types from 2 to 4. The central cavity or hilum large and very beautiful, opened in a star-like form, as in the corms of Colchicum autumnale. 19. Combined according to simple types from 2 to 4. The separate grains quite hollow, apparently cup-shaped. A marked form occurs in Radix ivarancusoe {Anatherum ivarancusw). 20. Firmly combined, from 2 to 12 in number, in very irregular groups, as in the rhizomes of Arimi 7naculattim. 21. A large number, often as many as 30, of small roundish grains, very loosely grouped. Frequent, for instance in the stem of the Bernhardia dichotoma. (6) Many smaller grains grown together upon one larger one. 22. In the pith of Sagua rumphii, etc., and generally in sago. Nageli's Classification of Starches from Different Sources. A. Grains Simple. I. Centric. Hilum in the mathematical center; lamellae always equal at two corresponding diametrically opposite points. Type 1. Spherical. When the grain is free both hilum and grain are spherical. Type 2. Lenticular. When the grain is free both hilum and grain are rountled; grains compressed; sometimes circular or ovoid; sometimes triangular or quadrangular. Type 3. Oval. When the grain is free both hilum and grain are oval to lanceolate-oval; occasionally kidney-shaped or somewhat curved; when on end they appear circular or somewhat compressed. Type 4- Spindle-shaped. Grain linear or lanceolate, tapering towards the pointed ends, or of equal width with blunt ends; when on end they appear almost circular. Type 5. Bone-shaped. Grain elongated and compressed from the narrow aspect, but linear spindle-shaped from the broad aspect, with enlarged laminated ends. II. Eccentric. Hilum usually more or less removed from the mathematical center of the grain; lamellse coarsest and finest at opposite ends of the grain, respectively. Type 6. Inverted cone-shaped. Grain on end almost circular; slender at the hilum end. Type 7. Cone-shaped. Grains on end almost circular; decidedly thicker and broader at the hilum end. Type 8. Wedge-shaped or compressed. Grains flattcnetl, of equal thickness throughout, or thicker but narrower at the hilum end than at the distal end. Type 9. Rod-shaped. III. Grains simple and structure obscure. Type 10. Structure not fully developed, or not identified owing to diminutive size of the grains. Lamellse, hila, cavities, fissures, and clefts seldom observed. B. Grains semi-compound. Type 11. Grains semi-compound. The component part-grains are enveloped partly or wholly by a common substance. CHIEF FORMS AND CLASSIFICATIONS OF STARCH-GRAINS. 67 C. Grains Compound. The component part-grains not enveloped by a common substance. I. Composed of fused part-grains. Type 12. Composed of fused part-grains. The part-grains are not separated by fissures, and even different grains may be fused with one another. II. Composed of separated part-grains. The part-grains separated by fissm-es. Type 13. Grains in 1 or 2 roujs. From 3 to 11 components arranged in 1 or 2 rows. Type 14- Equally divided grains of few components. From 2 to 10 or more almost equal-size part-grains which when separated have 1 curved surface and 1 or more pressure facets. Type 15. Unequally divided grains of few components. From 2 to 10 or more unequal sized firmly united part-grains, which when separated have 1 curved surface and sev- eral flattened pressure facets. Txjpe 16. Multiple grains. From 20 to many thousand firmly united part-grains which when separated are covered with pressm-e facets. Type 17. Hollow spherical grains. The part-grains are arranged in a spherical layer, as if a globular shell had been divided radially. For further details of the characteristics of the various types see Chapter V, page 197. Meyer's Classification of Starch-Grains. Mej-er, in his memoir Die Sttirkekorner {loc. cit.), criticizes Nageli's conceptions of the "pseudo-compound" and "true-compound" starch-grains (Chapter V), and holds that we must dismiss the idea of a "true-compound" grain in the sense held by Nageli because there are no starch-grains that have been formed by the separation of an originally single grain. The expression "pseudo-compound starch-grains" may also be rejected because it is now un- suitable, since if there are no tnie-compound grains there can be no false ones; furthermore, the grains under consideration are not individual starch-grains, but only simple starch-grains held together by a chromatophore substance. The "semi-compound" grains which according to NiigeU arise tlirough the division of the nuclei and the formation of systems of lamellae with- out any lines of separation between the cells is also incorrect because such starch-grains arise from several simple starch-grains being inclosed within a common starch-layer. Instead of the old name of compound grains, to avoid confusion, it is better to substitute the expression "complex" starch-grains; but Nageli's conception and term of "simple grain" may be retained. Meyer proposes some substitutes and some new terms in place of Nageli's, as follows : (a) Simple or monarch starch-grains. Grains which have but one hilum. lb) Complex starch-grains. Grains that are formed from several starch-grains which are so crowded in one chromatophore that they become enveloped within common starchy layers, and thus boimd into a single individual. Complex starch-grains may be diarch to polyarch. The approximately monotone starch-grains possess simple symmetrical forms; frequently they are spherical bodies, the shape depending upon the influence which the fluid chromatophore exerted upon the form and lamellae of growing starch-grains. Grains of a relative monotone type are found chiefly in the nutritive tissues of seeds, in which during the active growth of their cells and chromatophores an energetic and periodic solution of the starch-grains takes place, by means of which the central mass of the grain becomes relatively less dense, which in turn is followed by a uniform growth of the starch which ceases when the seed matures. Similar processes take place in many cotyledons which contain reserve material; also in numerous typical, colorless stor- age roots and bulbs which are filled with starch during their growth, and contain, if they vegetate normally, relative monotone starch-grains. Such organs during their development, and especially at the period when the reserve material is admitted, depend chiefly upon their neighboring structures for nourishment, and only in extreme cases of need can they draw upon the stored starch-grains. Examples of monotone starch-grains are found in the fleshy scale leaves of Adoxa, in the rhizome of Iris gerynanica, and in the mature tubers of Solanwn tvberosuvi. In the latter monotone starch-grains may develop within the parenchyma cells in 14 days. 68 DIFFERENTIATION AND SPECIFICITY OF STARCHES. Polytone starch-grains, which arc distinguished especially liy the periodic inequality of their lamellae, are found most abundantly and constantly in storage organs which live for several years, in consequence of which the carbohydrate passes through several periods of storage and disaj)pearance. They may be found in the rhizome tips of Adoxa in the spring of their second calendar year, and in 3-year-old bulbs of Hya- cinthus. These grains are also most readily formed in perennial plants, as observations in PeUionia have shown. (c) Solitary starch-grains. Grains which grow singly in one chromatophore. (d) Adelphous starch-grains. Those which grow along with other grains in one chromato- phore. They may be di- to poly-adelphous; up to 6-adeli)hous would be designated as oligo-adelphous. (e) Monotone starch-grains. Grains which during their entire life have imdcrgone periods of solution, leaving coherent traces of each of the lamellae formed during the period of storage. Perfect monotone starch-grains in which there still exists a trace of every lamella which has been formed and in which, even if eccentric, the lamellae are closed are exceedingly rare. This term is also to be used in all cases in which no distinct characters of a polytone type are present. (f) Polytone starch-grains. Grains that have during their development undergone two or more periods of solution in which numerous lamellae completely disappear or de- crease in •nidth, interspersed with other periods during which they developed as the relative monotone type. Polytone grains if they have an eccentric structure show a series of lamellae which are open laterally. The similarity of the monotone starch-grains at one period in the same plant part and of the various individuals of a definite plant part is a recognized phenomenon. This is the consequence of the specific nature of the chromatophore and cytoplasm and of the approximately similar biology of the latter. Monotone starch-grains which grow at different periods in one cell of a plant may, however, assume very divergent forms, since in the course of the life of the cell changes occur in the size, consistency, and chemistry of the chromatophores and cytoplasm. Thus, in the chromatophores of one cell of the scale leaves of Adoxa moschatellina of the first year's growth oialy monarch eccentric, conical starch-grains are found, while in the same cells of the second year usually only polyarch, almost centric, starch-grains are observed. The small variations in the form of the starch-grains found in the same plant part are not due alone to the fact that they were not formed at the same time, but also because every cell has its own biology, even every chromatophore has its individual properties. In different organs of the same plant grains may assume a great variety of form. Thus, the leucoplast of the tubers of Solanwm tuberosum form, besides complex and oligo-adelphous grains, mostly solitary, monarch, eccentric, conical, or oval grains, with a length of 200^/, and which have definite, rather delicate, irregular lamella. In the chloroplasts of the potato, large, solitary, inverted conical-shaped starch-grains occur which have at regular intervals refractive lamellae of equal width. The polytone starch-grains of one plant part, even in one plant cell, are in general less similar than the monotone grains, since in them the distinctions which originally existed are intensified through the variations caused by the marked solution alterations of the monotone grain. A. Monarch Starch-Grains. (a) Solitary starch-grains. I. Centric starch-grains. Hilum in the mathematical center; lamellae always equal at two diametrically opposite points (Nageli, Tyjje 1). Such centric grains, as already noted by Nageli (see Types 1, 2, and 3), occur only in seeds. It is therefore evi- dent that, just as in the seeds the cytoplasm is abundant and dilute, the chro- matophores lack density and have a tendency to expansion. CHIEF FORMS AND CLASSIFICATIONS OF STARCH-GRAINS. G9 A. Monarch Starch-Grains. — Continued. (1) Monarch solitary, spherical. (Niigeli, Type 1.) Hilum spherical in tlie center of the spherical grain; lamclliE of uniform thickness, forming complete circles. Examples of such grains are found in Sorghum vulgare and in Zca mays. In the latter, when the grains have reached two-thirds of tlieir growth, they are spherical with centric lamelhe. Grains which have just l)cen removed from immature endosperm cells show the distinct striations of the lamella; later, by pressure due to the growth of neighboring grains, the solitary grains become angular ; the lamellte of these angular grains are closed, but are finer at the flattened surfaces. (2) Monarch solitary, centric, lens-shaped. Hilum and grains rounded and compressed (also rounded-reniform or rounded-oval, compressed). (Niigeli, Type 10.) Examples: Starch-gTains from the seeds of Triticum, Sccale cereale, Hordeum vulgare. In the chromatophores of the endosperm of Hordeum small grains form early, and by the rapid growth of several grains in a relatively small clu'omatophorc become nuitually flattened. During the development of the endosperm cells the growth of the grain within the chromatoijhore is hindered, partly due to new formations, partly through the solution of small grains so that starch-grains of various shapes arise, such as laterally flattened, crescent-shaped, bean-shaped, etc. (3) Monarch solitary, centric, oval grains. Hilmn and grains oval and lanceolate-oval; circular in cross-section; lamellce equal at two diametrically opposite points, being coarsest at two poles. (Nageli, TjqDe 3.) Examples: Many grains from the cotyledons of Vicia faba and other Papilionacew. The starch-grains in the cotyledons of Cicer arietinum are instances of this form of grain. All the chro- matophores lie in the primordial utricle and usually contain 1 starch-grain, or rarely 2, which are flattened upon the sides in contact. Most of the starch-grains correspond in form to that of the chromatophores, which are disk-shaped, although of different densities. From time to time energetic solution takes place as a consequence of rapid growth, so that most of the grains are irregularly outlined and tuberculatetl in the early stages; later, the irregularities become less pro- nounced because there is less active solution. The cliromatophore surrounds the starch-grain almost uniformly as a green layer, the surface of the grain be- comes more and more smooth and increases relatively more in thickness than in width and breadth, since l>y the deposition of lamellte of uniform thickness the grain grows more rapidly in the thick diameter than in the longitudinal. In the mature grain a feebly refractive diffuse hilum is observed, the lamellis are strongly refractive, and the grains have become rounded and almost oval. (4) Monarch solitary, centric, rod-shaped, and cone-shaped grains. (Nageli, Types 4 and 5.) Examples: Starch-grains from the lactiferous vessels of the Euphorbiacea. Ob- servations upon Euphorbia vnjrsinites show that the young starch-grains from the latex are rod-shaped; most of them are irregularly corroded and have a feebly refractive line in the axis, which probably may be due to active fermentation and swelling. Very slightly corroded grains are also rod-shaped with bilateral eccen- tric lamellis. Since the younger lamellaj are laid down directly after the growth of the latex vessels terminates, they will be the most dense and hence probably ^vithstand solution better than those lying nearer the middle of the rod-shaped grain, and thus the explanation of the origin of the thicker ends may be similar to that observed in the case of Oxalis ortgiesi. The starch-grains from older in- ternodes are bone-shaped and somewhat flattened. The investigation of chro- matophores from material which had been carefully hardened and stained with fuchsin showed the grain completely enveloped in the substance of the chromato- phore, which, however, was more markedly massed at the two ends of the grain. II. Eccentric starch-grains. (5) Monarch solitary, eccentric, rod-shaped. Lamellte on one side the heaviest, and at the diametrically opposite side the finest. Grains are circular in transverse section; both ends of almost equal width and thickness. (Nageli, Type 9.) Ex- amples: In Dieffenbachia and Iris germanica. In Dieffenbachia beautiful mono- 70 DIFFERENTIATION AND SPECIFICITY OF STARCHES. A. Monarch Starch-Grains. — Continued. tone rod-shaped starch-grains arise by starch-grains growing simultaneously with the viscous chloroplast, and becoming unilaterally extended. The thinness of the stretched sides of the chloroplast results in the formation of eccentric lamel- la. When the chromatophore is first mature and a strong polytone gTowlh of the starch-grains prevails, the cliromatophore is expanded so that the starch- grain is somewhat broader at the base. (6) Monarch solitary, eccentric, conical. Lamellae are the densest at one side and the most delicate at the diametrically opposite side. Grains are conical; circular in transverse sections. Hilinn located at the narrower, less dense end. (Nageli, Type 6.) Examples: Starch-grains in the tuber of Solanum tuberosum and rhizomes of Adoxa. Starch-grains of Type 6 are found in chromatophores, which show a tendency to growth equally in all directions. Pure forms usually only occur when the monotone grain is half-grown. (7) Monarch solitary, eccentric, inverted-conical. Grains similar to those in Type 6, but with the hilum at the more dense end. (Nageli, Type 7.) Examples: Tuber of Solanum tuberosum, rhizomes of Adoxa and Iris gei-manica, etc. By a flatten- ing and then a sharpening to a point at the base of a grain of Type 6 with closed lamellse, such type may develop into one of Type 7 with lamellae open at the base. Grains of Type 7 are usually found in company with those of Types 5 and 6. (8) Monarch solitary, eccentric, flattened. Examples: Starch-grains in rhizomes of Zingiber officinale, Curcuma zeodaria, Maranta, pseudo-tubers of Phaius grandi- florus, etc. The starch-grains of Type 8 are formed in the same way as those of Types 6 and 7, only they are flattened. (6) Adelphous starch-grains. I. OUgoadelphous starch-grains. If several starch-grains grow simultaneously in a chro- matophore, they behave just as spherocrystals do when growing in an inex- haustible mother liquor. Starch-grains from Pellionia are an example. Only in the very earliest stage are the two starch-grains spherical, and, as is apparent, the spherical shape is the more pronounced the larger the chloroplast when the first grains start to form. If the chloroplast is still small when the starch-grains begin to grow, so that the grains develop along with the chloroplast and exceed it in growth, they are very soon prevented from increasing in size on the inner side, and both become flattened. The lamellae are heaviest within and below, and in purely monotone grains they are always closed, since the crystallization substance between the grains is furnished in the greatest quantity. Flattening of the grains results, as is readily seen if one considers that when two spherical grains grow side by side in a chromatophore, the chromatophore layer being thiimest where the two spheres come in closest contact with their surfaces. The entire process of growth of the diadelphous starch-grains is similar to that of monarch, solitary grains of Pellionia. II. Polyadelphous starch-grains. The polyadelphous starch-grains of a chromatophore, which are not easily distinguished from the diadelphous forms, are approximately similar in form and size, though the proportion of the diameter of the smallest to that of the largest is usually as 1 to 4. The greatest diameter of a chromato- phore filled with starch-grains which was measured by Nageli is 106^1. Such a chromatophore may, according to Nageli, contain between 10 and 30,000 grow- ing starch-grains. The form of the polyadelphous grains is mostly polyangular or rounded with centric structure. Some exceptions are found in the flattenotl forms of Arenaria and A. graminifolia and Drymaria cordata described by Nageli. The polyadelphous grains are found relatively seldom in rhizomes and roots, although there are some exceptions, but occur most abunilantly in the reserved food of seeds. The development of starch-granules in the chromatophore of the endosperm of Oryza sativa served as a good example of the formation of polya- delphous starch-grains. In every young seed of Oryza the nucleus lies in the abundant cytoplasm in the center of the cell, and is surrounded by small, scarcely CHIEF FORMS AND CLASSIFICATIONS OF STARCH-GRAINS. 71 A. Monarch Starch-Grains. — Continued. discernible Icucoplasts. In certain of the leucoplasts several starch-granules arise almost simultaneously, the number increases although it varies until cell division is complete. The chromatophores are rounded at first, and remain so as long as there is no pressure on them. When the starch-bearing chromatophores have attained about half of their ultimate diameter they begin to flatten against one another, since they are rather crowded in the cytoplasm. The starch-gran- ules, which are angular and much crowded when first recognized, gradually liccome more and more regular, since they check each other equally in growth. Distinct liimelkB are not present in the developing grain, since at first they are angular and corroded; and only a feebly refractive hilum, which corresponds to the earliest corroded stage of the grain, can be observed in the mature grain. B. Complex (Di- to Poly-arch) Starch-Grains. It can not be readily demonstrated that complex grains, like certain grains, for example, of Pellionia, are descended from adelphous grains. It has, however, proven that all plant parts in which complex grains are present also at times develop adelphous starch-grains which correspond entirely with the central lamellae of the complex grains. Furthermore, gradations between the adelphous and the similar complex forms can be found. Additional proof of the connection between adelphous and complex starch-grains is the fact that plant parts which produce few and irregular adelphous grains likewise have relatively few and irregular complex grains in their cells. In the storage scale leaves of Adoxa it is noticed that numerous adelphous grains form at first, but only complex grains are present later, and the layers of the latter resemble those of the former. After the two diadelphous starch-grains found in Pellionia have attained a certain size in the chloroplast, the chloroplast substance is entirely forced out between the contact surfaces of the grains, or is cut off from the remaining mass of the chromatophore by the growing trichites, so that no more starch-substance is stored in the region of the contact surfaces. If the chromatophore mass between the flattened surfaces be- comes inactive or disappears, the two starch-grains again grow into diarch form. Muter's Classification of Starches.* This classification, which is based upon histological and polariscopic peculiarities, is characterized by the designation of each group by some important type of starch : Potato group. Oval or ovoid granules, showing hilum and concentric rings clearly; cross and colors with selenite usually distinct. Legume group. Round or oval granules, hilum marked, rings faint, but rendered visible in cases by chromic-acid solution; cross and colors feeble. Wheat group. Rountl and oval granules, hilum and rings generally invisible; feebly marked cross and colors. Sago group. Truncated granules, hilum distinct, rings faint; cross and colors usually faint. Rice group. Polygonal granules, hilum distinct, rings faint; cross and colors usually faint. Kraemer's Classification of STARCH-GRAiNs.f This classification includes the starches of some of the more important vegetable drugs, together with a few commercial starches, and is based upon morphological and other characters. A. Simple spherical grains. (a) Not more than 5ii in diameter: Cimicifuga, Cypripedium, Frangula, Hydrastis, Leptandra, Piper, Prunus virginiana, Quassia, Quercus alba, Rhamnus purshiana, Spigelia, Viburnum opulus, and Viburnum prunifolium. •Organic Materia Medioa; quoted by Leffraann and Beam, Food Analysis. 1906. t Botany and Pharmacognosy, Philadelphia, 1907, 698. 72 DIFFERENTIATION AND SPECIFICITY OF STARCHES. A. Simple spherical grains. — Continued. (b) Not more than lOfj. in diameter: Calamus, Eunonymus, Gelsemium, Granatum, Quillaja, Sangiiinaria, Serpentaria, Tonka, Ulmus, Xanthoxyluin. (c) Not more than 15n in diameter: Apocynum, Cinchona, Colchici semen (in caruncle only), Convallaria, Sumbul, Valeriana, (rf) Not more than 20n in diameter: Glycyrrhyza, Phytolacca, (e) Not more than SOjx in diameter: Rumex, Stillingia. B. Compound spherical or polygonal grains. (a) 2 to 3 compound: Belladona radix (5 to 15/i), Sassafras (7 to 20/z), Veratrum viride (7 to 20m). (6) 2 to 4 compound: Aconitum (4 to 12/i), Cinnamonum (7 to 15^), Colchici cormus (7 to 20^), Ipecacuanha (4 to IAjjl, those of Carthagena ipecac being uniformly larger), Kramcria (20 to 30/i), Rheum (5 to 20ja), Sarsaparilla (7 to 20/i). (c) 2 to 6 compound: Podophyllum (5 to 12/i). (d) More than 6 compound: Capsicum (3 to 7|t), Cardamonum (1 to 4^1), Cubeba (1 to 4/j), Gossypii cortex (5 to 20ai), Mezereum (10 to 15/u), Myristica (5 to ly), Pimenta (7 to 10m), Rubus (3 to 7m). C. Ellipsoidal or ovoid grains. Althisa (10 to 20m), Geranium (10 to 15m), Glycyrrhiza (5 to 10m), Pareira (7 to 15m), Physo- stigma (25 to 40m), Rumex (10 to 20m), Stillingia (15 to 30m), Strophanthus (2 to 4m), Zingiber (15 to 30m). D. Grains of characteristic shape. Calumba (25 to 35m), Iris florentina (15 to 30m), and potato and other starches, such as arrow- root, wheat, corn, yam, canna, bean, pea, cassava. E. Altered grains. Guarana (10m), Jalapa (15 to 35m; also 2 to 3 compound grains), Tragacantha (2 to 10m), turmeric in masses (70 to 140m). F. Amylodcxtrin grains. Mace contains starch-grains which give a rcddisli color with iodine. Kraemer notes tliat leaves, herbs, and flowers do not as, a rule, contain starch. Winton's Classification op Starches.* Winton records that the forms of the grains are so numerous, even in the same variety, as to forbid accurate classification, but that the following are the more striking. : 1. Globular. The starch of the peanut and some grains of maize. 2. Lenticular. The large grains of wheat, rye, and barley. 3. Ellipsoidal. The starch of legumes. 4. Ovoid or pear-shaped. The starch of potato, canna, Bermuda arrowroot, yam, and banana. 5. Truncated. Most of the grains of cassava, batata, and sago. 6. Polygonal. The starch of maize, rice, oats, and buckwheat. Winton also gives an analytical key by Moellar to commercial starches that is based upon general histological characteristics. Many other classifications might be quoted, without, however, any material advantage. PROPERTIES OF THE STARCH-GRAIN IN RELATION TO MENDELISM. Gregory (The New Phytologist, 1903, ii, 226) found that the starches of round and wrinkled peas occur in two very different types. (See Plates 8, 9, and 10, figs. 47 to 56.) In the round seeds the jjeripheral cell-layers of the cotyledons contained a few oval starch- grains which did not e.xceed 0.06 mm. in the greatest diameter. In the third layer the grains reached 0.2 mm. in length, while the more deeply situated cells were crowded with oval grains measuring as much as 0.34 mm. in the greatest dimension. The grains were regular in shape, with a definite center surrounded by well-marked lines of stratification. In the ivrinkled peas the grains of the jieripheral layers were of about the same size as those of the round peas, but were of a different type, occurring in irregular spheres with several * The Microscopy of Vegetable Foods, New York, 1906, 645. STARCII-GUAIN IN RELATION TO MENDELISM. 73 centers, thus forming a compound Glucose Achroodextrin -^ Maltose-* Glucose \ Maltose -^ Glucose Reactions with iodine. Purple to a bluish violet. Reddish-violet, red, reddish- brown, or red- dish-yellow. No color reaction. No color reaction. In this schema it will be observed that a serial action is assumed by which one product is formed after another, in contradistinction to the theory of the coincident formation of dextrin and maltose and the splitting of these molecules. The different stages are char- SOLUBLE STARCH. 101 acterized by the chief decomposition products present and by the reactions with iodine. No account has been taken of residue. The scheme is, of course, modifiable under differ- ent conditions of experiment. Thus, the final product may be maltose and not maltose and glucose, or glucose; and instead of maltose there may be isomaltose; and in some reactions saccharose and other sugars would be present, etc.; according to som.e investi- gators either or both amylodextrin and maltodextrin would be included, etc. The color reactions of the second stage will vary in accordance with the nature of the decomposing agent and other conditions. SOLUBLE STARCH, KliVUKTEU STARCH, COAGULATED STARCH, "ARTIEICIAL STARCH," AND "ARTIFICIAL STARCH-GRAINS." SOLUBLE STARCH. The term soluble starch has been used so indiscriminately that it not only does not designate any specific body, but it may also apply to a substance so close to raw starch as to give the tjiaical blue reaction with iodine, or to one that is a mixture of starch and ery- tlu'odextrin, or to a specific body that has properties of both starch and erythrodextrin, or to solutions of dextrins or dextrin and sugar, etc., which are so far removed from starch as to give no color reaction with iodine. In other words, it has been used to signify the products of decomposition as they exist at any given stage of the process of saccharifica- tion from the first modification product to a mixture of achroodextrin, or to "malto- dextrin," in which there may not be a particle of starch, soluble or insoluble, present. Obviously the term soluble starch should be restricted to the first product wliich yields a blue reaction with iodine. Boiled starches are essentially pseudo-solutions which con- tain small and variable quantities of dissolved particles of starch, together with a relatively great mass of suspended particles which represent starch in many degrees of molecular condensation. Such pseudo-solutions may be rendered into perfect soluimis in proportion to the extent of solution of the suspended particles, so that by appropriate procedures a non-colloidal solution of starch can be obtained which can be filtered through a colloidal membrane or a Chamberland filter. Pseudo-solutions, such as starch-paste or less dense preparations, can be converted into a clear limpid solution, or true soluble starch, by minute quantities of amyloclastic enzymes, by small or large quantities of amyloclastic enzymes that have been modified by subjection to a temperature of 75° to 80° for 10 to 15 minutes, by the actions of dilute acids and other chemical agents, and by heating in the autoclave at about 140°. Starch-paste, starch pseudo-solutions, and starch non-colloidal solutions undergo, upon standing, or upon repeated heating and cooling, more or less reversion to insoluble forms of starch. This change becomes apparent at fu-st as a turbidity, which grows more marked as condensation proceeds, and which ultimately may be observed in the form of granules which have the appearance and general properties of starch-grains. This reversion product has been termed amylocellulose, or coagulated starch, or artificial starch, or artificial starch-graiiis. There seem to be many forms of this reverted substance which differ in solubiUty, digestibility, reaction with iodine, etc. A similar reversion may be caused by a specific enzyme that has been named coagulase, but the reversion products are not absolutely identical with those formed spontaneously (see page HI). Probably the earUest investigator to prepare soluble starch was Guibourt (Ann. de chim. et phys., 182S, xl, 183). This he did by pounding starch-grains in a mortar and placing m water. The rupture of the grains renders them partially soluble in cold water, giving rise to a gelatinous solution that yields a blue reaction with iodine. This observa- tion has been repeated by many investigators, and while it has been held by some that the gelatinization is due to heat formed diu-ing the pulverization, this assumption has been totally disproved (page 28), and it has, moreover, been clearly shown by the records 102 DIFFERENTIATION AND SPECIFICITY OF STARCHES. of a number of observers, as already stated, that the degree of solubility is proportional to the degree of comminution of the grains, and that complete comminution is equiv- alent to complete gelatinization by heat in so far as the actions of digestive agents are concerned. Fritzsche (Ann. d. Physik. u. Chemie, 1834, xxxii, 129), in liis study of the lamellse of the starch-grain, associates the greater density of the outer layer with the insolubility of the raw grains in cold water, as he found that when the grains are crushed some of the inner starch is dissolved in cold water. The solubility of a part of the raw grain in cold water has been confirmed by Jessen, Delffs, Knop, and others of the earlier investi- gators (p. 28), and later by the investigations of Brown and Heron, Meyer, and Maquenne. Uninjured raw starch-grains are as insoluble in idtro in solutions of amyloclastic enzymes as they are in cold water, provided the actions of bacteria are prevented. Brown and Heron (Ann. d. Chemie u. Phys., 1878, cxcix, 206) found that starch-grains which showed no change after a considerable time in the presence of malt extract were readily digested after they were crushed. Meyer (Die Starkekorner, etc., loc. cit.) noted that perfect grains of the starch of the potato and of Dieffenbachia seguina are dissolved after a time from within outwards in successive layers, but grains having clefts or fissures were eroded with channels, pits, cavities, etc. Maquenne (Compt. rend., 1904, cxxxviii, 375) compared the intensities of the digestive actions of diastase and dilute acids on raw starch, crushed starch, and starch-paste. By means, for instance, of malt extract at 55° he found the following percentages of soluble matter produced: Raw starch-grains, 2.8; raw starch broken by trituration, 94.8; and starch-paste, 102.2, the percentages being calculated on the quantity of original starch. The form of soluble starch that is obtained by simple solution of the comminuted raw grains is doubtless the nearest derivative of natural starch, but a similar or identical form may be prepared by the subjection of the grain to the temperature of complete gelatinization; or by a slow solution of raw starch by amyloclastic enzymes, dilute acids, etc., avoiding the formation of various decomposition products that are formed during rapid action; or by the liquefaction of starch-paste by amyloclastic enzymes, dilute acids, high temperatures, etc. The various products thus obtained are not to be regarded as being of uniform composition, as will be evident by the context, and they may be further individualized by modifications of certain of the methods of preparation. The form of so-called soluble starch that is made by boiling starch in sufficient water to make a liquid preparation is so universally known as not to require special notice. Starch-paste can be converted into liquid soluble starch without the formation of dextrin by heating in the autoclave or by careful restriction of the actions of amyloclastic enzymes, dilute acids, alkafies, etc., or by the use of modified enzymes. Syniewski (Ann. d. Chem. u. Phar., 1899, cccix, 282; 1902, cccxxiv, 212) placed a 5 per cent starch-paste made of potato starch in the autoclave, at a temperature of 140°, and obtained a clear solution which gave a blue reaction with iodine. This form of soluble starch is soluble in cold water and is without reducing action on copper solutions. He also prepared soluble starch by treating starch-paste for several minutes with malt extract. Lintner and Diill (Ber. d. d. chem. GeseUsch., 1892, xxvi, 2533) prepared soluble starch ("amylodextrin") by sub- jecting potato starch to the action of air-dried malt at a temperature of 70°, and arresting the action while iodine still yielded a blue coloration. Hot 40 per cent alcohol was added, and the soluble starch that separated upon cooling was finally purified by repeated frac- tionation with 30 to 40 per cent alcohol. Upon drying, a light white powder was obtained which was only slightly soluble in cold water, but very soluble in hot water. It yielded a blue reaction with iodine, but it did not reduce Feliling's solution. Pottevin (Compt. rend., 1898, cxxvi, 1218) found that when malt extract is kept at 79° to 80° for 15 to 20 minutes it loses its power of converting starch into sugar, but retains its power of liquefying starch-paste and of forming dextrin. Petit (Compt. rend., 1905, SOLtTBLE STARCH. 103 cxLi, 1247) discovered that malt infusions react toward tincture of guaiacum like solu- tions of compounds of iron or manganese. If the malt infusion yields only a faint guaiacum reaction, the addition of traces of alkali, or aeration and acidifying with lactic acid, causes it to give a full reaction. The preparation after tliis treatment has lost its power of sac- charifjdng, but not its power of liquefying starch-paste. Solutions ha\ang the power of liquefying starch-paste at 20° were made by treating a solution of commercial albumin with ferrous and ferric oxides. It was also found that the starch-liquefying action could be considerably intensified by the addition of lactic acid or asparagin. The liquefaction of starch-paste under pressure may be hindered by the presence of certain substances. Thus, Fernbach and Wolfif (Compt. rend., 1906, cxliii, 380) found that while salts wluch are neutral to methyl orange are without effect, salts that are alkaline are inhibitory, and that traces of alkali may be absolutely preventive. Green (The Soluble Ferments, London, 1901, 31) records that the "diastase of secretion" corrodes starch-grains and disintegrates them before solution, and rapidly liquefies starch-paste, most advantageously at a temperature of 50° to 55°, and that it will withstand heating to 70° without destruction. Ford (Jour. Soc. Chem. Industry, 1904, xxiii, 414) prepared soluble starch by previ- ous treatment of the raw starch with dilute alkali and acid, repeatedly washing, and then drying in the air. This purified starch was gelatinized, then liquefied at 79° to 80° by means of a trace of diastase, and then the limpid solution boiled to destroy further dias- tatic action. Ford also prepared a soluble starch that was practically ash-free by repeated precipitation of an acidified solution of starch with alcohol, with the addition of a very small quantity of potassium or sodium cliloride to prevent the starch from becoming milky, which results from the entire removal of acid or salts. Soluble starch prepared in this way is described as fairly pure and nearly neutral; that is, it does not give a reac- tion with rosolic acid or methyl orange, although somewhat acid to phenolphthaleine. O'Sullivan, Brown and Heron, and many other investigators have prepared soluble starch by arresting diastatic action at the proper time. A non-colloidal or perfect solution of starch was made by Fouard (Compt. rend., 1908, cxLVi, 285) by demineralizing the starch, and partial hydration at 80°. The preparation thus obtained can be filtered through a collodion membrane, a 5 per cent solution of col- loidal starch jdelding in tliis way a 2.74 per cent non-colloidal solution. Fouard states that this non-colloidal solution shows not a trace of polarization; that it gives an intensely blue reaction with iodine, appearing as a solution rather than as a suspended precipitate; that its \'iscosity is about that of a 1 per cent solution of sugar, and, therefore, nearly that of water; that it is unstable, becoming cloudy and forming a granular deposit; and that it is probably a heterogeneous system containing dissolved molecules together with starch- molecules in all degrees of condensation. In a subsequent study, Fouard (Compt. rend., 1908, C'XL^^, 978) found by ultra-microscopic examination that this non-colloidal solution stands intermediate between mineral colloids of insoluble elements and a perfectly disso- ciated salt-solution. Freezing causes a slight opalescence preceding the formation of granules. When a 5 per cent pseudo-solution is filtered tlu-ough several collodion mem- branes which differ in their density, by varying the proportions of alcohol the filtrate contains increasing proportions of colloidal starch, ranging from 1.518 to 2.365 per cent, and having an increasing rotatory power (0)0 = from 183° 15' to 191° 50', showing that the membrane acts as an analyzer. The state of perfect solution is changed by dilu- tion or evaporation in vacuo. The water therefore not only acts as a solvent, but also is a necessary factor in modifying the state of the starch. During the granular formation the solution shows a very slight acidity, such as might be caused by completely dissociated acid phosphate; and the electric conductivity slowly increases, corresponding to the for- mation of the granules and the liberation of mineral ions. 104 DIFFERENTIATION AND SPECIFICITY OF STARCHES. Table S. Duration Gramg Specific of change of extract rotatory in days. per 1000. power. 6 27.44 18S 2.9 21.C5 184.1 7.7 17.72 182.45 16.9 14.8 181.3 29.0 13.48 180.5 48.7 12.25 180.24 98 11.43 17S.2 174 11.04 178 Table 6. Number Grams Specific of hours of extract rotatory of filtering. per 1000. power. 6 27.48 188 30 29.61 186.3 78 32.13 184.15 126 35.5 184 In another inquiry, Fouard (Compt. rend., 1908, cxlvii, 931) investigated the altera- tions which occur in the pseudo-soUition upon standing. A neutral pseudo-solution con- taining 54.3 grams of starch per Uter at 15° was studied by removing portions at definite intervals ranging from a few hours to many days, filtering tlirough collodion membranes, and determining the amounts of solid matter and the specific rotatory powers. The prepa- rations, he found, undergo spontaneous gelatinization, and the process goes on during a period of several months. The results of these experiments, Fouard records, confirms liis former view of starch existing in different states of molecu- lar aggregation, each state having a definite and specific rotatory power. During the process of gelatinization the collodion membrane transmits these chfferent aggregates in definite order, the first to be checked being the most active polariscopically. The results of this series of experiments are shown in table 5. The rotatory power of the filtrates remained constant. In another series of experiments a similar pseudo-solution kept at 20° was filtered durmg a period of 126 hours, and samples taken after the lapse of 6, 30, 78, and 126 hours, respectively, with the results shown in table 0. The residual colloidal starch gelatinized on the filter during the sixth day, owing to concentration of the solution, causing the amount of starch in the filtrate to fall rapidly. In another inquiry, Fouard (Compt. rend., 1909, cxlvii, 502) found that the specific rotatory power of true solution of starch was reduced by the addition of small quantities of potassium hydrate, the larger the quantity of aUcaH the greater the effect, the rotatory power falling as low as (a)D =141. Neutralization instantly reverses the modification. He made further observations on the fractions obtained from pseudo-solutions at different stages of spontaneous gelatinization, and noted that as gelatinization advances the specific rotatory power of the dissolved starch falls towards the level of the rotatory power of maltose. He believes that the solu- tion of starch, whether by water or by the action of potassium hydrate, depends upon a reversible hydrolysis, with maltose as its ultimate term. Therefore, starch is simply a molecular aggregate of maltose in variable and unknown form. He concludes that no chemical combination takes place between the starch and alkaU, that starch has not acid functions, and that the action is essentially ionic. The presence of perfectly soluble starch in solution lowered the conductivity of the solution, but to a much less extent in the presence of ammonia than potassium. Before coagulation occurs, association of the starch with metalUc ions is beUeved by him to take place. The use of dilute acids to cause a reduction of starch-grain or starch-paste to soluble starch, or to render raw starch more soluble in cold water, was adopted by a number of the eai'Uer investigators, but it is probable that Bechamp (Compt. rend., 1854, xxxix, 653; 1856, xlii, 1210) was the first to use this method to obtain soluble starch, which substance he regarded as a special body occurring between common starch and dextrin. Since that time this method, and also various modifications of it, have been in continuous use, and no one, it seems, has so popularized it as Lintner (Jour. f. prakt. Chemie, 1880, xxxiv, 378), as is indicated by the comparatively very frequent references to "Lintner's soluble starch." His method is given as follows : MLx some good potato starch with enough 7.5 per cent hydrochloric acid to cover the starch. Allow tliis to stand at ordinary room temperature for 7 days, or for 3 days at 40°, when the starch will have lost its gelatinizing property. Decant with cold water until sensitive litmus paper shows no acid reaction. SOLUBLE STARCH. 105 and dry the starch in tlic air. This gives a preparation that is readily soluble in hot water, yielding a clear solution. A 2 per cent solution will remain clear or slightly opal- escent for two days, after wliich it will become cloudy, but if no further change takes place tliis will not interfere with its use for diastatic experiments. A concentrated (10 per cent) solution on cooling becomes salve-Uke. The solution will reduce only a minute quantity of Fehling's solution, so that this need not enter into consideration in using the starch for experimental purposes. Ten per cent hyth-ochloric acid will inunediately turn starch into soluble starch, differing in tliis respect from sulphuric acid, for if the latter be employed it wiU be necessary to use at least a 15 per cent solution and to conduct the experiment at 40°. Brown and Morris (Jour. Chem. Soc. Trans., 1889, lv, 449) subjected potato starch to the action of 7.5 to 12 per cent hydrochloric acid. With the former, the power of gela- tinizatiou was lost in 10 days; with the latter within 24 hours. This alteration is not accompanied by the least change of structure of the starch-grain or in the behavior of the starch-grain in polarized light. In hot water (60° to 70°) the grains form a perfectly limpid solution, and from the solution there separates, on cooling, a substance in the form of a white, flocculent, amorphous precipitate which has all of the properties of soluble starch prepared by the limited action of diastase or dilute acids on starch-paste at elevated temperatures. Both in solution and in the solid state this soluble starch is colored intensely blue in the presence of iodine. In solution it is without reducing power on copper solu- tions, and it has a specific rotatory power (a)j 3.86= +216. Even after 20 days the microscopic form of the granules is retained, notwithstanding the profound modification in the starch. Lintner's method was modified by Lake (Jour. Soc. Chem. Industry, 1894, xiii, 264) by centrifugaUzation of the starch-hydi'ochloric acid preparation. The starch is placed in an equal volume of 5 to 10 per cent hydrochloric acid and centrifugahzed, the action of the centrifuge causing a very intimate penetration of the starch by the acid. The acid is removed by washing with cold water and with dilute solutions of alkaUne carbonates, after which the preparation is centrifugahzed, and the starch collected and ch-ied at a temperature of 20° to 40°. Starch prepared in this way dissolves in hot water without gelatinization. The process may also be carried out with nitric, sulphuric, phosphoric, oxalic, lactic, acetic, and other liquid or volatile acids. Virneisel (Jour. Soc. Chem. Industry, 1899, xviii, 697) digested starch in a 1 to 2 per cent solution of mineral or organic acid, preferably sulphuric acid, at a temperature of 50° to 55° for about 12 to 15 hours. Dextrin and sugar were not formed. Blumer (Jour. Soc. Chem. Industry, 1903, xxii, 310) slowly heated starch in a 1 per cent solution of volatile acid for 5 to 6 hours at 115°, the acid afterwards being removed by distillation. The foregoing processes were modified by Cross (Jour. Soc. Chem. Industry, 1903, XXII, 1008) by heating the starch m a monocarboxylic acid with or without a dehydrat- ing agent, such as alcohol or a concentrated solution of neutral salt. For instance, starch dried at 100° is intimately mixed with one-thu'd to one-half of its weight of glacial acetic acid, heated in a steam-jacketed vessel for 1 to 2 hours, and the starch then freed from acid and dried. Formic or lactic acid may be employed, but preferably in the presence of a dehydrating agent to prevent gelatinization. Simiar procedures were adopted by Wotherspoon (Jour. Soc. Chem. Industry, 1904, XXIII, 29). For instance, th-ied starch is treated with glacial acetic acid to the extent of 10 to 50 per cent of its weight, and the preparation heated in a closed steam-jacketed converter until the starch is soluble in hot water. If an aqueous acid be employed, an inert dehydrating agent may be used, such as alcohol or brine, to prevent gelatuiization. The acid may be distilled off. 106 DIFFERENTIATION AND SPECIFICITY OF STARCHES. Tanret (Compt. rend., 1909, cxlvii, 1775) treated raw starch for 30 minutes with a cold 1 : 1000 hydi'ochloric acid, washing well with water, drying at 30° until the per cent of water was reduced to 17, then heating at 100° to 110° for 1 hour in a closed vessel. This solution was extracted with 25 per cent alcohol, which dissolved out about half of the starch. About one-fourth of the residue was soluble, and it had a rotatory power (a)o = + 188.6°. Upon the addition of 95 per cent alcohol to an alcoholic solution, precipitation occurred. The precipitate was washed with absolute alcohol, dried first over sulphuric acid, and then at 100°. The product was slightly soluble in boiling water, and had the following rotatory power : {a)^ = +208° to 210°. The insoluble portion gave an insoluble blue compound with iodine and resembled the granular amylose of Maquenne and Roux. The residue obtained by evaporating the alcoholic solution gave reddish-violet and red reactions with iodine. (See Dextrins, page 120.) Welwart (Chem. Zeit., 1907, xxxi, 126) states that the easiest means of making soluble starch is to boil a thin paste in a chamber with formic acid, which produces a more limpid preparation than acetic acid. The acid is subsequently expelled by boiling. Soluble starch may also be prepared by the action of sulphur dioxide at a temperature of 80° to 115°, as was done, for instance, by Berge (Bull. Assoc. Beige d. Chimistes, 1897, X, 444). He found that in the presence of either gaseous or liquid sulphur dioxide soluble starch is produced at the above temperature, while at higher temperatures dextrin and sugar may be formed (see pages 128 and 131). Skraup (Ber. d. d. chem. Gesellsch., 1899, xxxii, 2413) obtained by gentle acetylation of starch a derivative which on saponification yielded a substance which resembled soluble starch, giving a blue reaction mth iodine. Chlorine or chlorine-water were used by Siemens and Halske (Jour. Soc. Chem. Industry, 1898, xvii, 257). By this method the starch is made into a sludge and introduced into a vessel provided with stirring machinery. After raising the temperature to 45° the starch is treated with chlorine or chlorine-water, the product is then washed with water, and dried in the usual manner. The action of chlorine leads to a partial rupture of the coats of the starch-grain, and the halogen enters the grain and attacks boches wliich give to raw and bleached commercial starches an ethereal odor, which are subsequently removed by washing with water. The chlorine process was modified by Hartwig (Jour. Soc. Chem. Industry, 1905, xxiv, 1024), who subjected raw starch, especially corn starch, at a temperature of 50° to 86°, to the action of an excess of cUorine for 4 to 8 days, or as long as necessary to render the starch perfectly soluble in hot water. Gatin-Gruzewska and Maquenne (Compt. rend., 1908, cxlvi, 540) prepared soluble starch by adding to a 3 per cent starch-paste one-fourth its volume of a hot 40 per cent solution of potassium or sodiinn carbonate, and treating with one-thu'd its volume of alcohol. A precipitate of " amylopectin " (see page 113) is tin-own down. The filtrate contains a quantity of starch in solution. Or, 10 grams of potato starch are placed in 500 c.c. of a 1 per cent solution of sodium carbonate, sufficient water is added to make 1 liter, the preparation is neutralized with acetic acid, an equal volume of water added, and the whole set aside for 24 hours. The filtrate contains starch in solution and stains blue with iodine. Or, if starch be boiled with a concentrated solution of sodium citrate or sulphate, and the mixture be filtered, the filtrate will be found to be rich in soluble starch. Wolff (Jour. Soc. Chem. Industry, 1906, xxv, 437) treated starch at the ordinary temperature for an hour and a half in double its weight of a solution consisting of 2 to 4 parts per 1000 of potassium bichromate and about 15 per cent of sulphuric acid. The starch is then washed vmtil the excess of acid is removed, and then dried at a temperature of about 30°. The bichromate can be replaced by half its weight of permanganate. The starch thus prepared forms a paste which, when heated in the presence of traces of basic SOLUBLE STARCH. 107 matter, as, for instance, tap-water, becomes fluid and transparent. The liquefied paste gradually gelatinizes when cooled, but is readily liquefied by heating. If the prepared starch be washed in distilled water instead of tap-water containing traces of basic sub- stances, and dried at 30°, it may be converted into soluble starch by heating in the dry state at 80° to 100° for a few hours. Starch may be liquefied by boiling in a weak solution of aluminum chloride (Jour. Soc. Chem. Industry, 1901, xx, 492). The starch is boiled in the solution of aluminum chloride in a similar manner to that followed when dilute acids are employed, and the aluminum may be separated by the subsequent addition of sodium silicate. The use of aluminum chloride prevents the extraction of objectionable nitrogenous matter from the grains. Hydrogeii peroxide and ammonia were used by von Asboth (Chem. Zeit., 1892, XVI, 725) to prepare a soluble starch, which he called amylodextrin. This could be precip- itated by alcohol, and it was found to represent about 80 per cent of the original starch. Various other familiar oxidizing agents have been employed by different investigators, especially permanganate of potassium. Permanganate of potassium was used by von Siemens and Witt (Jour. Soc. Chem. Industry, 1896, xv, 366). The starch is stirred in water and a saturated solution of per- manganate is added until the pink color of the latter in the liquid persists. Manganese peroxide is deposited in state of extreme division on the grains, which assume a brown color. The starch is well washed, treated with a dilute solution of hydrochloric acid of a 0.5 per cent strength upward, according to the nature of the starch, then treated with alkali, and as soon as it gives a perfectly clear solution it is filtered off and freed from the manga- nese salt by sulphm-ous acid or bisulphite, then washed and dried. A number of oxidizing agents, including bleaching powder, chloric acid, magnesium dioxide, chromic acid, and potassium permanganate were used by Schmerber (Bull. Soc. Ind. Mulhouse, 1896, 238) to hquefy starch, but the last substance was found to give the best results. Equal quantities of starch and water are mixed thoroughly until the starch is in suspension, to which is then added one-fiftieth of the weight of starch of a warm 40 per cent solution of permanganate. The mixture becomes at first a redcUsh violet, turning to a dark brown. After 24 hours, to a solution consisting of about one-twentieth the weight of starch there is added hydrochloric acid in 5 volumes of water at intervals during constant stirring until the starch is decolorized, after which the starch is washed repeatedly to remove the acid and manganous chloride, and dried. The starch thus pre- pared yields a solution that is limpid and transparent. On standing the solution becomes cloudy, wliich cloudiness disappears upon heating. Schmerber's results were confirmed by DoUfus and Scheurer (Bull. Soc. Ind. Mulhouse, 1896, 241). Fernbach and Wolff (Seventh Internat. Congress Appl. Chem., London, 1909) Uquefied starch to the limpidity of water in 15 minutes by subjecting a 5 per cent starch-paste with a few drops of hydrogen peroxide and ammonia at a temperature of 70° to 75°. To each 5 c.c. of paste was added an amount of hydrogen peroxide equal to 0.005 gram of oxygen, and 0.004 gram of ammonia, but an excess of alkali was not found to be harmful, and it noted that other alkalis may be substituted. Weak acidity retarded liquefaction. Other hquefying substances were studied, especially the sulphates of iron and copper. In the hquefaction in the presence of ammonia some product in very small quantity is formed which partially neutralizes the ammonia. With a relatively large amount of hydrogen peroxide together with ferrous sulphate and sodium hydroxide the quantity of unreduced starch was very small, while with hydrogen peroxide alone a large quantity of unreduced starch remains. By subjecting 25 c.c. of the paste with 0.001 gram of ferrous sulphate, 0.042 gram of oxygen in the form of hydrogen peroxide, and 0.0006 gram of sodium hydrox- ide for 2 hours at a temperature of 75° to 78°, the preparation no longer gave a color reaction with iodine, and it had acquired an acid reaction and a high reducing power. The product 108 DIFFERENTIATION AND SPECIFICITY OF STARCHES. tlid not form an osazone, and therefore it was not regarded as being saccharine. The reducing substance differed essentially from dextrin because of its being precipitated from concentrated solution by copper sulphate, and therefore more like a gum. Thompson (Jour. Soc. Chem. Industry, 1904, xxiii, 29) treated starch in a neutral, alkaline or acid medium, preferably at high temperature, with more permanganate than necessary to oxidize all of the extractive impurities, the treatment being continued until all the starch was converted into a soluble form. Wolff and Roux (Compt. rend., 1905, cxli, 104G) modified the permanganate process; they also used bichi'omate and chlorine instead of permanganate, and they extended our knowledge of the properties of soluble starch. Starch was mixed with twice its weight of a 1 per cent solution of potassium permanganate containing 10 to 15 per cent of sulphuric acid, or 6 to 7 per cent of hydrochloric acid. The liquid becomes decolorized in about 2 hours, when the starch is filtered off, washed, and dried at 30°. A 5 per cent paste made with distilled water is little less viscid than ordinary paste, and generally behaves like ordinary starch-paste, but if treated at 70° in the presence of traces of basic substances such as ammonia, or hydroxides, or carbonates of alkalies or alkaline earths, etc., there occurs immediate liquefaction. This liquefying action takes place very slowly at ordinary temperatures, it is at its maximum at 70° to 75°, and it is somewhat different from the liquefying action of malt because of its persisting above a temperature of 80°. The hot solutions of this liquefied starch gelatinize as the temperature falls, and become a jelly at ordinary temperatiu-es, and again pass into a limpid solution upon heating. Unless the action of the permanganate is checked the starch is gradually broken down. Leitner (Zeitsch. f. angew. Chemie, 1890, 546) followed the progress of the actions by testing the preparation from time to time with iodine in the manner pursued when diastase is used. He recorded that the color reaction, which at first is blue, changes to violet, red-violet, red, and reddish-brown, which becomes weakened, until finally there is an absence of color response. The products, he states, are different from ordinary dextrins, as is shown by their forming precipitates with basic lead acetate and barium hydroxide, and by theii' acid reaction. They have a slight reducing power, and upon boiling expel carbon dioxide from carbonates. Glycerol breaks down starch into soluble starch and, if the action is continued or is carried on at a liigher temperature, dextrins and related bodies are formed. Zulkowsky (Ber. d. d. chem. Gesellsch., 1891, xxiii, 3295) found that starch heated in glycerine is first changed to soluble starch, and that as the temperature rises erytlu-odextrin and then achroodextrin are formed, together with a series of bodies which are increasingly soluble in alcohol and which could be separated by means of the chfferences in solubility. Difficulty was met with in removing the last traces of glycerine, owing probably to the formation of glycerides. This research was supplemented by another by the same author in conjunction with Franz (Chem. Centralbl., 1894, ii, 918; 1895, iii, 557), who studied somewhat in detail the varying products formed at different temperatures. At 190° soluble starch is produced which gives a deep-blue reaction with iodine and has a specific rotation of -(-188.3°. This starch is precipitated by alcohol, lime-water, or baryta- water. The alcohol precipitate, after drying and upon keeping, is gradually converted into the insoluble form. When kept in strong solution it gradually gelatinizes and becomes insoluble. At 200° erythi-odextrin is formed, and at 210° achroodextrin. Wlien the prepa- ration is subjected to prolonged heating related carbohydrates are produced, one resembling gum arable. Pregl (Monatsch. f. Chem., 1901, xxii, 1049) used 10 times the volume of glycerine and purified the product by filtering the solution into 60 per cent alcohol, collecting and drying the precipitate. The specific rotatory power of the aqueous solution was found to be (a)u = -1-191.26°. From this preparation he made and studied acetyl derivatives. SOLUBLE STARCH. 109 Iodine was employed l)y Rodcnwald and Kattein (Sitzungsbcr. d. Berliner Akad., 1S99, xxxiii-xxxiv, 628), wlio heated starch in Lugol's solution in the autoclave for 15 min- utes at 130°. A greenish-brown liquid is formed, which as stated consists essentially of iodide of starch and some sugar. The starch iodide, they found, can be obtained pure by dialj'sis and filtration. The quantity of iodine in combination was ascertained to be con- stant at 14.3 to 14.8 per cent. The starch iodide is decomposed by heating, and as the iodine is freed it is removed by a current of steam, leaving a clear solution. On cooling, granules separate which give a blue reaction with iodine, and when dried the granules are insoluble in cold water, but form a paste on boiling, and swell in potassium hydrate. Some of the grains measured 0.02 mm. in diameter. Formaldehj^de has a marked decomposing action on starch. Syniewski (Ann. d. Chem. u. Pharm. 1902, cccxxiv, 201 ; Jahr. ii. d. Fort. d. Tierchemie, 1902, xxxii, 100) put potato starch in 5 times its weight of 40 per cent formaldehyde, and at the end of two months the starch was found to be completely dissolved. The solution was opalescent, but showed no iodine reaction. Upon evaporation he obtained a homogeneous mass consisting of trioxymethylen and a crystalline condensation product. The addition of water caused a gradual disintegration of the latter, as was indicated by changes through a scale of colors from brown to red and violet and blue, owing to the separation of the formaldehyde. This cleavage could be accelerated by the presence of acid. The product present which gives the blue reaction with iodine he regards as amylodextrin or a body closely related to it. He notes that amylodextrin otherwise prepared is dissolved at once in formalde- hyde, with a transition of colors the reverse of that which is observed when the formalde- hyde combination is broken up; and the reverse play of colors is also found to occur when the solution is diluted. The amylodextrin thus prepared undergoes crystallization upon evaporation of the solution. Under the microscope, stars composed of needles were first seen, and these were surrounded by a substance arranged in radial concentric layers. These bodies were deceptively like starch-grains, and under the polariscope the similarity was fiu-ther showai by the double refraction and the interference figure. This amylodextrin was found to form a definite dark-blue iodine compound. Reichard (Zeit. ges. Brauw., 1908, xxxi, 161) confirmed Syniewski's views regarding the formation of a formaldehyde-amylodextrin compoimd by the action of 40 per cent formaldehyde at ordinary temperatiu^e. He showed that at higher temperatures the process goes on very rapidly, and that the temperature of gelatinization is reduced with an increase in the concentration of the formaldehyde. At 15° to 10° a 38 per cent solution of formalde- hyde gelatinizes the starch completely in two days. At 25° 1 gram of starch in 10 c.c. of the same solution of formaldehyde was completely gelatinized in 7 to 8 hours. The action is less rapid if the formaldehyde is neutralized. Potassium hydrate was used by B4champ (Compt. rend., 1854, xxxix, 653) to dissolve starch, and it has since been used by a large number of investigators and is one of the most valued agents when pure soluble starch is desired, because it, unlike dilute acids, diastase, glycerol, permanganate, and most other agents, does not give rise to dextrin or to any reducing substance; in other words, with a weak solution of alkali the decomposition of starch ceases at the stage of liquefaction. ToUens (Nachricht d. k. Gesellsch. d. Wis- sensch, etc., Gottingen, 1873, 762) also used potassium as well as sodium hydroxide to prepare soluble starch. Wr6blewski (Ber. d. d. chem. Gesellsch., 1897, xxx, 218) prepared soluble starch in the following way: 100 grams of rice starch are rubbed with a small quantity of a 1 per cent solution of potassium hydrate, and set aside for 2 to 4 hours, then more of the solution is added until the whole has attained a volume of 600 to 800 c.c. This jelly-mass is heated on a water-bath, stirred until fluid, and then boiled for 20 to 30 minutes, filtered, acetic acid added to slight acidity, precipitated with alcohol, dissolved and precipitated, redis- 110 DIFFERENTIATION AND SPECIFICITY OF STARCHES. solved in a small quantity of water, poured into a large quantity of alcohol under constant stirring, the precipitate washed in absolute alcohol and ether, and finally dried in vacuo. This method yields a snow-white preparation which contains little ash. It gives an intense blue reaction with iodine; it does not reduce Fehling's solution; and it is soluble in water to the extent of about 4 per cent. By the prolonged boiling of a dilute solution a small amount of reducing substance is formed. If a strong solution of alkali is used, as by Biilow (Archiv f. d. ges. Physiologic, 1895, lxii, 131), decomposition continues beyond the starch stage. Wr6blewski (Chem. Zeit., 1898, xxir, 375) reported a process for preparing a soluble starch by means of caustic potash that can be filtered through porcelain. On a small scale, the soluble starch can be prepared by rubbing 20 grams of starch in a mortar with 100 c.c. of cold water, pouring this into a 2-liter flask, adding about 1 liter of 0.5 per cent potassium hydrate, and boiUng under an inverted condenser for VA to 2 hours until it becomes limpid and of a pale-yellow color. The preparation is neutralized with acetic acid, pre- cipitated with an equal volume of alcohol, filtered, and filtrate washed successively with 50 per cent, 95 per cent, and absolute alcohol, and finally with ether, and then dried in vacuo. The snow-white powder thus obtained is soluble to the extent of about 3 per cent in water: and it contains about 0.4 to 0.6 per cent of ash. The less the ash the stronger the alcohol necessary to cause precipitation. The caustic-alkali processes for producing soluble starch were modified by Kantorwicz (Jour. Soc. Chem. Industry, 1904, xxiii, 1038; 1905, xxiv, 144) by the introduction of alco- hol, acetone, mixtures of alcohol and ether, etc., into the process to prevent the gelatinization of the starch while it is being rendered soluble by the action of the alkali. The starch is mixed with 50 to 90 per cent alcohol, and to this is added two-fifths the weight of the starch of a solution of caustic soda at 30° Baume. The mass thickens, and is set aside for an hour, then neutralized with acetic acid, and the resulting precipitate filtered off and dried. One part of the dried product with 10 parts of water yields a viscous, liquid-like paste. Both oxidizing agents and caustic alkalis have, as shown, been found valuable in ren- dering starch into a soluble form. Syniewski (Ber. d. d. chem. Gesellsch., 1897, xxx, 2415; 1898, xxxi, 1791) combined both by using sodium peroxide. 50 grams of sodium peroxide were dissolved in 500 grams of water and rubbed up with 50 grams of starch in 500 grams of water, and set aside for an hoiu-, after which the preparation was precipitated with 95 per cent alcohol, the precipitate dissolved in water, neutralized with acetic acid, and again precipitated with alcohol. By repeated solution and precipitation a snow-white powder was obtained which is ^^el•y soluble in hot water, and soluble to the extent of 12.5 per cent in water at ordinary temperature. It is colored blue with iodine, does not reduce Fehling's solution, and has a rotatory power at 20° of -|- 182.66° for a 2.5 per cent solution and -f- 189.5° for a solution of 12.5 per cent. In his second article he gives the rotatory power of a 10 per cent solution as (a)Di20 — +195.3°. Solutions containing more than 12.5 per cent could not be read in the polarimeter because of the deposition of a white precipi- tate, which precipitate, by analysis, he found to be formed from the soluble starch, pre- sumably by dehydration, and therefore a reversion product. This substance was also formed in solutions which gelatinized upon cooling, thus giving rise to a gelatinous mass identical with ordinary starch-paste. He describes the reactions with baryta-water, acetylchloride, and benzolchloride, and also the invertive action of water under pressure, and of alcohol, and diastase. A form of starch which possesses all of the properties of starch gelatinized at high tem- peratures may be prepared by treatment with ammonium thiocyanate, according to the process of the Arabel Manufacturing Co. (Jour. Soc. Chem. Ind., 1909, xxviii, 213). 100 parts of dry starch are mixed with 80 parts of a 50 per cent aqueous solution of this salt, to which 40 parts of alcohol are added. The product swells in cold water to form a paste. The thiocyanate can be removed by wasliing with alcohol, acetone, etc. REVERTED STARCH — ARTIFICIAL STARCH, ARTIFICIAL STARCH-GRAINS. Ill Soluble starch has also been prepared by the agency of sodium perborate (Stolle and Kopke, and Fritsche, Chem. Abstracts, Amer. Chem. Soc, 1909, iii, 1105), by using 1 to 2 parts to 100 of starch and heating at 30° to 40° for 5 hours. It may also be made in like manner by using a mixture of acetic and nitric acids instead of perborate. THE REVERSION OF STARCH-PASTE, SOLUBLE STARCH, AND AMVLODEXTRIN INTO COAGULATED AND INSOLUBLE FORMS OF STARCH. "ARTIFICIAL STARCH" AND "ARTIFICIAL STARCH-GRAINS." The reversion of soluble starch to an insoluble condition was noted by Brown and Heron (Ann. d. Chem. u. Phar., 1879, clxxxix, 266), who found that starch-pa.ste upon standing becomes opaque, owing to the formation of an insoluble form of starch. Salomon (Jour, f. prakt. Chemie, 1883, xxviii, 82) observed the same phenomenon in preparations of what he terms soluble starch (page 115). After precipitating and purifying, dissolving, and reduc- ing to the consistence of a syrup, and setting aside over night, there was deposited a white powder consisting of fine grains almost insoluble in cold water. A large part of the so-called soluble starch separated from a concentrated solution as a white powder. (See page 192.) The phenomena of the reversion of dissolved starch into an insoluble form were inves- tigated particularly by Rodenwald and Kattein, Wolff and Fernbach, jMaquenne, Roux, and Coombes. Rodenwald and Kattein (Sitzungsber. Kgl., pr. Akad. Wiss., Berlin, 1899, xxiv, 628) prepared a form of soluble starch by means of iodine-potassium iodide in con- junction with the autoclave, etc. (see page 109), which when hot is clear, or at most faintly turbid, and which upon cooling deposits a white substance having the form of "artificial starch-gi-ains." Both the filtrate and these artificial grains give the blue reaction with iodine. The grains when dried are insoluble in cold water, but are gelatinized upon boUing, and swell and form a paste in the presence of potassium hydrate. Wolff and Fernbach (Compt. rend., 1903, cxxxvix, 718) coagulated solutions of starch by means of amylocoagula^e, an enzyme which they discovered in the green seeds of cereals and associated with amylase in a large number of ripe seeds, germinating seeds, leaves, etc. They found that 5 c.c. of a 10 per cent infusion of air-dried malt is sufficiently strong to coagulate 100 c.c. of a 4 to 4.5 per cent solution of potato starch (prepared in the auto- clave at 130° for 2 hours) in 20 to 30 minutes at 15° to 25°. The stronger the solution of starch the quicker the coagulation and the denser and more coherent the mass. ^\'ith weak solutions the action of the amylocoagulase is liindered or prevented by the antagonistic action of diastase. The coagulum never exceeds 30 per cent of the starch present, and it is readily soluble in hot water. Malt extract looses its coagulating property if subjected to a temperature of 65° for five minutes. A solution of starch prepared by the restricted action of diastase was found to be not so suitable for coagulation as that prepared in the autocla\-e. Moreover, the reverted starches formed from the soluble starches prepared in these two ways are not identical. The coagulum, or reversion product, or amylocellulose, as it has been termed, was further studied by Maquenne, Fernbach, and Wolff (Compt. Rend., 1904, cxxxviii, 49), who found that the formation of amylocellulose in the coagulum is more rapid when pre- pared by the action of amylocoagulase on starch-paste heated at 120° for 15 minutes, or at 100" for 30 minutes, or liquefied by heating in the autoclave at 130° for 2 hours, than when amylocellulose is formed during the spontaneous reversion of starch-paste. The coagulum immediately after its formation contains very httle amylocellulose, but the latter increases and may reach 50 per cent of the total coagulum. The amylocellulose obtained by the reversion of starch-paste is saccharifiable by diastase, but not so with that formed by the action of amylocoagulase. Additional observations were recorded by Fernbach and Wolff (Jour. Fed. Inst. Brewing, 1904, x, 216) regarding the properties of amylocoagulase. They note again the antagonistic actions of diastase and amylocoagulase, and that if the action of the former 112 DIFFERENTIATION AND SPECIFICITY OF STARCHES. liredominates coagulation will not occur. Relatively high temperatures, they ascertained, are favorable to the action of diastase, while low temperatures are favorable to amylo- coagulase, and coagulation is favored as the temperature is lower and the duration of heating shorter during the preparation of the solution of starch. With a 2 per cent solution of starch, coagulation may occur at 8° but not at 20°; but with a 4 per cent solution coagulation takes place readily at 15° to 2.5°. The proportion of coagulum formed may be increased by various agents, and it was raised from 14.4 to 24.1 per cent with sodium hydrate at 15°. They note here that only a part of the amylocellulose is resistive to malt extract at 67°, and also that while the activity of amylocoagulase is destroyed by heating in the moist condition for 5 minutes at 63°, it is not destroyed in the dry state, and that it is present in kilned malts as well as in cured malts. During the same year two articles were published by Maquenne (Compt. rend., 1904, cxxxviii, 213 and 375), in which he recorded the results of his investigations on the pro- portions of amylocellulose formed from starch-solutions of different strengths. Such solu- tions were prepared by gelatinization of starch in a bath of boiling water, and afterwards by heating the preparation in the autoclave for 15 minutes at 120°. These solutions were set aside for 4 days at 9° to gelatinize, and then were saccharified by malt extract, and the results compared with those of similar solutions that were saccharified immediately after being taken from the autoclave. With the reverted solutions the amount of soluble matter formed was found not to increase in proportion to the amount of starch, contrary to what was found in the check solutions. Maquenne regards both raw starch and reverted starch as containing amylocellulose in different stages of condensation, that amylocellu- loses are obtained from comminuted grains by solution in cold water, that they are there- fore present in reverted starch-paste and in raw starch-paste, and that they are dissolved at high temperatures and therefore not apparent in freshly prepared starch-paste. Further studies of the reversion phenomena of soluble starch were made by Roux (Compt. rend., 1905, cxl, 440, 943, 1259). In the first of these articles he shows that the reversed action may proceed at any temperature between 0° and 150°, but that at the latter temperature the amylocellulose liquefies and then undergoes a degradation into a less complex form. He made the very interesting observation that, by the incomplete degradation of amylocellulose, "artificial starch-grains" can be produced which under the microscope show the peculiar structure of natural starches, and he believes that both natural and artificial starches are mixtures, and that they differ chemically owing to the existence of forms of variable degrees of condensation of the same nucleus. In the second article he describes three types of artificial starch of different degrees of conversion, which are completely soluble at 150°, 120° and 100°, respectively. The reversion was more rapid, he noted, in the least soluble; 4.3 per cent had reverted in 1 hour in the preparation soluble at 150°; while in the one soluble at 100° the percentage was 1.8. The rapidity of reversion was increased by the presence of acid or alkali in proper proportions. The reverted product could be redissolved only at a temperature not lower than the temperature at which the original starch was soluble and, since it had the same properties as the latter, he holds that it represents a return to the initial state. In the third article Roux compares the phenomena during saccharification of artificial and ordinary starches. Using maltase at a temperature of 56° for 4 hom-s, the apparent amounts of maltose formed were on an average 97.9 and 82.3 per cent in the artificial and ordinary starches, respectively. At 67° the percentages were 55.1 and 45 respectively. The dextrins formed from artificial starches, unlike those from natural starches, were almost completely soluble in alcohol. These investigations were supplemented and extended by Maquenne and Roux (Compt. rend., 1905, cxl, 1303; 1900, cxlii, 124), who found that artificial starches prepared by liquefaction of amylocellulose at 150° and yielding 96 to 98 per cent of maltose, AMYLODEXTUIN AND MALTODEXTRIN. 113 can by repeated reversion and liquefaction be made to yield 102 per cent of maltose. Ordi- nary starch-paste, they state, when measured by saccharifying and iodine tests, contains about SO per cent of amylocellulose or artificial starch. The remaining 20 per cent of the starch is composed of a mucilaginous substance which they refer to as amylopedin, to which the gelatinizing property of starch they believe is due. Amylocellulose they describe as being partly soluble in boiling water and completely soluble at 150°, colored blue with iodine, devoid of gelatinizing power, and entirely converted by diastase into maltose at ordinary temperatures. Amylopectin has the prop- erty of gelatinization ; it does not give a blue reaction with iodine; and is not saccharified by malt diastase, but merely converted into dextrins. They believe that amylopectin has the property of interfering with the reversion of amylocellulose, both in the starch- grain and in starch-paste. The action of liquefying diastase, they hold, is exerted upon the amylopectin, and consequently that such a diastase should be named amylopedinase. The saccharifying diastases, they state, act only on amylocellulose. In the second contribution these authors record that the conclusions formulated in the previous article must be slightly modified as regards the percentage of amylocellulose, and that starch-pastes in their ordinary condition contain about 90 per cent of amylo- cellulose and about 10 per cent of amylopectin, unless the latter may be formed by an especially active diastase with an attendant separation of a starch residue. Maquenne (Compt. rend., 1908, cxlvi, 317) confirmed and extended some of the preceding investigations in proof of a non-colloidal or perfect solution of starch. Demineral- ized starch, he states, forms absolutely limpid solutions which are not coagulated by elec- trolysis, and which are transmitted by membranes and through the Chamberland filter, as are salt solutions. Coombes (Ann. de chim. et phys., 1908, xlxxix, 280) found that natural starch con- tains from 80 to 85 per cent of starch and 15 to 20 per cent of amylopectin, the latter being insoluble but gelatinizable in water, not colored blue with iodine, and not at all or but sUghtly saccharifiable by ordinary diastase. He believes that the group of starches contain substances which do not differ in themselves, but in their degrees of condensa- tion, the less concentrated being stained blue with iodine, being soluble in hot water and in potassium hydrate, and being readily saccharifiable without the production of residual dextrins; but, on the other hand, the most concentrated are not stained blue with iodine and are resistant to the action of diastase. St. Jentys (Chapter II, page 58), by the agency of tannin, prepared artificial starch- grains which resembled the most beautiful natural grains, such as those found in Dioscorea and Carina; and by gallic acid he obtained grains resembling those of wheat, buckwheat, and Chinese sokyes. Crystalline products which do not reduce Fehling's solution and which are formed from starch by bacterial activity were examined by Schardinger (Zentralbl. f . Bakteriol., 1908, xxii, 98) . He found that a pure culture of Bacterium macerans gives rise to two crystalline bodies which can readily be separated, owing to their marked difference in solubility in cold water. The less soluble crystallized into hexagonal flakes or prismatic crystals, and gave a yellow reaction with iodine. This substance he believes may be identical with Meyer's amylodex- trin. The more soluble substance crystallized from alcohol in lanceolate needles. He tenta- tively gave the name "crystallized amylose " to this body. The addition of iodine produced gray needles which in one position appeared blue. Neither body is fermentable with yeast. AMYLODEXTRIN AND MALTODEXTRIN. The word amylodextrin, like the term soluble starch, has been used so indiscriminately that it impUes no definite body, simple or compound, and hence has been the cause of considerable confusion. Etymologically it signifies a body intermediate between starch 8 114 DIFFERENTIATION AND SPECIFICITY OF STARCHES. and dextrin, or a mixture of the two, and it has been used in both senses, but it has also been employed synonymously with soluble starch and maltodextrin, the former being the fii'st dissolution product of ordinary raw starch, while the latter is one of the later prod- ucts of digestion; and it has been given to various intermediate products between true soluble starch on the one hand and sugar on the other, products giving reactions with iodine ranging from a purple and bluish-violet to wine color, brownish-red, or reddish- yellow, and even to some giving no color reaction. It is obvious, from the foregoing, that the word does not designate any specific individual or mixture, but some indefinite starch- like or dextrin-like or dextrin-sugar substance or compound or mixture that may be formed or exist during the saccharification of starch at any time between the moment of solution of starch-grain and the final stage of dextrin digestion. The same statement (with modi- fications) holds true in regard to maltodextrin. In quoting the literature the terms amylo- dextrin and maltodextrin will be used as the various authors use them. It is probable that Jaquelain (Ann. de chim. et phys., 1840, lxxiii, 167) was the earliest observer to specifically note the existence of a body having the characteristics of amylo- dextrin; that is, of a body or compound or mixture having intermediate properties between soluble starch and dextrin, which gives a purple or violet reaction with iodine. He records that after treating starch with water at 150°, and then cooling, there separated round grains wliich gave a purple coloration with iodine. He also makes note of preparing a precipitate in the form of a powder that gave the same reaction. This or a closely related substance was probably described by Musculus (Compt. rend., 1870, lxx, 857), at first under the name of insoluble dextrm. In a later article (Bull. Soc. chim., Paris, 1874, XXII, 26) it is described under the name amidon soluble. W. Nageli (Beitrage z. naheren Kenntnis d. Stiirkegruppe, Leipzig, 1874) seems to have been the earliest investigator to make a serious study of amylodextrin; that is, of a body that stands between soluble starch (which gives a blue reaction with iodine) and erythrodextrin (which gives a red reaction). He treated raw starch with weak acids in the cold for many days, and found that a substance (amylodextrin) goes into solution which the acid soon changes into dextrin and sugar. The starch-grains undergo a slow but complete solution, the periphery being the last to dissolve; but the grain residues yield amylodextrin as long as they give a color reaction with iodine, even though the reaction be a yellow. Amylodextrin crystallizes in small flakes on evaporation, or by freezing, or by tlie addition of alcohol. According to Nageli the flakes consist of small needles which are associated in the form of radial aggregates which resemble crystals of inulin. It is insoluble in cold water, but is readily soluble in water at 60°, this solution remaining clear upon cooling. A freshly prepared precipitate by means of alcohol is readily soluble in cold water. It does not itself possess the property of dialyzability, but is dialyzable when accompanied by dialyzable substances. It has lower rotatory power than starch, but higher than erythrodextrin, and it hkewise stands between starch and erythi'odextrin in its behavior toward alcohol and baryta-water, and in its affinity for iodine. It consists of two modifications, neither of which in a soUd state is colored with iodine, but which when in solution become violet and red, respectively. The rotatory power he gives as (a) +175° -177°. The soluble starch described by Musculus and Gruber (Zeit. f. pliy- siolog. Chemie, 1878, ii, 177) was misnamed and corresponds with the amylodextrin of Nageli, but is not identical. These authors recorded the product as being insoluble in cold water, soluble in water at 50° to 60°, and in aqueous solution giving a wine-color reaction, or when dry a violet, yellow, or brown reaction in the presence of an excess of iodine. The specific rotatory power is given as (a) +218°. Herzfeld (tjber Maltodextrin, Halle, 1879) opposed the view of the splitting of the molecules of starch into dextrin and maltose according to the theory of Musculus, and held that there is a consecutive conversion of soluble starch into erytlu-odextrin, achroodextrin, AMYLODEXTRIN AND MALTODEXTRIN. 115 maltodextrin, anil maltose. At temperatures lower than 65° he states that starch is converted into achroodextrin and maltose, but at higher temperatures erythrodextrin and maltodextrin also are formed. The maltodextrin he assumes is composed of two dextrin groups and one sugar group, and that it has the formula of C6H12OG. The specific rotatory and reducing powers he gives as follows: (a) j =71.6°, K23.5. Much attention was later given to maltodextrin by -wirious experimenters. In the preparation of soluble starch, Salomon (Jour. f. prakt. Chemie, 1883, xxviii, 82) obtained amylodextrin. This was prepared by boiling 100 grams of potato starch for 2K hours in 1 liter of water containing 5 c.c. of sulphuric acid, then neutralizing, precipi- tating with alcohol, redissolving and reprecipitating, and, finally, boiling to a syrup and setting aside over night. The preparation after the boiling gave a violet reaction with iodine. After standing over night there was found a precipitate in the form of fine grains, which he states resemble Nageli's amylodextrin crystals. A watery solution of this deposit gave a blue or a reddish-violet reaction with iodine, the latter coloration being attributed by Salomon to an impurity in the form of dextrin. The specific rotatory power is recorded as (a)j+211.50°-211.97°, (o)d+189.98°-190.24°. Meyer (Botanische Zeitung, 1886, xlix, 697, 713), in studying the nature of starch- cellulose, states that amylodextrin constantly origmates during the treatment of starch- paste with dilute acids, diastase, pepsin, saliva, and in general all substances which cause a disaggregation of the starch-substance in the presence of water. But if the action of the splitting agent is very energetic from the beginning, the transformation of the amylo- dextrin is so rapid that no skeletons will be formed. The skeletons, he states, consist of pure amylodextrin only when they no longer stain violet or blue, but yellowish or reddish brown when allowed to stand in iodine solution for about 5 minutes. He also noted the formation of spherocrystals (occasionally of plates) which resembled starch-grains, except that the interference figure in polarized light was not orthogonal but diagonal; but both saliva and acid skeletons, he records, behave toward polarized light the same as intact grains. The crystals and the skeletons reacted identically to various reagents, including iodine, from which he concludes that they are an identical substance. Carl Niigeli (loc. cit.) differentiated the soluble granulose from the skeleton-like cellulose by subjecting raw starch to the prolonged action saliva, and Walter Niigeli (loc. cil.) found that as long as these skeletons showed any color reaction with iodine they were converted into amylo- dextrin. Griessmayer (Allgem. Bauer, und Hopfenzeit., 1887, xxvi, 147), in studying the real nature of starch cellulose, recorded that the skeleton-like substance of the grains is converted by dilute acid into amylodextrin. He digested raw starch in a 12 per cent solu- tion of hydrochloric acid for 100 days in the cold. The skeleton-like residue was washed free from acid and other extraneous substances. When dried it represented 30 per cent of the original quantity of starch. It was almost completely soluble in hot water, and by freezing spherocrystals of amylodextrin were obtained. Brown and Morris (Jour. Chem. Soc, 1885, xvii, 526; Ann. d. Chem. u. Pharm., 1885, XXIII, 72; Jour. Chem. Soc. Trans., 1889, lv, 449 and 462) devoted considerable attention to maltodextrin and made, among other studies, comparative investigations of amylodextrin, maltodextrin, and inulin. In their earlier contributions they disagree with Herzfeld (page 114) that maltodextrin is a mixture of dextrin and maltose, and they hold that his so-called maltodextrin was impure. They state that a mixture of dex- trin and maltose prepared so as to have the same optical activity and reducing power as maltodextrin is separable into its constituents by treatment with alcohol, whereas true maltodextrin is not thus separable; from a mixture of maltose and dextrin maltose may be fermented off, but not so with maltodextrin; when a mLxture of dextrin and maltose is subjected to malt extract a residue of dextrin is always left, while with maltodextrin there is no residual dextrin; and when a mixture of dextrin and maltose is subjected to dialysis 116 DIFFERENTIATION AND SPECIFICITY OF STARCHES. it splits into dextrin and maltose, while maltodextrin dialyzes intact. They give malto- dextrin the formula \ L'^t"^^"! • In their article published in 1889, they point out the close resemblance between the amylodextrin of NageU and the maltodextrin described in their eariier contributions, and they state that this substance is neither identical with soluble starch, as held by some; nor is it identical with the starch-cellulose of C. Nageli, as held by Meyer; nor is it a mixture of dextrin and maltose, as held by Herzberg. They also assert that while the composition of amylodextrin can be expressed in terms of a mixture of starch and dextrin, it is really a well-defined chemical substance ; that it does not possess the optical properties ascribed to it by its discoverer; that in composition it is analogous to maltodextrin, which may be represented by the formula | (q'^jj^'q") ' ^^'^ constituted by one amyloin or maltose group with six amylin or dextrin groups; that amylodextrin, like maltodextrin, is hydrolyzed immediately into maltose; that it is an entirely different substance from soluble starch, with which it has been confounded ; and that the first action of dilute acid on ungelatinized starch in the cold is to convert the starch-substance into soluble starch, which is gradually hydrolyzed into amylodextrin, a portion going at the same time into solution and being changed into dextrose; and that there is a great similarity between amylodextrin and inulin. The formulae and molecular weights they give as follows : Am lodex-trin ^^^'^^"°^'^ • Maltodextrin^ ^^'"^^^^"^ • Inulin j '^'-°--°''^ I (Ci2H2oOio)g' ' (Ci2H2oOio)2' I (Ci2H2oOio)4 Mol. weight 2286 Mol. weight 990 Mol. weight 1980 The purified amylodextrin had the following specific rotatory and reducing coefficients, (a)j3. so = 206.11, and k 3.86=9-08, respectively. A fm'ther study of the maltodextrins ("amyloins") was made by Morris and Wells (Trans. Inst. Brewing, 1892, v, 133) by means of analyses of fermenting yeasts at different stages of fermentation. They found, in experiments with Frohberry and Saaz yeasts, that the properties of these yeasts so differ that one may be used to determine free maltose and the other to determine the fermentable maltodextrin. The Saaz yeast, it was found, does not ferment maltodextrins, but the Frohberry yeast does. Schifferer (Neue Zeit. Rub. Zuck. Ind., 1892, xxix, 167; Inaug. Diss., Kiel, 1892) endeavored to prepare the maltodextrins of Herzfeld and of Brown and IVIorris, but without success, and in no case was he able to separate a body which resembled in any way the so-called amyloins (maltodextrins) referred to by the latter. He regards Herzfeld's malto- dextrin as being probably a mixture of 26 per cent of dextrin and 74 per cent of isomaltose; and the maltodextrin of Brown and Morris as a mixture of 67 per cent of dextrin and 33 per cent of isomaltose. The amylodextrin (really soluble starch) separated by Lintner and Diill (Ber. d. d. chem. Gesellsch., 1893, xxvi, 2533) was obtained by stopping the action of diastase while the preparation still gave a blue reaction with iodine, and precipitating with hot 40 per cent alcohol. The precipitate thrown down during cooling was purified by repeated fractionation with 40 and 30 per cent alcohol, yielding an extremely light white powder that was slightly soluble in cold water, but very soluble in hot water. It gave a deep- blue reaction with iodine, and its specific rotatory power was (0)0 = 196, and its formula (Ci2H2oC>io)54- From a 20 to 30 per cent solution they obtained spherocrystals. The products of starch transformation by the actions of several kinds of yeast were studied by Hiepe (Country Brewer's Gazette, 1893 and 1894; Jour. Soc. Chem. Ind., 1894, XIII, 267), but he failed to find any homogeneous precipitate ha^ing the properties of a body between dextrin and isomaltose, and he states his belief that the intermediate products between dextrin and sugar (the amyloins of Brown and Morris) consist of a mixture of dextrin, isomaltose, maltose, and glucose. AMYLODEXTRIN AND MALTODEXTRIN. 117 Meyer (Die Stiirkekorner, he. cit., page 47) studied the microchemical behavior of a-ainylose, /3-aiiiylose, and amylodextriii. He prepared what he terms amylodextrin from both amj^Ioses by means of malt extract, dilute acids, and other chemical agents. He noticed tliat amylose in solution and in crystals, and amylodextrin, have different degrees of affinity foi- iodine. When crystallized /3-amylose (in the form of starch-grains) and amylodextrin are added to a solution of amylose, and a trace of iodine is added, and then gradually more and more iodine, it will be seen, he states, that the solution will be colored first, then the starch-grains, and then the amylodextrin. Pure amylose solution is colored a greenish-blue with a shght excess of iodine; pure amylodextrin, even in a dilution of 1-6000, is colored brownish-red; a mixture of both solutions, according to the relative amount of amylose, will be colored blue, blue-violet, violet, red-violet, pure red, and a brownish-red. Billow (Archiv f. d. ges. Physiologie, 1895, lxii, 131) obtained what he describes as amylodextrin by subjecting potato starch to the action of a strong solution of potassium hydroxide, heating the paste thus formed in a water-bath, diluting, neutralizing with dilute acetic acid, precipitating with alcohol, dissolving in water, reprecipitating, etc. He also prepared amylodextrin by means of diastase and by dilute sulphuric acid. By precipi- tation with barium hydroxide he prepared several forms of amylodextrin-barium combi- nations, which varied in rotatory power from 191.1° to 205.4°, and also in the percentages of the amylodextrin and the barium in the combinations. Ost (Chem. Zeit., 1895, xix, 1501), following Lintner and Diill's methods, prepared maltodextrin, but found that it had a liigher specific rotatory power than that given by Brown and Morris, and that the method employed by the latter investigators to determine the reducing power yields fallacious results. Ling and Baker (Proc. Chem. Soc, 1897, clxxiii, 3) studied in detail the products formed by the limited action of diastase at 70°, and they separated two unfermentable substances which they prepared free from extraneous matters. These they state were maltodextrin a, which is identical with the maltodextrin of Brown and Morris, but having the properties (0)0 = +180 and 72 = 32.81; and maltodextrin /3, identical with Prior's achroodextrin III (see below), and having the properties (0)0 = +171.6 and 72=43. They believe that starch is broken down by diastase into a series of maltodextrins of decreasing molecular weight and rotatory power, but with increasing reducing power, and which have optical and reducing properties equivalent to varying mixtures of starch and maltose. According to Prior (Bayerisches Brauer Jour., 1896, vi, 385), the final products of the action of diastase are 3 achroodextrius (I, II, and III) and maltose. Lintner's isomal- tose he looks upon as being a mixture of acliroodextrin and maltose. The achroodextrius I and II correspond with the same dextrins of Lintner and Diill, but achroodextrin III is a dextrin which he regards as immediately preceding the formation of sugar. This substance he found has a specific rotatory power of (0)0 = +171 and a reducing power of 42.5 per cent of maltose. The dextrins showed differences in fermentability. By the action of yeast, glucose as well as maltose was formed. Wr6blewski (Ber. d. d. chem. Gesellsch., 1897, xxx, 2128; Chem. Zeitung, 1898, xxii, 375), in the report of his investigations of soluble starch, calls attention to the difference between soluble starch and amylodextrin. He states that soluble starch is to be regarded as the first product of the hydrolysis of natural starch; that it can be filtered through porcelain; that it does not reduce Fehling's solution; that it is colored blue with iodine; and that amylodextrin is a derivative of soluble starch and has the property of slightly reducing Fehling's solution, and is colored reddish-brown with iotline. The constitution and oxidation products of maltodextrin were studied by Brown and ?^Iillar (Proc. Chem. Soc, 1899, xv, II), w^ho state that while considerable quantities of this substance are present when the decomposition of starch has been arrested at the proper 118 DIFFERENTIATION AND SPECIFICITY OP STARCHES. time, its preparation in a pure state in sufficient quantities is a laborious process. They give it the following rotatory and reducing values : (0)0 = +18r-183°, and 22 = 42 to 43; and they state that it is completely hydrolyzed by diastase to maltose, and by acid to rf-glucose. The relation of maltodextrin to isomaltose was studied by Pottevin (Ann. de ITnst. Pasteur, 1899, xiii, 728), who asserts that the maltodextrin of Brown and Morris is not a tnie compound, but simply a mixture of dextrin and maltose, as previously claimed by Lintner and Dull. By fractional precipitation with alcohol he split up this maltodextrin fraction, and he asserts that the differences between mixtures of dextrin and maltose and maltodextrin in their behavior towards diastase and alcohol, and as regards dialyza- bility, are readily explained on the basis of differences in the form of the dextrin in the combination. The different dextrins differed in digestibility, solubiUty in alcohol and dialyzability, and he found that an artificial mixtm-e of maltose with a properly selected dextrin is absolutely identical with Brown and Morris's maltodextrin, which therefore he holds is a mixture and not a true compound. In studies of the constitution of starch and its derivatives, Syniewski (Ann. d. Chem. u. Pharm., 1899, 282) states that starch as well as the products obtained from it, such as the amylose of Meyer, the amylodextrin of Lintner and Diill, and the granulose of Nageli and various investigators, consists of many molecules of amylogen which are joined together in the form of a carbinolanhydrid union (see page 133). When amylogen is hydrolyzed, he assumes, all of the maltose molecules are split off one after another, and dextrin remains. By the continued action of malt extract the dextrin is split into isomaltose and glucose, and the isomaltose finally into glucose. All the products directly obtamed from starch by hydrolysis he designates dextrin. Those dextrins which originate from starch by carbinol- hydrolysis, and which therefore do not reduce FehUng's solution, he terms amylodextrin. The dextrin which originates from amylodextrin after the splitting off of all the maltose molecules he terms grenzdexirin. All of the dextrins between amylodextrin and grenzdextrin (those from which maltose may be split off) he terms maltodextrin. The dextrins which originate from grenzdextrin by the splitting off of glucose he terms glucodextrins. In another research, Syniewski (Ann. d. Chem. u. Pharm., 1902, cccxxiv, 201) studied an iodine compound of amylodextrin. He also made a solution of starch by heating starch- paste in the autoclave at 140°, and then subjected the solution to the action of malt that had previously been heated to 78° in the presence of 0.1 per cent of formaldehyde for 216 hours. The starch was completely converted into a body which the author terms grenz- dextrin II, and which he states is identical with the maltodextrin of Brown and Morris, and with the a-maltodextrin of Ling and Baker, and with the aclu-oodextrin of Lintner. By hydrolyzing grenzdextrin II with fresh malt extract it was split into a compound corresponding to the formula C24H42O21 and having the following rotatory and reducing powers: ia)r, 200 = 172.17°, R=A2.1. This substance he names j-maltodextrin. This r-maltodextrin is in turn resolved into maltose and dextrinose (Lintner's isomaltose). Hale (Amer. Jour. Science, 1902, xiii, 379) states that amidulin seems to stand between starch and erytlu-odextrm in regard to the solubihty of its iodide and in the degree of digestibility with saliva. Ling (Jour. Fed. Inst. Brewing, 1903, ix, 446) suggests that it would be well to use the term maltodextrin instead of dextrin when referring to intermediate dextrinous products of diastatic activity, because this would prevent confusion with dextrin prepared by torre- faction. He refers to the fact that Brown, Morris and Millar have stated that their mal- todextrin is completely converted into maltose by diastase, but he finds by subjecting a-maltodextrin (having the constants {a)u 3.93 = +180.5, and i?3.93 = 36.7) to the action of diastase that there are formed 90 per cent of maltose and 10 per cent of glucose. In the process of preparation of soluble starch, Ford (Jour. Soc. Chem. Ind., 1904, xxiii, 414) notes that if the treatment by dilute acid be carried too far there is a very con- AMYLODEXTRIN AND MALTODEXTRIN. 119 siderable production of maltodextrin ; that all soluble starches prepared in the heat by acid contain this substance; and it is to its presence, and not to the maltose or glucose, that the copper-reducing power of soluble starches is due. This is proved, he states, by the fact that yeast has no effect in reducing the amount of apparent maltose in a solution of starch which on dialysis does not yield sugar but a body having the properties of amylo- dextrin. He prepared soluble starch from arrowToot by Lintner's method, added water, and dialyzed for 2 days. The dialysate was concentrated, and precipitated twice with alcohol, and finally dissolved in water. The rotatory and reducing powers correspond to {a)D aba = +189) 'ffab3 = 14.4, which are the values given by Brown and Morris for amylo- dextrin. For practical purposes of diastasimetry the presence of amylodextrin in the starch is of no importance, as it is equivalent to soluble starch. The degradation products of starch brought about by the action of oxalic acid were studied by Griiters (Zeit. angew. Chem., 1904, xviii, 1169), who obtained achroodextrins I and II, maltodextrin r, maltose, dextrose, and a small amount of levulose. He found that the products of the action of oxalic acid are the same as those of diastase, except that the maltodextrin r produced by the acid is replaced by maltodextrin /S by the diastase, the latter showing different constants and behaving differently in the presence of diastase. He beUeves, however, that both forms occur simultaneously, but in varying proportions. He also notes that the divergent behavior of different isomaltose preparations toward malt extract indicates an occasional preponderance of the more resistant maltodextrin r as the lowest member of the dextrins present. By the use of chromic acid, Harz (Beiheft. z. botan. Centralbl., 1905; Woch. f. Brau, 1905, XXII, 721) prepared what he records as amylodextrin, and he found that (like the original starch) it did not behave as a uniform substance, but seemed to consist of a number of molecular groups which differ in complexity and density of internal structure. The same composite character he noted in erythrodextrin II that he prepared by the action of a 5 per cent alcoholic solution of hyckochloric acid on starch. Wlien, however, the achroodextrin stage was reached the products seemed to possess homogeneity. Moreau (Ann. d. d. Soc. Roy. d. Sc. m^d. e. nat. d. Bruxelles, 1903, xii, 117; 1905, XIV, 64), in following up his earlier work on the isolation of various products of digestion by precipitants, separated the different products by precipitation by barium hydroxide in aqueous or dilute alcoholic media, determining the progress of the reaction and nature of the precipitated bodies by the iodine reactions of the filtrates. Amylodextrin and ery- throdextrin were precipitated in aqueous solutions, while achroodextrin and sugar were precipitated only in the presence of more or less alcohol; but the limits of precipitability were sufficiently far apart to permit the separation of the several products by repeated fractionation by barium hydi-oxide. By this means he determined, in support of Mittel- meyer's theory of starch decomposition, that even in the earliest stages of digestion the starch molecule is immecUately broken down into tln-ee forms of dextrin and sugar. De- tailed directions are given for preparing pure amylodextrin from the dextrins of com- merce. Such amylodextrin, as well as pure erythrodextrin, were found to be absolutely devoid of reducing power in cupric solutions. The maltodextrin r, which was first described by Griiters, was further studied by Rheinfeld (Woch. f. Brau., 1906, xxiii, 510), who finds that its position in the series of disintegration products of starch is between the maltodextrin fi of Ling and Baker, or the achi'oodextrin III of Prior, and maltose. He subjected starch-paste to the action of diastase at a temperature of 70°, and allowed the process to go on until a red reaction with iodine no longer occurred. By a series of fractionations he obtained a form of malto- dextrin which had the constants (a)D= +167.7 and R = G0.1 per cent of maltose. Based upon glucose equivalents the constants were (a)D= +170-173 and i? = 61-64 per cent. He calls attention to the fact that this maltodextrin was fermented to the extent of 50 120 DIFFERENTIATION AND SPECIFICITY OF STARCHES. per cent by Frohberry yeast, wliile the same form of dextrin prepared by Griiters by means of oxalic acid was fermented only to the extent of 24 per cent. In a subsequent communi- cation (Z. Spiritusind, 1907, xxx, 371) Rheinfeld noted that a certain amount of the prod- ucts of hj^drolysis undergo polymerization and condensation when a solution is repeatedly evaporated. He carried on hydrolysis until no color reaction with iodine was obtained, and by repeated fractionation he prepared 5 specimens of maltodextrin 7. The constants he records as («)d= +103-167 and R = 5S—()2. Gruters' values were (0)0 = +160 and i? = 61. When determined as Ost's glucose values, when corrected, the values were {a)o = + 170 - 173 and 72 = 61 -64. Reichard (Zeit. ges. Brau. 1908, xxxi, 161) confirmed Syniewski's statement in regard to the formation of a formaldehyde-amylodextrin compound by the action of concentrated formaldehj^de. He studied the influences of different percentages of formaldehyde in relation to consecutive changes in the starch, and also its influence upon the temperature of gelatinization ; a gram of starch was completely gelatinized by 10 c.c. of a 37 per cent unneutralized formaldehyde in 7 to 8 hours at 25°. With weaker or neutraUzed formalde- hyde, and at lower temperatures, the action is slower. At 15° to 16° the starch was gelatin- ized in 38 per cent formaldehyde in 2 days. The formaldehyde preparations give at first a blue reaction with iodine, but when the gelatinous stage is reached the reaction is brownish-red; and the gelatinized starch dissolved in water yields a yellow reaction, which indicates a further reaction of the formaldehyde. Castoro (Gas. chim. Ital., 1909, xxxix, 603) heated pea starch for 5 hours in a 2 per cent sulphuric acid, filtered, and precipitated with alcohol. Upon treating the precipitate with water one part was found to go into solution and another to remain undissolved. The latter gave a blue- violet reaction with iodine, and corresponded with the amylopectin of Maquenne and Roux. Upon dialyzing the part in solution two fractions could be obtained, one giving a blue reaction with iodine and the other a violet reaction corresponding to that of ainylo- dextrin. By dialyzing a pseudo-solution of potato starch prepared by the agency of 1 per cent sulphuric acid, a diffusible portion was obtained that gave a violet reaction with iodine. He believes that the differences in the color reaction are due to differences in the size of the particles, the large particles becoming blue and the small particles violet. It is pointed out that analogous differences may be observed in gold-colloidal solutions. ERYTHROUEXTRIN, ACHROODEXTRIX, GRENZUEXTRIN, ETC. Although Vaquelin (BuU. de pharm., 1811, iii, 54) noted that when starch is subjected to torrefaction it is converted into a gum-like substance, and Vogel (Schweigger's Jour., 1812, V, 80) that starch is changed by weak acid into gum and sugar, and various experi- menters of the following twenty years that starch yields a gummy substance, it remained for Biot and Persoz (Ann. de chim. et phys., 1833, lii, 72) to demonstrate the distinguishing characters of this substance, and to give it the name by which it has continued to be known even to the present day. They believed that it existed as a constituent of the grain, and that it is liberated by boiling in water or by weak solution of acid, which disrupts the outer coating of the grain. From its strong dextro-rotatory action on rays of plane polarized light they named it dextrin. The same year, Payen and Persoz (Ann. de chim. et phys., 1833, liii, 73) prepared dextrin by the aid of diastase. They ascertained in these experiments that a number of substances were present in the preparations — one of them was insoluble in cold water, but soluble in hot water, and colored with iodine, and identical with the substance of the interior of the starch-grains; a second substance, which is soluble in both cold and hot water and in weak alcohol, but not colored Avith iodine, and in the nature of a gum; and a tliird substance, sugar, etc. This second substance is the body they named dextrin, and corresponds with the achroo- dextrin of the present. ERYTHRODEXTRIN, ACHROODEXTRIN, GRENZDEXTRIN, ETC. 121 Two years later, Payen (Ann. de cliini. et phys., 1836, lxi, 355 and lxv, 225) reported that the rotatory power of dextrin was equal to that of starch; that all starches have the same elementary composition (CgHio0.5) ; that all parts of the same grain,including dextrin, have the same elementary composition; and that the dextrins formed by dilute acid, diastase, and torrefaction are merely physical modifications of the same substance. Fiirstenberg (Ann. d. Chem. u. Pharm., 1844, lii, 417) dicovered that starches of cereals contain dextrin similar to that obtained by the action of dilute acid, or by diastase on starch, and that it is without reducing action on copper solutions. Blondeau de Carolles (.Jour. f. prakt. Chemie, 1844, 33, 439), Fehling (Ann. d. Chem. u. Pharm., 1845, lv, 13), and Kalinowsky (Jour. f. prakt. Chemie, 1845, xxxv, 193) found that by the action of sulphuric acid a number of compounds are formed, according to the time of action and strength of acid. Bechamp (Compt. rend., 185G, xlii, 1210) noted that a substance (erythrodextrin) may be formed that is intermediate between starch and dextrin (achroodextrin), giving a reddish coloration with iodine, and which he regards as an inversion product of starch. The assertion of Payen that the dextrins formed by the action of dilute acid, diastase, and torrefaction, respectively, are merely physical modifications of the same substance was strongly opposed by Mulder (Chemie des Bieres, Leipzig, 1858, 166), who asserts that they differ from one another in their reactions, especially with precipitants. In 1860, Musculus (Compt. rend., 1860, l, 785) refers to a gummy body formed by the action of diastase or acid wliich does not give a color reaction with iodine, which he did not isolate, but records as dextrin. He believed that the sugar formed during digestion is dextrose; that the sugar is not formed from dextrin, as was at that time and for many years believed, but that dextrin and sugar are formed coincidently from starch by the action of water, the proportion being 2 of dextrin to 1 of sugar. The following year he reasserted the proportion of dextrin to sugar, and also noted that this ratio exists as soon as the blue reaction with iodine ceases. He believed that the sugar prevents further change in the dextrin, but that at the same time it does not prevent the breaking down of starch. In a later article, JVIusculus (Ann. de chim. et phys., 1865, vi, 177) studied the character- istics of what he refers to as a "true dextrin," which is a body (achroodextrin) that does not give a color reaction with iodine, does not reduce copper solutions, and is not digested by diastase. In further studies (Compt. rend., 1869, lxviii, 267) Musculus gives a ratio of 1 : 1 and states that during the transition of starch by diastase and acid to "colorless dextrin" (non-color reacting dextrin) there is produced a modification, which he calls "insoluble dextrin," which is insoluble in cold water, which gives a violet to a wine-red brown or yellow reaction with iodine, is saccharified by diastase, and yields by acid less sugar than starch. The assertion of Musculus that the colorless dextrin is not digested by diastase was opposed by Payen (Ann. de chim. et phys., 1866, vii, 382), who to the contrary found it to be readily sacchai'ified. Nasse (Demateriis amtlaceis num in mammalium inveniantur disquisitio. Diss. Halle, 1866) reported a body under the name of " dextrinogen " which corresponds with the so-called colorless dextrin of Musculus. Three stages must be recognized, writes Griessmayer (Ann. d. Chem. u. Pharm., 1871, CLX, 40), in the process by which starch wliich has been boiled and set aside is finally transformed into glucose and other products: (1) After a few days the addition of a small amount of weak iodine gives a violet reaction, never a blue. The red of the iodine-dextrin reaction combining with the blue of the iodine-starch reaction makes violet. (2) After standing about 8 days the addition of iodine gives a red reaction, which is attributed to a substance he names dextrin I. By carefully adding a moderately concentrated solution of tannic acid, this dextrin can be precipitated, leaving the starch. Besides dextrin I, there is another dextrin present which does not give a color reaction with iodine. (3) After starch has stood for more than 8 days there occurs a time when the addition of iodine causes a red coloration which disappears izmnediately, but testing with Fehhng's solution 122 DIFFERENTIATION AND SPECIFICITY OF STARCHES. gives no sugar reaction. Later, the iodine will not yield a color reaction, while Fehling's solution is reduced. There is in solution, then, a substance that has great affinity for iodine, greater than dextrin I, which has lost its dextrin color-reaction characteristics towards iodine, yet without having acquired the distinguishing marks of sugar. Tliis dextrin he names dex- trin II. He believed that both dextrins coexist in fresh starch solution in small quantities. O'Sullivan (Jour. Chem. Soc. Trans., 1872, 579) also noted the existence of two dextrins, which he distinguished as a- and /3-dextrins, respectively, and which correspond with the dextrin I and dextrin II of Griessmayer. O'SulUvan found that both forms are changed into maltose by diastase or dilute acid, and that both have the same rotatory power, (a)j = +213. He prepared them almost free of reducing power, and suggests that if pure they would not reduce copper solutions. There appeared at the same time a very elaborate contribution by Briicke (Sitz. d. k. Akad. d. Wissensch., Wein, 1872, lxv, 3 Abth., 126) on the manner of digestion and assimilation of carbohydrates. He noted two dextrins, corresponding with those recorded by Griessmayer and O'Sullivan. The dextrin which gives a red reaction with iodine he termed erythrodexirin, and the one which did not give a color reaction he termed achroodcxtrin, which terms have continued in uni- versal use to the present. Besides these dextrins there was found a residue which Briicke named erythamylum, which he states consists of Niigeli's cellulose together with a substance which takes a red stain with iodine, which is already present in dry starch, but which is masked by the granulose and its blue reaction with iodine. A study of amylodextrin, together with a comparison of the properties of starch, amylodextrin, and dextrin, was made by W. Niigeli (Beitrage z. niiheren Kenntnis d. Starkegruppe, etc., 1874, Leipzig), in wliich he refers to the existence of two forms of dextrin, one of which is colored red or orange and the other yeUow with iodine. He asserts that there is no dextrin which does not become colored with iodine, and also that Mus- culus's theory of the breaking up of starch coincidently into dextrin and sugar does not hold. Three dextrins, a, /3, and r, were described by Bondonneau (Compt. rend., 1875, Lxxxi, 972, 1210). The a-dextrin corresponds with the dextrin I of Griessmayer, ery- throdextrin of Briicke, and the a-dextrin of O'Sullivan; while the |8-dextrin corresponds with dextrin II of Griessmayer, the achroodextrin of Briicke, and the /3-dextrin of O'Sulli- van. The next year, 0-Sullivan (Jour. Chem. Soc, 1878, i, 479) gave good evidence to show that the r-dextrin of Bondonneau does not exist, and that the latter was led into error because of the assumption that the sugar product is dextrose, and that the r-dextrin is merely an expression, not recognized by Bondonneau, of the difference between the reducing powers of maltose and dextrose. Nasse (Acliiv f. ges. Physiologic, 1877, xiv, 473) beUeved by the action of saliva or dilute sulphuric acid that there are formed dcx- trinogen (achroodextrin) and also a pecuUar form of sugar that is not grape sugar, and which he i^roposes to distinguish by the name amylum ptyalose. Thus far it had been clearly shown by a number of investigators that two very dififerent forms of dextrin are produced during the saccharification of starch, but Musculus and Gruber (Zeit. f. physiolog. Chemie, 1878, ii, 177) were the first to give evidence to lead to the belief that the so-called achroodextrin is not an individual, but a mixture of at least three achroodextrins which have different optical and reducing powers. Based upon the results of their research, they formulated a definite theory of the processes and products of starch saccharification, which with certain modifications has received a very general acceptance from that time to the present. According to these authors the derivatives of starch by the action of diastase or weak sulphuric acid are as follows : (1) Soluble starch. [Probably a peculiar form of starch or of erythrodextrin.] It is insoluble in cold water, and soluble in water at 50° to G0°. An aqueous solution gives a wine- red reaction with iodine. Dried in the air it will give a violet, j-ellow, or bro\vn with an excess of iodine. Its specific rotatory power is (a) = +218, and its reducing power 6. ERYTHRODEXTRIN, ACHROODEXTRIN, GRENZDEXTRTN, ETC. 123 (2) Erythrodextrin. It differs from starch in that it is soluble in cokl water, is not gran- uhir, and because in solid form or in solution it gives only a red color with iodine. Both soluble starch autl erythrodextrin they found to be readily affected by diastase. They did not succeed in obtaining pure erythrodextrin. (3) Achroudexfrin a. This does not give a color reaction with iodine. It is more easily converted into sugar by diastase than either soluble starch or erythrodextrin. Its rotatory power is (a) = +210, and its reducing power 12. (4) Achroodextrin /?. It is unaffected by diastase. Its rotatory power is (a) = +190, and its reducing power 12. (5) Achroodextrin y. It also is unaffecteil l)y diastase. Its rotatory power is (a) = +150, and its reducing power 28. (G) Maltose. Formula, C12H22O11+H2O. Rotatory power (tt) = +150, and its reduchig power 66. It is not affected by diastase. (7) Glucose. Formula C6H12O6+H2O. Its rotatory power is (o) = +56, and its reducing power 100. It does not undergo fermentation. The figures for the rotatory and reducing powers are stated to be only approximate, but they show a decrease of the former and an increase of the latter as decomposition proceeds, with the formation of substances of less molecular weight. Musculus and Gruber write that starch, before it appears in the form glucose, is changed into 5 isomerous bodies, i.e., erytlu'odextrin, achroodextrin a, achi-oodextrin (3, achroodextrin r, and maltose. They regard the starch substance as having the formula ?i (CioHooOio), in which nhas a value of not less than 5 or 6. Starch by absorption of water, by the addition of diastase or dilute acids, undergoes repeated splitting. At each subsequent splitting there appears besides maltose a new dextrin of less molecular weight than the preceding dextrin, that is, n becomes smaller at every stage until achroodextrin ;' results. The latter, through simple absorption of water, goes over into maltose, and by hydration and splitting this goes into 2 molecules of glucose. In 1879, Brown and Heron (Proc. Chem. Soc. Trans., 1879, xli, 596; Ann. d. Chem. u. Pharm., 1879, cxcix, 241), while accepting the theory of Musculus and Gruber of the breaking down of the starch-molecule by a series of hydrations and splitting-up processes, stated their belief that the starch- molecule can not have a simpler formula Table 7. than 10 (Ci2HooOxo) and that the first ac- tion of diastase is to separate by hydration one of these 10 groups, which is trans- formed into maltose, while the remaining 9 groups constitute the erythrodextrin a, or the first of the series of dextrins. By the addition of more water this dextrin is con- ceived to be spUt into maltose and another dextrin, erytln'odextrin /S, which consists of 8 groups. This in turn by hydration is split into maltose and another dextrin, consist- ing of 7 groups, and designated achroo- dextrin a, and so on by consecutive splitting and hydration until there occurs ultimately a complete conversion into maltose. According to this theory the number of dextrins is determined by the number of constituent molecules in the starch-molecule. There are therefore 8 possible dextrins, 2 erythrodextrins and 6 achroodextrins, derivable from a starch-molecule having the formula as above stated. As the action proceeds, the rotatory power falls while the reducing power rises, until finally both powers correspond to the prop- erties of maltose, as indicated in table 7. (See Brown and Morris, page 124.) The results recorded by Brown and Heron received sujjport in the investigations of Squu-e (Jour. Soc. Chem. Industry, 1884, iii, 397). He states that his experiments confirm No. of transformation. Specific rotation. CuzO reduced. Resulting dextrin. Soluble starch 1 2 3 216.0 209.0 202.2 195.4 188.7 182.1 175.6 169.0 162.6 156.3 150.0 0 6.4 12.7 18.9 25.2 31.3 37.3 43.3 55.1 55.1 61.0 Erythrodextrin a Erythrodextrin /3 Aciiroodextrin a Achroodextrin ^ Aciiroodextrin 7 Achroodextrin 1' Achroodextrin e Achroodex-trin C Achroodextrin '/ 4 5 6 7 8 9 . . . . Maltose 124 DIFFERENTIATION AND SPECIFICITY OF STARCHES. the results of these investigators as far as reaction No. 8 is concerned, namely, that at 60° or 63°, or even below, gelatinized starch undergoes a definite process of hydration which results in the production of practically 80 per cent maltose and 20 per cent dextrin. He states, however, that he did not succeed in obtaining trustworthy evidence of definite reactions representing the other numbers, nor sufficiently concordant results to warrant the assumption that at any given temperature the reaction can always be represented by definite formulae. The production of maltose he found to be continuous, and not by steps, and that there was never a complete disappearance of dextrin. Squire quotes Southy (Brewing, etc., London, 1877) as having noted that at lower temperatures the proportion of dextrin to sugar is increased. Almost coincidently with the appearance of Brown and Heron's article, O'Sullivan (Proc. Chem. Soc. Trans., 1879, xli, 770) reported that he had modified his views regarding the number of dextrins, he having now reached the conclusion that there are formed 1 erythrodextrin and 3 achroodextrins. He states that he separated them in a nearly pure state by means of precipitation with alcohol, and he designates them as follows : a-dextrin colored brownish-red with iodine; (3-dextrin I, /3-dextrin II, and /3-dextrin III, none of which gives a color reaction with iodine. O'Sullivan beUeved that the formation of dextrins is not due to the breaking down of the starch-molecule into simpler bodies, but to a rearrangement of the molecules, so that there are formed a series of substances of the same molecular weight, but differing in their behavior, owing to the change in con- stitution. He thought, therefore, that the entire starch-molecule is affected at once, in contradistinction to the serial degradation according to the theories of Musculus and Gruber and of Brown and Heron. Still another theory of the processes of saccharification was offered at this time by Herzfeld (Ueber Maltodextrin, Dissertation, Halle, 1879), according to which the trans- formation of the starch by diastase is by serial actions, giving rise consecutively to amylo- dextrin, erythrodextrin, and achroodextrin, the latter being split into maltodextrin and maltose. Below 65° it is stated that achroodextrin and maltose are formed, but at higher temperatures also erythrodextrin and maltodextrin. A process for the manufacture of dextrin was devised by Lauga (Jour. Soc. Chem. Industry, 1882, i, 513), which consists essentially of boihng glucose juice with concentrated phosphoric acid until dextrinization, at which time the preparation is cooled to about 50°, neutralized, and filtered under pressure through animal charcoal, etc. The products of the conversion of starch by organic and inorganic acids were examined by Salomon (Jour. f. prakt. Chemie, 1883, xxviii, 82), who concluded, contrary to other investigators at this time, that only one kind of dextrin results in the saccharification of starch. He took some very pure dextrin that was obtained from a solution from which starch had been removed. The solution of this dextrin gave a brownish-red reaction with iodine and, after purification by resolution in water and precipitation by alcohol, it gave a brownish reaction that disappeared immediately, and it failed to reduce Fehling's solu- tion. He also held that only dextrin and glucose are produced, no maltose. Schulze (Jour. f. prakt. Chemie, 1883, xxviii, 311) studied the influence of acetic acid on starch by treating starch with 20 per cent acid under pressure, and obtained, as he states, a dextrin like the a-dextrin of Bondonneau. Heating for 4 hours gave rise almost exclusively to this dextrin, but further heating caused the formation of more or less glucose. In 1885, Brown and Morris (Ann. d. Chem. u. Pharm., 1885, xxiii, 72; Jour. Chem. Soc, 1885, XLVii, 527) published what they state is to be looked upon as a continuation of the paper by Brown and Heron that appeared in 1879 (page 123). By this extended work they were enabled to add certain facts tending, as they believe, to a better knowledge of the products of the diastatic digestion of starch. The chief results of their investigation may be summarized as follows: ERYTHRODEXTRIN, ACHROODEXTRIN, GRENZDEXTRIN, ETC. 125 (1) When starch-paste is acted upon I)y malt extract above 40° the specific rotatory and reducing powers of the products indicate the existence of maltose and a non-reducing dextrin. (2) If the products of transformation arc fractionally precipitated by alcohol, the composition of the several fractions, as showoi by the optical and reducing powers, is capable of inter- pretation on the supposition that they consist of maltose and a non-reducing dextrin. (3) The authors look upon the foregoing as establishing a criterion of purity for any separated portion of the transformation products, and they ascribe any apparent departures from this rule, as regards substances described by other investigators, as due to im])uritics or to errors in determination. (4) The tendency of all starch transformations when subsequently acted upon by malt extract at 50° or 60° is rapidly to attain a state of equilibrium, which would correspond to the tenth equation Ci2HooOn=8H20 = 8Ci2H220ii+2Ci2H2oOio, corresponding to equation 8 of the series sho\vn in table 8, in which the theoretic amount of maltose contained in 100 parts of dextrins is given at each stage of conversion of starch down to No. 8. When all the products derived from starch are taken together, the reac- tion corresponds to an optical activity of (a), 3.86= +162.6, Table 8. and to /V3.86 = 49.3, and to a percentage composition of mal- tose 80.9 and dextrin 19.1. (5) The degradation of all of the higher transformations do^v^a to this point is due to the hy- drolysis of the more complex polymeric dextrins and malto- dextrin. (6) The dextrin and maltodextrin can be submitted to solution in water, to evaporation, and to frequent precipitation with alcohol -ivithout being hydro- lyzed. (7) This is proved by the fact that the transformation products after such treatment, when subsequently fractionated with alcohol, yield fractions with a mean value of («)j and K, coinciding with that of the original solution and also ))y the No. 8 equation. (8) The separated dextrins have not, as stated by 0 'Sullivan, different properties as re- gards their liehavior with malt extract from those they possess when in solution with the other transformation products simultaneously with them. (9) It follows that it is possible, by submitting a separated dextrin or a mixture of dex- trins to the action of malt extract at 50° to 60°, and determining the percentage of maltose they yield, to ascertain the actual or mean position of the dextrin in the poly- meric series — the actual position if it be homogeneous, or the average or mean posi- tion if it be a mixture. (10) By making use of this process it is possible to ascertain by examination of residual products of a beer or similar liquid, after primary fermentation is concluded, the values of (a)j and K of the original starch-products as transformed by the mashing process. (11) With the exception of No. 8 they alwaj's found, on fractionating such preparations with alcohol, that the dextrin is not homogeneous, but belongs partly to a higher and partly to a lower equation. (12) It follows from No. 11 that the whole of the starch-product in a transformation is not simultaneously affected, but that some portions are hydrolyzed faster than others and that a sharp line can not be drawn between the equations higher than No. 8. (13) The i^reparation of absolutely non-reducing dextrins is impossible by the mere precipi- tation with alcohol, even when aided by fermentation, but this can be accomplished by treating the dextrin with alkaline mercuric cyanide. (14) The dextrins are not directly fermentable by yeast, but require first to be hydrolyzed. (15) When the action of malt extract on starch-paste is limited there is always found, among the products of transformation, besides maltose and dextrin, a third body which has optical and reducing properties corresponding to an apparent composition of 34.6 per cent of maltose and 65.4 per cent of dextrin. Number of the stage of conversion. Constan combined (a) J 3.86 ta of the products. A' 3. 86 Maltose obtained in 100 of eacli of the dextrins down to No. S Soluble starch 1 0 216 209 202.2 195.4 188.7 182.1 175.6 169.0 162.6 '6!4 12.7 18.9 25.2 31.3 37.3 43.3 49.3 84.44 82.09 79.20 75.39 70.37 63.33 52.77 35.18 00.00 3 4 5 6 7 8 126 DIFFERENTIATION AND SPECIFICITY OF STARCHES. (16) This body is doubtless the same as that prepared in an impure state by Herzfeld, and described by him as maltodextrin. (17) Maltodextrin, they hold, is not a mixture of maltose and dextrin, as is proved by a number of facts. (18) While maltodextrin is unfermentablc by yeast it is converted into fermentable, crystallizable maltose by malt extract and by certain forms of Saccharomyces. (19) They believe that maltodextrin is not a mere hydration product of achroodextrin, but that it is produced from starch and the polymeric dextrins by the fixation of a molecule of water upon the ternary group (Ci2H2oOio)3 (of which there can not be less than 5 in the starch molecule) which results in the separation from the dextrin resi- due maltodextrin < (c^-^^^o^^i ■ This by fixation of two more molecules of water gives rise to a freely fermentable, crystallizable maltose. In 1889, Brown and Morris (Jour. Chem. Soc. Trans., 1889, lv, 449 and 462) reported the results of further investigations of the constitution of the starch-molecule, and also of the products of transformation. While adhering to their theory of the peculiarities of the groups of the molecule, they believe that the number of groups is distinctly larger than suggested in their previous work. In the first of these articles they study the amylodextrin of W. Nageli in relation to soluble starch, and the relation of amylodextrin to maltodextrin (see pages 115 and 116). In the second article they report their determinations of the molecular weights of carbohydi'ates, and state that the following hypothesis seems to them to be in accord with the facts : The starch-molecule may be pictiu-ed as consisting of 4 complex amylin groups arranged around a fifth similar group which constitutes a molecular nucleus. The first action of hydrolysis by diastase is to break up tliis complex and to liberate all 5 amylin groups ; 4 of these groups when liberated are capable, by successive hydrolyzations through maltodextrins, of being rapidly converted into maltose, while the central amylin nucleus, by closing up the molecule, withstands the influence of hydrolyzing agents and constitutes the stable dextrin of the low equation, which, as is known, is so slowly acted upon by subsequent treatment with diastase. The 4 readily hydrolyzable amylin groups are looked upon as of equal value and in their original state to constitute the so-called high dextrins, which can never be separated completely from the low dextrin by any ordinary means of fractionation. This hypothesis provides for intermediate maltodextrins and amylodextrins whose number is only limited by the size of the original amylin group. Each amylin group of the 5 has the formula of (C12H20O 10)20 ^^^^^ ^ molecular weight of 6,480, so that the entii-e starch-molecule, or, more correctly speaking, that of soluble starch, is represented by 5(Ci2H2oC)io)20) having a molecular weight of 32,400. They state that the dextrins are metameric and not polymeric compounds, as had already been suggested by O'SuUivan. A special study of dextrins, and with reference to the dextrin products of both enzymic and acid action, was made by Effront (Moniteur Scientifique, 1889, 513). He notes that dextrin can be obtained pure by destroying the sugar present by lactic acid fermentation ; and for the purpose of the determination of the sugar and dextrin he would destroy the sugars by ammonium hydroxide and sodium hypochlorite, and determine the dextrin by the polariscopic readings before and after the sugar destruction. (See also page 150.) Improved processes for preparing dextrin were devised by Schumann (Jour. Soc. Chem. Industry, 1888, vii, 335; 1889, viii, 295). By the first process 1 per cent of fixed acid is agitated and allowed to stand for 24 hours with cold starch in milky form. The water is then separated from the precipitated starch, and the latter is washed with fresh water until free of acid. The washed starch is once more reduced by water to a milky state at 15° Baume, and then boUed under pressure of 3 to 4 atmospheres with 0.5 per cent of saturated sulphurous acid solution until the fu-st trace of glucose can be detected in the product. The reaction is then stopped, the shght trace of sulphuric acid formed is fixed, and the syrup is filtered through animal charcoal and evaporated. In the second process ERYTHRODEXTRIN, ACHROODEXTRIN, GRENZDEXTRIN, ETC. 127 the starch is mixed to a thick cream with cold water and then treated with 1 per cent of its weight of sulphuric, hydrochloric, or nitric acid for 24 hours. The preparation is then washed free from acid. This prepared starch is either dried or again mixed with water to a cream, and heated to 160° to 170° in an oil bath, or by means of superheated steam, until all the starch is con^•erted. The solution is then refined. The percentages of dextrin, maltose, starch, etc., in commercial dextrin were studied by Hanofsky (Mittheil. d. k. k. Tech. Gew.-Museums, 1889, 56), the essential results being given in table 9. Table 9. Sample No. Maltose. Dextrin. Starch. Moisture. Ash. other organic matters. Acidity of 100 grams in c.c. of decinormal potash solution. 1 2 3 Per cent. Per cent. 4.25 47.78 10.90 30.75 3.75 29.46 Per cent. Per cent. 35.55 10.11 43.20 7.02 58.00 1 6.85 1 Per cent. 0.27 0.39 0.60 Per cent. 2.04 1.74 1.34 40.0 26.6 25.3 Contrary to the findings of a number of investigators mentioned in the preceding pages, Flourens (Compt. rend., 1890, ex, 1204), from experiments on the products of the saccharification of starch with acids, was led to the conclusion, from the entire agreement of the rotatory and reducing powers at the various stages of the process, that only one and not several dextrins are formed; and also glucose, but not maltose. In an article on the chemistry of starch and the nature of dextrins, Scheibler and Mit- telmeier (Ber. d. d. chem. Gesellsch., 1890, xxin, 3060), it is stated that dextrin could be readily purified by repeated precipitation with alcohol, or by osmosis. The dextrin thus obtained they look upon as a mixture of several dextrins. They prepared the hydrosazone by dissohdng the dextrin in the cold in the presence of phenylhydrazine. The hydrosazone is hydrolyzable by diastase, and is quite soluble in water. WTien a cold solution of dextrin was treated ■ndth sodium amalgam a product can be precipitated by alcohol which does not reduce Fehling's solution, nor dissolve in phenylhydrazine. This body, which appears in the form of a wliite powder, they named dextrite. It is saccharified by diastase or strong acids; it is not precipitated by basic acetate of lead or by lime-water; it does not reduce Fehling's solution, but acquires reducing power by treatment with diastase or dilute acid; its solution reddens litmus; it decomposes calcium carbonate; and it is soluble in phenyl- hydrazine. They call attention to the fact that a number of observers have made use of oxidizing agents to destroy sugars in the preparation of dextrins upon the erroneous assump- tion that the dextrin itself is not affected. The non-reducing dextrin obtained in this way they believe is probably a carboxylic acid of dextrin. They also state, in support of some observers and contrary to others, that pure dextrin reduces Fehling's solution. Leitner (Zeits. f. angew. Chem., 1890, 546) noted that permanganate brings about a conversion of starch into gummy substances which he states differ from dextrins by their acid reaction and by their yielding precipitates with basic lead acetate and barium hydroxide. Glycerine has been employed to prepare soluble starch by restricting its action, and by further action to convert the soluble starch into dextrins or dextrin-hke bodies. Zul- kowski (Ber. d. d. chem. Gesellsch., 1891, xxiii, 3295) subjected starch to hot glycerol, and found that it was soon completely broken down into erytlu-odextrin, achroodextrin, and a number of bodies of increasing solubility as the reaction proceeded. These various substances were separated by means of their different solubilities, but it was found difficult to remove the last traces of glycerine. In a later investigation by Zulkowsky in associa- tion with Franz (Ber. d. osterr. Gesellsch. z. Forderung chem. Ind., 1894, xvi, 120), 128 DIFFERENTIATION AND SPECIFICITY OF STARCHES. Zulkowsky's previous work was confirmed and extended. It was found, by heating starch-paste to 190°, that soluble starch was formed having a specific rotation of 188.3; at 200° crythrodextrin appeared. It gives a cherry-red reaction with iodine, and had a specific rotation of 181. At 210° another dextrin is formed, wliich is recorded as giving a brownish-yellow reaction with iodine and having a specific rotatory power of 173.5°. Both dextrins were found to reduce Fehling's solution. Wlien the preparation was heated at 210° for a long time two other products were formed, one of which resembled gum arable. Ost (Chem. Zeit., 1895, xix, 1501) prepared the dextrinous product according to Zul- kowski's process and found a glucose value of 90 per cent. He concluded that this substance is not a true dextrin. Bacillus amylobacter was used by ViUiers (Compt. rend., 1891, cxii, 435, 536) to con- vert starch into dextrins. Potato starch was made into a paste with tap-water; the paste put in a flask which it nearly filled; the starch inoculated at 100°; the flask closed with steriMzed cotton, and placed in an oven at 40°. Usually liquefaction had occurred by the end of 24 hours. After 2 to 4 days or more a red reaction with iodine was obtained. The preparation had the odor of butyric acid, of which it contained about 0.3 part for each 100 of starch. The chief product was dextrins. He states that it remains to be shown whether or not these dextrins are identical with those formed by the action of diastase or acid, and that it is remarkable that they are formed without an accompanying formation of either maltose or glucose. Villiers prepared the dextrins by filtration, drjang, resolution, and precipitation with alcohol. They occur in the form of a white, friable, hygroscopic mass, which consists of several dextrins having different rotatory and reducing powers, and different reactions with iodine. The dextrin having the highest rotatory power has the lowest reducing power and gives a red reaction with iodine, while those having lower rotatory powers and higher reducing powers do not give a color reaction with iodine. In Villiers's second article he states that besides the dextrins there is formed about 0.3 per cent of a carbohydrate that is dissolved in the alcohol used to precipitate the dextrins; that this is somewhat soluble in hot water and that it may be deposited in the form of fine crystals having the composition of C12H20O10+3H2O. It is dextro-rotatory, (0)0 = -f 159.4°; it is unfermentable; it does not reduce Fehling's solution; and it does not form an osazone. Another product had the properties of cellulose, and was converted into glucose by warm, dilute mineral acids. Dextrin-like products were obtained by Berge (Jour. Soc. Chem. Industry, 1892, XI, 448) by subjecting starch and various amylaceous substances in a dry state to the action of sulphm'ous anhydride. The substances are placed in a closed vessel into which sulphurous anhydride is introduced until all of the air is expelled, and the vessel then heated on an oil bath to a temperature of 120° to 190°, depending upon the nature of the product desu'ed. Upon the completion of the conversion the gas is perinitted to escape. As far back as 1846 Magendie showed that the blood has the power of converting starch into sugar, and more or less attention was given to this property by Claude Bernard, Hensen, Schiff, von Wittich, and other investigators. In 1892 Bial, under the guidance of Rohmann, and also Rohmann himself, studied the diastatic actions of the sera of the blood and lymph. Bial (Archiv f. ges. Physiologic, 1892, lii, 137; 1893, 156) found that these sera reduce starch to dextrin, maltose, and glucose. Rohmann (Archiv f. ges. Physiologie, 1892, LII, 157; Ber. d. d. chem. Gesellsch., 1892, xxv, 3652) subjected 100 grams of potato starch converted by heat into a paste with 5 liters of water, after cooling, to the action of 1 liter of bullock's serum, to which was added 100 c.c. of a 10 per cent alcohoHc solution of thymol to pre\-ent bacterial action. This preparation was kept at 32° for 24 hours, at the end of which time there were present dextrin, giving a brown reaction with iodine, acliroo- dextrin, glucose, and possibly maltose. After 10 hours' action a mixture of dextrin was present (which Rohmann calls porphyrodextrin), which gave a brown reaction with iodine, ERYTHRODEXTRtN, ACHROODEXTRIN, GRENZDEXTRIN, ETC. 129 together with some soluble starch. He states that mixtures of this body and soluble starch give iodine reactions ransinp; from a reddish-blue to a bluish-red, and that the so-called erytlirodcxtrin is not an individual, but a mixture of porphj'rodextrin and soluble starch. Schifferer (Inaug. Diss., Kiel, 1S92) unsuccessfully attempted the preparation of the so-called maltodextrins of Brown and Morris, but in no case did he succeed in obtaining any substance that had anj' resemblance to the amyloins of these authors (page 115); but he obtained dextrins, isomaltose, and maltose. Dextrins, he records, are not ferment- able, but they reduce Fehling's solution, and he believes that their number is not more than 2, and probably only 1 (achroodextrin), which does not give a color reaction with iodine. He thinks that other dextrins are mixtures of this with varying proportions of soluble starch, but his experiments failed to conclusively demonstrate this point. Rohmann had found that such mbctures always give a blue reaction, and therefore that erythrodextrin could not be made up of such a mixture (see above). Lintner and Diill (Ber. d. d. chem. Gesellsch., 1893, xxvi, 25-33) prepared amylo- dextrin, erythrodextrin, achroodextrin, isomaltose, and maltose. They hold that the theory of Brown and JMorris of the breaking down of amylin groups is untenable, and they con- tend that the process is serial, amylodextrin being the first product, this being split into erytlu"odextrin, this in turn into achroodextrin, and this into isomaltose, and this finally into maltose. Amylodextrin (Ci2H2oOio)54 is precipitated, they found, by alcohol in the form of a wliite powder, and that it can be obtained from a 20 to 30 per cent aqueous solution in the form of spherocrystals. It is easily soluble in hot water. Even a 10 per cent solution does not reduce Fehling's solution. Iodine gives with it a blue reaction. Its rotatory power is (a)D = +196. Amylodextrin, they state, is broken up into 3 mole- cules of erythrodextrin (CioH2oOio)54+3H20 = 3(Ci2H2oOio)i7.(Ci2H220ii). This is easily soluble in water, and hardly soluble in hot 50 per cent alcohol. It separates from a hot solution in dilute alcohol in the form of spherocrystals. It reduces Fehling's solution, and gives a reddish-brown reaction with iodine. The rotatory power is given as (a)D = +196. Erythrodextrin is in turn broken down into 3 molecules of achroodextrin (3(Ci2HooOio)i7.Ci2H220n)+6H20 = (Ci2H2oOio)5.Ci2H220ii). It is easily soluble in water, and almost insoluble in 70 per cent alcohol. Spherocrystals may also be obtained. Fehling's solution is reduced, R = 10. No color reaction occurs with iodine. The rotatory power is (0)0 = +192. It has a sliglitly sweetish taste. Achroodextrin is broken down into isomaltose according to the following reaction : 9[(Ci2H2oOio)5.Ci2H220n)] -I-45H2O = 54 Ci2H220n- It was not obtained in spherocrystals. It is readily soluble in water, and soluble in SO per cent alcohol and methyl alcohol, but insoluble in 95 per cent hot alcohol. The latter will dissolve 5 per cent of maltose. It tastes sweet. Concentrated solutions heated in water-bath turn yellow. Its rotatory power is (0)0 = +140, and its reducing power, R = 80. It ferments with yeast and malt with difficulty, and is convertible into maltose. The melting-point of its osazone is 150° to 153°. The four stages of decomposition, Lintner and Dlill hold, go on simultaneously, the energy of the diastatic process decreasing so that at a definite stage of the reactions, even at favorable temperatures, i.e., when two-tliirds of the achroodextrin has been converted into maltose, no more maltose is produced. They believe that the stoppage of the conver- sion of dextrin into sugar at a temperature of 70° is owing to the formation of a modified form of achroodextrin that withstands diastatic action. This has, however, been explained by Pottevin, Effront, and others by a modification of the properties of the enzyme when heated to this temperature. They hold that the amyloins, or maltodextrins, of Brown and Morris, etc., are merely mixtures of dextrin and isomaltose, and sometimes really identical with the latter, and that dextrins as well as starch are composed of isomaltose groups. In a later contribution (Zeitschr. f. Brau., 1894, xvii, 339) they describe a form of achroodextrin (achroodextrin II) which is similar to the maltodextrin 9 130 DIFFERENTIATION AND SPECIFICITY OF STARCHES. of Brown and Morris, which has the formula (Ci2H2oOio)3+H20, a specific rotatory power of (a)D = +183, and a reducing power, R = 2G.5 to 26.8. These figures do not agree with those of Brown and Morris. Conmiercial amyloins were examined by Hiepe (The Country Brewer's Gazette, 1893 and 1894; Jour. Soc. Chem. Industry, 1894, xiii, 267) with especial reference to their fermentability, the percentages of dextrin and sugars, and the existence of the so-called maltodextrin. After subjecting the amyloins to fractional precipitation by alcohol, they found dextrin, maltose, isomaltose, and glucose, but no substance having the properties of maltodextrin. Dextrin-like products of starch were investigated by Billow (Archiv f . ges. Physiologic, 1895, Lxii, 131), who prepared baryta compounds of amylodextrin, erythrodextrin, and achroodextrin, and determined their specific rotatory and reducing powers. A number of each of these compounds were studied. Various methods were pursued to obtain pure achroodextrin, such as treatment with phenylhydrazine, heating in alkaline copper solu- tion, precipitation with iron, fractional precipitation with barium hydroxide, and dialysis. Quite a nuniber of erythi-odextrins, having different rotatory and reducing powers, are recorded as having been obtained by these means. Meyer (Die Starkekorner, etc., loc. cit.) prepared by means of oxalic acid an erythro- dextrin that had a rotatory power of (a)D = +192°, and a reducing power R = 10. Lintner and Diill (Ber. d. d. chem. Gesellsch., 1895, xxviii, 1522) investigated the products of the conversion by oxalic acid by the methods they had previously employed in diastatic digestion. The chief difference noted was in the formation of dextrose as the final product of the action of acid, while maltose was the final product of diastase. With acid they obtained amylodextrin, erythrodextrin I, erytlu-odextrin II a, erytlu-odextrin II /?, achroodextrin I, achroodextrin II, isomaltose, and dextrose. With diastase they recorded amylodextrin, erytlirodextrin I, achroodextrin I, achroodextrin II, isomaltose, and maltose. The following were the color reactions of the dextrins with iodine: Erytliro- dextrin I, red- violet; erytlu-odextrin II a, red-brown with chlute iodine solution, but blue if the iodine is concentrated; erythrodextrin II (3, red-brown, even with concentrated iodine ; acliroodextrins, no color reaction. Ling and Baker (Proc. Chem. Soc. Trans., 1895, lxvii, 702, 739) obtained, by the action of diastase, a substance which in its rotatory and reducing properties resembled Brown and Morris's maltodextrin more closely than Lintner and Diill's achroodextrin II. They suggest that Lintner's isomaltose may consist of maltose and a simple dextrin, C12H20O10+H2O. They isolated a substance having the optical and reducing properties of "isomaltose," winch fermented slowly with beer yeast, leaving a residue which was not altered by diastase. This residue they state is a simple dextrin. Ost (Chem. Zeit., 1895, xix, 1501), in following Lintner and Diill's methods, found that a mixture of maltose and dextrin had properties corresponding to Lintner's iso- maltose, and he therefore concludes that Lintner's isomaltose is an impure maltose. Ost believes it uncertain whether Lintner and Diill's erytlirodextrin and achroodextrin are homogeneous substances; and he calls attention to the fact that the specific rotatory power of the so-called maltodextrin and achroodextrin II are not the same according to different observers, and he states that Brown and Morris's non-reducing dextrin does not exist. Ost believes that the dextrinous substance obtained by Zulkowski (page 108) is not a true dextrin, and he supports the statement of Musculus and Meyer, but in opposition to Rohmann and Schifferer, that erythrodextrins are mixtures of achroo- dextrin and starch. Prior (Bayerisches Brauer Jour., 1896, vi, 385) found in experiments with diastase 3 dextrins and maltose; and in some experiments with yeast also dextrose. He believes Lintner's isomaltose to be a mixture of achroodextrin and maltose. ERYTHRODEXTRIN, ACUROODEXTRIN, GRENZDEXTRIN, ETC. 131 Chlodounsky and Sulc (Sitzungsber. d. k. bohm. Gesellsch. d. Wissensch., 1896; Jahr. vi. d. Fort. d. Tierchemie, 1896, xxvi, 67) subjected starch-paste to the action of an extract of pancreas for 18 days at 38°. The gi-eater part of the starch was unchanged. The sohition was filtered and concentrated to a syrup, and then precipitated with 80 per cent alcohol, the precipitate being designated dextrin I. The alcohol was evaporated, and the solution again reduced to a syrup, and then precipitated with 90 per cent alcohol, yielding dextrin II. By subsequent treatment osazones were obtained. In the dextrin I fraction only achroodextrin could be found; and from the dextrin II fraction no dextrin could be obtained that was suitable for experiments, even after repeated purification. In other preparations with the glycerine extract of pancreas they state that erythro- dextrin could be detected. A method for preparing pure commercial dextrin was reported by Berge (Bull. Assoc. Beige d. Chimistes, 1879, x, 444), which consists of subjecting raw starch to a temperature between 80° and 115° in an atmosphere of gaseous sulphur dioxide. By this means Berge prepared dextrin containing as little as 0.95 per cent of sugar. He also used sulphurous acid in a liquid state, when he found that saccharification begins at about 100,° the most favorable conditions being a temperature of 1.35° to 140°, a pressure of 6 atmospheres, and 25 per cent of starch in a 3 to 6 per cent solution of sulphurous acid. Saccharification was complete in about an hour. Petit (Compt. rend., 1897, cxxv, 309, 355) prepared dextrin by subjecting boiled starch to the action of diastase at 70° for about half an hour, when it yielded a constant red reaction with iodine. He obtained the dextrin in the form of a wliite powder, which was non-hygroscopic and did not jdeld an osazone. This dextrin by further treatment was converted into saccharine products which differ somewhat according to the decomposing agent, the period of action, and the temperature. Lintner (Chem. Zeit., 1897, xxi, 737, 752) extended his previous investigations on the chemistry of starches, especially with reference to the dextrins and their isolation. He goes on to state that the purity of the product can be fairly well determined by the appHcation of cryoscopic and osazone tests, together with the rotatory and reducing powers, and the iodine reactions. Dilute alcohol he regards as the best agent for the isolation of the tlifferent dextrins, and he foimd that alcoholic barium hydroxide andalcohohc calcium- hydroxide solutions tend to bring about decomposition. The lower the molecular weight of the dextrin the greater the solubility, as a rule; but the dextrins were found to react upon one another when in solution, affecting each other's solubility. When strong alcohol is to be used he advises that only some of the total amount of alcohol be added to a hot dextrin solution, and that to this mixtm-e be added, with brisk agitation, hot alcohol of definite volume and strength, and that the preparation then be set aside to cool to room temperature. The products of the restricted action of diastase when boiled starch is acted upon at 70° were studied by Ling and Baker (Proc. Chem. Soc, 1897, clxxiii, 3; Jour. Chem. Soc, Lxxi, 1897, 508). Besides maltose they separated a substance which was isomeric with maltose (which they believe probably consists of maltose and a simple form of dextrin already described by them), together with two forms of maltodextrin. One of the malto- dextrins they look upon as being identical with the maltodextrin of Brown and Morris, while the other they identify with Prior's achi-oodextrin II. They hold that there is ample evidence to justify the conclusion that starch is broken down into a series of maltodextrins of decreasmg rotatory power and molecular weight and increasing reducing power, all of these bodies having optical and reducing powers equivalent to mixtures of starch and maltose. A number of erythrodextrins were obtained by Young (Journal of Physiology, 1897-8, XXII, 401) by subjecting soluble starch to the actions of dilute acids or enzymes, arresting 132 DIFFERENTIATION AND SPECIFICITY OF STARCHES. the action at the proj^er time, and then precipitating the special dextrin by means of a proper neutral saline. Preparations of different dextrins were also made from commercial dextrin. Tlu-ee forms of erythrodextrin were investigated. Erythrodextrin I was pre- pared from either commercial dextrin or a starch-digestion by saturating with magnesium sulphate, wasliing the precipitate with a saturated solution of this salt, dissolving the pre- cipitate in water, and then removing any soluble starch that may be present by half satura- tion with ammonium sulphate. The filtrate gives a bright reddish-purple (almost magenta) without any preliminary blue, irrespective of the quantity of iodine. This iodide of dextrin, they state, is thrown down upon saturation with ammonium sulphate in the form of a darker reddish-purple precipitate. Erythrodextrin II was obtained by saturating a solution of the mixed products with magnesium sulphate, filtering off the soluble starch and the ery- throdextrin I that are jirecipitated, and then saturating the filtrate with sodium sulphate at 33°. A solution of this last precipitate gives a reaction with iodine varying from a bright reddish-purple to mahogany-red. Young states that this dextrin may be made up chiefly of erythrodextrin I, but that this is improbable. Erythrodextrin Til was prepared by the spontaneous precipitation of the filtrate obtained after saturation as above of a solution of commercial erythrodextrin or a starch-digestion. This substance gives a red or purple reaction with iodine. If the salt had not dialyzed away previously and the iodine had been added, the filtrate, which at first is reddish-brown, becomes opaque, turbid, and brownish-black or of even a greenish tinge, and on standing a dark-blue precipitate is thrown down. If this precipitate is collected on a filter and dried it becomes reddish- brown; and if dissolved in water it forms a blue solution, which instantly changes to the red-brown characteristic color of this erytlirodextrin, the solution now behaving as an ordinary solution of iodide of erythrodextrin III. Young notes that glycogen gives a color reaction very closely similar to that of erythro- dextrin III, but that it can readily be distinguished from this dextrin by differences in its behavior with neutral salts. Achroodextrin was prepared by allowing diastase or dilute acid to act on starch-paste to a stage beyond the achromic point in relation to iodine. The action was stopped by boiling or by neutralization. The dextrins were precipitated by alcohol, collected, washed, dissolved in water, and then boiled on a water-bath to expel the alcohol. The solutions were cooled and treated in the usual way by saturation with neutral saline. Achroodextrin gives a very slight precipitate on saturation with annnonium sulphate, but it is doubtful, it is stated, if the precipitate is actually achroodextrin. Yoimg notes that erythrodextrin so-called is really a series of bodies quite distinct from either starch or the achroodextrin, and that the three products he distinguishes as erythro- dextrin I, II, and III give reactions closely similar to the corresponding products described by Lintner and Dtill and obtained by other methods. A method for separating dextrin from soluble starch was reported by Hefelmann and Schmitz-Dumont (Zeit. f. offentl. Chem., 1S9S; Chem. Centralbl., 1898, ii, 5G1) by which, for instance, 5 grams of commercial dextrin are dissolved in 250 c.c. of cold water and placed in a flask of slightly greater capacity, to which solution is added 18 c.c. of ether, and the flask stoppered and shaken. The contaminating starch separates as a flocculent precipitate, while the dextrin remains in solution. If sugar and soluble salts are present thej^ will remain in solution with the dextrin. Starch-paste was liquefied and converted into dextrin without the formation of sugar by Pottevin (Compt. rend., 1898, cxxvi, 1218) by the action of malt extract that had been kept for 15 to 20 minutes at a temperature of 79° to 80°. A 10 per cent solution of starch was subjected for 12 hours at G0° to malt extract that had thus been heated; and the solution was fractionated with alcohol into 3 portions. The first portion, or least soluble, gave with a trace of iodine a blue reaction, but a violet-brown with an excess; the second portion gave a reddish tint; and the third portion no coloration. In none was any sugar ERYTHRODEXTRIN, ACHROODEXTRIN, GRENZDEXTRIN, ETC. 133 present. Pottevin states tliat the dextrins differ only physically, and that the denser parts of the starch-grains yield a dextrin more difficult to saccharify than the other parts. A reversion product was obtained by Syniewski (Bcr. d. d. chem. Gesellsch., 1898, XXXI, 1791) in his experiments with soluble starch obtained by the action of sodium peroxide, which he beUeves is probably identical with a substance formed by the action of high pressure on potato starch. He noted that solutions containing more than 12.5 per cent of soluble starch deposit a wliite precipitate that is insoluble in cold water, and which when washed with water, alcohol, and ether, showed upon analysis that it is a deriva- tive from soluble starch by the removal of water. In a later communication (Ann. d. Chem. u. Pharm., 1899, cccix, 282) he holds: (1) The starch-grains consist of a homogeneous substance having the formula C'uHioOj. (2) That two forms of hydrolysis occur, carbinolhydrolysis and carbonylhydi'olysis, accord- ing as when in the combination with water there goes into solution an auhychid union between two cai-binol groups, or between two groups, one of which is carbonyl. (3) That the products from potato starch which do not reduce Fehling's solution are the results of carbonj'lhydrolysis. (4) That the simplest product of carbinolhydrolysis is amylogen (C54H96O48), and that the starch-molecule, as well as products of carbinylhydrolysis between starch and amy- logen, consists of an unkno^^^^ number of amylogen molecules in some form of an anhydrid miion. (5) That amylogen consists of 3 maltose groups in combination with 1 dextrin residue contain- ing 18 atoms of carbon, and that the latter is composed of 3 sugar residues, of which 2 are isomaltose residues. (G) That during the first stage of the hydrolysis of amylogen the maltose molecules gradu- ally separate, and the dextrin residue remains behind, but that this finally splits into isomaltose and glucose, and the isomaltose finally into glucose. (7) That the diastatic hytholysis of amylogens gives rise to intermediate products, the different stages being expressed as follows: amylogen = dextrin residue I = dextrin residue 11 = dextrin residue 111= isomaltose = glucose. (8) That all the products obtained from starch lay hydrolysis are to be designated dextrins. Those which originate by carbinolhydrolysis, and which do not reduce Fehling's solu- tion he terms amylodextriu. The dextrin which originates from amylodextrin after the splitting off of all the isomaltose molecules he names grenzdextrin. All dextrins between amylodextrin and grenzdextrin, those therefore from which isomaltose can be split off, he names maltodextrin. The dextrin which originates from grenzdextrin by the splitting off of glucose he names glucodextrin. (See pages 118, 135, and 147.) In studies of the saccharification of starch by the amjdase of malt, Pottevin (Compt. rend., 1898, cxxvi, 1218) dissents from the view of Brown and Morris that dextrin and maltose are formed coincidently and not successively, and that all dextrins have the same composition, molecular weight, rotatory power, and reducing power; but he supports Duclaux in the theory that dissimilarities of dextrins are due to physical differences. By heathig malt extract for 20 minutes at 79° to 80°, and treating starch-paste with it, the paste was liquefied quickly, the greater part going over into dextrin without forming any reducing .sugar. The paste was prepared by subjecting 10 grams of starch in 1 liter of water for half an hour at a temperature of 90°, and then for the same length of time in an autoclave at 120°. The paste thus obtained was sufficiently transparent to allow of the determination of its rotatory power, (a)j = -1-197.6°. Two liters of the paste were treated with 30 c.c. of the heated malt extract at 60° for 12 hours. Allowing for the correction required by the malt, the rotatory power was un- changed. A small quantity of iodine gave a violet reaction ; but with an adchtional amount of the reagent the color became a reddish-brown. The fluid was concentrated until it contained 10 per cent of solids, and it was then fractionally precipitated with alcohol. A 63 per cent alcohol precipitated a dextrin that was colored blue at first and then brownish- 134 DIFFERENTIATION AND SPECIFICITY OF STARCHES. violet, and was then converted into maltose by malt extract at G3.7°. Seventy per cent alcohol precipitated erythrodextrin from the filtrate. This dextrin was at first colored light red with iodine, and then a brownish-red, and yielded 82 per cent of maltose. In the filtrate achroodextrin was found, 95 per cent of which was converted into maltose. According to Pottevin, starch is changed into dextrin, and dextrin into sugar. He explains the simultaneous presence of dextrin and sugar in certain stages of the reaction by the fact that starch is composed of several constituents having varying degrees of saccharifiability. He treated raw wheat starch with different portions of malt until 10 per cent of the original amount was left, and then prepared starch-paste from this, and also from the normal raw starch, and found when both were treated in the same manner with diastase that the former yielded only 44 per cent of maltose, while the latter gave 75 per cent. In another contribution (Ann. d. I'Inst. Pasteur, 1899, xiii, 728) Potte\dn reports on the non-homogeneity of starch-paste, and he states that the interior more labile parts of the starch-grain are transformed almost instantly into dextrins which are wholly con- verted into sugar, and which are soluble in 60 to 70 per cent alcohol, while the more resistant portions of the grains jdeld dextrins that are only partially convertible into maltose and which are insoluble in strong alcohol. He shows that the diffusibility of various dextrins differs, and also that, owing to this, some of Brown and Morris's deductions regarding their maltodextrins are erroneous. He prepared mixtures of a properly selected dextrin and maltose which were absolutely identical with the maltodextrin of Brown and Morris, which is claimed by them to be an individual, but which, howe\'er, do not exist in chemical combination. Pottevin does not admit the existence of Lintner's isomaltose, and he holds that this body is merely a mixture of maltose and dextrin. In a third communication (Ann. d. I'Inst. Pasteur, 1899, xiii, 655) he maintains his statement in a former article that the starch- grain and also starch-paste are not homogeneous; that the less dense portions are readily converted into dextrin and tliis into sugar, while the more dense particles are converted slowly and not completely, and that there is always present at the end of saccharifica- tion a residue of stable dextrin. Pottevin studied especially the different phases of the reaction with reference to the properties of the enzyme used, and he holds that the enzyme is to be regarded as being a mixture, dextrin-fonxdng and sugar-forming, the one converting starch into dextrin and the other converting dextrin into sugar. Stable dextrin in its relations to maltodextrin and soluble starch was studied by Brown and Millar (Proc. Chem. Soc, 1899, xv, 13). They ascertained that when starch is trans- formed by active diastase at a temperature of 60° the reaction goes on very rapidly until a stage is reached when the rotatory power corresponds to (a)D = +153, and the reducing value R=80, at which time there is present maltose and a resistant dextrin having proper- ties corresponding to (0)0 = +195.7 and R = 5.7 to 5.9. The reducing power they believe to be inherent in the dextrin, as is indicated by the fact, as they state, that oxidation of the dextrin gives rise to dextrinic acid, a polysaccharid acid; and they believe that both the stable dextrin and the dextrinic acid are built up of Cg groups, and not Cjo groups, as in the case of maltodextrin. Thej' state that the dextrin molecule may be regarded empirically as composed of 39 CgHioOs groups in combination with a terminal C6H12OG group, or, more correctly, as a condensation of 40 glucose molecules with the elimination of 39 H2O. Petit (Compt. rend., 1899, cxxviii, 1176) supplemented the investigations above referred to. He pre])ared dextrin by the action of a 1 per cent of malt diastase on starch paste at 70°. The molecular weight of the dextrin he estimated to be 485, corresponding to the formula (CgH 1005)3. Its rotatory power was {a)r, = -1-166.6, and its reducing power, 72 = 18. Subjected to diastase at 50° to 55°, it was broken down into maltose and a residue ERYTHRODEXTRfN, ACHROODEXTRIN, GRENZDEXTRIN, ETC. 135 of the same dextrin. In a still later contribution (Conipt. rend., 1900, cxxxi, 453) Petit studied the dextrins formed at 50°, 60°, and 70°, respectively, which were separated by successive precipitations with alcohol. The fiRures given in table 10 are recorded. Table 10. Reducing power. Glucose equivalent. Rotatory power Molecular weiglit (by freezing). At 50° AtGO° At 70° 11.8 11.2 16.3 103.5 103.0 103.1 + 183° + 185° + 197° 1,096 1,310 723 The first two he regards as being probably identical, but quite different from the third. Differences in the products of diastatic and acid action were recorded by van Laer (Jour. Fed. Inst. Brew., 1900, vi, 162), who notes that not only are the sugar-products not identical, but also the intermediate bodies. (See page 152.) Acetyl derivatives were studied bj^ Pregl (JVIonatsch. f. Chem., 1901, xxii, 1049), who found a substance that gave a red reaction with iodine, and which he believes is an erythrodextrin. It is stated to reduce Fehling's solution and to have a reducing power equal to about 12.5 per cent of that of glucose. He could not identify it with any known dextrin, but states that it resem- bles a dextrin described by Syniewski, which was obtained by the action of malt extract. Baker (Jour. Chem. Soc, 1902, xviii, 134) investigated the actions of the diastase of ungerminated barley on soluble starch, in which the reaction was caused to continue for 13-2 hours at 50°, when it was found that dextrins and maltose were the only products. One of the dextrins obtained by precipitation with alcohol is recorded as giving a blue reaction with, iodine and to be acted upon by the barley diastase very slowly. After diges- tion for 90 hours at 45° to 50°, the product is stated to yield a blue reaction with iodine and to consist of unaltered dextrin with maltose and glucose. After 18 hours at 55°, by the action of malt diastase, this dextrin no longer gave a blue reaction with iodine, and the products consisted of achroodextrin and a considerable quantity of glucose. The dextrin formed by barley diastase differed from Nageli's amylodextrin, and he proposes, owing to its general beha\aor, to name it r-anujlodextrin. Erytlirodextrin was described by Hale as being one of the products formed by the primary action of iodine and oxidizing agents on starch (Amer. Jour. Science, 1902, xiii, 379). By titrating an arsenite or tartar emetic solution with iodine and impure starch, a loss of iodine occurs and a compound forms that gives a red reaction with iodine. This compound behaved like erythrodextrin, but it was not identical with it. Impure starch yielded also another substance which seemed to be intermediate between starch and ery- throdextrin. He states that amidulin stands between soluble starch and erythrodextrin and that the last is the first product of amidulin. In a further report on the constitution and decomposition products of starch, Syniewski (Rozfrawy ahademji umiejetnosci, Kjrakau, 1902; Jahr. ii. d. Fort. d. Tierchemie, 1902, XXXII, 98, 100) subjected a 5 per cent solution of potato-starch to a temperature of 138° to 140° for 12 hom's, and obtained amylodextrin. The hydrolysis of amylodextrin by malt extract at room temperature yielded a dextrin wliich he named grenzdextrin I. Malt extract acting on this dextrin produced maltose and an isomer of maltose wliich he named dexirinose. By heating malt extract for 15 minutes and then adding amylodextrin solution, and stopphig the action at the point when the solution no longer gave a blue reaction with iodine, the amylodextrin was found to have been converted into a form distinguished as grenzdextrin II. This he belie\es is identical with Brown and Morris's maltodextrin, 136 DIFFERENTIATION AND SPECIFICITY OF STARCHES. with Ling and Baker's a-maltodextriii, and with Lintner's achroodextrin II. Grenz- dextrin II is hydrolyzed by malt extract into maltose and a compound named y-malto- dextrin, which in tm-n may be broken down into maltose and dextrinose. From this and his previous investigations, Syniewski holds that the starch-molecule consists of 4 amylogen groups, its formula being CoicHsoyOxgo, each group consisting of 9 glucose groups united tlu'ough their 9 carbonyl groups, thus proving that only monocarbonyl groups exist among the glucose groups. The glucose groups, he concei\'es, are arranged into 1 dextrin group consisting of 3 rings of glucose groups linked together, and 3 maltose groups united to the dextrin group. The amylogen groups are conceived to be so combined in the starch-molecule that each is united to tliree others through 6 carbinol linkages, the carbinol hnkage existing between the dextrin groups (J-carbinol linkage) as well as between the maltose groups (?/t-carbinol linkage) of the amylogen groups. Amylodextrin, he states, is derived from the starch-molecule by the decomposition of ?M-carbinol linkage on the addition of 6 molecules of water. Heating starch-paste at 140° will not cause a decomposition of starch, but such decomposition can be brought about, he found, by diastase. Malt extract, by bursting the dextrin ring at fu-st, releases a carbonyl combina- tion between the two glucose groups of each dextrin group, and then gradually all the maltose groups are separated from the dextrin group, producing grenzdextrin I, which was made up of dextrin groups united tlurough d-carbinol linkages. The further action of diastase, he writes, causes the separating of one maltose gi'oup from each dextrin group of the gTenzdextrin, finally resulting in maltose and its isomer dextrinose, in the molecule of which the glucose groups are united by cZ-carbinol bonds. Malt extract at 78° likewise breaks the carbonyl bonds, primarily those which bring the glucose groups to the dextrin ring, thus forming grenzdextrin II, which is put down as being composed of 2 complexes (Cis) united by c/-carbinol linkage consisting of 3 glucose groups linked together through carbonyl bonds. It is conceivable, he states, that the separation of a maltose residue from this dextrin will produce )--maltodextrin and that dextrinose results from the fm'ther split- ting of tlie maltose group. Although the carbinol linkage seems to have great resistance to the hydrolytic action of diastase, dextrinose nevertheless finally breaks up into 2 molecules of glucose. Another coimnunication by Syniewski (ibid., page 109) reports the changes brought about in starch by subjection to the action of 40 per cent formaldehyde for 2 months. He investigated especially the iodine combinations of amylodextrin and concluded that each amylogen residue of the amylodextrin takes up 3 atoms of iodine, these atoms jorob- ably replacing the hych'oxyls of the primary alcohol group CHoCOH), each amylogen residue containing 3 such hydi-oxyls. That the hydi'oxyls of the primary and not the secondary groups are replaced by the iodine is, he states, confLrmed by barium compounds of amylogen, in which undoubtedly the primary hyth'oxyls are replaced by barium, the barium compound not combining with iodine, inasnmch as it shows no iodine reaction. This conception of the composition of iodine-amylodextrin lends probably, Syniewski holds, to the theory that the same alcohol groups of the amylogen residues also take part in the reaction with formaldehyde, the primary alcohol groups evidently being replaced by formaldehyde groups in the compound of amylodextrin and formaldehyde. By the hydrolytic action of added water, or by the addition of acids, one hydroxyl group after another will be set free, giving rise to color manifestations induced by the addition of iodine. Tliis change of color in the iodine reaction, which occurs also in the diastatic hydrolysis of starch, is due, he holds, to the gradual breaking away of the dextrin molecules from the maltose group, containing primary alcohol groups. It is therefore probable that the dextrins I and II, described in 1899, would show a red or brown color with iodine. In experimental studies of the processes concerned in the conversion of starch into sugar, Moreau (Ann. d. d. Soc. roy. d. Sc. med. e. nat. d. Bruxelles, 1903, xii, 117; Jahr. ERYTHRODEXTRIN, ACHROODEXTRIN, GRENZDEXTRIN, ETC. 137 ii. d. Fort. d. Tliierchemic, 1903, xxxiii, lOG) ascertained that precipitation of the prochicts of digestion by an aqueous solution of barium hydroxide does not throw down achroodex- trin, while with an alcoholic solution of barium hjalroxide the achroodextrin can not be separated from the erythroilextrin; but by precipitating the amylodextrin and erytliro- dextrin with an aqueous solution, and subsequently the achroodextrin in an alcoholic solution, the dextrin may be separated, and also a minute quantity of an undetermined dextrin-like substance. Repeated precipitation with barium hydroxide was found to free the dextrins from adherent sugai's. Prolonged digestion of starch always gave a residue that was not starch, dextrin, or cellulose. According to Moreau, dextrins do not exist as such in the starch molecule, nor are they produced by boiling a 1 per cent starch-paste, as maintained by Griessmayer. Moreau is of the opinion that Griessmayer's results were due to the presence of micro-organisms. In order to avoid misunderstanding in the use of terms applied to the intermediate products of the saccharification of starch, Ling (Jour. Fed. Inst. Brewing, 1903, ix, 446) proposes to discontinue the use of the word dextrin as applied to the products of diastatic and acid digestion, and substitute the word maltodextrin, which he thinks would not only specifically indicate the mode of origin, but also distinguish those products from torrefac- tion dextrin. Achroodextrin was prepared from peat, lichens, and moss by Reynaud (Jour. Soc. Chem. Industry, 1903, xxii, 567). A comparison of the products of potato starch with those of cereal starches was made by O'SuUivan (Proc. Chem. Soc, 1904, xx, 6.5), who states that no quantitative relationsliips were found between the percentages of maltose and dextrin of the former with those of the latter (see p. 148). The rapidity of hydrolysis by the action of dilute acids upon dextrin and maltose was studied by Noyes and several associates (Jour. Amer. Chem. Soc, 1904, xxvi, 266), who recorded that dilute hydrocliloric acid will hydrolyze dextrin only half as rapidly as it will maltose. A 2.5 per cent hydi'ochloric acid hardly affected the reducing power of dextrin, but the hydrolysis of maltose at 100° reached 90 per cent in an hour, and somewhat less than 95 per cent in 2 hours. The behavior of starch, amylodextrin, and erythrodextrin towards chromic acid was investigated by Hiii-z (BeUieft. z. Botan. Centralbl., 1905; Woch. f. Brau, 1905, xxii, 721). The author states that chromic acid forms combinations with these three substances that are analogous to those formed by iodine ; and that all three behave in such a way as to indi- cate that not one is a uniform substance, and that they really consist of a number of groups which differ in density or complexity of their molecular structures. Acliroodextrin seemed to be of apparently uniform composition. The method employed by Billow (p. 130) for the separation of the several products of digestion by aqueous and alcoholic solutions of barium hydroxide was improved by Moreau (Ann. d. d. Soc. roy. Sc. med. nat. d. Bruxelles, 1905; Woch. f. Brau, 1905, xxii, 37) . The progress of the reaction was determined by iodine. He found that amylodextrin and erythrodextrin were precipitated from aqueous solutions, but not achroodextrin or sugar, both of which were precipitated in the presence of alcohol. The limits of pre- cipitation were sufficiently far apart to permit of a complete separation of the several constituents by successive fractional precipitation by barium hydroxide. By this method he found in the digestion of starch by diastase, ptyalin, pancreatin, serum, or mineral acids that even diu'ing the very earliest stages of the decomposition processes all three forms of dextrin, as well as sugar, were present. He states that this confirms Mittelmayer 's theory of starch conversion, that is, that the starch molecule is at once broken down into these j^roducts, and that there may be some subsequent hydrolysis of the dextrins first formed. Moreau gives in detail processes for preparing amylodextrin, erj'throdextrin, and achroodextrin from commercial dextrin. The first two he prepared 138 DIFFERENTIATION AND SPECIFICITY OF STARCHES. absolutely free of reducing power, but the achroodcxtriii retained a slight reducing power, even after many successive precipitations. Because of the absence of reducing power in amylodextrin and erythrodextrin, Moreau believes that sugar does not constitute a part of the dextrin molecule. Many investigators have referi-ed to the existence of a dextrin residue that is not converted into sugar, but Fernbach and Wolff (Compt. rend., 1906, cxlii, 1216) found that if the residual dextrin is separated it may be saccharified by malt extract, and there- fore that if any dextrin exists which is not convertible into maltose it must be but an infinitesimal part of the original starch. The dextrinizing action of formaldehyde was referred to on page 109, in connection with Syniewski's investigations, the results of which have received confirmation in the investigations of Reichard (Zeit. ges. Brauw, 1908, xxxi, 161), who noted the succession of different color reactions \sdth iodine as the processes proceeded. Erythrodextrins were prepared by Tanret (Compt. rend., 1909, cxlviii, 1775) from a weak alcoholic solution of the insoluble precipitate (reverted starch) that formed in soluble starch on cooling, the quantity of erythrodextrins obtained representing about 8 per cent of the total matter dissolved in alcohol. A colorimetric method for deterinining the molecular weights of starch and its decom- position products and related carbohydrates was reported by Wacker (Ber. d. d. chem. Gesellsch., 1908, xli, 266; 1909, xlii, 2675). From these observations both erytlu-odextrin and achroodextrin have the values of 4 hexose groups. ISOMALTOSE, MALTOSE, GLUCOSE, SACCHAROSE, ETC. The discovery of sugar as a product of the activity of weak acids on starch was made by Ivirchhoff in 1811 (Schweigger's Journal, 1815, xiv, 389), and 3 years later he found that gluten and germinating barley also gave rise to sugar. He and his contem- poraries referred to tliis sugar as glucose. Over 30 years later Dubrunfaut (Ann. de chim. et phys., 1847, xxi, 178) isolated this body, and from a study of its properties he showed that it is not identical with glucose. This discovery seems, however, to have made no impression, since the literature of the subject shows that ^'arious in\'estigators up to 1872 refer to tliis sugar as glucose; and even Musculus based his theory of the hydrolysis of starch into dextrin and sugar upon the assumption that he was dealing with glucose. O'Sullivan (Jour. Chem. Soc, 1872, x, 579) rediscovered the sugar described by Dubrunfaut, and found that it had only 65 per cent of the reducing power of glucose, and named it maltose. Petit (Bull. d. 1. Soc. chim., 1875, xxiv, 519) records, after digesting starch-paste with diastase, that the preparation contained 5 per cent of dextrin and two kinds of sugar, one of which he states reduces an alkaline solution of copper oxide, while the other is without influence; the latter, after being boiled with 1 per cent sul- phuric acid, is saccharified, the resulting sugar being composed of about three-fourths of this non-reducing sugar. Sachsse (Sitzungsber. d. Naturf. Gesellsch., Leipzig, 1877; Jahr. li. d. Fort. d. Tier- chemie, 1877, vii, 60), in discussing the formula usually given for starch that was objected to by W. Nageli, who proposed for the accepted CeHjoOy the formula C3eH(3203i, states that it is the same as 6 (CgHioOs) ^-H20, and that tliis slight alteration is of interest analytically, as it takes into consideration the amount of starch in the relati\'e determina- tion in the conversion into sugar. If the formula read CqHkjOs (mol. weight 162), 180 molecules of dextrose would be derived from 162 of starch, or 100 : 90; but if it read accord- ing to Nflgeli, there would be 1,080 molecules of dextrose fi'om 990 of starch, or 108 : 99. Con-ect results, Sachsse states, are obtained by changing the formula to C3oH(;203i, while in using the other formula CgHioOs there was always an inexplicable dilTerence of 1 to 2 per cent. ISOMALTOSE, MALTOSE, GLtTCOSE, SACCHAROSE, ETC. 139 In an invest igalion of the products of activity of salivary and pancreatic enzymes in the digestion of glycogen, Seegen (Centralbl. f. med. Wissensch, 187G, xiv, 849) found that even though the glycogen was completely digested there was formed a much smaller amount of glucose than should be expected, only from 34 to 41 per cent of the calculated amount in the case of saliva, and from 45 to 48 per cent in case of the pancreatic extract; and he belie\-es that some other kind of sugar is produced, or some other form of decom- position product. Nasse (Archiv f. ges. Physiologic, 1877, xiv, 473) confirmed Seegen's work and ex- tended the investigations to starch. Nasse digested boiled arrowroot starch with filtered human saliva on a water bath at 40°, varying the quantity of starch and the period of digestion. In most cases the reducing power of the product amounted to only 48 per cent of that called for. Even when 0.3 gram of starch in 20 c.c. of water and 15 to 20 c.c. of saliva were digested from 6 to 7 hours, the reducing power was below 45 per cent of what should be expected. He holds that animal diastatic ferments do not convert starch into glucose, but into a form of sugar which he names amylum ptyalose, which when boiled ^nth cUlute sulphuric acid has its reducing power doubled. He also found achroodextrin. In experiments with glycogen he determined the presence of both ptyalose and acliroodextriu. Musculus and Gruber (Zeit. f. physiolog. Chemie, 1878, ii, 177) record both maltose and dextrose. The former, having the formula C12H22O1 1, a rotatory power of (a) = -f 150°, and a reducing power of 66, is believed by them to be formed from achroodextrin and to be convertible into glucose, each molecule being transformed by hydrolysis and cleavage into 2 molecules of glucose according to the following: Ci2H220ii+H20=2(C6Hi206), glucose having a rotatory power of (a) = -1-56 and a reducing power of 100. Nasse's assertion that a special form of sugar, and not glucose, results from the salivary digestion of starch was shown to be erroneous by Musculus and von Mering (Zeit. f. physi- olog. Chemie, 1878, n, 403), who ascertained that the conversion of starch by saliva is analogous to that by diastase, giving rise to acliroodextrin, maltose, and glucose. After repeated experiments they determined that under ordinary circumstances the quantities of maltose and glucose formed by the action of saliva or diastase are 70 and 1 per cent, respectively, of the original quantity of starch. Brown and Heron (Ann. d. Chem. u Pharm., 1879, cxcix, 241) using malt-extract recorded the production of maltose, and state that the results of their experiments indicate that even by the continued action on starch no glucose is formed, and that maltose is the end product, undergoing no further change. Pancreatic extract gave rise to both maltose and glucose. The contradictory results of Musculus and von Mering on the one hand by diastase, and of Brown and Heron on the other, as regards the formation of glucose by diastase, were discussed by von Mering (Zeit. f. physiolog. Chemie, 1881, v, 185), who attempted an explanation of the disagreement. Von Mering treated potato starch-paste with diastase for 4 hours at 00° to 70°, and then for 20 hours at room-temperature, and found by ordinary fractional precipitation of an alcoholic solution with ether that glucose was present. The prolonged action of diastase increased the amounts of both glucose and strongly reducing dextrins. Shorter action produced maltose, but no glucose. Maltose remained unchanged after 2 hours' treatment with diastase at 60°, but after 24 hoiu-s the reducing power was increased and the rotatory power decreased, indicating glucose forma- tion. Putrefaction, watery yeast-solution, and emulsin did not convert maltose into glucose; but by the prolonged action of saliva or pancreatic extract both maltose and glucose were formed, the pancreatic extract being energetic in the conversion of maltose into glucose. He states that diastase and saliva will convert starch-paste into 2 dextrins, one of which is broken down into maltose and a secondary glucose, while the other remains unchanged by the same ferments. 140 DIFFERENTIATION AND SPECIFICITY OF STARCHES. Jaquclain in 1S40 (loc. cit.) recorded that when starch is subjected to a temperature of 1G0° it is converted in part into dextrin and glucose. This statement was confirmed or contradicted by different observers so that Soxhlet (Reportorium f. analyt. Chemie, 1881; Jahr. u. d. Fort. d. Tliierchemie, 1881, xi, 80) was led to repeat the experiments to deter- mine the cause of the disagreement. Soxhlet found by a preliminary test that a given kind and quantity of starch at 149° yielded a larger amount of sugar the less the cjuantity of water present, from which he concluded that conunercial starch contains a saccharifying substance whose power is lessened by dilution. He also noted that hy subjecting speci- mens of the same starch for 5 hours at 149° entirely different proportions of sugar were formed in accordance with the amount of acid present. Potato starch and wheat starch showed acidity, 100 grams having an acidity equal to from 0.06 to 0.4 gram per 100 of sulphuric acid; but rice and corn starch usually exhibit an alkalme reaction. He notes also that even commercial starch has some reducing action, which may be due to the presence of preformed sugar. Salomon (Reportorium f. analyt. Chemie, 1881; Jahr. u. d. Fort. d. Tliierchemie, 1881, XI, 86) does not agree with Sachsse's statement regarding the formula for starch and the percentage of sugar obtainable (page 138). He found that the proportion of sugar conforms best with the formula CqHjoOs for starch. In the conversion of starch into dex- trose, according to the ecjuation a;(CoHio05)-|-H20 = CoHi20(3, the yield of glucose should be 111.11 parts of dextrose to 100 parts of starch, whereas according to Niigeli's formula the amount would be 109.09. Salomon saccharified potato starch by boiling with dilute acid according to tlie method of Sachsse, and obtained 111.16 and 111.11 per cent of glucose. In other communications (Jour. f. prakt. Chemie, 1882, xxv, 348, and xxvi, 324) he notes that while he obtained the proportion required on the basis of the formula being CgHioOs, the percentages found by other investigators in acid hydrolysis were lower — by Brunner 107, de Saussure 110, and von Allihn only 106 to 107. He reasons that these differences may be due to incomplete desiccation of the starch or to faulty methods in determining the quantities of sugar. Wliile he obtained the required theoretical quantity with potato starch, he recorded only 106.8 per cent with rice starch, which differ- ence he attributes to incomplete saccharification owing to an injurious effect of alkali upon the starch during the process of preparation. Salomon (Jour. f. prakt. Chemie, 1883, xxviii, 82) in a later investigation asserts Iiis belief that there is only one kind of dextrin formed during saccharification of starch; and that glucose, but not maltose, is formed by the action of sulphuric, oxaUc, citric, or tartaric acid, the conclusion as regards sugar being based upon the determination of the specific gravity and rotatory and reducing powers. The results of a study of wheat starch by Schulze (Jour. f. prakt. Chemie., 1883, xxviii, 311) support Salomon's conclusions regarding the formula of starch, the per cent of glucose, and of glucose being the sole sugar i)roduct. Schulze dried wheat starch and saccharified it by dilute hydrocliloric acid, and obtained by von AUilin's method (copper reduction) 110.986 of glucose per 100 of starch, corresponding almiost exactly to the formula CgHjoOs. By the specific-gravity method he recorded 111.4; and by the polariscope method, 111.85. By treatment with acetic acid under pressure he obtained a dextrin corresponding to the a-dextrin (ery throdextrin) of Bondonneau. Heating for 4 hours produced almost exclusively this dextrin, but by continued heating the dextrin partially goes over into glucose. The maximiun conversion of starch into sugar under varying conditions was investi- gated by von Allihn (Zeit. d. Ver. f. d. Rubenzucker-Industrie, 1883, 786; Dingl. Polyt. Jour., 1884, CCL, 554). Ten grams of anhydrous starch, containing 0.9 per cent of ash and 0.3 per cent of insoluble residue, were subjected to 100 c.c. of dilute hydrochloric acid of strengths varying from 1.33 to 10 per cent, and boiled for periods ranging from 2 minutes to 2}/^ hours. Table 11 is a statement of his results. ISOMALTOSE, MALTOSE, GLUCOSE, SACCHAROSE, ETC. Table 11. 141 Duration of boiling. Per cent of hydrochloric acid. 10 5 3.33 2 1.33 2 minutes 5 10 15 30 50 60 90 105 120 150 92.6 92.1 hV.7 89.6 87.4 9b!6 943 93.3 93.27 94.05 94.49 .... 84.94 93.68 95.05 94.89 87.85 92.87 93.84 94.65 Percentage of starch converted into sugar by boiling with the dilute acid. Von AJlilm recommends a 2 per cent solution of acid, and that the glucose, if wanted pure, be recrj'stallized from methj^l alcohol of a specific gravity of 0.816. Starch syrup was found by Sieben (Zeit. d. Vereins f. d. Rubenzucker-Industrie, 1884, 837) to contain 21.70 per cent of dexti-o.^^e, 15.80 of maltose, 41.96 of dextrin, 20.10 of water, and 0.3 of ash. Incidentally it is of interest to note that genuine honey contains neither dextrin nor maltose, and that Sieben found 34.71 per cent of glucose, 39.24 of levulose, l.OS of saccharose, 19.98 of water, and 5.02 of non-saccharine matter, and total per cent of sugar 75.02. Brown and Morris (Ann. d. Chem. u. Phar., 1885, ccxxxi, 72; Jour. Chem. Soc, 1885, XLVii, 527), by the action of malt extract at 50° to 60°, recorded at the end of digestion 80.9 per cent of maltose and 19.1 per cent of dextrin, and they hold that the properties of tlie products of the diastatic digestion of starch can be fully accounted for by the pres- ence only of maltose and a non-reducing dextrin. Lintner (Jour. f. prakt. Chemie, 1887, XXXVI, 481) looks upon maltose as being the sugar-product of starch digestion by diastase. Bourquelot (Compt. rend. 1887, civ, 71, 576) asserts that both maIto.?e and ghicose are products of enzymic activity. An important discovery bearing upon the explanation of the presence or absence of glucose among the products of digestion was made by Cuisinier (Chem. Centralbl., 1886, XVII, 614; Moniteur scient., 1886, 718), who noted that since glucose is formed in both normal barley and barley malt and in certain other cereals but not in others, there must be a special enzyme present which forms glucose. This hypothetical enzjane he named glucase. He found upon macerating corn-meal in water at a temperature of 50° that sugar was formed, and in order to determine the nature of the saccharifying agent, and also the character of the sugar, he carried out a series of experiments in which he found that corn contains an enzyme which converts starch into dextrins and glucose, and that the dextrins themselves are finally changed to glucose. He also showed that this enzyme converted maltose into glucose, and that it acts upon starch-paste as weU as upon raw starch. It has since been rendered evident that Cuisinier's enzyme was a mixture of diastase and glucase. The existence of such an enzyme as glucase was foreshadowed by the work of Brown and Heron (1880), who found that while diastase gave ri.se to maltose, extract of the pan- creas or of the small intestine not only formed maltose but also had the power of converting maltose into glucose. Von Mering (1881) confij-med Brown and Heron's statement of the action of extract of the pancreas. Bourquelet (1883) added further confirmation in experi- ments with extracts carried on under aseptic conditions; and he also obtained proof that the yeast plant and certain other low organisms secrete an enzyme which hydrolyzes maltose into glucose. Since the announcement of Cuisinier, it has been shown that glucase has probably a wide distribution in plant life. 142 DIFFERENTIATION AND SPECIFICITY OF STARCHES. Geduld (Wochenschrift f. Brauerei, 1892, viii, 620) also extracted glucase from corn by soaking the grist in cold water, filtering, and precipitating with alcohol. At temijeraturcs from 57° to 60°, 1 part converted 100 parts of maltose into glucose. This enzyme was pre- pared later by Beijerinck and others. The same or a similar enzyme has been shown to exist in the sera of the blood and lymjjh, in the saliva and pancreatic juice and succus entericus, and in yeast and malt, and in emulsin and certain other enzymes. Its presence or absence, as well as the amount present, serve at times to explain many discrepancies in the results of various observers as to the kind or kinds of sugar formed. In fact, diastatic enzymes as ordinarily prepared, especially animal enzymes, are apt to contain more or less glucase, and hence to give rise to more or less conversion of maltose into glucose (see page 143). Differences in the sugar product by the actions of enzymes and dilute acids, respect- ively, were noted by Effront (Moniteur sclent., 1887, 513). He states that maltose is always formed by the actions of enzymes and acids; that in the saccharification by malt the production of glucose is variable; that by the action of acid glucose is always formed with maltose in the earliest stages; and that in the earliest stages there is produced an almost constant ratio between the percentages of maltose and glucose (34 to 38 of maltose to 100 of glucose). Griessmayer (AUgem. Brauer- und Hopfenzeit., 1887, xxvi, 147), in his studies of the nature of the starch-cellulose of C. Nageh, refers to the presence of maltose and glucose after treating starch-grains with dilute acid until the skeleton-like substance (starch- cellulose) remained. Hanofsky (Mittheil. d. k. k. Tech. Gew.-Museums, 1889, 56), in analyses of commercial dextrin, makes note of the presence of maltose and also of the conversion of this sugar into glucose by further action of dilute acid. Brown and Morris (Jour. Chem. Soc. Trans., 1889, lv, 462) refer to maltose as the final product of the action of diastase. Lintner (Brauer u. Malzerkalender, 1890, xiii, 83), in his examinations of the action of diastase on raw starch from different sources at different temperatures, filtered the preparations and changed the dissolved substances into glucose by hydi'o- chloric acid. Flourens (Compt. rend., 1890, ex, 1204) subjected starch to the action of dilute sulphuric acid and found that the pro]3erties of the products of digestion as deter- mined by their rotatory and reducing powers are in accord with the belief that only one dextrin together with glucose is formed, but no maltose. The synthesis of a new glucobiose, now known as isomaltose, which is of importance as one of the end-jiroducts of the saccharification of starch, was reported by Fischer (Bor. d. d. chem. Gesellschaft., 1890, xxiii, 3687; 1895, xxviii, 3024). This sugar was obtained by subjecting 100 grams of glucose in 400 grams of hydrochloric acid of a specific gravity of 1.19 for 15 hours at a temperature of 10° to 15°. By precipitation with alcohol, a fioc- culent deposit was obtained which contained a small amount of isomaltose together with other substances. From the filtrate a precipitate was tlirown down by ether, which con- sisted of glucose, isomaltose, and another substance. By appropriate methods isomaltose was isolated and its properties studied. A body was prepared by Schiebler and Mittelmeier (Ber. d. d. chem. Gesellsch., 1890, xxiii, 3060; 1890, xxiv, 301) from the unfermented residue of starch sugar, which they state is identical with the isomaltose of Fischer that was prepared from glucose by the action of acid. This residue, referred to as gallisin, was purified by repeated precipitation with strong alcohol, and they found that it yielded an osazone which resembled the osa- zones of saccharoses, from which they concluded that gallisin contains a sugar. At first it was thought that this sugar was an intermediate product between starch-cellulose and glucose, but in their second contribution they recognize that the new sugar is formed from glucose. They subjected glucose to the action of 2.5 per cent sulphuric acid for 12 hours on a water-bath. The acid was then gotten rid of by barium hydrate, and the solution heated with phenylhydrazine acetate, when a large quantity of glucosone separated; ISOMALTOSE, MALTOSE, GLUCOSE, SACCHAROSE, ETC. 143 Table 12. SoUd glucose. Syrup A. Syrup B. Dextrose (fermentable) Isomaltose (reclucLng-iiower) = 84 per cent of that of maltose Dextrin 62.38 13.67 3.22 30.10 22.48 26.91 32.76 22.64 25.13 and on cooling the filtered solution another osazone was obtained which was identical with the osazone prepared from galHsin. Frequent mention is made of isomaltose in the subse- quent literature on the products of starch digestion. Tollens (Ann. d. Chem. u. Pharm., 1890, cclvii, 150) prepared from corn a syrup from which a sugar crystallized, which he states had all the properties of saccharose. In analyses of sweet corn at difTerent stages of growth he foiuid saccharose, glucose, dextrin, and other bodies. By the action of bulyric acid ferment {B. amylohacter) , Villiers (Compt. rend., 1891, cxii, 536) found among other products a small amount of a crystalline carbo- hydrate which was con\"ertetl into glucose by warm hydrochloric acid; and there was also another product which resembled cellulose, which was converted into glucose by acid. Lintner (Zeit. f. d. ges. Brauwesen, 1891, 281), in experiments with beer-worts, to which he added phenj^lhydrazine acetate, recorded that an osazone was formed which in crystal- line properties and melting-point was identical with the isomaltose of Fischer. Later, he and Dull prepared isomaltose from the products of the action of diastase on starch. The saccharine products of the fermentation of glucose by means of pure cultures of Sacch. cerevisioe and S. apunculatus were examined by Reinke (Zeit. f. Spiritusind., 1892, 79). His results were such as to indi- cate that only glucose was fermented. The dextrin present was determined bj' inversion with hydrochloric acid, while the percentage of isomaltose was determined by the residual reducing power after fermentation. The fig- ures are recorded in table 12. The existence of Cuisinier's glu- case (p. 141) was denied by Lintner (Wochenschr. f. Brauerei, 1889, v, 1038) because he failed to find glucose as well as maltose in experiments similar to those of Cuisinier. In a later research, Lintner (Zeit. ges. Brauw., 1892, xv, 123) acknowledges his error, and notes that while barley and wheat contain but little glucase, corn contains much. He prepared glucose by niLxing a thin starch-paste with corn-meal, and keeping the mbcture at 60° for 30 to 48 hours. The filtrate if concentrated will crystallize. The amount of glucose, he states, is small in comparison with maltose. In another article published at the same time, Lintner and Diill (Zeit. ges. Brauw., 1892, xv, 145) give a detailed method for the prepa- ration of isomaltose, by which a pure isomaltose can be obtained, and in quantity represent- ing 20 per cent of the original starch. This sugar they found to ferment with j^east much more slowly than maltose. It had the same specific rotatory power as maltose (a)^ = +140°, but its reducing power was found to be only 83 per cent of that of maltose. Its osazone melted at 150° to 153°, while that of maltose melts at 206°. The sole products of the action of diastase on starch are, according to Schifferer (Neue Zeit. Rub. Zuck. Ind., 1892, xxix, 167), dextrin or dextrins, isomaltose, and maltose. The maltodextrin of Brown and Morris he believes is probably a mkture of 07 per cent of dextrin and 33 per cent of isomaltose, while the maltodextrin of Herzfeld he regards as consisting of 26 per cent of dextrin and 74 per cent of isomaltose, and he thinks that Herzfeld must have unknowingly fractionated isomaltose from his so-called maltodextrin. Schifferer states that isomaltose is always formed during the saccharification of starch as long as dextrin is present; and that at the limit of the reaction the preparation does not have a reducing power corresponding to 80 to 81 per cent of maltose, but to 60 to 68 per cent. Glucase was investigated by Geduld (Wochensch. f. Brauerie, 1892, vii, 620), who found that this enzyme exists in ungerminated cereals in both soluble and insoluble forms, and in germinated grains in an insoluble form; that it does fiquefy starch-paste; that it has only a \-ery slight action on starch, a greater action on dextrins, and a very energetic 144 DIFFERENTIATION AND SPECIFICITY OF STARCHES. action on maltose, the end-product being dextrose; that the optimal temperature is between 56° and 60°, and that above 70° it is inert ; and that it is probably widely distributed and an important diastase. Geduld's results received sujiport in the investigations of Jalowetz (Wochensch. f. Brauerei, 1892, viii, 1264), who found evidence of the existence of glucase in cereals other than barley and corn. Morris (Trans. Inst. Brew., 1893, v, 132), however, failed to note any formation of glucose in experiments with maltose solution and aqueous extracts of corn, barley, rye, oats, and wheat, both malted and unmalted, with the exception of corn malt. He looks upon the results of Lintner and Jalowetz as fallacious because they ignored the existence of glucose and other sugars in the extracts they emj^loyed, and he believes that the enzyme is peculiar to corn and possibly other cereals as yet not examined, and that it does not occur as a normal or frequent constituent of barley or barley malt. Lintner and Diill (Ber. d. d. chem. Gesellsch., 1893, xxvi, 2533), in an investigation of the degradation of starch by the action of diastase, record that the products of digestion are amj'lodextrin, erythrodextrin, achroodextrin, isomaltose, and maltose. The so-called amyloins of Brown and Morris (page 115) they hold consist partly of mixtures of dextrin and isomaltose and in part identical with isomaltose. Isomaltose they found has a reducing power equal to 80 per cent of that of maltose; its rotatory power they give as (a)r, = + 140°, and the melting-point of its osazone 150° to 153°. Maltose they believe is formed from isomaltose, which they think leads to the conclusion that dextrins and starch are composed of isomaltose groups. The sugar products formed by the actions of animal enzymes were studied by Kiilz and Vogel (Zeit. f. Biologic, 1895, xxxi, 108). They used a 5 per cent solution of rice starch, and determined the sugars present by means of their osazones. They found that human parotid saliva formed isomaltose; that mixed human saliva formed isomaltose at first, then maltose, with traces of glucose; that dog's saliva formed isomaltose; and that pancreas of the ox also formed isomaltose. With liver glycogen instead of starch, parotid saliva formed 1 part of isomaltose to 2 parts of maltose; with muscle glycogen, saliva in small quantity yielded isomaltose with a little maltose and glucose, but a larger quantity ga^'e only maltose. With liver glycogen, pancreas gave isomaltose and a trace of maltose; and with muscle glycogen, there were formed isomaltose with a trace of glucose. Glucose, isomaltose, and maltose were found to be the sugar products of starch de- composition by Hiepe (The Country Brewer's Gazette, 1893, 1894; Jour. Soc. Chem. Indus- try, 1894, xiii, 267). Hiepe used several tj'pes of yeast, and noted that the percentages of isomaltose formed varied widely in relation to the kind of yeast. He shows that the amyloins are not indi\'iduals but mixtures. One of these bodies subjected to an elaborate process of fractional precipitation by alcohol showed the presence of glucose, maltose, isomaltose and dextrin. In a study of the differences in the products of starch conA^ersion by the action of acids and diastase, Lintner and Diill (Ber. d. d. chem. Gesellsch., 1895, xxviii, 1522) ascertained that isomaltose and glucose were produced with acid, while with diastase there were isomaltose and maltose. Lintner (Zeit. f. ges. Brauw., 1895, xvii, 173, 414) investigated the action of enzymes on isomaltose and maltose. He found that the conver- sion of isomaltose into maltose did not go on to completion, sometimes only 30 per cent being changed. He also notes that yeast powder gave rise to a very active formation of glucose, that a watery extract is less effective, and that isomaltose is more readily affected than maltose. He believes with Bau (Wochenschrift f. Brauerei, 1894, x, 1366; 1895, xi, 431) and Munche {ihidum) that isomaltose may consist of two stereoisomeric forms of iso- maltose which differ in fermentability. Prior (Bayerisches Brauer- Journal, 1895, 193) holds, to the contrary, that there are not two isomerides, and that Liutner's isomaltose is a homo- geneous substance. Lintner is of the opinion that the ferment which hydrolyzes maltose ISOMALTOSE, MALTOSE, GLUCOSE, SACCHAROSE, ETC. 145 and isomaltosc is not the same as tlie in^•el•tase which is also found in yeast, as it is less soluble than invertase and more nearly resembles glucase. Certain statements of Fischer, and of T.intner and Diill, were opposed by Ost (Chcmi- ker Zeit., 1895, xix, 1501, 1727). Ost investigated Fischer's isomaltosazone and concluded that it is a maltosazone, and he believes that maltose and not isomaltose is formed by hydi'o- chloric acid, as in Fischer's ex]icrinients. He followed Lintner and Diill's method for obtain- ing and purif}'ing isomaltose, and concluded that the substance was not homogeneous and that it is the nature of an impure maltose. Ost determined the rotatory power of maltose to be (a)D = +136.95° at 20°. He notes that this value differs from those of Meissl, Brown and Heron, Effront, and Herzfeld, but agrees with that of Parens and Tollens. Ulrich (Chemiker Zeit., 1895, xix, 1523), under Ost's direction, arrived at the same conclusion as Ost regarding the conversion of starch into maltose and not into isomaltose; and he also states that the melting-point of the maltosazone is modified by the method of preparation. Ling and Baker (Proc. Chem. Soc. Trans., 1895, xlvii, 702, 739) also investigated Ijntner's isomaltose, and obtained a substance yielding crystals resembling those of mal- tose and having properties resembling those of Lintner's isomaltose. The rotatory power was found to be +142.6° to +143.8°, and the reducing power 81.52 to 81.81 per cent of that of maltose. The melting-point of the osazone was 160° to 170°. In another instance an alcoholic extract of the products of diastatic digestion at 70°, upon treatment with phenjdhydrazine, jaelded a small quantity of glucosazone, and also an osazone that had the properties of Lintner's isomaltosazone. The latter osazone had a melting-point of 151°. They thought at first that this osazone was derived from a hexatriose, but in their second paper they withdraw this suggestion. Brown and Morris (Proc. Chem. Soc. Trans., 1895, lxvii, 709) agree with Ling and Baker that Lintner's isomaltose is not a chemical individual. They state, moreo\'er, that Lintner gives no evidence in any of his publications that the isomaltose prepared by him is identical with that of Fischer. They hold that Lintner's isomaltose can be split by fractionation with alcohol, or by fermentation, in such a manner as to indicate a mixture of maltose and dextrinous compounds of the maltodextrin or amyloin class; that the isomal- tosazone prepared by Lintner, upon which mainly he based his belief in the existence of isomaltose, is notliing but maltosazone modified by the presence of small but ^'arying quantities of another substance, such as dextrin.ous compovmds ; and that the only product of diastatic digestion capable of yielding an osazone is maltose. Jalowetz (Chemiker Zeit., 1895, xix, 2003) confirmed the statement of Brown and Morris that by mixing variable proportions of dextrin with maltose, maltosazones can be obtained which differ in melting-point, crystalline form, and general characteristics. In comparative studies of the actions of salivary, pancreatic, and intestinal enzymes, Hamburger (Archiv f. ges. Physiologic, 1895, lx, 543) showed that under certain condi- tions each will produce isomaltose, maltose, and glucose. He here found that two enzymes, diastase and glucase, are present in different proportions in the salivary, pancreatic, and intestinal secretions, and also in the blood. More diastase and glucase are found in the pancreatic juice than in the saliva, and less glucase than in the blood. In the saliva more diastase is found than in the blood and intestinal ferment, but less than in the pancreatic juice. Glucase predominates in the blood ferment, and therefore larger amounts of glucose than isomaltose and maltose are obtained by saccharification (see page 141). By the action of extract of pancreas on starch-paste, Chlodounsky and Sulz (Sitz. d. k. bohm Gesellsch. d. Wissensch., 1896; Jahr. u. d. Fort. d. Thierchemie, 1896, xxvi, 67) found that a sugar was formed which jdelded an osazone that was identified with glucosazone. In three contributions during this and the following year. Brown, Morris and Millar (Proc. Chem. Soc. Trans., 1896, xlviii, 242, 243; 1897, xlix, 4) take up a technical study of the specific rotatory and reducing powers of glucose, maltose, levulose, and invert sugar. 10 146 DIFFERENTIATION AND SPECIFICITY OF STARCHES. Further studies of the actions of diastase on starch were made by Ling and Baker (Proc. Chem. Soc, 1897, clxxiii, 3; Jour. Chem. Soc. Trans., 1897, lxxi, 508), in wliich they describe a substance isomeric with the isomaltose of Lintner. Its rotatory power was («)d = +156° and the reducing power 62.5. They beheve that it might consist of a simple dextrin together -with some maltose. Petit (Compt. rend., 1897, cxxv, 1899, cxxviii, 1176) found in experiments in which starch was subjected to a temperature of 70° with 1 per cent of diastase that a dextrin was formed which upon being treated with boiling dilute hydrochloric acid is within 3 hours comjDletely converted into glucose, but if the reaction is arrested at the end of a half hour, two osazones can be obtained by treating with phenylhydrazine and sodium acetate, one being glucosazone and the other corresponding to a diose, the latter being convertible into glucose by continued action. By further treat- ment with 1 per cent diastase at 50° to 55°, the dextrin formed yields two substances which differ in rotatory and reducing powers, the proportion of one being twice that of the other. Syniewski (Ber. d. d. chem. Gesellsch., 1898, xxxi, 1791), upon subjecting soluble starch prepared by sodium peroxide to the action of hydrochloric acid, found a quantity of glucose equivalent to 99.3 per cent of the starch, assuming, he states, that soluble starch has the formula C18H32O16. With malt extract at 65° for VA hours a yield of maltose equal to 82.7 per cent was recorded. Some of the discrepancies between the sugar determinations of different observers are due to a failm-e to recognize that sugars of various kinds may be present in the grains or seeds or in the diastatic extracts. In 1875, Kuhnemann (Ber. d. d. chem. Gesellsch., 1875, VIII, 202, 387) recognized in both germinating and ungerminated barley the presence of reducing sugars. He also isolated cane sugar and found that the quantity of sugar increased in germinating grains. His observations have had confirmation in the records of a number of investigators. O'Sullivan (Trans. Lab. Club, 1890, iii, 5) extracted barley with alcohol, and also extracted raflinose, and he showed that it is probable that both ger- minating and ungerminated grains may contain maltose, levulose, glucose, and saccharose. Brown and Morris (Jour. Chem. Soc. Trans., 1890, lvi, 458) found that during the germi- nation of barley the "secretion diastase" converts some of the starch into maltose, and that maltose is converted by the epithelium into saccharose. They also showed that the excised embryos placed in a solution of maltose changed maltose into cane-sugar. Griiss (Wochenschr. f. Brau., 1898, xiv, 81), in investigations of the sugars in barley and malt, ascertained that during the early stage of germination no erosion of the starch- grains occurs and that the grains are conserved for a time at the expense of stored-up sugars. Reducing sugar was found to be present in the aleurone layer under the furrow of the germinating grain. Cane-sugar was not found in the aleurone layer until after the begin- ning of germination, when it was detected there in larger quantity than in any other part of the grain. The reducing sugars in the endosperm were found to be in greatest quantity in the central part, and least in the cells contiguous to the aleurone layer. Gi'iiss believes that reducing sugar is converted into cane sugar in or near the aleurone cells. In another investigation, Griiss (Wochenschr. f. Brauerei, 1899, xvi, 519) made studies of the actions of two groups of oxidases and one group of diastases upon transitory starches. He notes among other things the formation, in the presence of air, of starch from saccharose. Ling (Jom-. Fed. Inst. Brew., 1898, iv, 187) determined the preformed sugars of malt by extraction with an alkaline solution. By treatment with phenylhydrazine he obtained an osazone which was probably glucosazone, and another which he records as being prob- ably a maltosazone. Differences in the products of the actions of acids and enzymes were pointed out by Johnson (Jour. Chem. Soc. Trans., 1898, lxxiii, 490), both in regard to final and inter- mediate products (see page 151). Pottevin (Compt. rend., 1898, cxxvi, 1218) treated wheat starch with successive portions of malt extract until 10 per cent of the original ISOMALTOSE, MALTOSE, GLUCOSE, SACCHAROSE, ETC. 147 amount of starch was left. He made starch-paste of this residue and also of the whole starch, and subjected both jmstes to the action of diastase under the same conditions. The former j-ielded 44 per cent of maltose and the latter 75 per cent. In another article (Ann. d. I'Inst. Pasteur, 1899, xiii, 728, 796) he discusses especially maltodextrins and Lintncr's isomaltose. Pottevin had previously found that the starch-grain is not homo- geneous, and that owing to this the dextrins yielded by different parts of the grain are not identical, the more resistant portions of the starch yielding dextrins which are only partially convertible into maltose. He found by mixing the proper kind of dextrin in proper proportion with maltose that a mixture could be obtained which seemed to be absolutely identical with the maltodextrin of Brown and Morris, and he therefore holds that the maltodextrin is not a chemical individual. He also contends with Brown and IMorris, and with Ling and Baker, that Lintner's isomaltose is not an individual. He states that it is merely a mixture of a non-reducing dextrin and maltose. In other communications during the same year (Ann. d. L'Inst. Pasteur, 1899, xiri, 665; Wochenschr. f. Brauerei, 1899, xvi, 641) Pottevin studied the phases of the processes of starch conversion, the heterogeneous character of the starch-grain, the differences in the dextrin and sugar products caused by variations in the constitution of starch, and the degi'ee of digestibility of the several components of the grains. Ground wheat starch was subjected for an hour to the action of malt extract, the solution filtered off, and fresh malt extract added to the undigested starch, this solution filtered off, and this process repeated four times, yielding four solutions. Examinations of these solutions showed that each successive portion of the dissolved products was more resistant to diastase. Dif- ferences in the starch-residues were also noted. Compared with the paste made from the whole starch, which yielded 80 per cent of maltose, a paste made from a residue represent- ing 9.5 per cent of the original starch yielded only 43 per cent of maltose. The denser the part of the grain the less digestible the starch, the less digestible the dextrin product, and the less the amount of the final sugax product. Notwithstanding that considerable evidence had already been offered to show posi- tively that maltodextrins are merely mixtiu'es of dextrins and maltose, Brown and INIillar (Proc. Chem. Soc. Trans., 1899, xv, 13) continued to study maltodextrin as though it w'ere an individual. They obtained a product of oxidation of maltodextrin which they term maltodextrinic acid. In an appendix they gave the methods they employed to deter- mine the yield of glucose by the hydrolysis of starch, maltose, maltodextrin, and maltobionic acid by oxaUc acid. (See page 152.) Syniewski (Ann. d. Chem. u. Pharm., 1899, ccix, 282), in his studies of the consti- tution of starch, holds, as already stated (page 133), that the simplest product of carbinol- hydrolysis is amylogen; that amylogen consists of 3 maltose residues with 1 dextrin residue, which latter in turn is composed of 3 glucose residues, of which 2 are isomaltose residues. In the first stage of hydrolysis of amylogen, all the maltose molecules gradually separate and the dextrin residues remain behind. By the prolonged action of malt extract this dextrin residue is split into isomaltose, and the isomaltose is finally converted into glucose. In a later article (Rozprawy akademji umiejetnosci, lirakau, 1902; Jiilu*. u. d. Fort, d. Thierchemie, 1902, xxxii, 98) he records that malt extract acting on grenzdextrin I produced maltose and another sugar, an isomer of maltose, which he terms dextrinose. It was found to have a rotatory power of (a)D2o= +141.41, and a reducing power of 84.5 per cent of maltose, and its osazone had a melting-point of 152° to 153°. He states that this sugar differs from Fischer's maltose and isomaltose, that it is probably identical with Lintner's isomaltose, and that it finally breaks up into glucose. Baker (Proc. Chem. Soc. Trans., 1902, xvin, 1177) studied the products of the action of barley diastase on starch. Barley diastase produced at 50° only dextrin and maltose during the first 1} 2 to 2 hours, but after 24 hours there was evidence of the presence of 148 DIFFEEENTIATION AND SPECIFICITY OF STARCHES. glucose. The dextrin isolated by precipitation with alcohol was acted upon slowly by barley diastase. At the end of 90 hours at 45° to 50° the product consisted of unaltered dextrin, maltose, and glucose. Barley cUastase was found to be without action on mal- tose; therefore the glucose must have come from dextrin. Ling (Jour. Fed. Inst. Brewing, 1903, ix, 446) noted that glucose is formed by the prolonged action of diastase that had been heated to 65°, and that the quantity formed by a restricted diastase did not exceed 12 per cent of the total products of hydrolysis. He also made the very interesting observation that upon continuing digestion the amount of glucose diminishes. He believes that a reversion of the glucose occurred through a synthetic action of the enzyme. Lintner's isomaltose, he suggests, may be a reversion product from glucose. Moreau (Ann. d. 1. Soc. roy. d. Sc. med. e. n. d. Bruxelles, 1903 xii, 117), by employ- ing malt extract, saliva, pancreatic extract, blood serum, or dilute hydrochloric acid as a hydrolyzer of starch and dextrin, found both maltose and glucose as sugar products. Both Rolfe and Geromanos (Jour. Amer. Chem. Soc, 1903, xxii, 1003) and Rolfe and Haddock (ibid., 1015) record that through hydrochloric acid hydrolysis both maltose and glucose are formed. Dierssen (Zeit. angew. Chem., 1903, xvi, 122), in experiments with oxaUc acid, found glucose and levulose, and also a disaccharid, but ne\'er maltose, and he states that if maltose is formed it must be converted immediately into glucose. The disaccharid corresponded with the properties of Lintner's isomaltose, excepting that it, unlike Lintner's isomaltose, is not converted by diastase into maltose. He does not believe it identical with Fischer's isomaltose, because it is fermentable, while Fischer's is not. Gruters (Zeits. angew. Chem., 1904, xvii, 1169) also studied the products of oxalic acid hydrolysis. According to him these substances consist of achroodextrins, maltodextrin, maltose, glucose, and a small amount of levulose. He failed in attempts to prepare a pure isomaltose, and found that mixtures of dextrin and maltose yielded osazones having the properties of isomaltosazone. He states his disbelief in the "isomaltose theory." A comparison of the products of the decomposition of potato starch with starches from cereals was made by O'Sullivan (Proc. Chem. Soc. Trans., 1904, xx, 65), who found that the percentages of maltose and dextrin obtained from potato starch bear no quantitive relation to those recorded in the experiments with cereal starches. This is in opposition to his previous finding (Jour. Chem. Soc. Trans., 1878, ii, 1410), and his statement has been shown by Ford and Guthrie (Jour. Soc. Chem. Ind., 1905, xxiv, 605) to be erroneous, and due to errors of experiment (see page 193). The hydrolysis of maltose and dextrin was in\-estigatcd by Noycs and his associates (Jour. Amer. Chem. Soc, 1904, xxvi, 266). They found that while the reducing power of glucose is scarcely affected by heating with a 2.5 per cent solution of hydrochloric acid, the products formed by the hydrolysis of maltose show a maximum reducing power after about 1 hour at 100° or after 20 minutes at 111°, and that further heating causes a decrease. The maximum reducing power corresponded to a hydrolysis of 96 to 98 per cent. Hydro- lysis seemed to be more complete in 2.2 to 4 per cent solution of product than in a 0.5 per cent solution. The rate of hydrolysis of dextrin was about half that of maltose. Jhe hydrolysis reached 90 per cent in an hour. The reducing power of the products obtained by the action of malt indicated a composition of 74 to 78 per cent of maltose and 26 to 22 per cent of dextrin. Such a mixture, from the results recorded, should, after 1 hour at 100°, give about 96 per cent of the copper oxide, which would correspond to complete hydrolysis. The amount actually found was 96.4 per cent for the 0.5 per cent solution, and 97.1 for the 2 per cent solution. Practically, the maximum effect is reached in about 1 hour at 100°. By dkect treatment of corn starch with 2.5 per cent hydrochloric acid they obtained in a 0.5 per cent solution a hydrolysis of 97 per cent in 1 hour and 98 per cent in 4 hours. DIFFERENCES IN PRODUCTS OF ACID AND ENZYMIC ACTION. 149 Roux (Compt. rend., 1905, cxl, 1259) made comparative studies of the saccharifica- tion of "artificial starclies" (page 112) and ordinary starch-paste by malt extract at various temperatures. After 4 hours digestion lie found dextrin and maltose, but no glucose. At 56° the average apparent amount of maltose from the artificial starch was 97.9 per cent, while from the ordinary paste it was 82.3 per cent. By fractionation the former yielded 30.4 per cent of crystallized maltose and the latter 30.4 per cent. At 67° the percentages of maltose from artificial and normal starches were 55.1 and 45, respectively. At 80° the artificial starch was unaffected. The dextrins of artificial starch, unlike those from ordinary starch, were almost completely soluble in alcohol. Fernbach and Wolff (Compt. rend., 1905, cxl, 1067) found that the energy of sacchari- fication is influenced materially by the condition of the starch, whether in liquefied or in paste form. Under the same condition of experiment the conversion products of liquefied starch that had been obtained by heating to 140° to 145° were greater than those of boiled starch as ordinarily prepared. They note that barley extract has a diastatic but not a liquefying power. Starch liquefied by a minute amount of malt extract heated to 70° beha\'ed the same as starch liquefied at high temperatures; therefore, the addition of a minute amount of malt extract (sufficient to cause liquefaction) to barley extract caused very energetic saccharification. A modified form of dextrose, provisionally termed (^-dextrose, was described by Roessing (Chemiker Zeit., 1905, xxix, 867), and obtained by him from starch syrups or pure dextrose that had been subjected under pressure to the action of dilute hydrochloric acid at tem- peratures of 130° or higher. Wlien sulphuric acid is used instead of hydrochloric, the tem- perature might be raised to 160° without giving rise to this abnormal product. Maquenne and Roux (Compt. rend., 1906, cxlii, 142) recorded that if a starch-paste and malt-extract mixture be nearly neutralized, a higher yield of maltose may be obtained. Fernbach and Wolff (Compt. rend., 1906, cxlii, 1216) showed, however, that the increased quantities can also be obtained without neutralization if a longer time of action be permitted. Neutralization, they state, therefore merely increased the velocity of the reaction. They record that any dextrin present, if separated, is saccharifiable, and tliat if there is present any dextrin that is not convertible into dextrose, it must represent only an infinitesimal fraction of the original starch. Croft Hill (Jour. Chem. Soc. Trans., 1898, lxxiii, 634), in experiments on reversible hydrolysis, states that glucose can under proper conditions be synthetized into maltose together with another sugar which he termed revertose. EmmerUng (Ber. d. d. chem. Gesellsch., 1901, xxxiv, 600, 2206, 2207, 3810, 3811) showed that revertose is isomaltose. This was confirmed by Armstrong (Proc. Roy. Soc, 1905, lxxvi, B., 592; The Simple Carbohydrates and Glucosides, 1910). Emulsin, an enzyme that forms isomaltose, was found in yeast by Henry and Auld (Proc. Roy. Soc, 1905, lxxvi, B, 568). Since yeast also contains maltase, which forms maltose, the occmTence of both isomaltose and maltose can be accounted for by the inde- pendent actions of the two enzymes, and not, according to Lintner and Diill, to the con- version of isomaltose into maltose (page 144). In fact, yeast extract may contain at least five diastatic enzymes, and emulsin at least three. DIFFERENCES IX THE PRODUCTS OF ACID AND ENZYMIC ACTION. The striking analogy that exists between the processes and products of enzymic and acid activity has not infrequently led to the assumption of a closer hkeness than really exists. It is assumed that both cause a degradation of the starch-molecule, and of the intermediate products of saccharification, by successive stages of hydrolysis, and that the stages as well as the intermediate and final products are essentially the same. But there are abundant reasons for believing that there may not only be certain specific differ- 150 DIFFERENTIATION AND SPECIFICITY OF STARCHES. ences in the products, but also more or less marked differences in the stages, at least in certain cases. In fact, not only are dissimilarities to be obser\'ed between the effects of enzymes and acids, but also between the effects of different enzymes and different acids. Different enzymes might be expected to yield somewhat different results, owing in part to differences in origin in plant and animal life respectively, but chiefly because enzymes, as ordinarily used or prepared, are seldom individuals, but composites of several distinct enzymes, varying in regard to kind and quantity and in their specific actions. If the progress of the action of animal enzymes (ptyalin and amylopsin) on starch be traced by testing from time to time with a very dilute iodine solution (1 per cent Lugol's solution), the blue reaction noted at the very beginning of digestion gives way to a purple, and this to a violet, this color persisting but becoming weaker and weaker, up to the time of the final disappearance of all color response with the iodine. With 'plant enzymes and with acid the blue reaction gi\'es way quickly to a red or brownish-red reaction which may ulti- mately pass into a reddish yellow, gradually fading away. This certainly indicates some differences in the intermediate products. Moreover, whether there is produced by enzymic action only maltose, or both maltose and isomaltose, or maltose and isomaltose and dex- trose, etc., will be found to depend solely or almost entirely upon the composition of the enzyme preparation. With a given enzyme maltose may be the sole sugar product, with another there may be isomaltose, or maltose and glucose, or isomaltose and glucose, etc., depending upon whether there be a pure maltase present, or emulsin, or maltase with glucase, or maltase with emulsin and glucase, etc., maltase converting dextrin into maltose, emulsin forming isomaltose, and glucase converting maltose and isomaltose into glucose, etc. The nature of the products formed by acid hydrolysis are modified by the nature of the acid, the temperature of digestion, etc. ; moreover, the products compared with those formed by enzymes and other agencies are by no means identical. Mulder (Chemie des Bieres, Leipzig, 1858, 16G; quoted by W. Nageli in Die Stiirkegruppe, etc., loc. cit.) found that the dextrins formed by the action of malt extract, sulphuric acid, and torrefaction, respectively, differed from one another in their reactions, especially in relation to certain precipitants. Soxhlet (Zeit. Spii'itindustrie, 1884; quoted by Johnson, Jour. Chem. Soc. Trans., 1898, Lxxiii, 490) states that the dextrins resulting from the actions of acids are quite different from those produced by diastase. Diastase, for instance, had no action on the dextrins formed and digestible by acid. Johnson (see page 151) has, however, shown that there is a slight action. Effront (Monit. Scient. f. 1887, 513; Jour. Soc. Chem. Ind., 1889, VI, 733) investigated the products of the saccharification of starch by extract of malt and by acid, and reached the following conclusions: (1) The course of the conversion of starch into dextrin and sugar is not the same with malt and acid. The saccharification by malt is attended by the decomposition of the starch-molecule into dextrin and maltose, while with acid the conversion is into dextrin and maltose, which in turn are converted into glucose. (2) The dextrins formed by malt and acid are not identical; those formed by malt are polymeric, while those formed by acid are not. (3) The dextrins in l)oth instances have the same rotatory power. (4) Maltose is always formed in the saccharification of starch by acid, and the quantity increases as saccharification proceeds. In the earlier stages of saccharification there is an almost constant ratio between the quantities of maltose and glucose formetl. This is about 34 to 38 of maltose to 100 of glucose, (f)) In the saccharification of starch by malt the formation of glucose is inconstant. While glucose almost always occurs in solutions of high gravity, it is only formed in liquids of low gravity if the malt extract employed be turbid. In another publication (Enzymes and their Applications, Trans, by Prescot, 1902) Effront states that the dextrins formed by the action of acid ha\'e a very low nutritive DIFFERENCES IN PRODUCTS OF ACID AND ENZYMIC ACTION. 151 With diastase. With oxalic acid. Amylodextrin Erythrodextrin I Amylodextrin Erythrodextrin I Erythrodextrin II a Erythrodex-trin II /3 Achroodextrin I Achroodextrin II Isomaltose Achroodextrin I Achroodextrin II Isomaltose Maltose Glucose value because pancreatic juice has a scant and incomplete action on them, wliile on the other liand the dextrins formed by malt are easily transformed by diastases, as shown by Soxhlet and Stutzer. A comparative study of the products of diastatic and acid degradation of starch was made by Lintner and Diill (Ber. d. d. chem. Gesellsch., 1893, xxvi, 2533; 1895, xxviii, 1522). In the first research they subjected starch to the action of air-dried malt, varying the amount, and stopping the digestion at the necessary point by boiling, and then repeatedly Table 13. fractionating the product with alcohol. In the second investigation they used oxalic acid instead of enzyme. Table 13 is a statement of the prod- ucts obtained in the two cases. They state that, judging from their experi- ments, acid and diastatic products of the same molecular weight would yield the same products and behave similarly upon the addition of more acid or enzyme. Even though the dextrins are alike, and even though the erythrodextrin II a and II /8 were also obtained with diastase, the only difference between the two hydrolytic processes would be that in the acid process no maltose but isomaltose is obtained and glucose appears as the end-product, and with the other maltose appears as the end-product. Eolfe and Defren (Jour. Amer. Chem. Soc, 1896, xviii, 869) found that only three carbohydrates are present during acid hydrolysis, namely, dextrin, maltose, and glucose. They believed that the products of acid and enzymic activities are the same. Petit (Compt. rend., 1897, cxxv, 309, 355) records that the saccharine products of dextrin digestion differ somewhat according to the decomposing agent, the period of action, and the temperature. The torrefaction dextrins of commerce can be distinguished, according to Henderson (Inaug. Diss., Munich, 1897; Jour. Soc. Chem., 1898, xvii, 591), from those obtained by the actions of diastase and acids by then- containing a more or less large propor- tion of substances of low rotatory and reducing power, so that the dextrins formed by these methods can not be considered identical. Henderson states that generally speaking the same dextrins are formed by the actions of diastase and acids, although with diastase dextrins are produced which resist further action of diastase. These dextrins, he believes, have the composition of achroodextrins I and II of Lintner and Diill. The possibility of isomaltose being included among the resistive products, he states, is to be recognized. The reactions with acids were found by Syniewski (Ber. d. d. chem. Gesellsch., 1898, XXXI, 1791) to proceed farther in the formation of glucose than with enzymes. Experi- menting with soluble starch prepared by the agency of sodium peroxide, he noted that with hydrochloric acid the amount of glucose obtained was equivalent to 99.3 per cent, whereas with freshly prepared malt extract it was 82.7 per cent of maltose. A comparative study of the products of acid and enzymic hydrolysis was made by Johnson (Jour. Chem. Soc. Trans., 1898, lxxiii, 490), who ascertained that the products in the two cases differ to a marked degree. He states that when the products of diastatic conversions are fractionated by means of alcohol, the precipitated portions have specific rotations which may vary from about (a)D3-86 = +150° to about 190°. These fractions are reduced to represent the following peculiarities: They are unfermen table by Saaz yeast, and theu' cupric-reducing powers and specific rotations can be expressed in terms of dextrin and maltose. Dextrin has a specific rotation of (0)03-86 = +1^5°, ^nd maltose (a)D3-86 = -[- 135.4°, so that no substance resulting from diastase hydrolysis can have a special rotation 152 DIFFERENTIATION AND SPECIFICITY OP STARCHES. of less than 135.4°. Wlien, however, fractions from the products of acid hydrolysis are separated the specific rotations vary between about (a) D= +190° and (0)0= +80°. These fractions contain no free glucose, because they are unfermentable by Saaz yeast; moreover, they can not be molecular aggregates of maltose and dextrin, since the specific rotation of a maltodextrin can not fall much below (a) ^ = + 150°, and we are in the presence here of fractions whose rotations can fall as low as 80°. There are therefore present in acid reactions intermediate substances which do not exist in diastase conversions. In the results of fer- mentation and dialysis tests he gives evidence to show that the fractions are definite compounds and not mixtures of glucose and other carbohydrates. Further proof that the intermediate products of acid and diastatic hydrolysis are not identical is shown by the difference in their behavior to diastase and acids. Soxhlet had already reported that diastase has no action on the dextrins produced by acid, but Johnson shows that it has a slight action, which he explains by the assumi^tion that it may attack mole- cules of starch that have not been acted upon by acid, although he admits that it attacks conversions which do not give a blue reaction with iodine. Diastase has no action on fractions having rotations of 114° and lower, but such dextrins do not offer much resistance to acids. In studies of the oxidation products and constitution of maltodextrin. Brown and Millar (Proc. Chem. Soc, 1899, xv, 11) found that this substance is completely hydrolyzed by diastase into maltose, and by acids into d-glucose. When maltodextrin is oxidized with mercuric oxide and barium hydi-oxide, the greater part of the jiroduct appears as a barium salt of a complex carboxylic acid, which they provisionally term maltodextrinic acid A. When maltodextrinic acid A is subjected to diastase it yields 40 per cent of maltose and 60 per cent of maltodextrinic acid B. Wlien acted upon by dilute oxalic acid, malto- dextrinic acid A yields 85.8 per cent of d-glucose and a simpler form of maltodextrinic acid. Van Laer (Jour. Fed. Inst. Brewing, 1900, vi, 162) notes that the differences in the products of acid and diastase are of im]iortance in brewing. He states that acid liydrolysis differs from enzymic hych'olysis not only in that the end-product of the former is glucose and that of the latter maltose, but also in that the intermediate products differ. In order to meet certain conditions favorable to brewing he proposes to partially saccharify grits by dilute acid, then neutralize, and finally complete saccharification by malt. In the saccharification of starch by acid it was found by Effront, as stated, that maltose is formed, but Lintner and Diill {loc. cii.) did not obtain maltose, and they look upon the absence of maltose in acid hydrolysis as being the essential distinctive feature to differen- tiate the products of acid and enzymic hydrolysis. Rolfe and Defren {he. cit.) found mal- tose present in both acid and enzymic hych-olysis. Rolfe and Haddock (Jom\ Amer. Chem. Soc, 1903, xxii, 1015) and Rolfe and Geromanos {^ihid., 1003) .support Rolfe and Defren. Rolfe, in a large number of determinations of the products of acid hydi'olysis of corn starch and potato starch, found conclusive evidence of the production of a reducing body other than dextrose, and Rolfe and Haddock made alcoholic fractions of a commercial glucose that had been made by the action of hydrochloric acid at a pressure of 2 atmos- pheres, and having a specific rotation of 126.5° and a reducing power of 0.575. From these fractions they prepared crystals of maltosazone and dextrosazone. Dierssen (Zeit. angew. Chem., 1903, xvi, 122; Jour. Soc. Chem. Ind., 1903, xxii, 312) used oxalic acid, and he agrees with Lintner and Diill in regard to the absence of maltose. Dierssen concluded that in acid hydrolysis glucose, levulose, and a disaccharid are formed. The rotatory and reducing powers of the latter, and the solubility, appearance, and melting- point of its osazonc correspond with Lintner's isomaltose, but it is not affected by diastase, while Ijintncr's isomaltose (which was prepared by diastase) was converted by this enzyme into maltose. Wliether or not the isomaltose obtained by Lintner by acid hych-olysis would have been converted by diastase was not noted. He states that it can not be identi- DIFFERENCES IN PRODUCTS OF ACID AND ENZYMIC ACTION. 153 cal with Fischer's isomaltose because Fischer's isomaltosc was not fermentable, while tliis sugar is fermentable; and, moreover, Fischer's was Isevo-rotatory, while his was dextro-rotatory. Diersscn goes on to state that from the results obtained no conclusions can be drawn with regard to the course of the diastatic hydrolysis of starch, but the fact is emphasized that the products in this case differ considerablj' from those yielded by acid hydrolj^sis. Additional evidence is also afforded by the tendency of glucose to form double compounils with other sugars, one molecule of the syruj)y isomaltose obtained by Dierssen yiekUng crystalline double compounds with any number of molecules of glucose from one upward. Griiters (Zeit. angew. Chem., 1904, xvii, 1169; Jour. Soc. Chem. Ind., 1904, xxiii, 875) also studied the final degradation products of the action of oxalic acid on starch. He states that these substances include achroodextrin I and II, maltodextrin ;-, maltose, glucose, and a small proportion of levulose. These he notes are the same as the products yielded by diastatic action, except that in the latter case maltodextrin r is replaced by maltodextrin /3, which exliibits different constants and also behaves differently towards diastase. He beheves that both these dextrins occur simultaneously, but in varying pro- portions, this view being supported by the fact that the conversion into maltose is some- times veiy imperfect, and at other times almost complete. The divergent behavior of various isomaltose preparations towards malt extract is also regarded as indicating that the more resistant maltodextrin r occasionally preponderates as the lowest member of the dextrin series. Like Lintner and Dierssen, he failed to isolate maltose in the pure crystalline state from syrups; nor could he accomphsh this from mixtures of maltose and dextrin, which, moreover, gave osazones resembling isomaltosazone. Differences in conditions under wliich hydi-olysis occm's may decidedly modify the products, as shown by Ski'aup (Ber. d. d. chem. Gesellsch., 1899, xxxii, 241.3). He noted, for instance, that usually, on acetylation of starch with mixtures of acetic and sulphuric acid, products of gi-eater complexity are produced by little acid and low temperatures than under opposite conditions. Of incidental interest in this connection, and in support of differences in the products of acid and enzymic hydrolysis, are the results of the experiments of Tebb (Jour. Physi- ology, 1897-8, XXII, 423) with glycogen, and of Loewi (^Vi-chiv f. exp. Path. u. Pharm., 1902, XLViii, 303), Henriques and Hansen (Zeit. f. physiol. Chemie, 1905, xliii, 417), and Abderhalden and Rona (Zeit. f. physiol. Chemie, 1904, xlvii, 530 and 1905, xliv, 200) with proteins. Tebb foimd in the acid hydrolysis of glycogen that the products are soluble glycogen, erythrodextrin, achroodextrin, and glucose, and that prolonged hydrol- ysis by saliva, pancreatic extract, and malt extract the only dextrins that could be sepa- rated in amount sufficient to work wath subsequently were of the achroodextrin variety, although evidence of erythrodextrin was noted. By Uver enzyme the intermediate dextrins resembled those produced by the amylolytic enzymes, except that a small amount of ery- throdextrin is constantly found in the earlier stages of the hydrolysis of glycogen. The reaction of liver enzyme differed from the other ezymes used in the nature of the final product obtained, which is, cliiefly at least, glucose and not maltose. Loewi found that an animal fed on the products of the prolonged self-digestion of pancreas may not only thrive but gain in weight. Henderson and Dean fed a bitch on the non-protein products of acid hydrolysis of lean beef together with lard and arrowroot, and record that their numerical results are essentially similar to those obtained by Loewi. From the ninth to the thu-teenth day of feeding, the animal was in nitrogenous cciuilibrium, and the body weight was maintained. In opposition to the statements of Henderson and Dean, both Henriques and Hansen and Abderhalden and Rona found that while animals may thrive on the enzymic jiroducts of protein digestion they do not live longer on the products of the acid hydrolysis of proteins than animals fed on a non-nitrogenous diet. 154 DIFFERENTIATION AND SPECIFICITY OF STARCHES. UNUSUAL PKOUUCTS OF THE DECOMPOSITION 01< STARCH. The usual products of the decomposition of starch thi-ough the agency of enzymes, dilute acids, bacteria, and other agents which bring about saccharification are erythro- dextrin, achroodextrins, maltose, isomaltose, and glucose, which products are variable quantitatively and qualitatively in relation to the conditions of experiment. The so-called amylodextrins and maltodextrins we are justified, from the literature of the subject, in regarding as mixtures. Such dextrins and sugars, as noted, may be looked for with cer- tainty luider ordinary conditions of experiment with most digestive agents, but certain unusual products may be formed under modified conditions or by certain agents. A num- ber of references have been made to such products, as, for instance: To Leitner (page 127), who found through the action of permanganate that gummy substances are formed wliich differ from dextrins; to Zulkowsky and Franz (page 127), who found by the prolonged action of high temperature with moisture that a substance is formed wliich resembles gum arable; to Zulkowsky (page 127), who by using hot glycerol found erytlirodextrin, achroo- dextrin, and a number of bodies of increasing solubility as the reaction proceeded; to Ost (page 128), who by the use of high moist heat obtained a dextrinous product that was not a true dextrin; to Roessing (page 149), who obtained from starch syrup by the action of dilute hydrochloric acid at high temperatures a modified form of dextrose which he pro- visionally named (^-dextrose; and to Roux (page 149), who prepared from artificial starch dextrins that were almost completely soluble in alcohol. Abnormal dextrins have been reported as existing in certain beers at times, and occasionally peculiar sugars or saccharine bodies have been recorded. Finally, if there be present contaminating enzymes a number of non-saccharine products of varied characters, ranging between sugars and COo and H2O, may be formed. DIFFERENCES IN THE DECOMPOSITION PRODUCTS OF DIFFERENT STARCHES. Sufficient evidence has been presented in Chapter II and in this chapter to justify the conclusion that starch is not a uniform substance, and that it exists in many isom- erous forms which differ in different species, different parts of the same plant, antl even in the individual matm-e grains, etc. It is therefore reasonable to suppose that the higher decomposition products, especially the so-called soluble starch and the dextrins, would be produced in corresponding differentiated homologous forms; in other words, that each form of starch-substance would yield peculiar constitutional forms of decomposition prod- ucts which would bear specific stereochemical relationships to the constitutional structures of the initial substances. Evidence in support of such a conception is found, for instance, in the constitutional differences of the inner and outer parts (the "granulose" and " cellu- lose") of the mature starch-grain as exhibited in the changes which ensue when the grain is subjected to gelatinization and pseudo-solution by heating in water, when comminuted grains are macerated in water, when the grains are subjected to certain aniline dyes and to digestive and various other reagents, and when normal and reverted starches are digested, etc. Wlien raw starch is heated in water to a proper temperature in relation to the kind of starch, swelling occurs, and in the course of time the grain becomes divided iiito two parts, constituting an outer portion that appears in the form of a sac or capsule, and an intra- capsular semifluid substance. After a time the capsule ruptures at one or more places, permitting the escape of the inner substance. By continued heating for some minutes, the length of time varying with the kind of starch, the sac undergoes a complete breaking down. The inner, more soluble, less dense part in case of all forms of starch, which yields a blue or purple or violet reaction in the normal state, yields an intense indigo-blue coloration with iodine, whereas the sacs almost always become a bluish-violet to a red-violet, showing of course a chemical difference between the inner and outer parts. The reaction of the inner DIFFERENCES IN DECOMPOSITION PRODUCTS OF DIFFERENT STARCHES. 155 part is the typical reaction of pure starch as umversally understood, but the explanation of the different reaction of the outer part might not be the same in accordance with vari- able conceptions of starch components and decomposition products. If one assumed that there exists an individual body between starch and erythrodextrins which is intermediate in character, and which possesses properties of both, the violet reaction can readily be accounted for, and as a corollary the outer or capsule part of the grain might with sufficient reason be regardetl as a transitional non-starch substance. There is, however, no necessity for the assumption of the existence of such an intermediate body, but there is sufficient e\idence to warrant the conclusion that such bodies as have been so described are modified forms of starch or mixtures of starch and dextrin. The violet reaction might satisfactorily be accounted for, as experimental observations have shown, and which is in accord with our views of the processes and products in the synthesis and analysis of starch, upon the basis of the presence of a small and variable amount of erythrodextrin in the outer coat. Ery- throdextrin is, as far as we know, the nearest individual to starch, and it is a natural con- clusion that, since the grain grows by external accretion, there may be present in the outer la3'er variable amounts of erytlirodextrin in the course of transition into starch. Erythro- dextrin gives a red reaction with iodine, and starch a blue reaction, and a combination of the two a purple, blue-violet, or red-violet, or intermediate gradations, in accordance with the proportions of these substances. The difference in color reactions between the inner and outer parts of the grain can not therefore of itself be taken to indicate any specific difference in the composition of these parts. There are, however, a number of facts which very strongly suggest constitutional differences, as, for instance, the differences in solubility, the differences in the reactions to anUine and other coloring agents, and differences in the degree of digestibility and in the products of digestion. Wlien comminuted grains are macerated in water the inner part goes into pseudo- solution, while the outer part remains undissolved in the form of suspended flakes, etc., and when raw starch is subjected to the prolonged action of weak acid, etc., the inner part is dissolved, leaving skeletons of the grains which from their polariscopic properties may be totally unchanged. This difference in solubility is in accord with the phenomena observed when the grains are subjected to moist heat, and it would seem that it can not be due in the least to the contaminating erythrodextrin, which is certainly more soluble than starch; nor is there evidence that it may be due to the presence of other substances. Differences in the reactions with aniline and other coloring agents have been referred to particularly on pages 55, 56, and 58 and will be considered further in Chapters IV and VI. Pottevin's experiments (pages 134 and 147) furnish strong evidence of differences in the constitution of the starch-substance in different parts of the starch-grain. He found, as already stated, that the denser external parts of the grain are less digestible than the less dense inner parts, and that the dextrin yielded by them is different, the dextrin frona the outer part being more difficult to saccharify and yielding a very much lower percentage of sugar under the same conditions of experiment, and in being insoluble in strong alcohol, and also differing in the degree of diffusibility. Roux (see pages 46, 112, and 149) recorded that not only were reverted and normal starches different in the degree of digestibility under the same conditions of experiment, but also that the dextrins of reverted starch, in accord with Pottevin, are, unlike those from ordinary starch, freely soluble in strong alcohol. (See on reverted starch, page 111.) Further evidence of differences in constitution that are suggested by variations in digestibility, etc., will be found in Chapter IV. Here it might be noted incidentally that in experiments by O'SuUivan it was found that the percentages of dextrin and maltose yielded by potato starch do not correspond with those from the starches of malt, barley, corn, and rice. It was shown, however, by Ford and Guthrie (page 193) that the differences were due to errors of experiment. 15G DIFFERENTIATION AND SPECIFICITY OF STARCHES. Admitting the existence of stereoisomeric forms of starch-substance, and that a priori there should be corresj^onding specific stereoisomeric forms of dextrins, we meet with a very obvious difficulty in the study of the specificities of these hypothetical decom- position products in relation to genera, species, etc., in the total lack of exact knowledge of the precise dextrinous products of digestion. It is manifest from the literature quoted in this and other chapters that our information as to the 'precise nature of the dextrins formed is nil; that most if not all experimenters have been working with impure bodies or mixtures; that our metliods of distinguishing tlifTerent erythrodextrins and different achroodextrins obtained from the same preparation, if such really have existence, are by no means certain ; and that it is therefore useless under present conditions to attempt the differentiation of corresponding dextrins from the starches of different species, except in a crude and inconclusive way, as, for instance, as done by Pottevin, Roux, and others, by taking the dextrinous products as a whole. THE SYNTHESIS OF STARCH. According to the well-known hypothesis of van't Hoff (Zeit. f. anorgan. Chemie, 1898, XVIII, 1) an enzyme gives rise only to such products in the analysis of a given substance as it will under appropriate conditions combine in synthetizing the same substance. In other words, if a given enzyme or mixtm'e of enzymes acts upon starch to decompose it tlarough a series of actions into a series of products such as erythrodextriu, aclu'oodextrin, maltose, and glucose, it or they will under properly modified conditions reverse the opera- tions and thus synthetize starch. The results of modern research lead unquestionably to the belief that the activities of enzymes underlie vital processes. Various plant and animal enzymes ha^'e been used to break down starch into the products mentioned, and these iDroducts correspond with substances found in plants during the formation of starch ; hence they are presumably intermediate bodies between starch and the immediate antecedents of glucose that are synthetized from carbon dioxide and water. Moreover, syntheses representing at least two of these hypothetical steps in starch formation have been carried out in vitro by Croft Hill (Jour. Chem. Soc. Trans., 1903, lxxxiii, 578) and E. F. Armstrong (Proc. Roy. Soc, 1905, lxxvi, B. 592), in which the maltose-glucose reac- tion was reversed, and by the author (Univ. Penna. Medical Bull., 1910, xxiii, 57; Proc. Soc. Exper. Biology and Med., 1910), in which the starch-erythrodextrin reaction was reversed, with slight evidence of an achroodextrin-maltose reversion and also of a maltose-glucose reversion. Luig (page 148) suspected a glucose reversion. Until comparatively recent years the energies of the chemist in the study of bodies which occur naturally only in plants and animals have been chiefly in the direction of analysis, so that he has worked essentially in a direction opposite to those in living matter in the syntheses of these substances; but during the last ten or twelve years especially he has come in touch with Nature's processes, owing to the eijochal investigation of Croft Hill, with the result of successes in the syntheses of carbohydi-ates, fats, and proteins which scarcely more than two generations ago would generally have been regarded as being im]irobable or impossible; and he has at the same time thrown extremely important light on the analytic and synthetic processes that take jjlace in living matter. The important discovery of Wohler, in 1828, that he had made urea by the interaction of ammonia and lead acetate, is the first on record of the synthesis in vilro from inorganic substances of an organic substance that is inherently peculiar to hving matter. Wohler evidently appreciated the great fundamental importance of his discovery, but it seems that this was far from being the case with his contemi)oraries, and that, as is often the case, an epochal in\'estigation remains wholly unappreciated for years until new inquiries bring it to the forefront. Even his teacher Berzelius failed to realize its importance. The progress in the synthesis of plant and animal substances from this period up to the present has been THE SYNTHESIS OF STARCH. 157 astonishingly slow and insignificant in comparison with the tremendous progress in other lines of cliemical investigation, and largely because of the ^•iolent methods generally emi^loyed in such work. Through ignorance or misconception of the processes in animals and plants the chemist was led too far away from the methods of Nature, but omng prima- rily to the basic work of Croft Hill the way has been opened for an endless amount of investigation which must bring results of incalculable value in the explanation of proto- plasmic processes. Apart from the work of Wohler, it seems that it was the discovery of Butlerow (Ann. d. Chem. u. Phar., 1861, cxx, 195; Compt. rend., 1861, liii, 145) that a saccharine substance (mcthylenitan, C7H 170(3) could be formed by the reaction of a solution of trioxylmethylene and hme-water, that led to Bayer's important statement (Ber. d. d. chem. Gesellsch., 1870, III, 68) that formaldehyde may be formed from carbon dioxide and water, and that theoretically by the polymerization of 6 molecules of formaldehyde there would result a hexose. Various more or less important modifications of Bayer's hypothesis and conceptions have been suggested, particularly by Erlenmeyer (Ber. d. d. chem. Gesellsch., 1877, x, 6.34), Bach (Compt. rend., 1893, cxvi, 1145, 1389; 1898, cxxv, 479), PoUacci (Bot. Centralbl., 1904, xcv, 425, and xc\^, 473; 1905, xcviii, 247), and Usher and Priestly (Proc. Roy. Soc, 1905-06, Lxx^^I, B, 369; 1906, lxxviii, B, 318). Since Bayer's investigations, interest has been aroused as to whether or not formaldehyde is to be regarded as a primary assimi- lative product; whether formal dehj'de can be formed from carbon dioxide and water in vitro; whether sugar can be produced from formaldehyde in vitro; whether a plant can thrive and take up formaldehyde from its ambient medium ; and whether connection can be traced between formaldehyde and the synthesis of sugar, dextrin, starch, and glycogen in the plant. Associated with these inquiries studies have been made, both in the plant and in vitro, in connection with the various saccharine and dextrinous substances which are con- cei\-ed to represent the main bodies in the analyses and syntheses of starch and glycogen. The mere fact that during the syntheses of starch or glycogen there are present intermediate bodies between carbon dioxide and water on the one hand and these polysaccharoses on the other, does not prove of course that such intermediate substances are utilized in the syn- theses, or that starch might not be formed at a single step from carbon dioxide and water, or that because of their seeming absence they may not actually be necessary and that they are made and instantly transformed; but the fact that a serial decomposition occurs in the organism and in vitro, and that the intermediate bodies formed during analysis corre- spond with those found in plants during the syntheses of polysaccharoses, is strong evidence of the formation of starch by the production of a definitely related series of intermediate bodies which show progressively higher and higher molecular weight and complexity of structure, and that such intermediate bodies as are found in plants may be regarded as representing specific steps in analysis or synthesis, in accordance with whether the plant is consuming or making starch or glycogen. If, therefore, formaldehyde can be formed in vitro from carbon dioxide, and sugar formed from formaldehyde, and if at the same time it can be shown that in the plant carbon dioxide and water are primary substances in synthesis, and that formaldehyde is a normal constituent of plants, and that a plant may thrive in a solution or in an atmo.sphere containing appreciable quantities of formaldehj'de and even assimilate this substance, it seems that we have reached the point where it is not a question as to whether or not for- maldehyde is a primary assimilative substance and an intermediate body in the synthesis of certain saccharoses, but to seek for the evidence to prove it. Attempts to produce formaldehyde in vitro from carbon dioxide and water were made by Maly (Ann. d. Chem. u. Pharm., 1865, cxxxv, 119), Royer (Compt. rend., 1870, lxx, 731), Bach (Compt. rend., 1893, cxvi, 1145, 1389; Chem. Centralbl., 1898, 11, 42), Lieben (Ann. d. Phys. u. Chem., 1895, xix, 463), Cohen and Jahn (Ber. d. d. chem. Gesellsch., 158 DIFFERENTIATION AND SPECIFICITY OF STARCHES. 1904, XXXVII, 2836), Berthelot and Gaudechon (Compt. rend., 1890, ex, 1690), Usher and Priestly {loc. cit.), and Fenton (Jour. Chem. Soc, 1907, xci, 687). In the investiga- tion of Bach, Berthelot and Gaudechon, Usher and Priestly, and Fenton formaldehyde was found, and in several forniic acid was noted. Various other substances have been used by chemists in the synthesis of formaldehyde, and other instances might be given which have bearing upon aldehyde formation in plants. If sugar can be reduced to aldehyde in vitro we are justified in assimiing that the reverse may be brought about under proper conditions. This finds confirmation in the investi- gations of Konig, Spieckmann and Olig (Jour. Chem. Soc, 1903, lxxxiv, 386), who found that bacteria (belonging to the type of Bacillus coli communis) in a solution of glucose actually form acetic aldehyde and other compounds; and in those by Renard (Ann. d. chim. e. phys., xvii, 321), who records that glucose when subjected to electrolysis in dilute sulphuric acid yields among its products trioxymethylene, which in turn is changed into formaldehyde, and that a reversal can be brought about so that formaldehyde is trans- formed into sugar. Other successful experiments in the synthesis of sugar in vitro have been reported by a number of investigators, as for instance, Loew (Jour. f. prakt. Chemie, 1886, xxxiii, 321; Ber. d. d. chem. Gesellsch., 1889, xxii, 475), Fischer (Untersuchungen u. Kohlenhy- drate ii. Fermente, Berlin, 1909 — a collection of papers that appeared in the Ber. d. d. chem. Gesellsch., from 1884 to 1908), Lob (Zeit. f. Electrochemie, 1907, xii, 282, Biochem. Zeit., 1907, XII, 78), Euler (Ber. d. d. chem. Gesellsch, 1906, xxxix, 39,45), Slosse (Bull. d. I'Acad. roy. d. Belg., 1906, xxxv, 547), and Berthelot (Compt. rend., 1903, cxxvi, 610). In some of these investigations intermediate bodies, such as a-acrose, a-acrosazone, f-fructose, i-mannitol, i-mannose, i-mannonic acid, rf-mannonic acid, Z-gluconic acid, and d-gluconic acid are recorded in the synthesis of glucose. The evidence, then, is conclusive that not only may saccharine substances be formed from formaldehyde, but also sugar that cor- responds with decomposition products of starch and glycogen. Glucose, as already stated, has been converted into isomaltose. Moreover, having glucose as an initial substance, one may obtain from it by appropriate methods in litro levulose, mannose, and other saccha- rine substances. Of incidental interest is the fact discovered by Mayer (Zeit. f. physiolog. Chemie, 1903, xxxviii, 135) that glycollic aldehyde is eliminated by rabbits in the form of glucose. The fact that aldehydases which oxidize aldehydes have been found in plants and also in the li^'er and other organs of animals is not without significance. Gautier suggested years ago that certain glucosides, such as arbutin and salicin, may be derived from formal- dehyde by the addition of hydrogen by molecular condensation and dehydration. It has also been shown that plant substances may be broken down with the formation of aldehyde, as for instance, salicin into helecin and this into salicylic aldehyde or glucose, and amygdalin into benzoic aldehyde, glucose, and hydrocyanic acid. Acetic aldehyde is produced in acetification, and in wines, etc.; and it is formed in the animal body as a decomposition product. There are quite a number of both plant and animal substances which by oxidation yield forms of aldehyde. In fact, instances of this character which have for their indication a suggestion that aldehydes are among the plant metabolites utilized in carbohydrate and other forms of metabolism might be considerably multiplied. In the animal organism starch is not merely reduced to the glucose stage, but ultimately to carbon dioxide and water. In other words, when diastatic enzymes cease their decom- posing actions other enzymes which cause further hydrolysis, or simple molecular splitting, oxidation, etc., come into play, by the agency of which a considerable number of products are formed, among which are included alcohol and its aldehyde. Whether or not aldehyde is formed in plants is a matter yet under discussion. Evidence in favor of the presence of this substance in plants has been offered m the investigations THE SYNTHESIS OF STARCH. 159 of Loew and Bokoruy (Archiv f. ges. Physiologic, 1881, xxv, 150; xxvi, 50), Reinke (Ber. d. d. chein. Gesellsch., 1881, xiv, 2144), Mori (Nuovo. Gio. Bot. Ital., 1882, xiv, 147), Pollacci (Atti d. iiistit. Bot. d. Univ. d. Pavia, 1900, vi, 45; Atti d. Real. Accad. d. Lincei, 1902, xvi, 199), Grafe (Osterr. bot. Zeitschr., 1906, xlvi, I), Kimpelin (Compt. rend., 1907, cxliv, 148), Usher and Priestly {loc. cit.), Gibson and Titherly (Annals of Botany, 1908, xxxii, 117), Bokorny (Archiv f. ges. Physiologic, 1908, cxxv, 467, and 1909, cxxviii, 565), and Scliryver (Proc. Roy. Soc, London, B., 1910, lxxxii, 226). Opposition to the conclusions of some of the ex])eriments has been offered by Loew and Bokorny {loc. cit.), Plancher and Ravenna (Atti d. Real. Accad. d. Lincei, 149, xiii, 459), Czapeck (Bot. Zeit., 1900, lviii, 153), Euler (Ber. d. d. chem. Gesellsch., 1905, xxxvii, 341), and Bokorny {loc. cit.). Wliile there may be reasonable doubt whether the presence of formal- dehyde in plants in detectable quantities has been conclusively demonstrated, it seems certain that it is formed, and that inasmuch as it is not in the nature of a storage substance it is at once transformed into a saccharine body, and therefore that scarcely more than traces could be expected to be found at any time. Another step in favor of formaldehyde constituting one of the primary metaboUtes used by plants in the elaboration of saccharoses has been demonstrated in the power of a plant to thrive in a medium containing formaldehyde and in its power to assimilate it. Work of this Idnd in showing one or the other or both of those phenomena has been carried out by Bokorny (Ber. d. d. chem. Gesellsch., 1888, xxi, 119; Archiv f. ges. Physiologic, 1908, cxxv, 467, and 1909, cxxviii, 565), Bouillac and Giustiniana (Compt. rend., 1903, cxxxvi, 1155), Treboux (Flora, 1903, lxxxxii, 49), Grafe and Vieser (Ber. d. d. bot. Gesellsch., 1894, xxvii, 431), and Usher and Priestly {loc. cit.). While it has been clearly shown that plants may tlirivc in a medium containing very small quantities of formalde- hyde, and that some of the investigators also show that formaldehyde is absorbed, and also that the plants may even thrive better in an atmosphere free from carbon dioxide if it contain formaldehyde than when it does not, it has only been infcreutially demonstrated that formaldehyde is used in the plant in the synthesis of saccharoses. The several links of the chain constituting the synthesis of starch and glycogen may be formulated hypothetically as follows : Carbon dioxide and water by deoxidation yield formaldehyde, and perhaps other aldehydes; aldehyde by polymerization and atomic re- arrangement yields sugars in the form of monosaccharoses ; monosaccharoses by dehydra- tion yield disaccharoscs; and these in turn by the same process yield polysaccharoses in the form of acliroodextrins; achroodextrins by some form of intramolecular rearrangement yield erythrodextrins ; erythrodextrins by some form of intramolecular rearrangment yield starches or glycogens. Dextrins, glycogens, and starches are closely related polysaccha- roses. Of the members of this chain the substances that are formed for storage purposes, and therefore such as might be expected to be found in large quantities in plants, are sugar, achroodextrin, erythrodextrin, glycogen, and starch. Of these, sugar and starch are quantitatively preeminently important, but in certain plants they seem to be replaced by glycogen or erythrodextrin. This scheme is, of coiu-se, modifiable in many ways, inasmuch as the exact processes and primary substances in starch formation can not be identical in all plants, or probably even at all times in the same plant. For instance, when saccharose is utilized in this synthesis the process can not be identical with that when maltose is used. The formation of saccharose must be on a somewhat different plan, inasmuch as this sugar is of a distinctly different constitution, as is instanced in it not giving many of the sugar tests, in its non-fermentability, and in its not reacting vnih phenylhydrazine. There are various reasons for believing that in the synthesis of starch there are formed thousands or even millions of stereoisomeric forms, each representing a homologue which differs from the others. In the first place, the differences in the plastids of different species IGO DIFFERENTIATION AND SPECIFICITY OF STARCHES. of plants and plant-parts justifies the asraimption of Meyer (page 51) that in accordance with these differences we ha^■e different biologic mechanisms. Hence it would seem to follow, for reasons stated on page 11, that corresponding metabolites should be modified in accordance with the modifications of the mechanisms which produce them. In the second place, it has been shown that saccharose, or cane sugar, can be obtained from starch as one of the products, and that the plant may utilize saccharose in the synthesis of starch. This holds good for maltose and other sugars which have a very difTerent chemical constitution from saccharose. If, therefore, sugars differing so in chemical constitution are used in the synthesis of starch, it seems probable that such constitutional differences tend to be carried throughout the whole series of sj^ithetic bodies and to be present ultimately in the starch or glycogen molecule. One can readily conceive how, through the operations of these two factors (differences in the Itting viechanism and differences in the building material), each species of plant, or even each individual, may produce a specific kind of starch; and, more- over, that this product may be modifiable under normal or abnormal conditions, owing to temporary or permanent modifications of the plastids, or of the food supply, etc. An instance of such a modification is found in the pecuhar form of glycogen found by Claude Bernard in paralyzed muscles — the glycogen of normal muscles yields a reddish or port- wine color reaction with iodine, while the abnormal glycogen yielded a blue reaction. Even after a given form of starch is produced it may undergo a spontaneous change into another stereoisomeric form (see page 9). SUMMARY AND CONCLUSIONS. One can not review the literature of the decomposition products of starches without a realization that our knowledge is exceedingly inexact and that we have scarcely reached the tkreshold of accurate information of the characteristics of the various homologous forms of starch, of the actual substances that are formed as intermediate products during the process of saccharification, and of the precise chemical processes involved in the several reactions. The available methods for the differentiation of homologous forms of starch have been used to but a very limited extent and (as in case of the differentiation of simi- larly related bodies) they are for the most part crude and almost wholly quantitative. The methods of measurement of the progress of the ^^arious steps in the saccharification of starch and the methods of separation of the intermediate bodies and the results of their study are quite as unsatisfactory. Exactly when any one stage begins and ends, when the last molecule of starch disappears and when the first molecule of achroodextrin appears, whether or not there exists any form of dextrin that will reduce copjier solutions or which through some obscure relationship with maltose will affect its reactions with copper, and whether there are multiple forms of erythrodextrin and of achroodextrin, are but a few of the fundamental questions that requu-e the final answer. In a word, the determination of the sequence of events and their relations and the exact products formed remains to be demonstrated, for as yet the only means we have of ascer- taining the degree of progress of enzymic saccharification that has even an approach to accuracy is the estunation of the amount of maltose, as for instance by the copper test, which is the best; but this may be attended by more or less fallacy because of the existence of an achroodextrin or other non-saccharine body which may take part in the reaction, or because of the presence of variable amounts of glucose which has a very difTer- ent reducing value, in addition to other distinctive properties. Finally, the assumption that the processes are those of successive hydrolysis does not take into account the fact that the liquefaction of raw starch or starch-paste is doubtless not one of hydrolj^sis but of adsorption; or that the conversion of starch into dextrin, and of a higher form of dextrin into a lower form, is in effect a depolymerization ; or that oxygen may be essential during a part of or throughout the entire group of reactions. SUMMARY AND CONCLUSIONS. 161 Wiih so unstable a foundation one can not go far in formulating conclusions that can luive more than a tentative value, and with this idea in view the following synopsis of certain especially important points may be made: (1) It may be conceded, from the literature herein cited, that the starch-substance is not a uniform body and that it exists in many stereoisomeric forms which vary in the starches of different plants, of different parts of the same plant, of different grains of the same starch, and even of different parts of the same grains; and that in accordance with these variations corresponding differences may be expected in the intermediate derivatives and reversion products in accordance with the stereochemic differences of the homologues. (2) That the first step ordinarily, though perhaps not essential, in the saccharification of starch is a liquefaction, and that this is followed serially by the production of erytlirodextrin, achroodextrin, maltose or isomaltose, and glucose, all of which processes, when once under way, go on together, the disappearance of one substrate after another eliminating tlie corre- sponding process, until an equilibrium of solution of maltose and glucose is attained; and that not only may the progress of the reaction be stopped at will, but also, by modifications of the processes, the stages of the degradation of the starch-molecule may be so specifically limited that they do not go beyond the liquefaction of starch, or the formation of dextrins, or the formation of maltose; in other words, these substances are the essential end-products of the reactions in the several cases, so that we may have liquefied starch without the forma- tion of any dextrin, or dextrins without a trace of maltose, or maltose without a trace of glu- cose. This serial action, which results in the formation of an individual substance at each step, finally disposes of the theory of Musculus of the coincident formation by hydrolysis of two substances in the form of dextrin and sugar, and their subsequent cleavage; and it is in support of the theory of Lintner and Diill. We may regard these markedly differentiated steps as representing the major stages, and assume that there are a number of substages. (3) That the processes involved in these major stages, as in the case of acids, are due to different functional properties of a single agent; or, as in case of enzymes, either to different functional properties of a given enzyme or to individual properties of different cooperative enzymes, or to both. Sufficient evidence has been offered to show that, at least in some instances, as in the reduction of starch to maltose and of maltose to glucose, two specific enzjines are required, one to yield the maltose and the other to convert this sugar into glucose. It seems likely that four enzymes (or four independent specific prop- erties) are necessary — one to liquefy the starch, one to convert the starch into dextrin, one to convert the dextrin into maltose, and one to convert the maltose into glucose. Were we to follow the decomposition processes in the body to the ultimate conversion of the sugar molecules into CO2 and H2O we should find with certainty that still other enzymes are involved, some carrying on very different chemical processes from those stated. (4) That all of the processes in acid and enzymic saccharification are those of hydra- tion, but not necessarily of hydrolysis. The swelUng or gelatinization and the formation of pseudo-solution and true solution can be fully accounted for on physico-chemical grounds upon the basis of the adsorption of water, which involves absolutely neither enzymic nor acid activity, nor hydi-olysis, and no hydration in the strictly chemical sense. The reversion of starch when in pseudo-solution or true solution to less soluble forms may be regarded as simply a manifestation of a reversal of the adsorption process and accompany- ing changes which we find paralleled in other of the so-called colloids, and which we obser\'e even in Uving colloidal matter, as in the case of the giving off and taking in of water by the muscle substance, which is assumed by some physiologists to be the essential mechanical part of the phenomena of contraction and relaxation. The conversion of starch into erythrodextrin, and of erythrodextrin into achroodextrin, if it be a process of hydrolysis, is of such a character that there is no jjermanent addition of water in the reaction,or, if the product be in the nature of a hydrate, the change is due to a 11 162 DIFFERENTIATION AND SPECIFICITY OF STARCHES. rearrangement of the atoms of the molecules so that in either case the product is of a depol- ymeric form. The conversion of dextrin into maltose, and of maltose into glucose, we are certainly justified in regarding as being due to hydrolysis, by which molecules of water are taken up and constitute not only a part of the process but also a part of the product. Starch is classed among the most typical of the so-called colloids, yet, like many or probably all of the so-called colloids, it may exist in crystalline form; likewise are many simple inorganic substances colloidal, as for instance silicic acid, platinum, gold, sih'er, ferric hydroxide, arsenious sulphide, aluminum hydroxide, etc. In fact, it seems probable from the advances of physical chemistry that any substance, according to conditions, may exist in either a colloidal or crystalloidal state. Hemoglobin in the erythrocyte exists in a colloidal state, yet when freed it is usually easily obtained in a crystalline state ; egg albumin and serum albumin are imder normal conditions typical colloids, yet they may be crystal- lized. Boiled starch is a typical colloid, but when demineralized and rendered by high moist temperature into a true solution it is thus transformed into a crystalloidal state or phase. The transition of colloidal into non-colloidal forms of starch and of dextrin into maltose or some other form or forms of sugar are instances of the readiness with which an alteration from one state to another can be brought about in the transition of intimately related substances, one into the other. It seems therefore, as is being generally recognized, that we are not dealing with colloidal or crystalloidal substances, but with corresponding states, phases, conditions, or forms. Hence, the transition of a typical colloidal state of starch into a crystalloidal state, and the reversal, are phenomena that were to be expected. In fact, the line of demarcation between these two classes of substances, or better, between these two states, is by no means so definitely defined as is generally believed, for not only are there substances which may be so arranged that every transitional stage may be filled in between the most typical colloid and the most typical crystalloid, but also it is found that the transition from one state to the other occurs with such apparent ease that the two states can not be so far separated as it would seem upon superficial investigation. This facility of transition is well illustrated in the ^'arious steps in the saccharification of starch : The starch-grain has a crystalline structure as well defined as the structure of sjjhero- crystals of many typical inorganic crystalloids; by the adsorption of water the crystallized starch is converted into a typical colloid that forms a typical /jscMf/o-solution ; by deminer- alization of the pseudo-solution and subjection to high moist heat, a true solution is formed, and hence the colloidal starch is transformed into a crystalloidal form; by the action of diastatic enzymes the starch in this crystalloidal state is converted into erytlu-odextrin, another colloidal substance normally existing in a colloidal state, but having certain col- loidal properties which distinguish it from starch; by further action of the enzyme this colloid erythrodextrin is converted into another colloid in the form of achroodextrin with the assumption by this substance of a further modification of properties ; from this dextrin normally existing in a colloidal state there is formed maltose, a typical crystalloid; and from maltose there may be derived glucose, which as we are accustomed to see it is a crys- talloid, yet (like iron, etc.) it may exist in colloidal form in combination with certain pro- teins. It goes without saying that, inasmuch as the processes of life are inseparably associated with protoplasm, and hence with the colloidal state of matter, that the interrela- tions of colloidal and crystalloidal states, together with the surprising readiness with which transitions from one state to another can be brought about, are phenomena of the greatest fundamental importance, and that every advance along these lines must bring us closer to the mechanics of protoplasm. Of no less fundamental importance are the transient and permanent effects of changes of external conditions on the constitution of starch, gelatin, fibrin, and such substances when in a colloidal state (page 96), for it seems that in these effects we have a primitive expression of the influence of changes in internal and external conditions upon the constitution of protoplasm, and hence upon its reactions; SUMMARY AND CONCLUSIONS. 163 in other words, that through the evolution of this basic property there have been brought about, and there is being brought about continually in all li\'ing matter, those various modifications of responsivity or irritability which we can readily observe if not explain. (5) That there is no justification for the assumption of the existence of amylodextrin and maltodcxtrinas indi\-idual substances which have properties intermediate between starch and dextrin, and dextrin and maltose, respectively. It seems to ha\^e been shown quite clearly that they are merely mixtures of variable composition depending upon the particular kinds of substances entering into them. For reasons that are perfectly obvious the use of the words amylodextrin and maltodextrin should be discontinued, and especially the former, because of its indiscriminate use in designating soluble starch, special forms of starch, mix- tures of starch and dextrins, etc. Likewise a clearer understanding of the dextrinous products of starch will be had by dropping multiple names for dextrins (nearly all or all of which are names for impure substances and unknown mixtures) and using exclusively erythrodextrin and achroodextrin, until at least these latter bodies have been satisfactorily studied. (6) That under various modifications of experiment abnormal forms of dextrin, and also gums, and also unusual saccharine products, etc., may be formed and that the proc- esses may be carried beyond the sugar stage by the presence of certain enzymes, bacteria, etc. (7) That the products of decomposition by high and dry heat, and by dilute acids and enzymes, and different acids, and different enzymes are not in all respects identical. The dextrins formed by torrefaction, acid action, and enzymic action are not absolutely identi- cal, and especially interesting is the comparison of the products of acid and enzyinic action with each other, and of these with those of torrefaction. In fact, it seems probable, had we adequate methods for differentiating the vast number of hypothetical stereoisomeric forms of starches and of dextrins, that it would be found that corresponding dextrins, as for instance some specific kind of erythrodextrin that is produced by mineral acid, organic acid, enzyme, or heat, would exhibit such differences as to show that we have as many stereoisomeric forms as we have types of means of producing them. In fact, one might go further and hold that the product of the plant enzyme will differ from that of the animal enzyme, and that as plant and animal enzymes differ in essential respects from one another, the product will be correspondingly modified. The conception held by many that dilute acids and enzymes merely increase the velocity of reactions seems dis- proved by the fact that the products of the reactions of these two classes of decomposing agents differ not merely quantitati^•ely but also qualitatively. It seems obvious, if the effects were merely those of an energizer, that no qualitative differences such as have been reported would be noted in the digestions by acids and enzymes. That they increase the velocity of reactions must be admitted, and likewise that any substance under appropriate conditions may act catalytically, but there are certain special properties which are attached to each catalytic body and its modes of action which must not be ignored. (8) Accepting the hypothesis of van't Hoff that an enzyme gives rise only to such products in the analysis of a given substance as it will under appropriate conditions com- bine in the synthesis of the same substance, and coupling this hypothesis with the concep- tion that the synthesis of starch in the plant is essentially fundamentally through the actions of enzymes, it is obvious that accurate knowledge of the processes in plants during this synthesis, and accurate knowledge of the processes in the analysis by the same enzymes in vitro, would be mutually helpful and corrective, the one checking the other. Hence the importance of comparative investigations and of advances in one investigation fore- telUng identical advances in the other. Thus, if in vitro by the agency of plant enzymes sugar may be reduced to aldehyde and this to CO2, and H2O, we may with confidence look for the reverse processes in the plant, which is essentially synthetic in contradiction to enzymic processes in vitro which are under the usual conditions of experiment essentially analytic; but the latter we may have reversed under appropriately altered conditions. CHAPTER IV. THE DIFFERENTIATION OF THE STARCHES FROM DIFFERENT GENERA, SPECIES, ETC. The differences in the histological characteristics of starches from different kinds of plants attracted the attention of some of the earhest workers, and as far back as 1834 Fritzsche (page 64) noted that not only were the forms very various, but often were so characteristic as to indicate the genus or family, or even the species. This observation received more or less confirmation in the investigations of Schleiden and others. In fact, the differences in the forms of the grains of certain of the familiar articles of commerce, such as cereals, beans, peas, potato, etc., are so marked as to have given the impression to superficial observers that the microscopical appearance is typical for each individual starch, whatever its source. That the starch of each species is specific in certain of its characters is obvious from the results of the present investigation, but it would be hazardous, as shown by the records of Niigeli and others (Chapter V), to rely solely, or even often, in any important measure, upon the histological features of the grains. In Plates 13, 51, 68, 75, and 78, figs. 77, 301, 407, 441, and 465, starches are pictured from species belonging to entirely different genera. If such grains were mixed it does not seem likely that the microscopist would detect that it is a mixture, yet such could readily be shown bj^ other means. In other words, the morphological method by itself may be entirely misleading, yet when coupled with other methods may have very great value. By the histological peculiarities it might be possible to state that this or that starch did not come from a certain species or genus, yet it might be impossible to assign it to its proper source, because of its seeming histological identity with certain other starches which may not have even the remotest relationship. The investigations along various indeijendent lines of inquiry have demonstrated that we have a number of means which collectively will not only enable us to differentiate the starches of different genera, species, varieties, and hybrids, but also of different parts of the same plant, indicating thereby a specific biologic relationship between the peculiarities of the starch-grains and the specialized plastids which form them, so that even in the same individual, if there are a number of groups of these starch-forming cells, each group being differentiated from the others, each will produce a form of starch which logically should have individual distinctive histological characteristics; but such characteristics would not of themselves necessarily imply stereochemic differences. Comparative investigations of starch have been pursued by various methods, such as the following: (1) The histological method, by which are studied the form and size of the grains, posi- tion and character of the hilum, characteristics of the lamella?, orientation, etc. (2) The proximate constituents and other features as regards general chemical compo- sition. (3) Color reactions with various reagents. (4) Reactions with aniline dyes. (5) Reactions with swelling reagents. (6) Temperature of gelatinization. 165 1G6 DIFFERENTIATION AND SPECIFICITY OF STARCHES. (7) The refractive index. (8) The reaction in polarized Ught. (9) Characters of the starch-paste and starch-solution. (10) The phenomena of digestibility. THE HISTOLOGICAL METHOD. The findings by the histological method ha^'e been referred to in the preceding chapters and will be (jnitc fully reviewed in subsequent chapters, especially in Chapter V, so that nothing need be stated at this place. PROXIMATE CONSTITUENTS AND OTHER FEATURES OF GENERAL CHEMICAL COMPOSITION. THE PROXIMATE PRINCIPLES. Starch-grains consist approximately of from 80 to 85 per cent of starch-substance, 15 to 20 per cent of water, and small and variable amounts of organic and inorganic substances, including fat, cell residues, cholesterin, dextrin, sugars, protein matters, phosphates, tannin, copper, and other substances. The figures in table 14 are by Konig (Die manschl. Nahrungs- u. Geniissmittel ; Die Rohstoffe des Pflanzenreiches, Wiesner, Leipzig, 1903, Bd. I, 580). (See Jessen, i^age 29.) Table 14. Kind. Starch substance. Water. Ash. Fat. Cell residues. Wheat 83.3 84.1 84.1 82.8 84.8 79.6 80.8 14 14 15.7 12.9 14.4 19.2 17.2 0.4 0.4 0.2 0.4 0.25 0.3 1.0 0.2 6.1 ' b!o4 0.3 0!05 ai Arrowroot Sago Potato (a) Potato (ft) Commercial starches, and starches as ordinarily prepared in the laboratory, contain large percentages of water and more or less impurities that can be removed by succes- sive treatment with dilute hydrochloric acid, alcohol, and ether, and subsequent drying. Salomon (Reportorium f. Analyt. Chem., 1881, 274; Jour. f. prakt. Chemie, 1882, xxv, 348, and XXVI, 324) recommends that desiccation be carried on at 120°, because at lower tem- peratures water is retained, while at higher temperatures the starch becomes discolored and decomposition processes set in. THE STARCH-SUBSTANCE. The starch-substance may be regarded, in the light of the more recent investigations, as consisting of a number of modified forms of a single substance, and it is probable that not only are certain of these modifications peculiar to the species, etc., but also that vari- ations exist in the kinds and proportions of these modifications in different grains, and even in a given grain from the starch of a given plant. Meyer {loc. cit.) records that the per- centages of a-amylose differ in different starches, as follows: potato 0.6, rice 0.9, corn 1, wheat 0.5, and arrowroot 2.5. Day (U. S. Dept. Agriculture, Office Expt. Stat. Bull. 202, 1908, 40) states that bhie-amylose represents the entire inside of potato, arrowroot, tapioca, and sago starches, and 90 jier cent or more of wheat, corn, rice, and barley starches ; that red-amylose constitutes the outer layer of starch-grains ; and that rose-amylose forms about 10 per cent of the inner part of wheat, corn, rice, and barley starches, but is absent from potato, arrowroot, tapioca, and sago starches (see page 58). See also references to the investigations of Kraemer, Salter, Denniston, and others in Chapter II. THE WATER OF STARCH. 167 Table 15. Kind. Wheat Rye Potato Corn Buckwheat. Rice Acorn Air containing Air containing 73 per cent 100 per cent of moisture. of moisture. 6.94 18.92] hi S ■ 10.01 19.36 is 10.33 20.92 10.53 19.55 1 <= n 10.85 20.02 10.89 10.84 U 11.96 22.98J £■" THE \V.\TER OF ST.4RCII. Starch is very hygroscopic, and the per cent of water is variable in relation to temper- ature, the amount of moisture in the air, and the character and amount of foreign sub- stances present. According to table 14 the percentages range from 12.9 to 19.2 for starches of different kinds, sago starch having the lowest and preparation a of potato starch the highest. Payen found that potato starch, after exposure for several days to an atmosphere saturated with moisture, contained 35 per cent of water. The freshly prepared starch may have 45 per cent of water. Soxhlet (Centralb. f. Agrikultur-Chemie, 1881, 554) recorded for air-dried potato starch 20 per cent of water, and for wheat and corn starch 16 per cent. Salomon (Jour. f. prakt. Chemie, 1882, xxv, 348; xxvi, 324) gives for air-dried rice starch 22.98 per cent of water. Saare (Zeit. Spiritu- sind., 1901, xxiv, 502, 512) found for wheat starch 9.9 to 15.3 per cent (mean 13.2) and about the same for corn. Hoffmann and Philippe (Woch. f. Brauerei, 1905, xxii, 71) gi\-e for air-dried potato starch 14.38 per cent. According to Dubosc (Dingl. Polj't. Jour., 1892, ccLxxxv, 213) the starch of the sago palm contains 12 per cent of water. Nos- sian (Jour. f. prakt. Chemie, 1861, lxxxiii, 41) showed the marked hygroscopic proper- lies of starch by subjecting starches from various sources, that had been dried at 100°, to atmospheres containing 73 and 100 per cent of moisture respectively at 17° to 20° (table 15). Archbold (Jour. Soc. Chem. Industry, 1887, vi, 83) gives the following data: Starch dried at 100° in vacuo is completely dehydrated. Starch dried at 15.5° in vacuo contained 10 per cent of water. Stai-ch dried at 20° in air containing 0.6 per cent of moisture contained 18.6 per cent of water. Stai'ch dried in air saturated with moisture contained 35.7 per cent of water. Carl Nageli (Die Starkekorner, etc., loc. cit.) noted that heat is de\'eloped upon the addition of water to drj^ starch. Fischer (Beihefte z. botan. Centralbl., 1902, xii, 227) states that the phenomenon is due to the chemical affinity of the starch-substance for water. UlUk (Zeit. f. d. ges. Brau., 1891, 565) dried potato starch at 120° and found that it had contained 12.1 per cent of water. Starch after having been exposed in an atmosphere saturated with moisture at 16° to 20° contained 37 per cent of water, and when mixed with water no rise of temperature occurred. When placed in an indifferent solution, such as a solution of sucrose, anhydrous starch took up 23.75 to 24.58 per cent of water, while the air-dried preparation absorbed from 18.98 to 19.1 per cent. The drier the starch the larger the quantity of water taken up and the greater the rise of temperature. Ullik, in experiments with 20-gram quantities of dried starch mixed with the same weight of water, recorded the changes in temperatures shown in table 16. Table 16. Preparation. Increase of temperature. 1. Anhydrous starch, dried at 120°. . . 2. Starch, dried at 90° 13.8° 12.0° 8.8° 3.0° 3. Starch, dried over sulphuric acid . . 4. Starch, air-dried Ullik's results received confirmation in the investigations of Hoffmann and Philippe {loc. cit.) and of Emslander and Freulich (quoted by Hoffmann and Philippe). 168 DIFFERENTIATION AND SPECIFICITY OF STARCHES. THE ASH. The phosphates of the starch-grains have attracted especial attention because of their giving, it is believed, an acid reaction to certain kinds of starches ; because of their assinned effect upon the rapidity of the decomposition of starches by enzymes ; and, finally, because of significant different percentages in starches from different sources and in different grains and parts of the same grain. The relation of the phosphates to the reaction of the grains will be referred to in a subsequent paragraph. The asserted effect upon enzymic processes has been a matter of dispute. Effront (Enzymes and their Applications, trans, by Prescott, 1902, 118) records that the addition of 0.5 gram of acid calcium phosphate to 100 c.c. of boiled starch increased the sac- charifying power of an infusion of malt 5.3 times, and that 0.7 gram of ammonium phosphate caused an increase of 6 times. It has been noted by a number of investigators that the addition of very small amounts of certain acids or acid salts is favorable to enzymic activity. Fernbach (Woch. f. Brau., 1900, xvii, 35; Jour. Soc. Chem. Ind., 1900, xix, 260) notes that Ling and other investigators found that diastatic activity is increased by small quan- tities of acid and that a distinction must be made between acid salts and free acids. In comparative experiments with increasing additions of acids, it was found that except- ing small quantities of acid phosphate the effect was injurious. Phosphates, he records, have a specific action on the velocity of saccharification. Neutral (dibasic) phosphates are alkaline to methyl orange, and they have a harmful effect on saccharification. Dias- tase was found to act best in a medium that has a neutral reaction to methyl orange, therefore acid phosphates (monobasic) are favorable, while free acid is harmful. Fernbach and Hubert (Compt. rend., 1900, cxxxi, 293) found that the primary action of alkaline phosphates is to increase the activity of malt diastase, and the secondary action is to decrease activity. They ascertained that during both primary and secondary actions the reaction of the preparations was acid to phenolpthaleine and alkaline to methyl orange. Ford (Jour. Soc. Chem. Industry, 1904, xxiii, 414), in reviewing the subject, states that in his opinion the favorable effect of the acid is not due to the acid alone, but to its neutralizing alkaline impurities, which have a powerful inhibiting effect on diastatic action, and that when these are more or less neutralized the action approaches its normal maximum, which takes place in a neutral medium, or at least one in which the free hy- drions or hydroions are at a minimum. Ford found that ordinary preparations of soluble starch invariably contain phosphates and possible traces of organic phosphorus compounds which are not removed by prolonged treatment with dilute alkali or acid, or washing with water. He holds that the acid phosphates constitute a negligible amount of acidity to starch-solutions, and that they may be looked upon as neutral salts. Maize, wheat, and rice starch contain less or are more easily freed from phosphorus compounds than potato starch, and Ford states that he had prepared from maize specimens which are close approximations to the pure substance. An especial degree of importance seems to have been attached to the phosphorus content of starch by the investigations of Fernbach (Compt. rend., 1904, cxxxviii, 428), who found that large grains and small grains of potato starch differ materially in the percentage of phosphorus. For every 100 grams of large and small grains he found the following proportions of PjO^, expressed in milligrams: 160-199, 143-158, 159-185, 160- 194, 178-226, and 138-215, the proportions in the larger heavier grains to the smaller lighter grains being 100 : 110 to 115. Regarding the smaller grains as being young grains (see page 180) he arrives at the conclusion that the mature starch-grain consists of a nu- cleus that is relatively rich in phosphorus which becomes during the course of growth covered with laj^ers free from j^liosjihorus, and which constitute the larger portion of the grain. The differentiation of the small young grains from the later deposited layers has also been shown by means of aniline dyes (page 57). THE REACTION OF STARCHES. 169 THE REACTION OP STARCHES. The reactions of commercial starches differ wadely, owing chiefly to different methods of preparation and variable contaminations. Soxhlet (Reportorium f. analyt. Chemie, 1881; Jahr. ii. d. Fort. d. Thierchemie, 1881, xr, 86), by titrating commercial starches wdth a standard solution of sodium hydroxide, determined that for every 100 grams of potato or wheat starch the acidity was equal to O.OG to 0.4 gram of sulphuric acid. For potato starch from 15 to 85 c.c. of a N/lO NaOH solution, and for wheat starch from 13 to 19 c.c. were required for neutralization. Rice and corn starches were, on the other hand, usually markedly alkaline. Meyer (Die Stiirkekorner, loc. cil.) states that the reaction of commercial starches varies, generally being acid, but sometimes alkaline. He I'ecords that he has not found a neutral preparation. Saare (Zeit. Spiritusind, 1901, xxiv, 502) noted that both wheat and corn starches are acid, wheat being the more acid. Ford (Jour. Soc. Chem. Ind., 1904, xxiii, 414) observed that the reactions of starch preparations, as determined by color indicators, are variable in accordance with the indi- cator. A solution of ordinary starch may be neutral to rosolic acid, acid to phenol- pthaleine, and alkaline to methyl orange. Therefore, in referring to starch as being neutral, acid, or alkaline, it is necessary to define in what sense the word is used. Litmus paper and litmus solution, he states, are useless for testing starch. A Table 17. number of starches from different sources were prepared bj' succes- sive treatment with alkali and di- lute acid, well-washed, and dried. Given quantities (15 to 20 grams) were gelatinized by heat, and upon cooling were liquefied at 79° to 80° by a trace of precipitated diastase, and then boiled. The preparation was made up to 500 c.c, then 1 c.c. of malt extract was added to each 70 c.c. of prepara- tion and the mixture digested at 40° for an hour, and then boiled and made up to 100 c.c. The reactions of 10 grams of preparation were found to be as given in table 17. From similar preparations of the same starches made by Lintner's method, the results appearing in table 18 were recorded. Table 18. Kind of starch. KeactioQ with rosolic acid. Arrowroot, Natal Neutral Acid; 0.17 c.c. N/100 NaOH Acid; 0.17 c.c. N/lOO NaOH Alkaline; 0.10 N/100 H.SO., Neutral Acid; 0.17 c.c. N/100 NaOH Alkaline; 0.30 c.c. N/100 HjSO, Neutral Rice Wheat Corn Potato Barley .4rrowroot, uiipurified Arrowroot, Lintner's soluble Kind of starch. Reaction with rosolic acid. Reaction with phenolpthaleine. Arrowi'oot « Rice Wheat Acid; 0.8 c.c. N/100 NaOH Acid; 3.7 c.c. N/100 NaOH Acid; 5.2 c.c. N/100 NaOH Acid; 4.6 c.c. N/100 NaOH Acid; 1.5 c.c. N/100 NaOH Acid; 0.5 c.c. N/100 HiSO, Alkaline; 1.0 c.c. N/100 H.SO, Alkaline; 2.0 c.c. N/100 H2SO4 Acid; 6 c.c. N/100 NaOH Acid; 27 c.c. N/100 NaOH Acid; 27 c.c. N/100 NaOH Acid; 25 c.c. N/100 NaOH Acid; 15 c.c. N/100 NaOH Acid; 5 c.c. N/IOO NaOH Acid; 10 c.c. N/lOO NaOII Com Potato Arrowroot t^ Arrowroot )' Potato I The acid property of starch was noted by Blondeau (Compt. rend., 1864, lix, 403), who states that starch forms with ammonia a combination having the projierties of a weak base. Rice starch demineralized by dilute hydrochloric acid and washed free from acid was found by Demoussy (Compt. rend., 1906, cxlii, 933) to have properties of a weak acid comparable to carbonic acid. He reports that it forms compounds with metallic hj'droxides, ammonia, and alkaline carbonates which are dissociable b}' water. He also 170 DIFFERENTIATION AND SPECIFICITY OF STARCHES. Table 19. noted that it has the property of absorbing neutral salines, etc. The asserted acid char- acter of starch is in accord with the observation of Ford and Guthrie (Chem. Soc. Trans., 1906, Lxxxix, 76), who state that soluble starch freed from impurities possesses feebly acid properties, as is shown by the fact that it is capable of combinino; with sodium hy- droxide to such an extent that the molecular conducti\'ity of the soilium hydroxide solution is reduced very considerably upon the addition of soluble starch. MISCELLANEOUS CHEMICAL DIFFERENCES. The proportions of insoluble residues are variable in relation to different kinds of starches, as will be noted by table 13 on page 166. W. Niigeli (Die Stiirkegruppe, etc., loc. cit.) states that potato starch leaves a larger quantity of insoluble residue than wheat starch, which is the reverse of the figures in Konig's table. Parow and Neumann (Zeit. Spiritsind., 1907, xxx, 561) worked out a simple method for the determination of the quantity of starch in cominercial products. One of the best-known methods is to digest the starch with malt extract, finish saccharification by dilute hydrochloric acid, and deter- mine the dextrose by Fehling's solution, using the theoretical factor 0.9 for calculating the dextrose into terms of starch. Parow and Neumann not onlj^ found this factor too low (0.95 being better), but also that the deter- mination could be made by a much simpler method by which dextrose can be determined polariscopically. In these experiments they as- certained that each of the starches examined has a specific factor, the differences depending upon degrees of saccharification, which of course are indexes of the quantities of undigested residue, were recorded. Parow (Zeit. Spiritusind., 190S, xxxi, 286) states that corn starch contains 0.44 per cent of protein, potato starch 0.26, and cassava starch 0.2.3. Little or no importance is to be attached to such data beyond the fact that they indicate the degree of contamination rather than specific differences in the composition of starches from different sources. Various laboratory procedures were pursued by Bloemendal (Pharmac. Weekbl., 1906, XLviii, 1249; Wochschr. f. Brau., 1909, xxxiii, 436; Chem. Ab. Amer. Chem. Soc, 1909, III, 1102). He studied the properties of potato, rice, wheat, and arrowroot starches as regards chemical composition, content of pentosans, heat of combustion, and specific gravity. The C and H contents of the four starches were remarkably constant, varying from 44.03 to 44.54, and 6.02 and 6.45, respectively, these figures corresponding to the familiar mean values, C44.4 Hq.o O49.4. The C and H values were the same in the small grains as in the large. The furfurol contents in the form of phloroglucids obtained by boiling with hydrochloric acid were 0.025 for arrowroot and 0.032 for rice for each 2.5 grams of starch. The quantity of pentosans was very small. Unimportant differences were noted in specific gravity, water-content, and heat values. Kind of atarch. Polarization factor for dry starch. Soliel-Ventzke scale. Circular degrees. Potato 2.872 2.938 2.944 2.918 8.288 8.478 8.497 8.420 Corn Rice Wheat The factors given in table 19 COLOR RE.\CTIOXS. IODINE. The use of iodine as a color-test for starch dates from the earlier observers, and the blue reaction is regarded as being specific, although Claude Bernard showed that an ab- normal glycogen may be formed in paralyzetl muscles wliich may give the same reaction, while others have reported that some forms of plant mucus and of cellulose also gave a blue reaction. The rapidity and intensity of the reaction to starches from different sources COLOR REACTIONS. 171 and also to different grains of the same starcli, and even of parts of the same grain, ha\e h)ng been known. Boiled starch and soluble starch with very rare exceptions (p. 53) alwaj's yields a blue reaction, but starch-grains, and the grain-sacs after boiling starch, may give a blue, violet, or reddish reaction, depending upon the source, or upon the amount of iodine used and other conditions. It is quite common, after boiling, for the capsidar part of the gelatinized grain to give a violet reaction, while the inner part becomes an indigo blue. A \\'eak iodine-iodide of potassium solution is usually employed, preferably a 1 to 2 per cent Lugol's solution. The differences in the behavior of starch, erytlirodextrin, and acliroodex- trin with iodine has rendered this agent of gi'eat value in the recognition and differentia- tion of these substances. The relati\'ely greater intensity of the blue reaction over the red with a mbiture of starch and erythrodextrin may cause the color reactions of the latter to be masked, unless the iodine be used cautiously. IODINE VAPOK. The use of iodine vapor to distinguish different kinds of starch was adopted by Dubosc (Chem. Zeit., 1904, xxviii, 1149). Iodine crystals are placed on a watch-glass which is put on a glass plate, on which also is the starch, and the whole is covered with a bell-jar and set aside for 24 hours. Corn starch was colored a blackish-violet, wheat starch a bluish-gray, sago starch a brownish-gray, and potato starch a yeUowish-gray, the intensity of the yellow being proportional to the amount of foreign matter. IODINE-CHLORAL HYDRATE. A modification of the iodine reagent as ordinarily employed is prepared by saturating a saturated solution of cliloral hydrate with iodine. The chloral hydrate usually causes swelling and bursting of the starch-grains, which is coupled with the iodine reaction. Green (The Soluble Ferments and Fermentation, Cambridge, 1901, 59) made use of this reagent to trace certain changes in the reserve starch in germinating pollen-gi-ains during the develop- ment of the pollen-tube. Green wTites that the plant whose pollen gave the most satisfac- tory results was Lilium pardalinum. The ripe pollen-grains when treated with a solution of iodine-chloral hydrate, are rendered transparent and the starch-grains are stained blue. Mixed with them here and there were a few grains staining like erythrodextrin. As the tube was put out from the grain these granules were gradually carried over into the pro- truding portion, and they flowed slowly down the tube as it extended. When the tube was as long as twice the diameter of the grain, if the iodine-chloral hydrate solution was added, the grains were found to be somewhat different in color, becoming sUghtly purple with iodine. With longer tubes, the grains, still traveling forward, showed this change more and more markedly, particularly near the tip of the tube. When a tube which had attained a length of 20 to 30 times the diameter of the grain was treated in the same way, the general effect of the iodine was very different. There were but few blue granules, which were in the part nearest the poUen-grain. The greater part of the length of the tube was studded with purple grains, and towards the tip they became nearly red. The starch was evidently in the process of digestion under the action of diastase, which other experiments had shown the same pollen to contain. The starch-grains did not change their shape nor show signs of corrosion even when seen under high magnification. The amount of cliloral hydrate was not sufficient to cause the swelling that is observed with strong solutions. IODINE-LACTIC ACID. By the addition of iodine to hot syrupy lactic acid, Lagerheim (Svensk. Farm. Tidskr., 1901; Jour. Soc. Chem. Ind., 1901, xx, 1245) prepared a reagent which, like the iodine- chloral hydrate combination, renders the plant tissues transparent and at the same time colors the starch, by which means he proposes not only to determine the presence of starch in situ, but also to differentiate between exhausted and non-exhausted tea-leaves. 172 DIFFERENTIATION AND SPECIFICITY OF STARCHES. PHENOLS WITH SULPHURIC OR HYDROCHLORIC ACID. It was found by Ihl (Chem. Zeit., 1887, xi, 19; Zeit. d. Alpincn. Oesterr. Apotheker- Vereins, 1888; Jour. Soc. Chem. Ind., 1887, vi, 306, and 1888, vii, 511) that phenols with sulphuric or hydi-ochloric acid give with carbohych-ates brilliant color reactions. If starch be moistened on a watch-glass with an alcoholic a-napthol solution, and if then a few drops of warm sulphuric acid are added, the starch is colored red-violet. Thymol, cresol, guaiacol, and pyrocatecol produce a splendid vermilion-red; resorcinol and orcinol give a yellow- red; whereas phloroglucinol gives a yellow-brown. The diffei'ent kinds of gum behave on the whole like starch. The actions of phloroglucinol on arabin is very characteristic. If arabinose be boiled with an alcoholic solution of this substance and hydrochloric acid, a fine cherry-red is produced. REACTKlXS WITH ANILINE DYES. Starch seems to have affinities for many aniline dyes. Investigations in this connec- tion have been referred to particularly on pages 55 to 58, and sufficient has been shown to prove that starch from different sources and even parts of the same grain do not react identically with a given stain. REACTIONS WITH VARIOUS AGENTS, WITH ESPECIAL REFERENCE TO THE DEMONSTRATION OF THE STRUCTURE AND COMPOSITION OF THE STARCHES FROM DIFFERENT SOURCES. Quite a number of reagents, such as potassium hydroxide, tannin, sulphuric and other mineral acids, acetic acid, concentrated solution of chloride of zinc, glj'cerine, etc., were used by the investigators prior to the seventies in studying the properties of starch-grains. Meyer (Die Starkekorner, loc. at.) made use of a number of such reagents besides enzymes, to iiring out the microchemical properties of a-amylose, /3-amylose, and amlylodextrin, including among these potassium hydroxide, calcium nitrate, 25 per cent hych-ochloric acid, and 3 per cent acetic acid. Kraemer (Jour. Amer. Chem. Soc, 1899, xxi, 650; Amer. Jour. Pharm., 1899, 174; Botanical Gazette, 1902, xxxiv, 341) made a comparative study of the structure of wheat, corn, and potato starches by means of the following agents: 1. Choral-iodine+iodine solution (equal parts of each). 2. Chlor-zinc-iodide solution. 3. Chloral solution (saturated), water, and glycerine (equal parts of each, to which iodine is added to saturation). 4. Calcium nitrate solution, 30 per cent. 5. Chromic acid solution, 15 per cent. 6. Saliva. 7. Taka-diastase. 8. Silver nitrate solution, 2 per cent. 9. Sulphuric acid, C. P. with 10 per cent of water. 10. Sodium acetate solution, .50 per cent. 11. Potassium hydroxide solution, 0.1 per cent. 12. Potassium nitrate solution, saturated. 13. Potassium phosphate solution, saturated. 14. Tannin solution. 15. Hydrochloric acid, 5 per cent. 16. Water. The results of his comparative study of wheat and corn starch are presented in the following quotation : Wheat. (1) Chloral-iodine +iodine solution causes the grains to become at first uniformly blue in color; swelling of the grains soon takes place and finally alternate light blue and blue layers are observed. (2) Chlor-zine-iodine behaves similarly to the pre- ceding reagent. Corn. (1) Chloral-iodine +iodine solution causes some of the grains to swell in 5 hours and others to show a tricheten arrangement of the layers; the grains do iiot appear to be swollen to the extent that the wheat grains are, and therefore show apparently a deeper color with the iodine. (2) Chlor-zinc-iodide brings out immediately the fissui'es or i)oint of growth, whirli is in marked contrast to the wheat starch; in the course of several hours the grains swell at one end, the portion showing the .swelling becomes light Ijlue, and finally almost colorless, while the other portion remains of a deep blue color; some of the grains finally disintegrate into several portions. REACTIONS WITH VARIOUS REAGENTS. 173 Wheat. — Continued. (3) Choral ;uul (glycerine .solu(ion behaves similarly to No. 1, but till' Krain is not coloreii aud the lacunar or fissures are more pronounced. (4) Calcium nitrate produces in 15 minutes a strong corrosion of some of the grains, and those not acted upon by the reagent in Ibis manner swell in 1 hour very ])erccptibly, then show a trichetcn-like* dcvelo])ment, and in 5 hours swell enormously, and finally burst. *IU-ferring to the trichites of Meyer (p. 77). (5) Chi'omic acid produces a similar eflfect upon the grains, but appears to be more ijronounced in its action. (6) Saliva causes in some grains the develoj)ment of prominent radiations and lamellae, in others a tricheten- like structure is developed; in 5 hours the grains give inihcations of corrosion, which in the course of 17 to 24 hours is verj' jironounced. (7). Taka-diastase, on the other hand, acts very slowly in comparison. In .'5 hours there is little or no effect observable, the trichetcn-like structure developing after this length of time and corrosion finally taldng place. (S) Silver nitrate has but little action upon the grains at first; in 5 hours the tricheten-likc development ap- pears, and later the grains swell and disintegrate. (9) Sulphuric acid acts almost immediately, causes the grains to become nearly transparent and irregular in outline, and a rapid solution takes place. (10) Sodium acetate causes some of the grains, in the course of 5 hours, to swell and others to become very much corroded. (11) Pofassium hydroxide very soon produces a swell- ing and rupture of some of the grains and in others the development of a prominent trichcten-like structure, and finally in both a slow corrosion. (12) Potassium nitrate causes almost immediately a swelhng and rupture of the grains, or a strong corrosion. (13) Potassium phosphate produces prominent fis- sures with the subsequent development of rather numer- ous tricheten-likc layers in some grains, in others there is a sweUing and rupture of the grains, with finally a gradual corrosive action on both. (14) Tannin produces a swelHng of the grains, to- gether with the development of rather large irregular lacuna;, and in 5 hours the grains become very much swollen and of irregular shape, after which disintegration and solution takes place. (15) Hydrochloric acid causes the appearance of prominent tricheten in some of the grains in a few hours, in others there is a tendency to swell, and both kinds finally divide into two or more parts. (16) Water at a temperature between 50° and 70° C. produces a marked effect upon the grains; those digested at a temperature between 50° and 55° C. for several hours, are swollen, and in many cases even ruptured; at 00° C. they show a prominent tricheten-like structure, which is scarcely visible at 65° C. and after digestion at 70° for 1 hour the grains become very irregular and swollen and are apparently not further affected by a temperature between 70° and 95°. Corn. — Continued. (3) Chloral and glycerine solution causes lenticular, somewhat irregular, or more or less star-shaped antl prominent lacuna; or fissures, and in the course of 21 hours in some grains jirominent radiations arc developed, whereas in others a marked swelling takes place. (4) Calcium nitrate makes the point of growth more visible as with pre-vious reagents, then strong radiations or a Irichcl en-like 8truct,ure develops in some grains, whereas in other grains the fissures develop into large radiating canals, which extend to the margin of the grain, the swelling continuing so that in 5 hours only the outline of the gi'ain is visible. (5) Chromic acid causes a prominent swelling of the more or less star-shaped point of growth, wliich continues to such an extent in some cases as to produce a rupture of the grain at one of the angles; in other cases there are numerous radiations, or a tricheten-like structure, devel- oped around the swollen fissures, which finally disappear as the grain swells and breaks down. (6) Saliva acts upon the grain very much like chromic acid and calcium nitrate, only instead of a swelling of the grain we have a rather slow corrosion in the course of 48 hours, following the pronounced development of fissures. (7) Taka-diastase behaves like saliva, only the cor- rosive action is more rapid. (8) Silver nitrate causes the formation of ijrominent and angular fissures, which become more or less circular in outline, aud near the periphery prominent radiations may develop. (9) Sulphuric acid produces in some of the grains marked development of an angular fissure wliich becomes circular to radiating in outline, whereas in others a corrosive action appears to begin at the periphery of the grain, followed by gradual solution of the entire grain. (10) Sodium acetate behaves very much like calcium nitrate. (11) Potassium hydro.xide acts similarly on the grain to calcium nitrate and sodium acetate. (12) Potassium nitrat<; differs very markedly in its action on corn from that on wheat, in that there is a strong development of radiating fissures which extend in many cases to the periphery, whereas in wheat there is a more pronounced swelling of the grains and an irregular corrosive action. (13) Potassium phosphate causes the development of a prominent lenticular or star-shaped fissure, which increases in size and in 17 to 24 hours there is a complete breaking down of the grain. (14) Tannin causes also the immediate production of prominent fissures which in 5 hours develop into largo canals, or circular portions, and there is finally a sep- aration of the grain into several parts. (15) Hydrochloric acid causes in some cases the development of large star-shaped or lenticular fissures, and in other cases in the course of but 20 minutes there is a marked swelling of the grain at one point, which continues until disintegration takes place. (16) Water between the temperature of 50° and 70° C. causes certain chai'acteristic features to be developed; the grains when heated for 90 minutes at 50° C. develop in most cases a rather pronounced circular fissure; on other grains there may be a swelling or lenticular or star- shaped fissure; at 55° to 60° C. the swelhng of the grain is more pronounced, and at 65° C. the remainder of the grains show a marked one-layered tricheten-like structure ; at 70° C. the markings have disappeared and the grains have become swollen to angular and irregular masses. 174 DIFFERENTIATION AND SPECIFICITY OF STARCHES. Potato. The boliavior of potato starch toward these reagents may be briefly summarized as follows: The first effect of chromic acid and other swelling reagents is to make the lamellEE more distinct; this is fol- lowed by the development of the crystalloidal character of the lamellae, which is most jironounced in those colored blue with iodine; this is fol- lowed by the production of small tracts or channels connecting the contiguous lamella;, particularly in the middle of the grain; succeeding this there is the formation of chan- nels which are larger and plume-like in appearance, the grain meanwhile swelling quite perceptibly, the mid- dle portion becoming clearer antl assuming a zigzag outline, between which and the periphery of the grain a nvmiber of crystalloidal lamelkc arise; the grain now becomes spheri- cal and marked by a number of con- centric lamellre near the periphery; the latter finally ruptiu-es and then follows a gi-adual solution of the gram, the jjeripheral layer sometimes recurving like the cutin layer of an epiilermal cell on treatment with sulphuric acid. Wheat. In wheat starch the develoiiment of the crystalloidal character of the lamella; is followed by the formation of narrow, interrujited or continu- ous, radial channels near the peri- phery of the grain, which are some- times connected with lamelte occur- ring near the midfUe of the grain; the grain meanwhile swells quite perceptibly, the center becomes clearer, the contents are crowded into crescent-shaped halves, wliich are still connected at the poles; the contents of each of the halves of the grain consist of crystalloidal lamella; in which are then produced small tracts or channels connecting the continuous lamellse; the halves in some instances finally separating and slowly dissolving. In some cases, on the other hand, there is a corrosion of the grain at the periphery, fol- lowed by gradual disintegration without the separation into halves. Corn. The first effect of reagents upon the corn starch-grain is to bring out the point of origin of growth; the latter becomes larger and in some cases more or less zigzag in outUne; between this and the peri- phery of the grain arise more or less interrupted or continuous radial channels (usually the latter); the crystalloidal structure of this grain develops slowly and is most i)ro- nounced when the grain has swollen to two or three times its normal size; at tMs stage we find that the center of the grain ha.s become clear and the point of origin of growth has become obliterated in some cases, and between it and the periphery occur numerous crystal- loidal lamell® similar to those ob- served in the potato starch; finally the peripheral layer ruptures and there is a gradual disintegration of the grain. Sometimes the grain appears to separate into as many parts as there were arms to the point of origin of growth, particu- larly when acted upon by saliva or diastase. The behavior of starch towards chromic acid was studied by Harz (Beiheft. z. bot. Centralbl; Woch. f. Brau., 1905, xxii, 721), who found that with chromic acid alone colors are obtained, varying from a yell o wish-grayish-green to an oUve-gi'een, and finally a golden-yellow, accorchng to the concentration of the solution. With a mixture of cliromic and sulphuric acids the colors were shades of green. The different kinds of starch, and also different grains of the same starch, varied greatly in their behavior, so much so that starch can not be regarded as a physically uniform sub- stance, the grains differing from one another according to the den,ser or looser constitu- tion of their ultimate complexes. Sodium salicylate has been reported by Lenz (Seventh Inter. Congress Appl. Chem., London, 1909; Jour. Soc. Chem. Ind., 1909, xxviii, 731) as a reagent to be employed in the niicrochemical differentiation of starches of different kinds. He states that if a trace of rye starch, in a hanging drop of a solution of 1 part of sodium salicylate in 11 parts of water, is examined under a magnification of 200, at the ordinary temperature, it will be found that after the lapse of an hour (more distinctly after 24 hours), most of the large granules have swollen ; only a small part resists the action of the salicylate and still shows the polarization cross between crossed nicols. In the case of wheat starch, only a few of the large granules become swollen; after 1 to 24 hours the outline of the unswollen wheat starch-gi'anules is sharply defined, and the granules, unlike those of rye starch, do not become flattened (starch of any kind which has been altered by storage in a moist condi- tion swells on treatment with the salicylate solution). Barley and millet starches swell to a small extent only. But few of the grains of oats, maize, rice, potato, bean, pea, lentil, and arrowroot starches become swollen. TEMPERATURES OF SWELLING AND GELATINIZAIION. When starch-grains are heated in water they begin to swell at temperatures usually between 45° and 55°, sometimes higher and sometimes lower, according to the source of the starch. As the temperature increases the grains lose their form and become gelatinous. GELATINIZATION, KEFRACTIVE INDEXES, AND REACTIONS IN POLARIZED LIGHT. 175 Table 20. Ivind of starch. Swelling begins. Gelatinization begins. Rye Corn Horse chestnut {/Esculus hippo- caslanum) Barley Chestnut (Castanea vcsca) Potato Rice Arum mamlatum Arrowroot {Maranla arund.).. . Tapioca {Jairopha utilis.) Arum csculenlum Sago (Sagus rumphii) Buclvwheat Acom Wheat 45.00 50.0 52.5 37.5 52.5 46.25 53.75 60.0 66.25 45.0 55.6' 57.5 50.0 50.0 55.0 Gelatinization complete. 55.0 62.5 forming ultimately a paste or a pseudo-solution of varying degrees of viscosity, according to the percentage and kind of starch and \-arious attendant conditions. Lippmann (Jour. f. jirakt. Choinie, 1861, Lxxxiii, 51) studied the temperatures at which swelling and at which gelatinization begin, and at which gelatinization is complete. The starch and water were put in a beaker in a water-bath and slowly heated, and the preparation was subjected to microscop- ical examination at proper in- tervals. His results are given in table 20. Lintner (Tollen's Handb. d. Kohlenh., ii, 207) records that potato starch suddenly goes into a paste at about 02° to 64°, while cereal starches un- dergo the same change at 75° to 80°. (See table 23, page 178.) According to Dafert (Meyer, Die Starkekorner, loc. cit. p. 134) rice starch gelatin- izes at 73°. ^\^lymper (Seventh Int. Cong. Appl. Cheni., London, 1909; Jour. Soc. Chem. Ind., 1909, xxviii, 806) recorded the temperatures of gelatinization of barlej^, corn, rye, potato, rice, wheat, and tapioca starches by subjecting them to a gradually rising temperature and examining them microscopically. The values differed in most cases from those of pre\ious observers, and also varied with the state of maturity of the gi'ains. The larger granules of any given starch were found to almost invariably succumb more quickly than the smaller granules to both wet and dry heat, and to diastase and mineral acids. Gelatinization may be brought about by various chemicals, such as potassium hy- droxide, chloral hydrate, chromic acid, etc. (see page 172). 56.25 58.75 57.5 62.5 58.75 62.5 58.75 62.5 58.75 61.25 58.75 62.5 66.25 70.0 62.5 68.75 63.75 68.75 66.25 70.0 68.75 71.25 77.5 87.5 65.0 67.5 REFRACTIVE INDEXES OF STARCHES. The refractive indexes of starches from different sources were determined by Ott (Osterr. bot. Zeitsclu'., 1899, xxxix, 313). The figures recorded are given in table 21. Table 21. Kind of starch. Refractive index. Kind of starch. Refractive index. Fritillaria n = 1.5040 1.5135 1.5200 1.5208 1.5212 1.5219 Barley n = 1.5220 1.5222 1.5245 1.5247 1.5293 Potato Corn Wheat Saco. . . Maranta Rye Tapioca Rice RE.\CT10XS IX POLARIZED LICHT. The behavior of starch-grains toward polarized light seems to have been discovered by Biot (Compt. rend., 1844, xviii, 795). Since then many observers have noted the differ- ences in the form and distinctness of the interference figure or "cross"; and also differences in the degree with which hght is transmitted by grains of the same and of different starches, 176 DIFFERENTIATION AND SPECIFICITY OF STARCHES. and ill the colors with a selenite plate. The point of intersection of the two parts of the cross usually corresponds to the position of the hilum, or in compound grains to each of the many hila. In grains having a centric hilum, the form of the cross corresponds with that of the Greek cross or cross of St. George, but in grains with eccentric hila, the figure corresponds with that of the Latin cross, the cross-arm being in any position between a short distance from the center and the almost extreme end of the part corresponding to the upright, and in such grains the arms become curved, sometimes so much so as to resemble the hanging branches of the weeping-willow tree, losing the appearance of a cross. Many illustrations will be seen in the accompanying plates. Muter {loc. cit.) found that the follow- ing starches give a well-marked cross : Potato, canna, maranta. Natal arrowroot, turmeric, mother-cloves, sago, tapioca, and cinnamon. The following do not give a well-marked cross: Wheat, barley, rye, acorn, and cacao. The following give a faint cross: Ginger, banana, nutmeg, sorghum, oat, and corn. The following give an indistinct cross : Pea, bean, and lentil. Rice gives a cross that is distinct and well-marked, and pepper shows a cross by high magnification. The polariscopical properties are destroyed by all agents which cause a swelling of the grains, but they are not affected after prolonged subjection of grains to dilute acid, which renders the grains soluble without destroying their skeletal structure. The employment of polarized light for the microscopic detection of the presence of foreign starches has been proposed by Gastine (Compt. rend., 1907, cxliv, 35; Jour. Soc. Chem. Ind., 1907, xxvi, 108). A small quantity of the starch is suspended in a drop of water, placed on a microscopic slide, and dried by exposure at room temperature and finally by heating for a few moments at 120° to 130°. The preparation is then mounted in Canada balsam and examined in ordinary polarized light, and also with the addition in the polariscope of a selenite plate. With ordinary polarized light, the grains of rice starch with a magnification of 300 appear brilliantly illuminated in the dark field and present a granitic structure. In the clu-omatic polarized light obtained by means of selen- ite, the blue and orange tints present a characteristic network of fines. Corn starch gives a siinilar network, but its meshes are much larger than those obtained with rice starch; corn starch also polarizes brilliantly. Millet, buckwheat, and rye starches, and many others, show appearances similar to those of rice and corn starches. The characteristic form of the grains of bean and pea starch and their very brilliant polarization render their determination easy. In the case of wheat starch, the groups formed of these grains are of very varied size and irregularly placed, giving a very characteristic appearance under the microscope; seen in ckromatic polarized light they do not present a symmetric network of fines. CHARACTERS OF STARCH-PASTE AND PSEUDO-SOLUTIONS FORMED BY STARCHES FROM DIFFERENT SOURCES. The characters of the starch-paste and also of the pseudo-solution vary with different kinds of starches. Tliis has been recognized in technical trades for many years, and on this account preference is given to certain starches in the stiffening, sizing, and finishing of fabrics. Potato starch yields a preparation by boifing that is poorly adapted to these pmposes, whereas wheat and corn starch yield an excellent product; but where a starch is normally not of high value in this respect it may be rendered so by various means, as, for instance, by the process suggested by Bellmas (Osterr. Chem. Zeitung., 1902, 366), which is to digest potato starch at 55° in a 2 per cent hydrochloric acid, then forirung a fimpid preparation by boiling. Various of the methods employed for making "soluble starch" and liquefying starch-paste may be used (see page 101). In the experiments of Saare and Martens (Zeit. Spiritusind., 1903, xxvi, 436) it was found that the length of time the starch was heated at the boifing-point had a considerable influence upon the stiffening power of PHENOMENA OF DIGESTIBILITY — RAW STARCHES. 177 lliL' prei)aration. Wlieat and corn starches did not attain their maximum stiffness until alter boilinjj; for f^onie miiuites; on the other hand, potato starch had its maximum stiff- ness when tlie boiling-point was reached, and lost in power very considerably by further heating. The non-homogeneity of slarcli-paste was particularly referred to by Pottevin (Ann. d. Inst. Pasteur, 1899, xiii, 728) (see page 134). Ling (Jour. Fed. Inst. Brew., 1903, ix, 440) found considerable difference in the conversions of starch-pastes and also of raw starch at temperatures below the point of gelatinization. The relative stiffening strength of starches prepared under the same conditions is, according to the Scientific American Cyclopedia of Receipts, taldng 100 as the standard of comparison: Pure dry rice starch, 100; rice starch No. 1, 95; rice starch No. 2, 91; pure chy corn starch, 87; corn starch, 85; rye starch, 81; oat starch, 80; acorn starch, 80; wheat starch, 80; barley starch, 78; Bermuda arrowi-oot, 75; Natal arrowroot, 73; pure potato starch, GS; and potato farina, 65. PHENOMENA OF DIGESTIBILITY. RAW STARCHES. It has long been established that starch-grains are in the nature of a reserve food of the plant and that they undergo ready solution in situ (presumably by the action of enzymes) when their derivatives are required for nutriti\'e purposes. While such dissolution takes place with ease in the plant, in seeds, in bulbs, etc., such is not the case in vitro, even though presumably the same enzymes are present, unless the outer protective coat- ings of the grains are injiu'ed so as to expose the starch-substance, especially the innermost part. It is true that there are many records to indicate that various kinds of raw^ starch are digestible in vitro with various degrees of ease or difficulty, but it is quite clear that when bacterial action has been prevented, and when the grains are iminjured, absolutely no digestion occurs. If, however, the grains have been eroded by bacterial or other action, or if the grains be fissured or broken so as to permit contact of the enzymes with the intra- capsular part of the grain, digestion proceeds with variable degrees of rapidity according to attendant conditions. So long as the grains are perfect, the coating, which varies in resistiveness in starches of different kinds, and also in grains of the same starch, serves as a perfect protective against the influences of diastatic enzymes in vitro, if bacterial action is prevented. The writer has subjected various kinds of starches to the actions of ptyalin, pancreatin, Taka-diastase, and malt diastase in a 1 per cent chloroform solution for over 12 months at optimal temperatures, without evidence of erosion or digestion of the grains, except in the case of injured grains; yet similar solutions of the same enzymes caused a practically complete saccharification of the boiled starches in 6 hours. It seems that the virtually absolute indigestibility of perfect starch-grains in vitro in the presence of enzymes is owing to the absence of some factor in the plant that in some way gets rid of the barrier presented by the coating. The non-digestibility of raw starch by enzymes was recorded as far back as 1835. Guerin-Varry (Ami. d. chim. e. phys., 1835, lx, 32) set aside a preparation of potato starch and malt extract in a sealed tube, at room temperature, for 63 days. At the end of this period there was not a trace of sugar, nor did the grains show any microscopic changes. Dubrunfaut (Ann. d. cliim. e. phys., 1847, xxi, 178) noted that while raw potato starch is unaffected by malt diastase, wheat, barley, and rice starches are affected. C. Nageli (Die Starkekorner, loc. cit.), in liis digestion experiments with saliva, by which he differ- entiated the so-called amylose and cellulose, found that the amylose was digested, lea,ving the skeleton cellulose. In fact, as the author has found, saUva becomes absolutely inert before any such action can take place. Hence it was not the sahva, but some other agent, that caused the effects noted. The inffuences of sahva on raw starch were also studied 12 178 DIFFERENTIATION AND SPECIFICITY OF STARCHES. Table 22. Kind of Btarch. Time of appearance of sugar. Potato 2 to 4 hours. IM hours. 30 to 60 minutes. 10 to 15 minutes. 5 to 7 minutes. 3 to 6 minutes. 2 to 3 minutes. Pea Wheat Oat Rye Corn by Hammarsten (Jahr. ii. d. Fort. d. Thierchemie, 1871, i, 187), who writes that the variation in the time required for changing starch-granules into sugar is due to differences in the starches, and also according as to whether the grains were whole or broken or gelatinized. Thus, in experiments made with human saliva, sugar was formed from the raw starches as shown in table 22. Hammarsten states that since the use of starch- paste made from these several starches showed no dif- ference in the sugar-forming time, it may be presumed that the unequal deposition of cellulose in different starches presupposes unequal resistance to the influence of the saliva and that it was therefore natural to ex- pect that starch-granules, wliich in a raw state were converted into sugar with difhculty, would react readily if after being pulverized they were treated with saliva. This supposition was confirmed, inasmuch as finely powdered potato starch was found to be rich in sugar after 5 minutes. Chewing starch-granules was then tried, with the result that sugar was formed in all of the above-mentioned starches in 1 to 4 minutes. The refractivity of raw potato starch to digestion was also noted by O'Sullivan (Jour. Chem. Soc. Trans., 1878, ii, 125), who records that in the presence of fresh malt extract the granules were unaffected after 24 hours. Baranetzky (Die starkeumbildenen Fermente in den Pflanzen, Leipzig, 1878, 40) places the order of digestibility of raw starches as follows: Buckwheat, wheat, bean, acorn, chestnut, potato, and rice. This does not accord with Hammarsten's records. Brown and Heron (Ann. d. Chem. u. Phys., 1879, clix, 20G) ascertained that starch-grains in the presence of malt extract do not undergo any change even after a considerable time although when in situ in germinating seeds they undergo more or less rapid solution. When, however, the grains were crushed by grinding in rough sand or broken glass, rapid solution occurred. The marked solubility of the broken grains was noted by Hammarsten and also by certain of the earlier investigators of starch. Recently Maquenne (Compt. rend., 1904, cxxxviii, 375) found that fully comminuted grains yield as much soluble matter by digestion at 55° with malt extract as starch-paste, and therefore that rupturing the grains is as effective as gelatinization by heat in rendering the raw starch digestible. Lintner (Brau. u. Malzerkalender, 1890, xiii, 83) carried out a series of experiments with various starches with diastase at different temperatures. In each experiment 2 grams of air-dried starch were subjected to the action of 50 c.c. of malt extract for 4 hours. Each preparation was then made to 100 c.c. by the addition of water, and then filtered to remove the undigested starch. The filtrate was saccharified by weak hydrochloric acid, and the sugar determined by Fehling's solution. From the sugar determinations the figures given in table 23 were estimated as showing the percentages of starch digested. Table 23. Kind of starch. Temperature of digestion. Temperature of gelatiniia- tion. 60° 55° 60° 65° Potato p.ct. 0.13 6.68 12.13 29.70 13.07 2 25 p.ct. 5.03 9.68 53.30 62.23 58.65 56.02 p.ct. 62.68 19.68 92.81 91.08 92.13 91.70 18 40 p.ct. 90.34 31.14 96.24 94.58 96.26 93.62 54 94 o 65 80 80 75 to 80 75 Rice Barley Wheat Malt (fresh) Malt (cured) Corn Rye PHENOMENA OF DIGESTIBILITY — RAW STARCHES. 179 The effects of enzymic action on raw starch were studied by Mej'er (Die Starkekorner, loc. cit.), \\\\o records that diastase as well as saliva can penetrate the porous starch-grains, as shown by the following obser\'ations : Fresh starch-grains from the endosperm of Ilordeinn distichum when i)laced in malt extract form many cracks which radiate from the center of the grains. The presence of these cracks is explained upon the assumption that the penetrating ferment causes a swelling of the grain. Starch-grains from Dieffenbachia, treated for 3 weeks with malt extract, when compared with untreated grains, show a degree of transparency which indicates that some of the substance has been lost, or that the grains have undergone some internal change. Diastase, he states, has the capacity to dissolve every laj'er of every starch-grain at ordinary room temperature if the action continues for a sufficient length of time. At 40° the ferment action is more powerful on intact grains than at 17°. At 40°, after 24 hours all the grains acted upon were dissolved. A pressure of 3 atmospheres on a preparation consisting of wheat starch and a solution of diastase did not accelerate the solution. In another place Meyer notes that perfect potato-starch grains are dissolved away centripetally in successive layers, gradually forming long, spindle-shaped grains ; but grains ha\-ing clefts or fissures were acted upon so as to give rise to channels, pits, cavities, etc. Similar effects were recorded with the starch of Dieffenbachia seguine. The non-uniform composition of different starch-grains was pointed out by Pottevin (Woch. f. Brau., 1899, xvi, 641) to be shown by the behavior of starch-grains acted upon by diastase at ordinary temperature when the process is carefully observed under the microscope. WTieat starch, he found, is attacked by diastase at ordinary temperature. Pottevin observed that the thickest grains were less rapidly affected than the thinner grains, and that the smallest polyhedral grains were not at all attacked. The outer layers were not digested, remaining as residues. He also obtained a number of residues by digest- ing starch-grains in the cold for an hour, obtaining the residues of undigested matter, at specified periods, and subjecting these residues in turn to digestion. He thus obtained a number of residues, each successive residue being more resistant than the preceding to diastase. All of these residues were gelatinized with hot water, and in comparison with whole starch yielded only little more than one-half the proportion of maltose. He believes that the starch-grain is heterogeneous, and that therefore starch-paste is heterogeneous, the part of the paste formed from the less dense parts of the grains being easily digested and the denser parts digested with difficulty and after prolonged action, the dextrin from the latter constituting the "residual dextrins." Effront (Enzymes and their Applications, trans, by Prescott, 1902, 128) states that a granule of starch is irregularly attacked by diastase, corrosion occurring in very different directions and places. The manner of corrosion, he found, arises from the inequality of resistance of the surface of the grains, so that the difference existing in the compactness of the A'arious parts of the grain is, on the whole, the initial cause of the variations in the resistance to diastatic action. Potato and barley starch he writes are both composed of non-homogeneous grains whose layers differ in degree of compactness ; but in potato starch more resisting layers are found than in barley starch. Gelatinization, he records, does not change the properties of starch, which are owing to the variations in the compactness of the layers; therefore the more coherent particles form a paste that is more difficult to liquefy, and yield a dextrin of greater resistance than the less coherent particles. Ling (Brit. Assoc. Report, 1903; Jour. Soc. Chem. Ind., 1903, xxii, 1058) found in his investigations of the action of diastase on starch-grains of raw and malted barley that the starches of cereals differ from the starch of potato in being readily attacked by diastase in an ungelatinized condition. He carried out a series of experiments with mashes of barley and malt starches of various origin, the starches being mixed with the diastase preparation in the dry state and mashed with water at various temperatures for 2 hours. The figures in table 24 illustrate the results. 180 DIFFERENTIATION AND SPECIFICITY OF STARCHES. Table 24. Kind of starch. Mashing temperature. (a) u =3.93 fl3.93 Kind of starch. Mashing temperature. (a) ,,=3.93 K3.93 Barley, 1899 sample. . Do ° C. 60 65.5 71 60 65.5 71 60 65.5 ° E. 140.8 143.4 145.1 150.7 152.3 163.8 150.9 156.1 ;;. ct. 88.7 88.2 80.6 84.2 79.0 77.8 84.4 80.6 j Kilned malt sample. . Low-dried rualt Do Do Barley starch (iiaste) Potato starch (]jaste) Do Do " C. 71 60 65.5 71 65.5 60 65.5 71 ° E. 161.3 151.6 152.5 165.7 168.6 154.5 155.6 167.1 V-ct. 67.2 85.3 S3.3 66.6 52.4 81.8 71.5 55.3 Do Barlev, 1002 sample. . Do Do Kilned malt sample. . Do Fcrnbach (Compt. rend., 1904, cxxxviii, 428) ascertained that the small, liglit grains as compared with the large, heavy grains of potato starch contain a higher percentage of phosphates, and he concluded that the grains consist of a nuclear portion relatively rich in phosphorus, upon which are superposed layers free from phosphorus. In digestion experiments he noted differences in the behavior of the grains and parts of grains in accord- ance with this peculiarity. Wliole starch-grains were found by Maquenne (Compt. rend., 1904, cxxxviii, 375) to be but little affected by malt diastase. Only 2.8 per cent of the raw starch was dissolved after treatment with malt extract at 55°. The same amount of starch after boiling and subjected to the same quantity of malt, the same temperature and period of time, was completely saccharified. Under the latter conchtions of experiment, finely comminuted raiv grains were saccharified to the extent of 94.8 per cent, or almost equivalent to gelatinized starch. Day (U. S. Dept. Agriculture, Office Expt. Stas. Bull. 202, 1908, 24), in experiments with wheat, corn, barley, rice, potato, and arrowToot starches, found that raw starch digests slowly; that the cracked grains digest more rapidly than the solid grains; that the outer layer is less digestible than the inner; that the six kinds of starch fall into two divisions according to digestibility, the starches of wheat, corn, barley, and rice being more digestible than those of potato and arrowroot. (See pages 166, 189, 190, and 192 for further reference to the investigations of Day.) Hanausek (The Microscopy of Technical Pi'oducts, trans, by Winton, 1907, 37) describes erosion changes in potato-starch grains such as may be obser\^ed in a sprouting potato. The specimen examined he thinks had been acted upon by diastase. Some of the grains were marked by clefts and fissures which passed through or radiated from the hilum; others had an irregular sac-like nucleus; and others were corroded or finely pitted, displaying cavities, channels, etc. In a specimen of wheat starch he found grains which were partly eaten away, mostly on one side, but the appearance was different from that of the sjirouted grain caused by diastase. He attributes the erosion to the action of acid used in the preparation of the starch. Eroded grains in situ he states display delicate concentric rings and more or less distinct central fissures. Often the rings in parts of the grain (in the same radial direction) are very distinct, while in the intermediate parts they are scarcely evident. After long-continued action of the ferments the grains become hollowed in the center and show numerous canals resembling the bm'rows of insects. Han- ausek also observed the grains from sprouted rye, which he states present a very different appearance. The outer layers, he records, are radially striated, and separated from the inner by a marked circular crack, while the central portion is irregularly fissured, the original clefts about the lulum being enlarged. It seems to be generally believed that starch-grains of different sources undergo erosion in specific ways peculiar to the plant source. If the character of erosion is peculiar to each kind of starch, it indicates specific forms of non-homogeneity of the granules which are attributable to peculiarities of the metabolic processes which give rise to the starch-grain ; hence, is a matter of considerable importance in this research. PHENOMENA OF DIGESTIBILITY — RAW STARCHES. 181 Several observers have studied the phenomena of erosion of the starch-grains m sihi, and a number have made comparisons of the changes which occur in situ and in vitro, and there are also records of comparisons of the differences in the erosion phenomena exhibited by the grains of different plants. Baranetzky (Die stiirkeumbildenen Fermente in den Pflanzen, Leipzig, 1878) studied the starches of Polygonum fagopyrum, Phaseolus multiflorus, MirahiUs julapa, Quercus pedunculata, /Esculus hippocastanum, Solanum tuberosum, Fagopyrum esculenlum, Triticum sativum var. vulgare, and Oryza saliva. He records that these starclies differed in their susceptibility to tlie action of ferments. Tiie grains of buckwheat were the most easily affected; wheat and bean coming next; those of potato, and especially rice, being the least easily affected. None was found to entirely resist the influence of ferments. A given kind of starch was noted to be affected not only by the ferments of the same plants but also by any kind of a starch-reducing ferment, all kinds affecting it similarly. The experiments by Baranetzky include the actions of ferments of barley-malt, of germinated seeds of Polygonum fagopyrum, MirahiUs jalapa, Phaseolus multiflorus (coty- ledons), Esculus hippocastanum (cotyledons); of sprouting tubers of Solanum tuberosum and Beta vulgaris; of stems and leaves of Phaseolus multiflorus; and of leaves of Eriobo- trya japonica, Acanthus cordifolia, and Tradescantia discolor. The solution of the starch- grains by the action of ferments was found to run parallel with the formation of starch- paste. Since various starches differ in their resistance to the action of ferments, weak ferments are limited in their action to grains that are easily affected, while energetic ferments are required to influence such ferment-resisting grains as those of Solanum and Oryza. The phenomena of the solution of the starch-grains are stated to refer only to the actions of ferments, and not to any other influence. Bacteria, he notes, will naturally form in a solution that has been allowed to stand, yet in the easily affected grains signs of erosion appear in about 24 hours, at the time when, as he states, the fluid is still free from organisms. Nor can the solution of the grains be due to the action of acids, because grains of potato starch subjected to the macerating action of weak hydrochloric or acetic acid for 33 days were entirely unchanged. Although the grains of horse-chestnut were rendered transparent in a month after treatment with very weak acetic acid, this change only slightly resembled that produced by the action of ferments. Baranetzky writes that up to that time the changes which the starch-grains undergo by the action of ferments ha\'e been observed only in the living plant, and that experimenters had not been able to produce them artificially. Phaseolus Multiflorus. Baranetzky records that the dissolution in vitro of the starch-grains of Phaseolus multiflorus always proceeds from the inside toward the periphery. After the ferment has been allowed to act for 24 hours, it will be observed that some grains have a sort of furrow, having the same shape as the grain, and which at first appears to be fifled with a granular mass. This furrow widens and becomes entirely transparent, and its walls (lamellse) break at several places, forming an outlet. It now becomes evident that this apparent furrow is composed of a uniform, very transparent substance which remains in the place of the dissolved starch-grains. The unchanged, dense mass of the grain gi'adually grows less until finally only tiny flakes and streaks are left, which are irregularly distributed on the periphery of the remaining skeleton. This skeleton is a clear, transparent disk, with a delicate but distinct contour, and completely retains the form and size of the unclianged grain, and the lamelhc are frecjuently even more marked than in the intact grain. In treating starch-grains in this state with iodine, Baranetzsky found that all the denser parts give the usual iodine reaction, while the transparent parts remain colorless, even after the prolonged action of the reagent; that the skeletons therefore no longer con- 182 DIFFERENTIATION AND SPECIFICITY OF STARCHES. tain "granulose," and that it is easy to prove that they consist of pure "cellulose," for if Lugol's solution be used instead of iodine the skeletons in a short time become brownish, then a copper-red, and finally gradually go over into violet. Thus, he states, the actions of the plant ferment, Uke those of saliva, at first cause the granulose of the starch-grains to dissolve. The cellulose of the starch-grain of the bean was not so insoluble, because after prolonged action of the ferment the grains become more transparent and delicate, so that the grains finally can be distinguished only when stained; their contours, which at first are distinct, become obscure, and the skeletons finally disappear altogether. Powerful fer- ments brought this about in 4 to 5 days. In the gernfinating seed of Phaseolus mulliflorus the same phenomena were observed, except that frequently some of these grains were found which did not dissolve from the inside but were dissolved evenly, so that the transparency of their skeletons gradually became visible throughout the entire grain. Similar appearances were also observed in artificially dissolved grains where the action of the ferment was slow and weak, although an actual formation of a skeleton never took place. This deviation from the ordinary process of solution, Baranetzky believes, is probably due to a non-penetration of the ferment evenly into all parts of the grain. The fresh starch-grains of the scarlet bean almost always show, he states, a complicated system of furrows and fissures, of which several extend to the periphery. In ordinary cases the ferment probably forces its way through these fissures into the inside of the grains and begins its action before the out- side layers have been in the least affected by it. Accordingly, it is found that the inner furrow usually follows the direction of tlie fissure, and widens toward the inside to reach the periphery of the grain at these places. While the dissolution of the granulose takes place almost uniformly from within outward, the grains of some of the seeds of Phaseolus deviated from this rule. In these cases the granulose disappeared very unevenly, and the remaining skeletons, although consisting of pure cellulose, were nevertheless usually substantial, and their contours almost as well-defined as those of the intact grain. These skeletons, however, were finally dissolved by the prolonged action of the ferment. Baranetzky states that this would seem to prove that the starch-grains of the same plant may differ in their cellulose content, since it does not seem likely that these deviations can be attributed to the differences in the solubility of the granulose and the cellulose of the different starch-grains. QuERCus Pedunculata. This starch behaves similarly to that of Phaseolus. The process always begins in the inside of the grain, Vv'ith the difference that the outlines of the gradually increasing clear spot are not distinct, and gradually pass into the grain substance. As the process of solution reaches the periphery, only the granulose disajijiears, the cellulose remaining behind in the form of a sharply outlined skeleton, as in Phaseolus. The cellulose reaction is not so easily produced as in Phaseolus. Lugol's solution causes a yellowish color at first, v.'hich finally goes over into copper-red and violet. The cellulose skeletons are also completely dissolved. SoLANUM Tuberosum. In the grains of potato starch the process of solution begins at the periphery and extends to the inside of the grains, where small round or irregular clearly outlined spots make their appearance. Seen in profile, these spots appear like canals entering into the inner part of the grain and arc filled with a soft transparent substance. Fine but distinct lamella) may often be seen in the latter. Sometimes those places on the outside of the grain at which erosion takes place simultaneously are numerous and irregular, and grad- ually merge into each other, so that the surface of the grain is covered with complicated markings. The eroded places grow deeper towartl the inside of the grain, and then widen PHENOMENA OF DIGESTIBILITY — RAW STARCHES. 183 out, SO that the process of solution may now take place in the opposite direction — that is, from the inside to the periphery. Otherwise, the behavior of potato starch is like that of Phaseolus grains. The remaining cellulose skeleton shows sharp outlines at first, and more or less distinct lamella^ especially in the larger grains ; the cellulose skeletons appear to be more substantial, and the outer layers are decidedly more dense than those in the inside of the grain. The cellulose reaction with Lugol's solution is not so easily obtained with the skeletons of potato starch. At first they are stained only a pale yellow, and it takes 30 to 60 minutes to produce the violet reaction. The outside dense layers are less easily stained than the other parts, and usually continue to show a brownish color when the violet has already begun to appear in the inner parts. The solution of the cellulose skele- tons appears to take place more rapidly here. Some grains were noticed in which one part was still unchanged, while the skeleton of the other part had almost entirely dissolved. In the few sprouting tubers of potato examined it was found that solution seems to begin in the center of the grain, and the process to go on evenly from this point, so that the grain is changed into a thin-walled shell, with the inner layer of the wall indistinct. These deviations are stated to be unimportant in the process of solution. iEsCULUS HiPPOCASTANUM. The starch-grains of the horse-chestnut are changed outside the cell in the following manner: At that part of the starch-grain farthest removed from the hilum, where the otherwise not very distinct lamellae appear particularly distinct and numerous, a trans- parent triangular space is developed which penetrates deeper into the inner part of the grain, and reaches and surrounds the hilum; in the meanwhile, some single lamellae or groups of lamellae dissolve more quickly than others, causing the latter to appear as dark cross-stripes, and also giving the grain a peculiar striated appearance. The more dense and unchanged portions belong to the periphery of the grain, so that after the process of solution in the inner part has sufficiently progressed, the grain floats in the solution and appears like a bubble whose walls are riddled with numerous irregular holes. Finally, only isolated portions of the latter are left, which, however, retain their former position, so that the outlines of the former starch-grain can still be recognized. This goes to show, wTites Baranetzsky, that in this case also, after the granulose has dissolved, a residue is left which undoubtedly is cellulose, but which is so delicate that it is hardly visible. The interstices sometimes take a violet color with Lugol's solution and sometimes remain colorless. In the cotyledons of germinating seeds of Msculus hippocastanum, starch-grains are very often found which, while unchanged on the inside, have a sort of pitted appearance on their outer surface. Seen from the flat side the grains usually show a small cavity at the end opposite the hilum, and from which the starch-substance has entirely disappeared. Such cavities also sometimes develop on the flat side of the grain, and when seen from above look clear and round. Seen from the side of the grain, the contours of these cavi- ties are always distinct, proving that the solvent action of the ferment is merely a super- ficial one and does not penetrate into the substance of the grain. Very often conditions are found in which the greater part of the grain has disappeared while the remaining clearly outlined part is unchanged. How this superficial process of solution continues was not determined. During germination of such seeds as those of yEsculus hippocastanum, only a very small part of the grain is used up. It would therefore seem that grains which so energetically resist the action of the ferment undergo no change. Usually, however, the germinating seeds of the horse-chestnut dissolve from the inside, resembhng the process described in the grains of potato starch. The starch-grains of one and the same seeds are not always affected in the same way, so that it would be necessary to examine a number of seeds in order to observe typical forms of the two extremes. 184 differentiation and specificity of starches. Polygonum Fagopyrxjm. The small grains of this plant were not easily observed. Under magnification of 300, in the first stages of erosion the grains have a striped appearance, with fine, dark, dotted lines radiating from the center to the periphery. Magnified 500 times, it can clearly be seen that the change begins on the outside. Numerous narrow canals are found on the peri- phery which penetrate into the center of the grain ; the canals appear lighter than the other parts, due to the lamellas of the gi-ain, which do not show under a lower power. The sub- stance in the center of the grains dissolves more easily, and the canals therefore run together here into a small hollow, and the process of solution now continues centrifugally. Finally, only a ring of isolated small grains is left which is evidently held together by the remaining invisible mass of cellulose. Fagopyrum Esculentum. The same process takes place in the germinating seeds of buckwheat, but among all the seeds examined those of buckwheat were most easily affected by the ferment. Only small portions of the grains were left after 24 hours' action of the ferment treated with strong formic acid, and in 48 hours no trace of the starch-grain was left. The minimum of the time reaction is probably only a few hours. Triticum Sativum var. Vulgare. Baranetzky found that the starch-grains of wheat are almost as easily affected as those of buckwheat. As is well known, there are two kinds of starch-grains in the seeds of wheat, the large, flat, crescent-shaped and the small, more or less spherical. The erosion process described is that of the large grains. The first changes are Uke those of Polygonum, the same striped appearance, and the same radiating lines, but the latter not always ex- tending to the center. Similar canals also run from the flat sides into the inner part of the grain; on the surface of the grain these canals appear as small, clearly outlined circles, greatly resembling the round tufts of the parenchymatous cells. In the meantime tlie clear parts of the grain grow wider, so that the unchanged portions ha\'e the appearance of dense radial bands in the clear, concentric-layered grain. At the same time the clear portions begin to melt away from the outside, but the bands do not keep pace with tliem, the ends of the latter usually protruding more or less freely. After these bands are finally absorbed only a small disk remains, which is formed from the central part of the grain where the lateral canals did not penetrate. This disk now dissolves at its upper surface and at the walls of the canals which have penetrated from the sides of the grain into its center. At last only an u-regular lump, riddled with holes, is left of the entire starch- grain, and this also finally dissolves. Very often the process of solution begins with the formation of concentric spaces which evidently correspond to the dissolved single groups of lamella;. In such cases the grain usually dissolves from the outside, without previous formation of canals. Usually the formation of the centric spaces combines in different ways with the radial canals, producing varied and complicated appearances in the process of erosion. In rare cases, starch-grains may be found which dissolve equallj^ in all parts. They then become very transparent and show 2 to 3 concentric zones of varying density, gradually merging into one another. The final disappearance of such grains takes place either by a melting away from the outside or by a gradual complete solution. No cellulose skeleton was found in any of the starch-grains of Triticum after the granu- lose had been extracted. Watery solutions of iodine always have a violet reaction, even in those parts of the grains which were on the point of disappearing. This, Baranetzsky states, need not necessarily mean that the grains of Triticum contain no cellulose, for the same results should be obtained if the cellulose were more soluble, or at least equally as soluble, as the granulose of the grains in question. PHENOMENA OF DIGESTIBILITY RAW STARCHES. 185 The starch-grains of germinating seeds of wheat were found to be variously affected, so that all of the phenomena described above could be observed in them. It is worthy of remark that in one and the same seed the process of solution takes place in a definite way, the usual course being as follows: A concentric fissure develops at a distance from the rim of the disk ; in the middle of the space surrounded by this fissure a system of irreg- ular, branched fissures arise, giving this portion the appearance of falling apart, while the ring of the surrounding border is vuibroken. The latter is absorbetl in its entirety, and radial stripes may frequently be noticed in it. The deviations from the usual course exhibited by some grains explains the contradictions by Sachs (Botan. Zeitung., 1862, page 147) and by Gris (Annales de science naturelles (Botanique), 1860, xiii, 110) in their descriptions of the same process. Sachs's statement that the affected grains first lose their granulose, leaving a skeleton which becomes wine-colored when treated with iodine, could not be verified. In the germinating seeds, as well as in the tests outside the cell, the iodine reaction was always the same, i.e., a blue-violet color was obtained until the very disap- pearance of the grains. MiRABILIS JaLAPA. As is well known, these grains consist of numerous part-grains (p. 273) which separate easily, and in their isolated state look like very minute spheres showing a Brownian move- ment. It is hardly possible to follow the process of erosion in such gi-ains, yet on the whole it is safe to assume that they also dissolve through the action of the ferment. Their contours become vague and indefinite, and the single part-grains gradually change into indistinguishable dark points, which finally become invisible. In the grains or their frag- ments that have not fallen into the single minute grains, the part-grains lying on the l^eripherj' of the groups are affected first, while the others remain unchanged. If such groups are not \eiy large, then all their part-grains finally are changed, and the whole fragment takes on the appearance of a fine, granular, semi-transparent mass. Oryza Sativa. In size and form, Baranetzky writes, these grains closely resemble those of buckwheat, but show an extreme difference in their solubility, being the least easily affected by fer- ments. The erosion process here appears to begin in the middle of the grain, and solution takes place peripherally, the last stages in the process thus resembling those of buckwheat grains. Baranetzky notes that close observation of the process of erosion of the grains of different starches at once shows variations in their structure. In most of the starches examined (bean, buckwheat, potato, and rice) erosion takes place more readily in the inner layers than in the outer ones, agreeing perfectly with Niigeli's theory on the structure of the starch-grain. When the soft hilum is eccentric the inner cavity usually si)reads toward the hilum side. In wheat starch the structure of the grain seems to be symmetrical, as is shown by the concentric layers which make their appearance as the grains dissolve, the soft hilum being in the exact or mathematical center. Nevertheless, the central part of round stai ch-grains does not dissolve as easily as the outer part. The radial canals in which erosion begins usually stop at a distance from the center of the grain, final solu- tion usually taking place by the gradual melting away from the outside. The greater solubility of the outer part of the grain might be due to a difference in chemical composi- tion, but the presence of cellulose could not be proved in this case, and there would be no other reason for supposing any chemical modification in the granulose. Baranetzky records that the formation of the canals by which the ferments reach the inner part of the grain is worthy of notice. The possibility alone of this formation proves, he states, that not only different concentric layers, but also different parts of the 186 DIFFERENTIATION AND SPECIFICITY OF STARCHES. substance of the starch-grains, may be unequally affected by ferments from the outside toward the inside, and that therefore they must differ in their density. Special emphasis, he believes, should be laid on the fact that the contours of these canals are sharply out- lined, and that they usually remain narrow, while in the meantime the concentric layers successively disappear. This may be clearly seen, especially in the seeds of wheat, where different parts of the grain remain unchanged while others are already transformed into a cellulose skeleton. In the ferment-resisting grains of potato and rice starch, after the action for 24 hours of a strong ferment, single eroded grains can nearly always be found, the number increasing after a few days; but the majority of the grains remain unaffected in spite of repeated renewal of the ferment. This is most strikingly exhibited in the scarlet bean and horse-chestnut, where a few single grains were found to be entirely intact at a time when all the others were on the point of disappearing. It is an interesting fact, writes Baranetzky, that the structure of the grain varies in single seeds of the same species of plants, as is undoubtedly proved by the differences of the grains in their processes of erosion. It would seem, however, that during germi- nation the majority of starch-gi'ains in one and the same seed are affected alike, but differently in different seeds. It seems of little moment, he found, whether the entire sub- stance of the starch-grain is permeated by the ferment, or whether the ferment action is merely a superficial one. It is more than likely that the process of solution depends on the concentration of the ferment solution. A highly concentrated ferment, he belie\^es, will act so energetically that the outer layers of the grain are affected before the inner parts have been permeated by ferment; the character of the colloidal ferment must neces- sarily retard the latter procedure. On the other hand, if the mother-liquor surrounding the starch-grain contains, as is usual, only a small amount of ferment in the living cell, then the ferment gradually permeates the entire mass of the starch-grain and the central soft portion is the first to come under the influence of the dissolving agent. The unequal power of resisting ferments exhibited by different starch-grains can not, he believes, be attributed to their cellulose content, for Phaseolus grains are an example of grains which dissolve easily and leave a skeleton of cellulose behind after the granulose has been dis- solved out of them. On the other hand, the more resisting grains of /Esculus hippocastanum appear to contain much less celhdose. Baranetzky concludes that the differences in the behavior of different grains toward ferments must be due to specific differences in the struc- ture of the starch-grains of different plants. Important in tliis connection is an investigation of Wortmann (Zeit. f. physiolog. Chemie, 1882, vi, 287) on the actions of bacteria on raw starch. His conclusions are as follows : Bacteria possess the power of producing the same changes as diastase in starch- grains, starch-paste, and soluble starch; they dissolve different starches with varying rapi(hty; they will attack starches only in the absence of other carbohydrate nutrients, and only when not deprived of air; the action of bacteria on starches is due to a ferment secreted by them for this purpose, which, like diastase, can be precipitated with alcohol and is soluble in water; the action of this ferment is purely diastatic; the ferment in itself is capable of exerting its influence in the absence of acids ; bacteria also secrete the ferments which are active in neutral starch-solutions; they act more energetically in weak acid solutions; the manifestations of erosion agree exactly with those described by Baranetzsky, as being caused by diastase. Wortmann's basic experiments were made with wheat starch ill a preparation consisting of 1 or 2 drops of liciuid from putrefying beans or potatoes (in which the Bacterium termo is abundant) in a 1 per cent solution of equal parts of sodium chloride, magnesium sulphate, potassium nitrate, and acid ammonium phosphate. Erosion begins at room temijerature in 5 to 7 days. Another, and perhaps the most important, investigation on the phenomena of erosion of the starch-grain was made by Krabbe (Jalir. f. wissensch. Botanik, 1890, xxi, 520), in PHENOMENA OF DIGESTIBILITY — RAW STARCHES. 187 which were studied (ho alterations of starches of various kinds in situ and m vitro. The following data are from Krabbe's article: Triticum Sativum var. Vulgare. During the process of erosion of the starch-grains of wheat, wedge-shaped and coni- cal indentations are formed with their bases turned toward the circumference of the grain. They are distinguished from the rest of the substance by their weaker refractive power, and when seen in profile appear to have more or less dense lamellae, which can not be seen in the intact grain. These indentations are in reality pore-canals of unequal length pene- trating to the center of the grain. The formation and successive lengthening of these canals is not caused by the dissolving action of the diastase, as is proved by the distinct outlines of the pores in all stages of development. The starch-substance remains unchanged during the formation of the canals; swelling agents and iodine affected the eroded grain the same as the intact one. Ferments therefore can not penetrate into the intercellular spaces of the starch-grain, and there can be no question of their dissolving action. As the pore-canals lengthen they usually become branched, thus setting up intercommunication at various places in the cell. Observation proves that these canals originate in the outer surface of the grain, the canals within the grain being secondary branches of older canals which have an opening on the outside of the grain. Superficial observation of grains about to dissolve often revealed some grains which looked as though large portions of the cir- cumference had been absorbed, but closer examination showed that this appearance is due to the same factors which cause the formation of the pore-canals, and not to the action of the ferment. In cross-section the wedge-shaped and conical canals on the circumference of the grains are at first circular, but as they widen they become elliptical. This is due to the lateral merging of two hitherto parallel pores, a phenomenon which can be observed in all its stages in the starch-grains of Triticum. In consequence of this, fissures from the periphery of the grain, encircling about one-third of the circumference, give rise to the mistaken supposition that erosion had taken place here. There were evidently such starch- grains, in which similar coalescence had taken place, which led Baranetzky to the erroneous conclusion that the diastatic ferment penetrates the starch-substance and acts as a dis- solving agent. Other phenomena may be observed in Triticum, which also might lead to incorrect inferences. In some grains passages running parallel to the concentric layers arise which look as though they had their origin in the inner part of the grain without regard to the outside of the grain, but it can be seen that these passages communicate with openings on the outer surface. The origin of these spaces can be traced to ferment passages which come from outside. Since these latter are capable of entering the grain in all directions, it will not be surprising to find now and then some of these passages running parallel to the concentric lamellae. Thus the different manifestations in processes of erosion in Triticum depend primarily on the width, the branching, and the manner of inter- communication of the pore-canals, and in a measure on the unequal distribution of the pore-canals in the eroded grain. As a rule, the pores on the margin develop earlier and in greater numbers on the flat side of the grain, but any number of variations from this rule may occur; and it is easy to conceive that the unequal arrangement and distribution of the pores in different grains can produce widely deviating characteristics in the process of erosion. HoRDEUM Vulgare and Secale Cereale. The process of erosion is similar to that in Triticum. No appearance of diastase penetrating the substance could be determined; and the very distinct outlines of the pore-canals and the rarer occurrence of lateral coalescence of pores than in Triticum confirm the theory that diastase does not act in the manner of dilute acids. 188 differentiation and specificity of starches. Zea Mays. As in the above, canals also form on the flat sides of the grain, which communicate with those coming from the margin of Zea mays grains, and like barley and wheat starch grains, the grains are ruptured by these ninnerous canals, which finally cause the complete dissolu- tion of the grain. This is true not only of Graminacea: but also of a large number of other plant orders. In Polygonum fagopyrum, Rheum rhaponticum, Polygonum bistorta, Convolvulus (root), Adoxa moschatelUna (rhizomes), Galanthus nivalis (scales of bulbs). Narcissus poeii- cus, Tulipa, and Gcsneriana (scales of bulbs), the tlissolution of the starch-grain by pore- canals may be observed. Hyacinthus Orientalis. The distinguishing characteristic of erosion in these starch-grains seems to be that the formation of pore-canals may be combined in various ways with the development of inner cavities, each grain thus presenting a different picture. While some grains are eroded by numerous ramified passages, almost equal in uidth, others show merely distinctly outlined cavities of various shapes which naturally have an opening for the ferment at any indifferent part of the surface of the grain. These grains are a striking example of the fact that the diastatic ferment during its action on the starch-grain does not penetrate the starch-sub- stance. The skeleton remaining after the action of the ferment, as well as the tiny broken particles resulting from the breaking up of the grain, show all the characteristics of normal starch-substances. Phaseolus Multiflorus. In these grains also the diastatic ferment enters the grain from without, and the solution of the starch-substance takes place from within after the ferment has passed through these canals into the inner part of the grain. Before germination has begun, the large grains of the cotyledons of Phaseolus have a distinct inner cavity, formed by the basal parts of radial fissures. These fissures differ from ferment passages in that they usually end in a sharp point on the margin of the grain, while the latter, as a rule, have a rounded apex. Wlien the fissures begin to widen and their ends become round we know that the ferment has penetrated the inner part of the grain. They perform as ferment passages the same function in the erosion process as soon as conmiunication has been established between the two. In grains with numerous fissures, the process of erosion may at once appear to be determined by them, inasmuch as they take up the ferment from outside and conduct it in definite passages. This is the case with Phaseolus multiflorus. Wlien this has reached the inner cavity it naturally spreads, and as the hole enlarges dissolution must take place from within. Starch-grains in which erosion takes place from the out- side are usually comparatively large, with an eccentric hilum and distinct lamellae, the latter less numerous at the hilum than at the latter end. The watery lamella) at this part of the grain usually end blindly at the sides, wedging themselves in between the more dense lamella;, usually at some distance from the outer surface, and never known to have free ends. LiLiuM Candidum. The ferment attacks the entire outer surface of the grain with equal intensity, not affect- ing the broad, densely lamellated end more than the hilum end. As soon as ferment action begins, the lateral increase of the outer lamellae must cease, and we then have starch-grains with more or less substantial lamelhe and free ends, according to the intensity of the diastatic action. The sides of these grains exhibit the same characteristics of erosion as those described in Graminacea, the ferment coming from the peri])hery and acting more energetically on the watery layers than on the dense ones, jn'oducing similar bands and protuberances. In germinating seed of Lilium candidum, Phaius, Lathraa, and Orobanche not all of the gi-ains dissolve from the outside. In many grains the margin remains intact until PHENOMENA OF DIGESTIBILITY— RAW STARCHES. 189 the developmcMit of the pore-canals. This is very often accompanied by the formation of an inner caA'itj\ In all such cases the starch dissolves from within, as is usually observed in the smaller grains. The characteristics marking the processes of erosion in Lilium candidum, Lathrcca dandestina, Orobanche, and Phams grandijlorus may be grouped as follows: (1) Erosion usually takes place from the outside; (2) in large grain special local points of erosion often occur in the form of grooves or crater-like depressions; (3) in contradistinction to the above, the small grains usually dissolve from within, inasmuch as pore-canals are developed which by extending to the inner part of the grain often produce a cavity. Erosion of Starch-Grains in Non-Typically Reserve-Storing Parts of the Plant. In parasitic plants like Neoilia 7iidus-ai*is and different species of Lathroea and Oro- banche, during the de^•elopment of the buds a large amount of starch de^•elops in the cells of the flower-stalks. As soon as fructification has taken place and the seed begins to develop, tliis starch disappears from these parts in the same way as from the true reserve receptacles. In the rather small grains of Neoitia nidus-aids the grains are usually attached by pore-canals and are broken by ramification of the same. In Lathrwa and Orobanche the process is as previously described. In non-parasitic plants, as well as in those having oilj' seeds, such as Linum raplianus, Sinapis, Papaver, etc., the same erosion character- istics appear in more or less marked degree. Erosion of Starches Outside the Plant by Action of Watery Extracts of Diastase and by Bacterial Fluids. Baranetzky (page 181) has recorded that most starches dissolve outside the plant in watery solutions of diastase, and Wortmann (page 186) has pro\'cd that bacteria are capa- ble of attacking starch-gi-ains. In Graminacea} Krabbe found that erosion is alike in germinating seeds in extracts of diastase and in bacterial fluids. Slight deviations were noticed in the bacterial fluids. In potato starch, on the other hand, the process of erosion in extracts of diastase and in bacterial fluids differs from that in sprouting tubers. Uni- form dissolution of the grains from without, which is a rule in tubers, seems never to take place outside the jilant in extracts of diastase or bacterial fluids. In the bacterial fluid the starch-grains dcAclop dej^ressions like holes bored into the outer sm'f ace of the gi'ain, and these depressions increase in circumference and depth with the continued act ion of the bacteria ; they naturally merge into each other as the starch-substance separating them gradually dissolves, thus producing large irregular depressions on the outside of the grain. The action of extracts of diastase outside the plant produces superficial irregular points of erosion which soon deepen and enter the inner part of the grain and assume chfferent forms. As a rule, the eroded grains are filled with a network of canals, and inner cavities are also usually formed. Sometimes superficial fissures have been observed in the diastatic action in the grains of potato starch, the ramification of which gave the grains a reticulated appearance. Such branched fissures were observed in the eccentric and distinctly lamcllated grains of Friiillaria imperialis. The fissures begin at the wide end of the grain and lengthen and ramify until they completely cover the outer surface. As they lengthen they also become deeper, and if the chastase is allowed to act long enough they finally cut the grains into pieces. Day {loc. cit.) observed the effects of unfiltered saliva on potato, arrowroot, corn, rice, wheat, and barley starches. After 14 hours at 40° (the preparations each containing a few drops of thymol solution) in most of the corn and rice grains the center part was apparently all dissolved, leaving a thick shell ha^•ing deep clefts along what were the edges of the polyhedral grain; in many of the barley and wheat grains there was no change except corrosion of the surface, but in many others there remained nothing but hollow 190 DIFFERENTIATION AND SPECIFICITY OF STARCHES. shells, thinner than in the corn, and with pitted surfaces and cracks around the edge, much like the openings of oyster shells. At the end of 38 hours the arrowroot and some potato grains showed surface erosions. Some of the potato grains showed change at the small end, this part not reacting with iodine, while the remaining part stained a deep blue. BOILED STARCHES. The results obtained by various investigators in their digestion experiments with boiled starch are quite as conflicting as those with raw starch. It has long been believed that certain kinds of boiled starches are more digestible than others, and it is owing chiefly or solely to this that arrowroot starch is given preference by the physician over other starches in infant feeding and in the diet of the sick-room. Its superior qualities, however, as will be noted from the following pages, are not owing to the factor of digestibility, but to the absence of certain contaminations which render certain other starches of commerce, nota- bly potato starch, unpalatable. In a word, all kinds of boiled starches as ordinarily pre- pared and subjected to the same conditions of experiment are of equal digestibility and yield the same products, quantitatively and qualitatively. In the experiments by Hammarsten (loc. cit.) with saliva on normal raw starch there were material variations in the time of the formation of sugar in the case of different kinds of starch, the interval ranging from 2 to 3 minutes to the other extreme of 2 to 4 hours. Wlien, however, the grains were comininuted by chewing, which is equivalent, as previously shown, to partial or complete gelatinization in proportion to the degree of comminution, sugar was formed in all of the starches in 1 to 4 minutes. Starch-pastes were found by Hammarsten to show no difference in sugar-forming time. Levberg (Inaug. Disser., St. Petersburg, 1874; Ber. d. d. chem. Gcsellsch., 1874, x, 76) gives the order of digestibility of boiled starches, as determined by the quantity of glucose, as follows: ArroAvroot, potato, wheat, and rice — the first being the most digestible. He made one series of experiments in which he determined the quantity of glucose formed when given quantities of saliva and starch were used. With 9 c.c. of saliva he obtained from potato 60.3 per cent and from arrowroot 59.62 per cent; with 10 c.c. of saliva he records from rice 55.76 per cent; and with 16 c.c. of saliva with wheat starch the yield was 62.87 per cent. In other observations he found, with standard quantities of saliva and starch, that the maximum results were recorded as follows: Arrowroot 8 hours, potato 9 hours, wheat 12 hours, and rice 14 hours. Salomon (Jour. f. prakt. Chemie, 1882, xxvi, 324), upon the basis of the amount of sugar formed by the digestion of potato and rice starches, found that these starches are identical in degree of digestibility. O'SuUivan (Jour. Chem. Soc. Trans., 1878, ii, 141) conducted most of his experiments with potato starch. In his earlier experiments he satisfied himself that different boiled starches (potato, rice, wheat, corn, etc.) give the same results, quantitatively and quali- tatively, but in a comparatively recent article (Proc. Chem. Soc. Trans., 1904, xc, 65), in which he reports the results of six series of experiments with potato, malt, barley, corn, and rice starches, using extract of malt or diastase, it is stated that the percentages of dextrin and maltose yielded by potato starch do not correspond with those recorded with the other starches, and therefore that the products of potato starch can not be taken to indicate the products of the other starches. It has since been found that the difference noted by O'Sullivan is due to errors of experiment (see Ford, page 193). The digestibility of all kinds of starch is stated by Butyozin (Inaug. Dissert., St. Petersburg, 1887) to be increased by the length of time they are boiled, and that under given conditions the starches of millet, buckwheat, rice, and peas show ease of digestibility in the order given. In opposition to this author. Day {loc. cit.) records that potato, arrow- root, and probably tapioca and sago starch-pastes are not rendered more digestible by prolonged cooking; and in confirmation, that cereal starches (wheat, corn, rice, barley) PHENOMENA OF DIGESTIBILITY — BOILED STARCHES. 191 become more digestible. According to Day, the raw starch-grains contain three substances which, owing to their different color reactions with iodine, may be distinguished as blue- avujlose, rcd-amylose, and rose-amijlose. Blue-amylose is identical with the substance known as granulose or the j3-amylose of Meyer; red-amylose constitutes the outer layer of the starch-grain and is more difficult of digestion than blue-amylose; rose-amylose forms about 10 per cent of the inside of cereal starches, and is not found in potato, arrow- root, tapioca, and sago starches, and is the least digestible of the three. Blue-amylose constitutes the entire inside of potato, arrowroot, tapioca, and sago starches, and 90 per cent or more of the inside of cereal starches, and is not rendered more digestible by boiling. The digestibility of red-amylose is not increased by boiling, but rose-amylose is slowly changed into blue-amylose by cooking and thus rendered more digestible. Day calls attention to the fact that the skin formed in boiled starch contains reverted starch that gives no color reaction with iodine, and which is very slow of digestion. Finally, starch- paste made below the boiling temperature of water is as easily digested as that which has been boiled for a few minutes. The use of impure gluten for the saccharification of starch, which dates back to Kirch- hoff (1814) and Mathieu de Dombasle (1814), was resumed by Reychler (Ber. d. d. chem. Gesellsch., 1889, xxii, 414), who made preparations of what he termed "artificial diastase." These were prepared by means of dilute acids or acid salts (hydrochloric, phosphoric, acetic, tartaric, and lactic acids, and acid sulphate of potassium, and acid phosphates of alkalies were found most desirable). Well- washed, fresh gluten was subjected to the action of the solvent at 30° to 40° for 4 to 5 hours. Reychler notes that the preparations of wheat flour act like diastase, and that barley contains a diastatic constituent. In a number of experiments with barley, wheat, and corn, the saccharifying action was found to be greatest in the case of barley, and the intensity of action was increased by the addition of potato starch. Grierson (Pharm. Jour. Trans., 1892, xxiii, 187) subjected given amounts of boiled starch, water, and pancreatic extract to a temperature of 37° to 38°, and determined the i^eriod of the reaction with iodine to show the differences in the degrees of digesti- bility of different starches. Corn starch yielded a blue reaction after 20 hours, and wheat and rice starch after 2 hours. Tapioca was colored a weak green after 30 minutes; but tous-le-mois (from Canna eduUs), Bermuda and St. Vincent arrowToots, and potato starch showed no reaction with iodine after 19 minutes. Oat and wheat flour gave a starch reac- tion after 80 minutes. He concludes that tous-le-mois, arrowroots, and potato starch are best for patients with weak digestions. A hne of investigation similar to that followed by Grierson, but covering a somewhat wider field, was pursued by Stone (The Carbohydrates of Wheat, Maize, Flour, and Bread, and the Action of Enzymic Ferments upon Starches of Different Origin, Bull. No. 34, Office Expt. Sta. U. S. Dept. Agriculture, 1896), who carried out a plan to yield comparative rather than absolute results. Care is stated to have been taken to secure identically the same physical condition of aU the starches, exposure under constant conditions of tempera- ture and dilution, and to a uniform solution of diastase. As an index of the energy of the action of the enzyme, a solution of iodine-iodide of potassium was employed, the test being made byremoving 0.5 c.c. of the starch preparation to a watch-glass or porcelain test-plate, and adding a drop of iodine. When the blue reaction could no longer be obtained after repeated tests the end of the experiment was recorded. Malt diastase, salivary enzymes, pancreatic enzymes, and Taka-diastase were used, together with potato, sweet potato, corn, rice, and wheat starches. The main conclusions derived from his investigations are as follows : (1) The starches of sweet potato, maize, rice, and wheat vary greatly in their susceptibility to the action of enzymes. (2) This variation reaches such a degree that under precisely the same conditions certain starches require 80 times as long as others for complete solution or saccharification. 192 DIFFERENTIATION AND SPECIFICITY OF STARCHES. (3) This variation is exhibited toward all of the common cnzymic ferments studied, viz, dias- tase, ptjalin, pancreatin, and Taka-diastase, in the same relative order, with slight exception. (4) This order, beginning with the starch which is most easily changed, is, for vialt extract, sweet potato, potato, wheat, and maize; for saliva, potato, sweet potato, maize, rice, and wheat; for pancreatic fluid, potato, sweet potato, and maize, with wheat and rice unchanged; for Taka-diastase the potato was more quicklj^ changed than any other. (5) It seems reasonable to assume that the same relative degree of susceptibility exhibited bj- these starches in the experiments described would still obtain when they are sub- jected to the action of the same enzymes in the process of digestion. Lindet (Bull. Soc. chim., 1902, xxvii, 634) records that the starch of stale bread is less digestible than that of fresh bread. Roiix (Compt. rend., 1904, cxxxviii, 1356), in his studies of reverted starch, states his belief that stale bread may contain this substance, and Day (loc. cit.) several years later reports its presence in the skin formed on boiled starch. Effront (Enzymes and their Applications, trans, by Prescott, 1902, 128) assumes that both raw and boiled starches of different kinds differ in digestibility because of inherent differences in the starch, which are owing to differences in the compactness of the layers. He states that potato starch and barley starch are both composed of non-homogeneous granules cUffering in degree of compactness of the layers which compose them. In the granules of potato starch more resisting layers are found than in the granules of barley starch, and from chfferent kinds of starch there are obtained pastes which saccharify with more or less chfficulty. It must therefore be assumed that the difference in compactness between parts of the same granule does not disappear when the starch gelatinizes. The more coherent parts of the granules will form a paste more difficult to liquefy (see page 179). Liquefied starch was found by Fernbach and Wolff (Compt. rend., 1905, cxl, 1067) to yield a larger quantity of products of digestion than ordinary starch-pastes. The pastes of cereal starches, they found, were more readily saccharified than the paste of potato starch. In another contribution {ihid., page 1547) they state that the starch from green peas differs from other starches by a high percentage of amylocellulose that is not saccharified. They found that if the freshly boiled starch is at once subjected to diges- tion, complete saccharification occurs, but if allowed to stand an unsaccharifiable amy- locellulose is thrown down. I'ea starch in its natural condition, they state, is analogous to i^otato starch that has been coagulated by amylocoagulase. Roux (Compt. rend. 1906, 95) subjected starch-pastes prepared at 100°, 120°, and 150° to the action of malt extract at 56°, and found that the yields of maltose from different starches were not the same. Calculated on the basis of dry starch, the quantities were: Potato 83.0 per cent, corn 83.3 per cent, wheat 87.1 per cent (prepared at 150°), rice 85.2 per cent, pea 83.8 per cent, tapioca 81.5 per cent. The yields from starches prepared at 100°, 120° and 150° differed but slightly from each other. Ling (Jour. Fed. Inst. Brewing, 1903, ix, 446) found that soluble starches prepared under different concUtions do not yield identical results under the influence of diastase. In an investigation of soluble starch of various origins, means of preparation, and properties. Ford (Jour. Chem. Ind., 1904, xxiii, 414) recorded that there is no doubt that preparations of soluble starches do differ in certain physical characteristics, and that he is of the opinion that when different specimens give different maltose productions with diastase, it is not the starch which causes the variations, but the impurities present in the specimen. Solutions of soluble starch, from whatever source, when equally pure, will give the same maltose production when acted upon by equal amounts of diastase under the same conditions of temperature, etc. The starches experimented with, except barley, were bought commercially and purified by treatment with dilute alkali and acid, being well washed and air-dried. Portions of 15 to 20 grams were gelatinized, and then liquefied at 79° to 80° by means of a trace of precipitated diastase, boiled when limpid, and then PHENOMENA OF DIGESTIBILITY — BOILED STARCHES. 193 made up to 500 c.c. To eacli 70 c.c. of each starch-solution was added 1 c.c. of malt extract at 40°, and the preparation kept at 40° for 1 hour. The preparation was then boiled and made up to 100 c.c. The results in all cases, corrected for reduction of starch-solutions and malt extract, are shown in table 25. Table 25. Kind of starch. CuO per 100 c.c. gram 0.53 0.52 0.52 0.52 0.54 0.52 0.47 0.54 Rice Wheat Maize Potato Bailey Arrowroot, unpurified Arrowroot, Lintner's soluble Table 26. Kind of starch. CuO per 100 o.c. Arrowroot q: gram 0.54 0.53 0.50 0.52 0.53 0.54 0.47 0.44 Rice Wheat Maize Potato Arrowroot ) Potato I It is e\-ident, as Ford states, that the origin of the starch has, under these conditions, no influence on the results when the starches are equally purified. From the same starches soluble starches were prepared according to Lintner's method, and similarly tested, with the results shown in table 26. The specimen designated potato I was made from a "purest commercial farina" by gelatinization and subsequent liquefaction with a trace of diastase. The solution after con- centration was precipitated and washed with tap-water and twice with alcohol. The result, Ford states, shows well how impurities cling to the starch. The slight differences shown in the table, he holds, may be accounted for by variations in the impurities, and one may safely infer that equally purified preparations of Lintner's soluble starch from starches of different origin will give the same maltose production with equal quantities of diastase. The above preparations, though exhaustively washed with distilled water, were by no means pure. Further studies of the quantitative identity of the decomposition products of starches of various origins were made by Ford and Guthrie (Jour. Soc. Chem. Ind., 1905, xxiv, 605). The starches were prepared by these investigators, with the exception of arrowi'oots, which were bought and then subjected to purification in the usual manner. All of the starches were purified by treatment with 0.2 per cent caustic soda and hydrochloric acid, washed with distilled water, and extracted with alcohol and ether when necessary; then air-dried and further dried at 25° to 30°. In each experiment 5 grams of starch were gela- tinized as usual, cooled to 05°, and converted during VA hours with 25 c.c. of malt extract that had previously been heated for 10 to 15 minutes at 65°. The converted preparations were made up to definite volumes and the specific gravity, optical rotation, and copper- reducing power were observed. Their experience convinced them of the validity of "the law of relationship" pronounced by Brown, Morris, and Millar (Jour. Chem. Soc, 1897, Lxxi, 115) ; hence, determinations of the solids and their optical rotation is all that is neces- sary. The following rotatory values were recorded (table 27) : Table 27. Kind of starch. Potato (controls of same specimen) Pea Barley Wheat Oats 13 (0)r 161.0 160.7 159.5 160.7 161.0 160.8 161.0 162.0 Kind of starch. Corn Tous-les-mois (Canna edulis) Arrowroot (Maranta) Rice Lentil Banana Barley, m.alted Barley, malted Mt 160.0 159.7 161.0 161.5 160.0 159.7 160.7 160.8 Kind of starch. Potato (commercial) Potato (purified). . . . Rye Buckwheat Chestnut Pari Millet (o)i 163.4 161.0 163.2 164.0 166.9 163.7 162.6 194 DIFFERENTIATION AND SPECIFICITY OF STARCHES. Table 28. Kind of starch. ("'dj.h E =3.93 Maize 174.7 175.6 174.0 175.4 174.4 175.0 174.5 40.5 43.8 43.8 40.7 45.0 44.0 43.9 Wheat Potato Arrowroot Barley, malted. . . Rice Potato Ta BLE 29. Kind of starch. ("Jo 3. 93 R=3.93 Banana Oat Barley Potato 153.1 155.4 154.3 154.2 156.4 74.0 72.6 74.5 73.3 74.4 Pari They state that the last fi^'e starches were obviously impure, and that they did not yield homogeneous pastes. Successive purifications lowered the angle in each case. The chestnut starch contained evident traces of something other than starch. In another series of experiments they used precipitated diastase in the proportion of 0.2 gram per 5 grams of starch at 60° for 1 hour, with the results shown in table 28. In another series, with 0.4 gram of precipitated diastase to 5 grams of starch for 2 hours at 60°, they recorded the results shown in table 29. In an earlier investigation. Ford showed that metallic and other impurities influence amylolytic activity, and that dried starch readily absorbs sub- stances which may interfere with hydrolysis. Ford and Gutlirie state that the contracUctory results of O'Sullivan and themselves are to be attributed to impurity of the potato starch used by 0'Sulli\'an. A specimen of this starch was obtained from O'Sul- livan, and was found to contain copper to the extent of 0.035 per cent. They determined experimentally that this amount of copper is sufficient to bring about a marked reduction in maltose formation. It seems, therefore, cjuite clear from the results of the investigations of both Ford, and Ford and Guthrie, that the source of starch is without influence on the products of decomposition if the starches are suffi- ciently purified and freed from substances which influence amylolytic action. That all boiled starches, whatever the source, yield the same decomposition products, quantitatively and qualitatively, under the same conditions of experiment, has been shown by the author's work. To obtain uniform results it is essential to secure as nearly as pos- sible the same conditions of experiment. It seems to be a universally accepted belief that enzymes in a 1 to 2 per cent solution of chloroform, or in other antiseptic solutions which are inert towards the enzyme, do not undergo rapid impairment in energy. While this is in a large measure justified in regard to plant enzymes, such is not the case with amylopsin and ptyalin and probably all animal enzymes. Even within 24 hours such diastatic enzymes show deterioration, which steadily increases with time. Therefore, in making compar- ative experiments the solution of enzyme should be made immediately before using. It is also of importance that none of the starch be lost after boiling, by adherence to the sides of the vessel, and that other specific standard conditions be maintained. A large number of experiments were carried out with Merck's "pure ptyalin," "pure pancreatin," and "medicinal malt diastase," and also with Parke, Davis & Company's Taka-diastase. These enzymes showed marked differences in energy, Taka-diastase being the most ener- getic, pancreatin a shade less energetic, malt diastase distinctly less energetic than the foregoing, and ptyalin comparatively weak. It is, therefore, manifest that, in making comparative studies of different starches, a given enzyme must be used throughout. More- over, as the different preparations of the same enzyme differ widely in energy and also in their content of glucase, etc., the same preparation must be used throughout for control experiments made to permit of proper corrections. A number of experiments were made with starches prepared from Solanum tuherosum, Zea mays, Batatas edulis, Canna roscoeana, Canna loarszewiczii, Canna musa^folia, Canna edulis, Maranta arundinacea, Freesia refracta var. alba, Dieffenbachia seguine, and other starches derived from widely separated genera, orders, etc. In each experiment 0.5 gram of starch was boiled in 35 c.c. of water for 1 minute, the preparation cooled, the starch SUMMARY AND CONCLUSIONS. 195 adherent to the walls of the beaker washed out, and the preparation made up to 50 c.c, and then warmed in a water-bath to 37°. To 5 c.c. of water was then added 0.75 gram of pancreatin rubbed up to a paste in a mortar, and water gradually added to make 25 c.c, and this preparation likewise placed in a water-bath warmed to 37°. The starch and enzyme preparations were then mixed antl jilaced in a moist chamber having a maintained tem- perature of 37°. At 15-minute intervals during the first hour the preparation was shaken, and 5 c.c. removed, quickly boiled, and tested with a 2 per cent Lugol's solution and with Purdy's solution. Two immediately subsequent observations were made at half-hour intervals, and subsequently at 1-hour, and later at 2-hour intervals, the entire period of actual experiment covering G hours. The record given in table 30 may be taken as being practically absolutely identical with that of every starch, no matter what its source, the de\'iations from this being absolutely insignificant, as shown by the sugar determina- tions, which vary only in the third decimal figure. The sugar determinations were made in terms of maltose. The pancreatin contained an appreciable amount of glucase, giving rise to the conversion of some of the maltose into glucose, thus yielding higher values than had the sugar consisted solely of maltose. Table 30. Conditions of experiment. Time from beginning of experiment. Color reaction with iodine. Weight of sugar (mal- tose+gIuco8e) in terms of maltose. Kind of starch: SoLanuni tubcro- ^ sum. 50 c.c. of 1 per cent boiled starch, with 0.75 gram of pan- creatin in 25 c.c. of water, both having been heated to 37°, and • then mixed, and placed in moist chamber at 37° constant tem- perature. Sug.ar determination by Purdy's solution. hrs. min. 0 15 0 30 0 45 1 00 1 30 2 00 4 00 6 00 gram. 0.334 0.362 0.396 0.409 0.423 0.436 0.488 0.528 Very weak violet Faint violet Fainter violet Trace of violet Barest trace of violet. . Do No color reaction These results coupled with those of Ford, and of Ford and Guthrie, leave no doubt that under the same conditions of experiment all boiled starches, from whatever source, are practically absolutely identical in digestibility. SUMMARY AND CONCLUSIONS. Of the many methods and reagents used in the differentiation of different starches, different grains of the same starch, and different parts of the same grain, it is obvious that their relative values extend within very wide limits. Many of the results recorded are fallacious, owing to the presence of foreign matter or of other incidental conditions; and very frequently the reports of one observer are not confirmed, or are absolutely contra- dicted, by those of others, even when the same method or reagent had been employed. In many instances the cause of the different findings has been rendered quite obvious, as in the discrepancies in the reports on the digestibility of boiled starches; but in others it is not clear, as, for instance, in the repeated statements of the digestibility of raw starch under conditions in which it is asserted bacterial action had been prevented. There will be found, therefore, in the following paragraphs several statements ^^•hich in the light of these contradictions may be regarded as being tentative. The following statements cover in a brief way the more essential points embraced in the literature referred to in this chapter : (1) That the histological method is of great value in the differentiation of starches from different sources, different grains of the same starch, and different parts of the same grain, but that this method of itself, if solely depended upon, to diagnose different kinds 196 DIFFERENTIATION AND SPECIFICITY OF STARCHES. of starch, may be insufficient and even absolutely misleading. Therefore, for the positive identification of any given kind of starch other methods of differentiation should be coupled with the histological method. (2) That young grains and mature grains may show differences not only in their histo- logical characteristics, but also in their chemical and other reactions. The young grain serves as a nucleus for the deposition of starch during the development of the mature grain, it lacks lamellar structure, it shows a higher phosphorus content, it is more resistant to the actions of both dry and moist heat, it is more resistant to enzymes and dilute acids, and it exhibits differences in behavior towards iodine and aniline dyes. Likewise, the outer coat of the starch-grain shows differences from the inner part. (3) That the starches from different somxes, when subjected to the ordinary labora- tory procedures of differential study, show no differences in elementary composition; that different starches show different percentages of water, starch-substance, kinds of starch-substance, digestible matter, fat, furfurol contents, protein, ash, etc. ; that unim- portant differences may be observed in specific gravity and in heat values; that starches differ in the degree of acidity, and that the reaction is variable in accordance with the indicator, as, for instance, acid to phenolpthaleine, alkaline to methyl orange, and neutral to rosolic acid, and that different starches may react differently with the same reagent; that phosphates constitute a more or less important element in the digestibility of starches, not only of themselves directly, but also in the neutralization of alkaline impurities. (4) That the color reactions of different starches, different grains of the same starch, and different parts of the same starch-grain may not be the same. (5) That the different swelling reagents show not only quantitative and qualitative differences in relation to different kinds of starches, and even to different grains of the same starch, but also that some at least of these possess individual characters by which their reactions may be distingiiished from those of others. (6) That the starches of different kinds show quite a range of temperature at which gelatinization begins and at which it is complete, and that in a given specimen the larger grains tend to gelatinize at lower temperatures than the smaller grains. (7) That the reactions in polarized light, without and with a selenite plate, may vary markedly, not only in starches from different sources, but also in different grains and in different aspects of the same grain. (8) That the starch-pastes and pseudo-solutions obtained from different starches are not identical in their stiffness and in their ijenetrability in relation to fabrics; that the stiffness of the paste is affected by the length of the period of boiling, a longer period increasing the stiffness of the pastes of the certain starches, but decreasing it in others; and that gelatinized starch tends to undergo re\-ersion, and therefore becomes less digestible. (9) That if uninjured starch-grains are subjected in vitro to the actions of enzymes under strictly aseptic conditions, absolutely no digestion occurs; but if the grains be broken, eroded, or cracked, etc., or if bacterial action is not prevented, a more or less rapid erosion takes place, which is more or less peculiar to the kind of starch and corresponds with the peculiarities of the erosion phenomena observed when the starch is in situ; that oxj'gen may be necessary in the saccharification of starch, as has been indicated in enzymic and bacterial processes. (10) That boiled starches, whatever their source, are of equal digestibility, yielding quantitatively and qualitatively the same saccharine products, provided the conditions of experiment are the same. Different starches may nevertheless have different values as articles of diet, owing to the presence in some of unpalatable or other forms of contami- nation. Corn starch, for instance, should be equally as good as the comparatively expen- sive arrowroot, but potato starch, which is of equal digestibility, has a comparatively low value because of its being or becoming unpalatable. CHAPTER V. A SYSTEMATIC SUMMARY OF THE GROSS HISTOLOGICAL PROPERTIES OF STARCHES FROM VARIOUS SOURCES. NAGELI'S CLASSIFICATION OF STARCHES FROM DIFFERENT SOURCES. Nageli, in his elaborate memoir on Die Starkekorner (loc. cit.), recorded the results of the histological examinations by Raspail, Payen, Soubeiran, Schleiden, Bischoff, Criiger, Walpers, Berg, Miinter, Fritzsche, Harting, and others, including his own observations, of over 1200 starches from different sources. These starches he arranged in 3 classes, 5 sub- classes, and 17 types. As this pubhcation has long since been relegated to the anti- quaria, as virtually no notice is to be found in recent works of this part of his verj' laborious investigations, as the memoir is a rarity even in the largest of our libraries, and as this data has especial importance apropos of the present research and is of general botanical and biological interest, it seemed very desirable to include these valuable records in the present memoir. In recent literature a large number of descriptions of the forms of different starches will be found, but as such publications are generally readily available these refer- ences have not been included. The following text is a free translation of Chapter XII, almost in its entirety, of Niigeli's memoir; but many changes have been made in the Latin names, in part in the correction of typographical errors, and in part to bring the matter up to date. Nageli states that the different types of starch-grains merge almost imperceptibly into one another, and therefore that the boundaries are necessarily arbitrary ones; and, moreover, that very similar starch-grains are often placed far from one another in the different types. The author states that the grains might always have been arranged systematically within the individual classes according to their shape and structure, h\it that such a comparative classification would scarcely be possible without a complete col- lection of microscopical preparations. He therefore thought it better to make a classifica- tion according to other principles. In each group the seeds were considered separately from other parts of the plants and the latter were enumerated in systematic order. The natural relationships often determined to what type the grains belong. Thus, the starch- grains from all the Bromus seeds, although rather varied, are placed together; the grains of the seeds of Alis7>mcece are classified in one tj^pe, while those of Butomacece are placed in another, although their grains in general are very similar. Even if the starch of the species of one order completely agrees with that of the other order they are thus classified, because on the whole the grains of Alismacece follow more closely a given type, and those of Butomacece another type. It follows from this, he states, that the descriptions are somewhat more diffuse, less sharply drawn, and correspondingly less diagnostic than they would be in case of a comparative systematic arrangement. They have throughout not the significance of differential characters, and this the less so since it has not been deter- mined which are specific and which are merely individual characteristics. Especial con- sideration has been given to the relative size, mostly to linear dimensions. The following is Nageli's classification of starches from different sources: A. C.RAINS Simple. I. Centric. Hilum in the mathematical center; lamellae always equal at two corresponding diametrically o]ipo.site points. Type 1. Spherical. When the grain i.s free both hilum and grain are spherical. 197 198 DIFFERENTIATION AND SPECIFICITY OF STARCHES. A. Grains Simple — continued. I. Centric — continued. Type 2. Lenticular. When the grain is free both hilum and grain are rounded; grains compressed;* sometimes circular or ovoid; sometimes triangular or quadrangular. Type 3. Oval. When the grain is free both hilum and grain are oval to lanceolate-oval; occasionally kidney-shaped or somewhat curved; when on end they ajipear circular or somewhat compressed. Type 4. Spindle-shaped. Grain linear or lanceolate, tapering towards the pointed ends, or of equal-width blunt ends; when on end they appear almost circular. Type 5. Bone-shaped. Grain elongated and compressed from the narrow aspect, but linear spindle-shaped from the broad aspect, with enlarged laminated ends. II. Eccentric. Hilum usually more or less removed from the mathematical center of the grain; lamelloe coarsest and finest at opposite ends of the grain, respectively. Type 6. Inverted cone-shaped. Grain on end almost circular; more slender at the hilum end. Type 7. Cone-shaped. Grains on end almost circular; decidedly thicker and broader at the hilum end. Type 8. Wedge-shaped or compressed. Grain compressed,* of equal thickness throughout, or thicker but narrower at the hilum end than at the distal end. Type 9. Rod-shaped. III. Grains simple and structure obscure. Type 10. Structure not fully developed or not identified, owing to diminutive size of the grains. Lamella?, hila, cavities, fissures, and clefts seldom observed. B. Grains Semicompound. Type 11. Grains semicompound. The component part-grains are enveloped by a com- mon substance. C. Grains Compound. The component part-grains not enveloped by a common substance. I. Composed of fused part-grains. Type 12. Composed of fused part-grains. The part-grains are not separated by fissures, and even different grains may be fused with one another. II. Composed of separated paii-grains. The part-grains separated by fissures. Type 13. Grains in 1 or 2 rows. From 3 to 11 components arranged in 1 or 2 rows. Type 14. Equally divided grains of few components. From 2 to 10 or more almost equal sized part-grains which, when separated, have one curved surface and one or more pressure facets. Type 15. Unequally divided grains of few coynponents. From 2 to 10 or more unequal sized, firmly united part-grains, which when separated have one curved surface and several flattened pressure facets. Type 16. Multiple grains. From 20 to many thousand firmly united part-grains which, when separated, are covered with jiressure facets. Type 17. Holloio spherical grains. The part-grains are arranged in a spherical layer, as if a globular shell had been divided radially. , For further details of the characteristics of the various types, see as follows: Type 1, page 198; type 2, page 203; type 3, page 207; type 4, page 212; type 5, page 213; type 6, page 213; type 7, page 214; type 8, page 221 ; type 9, page 229; type 10, page 232; type 11, page 251; type 12, page 252; type 13, page 254; type 14, page 255; type 15, page 268; type 16, page 273; type 17, page 293. Type 1. Grains Simple, Centric, Spherical. Grains spherical or oval-spherical, and more or less polyhedral when crowded. Hilum in center, spherical. Lamellfc of equal thickness; stria? of the dried grain radiate in all directions. As far as known, Nageli states, these grains exist only in seeds. In the other parts of the plant, the underground parts for example, spherical grains with a central hilum may occur. Judging from their * Nageli, in his descriptions, uses the words zusaminengedriiekt (compressed) and abgeplattet (flattened), but since they are botli appHed to the effect of pressure from abore they may be regarded as synonymous, and this has been adopted at times in the translation. TYPE 1. GRAINS SIMPLE, CENTRIC, SPHERICAL. 199 rather diminutive size, they may be imdeveloped forms of an eccentric type, and the fact that eccentric grains are found in rchited plants makes the liling twice as long as broad; later, mostly polyhedral with sharp edges; hilum distinct, very rarely with delicate lamelliB surrounding it. Size about IG to 2\fi. The dried grain usually has a central cavity with radial fissures. In the fresh, not fully developed grains several short radial fissures also occur, as well as grains with granular outer surfaces. Irregular compound grains of 2 to 6 part-grains are often found among the simple ones. The grains in the outer part of the seed are somewhat smaller (about IG/i) than in the inner part (about 21//). According to Payen (Ann. Sc. Nat., 1838, ii, p. 23), the starch- grains of maize average 30/x and the horn-like part of the seed contains crowded polyhedral grains, while the inner mealy parts inclose more rounded, loosely arranged ones. Coix lacryma Linn. {Graminacece.) Dry seed. — Grains spherical, frequently more or less angular owing to pressure ; no lamella; ; solid or with a small central cavity having sometimes numer- ous fissures radiating from it. Size about 12 to 16yu. Paspaln7n dilatalum Poir. {Graminacece.) Dry seed. — Grains rounded or oval, many of them angu- lar or polj'hedral owing to pressure; no lamellae; sometimes with small central cavity. Size about 6 to 7/i. Paspalum platycaule Poir; P. complanalum Nees. {Graminacece.) Dry seed. — Grains as in P. dila- talum, the larger ones with central cavity from which several short fissures radiate. Size about Sn, rarely up to 12/i. Paspalum stoloniferum Bosc; Maizilla sloloniferum Schlecht. {Graminacece.) Dry seed. — Grains angular with roimded corners to polyhedral, with sharp edges and angles; the larger ones with a central cavity from which short fissures sometimes radiate. Size about 12/i. The grains which have fallen out of the cell are often clumped together, resembling compound grains. Amphicarpum purshii Kunth. {Graminacece.) Dry seed. — Grains spherical to almost polyhedral; no lamella?; the majority with single short fissures radiating from the center. Size about 19/i. Olyra paniculata Swartz. {Graminacca\) Dry seed. — Grains polyhedral, filling the cells; the largerones mostly with a large or a small cavity, and rarely with single, delicate fissures. Size about 12/i. Oplis)nemis colonus Humb., Kunth.; Panicum colonum Linn. {Graminacece.) Dry seed. — Grains rounded, sometimes angular owing to pressure, with small central cavity from which fissures radiate. Size about 10/i. Oplisrnenus frumentaceus Kimth.; Panicum frumentaceum, Roxb. {Gram,inacece.) Dry seed. — Grains spherical or oval-sj^herical, sometimes angular owing to pressure; the largest ones some- times somewhat shrunken; with a central cavity, rarely with short radiating fissures. Size about 14 to 18/i. Setaria glauca Brauv.; Panicum glaucum. Linn. {Graminacece.) Dry seed. — Grains spherical or oval- spherical, often angular owing to pressure; often with small central cavity, with radiating fissure; sometimes appears to be split into several part-grains by these fissures. Size about 8/i. Single grains with granular surfaces. Setaria italica Beauv.; Panicum itcdica Linn. {Graminacece.) Dry seed. — Grains during early devel- opment spherical or oval-spherical, but later polyhedral with sharp edges; lao lamellae; often with distinct central hilum. Size about 14/^; according to Payen sometimes 16//. Fre- quently some of the grains have granular surfaces, probably the result of abnormal solution. Compound grains of 2 to 3 part-grains appear with the single ones; these can best be observed in not fully developed seeds. Setaria flava Kunth; Panicum alopecuroideum Schreb. {Graminacece.) Dry seed. — Grains as above; increasing in size from the surface to the middle of the endosperm. Isachne au.stralis R. Br.; Panicum antipodum Spreng. {Graminacece.) Dry seed. — Grains angular with roundish corners to polyhedral (filling the cells); the larger ones usually with central cavity and frequently with radial fissures. Size about 13/i. 200 DIFFERENTIATION AND SPECIFICITY OF STARCHES. Panicum niiliaccum Linn. (Graminacccc.) Dry seed. — Grains spherical or oval-spherical, fre- quently somewhat polyhedral owing to pressure; usually with small central cavity from which sometimes a few fissures radiate. Size about 7ju; according to Payen, lOfi. Panicum acuminatum Swartz. (Graminaceae.) Dry seed. — Grains angular, with rounded corners, to polyhedral (filling the cells); usually with a cavity and sometimes with radial fissures, giving the single grains the appearance of compound grains. Size about 12 to 15yu. Panicum tonsum Steud.; Tricholcena tonsa Nees. (firaminacem.) Dry seed. — Grains angular with rounded corners, or polyhedral with blimt, or, more often, sharp corners and edges; the larger ones hollow in the center. Size about 12/i. Panicum hoffmannseggii R. and Sch. (Graminacew.) Dry seed. — Grains angular with rounded corners, or polj'hedral; frequently with large or small angular cavity. Size about lOju. Helopus annulatus Nees. {Graminacew.) Dry seed. — Grains spherical or somewhat angular owing to pressure; with central cavity from which fissures sometimes radiate. Size about 7 to 8/x. Pennisetum longistylum Hochst. {Graminacew.) Dry seed. — Grains roundish, frequently somewhat angular owing to pressure; usually with a smaller or larger central cavity from which delicate short fissures radiate. Size about 15 to 20/i. Pennisehtm cenchroides Rich. {Graminacew.) Dry seed. — Grains spherical, sometimes angular, often with a central cavity and delicate radial fissures. Size about 15/x. Some have a granular surface. Other seeds designated Gymnothrix cenchroides R. and Sch., have spherical to rounded-oval, frequently somewhat angular, grains with central cavity and numerous fissures. Size about 11^. Penidllaria spicata Willd.; P. pluckenetti Link; Pennisetum typhoideumRlch. {Graminacew.) Dry seed. — Grains usually spherical, more or less polyhedral, owing to pressure; frequently with a small central cavity, rarely with radial fissures. Size about 12 to 15/^. Anthephora elegans Schreb.; Cencliruslwvigatus Trin.; Tripsacum hermaphroditum Linn. fil. {Gram- inacew.) Dry seed. — Grains spherical, usually angular owing to pressure, and sometimes polyhedral with sharp corners and edges; sometimes delicate concentric lamellie; usually a small, central cavity from which some or at times numerous deep fissures radiate. Size about 36/Li, rarely as much as Sl^i. The grains are rarely compressed, sometimes the larger ones fall to pieces. Among them are found some compound grains with few divisions, and some grains that have separated. The separated grains have one hemispherical and one or two plane surfaces, which distinguish them from the simple, polyhedral grains, wliich are more or less flattened on all sides. Lappago racemosa Willd. {Graminacew.) Fresh endosperm. — Grains polyhedral, with rather sharp edges and angles; the larger ones hollow. Size about 9(U. Although the grains are sometimes clumped together, they nevertheless appear to belong to the simple type. Lopholepis ornithoccphala Decsn. {Graminacew.) Dry seed. — Grains angular, with rounded corners to polyhedral; with, small, rarely large, usually stellate angular cavity. Size about ll^u. The protoplasmic cells are crowded with starch-grains ; there are no indications of compound grains. Centotheca lappacea Beauv.; Cenchrus lappaceus Linn. {Graminacew.) Dry seed. — Grains spherical, sometimes slightly angular owing to pressure; with or without indistinct lamella'; with central cavity from which fissures usually radiate. Size about 25 to 30yu. Beckera petiolaris Kochst. {Graminacew.) Dry seed. — Grains rounded to nearly polyhedral; with central cavity and radial fissures. Size about llji. Ampelodesmos tenax Link.; Arundo tenax Vahl. {Graminacew.) Dry seed. — Grains spherical, angular with rounded corners, or polyhedral; usually with small or large central cavity, and some- times with radial fissures. Size about 18 to 21fi. Pappophonum nigricans R. Br. {Graminacew.) Dry seed. — Grains circular or ovoid; the smaller ones spherical; the larger ones compressed to about one-half their width; with a central cavity and radial fissures. Size about 25 to 30/i. Notwithstanding the pressure facets the larger grains Ijelong to the spherical rather than to the lenticular type, because they show the same radial fissures in their narrow aspect as in the Ijroad one. Gymnopogon foliosus Nees. {Graminacew.) Dry seed. — Grains angular with rounded corners to poly- hedral; with a small or large square cavity or one with radial fissures. Size about 16^. These starch-grains iloubtless belong to the simple spherical type. No indications of compound grains were present even within tlie cells. TYPE 1. GRAINS SIMPLE, CENTRIC, SPHERICAL. 201 Uniola latifolia Michx. {Graminacece.) Dry seed. — Grains rounded, usuallj' more or less polyhedral owing to pressure, slightly if at all compressed, usually hollow. Size about 20ju. Chusquea cumingii. Nees. (Graminacece.) Dry seed. — Grains rounded, angular with rounded corners to polyhedral; the larger ones with central cavity and some with radial fissures. Size about 15/i. No indications of compound grains within the cell. Orthodada laxa Beauv. {Graminacece.) Dry seed. — Grains angular with rounded corners, or poly- hedral; lamellae, if any, delicate, homogeneous, or very slight. Size about 35 to 40/i. The prcjtoplasmic cells, the outlines of which are often quite indistinct, are filled with two kinds of grains of unequal size, the smaller lying in between the larger. Hemarthriu fascicidata Kunth. {Graminacew.) Dry seed. — Grains spherical or oval-.spherical, usually polyhedral, or sometimes angular owing to pressure; the larger ones with usually a small central cavity and several radial fissures. Size about 15/i. Rotthcella arundinacea Hochst. {Graminacece.) Dry seed. — Grains spherical or oval-spherical, some- times angular owing to pressure; the larger ones usually with a small central cavity, and frequently with several radial fissures. Size about lO/i. Some compound grains of few com- ponents (tj'pe 14). Manisuris granularis Swartz. (Graminacece.) Dry seed. — Grains rounded or angular with rounded corners, frequently spherical or oval-spherical; the larger ones with a central cavity, and usually with a few radial fissures. Size about 12ii. Andropogon dissitifloriis Mich.x. {Graminacece.) Dry seed. — Grains sjiherical or oval-spherical, and angular with rounded corners; with a central cavity and radial fissures. Size about 15;u. Single separated-grains with one curved surface and one to three pressure facets indicate the former existence of compound grains. Andropogon contortus Linn.; Heteropogen contorlus R. and S. (Graminacece.) Dry seed. — Grains rounded or angular to polyhedral owing to pressure; with a small central cavity, and some radial fissures. Size about 15 to 21yu. Andropogon dircrsiflorus Steud. (Graminacece.) Dry seed. — Grains angular, with rounded corners, to polj'hedral; with small central cavity and a few radial fissures. Size about 16 to 21yu. Andropogon leucostachyus Humb., Bonp.; Hypogyniumcajnpestre 'Nees. (Graminacece.) Dry seed. — • Grains spherical, rounded-oval, or angular with rounded corners; with central cavity and radial fissures. Size about 20yu. Among them isolated compound grains consisting of a few components may be found; also separated-grains with one curved and 1 to 3 plane surfaces; cavity with single radial fissures. Andropogon iscluemum Linn. (Graminacece.) Dry seed. — Grains spherical, or with rounded corners, or polyhedral; with small central cavity and a few radial fissures. Size about 13;u. No compound grains; a few separated-grains, as in Andropogon leucostachyus, appear to be present. Andropogon umbrosus Hochst. (Graminacece.) Dry seed. — Grains rounded, more or less angular to polyhedral, owing to pressure; with small central cavity and single radiating fissures. Size about 15 to 19^. Andropogon laguroidesTiC (Graminacece.) Dry seed. — Grains spherical, oval -spherical, and angu- lar with rounded corners; with large or small cavity and ridial fissures. Size about \T^^l. Some separated grains with 1 to 3 pressure facets appear to be present. Andropogon argenteus DC. (Graminacece.) Dry seed. — Grains spherical, frequently somewhat angular o^sdng to pressure; with rather large cavity. Size about 14ju. Seeds not entirely ripe. Andropogon cernus Roxb. ; Sorghum cernus Willd. (GraminacecE.) Dry seed. — Grains as in above, solid or with small central cavitj' and single radiating fissure; sometimes split by the fissures into what appear to be separate grains. Size about 13;u. Some grains have a granular surface. Andropogon sorghum Brot.; Sorghum vulgare Pers. (Graminacece.) Dry and fresh seed. — Grains sjiherical or oval-spherical, sometimes more or less polyhedral owing to pressure; fresh, with distinct central hilum, rarely with radial fissures; dry, usually with central cavity from which fissures radiate. Size aijout 11 to 15/j. According to Payen, the diameter is 30/i. Among these are found compound grains of 2 to 4 or 5 part-grains. Andropogon aciculatus Retz.; Chrysopogon aciculatus Trin. (Graminacece.) Dry seed. — Grains spher- ical or rounded-oval; with large or small central cavity from which frequently fissures radiate. Size about llju. Among these are found some doublets and triplets. 202 DIFFERENTIATION AND SPECIFICITY OF STARCHES. Andropogon ncpalcnsis Hort. berol. {Graviinacece.) Dry seed. — Grains spherical, occasionallj' by means of pressure somewhat angular; with small central cavity from which single or several fissures radiate. Size about 13 to 17^. Anthestiria cymbaria Roxb. ; Andropogon cymbarius Linn. (Graminacece.) Dry seed. — Grains rounded, or rounded-oval, sometimes somewhat angular to polyhedral owing to pressure. Size about 20 to 25m. Anthestiria pseudo-cymbaria Steud.; Andropogon cymbarius Hochst. {Graminacece.) Dry seed.^ Grains rounded, angular with rounded corners to polyhedral; with central cavity and radial fissures. Size about 32 to 37/i. Anthestiria laxa Andr. {Graminacece.) Dry seed. — Grains spherical to oval, often more or less angular owing to pressure; with small central cavity and marked radial fissures. Size about 16 to 20/i. Androscepia gigantea Brongn.; Apluda gigantea Spreng. {Graminacem.) Dry seed. — Grains spherical or oval-spherical, many of them with rounded corners, or even polyhedral, owing to pres- sure; usually with a small cavity and several radial fissures. Size about 20 to 26/i. Imperata arundinacea Cyrill. {Graminacece.) Dry seed. — Grains sjiherical or oval-spherical, angular with rounded corners; with central cavity and radiating fissures. Size about 13 to 17^- Among these are found separated-grains with one curved and one to four plane surfaces; hollow, and with several radial fissures, size about 3 to 11^- Compound grains of 2 to 5 equal part-grains are also found. Saccharum spontaneuvi Linn. {Graminacew.) Dry seed. — Grains spherical or oval, or rounded- angular; with a large or small cavity. Size about 9/i. Among these are some separated- grains with one curved and one to three plane surfaces; also hollow. Size about 7/n. Com- pound grains of few equal-sized part-grains are occasionally present. The seeds were not quite ripe. Erianthxis ravennce Beauv.; Saccharwn ravennce Murr. {Graminacece.) Dry seed. — Grains spherical or oval-spherical, some of them almost polyhedral owing to pressure; with central cavity, usually large, and radial fissures. Size about 21/i. Among them numerous separated-grains with one curved and 1 to 4 plane surfaces; also with a cavity and fissures. Size 4 to 15ju. Few compound grains of 2 to 5 equal-sized part-grains are occasionally present. Zoysia tenuifolia Willd. {Graminacece.) Dry seed. — Grains spherical, oval-spherical, angular with rounded corners or almost polyhedral; usually with a central cavity, and frequently with radial fissures. Size about lUfj.. Hohcnbergia strobilacea Schult. fil. {Brojueliacece.) Dry seed. — Grains splierical or oval-spherical, frequently somewhat angular owing to pressure, sometimes irregular; no lamellae; often with a small central cavity without fissures. Size about 15/i. Among them compound grains of few, equal-sized part-grains (type 14). Billbergia zebrina Lindl. {Bromeliacece.) Dry seed. — Grains rounded, sometimes irregular; no lam- ellae; with central cavity from which fissures usually radiate. Size about 15/i. Among them compound grains of few equal-sized part-grains (type 14). Pitcairnia (dbuccefolia Schrad.; P. punicea Lindl. {Bromeliacece.) Dry seed. — Grains more or less spherical, occasionally somewhat irregular; without lamellaj, mostly with large central cavity from which fissures sometimes radiate; some of them somewhat shrunken, due to drying. Size about 14 to 18//. Among them some compound grains of few parts (type 14). Pterostegia drymarioides F. ]\L {Polygonacece.) Dry seed. — Grains rounded, more or less angular owing to pressure; with small or large central cavity. Size about lO/i. Oxyria digyna Campd. {Polygonacece.) Dry seed. — Grains angular with rounded corners to poly- hedral with rather sharp corners and edges; the small ones solid, the larger ones with small central cavity, and the largest ones with single radial fissures. Size about lO/u. The grains as they come from the cell frequently adhere to each other in groups or in rows. Rheum hybridum Ait.; Rheum rhaponticum Linn. {Polygonacece.) Dry seed. — Grains exactly spherical, sometimes somewhat angular owing to pressure; without lamellae; with small central cavity, the largest sometimes have a few very short radial fissures. Size about 13/u. Polygonum linctorium Lour. {Polygonacece.) Dry seed. — Grains rounded, somewhat angular owing to pressure; with large or small hole. Size about 13/i. TYPE 2. GRAINS SIMPLE, CENTRIC, LENTICULAR. 203 Polygonum oricntale Linn. {Polygonacecc.) Fresh seed. — Grains spherical, or somewhat angul.T,r owing to pressure; hihim usually distinct. Size about lOyu. Croups of adhering grains and also compound grains occur, the latter not easily distinguished from the former. Some semicompound forms with several hila were also observed. Fagopyrum cymosum Meisn. {Polygonacecc.) Fresh seed. — Grains spherical or oval-spherical, usu- ally more or less polyhedral owing to pressure; no lamella;; the larger grains with a distinct central hilum; when dry, with a cavity, not often with radial fissures. Size of the spherical about 17/i and that of the oval about 2lfi. Fagopyrum esculentum Moench.; Polygonum fagopyrum Linn. (Polygonacece.) Dry seed. — Grains spherical or oval, frequently more or less angular or even polyhedral owing to pressure; with small central cavity from which, at times, single short fissures radiate. Size about 10 to 12/i. Among them compound grains with few part-grains in one or two rows. Emex spijiosa Cambess. {Polygonacew). Dry seed. — Grains spherical, or angular with rounded corners owing to pressure; the larger ones with a small central cavity from which frequently delicate fissures reach the surface. Size about 8n, rarely lOyu. Rumex ratientia Linn. (Polygonacew.) Dry seed. — Grains spherical, or more or less slightly angular owing to pressure; usually with large cavity. Size about 12 to 14;u. Tragopyrum lanceolatum Biclirst. (Polygonacece.) Dry seed. — Grains rounded, almost angular, occasionally with small central cavity. Size about Oyu. Airaphaxis spinosa Linn. (Polygonacece.) Dry seed. — Grains spherical, frequently more or less angu- lar owing to pressure; with small central cavity. Size about 6 to S/x. Antigonon species from Guatemala. (Polygonacece.) Dry seed. — Grains spherical, rarely oval, usually more or less polyhedral owing to pressure; some with small central eavit}^ and single short radial fissures. Size about 13 to llfx. Among them comjjound grains consisting of 3 to 5 part-grains disposed in rows. Pisonia aculeata Linn. (Nyctaginiacece.) Dry seed. — Grains spherical or polyhedral, frequently with one curved and several pressure facets; the larger with a central cavity, and at times with several short radial fissures. Size about 13/u. Lhidoubtedly simple grains. A few compound grains composed of 2 to more than 12 part-grains are present. Nepenthes destillatoria Linn. (Nepenthaceae.) Dry seed. — Grains spherical, or with rounded corners to almost polyhedral; with a large or small central cavity, and often with radial fissures. Size 10 to 13m. Acanthus mollis Linn. (Acanthacece.) Fresh and dry cotyledons. — Grains spherical to oval, not at all or slightly compressed; with delicate, indistinct lamellse; fresh, with central spherical hilum; dry, with marked radial fissures, which extend almost to the periphery, the fissures being equallj' marked in all aspects. Size about 60^. Among them are semicompound and compound grains with few equal or unequal size part-grains due to the splitting of the hilum as well as to the breaking away of corners. The starch-grains and cell-tissue of Acanthus, like the seeds, resemble tlie starchy seeds of the Papilionacew. Drosera longifolia Linn. (Droseraceas.) Dry seed. — Grains spherical or oval-spherical, rarely oval, many of them more or less polyhedral owing to pressure; with a central cavity and several marked radial fissures, some of which frequently are short, others extend to the periphery, and appear to split the simple grain into part-grains. Size about 18/j. In unripe seeds rounded grains only occur, while in the ripe seeds many polyhedral grains are seen along with them. Drosophyllum lusitanicum Spreng. (Droseracece.) Dry seed. — Grains spherical or nearly oval, but rarely oval, and many of them more or less polyhedral owing to pressure; with small central cavity and radial fissures. Size about 15 to 19^. Type 2. Grains Simple, Centric, Lenticular. Grains circular, rounded-oval, kidney-shaped, or triangular, or 3 to 4 angular with rounded corners and pres.sure facets; more or less polyhedral when crowded. Hilum in the center, usually shaped like the grain, but much thinner in comparison. Lamellae usually of equal thickness at points opposite to hilum, but thickest toward the margin of the grain. During the process of drying a fissure coinciding with the largest plane is nearly always formed, and from which radial fissures issue, which are usually invisible on the broad aspect but appear as 204 DIFFERENTIATION AND SPECIFICITY OF STARCHES. dark stripes on the narrow side. Tlicse grains are usually symmetrical on all sides, although the two sides of the margin sometimes develop unequally. This type passes over into the centric- spherical as well as into the centric-oval type. It is found with certainty only in spores and seeds. Although round grains which have been pressed together do occur in the underground parts of the plant, the absence of a hilum and of lamellae makes their structure doubtful. Compound grains are never or very rarely found along with the simple ones. (Edogonium landsboi-oughii Hass., Kutz. (Algcv.) Dry spores. — Grains rounded, triangular with rounded corners, or oval, frequently irregular; compressed to al)out one-half or more of the width, witli a longitudinal slit on the narrow aspect. Size about 12/j. Similar or somewhat smaller starch-grains sometimes form a partial or complete wall within the vegetative cells. Some grains approach the centric-oval (type 3). (Edogonium vesicatmn Vauch., Link. (Algce.) Dry spores. — Grains rounded or rounded-oval, fre- quently somewliat angular or irregular; the wider ones compressed to a httle more than one-half of their width. Size about Sfi. The more starch the less oil present. (Edogonium echinosperinum A. Braun. (Algce.) Dry spores. — Grains as in preceding. Size about 7/i. BulbochcBte sphcerocarpa A. Braun. (Algce.) Dry spores. — Grains circular to oval, frequently some- what angular or irregular; the larger ones compressed to about one-half or more of their width, with a long slit on the narrow aspect. Size about 10 to 13^. Approaching the centric-oval (type 3). The grains in the cell are almost equal in size and are arranged close together in a single layer along the primordial sheath, within which the lumen is filled with red oil, but very often it is not visil^le until part of the oil has been expressed. Similar grains, but some- what smaller in size, often form a partial or complete wall within the vegetative cell. Bulbochcete setigera Roth. (Algce.) Dry spores. — Grains rounded, mostly polygonal; the larger ones compressed to about one-half or more of their width; spindle-shaped when seen from the narrow aspect. Size about 11/i. Nitella. (Characece.) Spores. — Grains rounded, usually irregularly angular, the angles sometimes api^earing almost loliular; seldom regularly 4 to 6 sided or compressed to about one-half to one-third of their width; rarely with delicate lamellte, and a central hilum; dry, without fis- sures. The grains in a spore are, as a rule, of equal size, and arranged in a simple layer within the inner surface of the membrane, and so crowded that by pressure they become polygonal. Nitella syncarpa Kutz. Fresh. — Size about 70^. Nitella Jlexilis Ag. Dry. — Size 17 to 50-60;u. Nitella fasciculata A. Braun. Dry. — Size about 14 to 56/i. Nitella hyalina Kutz. Dry. — Size about 15 to 45/^. The fissures, which are seen from both aspects, are undoubtedly due to pressure. Nitella translucens Pers. Dry. — Size about 7 to 40//. Nitella batrachosperma A. Braun. Dry. — Size 10 to 45/j. Nitella tenuissima Kutz. Dry. — Size about 13 to 42ju. Nitella gracilis Ag. Dry. — Size about 10 to 40/i. Nitella exilis A. Braun.; A'^. flabellata Kutz. Dry. — Size about 15 to 45//. Chara. (Algce.) Spores. — These contain a wall of starch which consists of two kinds of grains — large, rounded and compressed grains, and small grains of indistinct structure (type 10). The larger grains are arranged with their broad side along the wall, and form a simple layer, while the smaller grains fill in the spaces between the larger ones. On crushing the spore the oil within comes out first, then the small starch-grains. Increased pressure will finally release and expel the larger grains. For this reason the latter very often are injured, and have a number of more or less marked fissures radiating from the center. In order to obtain them uninjured it is best to cut the spores. Chara factida A. Braun. Fresh. — Grains rounded, at times somewhat irregular; thickness two-thirds their width; lamellae numerous, usually 3 of them very distinct at regular intervals; hilum large, circular when seen on one side, and elliptical on the other. Size about 65/1. Chara hispida Linn. Dry. — Grains almost round, rounded-oval, often exactly circular, com- pressed to about one-half or more of the width; when seen from the narrow side oval, fre- quently with a longitudinal slit; lamella; numerous, very delicate; hilum circular, when seen from the broad aspect, but elongated from the narrow aspect. Size about 80 to lOO/x. TYPE 2. GRAINS SIMPLE, CENTRIC, LENTICULAR. 205 Chara baueri A. Braun. Dry. — Size of the grain about 65/Lt. Cham alopccuroidca Del. Dry. — Grains circular, or angular with round corners; compressed to about half their width. Size about 60/li. Chara barbata Meyen. Dry. — Grains circular; compressed to about half or more of their width; with numerous distinct lamellae. Size about 90/i. In some of the grains fungi threads were found which affected the starch-grains. Chara fragilis Desv. Dry. — Grains rounded or circular; usually with beautiful, delicate lamellae. Size about 98yu. Chara aspera Willd. Dry. — Size about 100^. Chara contraria A. Braim. Dry. — Size about 70(U. Chara gymnophylla A. Braun. Dry. — Size about 68/i. Chara crinita Wallr. Dry. — Size about 70^. Chara coronaia Ziz. Dry.- — Size about 66yu. Pilularia globulifera Linn. {Marsiliacem.) Dry gymnospores. — Grains rounded, rounded-oval, rarely somewhat irregular; compressed to a little more than one-half; seen in the narrow aspect, a longitudinal slit or an elongated hilum, around which lamellae can rarely be distinguished. Size about 24 to 50/i. The grains approach the centric-oval (type 3). Pilularia minuta Du Rieu. {Marsiliacece.) Dry gymnosjiorcs. — Grains rounded to almost oval, the Tvider ones compressed to about one-half or more of their width; sometimes with a longitud- inal slit seen in the narrow aspect; rarely with several indistinct lamella;. Size about 17 to 56^. The grains very closely approach tj-pe 3. Ephedra distachyia Linn. {Gnetacea.) Dry seed. — Grains circular, sometimes rounded-oval, kidney- shaped, or triangular \vith rounded angles; compressed to about two-thirds their width; a marked lengthwise slit on the narrow side; no lamellae. Size about 17/j. Ephedra alata Desne. {Gnetacea:.) Dry seed. — Grains circular to oval, often triangular with rounded angles, or oval pear-shaped; about five-sixths to as broad as long; two-thirds to three-fourths as thick as long; no lamella;; in the wide aspect sometimes seen with indistinct radial fissures; on the narrow side, with a long slit from the two ends of which single oblique fissures radiate, or vAth 2 curved fissures which are joined at their convex curvatures. Size about 27/u. Ephedra fragilis Desf. {Gnetacece.) Dry seed. — Grains as in the preceding, though somewhat larger, and frequently with distinct radial fissures in the broad aspect. Size 30^1. SHpagigantealjEiga.sc. (Graminacece.) Dry seed. — Grains circular to rounded-oval; compressed to one-half and two-fifth.s of their width; no lamellae; sometimes ynih a longitudinal slit in the narrow aspect. Size about 28/j. These grains seem to occur more frequently in the inner part of the large cells of seeds. On the whole the principal bulk of starch consists of small, broken separated-grains (type 16). Heteranthelium piliferum Hochst. {Graminacece.) Dry seed. — Grains circular and triangular with rounded angles to kidney-shaped and oval; lamellae very delicate, if any; the circular ones compressed to a little more than or over one-half their width, ^vith a longitudinal slit when viewed from the narrow aspect. Size 20 to 25/i. Among them numerous small, rounded grains, as in the Hordeacem. Hordeacece. Seed. — The inner tissue contains niunerous small grains along with large centric-lentic- ular ones; the outer cells are completely filled with the small grains (type 10). Triticum turgidum. {Graminacea;.) Fresh seed. — Grains circular, oval, or irregular; from three- fourths to just as broad as long; one-third to one-half as thick as broad; sometimes with distinct hilum; rarely with lamellae; lanceolate, elliptical, or plano-convex, when seen in the narrow aspect. Size about 24/i. At times the outer surface has a reticulate appearance caused by erosion of the grain. Triticum monococcum Linn. {Graminacece.) Dry seed. — Grains as in the preceding, but somewhat smaller. Size about 30ju. Triticum dicoccum Schrank.; T. amyleum Sering. {Graminacece.) Dry endosperm. — Grains as in the preceding. Size 27ju. Agropyru?ti rigidum R. and S.; A. cristatum R. and S.; Triticum rigidum Schrad; and T. cristatum Schreb. {Graminacece.) Dry seed. — Grains rounded or oval; compressed to about one-half to two-fifths their width; without distinct lamellae; sometimes with a longitudinal slit in the narrow aspect. Size 33m. 206 DIFFERENTIATION AND SPECIFICITY OF STARCHES. Secale cereale Linn. (Graminacece.) Dry seed. — Grains circular, oval or somewhat irregular; from three-fourths to as broad as long; one-half to one-third as thick as broad; rarely with single lamclliB, or with distinct hilum; elliptical or oval in the narrow aspect. Size iS/x. Some grains with reticulated outer surfaces. Elymus engelmanni Hort. (Graminacea:.) Dry seed. — Grains circular, rounded-oval or kidney- shaped; three-fourths to as broad as long; the larger ones three-fifths to three-fourths, those of medium-size two-fifths to one-half as thick as long; without lamellae; sometimes with a small oblong central cavity. Size 2Qn. Elymus hystrix Linn.; AspreUa hystrix Willd. (GraminacecB.) Dry seed. — Grains rounded, homo- geneous and solid; compressed. Size 22ii. The outer surface frequently shows reticulate or nodule-like markings. Hordeum vulgare Linn. {Graminacea;.) Dry seed. — Grains circular or oval; three-fourths to as broad as long; one-half to three-fifths as thick as broad; homogeneous and solid. Size 3.5/i. Hordeum bidbosum Linn. {Graminacece. ) Dry seed. — Grains circular, rounded-oval, triangular to quadrangular; one-half to three-fifths as thick as broad; without lamellae; rarely with dis- tinct, small hilum. Size 20 to 23/i. Hordeum himalayense Ritter; Critho wgiceras E. Meyer. {Graminacece.) Dry seed, endosperm. — Grains rounded, rarely kidney-shaped or oval, frequently irregular; compressed; no lamellae; some- times with a small, rarely a large, central cavity. Size 26 to 33yu. ^gilops caudata Linn. {Graminacea;.) Dry seed. — Grains circular, rarely oval, lenticular-shaped; compressed; homogeneous and solid. Size 30/u. JEgilops triuncialis Linn. {Graminacea;.) Dry seed. — Grains as in the preceding, sometimes with small central cavity, and with a single lamella. Size 28yu. Braconnolia elymoides Goodr. {Graminacea.) Dry seed. — Grains rounded or oval; compressed to a little more than one-half the width or over; homogeneous and solid. Size 28^. Lileae subulata Humb., Bonp. {Naiadacem.) Dry embryo. — Grains rounded-oval, or oval, and almost terete; two-thirds to almost as broad as long; oval or elliptical when seen in the narrow aspect, and with a distinct longitudinal slit. Length 13^, width 11/i, and thickness 7 and S/i. The grain approaches the centric-oval type (type 3). Much protoplasm and some oil may be found with the starch. Triglochin barrelieri Lois. {Naiadacem.) Dry embryo. — Grains usually oval, the wider ones compressed ; homogeneous. Size about Q/x. The grains approach the centric-oval type (type 3). Little starch, much oil. Scheuchzeria palustris. {Naiadacece.) Dry embryo. — Grains circular or oval, sometimes irregular; compressed to about one-half to one-third their width; homogeneous. Grains 9;u. Starch, oil, and protoplasm. Alisma ranuncidoides Linn. {Alis7naceoe.) Dry embryo. — Grains rounded-oval, or oval three-fifths to almost as broad as long; two-fifths to three-fifths as thick as long; without lamellip; sometimes with a distinct hilum; when seen from the narrow side, slightly contracted in the middle, rarely with a lengthwise slit. Size 20ju. The grains approach type 3. Some oil with the starch. Sagittaria ,sagitta;folia Linn. {Alismacea; .) Dry embryo. — Grains rounded, or triangular with rounded angles, or oval, and oval pear-shaped, the wider ones compressed at about one-half their width; many with a longitudinal slit in the narrow aspect. Size 25yu. Among them some doublets. Little oil with the starch. Actinocarpus damasonium Smith. {Alismacece.) Dry embryo. — Grains rounded or oval, occasionally kidney-shaped; two-thirds to almost as broad as long; the wider ones compressed at two- thirds or over ; rarely with few lamella? and wth distinct hilum ; a few very delicate fissures are sometimes visible in the broad aspect, and an indistinct slit can frequently be seen in the narrow aspect. Length 20 to 24,u, and width 16yu. The grains closely approach type 3. Luzula and Juncus. {Juncacece.) Seed. — Dry and ripe seeds only were examined. Their starch- grains are undoubtedly simple, they can be seen within the cell equally distributed through- out the lumen, and there are no indications of compound ones. The grains are compressed and occasionally a central cavity can be distinguished. Therefore they most likely belong to the centric-lenticular type. The surface is more or less polyhedral, and frequently very irregu- lar. This is no doubt due to the shrinking which the substance, being very soft in a fresh state, undergoes in drying, and also to the flattening caused by the crowding of the grains. TYPE 3. GRAINS SIMPLE, CENTRIC, OVAL. 207 Luzula nivea Desv. Dry seed. — Grains rounded, more or less angular, frequently with sharp angles; homogeneous; the larger ones compressed at about one-third their width. Size 11 to 13/li. Luzula forsteri Desv. and L. muUifiora Lejeum. Dry seed. — Grains rounded, occasionally somewhat angular, compressed. Size 20^. Thickness sometimes only 1.5 and 2^. Juncus baltieus Dethard. Dry seed. — Grains rounded, many of them somewhat angular, with blunt, rarel}' sharp, angles; somewhat compressed, extending halfway; homogeneous or granular; sometimes with small central cavity. Size 9 to 12/j. Juncus glaucus Ehrli. ; J. effusus Linn. Dry seed. — Grains rounded, quadrangular with rounded corners, or polyhedral with sharp angles, usually irregular; compressed at about one-third; homogeneous or granular; sometimes hollow. Size 9 to 12/x. Juncus bulbosus Linn.; J. comprcssus Jacq. Dry seed. — Grains rounded or roundish-angular, or often irregular with blunt or sharp corners; compressed; homogeneous or granular. Size IB^i. Juncus acutijlorus Ehrh. ; J. sylvaticus Reichard, var. inacrocephalus. Dry seed. — Grains rounded, frequently more or less angular; compressed, sometimes to about one-half the mdth; homo- geneous or granular. Size about ISju. Salsola soda Linn. {Chenopodiacew.) Dry cotyledons. — Grains rounded to oval, usually angular and irregular; the broad ones laterally compressed at about one-half or over; homogeneous; with a long slit when seen from the narrow side. Size 11 to 13^. The grains approach type 3. Among them some compound grains of two or many part-grains. Seeds obtained from various places were examined. Those whose starch-grains have just been described were green, as were also the tissues of the cotyledons. Some of the starch-grains were also green, proving that they were formed in the chlorophyl. These seeds were rich in starch, and also contained much oil. On the other hand, in seeds of a bro^ii color with colorless tissues in the cotyledons, considerably more oil and remarkedly little starch was found. The latter are a little smaller (up to 8m), rounded oval, or angular with rounded corners, and are compressed. The seed of other species of Salsola soda examined contained no starch, possibly oil replacing the starch in ripe Salsola soda. Trapa natatis Linn. {Onagonocew .) Dry cotyledons. — Grains circular, rounded-oval to triangular to quadrangular; one-third to one-half as thick as broad; usually with indistinct and scant lamelliE; frequentlj' with a small central cavity from which delicate fissures radiate; with a marked slit seen in the narrow aspect. Size to 32/x. Type 3. Grains Simple, Centric, Oval. Grains oval or oblong, with equal blunt ends, rarely with unequal or pointed ends; seen on end, circular or compressed; more or less angular when crowded. Hilum in the center, and in a general way of the same shape as the grain, but relatively much longer and more compressed. Lamel- lae, as a rule, alike at two opposite parts, most marked at the ends. The true terete grains, when dried, show fissures which radiate in all directions from the longitudinal axis; the compressed forms, almost without exception, show a fissure which coincides with the largest plane, which fissure when seen from the narrow side appears as a dark median line, and from the broad aspect has a very indistinct or almost invisible border. The fissure is sometimes crossed at right angles by another, but more frequently onlj' cross-fissures radiate from it. The grains are usually symmetrical on all sides, but the compressed forms show a decided inclination, which is more marked along one margin than along the opposite, and they therefore become plano-convex, kidney-shaped, elliptical, or triangular with rounded angles. This type, going over on one side to the centric- spherical and on the other to the centric-lenticular, is found with certainty only in spores and seeds, and seems to be entirely wanting in other parts of the plant. Semicompound and compound grains are occasionallj' found among the simple ones, and they usually consist of few, and, almost without exception, part-grains of equal size. Chara aspera Willd. {Algoe.) Dry bulbils of the lowest nodes of the stem. — The bulbils consist of a single cell, with a structureless membrane, and filled with two kinds of starch-grains : (1) Large grains, spherical or oval; two-thirds to as broad as long; the widest ones compressed to three-foiu-th.s of the grain; usually without distinct lamella; with central, spherical, or oblong hilum; usually with marked fissures radiating from the center; the compressed forms have a slit coinciding with the plane. Length 40 to 100^, width lOfx, thickness to bOfi. 208 DIFFERENTIATION AND SPECIFICITY OF STARCHES. (2) Small grains rounded, more often oblong or oval; usually distinctly convex at one end and less so or somewhat concave at the other; occasionally reniform or triangular; the wide forms are slightly compressed. Size 4 to SOfi. The classification of these grains is somewhat uncertain. The lamellae of the spherical grains arc mathematically centric. The structure of the large oval grains agrees with that of the Legimiinoscc. The smaller grains resemble the largest in form at least, but the almost central hilum seems to be more spherical. Perhaps the larger and smaller grains of Leguminosoe also differ as in the case with the seeds of Chara. Among the grains described doublets occur which are composed of one large and one small component. Marsilea ■puhescens Ten. (Marsiliacece.) Dry gymnospores. — Grains oval, or rounded triangular to oblong, and pear-shaped, at times somewhat irregular; the \vider ones compressed to about one-half; hilum oblong, lamellse distinct, frequently with a longitudinal slit that is seen distinctly in the narrow side, and from which cross-fissures frequently radiate. Length 30 to llOyii, width 60yu, thickness 35,u. Among them some compound grains (type 11). Marsilea -puhescens Ten. (Marsiliacece.) Dry atidrospores. — Grains elliptical, rounded-triangular to oblong, and pear-shaped; two-thirds to as broad as long, to 4 times as long as thick, the broader ones compressed to about one-third; sometimes with a slit seen in the narrow aspect. Length lOyu, breadth 7(1. Globularia piluUfera Linn. {Globularisacew.) Dry androspores. — Grains rounded, or rounded-tri- angular to oval, or pear-shaped; compressed to about one-half or more; a longitudinal slit in the narrow aspect. Size 7ai. Butomus umbellatus Linn. (Alismaceoe.) Dry embryo. — Grains rounded, reniform to oval, and pear- shaped; the wider ones compressed to about one-half; a longitudinal slit on the narrow aspect. Length ll/u, width 10/u. The grains closely apjiroach the centric-lenticular type (type 2). Limnocharis plumicri Rich. [AUsmoceoi.) Dry embryo. — Grains circular to oval, occasionally some- what irregular; compressed to about one-half the width, without lamellse; frequently with distinct hilum; on the broad aspect, usually with 2 to 7 marked fissures radiating from the center to the margin; a length^\^se slit is often seen on the narrow side. Length 21//, width 17;u, thickness 8/u. The grains approach type 2. Philydrum lanuginosum Gaert. {Philydacece.) Dry seed. — Grains elliptical or ovoid, sometimes almost reniform, frequently somewhat irregular; one-half to equally as broad as long; some- times slightly compressed; no lamellse; usually with an indistinct slit which in the compressed grains coincides \vith the greatest plane, and from which indistinct cross-fissures sometimes radiate. Length 22/li, breadth 14/i, thickness 12/z. Doublets and trijilets are often foimd. Reussia triflora Endl. [Pontederiacca..) Dry seed. — Grains oval to lanceolate, very often more or less curved, usually elliptical or spindle-shaped, sometimes drawn into sharp points, and more or less regular; 1.5 to 4 and 6 times as long as broad; the broad ones compressed, the smaller ones terete; no lamellse; an indistinct longitudinal slit is seen in the narrow, and at times also in the broad aspect. Length 30 to 45/.1. A few compound grains have small number of part-grains. Heteranthera limosa Vahl. (Pontederiacece.) Dry seed. — Grains elliptical to oblong, usually some- what curved or reniform, sometimes triangular, and frequently irregular with protruding angles or humps; occasionally pointed at one or at both ends; 1.33 to 3 times as long as broad; the broader ones compressed; with a central cavity or fissure. Length 16 to 22/i, breadth IOm, thickness 8m. The grains resemble those of Reussia, but are shorter, more compact, and more torulose. Ponlederia sp. (Pontederiacew.) Dry seed. — Grains as above, but shorter, less pointed, and more polyhedral. Eichhornia tricolor M. Seubert. {Pontederiacece.) Dry seed. — Grains as in Ponlederia. Stratiotes aloidcs Linn. {Hydrocharitacea:.) Dry embryo. — Grains elliptical, oblong, rarely curved, or almost reniform; 1.5 to 3 times as long as thick; the majority compressed; usually with a longitudinal slit seen in the narrow aspect. Length about 13/x, thickness about 8^. Damasonium indicum Willd.; Ottelia alismoides Pers. {Hydrocharitacea;.) Dry embryo. — Grains oval or elliptical, frequently somewhat kidney-shaped; one-half to almost equally as broad as long, the broad ones laterally compressed to about one-half, the narrow ones slightly or not at all compressed; no lamelke; frequently with a longitudinal slit on the narrow aspect. Length about 20 to 27m, breadth 18^. The grains approach the centric-lenticular form (type 2). TYPE 3. GRAINS SIMPLE, CENTRIC, OVAL. 209 Hydrocharis morstis-rance Linn. (Hydrocharitacew.) Dry embryo. — Grains rounded to oval. Length lOfi. Approacliing contric-lcnticular form (tj^pe 2). Lachnanthcs tindoria Ell. (Ha-modorncece.) Dry seed. — Grains elliptical to lanceolate, frequently slightly curved, sometimes triangular or reniform, usually more or less irregular and toru- lose; one-fifth to almost as broad as long; the broad ones slightly compressed; no lamellae; with frequently a very large cavity with a longitudinal slit. Length 20 to 28ju, breadth 11 fi. It is doubtful whether these grains belong to this tjTDC. Naias major Roth.; N. marina Linn. {Naiadacea;.) Dry embryo. — Grains oval, three-fifths to four- fifths as broad as long; the broad ones compressed; without lamellse; a loiagitudinal slit from which several fissures usually radiate may be seen from the narrow aspect, and sometimes also from the broad aspect. Length 30/u. Zostera marina Linn. (Naiadacea:.) Dry embryo. — Grains pressed into true polyhedrons having sharp edges and angles, filling the cells like a parenchyma; one-half to as thick as long; no lamellse; with fissures, which in the isodiametric grains usually radiate from a central cavity and in the oblong ones proceed from a longitudinal slit at right angles to it. Length 30 to 35^. The grains approach centric-spherical (type 1). The grains in the outer parts are of less size, those in the outermost being only 4 to 6/u. They are rounded or oval, and more or less angular. These grains show very distinctly that they belong to the simple type. Ruppia maritima Linn. (Naiadacea:.) Dry embryo. — Grains rounded to rounded-oval, frequently somewhat angular; wthout lamellse; with a rounded, oblong, or irregular cavity, and usually with irregular radial fissures. Length 25^. Seems closely to approach the centric- spherical type (tjT^e 1). Compound grains of few and usually irregular part-grains are also found (type 15). Zajinichellia pedicellata Fries. (Naiadacea.) Dry embryo. — Grains rounded to oval; half to as broad as long; the broader ones compressed to about one-half and one-third their width; without lamellse; longitudinal slit marked in the narrow side, also sometimes shows indistinctly in the broad aspect. Length 18/j, breadth 15/:. Althenia filiformis Petit. (Naiadacew.) Dry embryo. — Grains elliptical to oval, sometimes reni- form; half to as broad as long; the broader ones laterally compressed to about half or more of their width; lamellse indistinct; longitudinal slit is seen in the narrow aspect. Length 18;u, breadth 15/u. Potamogeton juitaiis Linn. (Naiadacew.) Dry embryo. — Grains oval or elliptical; three-fifths to three-fourths as broad as long; the broader ones compressed to about half their width; the narrower slightly compressed; no lamellse; with a longitudinal slit on the narrow side. Length 30m, breadth 38^. Potamogeton prcelongus. (Naiadacece.) Dry embryo. — Grains elliptical or oval; compressed; no lamellse; sometimes with a longitudinal slit on the narrow side. Length 15^, breadth 15/^. Calla palustris Lima. (Aroidea:.) Dry seed. — Grains spherical or oval, frequently irregular; some- times slightly compressed; half to as thick as long; with a small central cavity, which in the spherical form is ahnost round, in the oval forms oblong. The grains approach the centric- spherical (type 1). Anthurium acaide Sweet.; Pathos acaulis Linn. (Aroidece.) Dry seed.— Grains spherical or oval, sometimes slightly angular, due to pressure; three-fifths to as broad as long; frequently with a small central cavity. Length 8 to 10m. The grains approach the centric-spherical (type 1). Ceratophyllum svbmermm Linn. (Ceratophyllacem.) Dry cotyledons. — Grams rounded to oval; two- thirds to as broad as long; slightly compressed; no lamellse; a longitudinal slit is seen on the narrow side, and a delicate one is sometimes visible in the wide aspect; also simple radial fissures. Length 23m, ■«idth 20m. The grains approach the centric-spherical (type 1). Nelumbiiim speciosum Willd. (Nymphacece.) Dry cotyledons.— Grmns oval or elliptical, sometimes one margin more curved than the others, or even plano-convex, many of them with 3 or 4 angles; the -n-ide ones slightly compressed; no lamellse, or single and indistinct; a marked slit is seen on the narrow side, less marked on the broad aspect. Length 20m. The grains de- crease in size toward the smface of the cotyledons, lumule contains oil and starch. The grains small, rounded, somewhat irregular. Size 5 to 6m. (Syn. Nelumho nucifera Gsertn.) 14 210 DIFFERENTIATION AND SPECIFICITY OF STARCHES. Nchnnbium lulcuin Willd. {Nchivibo lutca Pits.) Dry seed. — Grains as in the preceding, tliough on the whole somewhat broader. Ijength 20(U. Cistus vulgaris Spach. var. C. crclicus Linn. (Cislnccw.) Dry seed. — Grains oval, elliptical, oblong, blunt-triangular, and at times almost rcniform; two-thirds to almost as broad as long; the broader ones compressed; no lamellse; usually a longitudinal slit is seen in the narrow aspect. Length 21/i. Helianthemum wgyptic^im Mill. (Cistacew.) Dry seed. — Grains oval, very often reniform or roimded-triangular; one-half to two-thirds or seldom almost as laroad as long; slightly compressed; no lamellse; with an oblong cavity, or a longitudinal slit on the narrow side. Length 18ju. Lechia thymifolia Mich.x. (Cistacece.) Dry seed. — Grains rounded-oval to elliptical, often broadly triangular or almost reniform, two-fifths to almost as broad as long; the broader ones slightly compressed; no lamellie; usually with a longitudinal slit which appears narrower, but more marked when seen from the narrow side. Length 12fi, rarelj' 20/x. Bixa orellana Linn. (Bixaceoe.) Dry unripe seeds. — Grains rounded-oval to elliptical, sometimes blunt-triangular and almost reniform; the long ones two-fifths to one-half as broad as long, and the shorter ones as broad as long; the broader ones compressed about two-thirds of their width; few delicate or no lamella; with a long slit-like cavity which is more or less distinct in both aspects ; or with a narrow slit and short transverse fissures seen only from the narrow side. Length 30 and 3LV. Mangifera species (probably M. indica) Linn. (Avacardiacew.) Dry cotyledons. — Grains oval to oblong, sometimes blunt-triangular, rarely almost rcniform; two-fifths to as broad as long; the broad ones compressed; no lamella; often with a slit-like cavity which is usually more distinct in the narrow aspect, and from which frequently very delicate cross-fissures radiate. Length 30^- Some doublets among them. Anacardium occidentale Linn. {Anacardiacce.) Dry cotyledons. — Grains spherical or ellipsoidal, often on one margin curved or even plano-convex; two-thirds to as tjroad as long; the widest ones sHghtly compressed; no lamellae; the spherical grains have a small central cavity, and the oval ones an indistinct longitudinal slit. Length lOyu. Among them compound grains of 2 to 3 part-grains. Lotus edulis Linn.; Krokeria edulis. (Leguminosce.) Dry cotyledons. — Grains spherical-oval or rcni- form; three-fifths to almost as broad as long; the broad ones compressed, the narrower ones sHghtly compressed, few delicate, sometimes indistinct lamellae; a delicate longitudinal slit with very indistinct radial fissures in the broad aspect; the slit as well as the radial fissures are more marked in the narrow aspect. Length 45;u. Caragana altagana Poir. {Leguminosce.) Dry cotyledons. — Grains rounded or oval; three-fourths to equally as broad as long; slightly compressed; occasionally with a longitudinal slit on the narrow side. Length 8;u. Caragana arboreseens Lam. (Leguminosce.) Dry cotyledons. — Grains rounded or oval; three-fifths to almost as broad as long; slightly compressed; with indistinct cavity seen on the broad aspect; on the narrow side a distinct longitudinal slit. Size 8 to lOyu. Cicer arietinum Linn. {Leguminosce.) Fresh and dry cotyledons. — Grains almost spherical to ellipti- cal; twice as long as thick; with few, very indistinct lamella and an oval or oblong hilum; dry, with a longitudinal slit and delicate radial fissures. Length 30/i, thickness 23yu. Among them compound grains of few equal-sized part-grains (see type 14). Pisum sativum Linn. {Leguminosce.) Dry cotyledons. — Grains oval to almost reniform, and blunt- triangular; half to almost as broad as long; the l)road ones compressed. Lamelke distinct, usually 5 to 7; an indistinct cavity with delicate radial fissures is seen on the broad side; a marked longitudinal slit on the narrow side. Length 65/j. A few doublets are found among them. Ervum lens Linn. {Leguminosoi.) Fresh and dry cotyledons. — Grains oval to triangular with rounded corners; half to almost as thick as long; with 2 to 3 lamella;; and ovate or oblong hilum; dry, with delicate radial fissures in the broad aspect, and with a distinct slit in the narrow asjiect, but rarely with cross-fissures. Length 40/u, width 30^. According to Payen, the length is 67m- Doublets with equal parts are often found among them (type 14). TYPE 3. GRAINS SIMPLE, CENTRIC, OVAL. 211 Ervum agrigentinuni Cups. {Leguminosce.) Dry cotyledons.— Grains roundod-oval to oblong, occa- sionally-conical or reiiiforin, frequently irregular and olitusc-angled; half to as thick as long; the wider ones slightly comjiressed ; lainelkc quite distinct (2 to 5); a longitudinal slit with numerous radial fissures. Length 50ju, breadth 35^. Among them some doublets. Vicia caJcarata Desf. (Leguminosce.) Dry cotyledons. — Grains rounded to oblong, occasionally reniform or triangular, fr(>(]uently irregular; three-fifths to as long as broad, the wider ones compressed to about half; lamellte indistinct, if any, sometimes more distinct in the narrow aspect; lengthwise slit hariUy visible in the broad aspect, but distinct and sometimes with several cross-fissures on th(! narrow aspect. Length 35 and 45(U. V. saliva Linn. Dry cotyledons. — Grains as in the preceding, one-half to equally as broad as long. Length 32^. Faba vulgaris Still.; Vicia faba Linn. {Leguminosm.) Dry cotyledons. — Grains rounded-oval to oval and reniform; half to as broad as long; compressed to about one-half or more; occasionally 1 or 2 indistinct lamella;; fine, irregular, especially radial fissures are seen in the broad aspect; distinct longitudinal slit from which fissures sometimes radiate is seen on the narrow side. Length 45 to 50/z; width 34/i. According to Payen, the grains attain to Ibii. Lathyrus sativum Linn. {Leguminosce.) Dry cotyledons. — Grains rounded, oval, reniform, frequently irregular; half to as long as broad; the broader ones compressed to about two-fifths; some- times ■with delicate lamelke, longitudinal slit and fissures, indistinct or fine from the broad aspect, well-marked in the narrow side. Length 56/li, width 34ju. Lathyrus nissolia Limi. {Leguminosce.) Dry cotyledons. — Grains rounded to oval, frequently reni- form or obtuse blunt-triangular or somewhat irregular; half to as broad as long; the laroad ones compressed to about two-thirds; lamella? rare, very fine; longitudinal slit marked in the narrow aspect; in the broad view fine or indistinct radial fissures. Length 35/j, breadth 26p, thickness 22^. Lathyrus aphaca Linn. {Leguminosce.) Dry cotyledons. — Grains rounded to oval, frequently angular with rounded corners, or somewhat reniform; three-fifths to as broad as long; the broader ones compressed to about half their width; lamellte very fine, or rare; longitudinal slit well marked in the narrow aspect, with almost invisible radial fissures on the broad aspect. Length 37yii. Orobus niger Linn. {Leguminosce.) Dry cotyledons. — Grains rounded to oval, frequently somewhat angular; three-fifths to as Itroad as long; the wider ones compressed to about half their width; lamelliB rare, or indistinct; longitudinal slit marked in the narrow aspect, but indis- tinct from the broad aspect. Length about 29yu. Orobus lathyroides Linn. Dry cotyledons. — Grains as above, frequently rounded-triangular, or reniform. Length about 27^. Onobrychis caput-galli Lam. {Leguminosce.) Dry cotyledons. — Grains roianled or oval, occasionally somewhat angular; tliree-fourths to as broad as long; slightly compressed; no lamella;; with distinct longitudinal slit seen in the narrow aspect. Length L5yu. Onobrychis sativa Lam. {Leguminosce.) Dry cotyledon s.^GraHns rounded or ovoid; three-fourths to just as broad as long ; slightly compressed ; longitudinal slit on the narrow side. Length about 1 l/i. Catiavalia obtusifolia DC. {Leguminosce.) Dry cotyledons. — Grains rounded to elliptical, occasionally triangular or reniform, or somewhat irregular; half to as broad as long; the wider ones com- pressed to about one-half, and the narrow ones very slightly compressed; lamella? indistinct or rare; longitudinal slit and radial fissures fine on the broad side, marked on the narrow side. Length about 30 to 35/.t, breadth 24/i. Phaseolus multijlorus Lam. {Legummosce.) Dry cotyledons. — Grains rounded to oval, often tri- angular; three-fifths to as broad as long; the broad ones slightly compressed; no lamella;; occasionally with radial fissures; longitudinal slit distinct on the narrow aspect, indistinct from the wide aspect, rarely with a transverse slit. Length about 35//. Phaseolus vulgaris Linn. Dry cotyledons. — Grains as above; usually with marked fissures on the broad side and a distinct slit on the narrow side. Size 40^. According to Payen, 63yn. Phaseolus aureus Hamilt. Dry cotyledons. — Grains as in Ph. multijlorus, but somewhat larger; sometimes with indistinct lamelliE; fissures less marked. Size 55//. Phaseolus saponaceus. Dry cotijledons. — Grains as in Phaseolus multijlorus, but larger; 3 to 6 lamellae within the margin usually \nsible, very delicate, and crowded; longitudinal slit and fissures marked. Length 50 to 56//. 212 DIFFERENTIATION AND SPECIFICITY OF STARCHES. Vigiia glabi-a Savi. (Lcgumivosa\) Dry cotyledotis. — Grains rounded to elliptical, sometimes almost reniform; half to as broad as long; the broad ones slightly compressed; nxmierous fine lamel- la^; longitudinal slit moderately fine on the broad aspect and with radial fissures; slit mostly without fissures and more marked on the narrow side. Length about 40 to 48aj. Dolichos monachalis Brot. (Leguminosm.) Dry cotyledons. — Grains oval, almost reniform, or rounded- triangular; half to almost as broad as long; the wide ones slightly compressed; no lamellae; with longitudinal slit, and frequently with radial fissures. Length about 27^, width about 20ju. Doublets are occasionally found. Lablab vidgaris Savi. (Legwninosw.) Dry cotyledons. — Grains rounded or oblong, often reniform; half to as broad as long; the wider ones compressed to about one-half; no lamella; a longi- tudinal and sometimes a transverse slit, besides numerous fissures running irregularly when seen on the broad aspect; in the narrow aspect the fissures usually radiate from the two ends of the longitudinal slit. Length 25 to 29//. Drepanocarpus lunatns Meyer. {Legujninosw.) Dry cotyledons. — Grains rounded to oval; often tri- angular and somewhat reniform or irregular; tlirce-fifths to as broad as long; the broad ones slightly compressed; without lamellie; occasionally numerous irregular fissures, more often a longitudinal slit from which few or many fissures radiate. Length about 45/^. Type 4. Grains Simple, Centric, Spindle-shaped. Grains linear or lanceolate, either narrowed toward the ends (spindle-shaped) or with ends of equal width, or even somewhat broadened and blimt (rod-shaped) ; almost circular in cross-section. Hilum and lamella! invisible. This type occurs only in the latex of native Euphorbiacece, and prob- ably is but a transition to type 5. Euphorbia lathyris Linn. (Euphorbiacece.) Fresh latex from the stem.. — Grains spindle-shaped or rod-shaped, often somewhat irregular; circular in cross-section; 4 to 8 times as long as wide; frequently broadened at the ends, as if swollen; no lamellee; with a long depression in the linear axis, and with numerous fine, short, transverse fissures. Length 55m, width lOyu. The fissures are visible if the starch-grain be examined in the unchanged latex or in water. Euphorbia palustris Linn. (Euphorbiacece.) Fresh latex from the stem. — Grains rod-shaped, or cylin- drical spindle-shaped, thickened toward the middle; the ends are frequently rounded into small heads. Length 17 to 35/!, width 3 to Gju. Euphorbia virgata W.K. (Euphorbiacea;.) Fresh latex. — Grains as in the preceding, but not much thickened toward the middle. Length 32ju, width 5yu. Euphorbia dulcis Linn. (Euphorbiacece.) Fresh latex from the stem. — Grains rod-shaped, ends rounded, 6 to 10 times as long as wide. Length about 17 to 28;u, width 2.05 to 3.05/^. Euphorbia procera Biebrst. (Euphorbiacece.) Grains as in the preceding. Length 32yu, width 4^. Euphorbia epithymoides Linn.; Euphorbia fragifera Jan. — Grains as in the above, 7 to 12 times as long as thick. Length about 42yu, thickness 3 to 5^. Euphorbia cyparissias Linn. (Euphorbiacece.) Fresh latex from the stems. — Grains rod-shai^ed, or cylindrical spindle-shaped, ends round or blunt, usually broadened, rarely narrowed; 6 to 10 times longer than thick. Euphorbia nicceensis AIL; Euphorbia glareosa Biebrst. (Euphorbiacece.) Fresh latex from the stem. — Grains rod-shaped or cylindrical spindle-shaped; usually with blunt, equally broad, capitate ends, so that the whole grain appears almost bone-shaped; slightly terete or compressed at the widened portion. Length 28fi, width at some places 11^. Type 5. Grains Simple, Centric, Bone-Shaped. Grains elongated, compressed; linear spindle-shaped in the narrow aspect; at first spindle- shaped in the broad aspect, afterwards moderately broad toward the middle, and with very much broadened spatulate, sometimes lobate, split ends. Hilum invisible; lamella; usually indistinct. This type is found only in the latex of arborescent Euphorbiacece. The middle part may be almost wanting in the short grains; and the \vidth of the ends may even exceed the entire length of the grain. In the longer grains lateral lobes along the plane of the widest growth may also appear. The two ends sometimes develop unequally. TYPE (). GRAINS SIMPLE, ECCENTRIC, INVERTED, CONE-SHAPED. 213 Euphorbia ncreijolia Liim. (Euphorbmccae.) Fresh latex. — Grains in tiie broad aspect bone-shaped, witli roinulcd, hroadciu'd, and usually lobate ends; two-nintiis to one-third as broad as long about the uiidtlle, and half as broad as long at the ends. In the narrow aspect, rod-shaped or narrow-spindle-shaped, frequently curved; one-fifth to one-seventh as thick as long; usually without lamellae; with a channel-like cavity along the median line from which short transverse fissures radiate. Length about 50;u. Some forms with lobate processes in the middle of the grain may also occur, in which case they are two-thirds as broad as long. The fissures are already present in the grain in the latex and do not alter in water. Type 6. Grains Simple, Eccentric, Inverted, Cone-Shaped. Lamellae more numerous and coarser on one side, and fewer and finer on the diametrically opposite side. Grains more or less conical, almost circular on cross-section. The eccentric hilum is toward the slender end. The lamellae, which completely encircle the grain, are interspersed at the thickened side with incomplete lamella, but rarely in great numbers. On drying, fissures radiating from the central lamelliE are formed which are usually turned toward the distal end, and have a funnel-like arrangement. This type usually has no well-marked characteristics, and through the grain with equal ends easily merges into tj-pes 7 and 9; and again through these grains to those which are slightly compressed at the distal end into type 8. This type (6) may possibly occur more often than would appear from the specimens described, of which the potato is the best example. At a first glance some grains seem to belong to this type, while in reality they are flattened grains belonging to type 8. Saururus cernuus Linn. {Piperacece.) Root-stock. — According to Leon Soubeiran (Journ. Pharm., 1854, XXV, 100) the grains are rounded or oval; size 10 to 40/i; the majority with distinct centric lamellae and a hilum. The drawing by Soubeiran represents the hilum as being very eccentric, and situated at the narrower end, so that the grains apparently belong to this type. Ipoma:a purga Schlecht. Jalap. {Convolvulacece.) Dry tubers. — Grains spherical or egg-shaped; three-fourths to as long as broad; 4 to 7 lamellae, rarely distinct; frequently with single radial fissures in the interior; eccentricity of hilum usually one-fourth to one-seventh. Size about 35ju. According to Leon Soubeiran (Journ. Pharm., 1854, xxv), the starch-grains in the root of Batatas jalappa Choisy are 30 to 70 and SO/x in size; many are somewhat round or oval; some indistinctly triangular; others elliptical, but at the distal end cut off at right angles to the longitudinal axis; with distinct lamelte and eccentric hilum; in place of the hilum radiat- ing fissures are frequently found. Nageli queries: Can this be understood as a true jalap? Solarium tuberosum Linn. (Solanacew.) Fresh and dry tubers. — GraiiLS more or less egg-shaped, sometimes irregular, usually two-thirds to four-fifths as wide as long; lamellte chstinct, the majority of them complete; hilum, as a rule, at the narrow end, occasionally at the broad end and one-fourth to one-sixth eccentric; dry starch — usually with a small cavity in the central lamellae, and occasionally a few very short fissures, usually in the direction of the longitudinal axis. Length 70 to 90//. According to Payen, the grains in the large Rohan potato attain a length of 185^, and in some other kinds of potatoes 140/t. Among them are semi-compound grains comprising 2, rarely 4, part-grains, and also compound grains of 2 to 3 part-gi-ains. NiphcRa oblonga Lindl.; Achimenes alba Hort. (Gesneracece.) Dry scales from the root-stocks. — Grains oval to lanceolate; in transverse section usually almost circular; with numerous, delicate lamellae, of which only a few of the innermost are complete; hilum at the smaller end, one- fifth to one-twelfth eccentric; the end opposite the hilum is occasionally broadened into a somewhat compressed knife-hke edge. Length 30^. Dicentra formosa Walp; Diclytra formosa DC. {Fumariacece.) Dry root-stocks. — Grains oval, pear- shaped, 3 to 4 angles; usually more or less irregular; half to as broatl as long; the broader ones slightly compres.sed; lamellae very fine or invisible; hilum at the thin end, about one- eighth eccentric; in place of the hilum there is usually a small cavity from which several fissures radiate, forming a funnel-like groove. Length about 2()/(, breadth 20ju. Some grains are obtusely triangular, inasmuch as one angle is formed by an outer system of lamellae. Length 26/u, breadth 20/j. 214 DIFFERENTIATION AND SPECIFICITY OF STARCHES. Oreodaphne exaUala Nees; Laurus exallalus Sieb. PI. martin. 100. (Lauraccce.) Dry cotyledons. — Grains rounded-oval to elliptical, frequently egg-shaped, frequently somewhat irregular or unsymmetrical; half to as broad as long, the broader ones slightly compressed; no lamellae; hilum about one-seventh eccentric, usually at the narrow end, rarely at the broad end, and occasionally with a longitudinal slit; sometimes with several funnel-shaped fissures radiating from the center of the lamellae. Length about 21^. The two ends of the grain are very often alike in width. Laurus nobilis Linn. (Lauracece.) Dry cotyledons. — Grains rounded, egg-shaped, frequently some- what irregular, or angular owing to pressure; no lamellae; hilum slightly eccentric, frequently at the narrow end; in place of the hilum there is usually a cavity with more or less numerous short radial fissures, frequently with one marked transverse fissure which divides the grain almost in half. Length 23fi, thickness 19//. Among them some compound grains of few, equal components. (See type 14.) Mucuna -pruriens DC. {Leguininosce.) Dry cotyledons. — Grains oval or elliptical, usually more or less irregular, the smaller ones almost round; usually half, the longest ones only two-fifths, as broad as long; generally circular in transverse section; the broader ones slightly compressed; no lamellte; in the place of the one-seventh eccentric hilum, a small cavity is found from which delicate fissures pass out in the form of a funnel-like groove. Length 17 to 21/x. The center of these is found at the tapering end; the posterior end being more or less broadened and compressed, sometimes appearing to be cut off, leaving a doubt whether this is a transi- tion type going over into type 8 or whether it is a separated-grain. Mucuna wens DC. {Leguminosm.) Dry cotyledons. — Grains egg-shaped to rounded-oval, one-half to almost as broad as long, circular in cross-section; no lamellae; at the narrow end a small cavity instead of the one-half eccentric hilum, from this cavity two fissures diverge toward the distal end; and also usually one or more distinct transverse fissures, and sometimes also a few short ones, extend towards the hilum end. Length about 16 to 28/i. Among them some doublets, triplets, and some separated-grains. Type 7. Grains Simple, Eccentric, Cone-Shaped. Lamellae heaviest and most numerous on one side, and thinnest and least numerous on the diametrically opposite side. Grains more or less conical, almost circular in cross-section. The eccentric hilum toward the broad end. Only a few of the innermost lamelke, as a rule, are complete; the outer ones at the thickened radius seem to be incomplete. On drying, radial fissures are formed which diverge from the center of the lamellae, chiefly toward the distal end, among which one is often noted as coinciding with the median line. Perfect specimens of this type occur very rarely. Generally grains are present in which the distal end (distal to hilum) is tapering, but more or less broadened, approaching the wedge-shaped type (type 8), and sometimes, though rarely, this same end is condensed, or of equal thickness, so as to approach the inverted conical type (type 6) or the rod-shaped type (type 9). Schcenus mucroiiatus Linn.; Cyperus wgyptiacus Gloxin. (Cyperacece.) Dry creeping root-stocks. — Grains spherical and blunt-triangular to elliptical; frequently conical, or somewhat curved with beak-shaped distal end; two-fifths to as thick as broad; no lamelke; instead of the one-fourth to one-fifth eccentric hilum a small cavity is found from which extend single, short fissures. Length 21/j. Among them, isolated doublets, of unequal halves arising from the cutting off of the tapering distal end. Cyperus esculentus Linn. {Cyperacece.) Dry root-stocks. — Grains usually conical, rarely rounded or oval; slightly curved at the thin end and frequently with two beak-shaped processes; one- half to almost as broad as long; the broader ones laterally compressed to about two-thirds; no lamelke; hilum at the broad entl, one-third to one-fourth eccentric; occasionally a small or a slit-like cavity. Length about 12 to 14/i. Commelina hirsuta R. Br. {Commelinaceoe.) Dry tubers. — Grains pear-shaped, or elliptical, or elongated-conical, rarely triangular, frequently somewhat irregular; usually half as broad as long; lamellae numerous and distinct; mostly thickened at the hilum end, the distal end usually i)ointed; and in the triangular grains broailened and squared. Instead of the hilum, a small cavity with radiating fissures; eccentricity about one-tenth. Length about 42//. Some of the grains approach the wedge-shaped (type 8) and the rod-shaped type (type 9). TYPE 7. GRAINS SIMPLE, ECCENTRIC, CONE-SHAPED. 215 Triglochin barrelieri Lois. (Naiadaceoe.) Dry root-stocks. — Grains usually oval-lanceolate or conical, rarely rounded or hlunt-triangular, sometimes curved, frequently irregular; one-fourth to as broad as long; tlie broader ones slightly compressed; lamellae rare and indistinct; usually a longitudinal slit, from which single fissures sometimes diverge laterally. Length about 38/u. The hilum is sometimes at the broad end. On the whole, the majority of the grains show no distinct tjqje. Among them compound grains of few, unequal parts (see tj^DC 15). Very rich in starch. Scilla peruviana Linn. {Liliacccc.) Dry bulbs. — Grains rounded or oval pear-shaped; one-half to almost as broad as long; lamelke delicate, the innermost ones complete, the outermost ones unilateral; hiliun end thickened, the distal end narrow and obtuse, at times cut off, rarely broadened and squared; instead of a hilum, frequently a small cavity from which single short fissures radiate; eccentricity of hilum one-half to one-fourth. Length about 16/*, width 36/i. Among them some compound gi-ains of few equal or unequal parts. Orniihogalurn umbellatwn Linn. {Liliacew.) Dry bulbs. — Grains oval-rounded, rounded pear-shaped, and oval; no lamellse; thickened at the hilum end; frequently a small cavity and some- times single, ■ftith fine fissures; eccentricity about one-third. Length about 2-ifi. Among them a few doublets. Paris quadrifolia Linn. {Liliaceoe.) Dry root-stocks. — ^Grains rounded, elliptical, oval pear-shaped, or reniform, half to almost as broad as long; the broad end slightly compressed, thickened at the hilum end; a small cavity or longitudinal slit; eccentricity about one-quarter. Length about 13//, width lO/n. Rich in starch. Trillium rhomboideum Michx. (Liliacece.) Dry root-stocks. — Grains rounded, rounded triangular to oblong, and pear-shaped; frequently somewhat irregular; one-half to about as broad as long, the broad ones slightly compressed; thickened at the hilum end; a small cavity, or more often a longitudinal slit, instead of the hilum; eccentricity about one-quarter. Length about ll/i, width 9/1. Rich in starch. Billbergia amoena Lindl. {Bromeliacece.) Dry root-stocks. — Grains rounded-oval to elongated pear- shaped, one-half to two-thirds as broad as long, the broad ones slightly compressed; thick- ened at the hilum end; the hilum has a small cavity from which several marked fissures radiate; frequently also one short transverse fissure; eccentricity one-thiril and one-fourth. Length about 21/i. Among them compound grains of few usually equal-sized components (see tj^pe 14); also some semi-compound grains of two small " pai-t-grains." Zostera nana Roth. {Naiadacew.) Dry root-stocks. — Grains conical, or from one aspect a flattened oval, and from the other aspect cone-shaped (the distal end broadened and squared) ; usually almost twice as long as broad; lamellte indistinct, or none; hilum at the broad end, eccen- tricity about one-sixth; usually a small cavitj' from which several short fissures pass out. Length about 32/i. Richardsonia scabra Linn. (Rubiaceoe.) Dry roots. — Grains spherical to oval and conical; frequently 1.5 times, rarely twice as long as broad; almost circular in cross-section; at times with a few inchstinct lamellae; hilian at tlie broad end; rarely more eccentric than about one-sixth; occasionally a small cavity. Length about 21/^. Also compound grains of few, mostly equal components. (See type 15.) Vinca minor Linn. (Apocynacece.) Dry stolons. — Grains rounded to oljlong and conical; about one to three times as long as broad; occasionally broadly triangular, about 1.5 times as broad as long; frequently more or less irregular; sometimes with one or two protruding pointed angles; hilum at the broad end; the distal end rarely broadened into a knife-like edge; occasionally with a small cavity. Length about 13/i. Also some doublets of unequal components. Sim- ilar simple starch-grains occur also in the roots of Vinca minor, besides a few larger ones (about 16/i), which are somewhat broader (at most twice as long as broad) and also rather thicker, and always with blunt ends. Among these are found many compound grains of few almost equal components. Symphytum bulbosum Schimj). (Buraginacew.) Dry tubers. — Grains round to egg-shaped and oval- conical; 1.5 times, rarely twice, as long as broad; lamella? none or very ilelicate; thickened at the hilum end, or with tlie two ends of similar thickness, or the distal <;ud less dense and knife-like; usually with a small cavity and radial fissures; eccentricity about one-half or more. Length 32 to 44/i, thickness 24 to 30/^. 216 DIPFERENTIATrON AND SPECIFICITY OF STARCHES. Symphytum tuberosum Linn. (Boraginacece.) Dry root-stocks. — Grains similar to the preceding, somewhat smaller and less well-developed, spherical to oval; 1.33 to 1.5 times as long as broad; no lamella;; hilum end rarely distinctly broadened; a small cavity and occasionally short, delicate, radial fissures; eccentricity about one-half. Length about 18;u, thickness about 13;u. Also compound grains of few, usually equal components. (See type 14.) Nagelia zebrina Regel. ifihamnaceae.) Fresh scales of the root-stocks. — Grains usually conical, some- times oval-cylindrical, usually twice as long as broad; circular in cross-section; the Ijroad ones slightly compressed; thickened at the hilum end, distal end narrowcil, sometimes broad- ened, obliquely cut-off, or square-shaped; lamella; and hilum distinct; eccentricity about one-sixth. Length about 45ju, width about 33;u. Very often the grain is thickened from the hilum end to about the middle, and from there on it narrows down to a cone, thus represent- ing the appearance of a spindle with the two parts unequal; they approach the rod-form (type 9). Gloxinia speciosa Lodd. (Gesneraceoe.) Fresh tubers. — Grains oval, conical, or oblong, about three times as long as broad; frequently with a lateral appendage (composed of a special external system of lamellae); hilum and lamellaj usually distinct, but delicate; thickened at the hilum end; sometimes the two ends are equally thick, the distal end rarely broader and compressed; eccentricity about one-fourth. Length about 48ju, width 22/j. Some of the grains approach the rod-form (type 9). Gloxinia hirsuta Lindl. {Gesneracece.) Fresh tubers. — Grains similar to the preceding, oval or conical; lamellae not always distinct; hilum usually at the broad end; frequently single, radial fissures. Length about 42/n. Also some doublets and semi-compound grains with 2 to 3 components. Orobanche sp. {Scrophulariacece or Orobanchaceoe.) Dry root-stocks. — Grains spherical to oval, and truncated-conical; 1.5 times, rarely twice, as long as broad; lamellae indistinct; thickened at the hilimi end, rarely broader and thinner at the distal end; a cavity, and marked more or less numerous radial fissures; eccentricity about one-third. Length about 46//, width about 38ju. Also some compound grains of few usually equal-sized components. (See type 14.) Orobanche procera Koch. {Scrophulariacece or Orobanchaceoe.) Dry root-stocks. — Grains rounded- oval to conical and oblong; 1 to 2.5 times as broad as long; lamellae none or delicate; thick- ened at the hilum end; distal end sometimes broadened with a knife-like edge; a small cavity, and frequently very short radial fissures; eccentricity about one-fourth to one-fifth. Length about 28m, width 20//. Some of the grains approach the wedge-shaped type (type 8) and others the rod-shaped type (type 9). Some doublets and triplets are present. Lathrcea squarnaria Linn. {Scrophulariacece or Orobanchacece.) Dry scales of the root-stocks. — Grains oval, conical, elliptical, frequently somewhat irregular; almost circular in transverse section; rarely compressed; as a rule, twice as long as broad; occasionally broader than long; lamellae distinct, the innermost ones complete and rather unsystematically arranged, the outer ones unilateral and crowded; usually thickened at the hilum end, distal end frequently drawn into a conical point or broadened into a knife-like form; radial fissures, especially in the direction of the "verdickungshalbmesser;" eccentricity one-fifth and one-eighth. Length about 125/j. Cyclamen hederifolium Ait. {Pi-imulacece.) Dry bulbs. — Grains triangular, oval, lanceolate, usually conical, often irregular or slightly curved; one-fourth to as broad as long; the broader ones compressed to about one-half, but the smaller not compressed; lamellae none, or indistinct; thickened at the hilum end, distal end narrow, sometimes (h-a^\Ti into a point, sometimes broadened and knife-like; often a cavity, from which short, delicate fissures rarely pass out; eccentricity about one-tenth. Length about 36/i, width about 24/i. Triangular grains frequently occur, with two sets of lamellae at right angles to one another, of which the outer set forms the upper angle of the triangle and the inner one the base; the hilum and the greatest thickness are found in one angle of the base. Dodecatheon meadia Linn. {Primulacece.) Dry roots. — Grains triangular, oval, oblong, conical, elliptical, or spindle-shaped with the two unequal parts; frequently more or less irregular; 2 to 4 times, rarely equally, as long as broad, in the latter case compressed to about one- half; usually without lamellae and hilum. Length about 2G//. The grains probably belong to this class, as their structure seems to resemble that of the cyclamens. Some doublets are found among them. TYPE 7. GRAINS SIMPLE, ECCENTRIC, CONE-SHAPED. 217 Carum bulbocaslanum Kocli.; Hunium bulbocastamtm Liiin. {Lhnbellifera;.) Dry lubcrs. — Grains spherical-oviil or elliptical, frequently thicker at one end, very often reniform or triangular, the majority more or less irregular, some with one or more mammary processes; no lamellse; rarely with a cavity, and short, radial fissures at the thick end, instead of a hilum, frequently \\nth a longitudinal slit; eccentricity about one-fourth. In one sample from Zweibriicken the grains are about 26^ in length, and mostly one-half to two-thirds as thick; in one sample from Zermatt the grains are smaller, about 18/i, and mostly two-fifths to two- thirds as thick as long. Among them compound grains of few, usually unequal components. (See type 15.) Adoxa 7noschatellina Liim. (Caprifoliacece.) Dry root-stocks. — Grains spherical to oval, lamellje rare and indistinct; occasionally a small cavity with several short, delicate radial fissures, usually at the broad end; eccentricity about one-third. Length Zlfi. Among them some compound gi-ains of 2 to 3 equal or unequal components. Umbilicus ■pendulinus DC. (Crassulacece.) Dry root-stocks. — Grains rounded, oval, conical, 3 to 4 angles; usually irregular; three-fifths to as broad as long; the broader ones compressed to about three-fourths; lamellae incUstinct, only the innermost ones complete; thickened at the hilum end; radial and sometimes also irregular fissm-es; eccentricity one-fourth to one-fifth. Length about 50/x. The young grains are conical; the older ones are frequently distorted and knife-like at the distal end. Cephalotus follicularis R. Br. (CephalotacecB.) Dry root-stocks. — Grains roimded, and rounded-tri- angular to oval, and oval-conical; no lamellse; at times a delicate longitudinal slit, occa- sionally with a small cavity instead of the hilum ; tliicker at the hilum end; distal end narrowed, sometimes spread out into a knife-hke edge; eccentricity about one-third to one-fifth. Length about 18 to 21/Li. Cocculus palmatus DC; Menispernum palmatum Lam. {Menispermacece.) Dry roots. — Grains oval, oblong, conical; rarely 3 to 4 angles; rounded and sometimes lobate; usually irregular; half to as broad as long; the broader ones slightly compressed; with 3 to 8 usually complete lam- ellae; cavity instead of the hilum, usually with marked radial fissures at the broad end; eccentricity one-half to one-fourth. Length 90^. According to Payen, they are sometimes as large as 180/:. Among them doublets and triplets of equal or unequal sized components. The gi-ains are usually stained yellow in the cavity and fissures. The coloring disa])pears in water. Ranunculus garganicus Ten. {Ranunculacew.) Dry, thickened roots. — Grains spherical to short- ened-conical; two-thirds to as broad as long; in the spherical ones a small, usually central cavity instead of the hiltim; in the other grains the cavity is about one-eighth eccentric. Length about 14//, wdth about lO/i. Starch plentiful. Ranunculus bulbosus Limi. {Ranunculaceoe.) Dry tubers. — Grains rounded, oval, conical, frequently irregular; half to almost as broad as long; no lamellae; a small ca\-ity instead of the hilum at the wide end; eccentricity about one-fourth. Length about 12/u. Some compound grains of few, usually unequal components. (See type 15.) Starch plentiful, filling the cells. Anemone ranunculoides Liim. (Ranunculacece.) Dry root-stocks. — Grains usually oblong or pear- shaped; generally three-fifths to tlu-ee-fourths as broad as long; no lamellse; at the wide end a small cavity from which a few delicate fissm-es sometimes radiate, or a longitudinal slit; eccentricity about one-third to one-fourth. Length about 13/:. Among them compound grains of few, equal or unequal components. (See type 15.) Starch plentiful, entirely filling the cells. Aconitum anthora Linn. (Ranunculacew.) Dry napifrom roots. — Grains mostly pear-shaped, rarely oval, or rounded-reniform, or rounded-triangular, frequently somewhat irregular; no lam- ellae; broadened at the hilum end, distal end in the triangular gi-ains is broadened and squared; narrowed in the other forms; instead of the hilum, a small cavity from which several fissures radiate; eccentricity about one-third to one-fifth. Length about 38/t. Compound grains of few, equal or unequal parts. (See type 15.) Pceonia officinalis Retz. {Ranunculacecc.) Fresh root-stocks.— GT&ms spherical to oblong and conical; sometimes somewhat irregular, 1 to 2 times as long as broad; lamellse few (2 to 4), delicate, rarely distinct; hilum also intlistinct, eccentricity two-ninths. Length about 28/t, width about 19/:. Many compound grains of few, equal and unequal components. (See type 15.) 21S niFFERENTIATION AND SPECIFICITY OP STARCHES. Corydalis cava Schweigg., Koert.; Corydalis bidbosa Pers. {Fumariacem.) Fresh and dry tuberous root-slocks. — Grains rounded to oblong, often conical to reniform-triangular; generally some- what irregular; two-fifths to as broad as long; the broader ones slightly compressed; no lamellae; hilum at the broad end, one-fourth to one-fifth eccentric; in the triangular forms the distal margin is squared, so that the grain may be as much as 3 times as long as broad. In the dry grains either a longitudinal slit or a large or small cavity \vith single radial fissures. Length about 17 to 21^. Many grains approach the wedge-shaped type (type 8). Among them some compoimd grains of 2 to 4 usually unequal components. Cardaminc granulosa All. (Brassicaceoe.) Dry root-stocks. — Grains rounded, oval, conical, frequently more or less irregular; the broader ones slightly compressed; usually homogeneous; rarely with a small cavity at the broad end, and with a delicate longitudinal slit; eccentricity about one-third and one-fourth. Length about 14/j. Some grains appear to approach type 8. Among them some doublets and triplets. Dentaria digitata Lam. (Brassicacece.) Fresh and dry scales of the root-stock. — Grains oval or pear- shaped, almost circular in transverse section; two-fifths to almost as broad as long; lamellae delicate (5 to 7), only the innermost ones complete; hilum at the broad end, one-fourth to one-sixth eccentric; the distal end frequently elongated, bluntly cut and squared. In the dry grains a small cavity, and occasionally a few delicate fissures. Length about 32/i, width about 25/u. Among them some doublets. In the young scales the starch-grains are about ItV long, and usually 2 to 3 times as long as thick, and homogeneous. Dentaria polyphylla W. K. (Brassicacece.) Dry scales of the root-stocks.- — Grains rounded-oval and blunt-triangular to oblong and elongated-conical; two-fifths to almost as broad as long, the broader ones slightly compressed; lamellae distinct, only the innermost ones complete; usu- ally at the broad end, instead of the hilum, a small cavity is found from which frequently a few short fissures radiate; eccentricity about one-fourth and one-seventh. Frequently the distal end is broader, but somewhat thinner, but at times thicker than the more dense hilum end; very often it is broadened by a lateral layer of lamellae, the deposit forming a right or an obtuse angle with the axis of the grain. Length about 60n, width aljout 40(U. Some grains ajiproach the cuneiform and some the rod-shaped type (type 9). Nymphcea alba Linn. {Nymphoeacea;.) Fresh root-stock and roots. — Grains ovoid, oval and conical; twice as long as tliick; hilum at the thick end, one-fourth eccentric. Length about 14//, thickness about 10/i. Among these are compound grains of few, equal or unequal compon- ents. (See type 15.) Stellaria bidbosa Wulfen. (Caryophyllacew.) Dry tubers. — Grains spherical, oval, or conical; 1 to 2 times as long as thick; at the thick end, instead of a hilum, a small cavity is found, one- fourth eccentric; the distal end narrowed, or as liroad or somewhat broader than the hilum end, and compressed. Length about 10/i, with compound grains of few usually equal com- ponents. (See type 14.) Althcea officinalis Linn. (Malvacem.) Dry roots. — Grains rounded, oval, oblong, conical, frequently somewhat curved or reniform or rounded-triangular, frequently somewhat irregular; two- fifths to as broad as long; or the l)roader ones slightly compressed; no lamellae; rarely only a cavity at the thicker end instead of the hilum, eccentricity about one-fourth; more often an irregular, longitudinal slit. Length about 21^, width 31/:. Among them few doublets and triplets. Euphorbia dulcis Jacq. {Euphorbiaceoe.) Dry root-stocks. — Grains rounded, rounded-triangular, oval, mostly conical, some over twice as long as thick, circular in transverse section; others twice as broad as long; comjjressed and squared at the distal end; eccentric; several radial fissures or a longitudinal slit present. Length about 13/<. Among them are found some doul)lets and triplets. Starch plentiful. Geranium phmim Linn., var. lividum L'H(!rit. {Geraniacew.) Dry root-stocks. — Grains rounded- triangular, oval, oblong, or conical, more or less irregular; two-fifths to almost as broad as long; the narrow ones circular in transverse section, the broad ones slightly compressed, thickened at the hilum end, the distal end somewhat broader but comjjressed; a small cavity is found from which a tlelicate longitudinal slit or radial fissures diverge; eccentricity about one-eighth. Length about 18//, wiiltli 13/j. TYPE 7. GRAINS SIMPLE, ECCENTRIC, CONE-SHAPED. 219 Geraninm sijlmtiam Linn. {Gcraniacece.) Dry root-stocks and roots. — Grains as in the preceding; most of them conical; many also triangular, about twice as broad as long; flattened at the distal margin; hilum end always thickened. Length about 21^. Transition into wedge- shaped tyjie (type 8). Oxalis acetosella Linn. [O.ralidacece.) Dry scales of the root-stocks. — Grains rounded-triangular, oval, conical, oblong, frequently more or less irregular; 0.33 to 1.5 times as broad as long; the broader ones compressed to about half the width; hilum end usually thickened, sometimes both hilum and distal ends of equal thickness; a small cavity with delicate radial fissures or a longitudinal slit is present. Length about 22//. Transition into the wedge-shaped type (type 8), antl to the rod-shaped type (type 9). Some doublets and triplets. Orobus alius Linn. (Leguminosa.) Dry thickened roots. — Grains oval or conical, usually L5 times, rarely twice, as long as thick; at the thicker end instead of a hilum a small cavity is present, sometimes with single radial fissures; eccentricity about one-seventh. Length about 16yu. Among these are found compound grains, of few usually unequal components. (See type 15.) Apios tube)-osa Moench. {Lcguminosce.) Fresh tubers. — Grains elliptical, oblong, usually somewhat irregular; 1.33 to 2.5 times as long as thick, circular or slightly compressed in transverse section; lamellae indistinct; hilum often visible, small, mostly at the thick end, one-third to one-seventh eccentric. Length about 30/^. Among them are found compound grains of 2 to 4 components. Encephalartos spiralis Lehm. {Cycadacece.) Dry embryo. — Grains rounded-oval or conical, 1 to about 1.5 times as long as thick; vnih. single lamellie; at the thicker end, instead of the hilum, a small cavity is present, from which delicate, short fissures radiate; eccentricity about one- third and one-fourth. Length 30 to 35ju, width 25 to 30ju. Among these are found compound grains of few usually equal components. (See type 14.) All forms of transition into the simple type may be found among the separated grains. Visciim album Linn. {Lwanthacew.) Fresh seed and embryo. — Grains rounded to rounded-oval, usu- ally without lamella; hilmn often distinct, about one-third to one-fourth eccentric. Length about 2-ifi. Also compound grains of 2 to 4 components. The starch-grains in the embryo are like those in the seed, though smaller. Loranthus europoeus Linn. {Loranthacece.) Dry seed and embryo. — Grains rounded, oval-elliptical, or conical; frequently slightly curved or irregular; no lamelhe; hilum at the thicker end, one- third to one-fourth eccentric; instead of the hilum, a small cavitj^ with several radial fissures may be present, rarely a longitudinal slit. Length about 18ju, width 12yLi. There also are present numerous compound grains of 2 to 4 and more components. Starch plentiful in the seed; also in the embryo, though the grains are smaller. Psittacanthus vellozianus Mart. {Loranthacea;.) Dry embryo. — Grains oval, or conical, 1.5 times to almost twice as long as thick; rarely with distinct lamellae, the outermost ones complete; instead of a hilum, a small cavity is frequently present, from which sometimes rather short, delicate fissures radiate; eccentricity about one-foiu-th; the hilum end usually thicker, fre- quently distinctly narrower, in which case the distal end also becomes somewhat narrower. Length 35 to 40/j, thickness about 30ju. The starch is a transition form to the inverted conical type (type 6) and the rod-shaped type (type 9). Carolinea princeps Limi. (Malvacew.) Dry cotyledons. — Grains spherical, spherical-oval, or short- ened-conical; lamellaj indistinct; at the thicker end, instead of the hilum, a small cavity is found; eccentricity about one-half to one-third. Length about 25/i. There are also present compound grains of few usually unequal components. (See type 15.) Sterculiacece. Dry seeds. — These starch-grains greatly resemble those in the cotyledons of the Legu- minosce [Phaseolus, Vicia, etc.). They are oval, reniform, rounded-triangular, the broader ones slightly compressed, a fissure coinciding with the largest diameter is found in the flat grains, and on the circular ones a cylindrical canal. These at first on this account were classified by Niigeli among the centric-oval type (ty])e 3). Triangular forms also occur with fissures radiating from an eccentric point, as well as conical ones in which this point is at the thickened end. It seemed to Nageli therefore more likely that these starch-grains are of an eccentric structure and that they belonged partly to the conical type (type 7) and partly to the wedgtvshapeil type (type 8). Investigation ujion the fresh grains Nageli states must decide this point. 220 DIFFERENTIATION AND SPECIFICITY OF STARCHES. Abroma angushwi Linn. fil. (Sleradiacew.) Dry seed. — Grains rounded-oval to oblong; sometimes curved, or reniform, or triangular, rarely conical, often irregular; two-fifths to as broad as long; the broader ones compressed; the narrower ones terete; no lamellse; frequently from the narrow aspect a longitudinal slit is distinctly visible, and sometimes a small cavity with radial fissures is found at the thickened end; eccentricity one-third and one-fourth. Length 30,u, width 20ju. Compound grains of few components are present. Starch and oil present. Wallheria indica Limi. {Sterculiacece.) Dry seed. — Grains rounded-oval to elongated-oval, fre- quently curved, reniform, or triangular, or even irregularly triangular or quatlrangular, and a few tetrahedrons; two-fifths to almost as broad as long; the broader ones slightly compressed, and usually with a longitudinal slit, the narrower ones terete and with a channel-like furrow, also sometimes with several transverse or radial fissm-es; lamelliE rare, or very delicate. Length about 30 to 35^, width 25fi. Some compound grains of few components are present. Melochia pyramidata Linn. (Sterculiacece.) Dry seed. — Grains rounded to elongated-oval, some- times reniform or rounded triangular; the broader ones compressed to about half their width; lamellae delicate, or none; from the narrow aspect a longitudinal slit is usually very distinct, sometimes transverse fissures are also present. Length about 30 to 36/i, width about 15/x. Some doulslets and triplets. Riedlea corchorifolia DC; Melochia corchorifolia Linn. (Sterculiacece.) Dry seed. — Grains rounded- oval to oblong; frequently curved, reniform, or triangular, very often somewhat irregular; two-fifths to as broad as long, the broader ones slightly compressed; the smaller ones terete; no lamellae; from the narrow aspect usually a clearly defined longitudinal slit, and occasion- ally many transverse fissures. Length about 31^, width about 19m. Hermannia alihccefolia Linn. (Sterculiacece.) Dry seed. — Grains rounded to oblong, sometimes coni- cal or rounded-triangular, the longer ones one-third to one-half, the shorter ones half to as broad as long; the broader ones compressed to over half their width, the smaller ones not compressed; no lamellae; with a longitudinal slit somewhat more marked from the narrow aspect. Length about 20yu, rarely 28/;. Hermannia nemorosa Eckl. (Sterculiacece.) Dry seed. — Grains rounded to elongated-oval, at times somewhat conical or triangular ; half to as broad as long; the broader ones slightly compressed; no lamellae; a longitudinal slit is usually observed. Length about 18/i. Melhania didyma Eckl., Zeyh. (Sterculiacece.) Dry seed. — Grains rounded-oval to elongated-oval, frequently curved, reniform, triangular, rarely conical, often somewhat irregular; two-thirds to as broad as long, the broader ones slightly compressed, the narrower ones terete; no lamellae; from the narrow aspect usually a clearly defined longitudinal slit, and occasionally a single transverse fissure, are observed. Length about 30 to 35/f, width 20 to 25/i. Melhania erthroxylon R. Br. (Sterculiacece.) Grains as in the preceding. Eriolcena species. (Sterculiacece.) Dry seed. — Grains rounded-oval, oval, rarely elongated-oval, many reniform or triangular, few conical, one-half or rarely two-fifths to as broad as long, the broader ones slightly compressed; lamellae very delicate, or none; u.sually a longitudinal slit, and sometimes single transverse fissures are present; at the thickened end a small cavity from which radial fissures diverge; one-half to one-third eccentric. Length about 33/;, width about 27/j. Visenia tomemtosa R. P. (Sterculiacece.) Dry seed. — Grains rounded to oblong, some cm-ved, or reniform, or triangular; one-half or rarely two-fifths to as broad as long; the broader ones slightly compressed; no lamellae; usually \vith a longitudinal slit, sometimes with transverse fissures. Length about 28^, width 20^. JEsculus hippocastamim Linn. (Sapindaceoe.) Fresh and dry cotyledons. — Grains usually conical, occasionally oval or triangular or angular; frequentlj- irregular obtuse angular forms; half to as broad as long; lamella- and hilum indistinct, or delicate; thickened at the hilum end, narrowed at the distal end, having a circular transverse section, or broadened and thinned and angular (the width may thus exceed the length of the gTain); in dry grains there is present a small cavity and 1 to 4 radial fissures, especially in the direction of the long axis; eccentricity about one-seventh. Length 29/i, rarely 30/u. There are present a few semi- compouiul and compound grains, each with 2 to 4 components. TYPE 8. GRAINS SIMPLE, ECCENTRIC, CUNEIFORM, OR FLATTENED. 221 Aniyris sylvatica Jacq. (Burseracew.) Dry cotyledons. — Grains rounded, oval, conical, two-thirds to almost as thicl< as long; no iainclla?; at the thiciv end, instead of the hihun, a small cavity with single fissiu'os is observed; eccentricity one-third and one-fourth. Length about 17ju, width 13/i. There are also present compound grains of 2 to 4 (rarely more) equal or unequal components, and numerous separat<^d grains (4 to 12) with 1 curved surface and 1 to 3 pres- sure facets; with a small cavity and radial fissures. Amyris species. (Burseraceoe.) Dry cotyledons. — Grains rounded, oval, ellipsoid, shortened-conical, occasionally somewhat angular; half to as broad as long; the larger grains have a small cavity and single radial fissures. The cavity central, if one-half eccentric, then it is at the thicker end. liCngth about 15^, width about 12yu. Among these are compound grains of 2 to 4 or more equal or unequal components witli one curved surface and 1 to 3 pressure facets ; a small central cavity and single radial fissures are present. Cells are entirely filled with starch; little or no oil. Peganum harmala Linn. {Rxdacew.) Dry unripe seeds. — Grains rounded, oval, shortened-conical, frequently more or less irregular; no lamellae; instead of the hilum, sometimes a small cavity, occasionally with short radial fissures, usually near the thick, rarely at the thinner end. Length about 14ju. Compound grains of 2 to 5 equal or unequal components. Starch plen- tiful in unripe seeds; ripe ones contain oil, but no starch. Memecylon capense Eckl. (Melastojnacece.) Dry cotyledons. — Grains conical, oval, frequently unsym- metrical, and oblique or curved; one-half to almost as thick as long; no lamellae; a longitudinal slit Math single transverse fissures; or a small cavity with radial fissures; the cavity is at the thick end, and eccentric about one-half to one-fourth. Length about 16 to 20/j, thickness about 15m. There are also present compound grains of few usually equal components. (See type 14.) Memecylon amplexicaule Roxb. {Melastomacece.) Dry cotyledons. — Grains as in the preceding, but less conical and more symmetrical. Also similar compound grains. Syzygiuni guineense DC. {Myrtacece.) Dry cotyledons. — Grains oval-spherical to elongated-oblong, occasionally somewhat irregular; two-fifths to almost as thick as long, the broader ones slightly compressed; lamellre numerous, delicate, sometimes not distinct, the innermost ones complete, the outer ones unilateral; instead of the hilum, there is a small cavity from which single fissures radiate, chiefly in the direction of the long radius; hilum end usually thick- ened, the distal end sometimes narrower and thinner, and sometimes wider, also either taper- ing or of equal thickness. Length about 34 to 42//, thickness about 25/^. Carophyllus aromaticiis Linn. (Myrlaccw.) Dry cotyledons. — Grains oblong or elongated-conical; usually two-fifths to three-fifths as broad as long; the broadest ones slightly compressed; lamella; delicate; only the innermost ones complete; a small cavity is usually present instead of a hilum, and frequentlj^ a lengthwise slit with single olilique fissures; thickened at the hilum end, distal end narrowed and at the same time pointed or rounded, or rarely broadened and compressed into a knife-like edge, and besides often cut off obliquely; eccentricity about one-fifth to one-sixth. Length about 36^. Jambosa vulgaris DC. (Myrtacece.) Dry cotyledons. — Grains usually conical, and circular in trans- verse section; the longer ones about half as thick as long; the shorter ones thicker and some- times rounded-triangular; usually without lamellae, more rarely with few indistinct lamellae, only the innermost ones being complete; instead of a hilum, a cavity is observed with a longitudinal slit, and sometimes with radial fissures; the cavity is at the thicker end, about one-sixth eccentric. Length about 30 to 34^. Amphiairpwa monoica Nutt. (Legmninosce.) Dry cotyledons. — Grains rounded-oval or conical; half to as broad as long; the broader ones sometimes slightly compressed; instead of a hilum, a small cavity with single, delicate, short fissures is usuallj^ observed; thickened at the hilum end; distal end tapering and conical, rarely broadened and squared. Length about 44/^. Semicom- pound grains of 2 to 5 part-grains and compound ones of few usually equal components. Type 8. Grains Simple, Eccentric, Cuneiform, or Flattened. Lamellae coarsest and most numerous on one side, and finest and fewest on the diametrically opposite side. Grains usually Ijroader at the distal end, at the anterior end narrowed and compressed; either equally thick througliout, or more frequently thicker at the anterior end, and at the distal 222 DIFFERENTIATION AND SPECIFICITY OF STARCHES. end flationod, ixnf^ulur, and squared. Hilum usually toward the narrow, thick end. As a rule, only a few of the innermost laniellir are complete; all the outer ones apix-aring incomplete. On drying, the grain usually develops a cleft, coinciding with its greatest plane, which is visible from the narrow aspect. From the broad aspect several fissures are seen radiating from the hilum chiefly toward the distal margin; instead of the latter, one longitudinal or one transverse slit, or both combined, occasionally are observed. Pure types of the grain rarely occur, they are usually found mingled with others, mostly with the cone-shaped type (tyjje 7), less frequently with the rod-shaped type (type 9), and the inverted cone-shaped type (type 6). Erythronium dens-canis Linn. (Liliaceoc.) Dry bulb scales. — Grains wedge-shaped from the broad aspect, rounded circumference, 0.6 to 1.5 times as Ijroad as long; cone-shaped from the narrow aspect; hilum end is thickened, often somewhat protruding, the distal end is usually curved; laraellce none, or very delicate; instead of the hilum a small cavity may be observed from which several delicate fissures emerge; eccentricity about one-ninth. Length about 34/x. Tulipa gesneriana Linn. {Liliacew.) Fresh bulb scales. — Grains rounded, cuneiform, almost as broad as long; lamellse and hilum about three-fifths as thick as broad, and are rarely distinct; eccen- tricity about one-ninth. Length about 30/i, width about 27jjl. According to Raspail the size may reach 50^. Tulipa sylvestris Limi. (Liliaccce.) Fresh bulb scales. — Grains rounded-triangular or oval-cunei- form, and oval from the narrow aspect, elongated-conical, usually one-half to two-thirds as broad as long; compressed to about one-fourth or more of their \vidth, hilum end often some- what protruding and thickened, lamellae and hilum delicate or indistinct, eccentricity about one-tenth. Length about 50^1. The smaller grains are usually oblong, and formed like little oval rods. Frilillaria meleagris Linn. (Liliacew.) Scales of dry bulbs. — Grains oval or rounded-triangidar, occasionally somewhat protruding at the hilum end; the broader ones compressed to about one-half; the distal end of equal thickness or somewhat thicker than the hilum end; lamellae delicate; occasionally a cavity, rarely with several short very delicate radial fissures, is found instead of the hilum; eccentricity about one-sixth. Length 21yu. Lilium candidwn Linn. Bulbes de Lis. (Liliaceoe.) Bulbs. — ^According to Payen (Ann. Sc. Nat., 1838, II, p. 17; pi. 4, fig. 5), the grains are oval, or oval-triangular; almost twice as long as broad; lamellae delicate; hilum, which is sometimes double, at the narrow end; eccentricity about one-tenth. Length about ll.'j^u. The fully developed grains are irregular and often rough, with fissures radiating from the hilum. Liliuvi bulbiferum Linn. (Liliacew.) Bidbs. — According to Scbleiden (Grundzuge, 3d ed., auf. 183, fig. 7), the grains are musscl-shell-shapcd and broader than long; from the small aspect they are oval; lamella; and hilum distinct; eccentricity one-fifth to one-sixth. Muscari bolryoides Mill. (Liliacew.) Scales of dry bulbs. — Grains rounded, mostly with triangular, rarely rhomboid or irregular, outlines; 0.75 to 1.33 as broad as long; compressed, and from the narrow aspect, conical in form; thinner toward the distal end, or at times of equal thickness; lamellae usually distinct, but delicate; sometimes a small cavity instead of a hilum; eccentric- ity about one-sixth; from the narrow aspect, a longitudinal slit is usually present. Length about 30/i. Hyacinthus orientalis Linn. (Liliacew.) Scales of fresh bulbs. — Grains oval, cuneiform, triangular, more or less irregular, frequently -with protruding angles; the distal end, wliich is thinned and squared, is as Inroad as, or broader, than the hiliun end; lamelUe more or less distinct; hilum at the thicker end; eccentricity about one-fifth. Length about 45/i. Some semi- compound (see type 11) and compound grains arranged in 1 to 2 rows (see type 13). Simple grains were almost exclusively found by Payen (Ann. Sc. Nat., 1858, ii, p. 22). Scilla aulumnalis Linn. (Liliacew.) Scales of dry bulbs. — Grains rounded, oval, and triangular; instead of the liilum a small cavity with single short fissures may frequently be observed at the thicker end; about one-fourth eccentric. Length about 24yu. These are transition forms to the cone-shape (type 7). Dioscorea batatas Desne. (Dioscoreacew.) Fresh tubers. — Grains rounded, rounded pear-shaped, usu- ally with 3 to 4 and 5 rounded angles, frequently somewhat irregular; 0.5 to 1.5 times as broad as long; compressed to about one-half to one-tbird of their ■nidth; from the narrow aspect TYPE 8. GRAINS SIMPLE, ECCENTRIC, CUNEIFORM, OR FLATTENED. 223 coiii^-sluipod ; tliiiiiuT towjirds the distill end; lamellae delicate or indistinct; hiluni frequently invisible, approaching the thicker end; eccentricity one-sixth to one-eighth. In large tubers (20 em. long by 2 em. thick) the grains are larger and relatively more compressed, and have ratlier more distinct Uuncihe. Length about 75ix, width 45^, thickness about 25;u, average about GSyu. In small tubers (14 nnn. in length, .ind of equ.il thickness) the starch-grains are smaller, relatively thicker, and with less distinct lamella. In the narrow (2.5 by 4 mm.) part of the root which bears these small tubers, grains are almost without lamelliB, more rounded, .and their size is about 18 to 24/i. Dioscorea saliva Linn. {Dioscoreacece.) Rool-stocks. — According to Leon Soubeiran (Journ. Pharm., 1854, XXV, 181) the smallest grains are usually spherical or oval, the larger ones pear-shaped or elongated, the largest ones indistinctly triangular; lamella; and hilum none. Length 40 to 50/<, diameter 10 to 20ft. According to Raspail, the size is up to GOyu. Dioscorea alala Linn. (Dioscoreacew.) Root-stocks. — According to Payen (Verhandl. der Paris Acad- emy, 1847, July 26), the grains are irregular, spherical, and without Lamella;; 2 to 13 adher- ing to each other. Galanthns 7iivalis Linn. {Amaryllidaccce.) Fresh bulb scales. — Grains almost circular and rounded- triangular to oval, frequently broader than long; thickened and narrowed at the hilum end; distal end broad and thinned ; lamellte rather distinct, the inner ones complete, the outer ones unilateral; hilum one-half to one-fourth eccentric, sometimes with a few short radial fissures. Length about 28ju. Among these are found some isolated, semi-compound grains. At the position of the hilum several part-grains are found lying beside one another, causing the proximal end to increase in width; also some compoimd grains with few, unequal components (sec tJ^^e 15), and some with 2 to 5 equal components, mostly arranged in one row. Leucoium vernum Linn. {Ainaryllidacece.) Fresh scales of the bulbs. — Grains rounded, triangular wth rounded angles, usually broader than long to twice as broad, rarely somewhat longer than broad; strongly compressed; hilum almost in the middle of the thicker margin, from which point a large mammary process projects; the distal margin is less thick and squared. Lamellae delicate, often indistinct; eccentricity one-third to one-ninth. Length about 25 to 30m, width 40 to 50^. Phaius grandiflorus Lour. ; Bletia tankertnllew R. Br. (Orchidaceoe.) Pseudo-tubers. — According to Leon Soubeiran (Joimi. Pharm., 18.54, xxv, 181), grains generally oval or triangular, some much elongated, others with rugged protuberances; lamella; distinct; no hilum. Width 40 to 50/u, length 100 to 200yu. Judging from the outline sketches, the grains nuist Iw nuicli flattened. According to Schleiden (Grundzuge, 3, auf i., p. 183, fig. 8), the grains are oval or rounded- conical, almost twice as long as broad; lamella; distinct, frequently with an outer lateral group of lamellic; hilum at the narrow end, one-sixth to one-tenth eccentric. Zingiber officinale Rase. {Zingibcraceoc.) Dry root-slocks. — Grains rounded-cuneiform, oval, 4 and 5 angular, often somewhat irregular; half to as broad as long, the narrower ones compressed to about one-half or one-third, and the broader ones to about one-fourth or one-fifth of their width. The hilum end narrowed, triangular, or with a protuberance, which is sometimes turned laterally; lamellae invisible; hilum frequently indistinct; instead of the hilum, there is often a small cavity with 1 or 2 very short fissures; eccentricity one-eighth to one-eleventh. Length about 45^, width about 30 to 40m, thickness 7 to 9ju. Curcuma zedoaria Salisb. {Zingiber acew.) Dry tubers. — Grains oval, elongated, more or less trian- gular, occasionally irregular, two-fifths to two-thirds as broad as long; compressed to about one-half or one-third of their width (one-third to one-sixth as thick as long) ; from the nar- row aspect of equal thickness throughout the entire length, with rounded ends; from the broad aspect the distal end is narrowed-triaugular or protruding, at times turned laterally, frequently with a pointed wart-like protuberance; lamelhc numerous, delicate, and incom- plete; instead of the hilum a small cavity is rarely present; 0.04 and 0.03 eccentric. Length about 70m, thickness about 12^. Curcuma leucorrhiza Roxb. {Tikhur-flour, partly East India arrowroot; Travancora starch.) (Zingih- eracece.) Root-stocks.— According to Walpers (Bot. Zeit., 1851, 337), Schleiden (Grundzuge, 3d ed., I, p. 185, fig. 11), Berg (Pharmacognosie, 481), and Leon Soubeiran (Journ. Pharm., 1854, xxv, 178), the grains are oval, elliptical, ovoid, elongated-oval, or almost spatulate or elongated-triangular; towards the base suddenly narrowed, short-pointed, or drawn out 224 DIFFERENTIATION AND SPECIFICITY OF STARCHES. into a rather elongated blunt point; strongly compressed to about one-fourth of their width; lamellae none, or numerous and delicate; occasionally a small transverse fissure in the hilum; hilum about one-seventeenth eccentric. The grains are as large as, and occasionally larger than, those of potato starch. According to Soubeiran, width 20ju, length 60 to 70/u, and thickness 10 u. Starch-grains of Tikhur-Mehl flour are about 72fx long and 38/i broad, thus they are 1 to 3 times as long as broad; 6 to 8yu thick; sometimes from the broad aspect they are oblique or irregular, and broadened toward the distal rounded end; while at the proximal end the}' are usually narrow and triangular; this end may occasionally bo either almost l)lunt or the angle may protrude; even the place where the hilum is located may occasionally be more prominent and wart-shaped ; in the long, narrow aspect the grains are rod-shaped, and of cither almost equal thickness throughout or slightly thicker in the middle, with rounded or blunt ends; the hilum is rarely indicated by a small cavity about one twenty-fourth eccentric; lamellse deli- cate and numerous. Curcuma longa Linn. (ZingiberacecB.) — According to Munter, grains as in preceding. Curcuma anguslifolia Roxb. (Zingibcracece.) Root-stocks. — According to Leon Soubeiran (Joum. Pharm., 1854, xxv, 178), the grains are triangular, with rather blunt, not flattened, angles; lamellse and hilum visible, but indistinct, grains very unequal in size, from 5 to 30m, small ones numerous, many split and torn. Soubeiran concludes from this that Travancora starch can not be derived from Curcuma angustifolia. Hedychium flavescens Carey. (Zingiberacece.) Root-stocks. — According to Fritzsche (Poggendorff's Annal., 1834, xxxii, 142, taf. ii, 40-49), the gi-ains are oval to lanceolate, narrow at the hilum end, or drawn out to a point; one-third to one-half as long as broad; very strongly compressed; lamellse numerous, delicate; hilum invisible. Also semi-compound grains are present, \vith 2 to 4 part-grains, which at the hilum end are completely separated by fissures, but at the distal end are surrounded with the common lamellse of the grain. Maranta arundinacea Linn. {Jamaica West Indian arrowroot, partly East Indian arrowroot; Mar- anta starch.) (Marantacece.) Dry stolon. — Grains rounded to oval, rarely oblong, usually more or less irregular; often triangular; two-fifths to twice as broad as long; the oblong and oval grains are about circular in transverse section; many are pressed in until they are about one-half as thick as broad; thicker at the hilum end; from the broad aspect the distal end is either narrower or broader than the hilum end (to about twice as broad as the length of the entire grain), the distal end is always thin; lamellaj delicate; instead of the hilum a small cavity is frequently found, cither with a rather delicate, short, transverse fissure, or with 3 to 4 short radial fissures, or rarely with a longitudinal slit; eccentricity one-half to one-sixth. Length 40 to 50//, width aliout 40/i. According to Soubeiran, the length is 60 to lOfx. Occasionally a smaller part-grain rests upon a grain, as if the grain had been cut off at one corner. Jamaica arrowroot sp. — Related to the foregoing starch is a specimen which was sent to Niigeli from England as a variety of potato starch, but which, however, was found to be a kind of Jamaica arrowroot. The grains are oval, elongated-conical, triangular to quadrangular, usually more or less irregular; two-fifths to twice as long as broad; the hilum end thickened, the distal end broadened; compressed, and almost obliquely angular; lamellse quite distinct; delicate, the outer ones frequently forming a special lateral group ; the grain is usually solid, instead of the hilum there is occasionally observed either a short longitudinal or transverse slit or very short radial fissures, which form a triangle, or an oblique-angled cross or more rarely a right-angled cross; eccentricity about one-fifth. Length about 50fi. This starch in some respects acts like potato starch, or as a variety derived from it, swelling up and passing into solution when cooked, dissolving if subjected to moist heat, and soluble when treated with saliva. Maranta starch is easily distinguished, however, from potato starch, with which it is often adulterated, by means of the smaller, often obliquely triangular grains (scarcely two-thirds as large as potato starch-grains) ; by the many wedge-shaped, compressed forms with thickened hilum end and thinned angular distal margin (in the potato starch the hilum is usually found at the thinner end); by tiie more delicate, less distinct, narrower lamellse and the short slits or radial fissures which develop when the grain is dry (of which, as a rule, the potato starch shows no indication). TYPE 8. GRAINS SIMPLE, ECCENTRIC, CUNEIFORM, OR FLATTENED. 225 Canna. (Cannacece.) The root-slock. — The starch-grains are distinguished by the marked eccen- tricity of the hiluin, by the unilateral, incomplete, usually very distinct, coarse lamellae, and by their more or less mussel-like form, which is broadened and flattened, and at the hilum end somewhat protruding. In different parts of the same root-stock, and even in the same tissue, the grains present great variations in shape. In the following notes Nageli commu- nicates the results of his researches upon a series of species of Canna without regard to the question whether the variations are of specific or of merely individual value. Canna pediinculata Sims. Fresh root-stock. — Grains oval to transverse-oval; at the hilum end usually narrowed (rounded cuneiform) or even slightly protruding; sometimes with lobate pro- jections; half to twice as broad as long; the broader ones compressed to about one-third of their width; lamellae indistinct; hilum often indistinct, and found either in the protruding blunt end, or in a small projecting wart-like structure; neither fissures nor cavity are observed in the dry grain. Length about GO/i. There are also compound grains with components arranged in a single row. (See type 13.) Canna picia Hort. Fresh root-stock. — Grains oval to transverse-oval, triangular, usually protrud- ing at the hilum end; compressed to about one-half to one-third of their width, lamellie are numerous and distinct. Length about 112/x. In many grains there is an internal system of lamellae, the longitudinal axis of which deviates from that of the outer system of lamellae, mostly by about 90° or somewhat less; occasionally about 2 to 3 small part-grains lie beside one another at the hilum end. Canna linkii Bouch^. Fresh root-stock. — Grains rounded, or transverse-oval, once to almost twice as broad as long; compressed to one-half, rarely one-foiu'th, of their width; lamellje numerous and distinct, the lateral ones incomplete; hilum in the slightly elongated, blunt hilum end; eccentricity one-sixth to one-tenth. Length about 70^1, thickness 15 to 25^. Also a small number of semi-compound grains are present, which, at the hilum end, have 2 to 3 hila lying beside one another; and also scattered compound grains consisting of a few components. Canna vitatia Hort. Fresh root-stock. — Grains similar to those of Canna linkii; usually rounded or transverse-oval, 1 to 1.5 times as broad as long; distal margin usually rounded; proximal margin rather straight, on both sides of which are two blunt often somewhat prominent angles; in the center is a wart-like protuberance in which the hilum is found. Length about TSyu- Canna altenstdnii Bouche. Fresh root-stock. — Grains similar to those of Canna linkii; 0.66 to 1.5 times as broad as long; lamellae nearly always indistinct; upon drjing often a short, somewhat curved, transverse slit appears in the hilum. Length about 70;u. Canna variegata Hort. Fresh root-stock. — Grains as in Canna linkii; at the hilum end mostly slightly projecting; 0.66 to 1.5 times as broad as long; lamellae numerous, quite distinct. Length TOm- Canna floribunda Hort. Fresh root-stock. — Grains similar to the preceding; mostly rounded. Length about 95;u. Canna albiflora Hort. Fresh root-stock.— Giaivvs similar to those of Canna variegata; mostly more or less rounded; somewhat narrowed and projecting at the proximal blunt end; 0.33 to 1.5 times as broad as long; compressed to about one-third of the width. Canna ramosa Hort. Fresh root-stock. — Grains as in the preceding species; frequently somewhat irregular; 0.66 to 1.5 times as broad as long; eccentricity about one twenty-second. Length about 105^. Many grains with an inner system of lamells, the longitudinal axis of which deviates about 90° from that of the external system. There are also several semi-compound grains with 2 to 3 small part-grains lying next to each other at the hilum end; and also some compoimd grains. Canna elegans Hort. Fresh root-slock. — Grains triangular with blunt or rounded corners, sometimes mussel-shell-shaped, with protruding hilum end; 0.75 to 1.5 times as broad as long; compressed to one-third of their width; lamellae few; hiliun one-sixth and one-tenth eccentric. Length about 40/i. Some semi-compound grains are present with 2 to 4 part-grains lying next to each other at the hilum end; and also compound grains consisting of 2 to 5 almost equal components. Canna speclabilis Hort. Fresh root-stock.— Grains rounded, oval, sometimes rounded cuneiform; more or less protruding at the hilum end; three-fifths to just as broad as long; compressed to about one-third, occasionally to one-fourth of their width; lamella; numerous, distinct, 15 22G DIFFERENTIATION AND SPECIFICITY OF STARCHES. incomplete; hilum in the protruding blunt end, one-eighth to one-thirteenth eccentric; no fissures in the dry grains. Length about 120^, width 20 to SO/u. Some grains with projec- tions on the surface; others with 2 groups of lamcllsc, the longitudinal axis of which changes from 0° to 180°. Semi-compound grains are present with 2 to 3 hila Ij'ing next to each other at the hilum end ; also some compound grains. Carina limbata Rose. Fresh root-stock. — Grains rounded, oval, oblong, and conical; frequently irregular obliquely angular; half to as broad as long; the broad ones compressed to about one-half their wdth; lamella; and hilum frequently indistinct. Length about 95^. Some grains with an inner group of lamellae. Some semi-compound grains with 2 or more hila lying behind one another; and also compound grains are present. Carina lanuginosa Bosc. Fresh root-stocks. — Grains oval; very seldom elliptical, or 3, 4, or 5 angular, with blunt angles ; one-half to almost as broad as long; slightly compressed; lamella? numerous, complete; hilum in narrow, often wart-like protruding upper end; eccentricity one-thirtieth to one-fortieth. Length about lOOfi. Compound and semi-compound grains are present. The latter with 2 to 3 small components at the proximal end, rarely in the center (in this case the grains are small, about 4 times as long and almost as thick as broad). Canna coccinea Ait. Fresh root-stock. — Grains similar to preceding; one-half to fully as broad as long; the broader portion compressed to about half the width; the smaller are terete; no fissures after desiccation. Length IQix. Several semi-compound grains with 2 to 3 hila lying beside one another are present; and also compound grains with components arranged in one row (see type 13). The African arrowToot (of Canna coccinea, according to Berg, Pharma- cognosie, 483) grains are sheath-shaped, ovoid, or elongated, provided with distinct lamella, and hila located below one another. Frequently two grains grow together, which often reach twice the length of the potato starch-grain. Canna lagunensis Lindl. Fresh root-stock. — Grains oval, rounded, oblong, more or less pear-shaped or almost violin-shaped, chiefly irregular, and often somewhat curved or oblique; mostly about one-half as broad as long; the smaller ones terete, the broader ones compressed to about one-half or more of their width; lamellae very numerous, incomplete; hilum frequently indistinct, located at the narrow, sometimes projecting, wart-like end, about one-fortieth to one-seventh eccentric. In many grains an inner lanceolate or spindle-shaped ladder-like group of lamellae extend from the hilum towards the distal end, to more than one-half to two-thirds the length of the grain. In the symmetrical grains it has a median and in the oblique and curved grains a lateral position. Semi-compound grains with two or more part-grains lying in a row are present; in addition there arc compound grains with part- grains in one row (see type 13), and also some which consist of one large and one or two small components. Canna indica var. aureo-vittata Hort. Fresh root-stocks. — Grains oval or oblong; rarely triangular or spindle-shaped; one-third to two-thirds as broad as long; the broadest ones compressed to half their width; lamelliE distinct, almost all of them incomplete; hilum end narrowed, or broad and rounded; hilum about one-twelfth, rarely one-twentieth, eccentric; occasionally a few short fissures radiate from the hilum. Length about 130;^. Semi-compound and com- pound grains, both consisting of 2 to 3 components, are rarely present. Canna cubensis Hort. Fresh root-stock. — Grains as above, sometimes the proximal (hilum) end pro- truding and blunt. Length about 80/x. There are some grains which have a special internal system of lamellae. Canna heliconiwfolia Hort. berol. — Grains distinguished for their very irregular shape; strongly compressed; with lamellae. Length 65/i. Many semi-compound grains are present having a row of components along one margin. Canna edulis Ker. — According to Fritsche (Poggendorf's Annal., 1834, xxxii, 141 ; Taf. ii, 37-39), the grains are much flattened, almost as broad as long; lamellae distinct; several semi-compound grains with 2 or more components. According to Leon Soubeiran (Journ. Phar., 1854, xxv, 179), some of the grains are very small, spherical or oval ; others pear-shaped, circular, rounded or imperfectly triangular; the larger ones strongly compressed; lamellae and hilum distinct. Size 40 to 80 or 90/u. According to the drawings, the grains are twice as long as broad, and have numerous tubercular prominences and depressions on their margins. The hilum is very eccentric, and is found at the narrow end. TYPE 8. GRAINS SIMPLE, ECCENTRIC, CUNEIFORM, OR FLATTENED. 227 Canna glaucn Linn.^ — According to Berg (Pharmacognosie, 481) the grains are very irregular; usually (lisk-sliaped, or iiii(>qually curved on both surfaces; ovoid, quadrangular, cuneiform, crescent-shajietl, or reniform. They are as large as grains of the potato starch; the hilum lies at a jioint which in tlie reniform grains is emarginate, in the quadrangular grains often in the middle, and surromided l\v numerous concentric lamelke. Canna giganlea Desf. — According to Payen (Ann. Sc. Nat., 1838, ii, IG), the grains are pear-, flask-, and retort-.shaped, and attain a length of 175/^; compressed to three-fourths or two-thirds of their width; hilum at the narrow end; lamellae delicate. (See also p. 229.) Canna discolor Lindl. — According to Payen (Ann. Sc. Nat., 1838, 10), the grains are rounded- shield-shape, more or less elongated-conical; hilum at the narrow, often somewhat protrud- ing, end, or sometimes between two more or less marked projections; lamellae numerous. Length about 150/u. Hydrophyllum virginicum. (Ilydrophyllacece.) Dry root-stocks and roots. — Grains rounded-oval to elongated-oval (elliptical), more or less triangular; half as broad as long; two-ninths to two- fifths as thick as long, hilum end narrow and thick; distal end broad, thin, and squared; lamellae none, or indistinct; instead of a hilum, a small cavity is usually present; eccentricity of hilum one-sixth. Length 28^1, width 20(U. Among these grains are some compound forms consisting of 2 to 5 equal or unequal components. Scheeria nwxicana Seem. {Gesneracexe.) Fresh scales of the root-stock. — Grains oval, broadly-conical, rounded-triangular, rarely oblong; half to as broad as long; the broader ones compressed to about half their width; from the narrow aspect they are conical; lamellae usually distinct; the hilum end is narrow and thick; the distal end is broad and squared; eccentricity of hilum about one-ninth. Length about 40yu, width about 33;u. Sdadocalyx imrszewiczii Kegel. {Gesneracea.) Fresh scales of the root-stocks. — Grains rounded, rounded-triangular, or oval; 0.6 to 1.25 times as broad as long; the broader ones compressed to about one-third of their width; from the narrow aspect they are oblong, with both ends equal; lamellae distinct. Length about 50;u, width about 38/i. Isolated semi-compound grains with several hila, and occasionally doublets, are also present. Achimenes hirsuia DC; Locheria hirsula Kegel. {Gesneracew.) Fresh scales of the root-stock.- — Grains oval, elongated- triangular, conical; one-half to three-fourths as broad as long; the broader ones compressed; the narrower ones terete; from the narrow aspect both ends are either alike or the hilum end may taper; lamellae delicate. Length about 50ai. Transitional forms to the rod-shape (type 9). Isolonia vestitmn Benth. (Gesneracecs.) Fresh scales of the root-stocks. — Grains oval, broadly-conical, 3 to 4 angles with rounded corners; frequently somewhat irregular; 0.66 to 1.5 times as broad as long; the broader ones compressed to about one-third or more of their width; lamella usually indistinct. Size about 28yu. Some isolated doublets are also present. Tudcea regelii Heer. (Gesneracece.) Fresh scales of the root-stocks. — Grains oval, usually rounded- triangular or mussel-shell-shaped; 0.66 to 1.5 times as broad as long; compressed to about one-half or more of their width; from the narrow aspect the proximal and distal ends of almost equal thickness; lamellae usually distinct; hilum about one-seventh eccentric. Among them are found semi-compound or wholly compound grains, consisting of 2 to 3 equal or unequal components. Tudcea picta Desne. {Gesneracece.) Fresh scales of the root-stocks. — Grains rounded, oval, triangular, often somewhat irregular and angular; the distal end usually broad, occasionally somewhat oblique; half to as broad as long; the broader ones compressed to about half their width; from the small longitudinal aspect cone-shaped, with thickened hilum end; lamellae numerous, delicate, only the innermost ones complete; hilum about one-fifth to one-ninth eccentric; occasionally several short fissures radiate from the hilum. Length about 50m, width 35/i. Some semi-compound grains with two hila are also present. Trevirania longiflora Kegel. {Gesneracece.) Fresh scales of the root-stocks. — Grains oval, conical, rounded, 3 and 4 angles; half to as broad as long; the broader ones compressed to one-half, rarely one-third, of their width, the narrower ones not at all or slightly compressed; lamellae indistinct; hilum also often indistinct, located at the thicker end (in the broad, compressed grains both ends are frequently of almost equal thickness). Length about 48ai, width 35^. Transitional forms to the cone-shape (type 7). 228 DIFFERENTIATION AND SPECIFICITY OF STARCHES. Saxifraga granulata Linn. (Saxifragacew.) Dry scales of bulbils. — Grains rounded, roundod-tri an- gular to oval, with narrow proximal and broad distal ends; 0.6 to 1.33 as broad as long; the broad ones compressed to half their width ; from the narrow aspect mostly cone-shaped with thickened hilum end; frequently with many distinct lamellae; instead of the hilum there is a small cavity from which in the broad aspect several radial fissures proceed chiefly towards the distal end; in the narrow aspect a longitudinal slit is found. Length about 28ix, width about 25/i. Rammculus aconitifoliiis Linn. (Ranunculacece.) Dry roots. — Grains rounded, rounded-triangular, mostly rounded or reniform, usually twice as long as broad; compressed to about half the short, flattened diameter; one side (in the kidney-shaped grains the concave) thickened, the opposite one thinned, knife-like; hilum approaching the thickened margin; often in place of the hilum there is a small cavity, from which in the broad aspect proceeds several radial fissures or sometimes one transverse fissure (parallel to the proximal margin) ; in the small aspect there is often a distinct longitudinal slit; eccentricity of hilum about one-fourth to one-fifth. Length about 42yu. Some roots are poor, others rich in starch. Ficaria ranunculoides Moench.; Ranunculus ficaria Linn. (Ranuncidacece.) Fresh and dry thickened roots. — Grains rounded, rounded-triangular, and broadly-conical, and occasionally some- what irregular; at the proximal end usually narrowed, and at the distal end broadened; almost as long as broad; one-half to two-thirds as thick as broad; from the narrow aspect of equal thickness throughout or tapering towards the proximal end; lamellae delicate or invisible; hilum one-fourth to one-sixth eccentric, indistinct in fresh grains, usually indicated bj' a small cavity in dry grains. Length about 33/j, width about 30/^. Some semi-compound grains (see type 11); also a few doublets and triplets. Oxalis pentaphylla Sims. (Oxalidacece.) Dry scales of the S7nall bidbs on the root-stock.- — Grains tri- angular vrith blunt angles, mussel-shell-shaped, with 4, 5, or 6 angles, rarely oval ; two-thirds to almost as broad as long; lamellae usually visible, but very delicate, only the innermost ones distinct; hilum often indistinct, one-ninth to one-twelfth eccentric. Length about 60^, width about 58/i. Oxalis lasiandra Zucc. (Oxalidacew.) Dry scales of the small bulb. — Grains rounded or rounded- oval, frequently 3, 4, or 5 angles; sometimes rather irregular; 0.66 to 1.66 as broad as long; compressed to about one-half or more of their breadth; from the broad aspect the hilum end projects into an oblique angle, or into a wart-like outgrowth; from the narrow longitudinal aspect both ends alike, or the proximal end thicker than the distal end; lamellce delicate or indistinct; frequently instead of the one-eighth eccentric hilum a small cavity is observed with two short, delicate fissures, which usually resemble a curved cross-fissure. Size about 28^. Oxalis crenata Jacq. {Oxalidacece.) Bulbs. — According to Payen (Ann. Sc. Nat., 1838, ii, 17, pi. 6, fig. 3), the grains are cylindrical or somewhat conical in shape; over twice as long as broad; lamellae distinct; hilum about one-third eccentric. Size about 100^. A few semi-compound grains with double hila, and isolated doublets, are present. From the description and drawings this type can not be distinguished with certainty. Probably the compressed grains belong here (otherwise they hold a medium position between the inverted-cone-shaped and rod-shaped types). Rhizophora mangle Linn. (Rhizophoraceoe.) Dry radicle. — Grains oval, conical, triangular, quad- rangular, or of irregular form; half to as broad as long; the broader ones compressed to one- half or over; in the broad aspect, the proximal end is occasionally narrowed, the distal end broadened; on the narrow aspect, often conical with thickened proximal end and tapering distal end; in the narrow aspect a longitudinal sht is frequently present; sometimes there is a small cavity instead of the hilum, about one-fourth eccentric. Length to 24 to 28/i, width to 20/i. The structure of most grains is uncertain. Some belong here on account of their cuneiform appearance; others seem to approach the conical type (type 7). Globba marantina Linn. {Zingiber acece.) Dry seed. — Grains rounded-oval to oblong, narrowed chiefly towards one end which projects in a small papilla, frequently unsymmetrical or arched, besides many irregular and angular forms, mostly 3, 4, and 5 angles; 0.33 to 1.33 as broad as long; the broader ones compressed to about one-third and one-fifth their width (to 4 and 5 times as long as thick); from the narrow aspect, the grains are elongated conical; without TYPE 0. GRAINS SIMPLE, ECCENTRIC, ROD-SHAPED. 229 lamellae, hilum, and fissures. Length to 30/j, width to 25m, thickness to 7 and 8n. The struc- ture of these grains is doubtful. No hilum was plainly discovered after the grain had been roasted to a yellowish color, boiled in water, and treated with sulphuric acid. Several times, however, Nageli thought lie detected a small cavity near the narrow end. He states that if the grains belong here, the hilum must have a very eccentric position. Carina gigantea Desf.; Canna indica,Liini. {Cannacece.) Dry seed. — Grains scale-Hke, rounded-oval to elongated, usually more or less irregular, often with crenate margin; one-third to as broad as long; very strongly compressed; usually homogeneous, without lamellae, cavity, or fissures, but occasionally with one or more refractive bands in the median line. Length about 18 to 21yu, thickness L5 to 2/;. The fresh grains of Canna indica are like those just described, somewhat undulating at the margins, entirely homogeneous. Length about 21//, thickness 3 to 4/i. Nothing can be seen of a hilum in the fresh or dried starch-grains. After slight roasting, a small rounded cavity or gas-bubble very near one end is occasionally observed. This is probably the position of the hilum, the eccentricity of which is one-sixth and over. Nevertheless this description is not wholly safe, since sometimes in grains more thoroughly roasted two, three, or four such hollow spaces may be observed. (See also p. 227.) Maranta ramosissima Wall. {Marantaceoe.) Dry seed. — Grains rounded-oval to oblong, frequently somewhat broader towards one end, mostly more or less irregular and tuberculated; two- thirds to almost as broad as long; the broader ones compressed to about one-third of their width (about 3 or 4 times as long as thick), from the narrow aspect frequently with a delicate longitudinal slit; almost homogeneous. Length about 16^, breadth about \Qti, thickness 3 to 4 //. Nageli states that he places these grains here, in which neither lamellae nor hilum are visible, on account of their analogj' with the canna. Among these are found a few distinctly compound grains with numerous coalesced components, similar to those of Maranta sp. (see type 12). It is uncertain whether the indented forms are not accidentally compound; from the narrow longitudinal aspect many of these same forms are also slightly torulose. Type 9. Grains Simple, Eccentric, Rod-shaped. Lamellae coarser and more numerous on one side, finer and less numerous on the diametrically opposite one. Grains mostly elongated, terete, or somewhat compressed; at both ends of almost equal breadth and thickness. On drying, fissures are observed radiating from the hilum, chiefly in the direction of the longitudinal axis. In the terete grains two longitudinal slits are noticeable, sometimes crossing each other at right angles; in those which are compressed there is a slit which coincides with the greatest plane. There are many transition forms to the conical (type 7), and to the inverted-conical (type 6). Angiopteris. (Marattiacece.) Stems of leaves (fronds). — According to Harting (Rech. sur. I'anat. rOrganogenis et I'Histiogenia du genre Angiopteris, plate vii, 8, 9), the grains are oval to cyhndrical, the longer ones 4 times as long as broad; distinct, incomplete, lamellae; hilum very eccentric; the hilum end occasionally narrowed, but more frequently somewhat broader than the distal end. Length about 70/u. Hemerocallis fulva Linn. (Liliacece.) Fresh root-stocks. — Grains usually elongated-oval, 1.5 to 2.5 times as long as broad; the hilum end somewhat smaller, and almost circular in transverse section, the distal end rather broader and compressed; no lamellae; hilum about one-sixth to one-eighth eccentric. Length about 25//. These forms resemble the cuneiform type. Tamus communis Linn. (DioscoreacecE.) Dry root-stock. — Grains of very varied and very irregular shape, rounded to elongated-conical, and rod-shape; about 6 times as long as thick; the broader ones somewhat compressed; lamellae usually distinct, delicate, and numerous; hilum often invisible, can sometimes be recognized by means of a small cavity; hilum end frequently thickened; the distal end of equal width or even (in the oval and rounded-tri- angular grains) broadened and squared. Length about 52/i. This type is not well pronounced. There are many transitional forms to the cuneiform type (type 8). Among these are grains with an internal ladder-like system of coarse lamella. Furthermore, there are semi-com- pound grains with two inner large components, and compound grains of two to four compo- nents. Moreover, the grains in the course of solution exhibit very diverse and peculiar forms. 230 DIFFERENTIATION AND SPECIFICITY OF STARCHES. Vallisneria spiralis Linn. (Hydrocharilacece.) Dry root-stocks and runners. — Grains oval, elongated, conical, more rarely triangular with rounded corners, or pear-shaped; mostly 1.5 to twice as long as thick, lamellse frequently distinct, all incomplete; instead of the hilum a small cavity with short radial fissures is observed; hilum end rarely thiimer, sometimes of equal thickness, usually thicker than the distal end; the latter is sometimes broadened, thinned and squared; eccentricity about one-seventh. Length about 35/j. 7ns sambucina Linn. {Iridacem.) Fresh root-stock. — Grains are elongated or oval, frequently 3 to 4 angles, with protruding rounded angles; a little longer than 3 times as long as broad; slightly compressed; one-half to two-thirds as thick as broad; lamellae delicate, rarely visible; sometimes several (2 to 5) radiating fissures emerging from the hilum. Length about 27/i, width about 16/i. The type of these grains, like those of the other species of Iris, is not clearly defined. Transition forms to the other eccentric types are numerous. Iris florentina Linn. (Iridacece.) Dry root-stock. — Grains spherically-oval, oval, rarely oblong; often narrowed at one end; circular in transverse section; a little longer than twice as long as thick; frequently several (3 to 5) distinct lamella on the side of the long radius; frequently instead of the hilum a very small cavity is observed, from which a few radial fissures rarely emerge; about one-fourth to one-sixth eccentric. Length about 25/i, thick- ness about I9fi. Transitional forms to the other eccentric types. Several doublets are present. Iris pallida Lam. (Iridacece.) Fresh root-stock. — Grains spherical, rounded-triangular, oval, or conical, usually slightly compressed in transverse section; a little longer to scarcely twice as long as broad; lamellse delicate and rarely visible; usually instead of the hilum a small cavity with 2 to 5 radiating fissures is observed. Length about 19/i, width about 13ju. Transitional forms to the other eccentric types. Several doublets are present. Himantoglossum hircinum Rich. (Orchidaceas.) Dry bulbs. — Grains oval, rod-shaped, conical, rarely rounded or rounded-triangular; most of them circular in transverse section; the broadest ones slightly compressed (about half as thick as long); lamellae rather numerous, distinct and incomplete; sometimes instead of the hilum a small cavity is observed from which several very short and delicate fissures radiate; about one-eighth eccentric; both ends usually of equal thickness, occasionally the hilum end is thicker or thinner; the distal end often shortened, and if broader than the proximal end, it is slightly compressed. Length about 35;u. Some grains have an almost central hilum; others are semi-compound, and with two components. Schacht (Microscon, 2 Aufl., pag. 48, fig. 2, / and g) shows drawings of two starch-graina from the bulbs of Himantoglossum, unlike any that Nageli saw; most likely, NageU states, they are in process of solution. Hedychium gardnerianum Wall. {Zingiber acece.) Root-stocks. — According to the description and drawings of Bischoff (Bot. Zeit., 1844, 388) the grains are 2 to 4 times as long as broad; often distorted or sharply indented; both ends alike or thickened at one end, either clavi- form or capitate; plainly striated or repeatedly constricted. Length about 50;u. All the grains that are illustrated are probably about to pass into solution. This view is supported by the illustrations, which Fritzche (Poggendorff's Annal., xxxii, Taf. ii, figs. 50 to 53) gives of the grains from the root-stocks of Hedychium hirsutum in process of solution. Alpinia galanga Swartz. {Zingiber acece.) Dry root-stocks. — Grains lanceolate, conical, or rod-shaped, frequently somewhat irregular; mostly 2.5 to 4 times as long, occasionally a little longer than broad; the narrower ones terete, the broader ones slightly compressed; lamellae delicate or indistinct, only the innermost ones complete; usually instead of the hilum, a small cavity is observed; one-seventh to one-seventeenth eccentric; the hilum end often thicker; occa- sionally the distal end broadened and also thinned and squared. Length to 55/i, width to 25/i. Several grains show traces of changes by moist heat. Among these isolated doublets are rarely found. Costus speciosus Smith. {Zingiberacece.) Root-stock. — According to Fritzche (Poggendorff's Annal., 1834, XXXII, Taf. ii, fig. 32), the grains are almost cylindrical, about 4 times as long as thick; with distinct lamellae; hilum about one-eighth eccentric. Costus spiralis Rose; Costus comosus Rose. {Zingiberacece.) — According to H. Criiger (Bot. Zeit., 1854, Taf. II, figs. 20 and 23), the grains are as above, but somewhat narrower at the hilum end. TYPE 0. GRAINS SIMPLE, ECCENTRIC, ROD-SHAPED. 231 Musa paradisiaca Linn. {Musaccw.) Fruit. — According to H. Criiger (Bot. Zeit., 1854, Taf. ii, fig. 1), the grains are oval and elongated (in the interior of the fruit) to linear (in the rind of the fruit); 1.5 to 10 times as long as broad; with distinct lamellaj; hilum apparently at the thicker end, one-half to almost one-eleventh eccentric. Caladium seguinum Vent.; Dieffenhachia seguine Schott. (Aroidece.) Root-stock. — According to Schleiden (Grundzuge, 3. Aufl., 1 fig., 2d on p. 184) and H. Criiger (Bot. Zeit., 1854, pi. ii, figs. 8 and 15) the grains are rod-shaped; 2 to 10 times as long as thick; frequently with a lateral wart-like or lobe-like process formed by an outer system of lamelloe which is superimposed at right angk's to the longitudinal axis; with distinct lamellse; hilum end sometimes slightly thicker, sometimes slightly thinner, than the distal end; eccentricity about one-sixth. Philodendron grandifolium Schott. {Aroidew.) Root-stock. — According to H. Criiger (Bot. Zeit., 1854, Taf. 11, figs. 7 and 13), the grains are elongated to lanceolate, or rounded-triangular (and then probably compressed); sometimes bilobate; 3 to 4 times as long as broad; with distinct lamellae; hilum at the broader end, about one-sixth or more eccentric. Achimenes tubiflora Nicholson ; DolicJiodeira tubiflora Hanst. ; Gloxinia tubiflora Hook. (Gesneriacece.) Fresh bulbs. — Grains rod-shaped, the smaller ones oval or conical; many with one (rarely with two) large or small lateral wart-like appendage, which is found occasionally at the proximal end, sometimes in the middle, but chiefly at the distal end; about 5 times as long as thick; terete; lamellae numerous, distinct, almost all incomplete; instead of the hilum a small cavity with one small transverse fissure, or with two short fissures directed towards the distal end ; about one- eighteenth eccentric ; two ends usually alike in thickness, the hilum end rarely somewhat thicker (in the small grains, as a rule, the hilum lies in the thicker end). Length about Q6iJ., thickness without the appendage about 16/n, but with the appendage almost double the size. Guthnickia atrosanguinea Regel. (Gesneriacece.) Fresh scales of the root-stock. — Grains rounded-oval to oval-cuneiform, and cylindrical-oval; about 2.5 times as long as broad; circular in transverse section; the broadest end slightly compressed; lamellae distinct; frequently an external lateral system of lamellae; both ends of nearly equal thickness; the hilum end occasionally somewhat thicker or thinner than the distal end. Length about 35ai, width about 22/^. Among these are some semi-compound and some completely compound grains of two components. Plectoponia fimbriatum Hanst. (Gesneriacece.) Fresh scales of the root-stock. — Grains usually oval, less frequently triangular-oval; the broader ones slightly compressed, the narrower terete; both ends of almost equal thickness. Length about 28;u. Seemannia temifolia Regel. (Gesneriacece.) Fresh scales of the root-stocks. — Grains mostly elongated- oval or cylindrical-oval; more rarely conical or somewhat irregular; usually 2 to 3 times as long as broad; terete; the broadest ones slightly compressed; lamellae usually distinct; often a lateral wart-like process consisting of a separate system of lamellae (as in Dolichodeira) ; the two ends are usually of nearly equal thickness; although the hilum end may be either thicker or thinner than the distal end, which latter is sometimes broadened into a blunt, knife-like edge. Length about bOfx, thickness about dO/x. Dentaria enneaphyllos Linn. (Brassicacece; Cruciferce.) Dry scales of the root-stock. — Grains oval, conical, oblong, oval-triangular, frequently somewhat irregular; 1.5 to 3 times as long as broad; the broader ones compressed to about half their width, the narrower terete; without lamellae; instead of the hilum a small cavity is often observed from which a dehcate fissure may radiate; the hilum end frequently thicker, sometimes of equal thickness, rarely thinner than the distal end. Length about 42^, width about 25/i. Krameria triandra R. P. (Polygalacece.) Dry roots. — Grains mostly rod-shaped or oval, sometimes conical or club-shaped, frequently arched or even irregular; also with single knob or lobe- like processes; terete; about 5 times as long as thick; the broader ones slightly compressed; about 3 times as long as broad, and 5 times as long as thick; lamellae rare and indistinct; occasionally instead of the hilum a small cavity is observed from which short fissures occa- sionally radiate; about one-eighth eccentric; hilum end either narrower or broader than the distal end. Length about 42/j. Among these are compound grains of few and mostly equal- sized components (see type 14). Also rounded simple grains are often found, and indeed sometimes almost exclusively, while in some parts of the tissues the oblong grains and in still other parts the compound grains appear in greater numbers. Many of the apparently round, simple grains on closer investigation proved to be separated-grains. 232 DIFFERENTIATION AND SPECIFICITY OF STARCHES. Colocasia odora Brongn. (Aroidea.) Dry seed. — Grains rounded to oval-conical, mostly somewhat irregular; half to just as broad as long; the narrower forms not compressed; the broader ones compressed to about half their width; no lamella; instead of the hilum there appears fre- quently a small round cavity from which a single fissure may emerge, more often a slit-like cavity, which from the broad aspect appears delicate and from the narrow aspect more clearly defined; hilum end in the small forms is broader than the distal end; eccentricity about one-third and one-fourth. Length about 22 to 26m, width about 20/j. The structure is not very clearly defined, frequently approaching the conical type (type 7) and the cuneiform or compressed type {type 8). Doublets and triplets are also found. Mammea americana Liim. {Guttiferm.) Dry cotyledons. — Grains mostly oval or elongated-oval, rarely almost rounded or conical; usually twice as long as broad; circular in transverse sec- tion; the broadest and at the same time somewhat irregular forms are seldom slightly com- pressed; no lamellae; instead of the hilum a small cavity is usually observed, generally with a cross-fissure and several radial fissures ; ordinarily the grains are of almost equal width through- out their entire length ; now and then narrowed at either the proximal or distal end, or even at both ends; eccentricity mostly one-third to one-fourth (seldom one-fifth and one-sixth). Length about 55^, thickness about 35yn. Rich in starch; poor in oil. Type 10. Grains Simple, Structure Obscure. Nageli places here all the species of starch, the structure of which, mostly on account of their diminutive size, is either not fully developed or not identified, and hence remains doubtful as to which of the foregoing types they belong. The grains, as a rule, are undoubtedly simple. Now and then some separated-grains which are no longer recognized as such may occur. Lamellae, hilum, cavities, fissures, and clefts are seldom mentioned, because they have not often been demonstrable. Vaucheria tuberosa A. Braun. (Algce.) Dry tuberous swollen ends of root-like organs. — Grains (all simple) oval to elongated, lanceolate, mostly somewhat unequal laterally, or elliptical and constricted in the middle, reniform, frequently somewhat irregular; 1.5 to 3.5 times as long as broad; the broadest ones compressed to about half the width ; sometimes from the narrow aspect a deli- cate longitudinal slit is observed. Length about 13/i, width about 7ju, thickness 2.5 to 3.5m. Equisetum hyemale Linn. {Equisetacece.) You7ig tubers. — According to Leon Soubeiran (Journ. Pharm., 1854, xxv, 182) the smallest grains are spherical, the larger ones elongated, and the largest ones pear-shaped; there are many very irregular forms with dentate projections on the circumference; lamellae and hiliun rare and indistinct. Length 50 to 60m, width 10 to 30m. Olfersia undulata Presl. {Filidnem.) Dry root-stock. — Grains rounded or rounded-triangular, occa- sionally oval, not at all or very slightly compressed, many with a small or a large cavity. Size about 10m. Poly-podium vulgare Linn. ; Radix polypodii. (Proteacem.) Dry root-stocks.— Grains rounded; rounded- triangular, oval, frequently somewhat irregular; compressed to about half or more of their width; many are thickened at the more convex margin or at one angle of the triangle, and at the opposite broad margin thinned and squared ; without lamellae, hilum, or cavity (even after roasting and boiling, the hilum is not visible). Size about 9 to I 1m. Many of the grains agree closely in their form with those of the cuneiform type (type 8). Polypodium distans Kaulf. {Proteacece.) Dry root-stock.- — Grains rounded or rounded-triangular, rarely oval; compressed to about one-third of their width; elongated from the narrow aspect. Size about 14m. Adiantum capillus-veneris Linn. (Filidnece.) Dry root-stocks. — Grains rounded to oblong, mostly oval, sometimes somewhat irregular; narrow ones little or scarcely compressed, the broader ones to about half their width; from the broad aspect indistinct fissures are occasionally observed; from the small aspect the grains are elongated-oval, and generally have a distinct longitudinal slit. Length about 30m, width about 20m. Iu spite of the size of these grains, it is doubtful to which type they belong. In form and median cleft they resemble the centric- oval and centric-lanceolate types. Asplenium vriarinum Linn. {Filicinece.) Dry root-stock. — Grains rounded or oval, often somewhat irregular; the broad ones compressed to about half their width; the majority have a large cavity. Size about 1 1m. Isolated doublets are .n,lso present. TYPE 10. GRAINS SIMPLE, STRUCTURE OBSCURE. 233 Scolopendrium officinarum Swartz. {FilimiecE.) Fresh root-stock. — Grains circular to oval, some- times rather irregular; compressed to about half their width. Size 9/u. Diplaziwn plantagineuyn Swartz. {Filidnece.) Dry root-stock. — Grains rounded or oval, most of them only slightly compressed, the broad ones rarely to about half their width. Size about 15^1. Polystichum thelypteris Roth. {Filicineoe..) Dry root-stock. — Grains rounded to oval, sometimes rather irregular, rarely somewhat angular; the broad ones compressed to about one-third and one-fourth of their width; from the narrow aspect elliptically oblong and elongated lanceolate; both ends of equal thickness; sometimes with one longitudinal cleft. Size about 17^. These grains, like those from the root-stocks of several other Filicineoe, resemble very closely those of the centric-lenticular type. Aspidium filix-mas Swartz.; Radix filicis. {Filicineoe.) Dry root-stocks. — Grains rounded to oblong, rounded-triangular, reniform, often somewhat irregular ; the broad ones compressed to about one-half and one-third of their width; from the narrow aspect sometimes with a longitudinal cleft. Size about 19/i. A very few compound grains consisting of 2 to 3 components are also observed. Cystopteris fragilis Bernh. (Filicineoe.) Dry root-stocks. — Grains rounded, oval, pear-shaped, fre- quently irregular and somewhat angular; the broad ones compressed to about half their width; somewhat thicker at one end, and thinned and squared at the opposite end. Size about 9n. Cystopteris bulbifera Bernh. {Filicineoe.) Dry tuber-like bulbils on the /ronds.— Grains rounded or rounded-oval, rarely somewhat angular, as the result of pressure; compressed to about half their width; either equally thick throughout or thicker on one side. Size about 7ju. Ophioglossum vulgatum Linn. {Filicineoe.) Dry root-stocks and roots. — Grains rounded or oval, frequently somewhat irregular; the broad ones compressed to about half their width; from the narrow aspect elongate or oval; often a longitudinal slit. Size about 12/j. Starch and oil present in the root-stocks; starch only in the roots. Botrychium Lunaria Swartz. {Filicineoe.) Dry root-stock and root. — Grains rounded, the larger ones somewhat hollow and compressed. Size about 7/i. No oil in the root-stock. Molinia coerulea Moench. {Graminacece.) Dry roots. — Grains oval, oblong, conical, rarely rounded- oval; broad ones distinctly compressed. Size about S/x. Carex arenaria Linn. {Cypcracece.) Dry root-stocks. — Grains rounded, oval, rounded pear-shaped, sometimes slightly compressed; rarely with a small cavity. Size about 12yu. Among these are observed compound grains of few, unequal components (see type 15). Starch quite plentiful. Carex hirta Linn. Dr^/.^Grains similar to those in the preceding species, rounded to oval, rounded pear-shaped; the broadest ones compressed at about half their width. Size about 12;^. Among these are observed compound grains of few, unequal components. (See type 15.) Starch plentiful. Carex intermedia Good.; Carex distichia Huds. Dry. — Grains roimded to oval; broad ones slightly compressed, the larger ones with a cavity. Size about 8^. Starch not plentiful. Carex atrata Linn. Dry. — Grains rounded. Size 1.5 to 3/i. Some of them are probably separated- grains. Starch not plentiful, found in rather thick-walled cells. Carexbicolor AW. Dry. — Grains rounded to oval, rounded pear-shaped. Size about 5/i. Among these grains are observed some compound grains of few mostly equal components. (See type 14.) Cyperus polystachyos Rottb. {Cyperaceoe.) Dry root-stocks. — Grains spherical to oval and oval pear-shaped. Size about 8//. Among these grains are observed some doublets and triplets. Starch not plentiful. Abilgaardia monostachya Vahl. {Cyperaceoe.) Dry root-stocks. — Grains rounded to oval, and oval pear-shaped; a small cavity in the larger ones. Size about 9/i. Among these grains are found compound grains of few equal and unequal components. (See type 14.) Scirpus triqueter Linn. {Cyperaceoe.) Dry root-stock. — Grains rounded, frequently triangular with rounded corners, oval, conical, and frequently somewhat irregular; the broad ones com- pressed to one-third their width ; many with a small cavity. Size about 18/i. In some conical grains the cavity is at the thicker hilum end, and in some triangular forms the distal end is thinned and squared, so that the type seems to belong now to the eccentric-conical, now to the cuneiform. In most of the grains no hilum is visible. Starch plentiful. 234 DIFFERENTIATION AND SPECIFICITY OF STARCHES. Scirpus pungens Vahl.; Scirpus rothii Hoppe. (Cyperacecs.) Dry root-stock. — Grains similar to the preceding; usually round, and frequently with 1 or 2 small wart-like or angular processes; the broad ones compressed to one-third or more of their width. Size about ll/u- Eriophorum capitatum Host.; Eriophorum scheuchzeri Hoppe. {Cyperacece.) Dry root-stock. — Grains roimded pear-shaped, oval, and triangular. Size about 7/i. Some doublets and triplets are also observed. Triglochin maritimum Linn. (Naiadaceae.) Dry root-slock. — Grains spherical, oval, rounded pear- shaped, frequently with a small cavity. Size about 8^. Starch quite plentiful. Scheuchzeria palusiris Linn. {Naiadacece.) Dry root-stock. — Grains rounded or oval, frequently somewhat irregular. Size about 4;u. Starch quite plentiful. Alisma plantago Linn. (AKsmaceai.) Fresh root-stock. — Grains rounded, rounded-oval, sometimes rather irregular. Size about 7fi. Among these are large separated-grains with 1 to 3 pres- sure facets. Size about 2 to 5/i. Starch and oil plentiful in the upper part of the root-stock; n J starch in the lower part. Butomus umbellatus Linn. (Alismacem.) Dry root-stock. — Grains rounded and triangular with rounded corners to oblong and cuneiform, frequently more or less irregular; compressed to one-third or more of their width; frequently from the broad aspect a cavity is observed, and from the narrow aspect a longitudinal slit. Size about 12;u. Starch quite plentiful. Juncus bulbosus Linn. ; Jimcus compressus Jacq. (Juncacew.) Dry root and root-stocks. — Grains usu- ally rounded or slightly angular, rarely oval or rounded pear-shaped; compressed to half their width; many have a small cavity. Size about 7/i. Not much starch in the root-stock; more in the roots. Juncus balticus Dethard. {Juncacew.) Dry root-stock. — Grains rounded to oval and pear-shaped. Size about 7fi. Among these grains are compound grains of 2 to 5 equal or unequal compo- nents. Starch not very plentiful. Luzula spadicea DC. (JuncacecE.) Dry root-stock. — Grains rounded or oval, frequently somewhat angular. Size about 8fi. Among these are observed compound grains of few usually equal components (see type 14). Starch not plentiful. Tofieldia calyculata Whalenb. (Liliaceoe.) Dry root-stock. — Grains rounded to oval, rounded pear- shaped; many slightly compressed; in the larger ones a small cavity is frequently observed. Size about Gfi. Veratrum album Linn. (Liliacece.) Dry root-stock.- — Grains rounded or, when crowded in the cell, somewhat angular; usually there is a large or a small central cavity from which single fissures sometimes radiate. Size about 13 to 15/i. Among these are found compound grains of 2 to 4 usually equal components. Possibly the greater part of the starch is found in the compound grains. In the root fibers there are grains of equal to somewhat larger size, and having the form of compound grains of type 15. Bulbocodium vernum Linn. {LiliacecB.) Dry corm.- — Grains spherical, oval, or rounded pear-shaped. Size 6 to 8m. Among these are compound grains of few usually unequal components. (See type 15.) Gagea stenopetala Rchb. (Liliacem.) Scales of dry bulbs. — Grains spherical, nearly all of them flat- tened on one side; instead of a hilum, usually a small cavity is observed from which radial fissures emerge; hilum usually either near the middle of the flattened side, or almost central in the completely spherical forms; frequently a curved slit is observed running parallel to the flattened margin, the curved portion of which is directed towards the squared edge. Size 17 to 26m. The grains look like separated-grains, resembling a segment of a circle. Gagea lutea Schult. {Liliacece.) Scales of dry bulbs. — Grains spherical or oval-spherical, frequently somewhat angular or irregular; a small almost central cavity and radial fissiu-es; lamellae none, or very delicate, and concentric. Size about 22^. Some compound grains of few un- equal components are also present. (See type 15.) Narthecium ossifragum Huds. {Liliacece.) Dry root-stock. — Grains rounded to oval and pear-shaped, frequently somewhat irregular. Size about 17^. Compound grains with few, usually un- equal components are also present. (See type 15.) Smilax china Linn. ; Radix china:. {Liliacece.) Dry root-stocks.- — Grains more or less angular, as a result of pressure, frequently polyhedral with sharp angles and edges, also with flat surfaces; isodiametric or longer in one diameter; instead of a hilum a small central cavity is observed TYPE 10. GRAINS SIMPLE, STRUCTURE OBSCURE. 235 from which a ftnv fissures radiate; lami'lUe rarelj' distinct. Size 40 to 50/i. The grains fill the cells. Originallj' the grains are almost spherical, but become angular as a result of pres- sure. In some grains the surface is flattened only here and there. Among these grains are some compound grains of few equal components. (See type 14.) Conoslylis involucrala End!. {Hwmodoracem.) Dry roots. — Grains rounded, angular with rounded comers, or sharply polyhedral; sometimes they have a small central cavity from which single fissures rarely radiate. Size about lift. Grains completely fill the cell and become angular as a result of pressure. Galanthus plicatus Bierbst. (Ainaryllidacece.) Scales of dry bulbs. — Grains rounded, oval, reniform, frequently somewhat irregular; compressed to one-half or over of their width; from the narrow aspect a longitudinal slit is frequently observed. Size about 42jli. Hilum is usually central. Sternbergia lutea Ker. (Amaryllidaceoe.) Scales of dry Mdbs. — Grains rounded, oval, reniform, rounded-conical, often somewhat irregular and with blunt angles; compressed to about half their width; instead of the hilum there is a small central or about one-fourth eccentric cavity, from which radiates some very distinct usually cruciform fissures; from the narrow aspect a longitudinal cleft is observed; sometimes isolated lamellse are present. Length about 63;^, width to 40/1. Among these are found equal or unequally divided doublets. In fresh bulbs from the Botanical Garden in Zurich (the starch above described is from a plant at Ceph- alonia) the grains are smaller (about 38;u in size), rounded-reniform, and rounded-triangular to oblong, frequently more or less irregular; they have a small central or slightly eccentric cavity and very short fissures. Narcissus poelicus Linn. {A maryllidacece.) Scales from fresh bulbs. — Grains rounded, with triangu- lar, rhomboidal, and quadrangular outlines; oval, reniform, or oval-cuneiform, frequently more or less irregular; two-fifths to as broad as long; the broad ones compressed to half or more of their width; lamellse and hilum are distinct in only a few of the larger grains. Size about 21/z. Some of the grains undoubtedly belong to the eccentric, cuneiform type (type 8) ; they are broader than long; the hilum is one-sixth eccentric. Among the grains above men- tioned there are a few compound grains of 2 to 4 components. Ccelogyne fimhriata Lindl. {Orchidacecc.) Fresh pseudo-tubers. — Grains usually spherical, without lamellse; frequently with a small cavity from which a few (1 to 4) short or long fissures radiate. Size about 27ix. Orchis mascula Linn. {Orchidacece .) Fresh tubers. — Grains spherical, rounded-triangular, or rounded- conical; the broad ones slightly compressed; hilum rarely visible, located at the broader end, two-fifths to two-sevenths eccentric; lamella? none, or unilateral, at the narrow end. Size about 23^. Among the above are some compound grains of 2 to 5 components, and also isolated semi-compound ones. Orchis globosa Linn. Dry. — Grains are spherical, and have a small central cavity. Size about 14^. Also some compound grains of few components are observed. Starch quite plentiful. Orchis militaris Linn. Fresh 7iot fully mature tubers. — Grains spherical. Size about 12ju. Doublets are also rarely observed. Orchis latifolia Linn. — According to Payen (Ann. Sc. Nat., 1838, ii, pi. 6, fig. 19) the grains are spherical, oval, or conical; two-thirds to twice as long as broad; lamellie distinct; hilum at the thicker end, one-fourth eccentric. Size about 45/i. Among the above are some semi- compound grains of two inclosed components. Platanthera bifolia Rich. {Orchidacece.) Tubers. — According to Payen (Ann. Sc. Nat., 1838, ii, pi. 6, fig. 18) the grains are almost spherical, oval, conical, broadly triangular; one-half to almost twice as long as broad; lamella distinct; hilum about one-fourth eccentric, in the elongated forms located at the thick end. Size about 45;u. The broad grains appear to increase in width at the distal end by means of two prominent angles bounding the less dense and squared edge. Calla palustris Linn. (Aroidece.) Dry root-stock. — Grains spherical or oval; the larger ones occasion- ally with a small cavity or cleft. Size about 0 to 8/x. Among the above are some doublets, usually of halves. Acorus calamus Linn. (Aroidece.) Dry root-stock. — Grains rounded, angular with rounded corners, oval or oval-conical; one-half to almost as thick as long. Length about 14yu, thickness about 9m. Among the above are some compound grains of 2 to 3 equal or unequal components. 236 r>IFFERENTIATION AND SPECIFICITY OF STARCHES. Dorstcnia hrasiliemis Linn.; Radix contrajervce Moroen. (Arlocarpacew.) Dry rool-siack. — Grains spherical, usually more or less angular; the larger ones have a small central cavity. Size 6 to 7m- The cells are filled with starch. Some of the grains are simple; the majority, how- ever, are the separated-grains of compound ones (see type 16). Parietaria diffusa Mert. and Koch. (Urticacece.) Dry root-stock. — Grains spherical or rounded- oval; the larger ones hollow; Size about 7fi. Among the above are some compound grains of few, equal components (see type 14). Generally poor in starch. Polygonacece. Underground parts. — Since the seeds of Polygonacece contain centric starch-grains, it was desirable to examine the structure of the grains from other parts of the plant. Although a number of roots and root-stocks were examined, the question could not be definitely decided. In the analogous structure and in the frequent occurrence of a similar median slit, the grains resemble those in the seeds of Leguminosce. They belong, however, to another type, as is particularly illustrated by the compound grains. Very often small particles are cut off from simple grains; and the large components sometimes show a distinctly eccentric, spherical hiliun. Many grains have a swollen appearance. Polygonum viviparurn Linn. {Polygonacece.) Dry root-slock. — Grains rounded, rounded-triangular, oval, oblong, elliptical, pear-shaped, or reniform, frequently irregular, occasionally somewhat curved, and of ten with protruding corners or wart-hke process; broad ones compressed; from the narrow aspect they have a central cavity or a distinct longitudinal slit. Size about 14/i. Among the above are some doublets. Polygonum bistorta Linn.; Radix bistortce. (Polygonacece.) Dry root-stock. — Grains rounded, rounded- triangular ; oval, reniform, pear-shaped, rod-shaped or irregular, occasionally curved, frequently with papillary protuberances on the circumference; the broad ones compressed; one-quarter to as broad as long; many have a longitudinal slit or a large central cavity. Length about 17/u, width 11/i. Among the above are some doublets consisting of frequently elongated, equal halves. The simple and the compound grains resemble those of Polygonum viviparum. Polygonum alpinum All. (Polygonacece.) Dry rootlets of young shoots. — Grains rounded to oblong, elliptical, sometimes reniform or irregular; the broad ones compressed; the larger ones have a cavity. Size about 9/i. Little starch. Polygonum aviculare Linn. (Polygonacece.) Dry root. — Grains spherical or rounded-oval; the larger ones with a cavity. Size about Ifi. Polygonum convolvulus Linn. (Polygonacece.) Dry root. — Grains spherical to oval. Size about 8/t. Poor in starch. Rumex obtusifolius; Radix lapathi acuti. (Polygonacece.) Dry root. — Grains oval, rounded-triangular, elliptical, lanceolate or lanceolate spindle-shaped; one margin frequently more convex than the other, or even somewhat curved; one-third to as broad as long; not at all or slightly compressed; some grains have a longitudinal slit. Length about 16 to 20/i, width 11//. Lamellae and hilum invisible; the latter is probably toward the thicker end. Rumex sanguineus Linn. (Polygonacece.) Dry root.- — Grains rounded to lanceolate with either blunt or pointed ends; frequently with one strongly convex and one straight or slightly concave margin; sometimes triangular and irregular; 5 to 6 times as long as broad; the broad ones slightly compressed; from the narrow aspect there is a distinct longitudinal slit. Size 20 to 24yu. Among the above some doublets are observed. Rumex crispus Limi. (Polygonacece.) Dry root. — Grains oval, oblong, elliptical, pear-shaped, very often triangular or somewhat reniform with elevated ridges, often irregular, with either blunt or pointed angles; the broad ones distinctly compressed; from the narrow aspect the majority have a longitudinal slit. Size about 20 to 32/i. Rumex acetosa Linn. (Polygonacece.) Dry root-stock. — Grains spherical to oblong and conical; the larger ones with a cavity nearly the same shape as the grains. Size about lO^i. Among the above are compound grains of few, usually unequal, components (see type 15). Rumex maritimus Linn. (Polygonacece.) Dry root. — Grains rounded, oval, oblong, conical, the broad ones slightly compressed; some grains have a narrow slit-like cavity. Among the above are some compomad grains of few equal or imequal components (see type 15). Rumex arifolius All. (Polygonacece.) Dry root-stock. — Grains spherical or rounded-oval, sometimes slightly angular; the larger ones have a central cavity. Size about 9ju. Also some compound grains of few equal or unequal components (see type 14). TYPE 10. GRAINS SIMPLE, STRUCTURE OBSCURE. 237 Aristolochia clcmatitis Lam. {Arislolochiaceoe.) Fresh root-stock. — Grains usually spherical; also some lamella;; hilum usually indistinct, about one-third eccentric. Size about 11//. Also some compoimd grains of 2 to 4 equal components. Size about lOfi. Asarum europceum Linn. (Aristolochiacece.) Dry stolons. — Grains spherical or rounded-oval. Size about G/x. Also some compound grains of 2 to 4, rarely 6, equal components. Size about ll/j. Moderate amount of starch; oil plentiful. Plantago maritima Linn. (PlantaginacecB.) Dry root-stock. — Grains spherical to oval; rarely with a small cavity. Size about ll;u. Also some compound grains of few usually unequal compon- ents (see type 15). Plantago media Linn. {Plantaginacem.) Fresh root-stock. — Grains rounded. Size about 7fi. Also some compound and scparated-grains are observed (see type 15). Valeriana officinalis Linn. (Valerianaceos.) Fresh root-stock and stock. — Grains spherical; occasion- ally wth a central hilum. Size about 9/i. Also there are some compound grains of few equal components (see type 15). Valeriana saliunca All. (Valerianacece.) Dry root-stock. — Grains spherical or rounded-oval. Size 4 to 6/t. Among the above are compound grains of few usually equal components (see type 14). Starch quite plentiful. Valeriana tuherosa Linn. (Valerianacece.) Dry root-stock. — Grains spherical; large or small central cavity, frequently with several radiating fissures. Size about IS/i. Also a small number of compound grains of few almost equal components (see type 14). Dahlia variabilis Desf. (Composita.) Fresh tuberous roots. — Grains circular or rounded-oval; the large ones compressed to one-fourth and more of their width; they are homogeneous. Size about 40(1. Judging from the shape, this starch seems to belong to the centric-lenticular type. Cinchona sp.; Cortex china Huanaco. (flubiacece.) Dry bark. — Grains rounded, irregular, very often with curved surfaces. Size about 13/i. Most of the grains look as if they had been affected by moist heat. Starch scarce. Swertia perennis Linn. (Gentianacew.) Dry root-stock.— Grains rounded, frequently spherical, rarely with a small central cavity. Size 6 to 8fi. Among the above are compound grains of few nearly equal components (see type 14). Omphalodes verna Moench. (Boraginaceoe.) Dry root-stock. — Grains rounded, rarely with a small cavity. Size 5 to 7(U. Also compound grains of few nearly equal components (see type 14). Starch scarce. Convolvulus soldanella Linn. {Convolvulacecc.) Dry stolons. — Grains spherical, rarely oval; few with single lamellae; a small cavity is observed instead of the hilum, central to one-half eccentric. Size about 17ju. Among the above are some compound grains of few equal or imequal com- ponents (see type 17). Convolvulus lineatus Linn. (Convolvulacew.) Dry root-stock. — Simple and compound grains (see type 14) as in the preceding species. Size of the simple grains about 14/z. Convolvulus iinperati Vahl.; Batatas littoralis Chois. {Convolvulacecc.) Dry stoZo?is.— Grains spherical or oval-spherical; rarely with single delicate lamellae. Size about \.9tx. Among the above are some compound grains of few usually equal components (see type 14). Ipomoea turpethum R. Br. {Convolvulacecc.) Roots. — According to Leon Soubeiran (Journ. Pharm., 1854, XXV, 91), the grains are oval or indistinctly triangular; without hilum or lamellae; rarely with either straight or radial fissiu-es; several grains are somewhat tuberculated and apparently compound. Size 20 to 30 or 40/i- Polemonium reptans Linn. {Polenioniaceoe.) Dry root-stock.— GxsJms rounded to oval, about twice as long as broad; the broad ones slightly compressed; some grains have a small cavity. Size about Syu. Among the above are also some compoimd grains of few nearly equal components (see tj-pe 14). Atropa belladonna Linn. {Solanacece.) Dry root. — Grains rounded, oval, oblong, elongated-conical, frequently somewhat irregular; two-fifths to as broad as long; occasionally with indistinct and few lamellse; mostly at the narrow aspect instead of the hilum a small cavity is observed; eccentricity one-half to one-sixth. Length about 14 to 18m, width 13/i. Among the above are some compound grains of few equal components (see type 14). Scrophularia nodosa Linn. {Scrophulariacece.) Dry root-stock. — Grains rounded to oblong and con- ical. Size about 9/i. Compound grains of few equal or unequal components (see type 14). 238 DIFFERENTIATION AND SPECIFICITY OF STARCHES. Gratiola officinalis Linn. (Scrophulariacece.) Dry root-stock. — Grains rounded to oval. Size about 6/1. Also compound grains with a few equal or unequal components arc observed (see type 14) . VeroJiica austriaca Limi. (Scropkulariacew.) Dry root. — Grains spherical to oval; the large ones with a small central cavity. Size 8 to lOju. Also some compound grains of few equal or un- equal components (see type 14). Wulfenia carinthiaca Jacq. {Scrophulariacece.) Dry root-stock. — Grains rounded to oval. Size 5 to 7fi. Also some compound grains of few equal or unequal components (see type 14). Pediadaris harrelieri Rechb. (Scrophulariacem.) Dry root. — Grains spherical or rounded-oval; the larger ones with a small central cavity. Size about 9/x. With these are found compound grains of few almost equal components (see type 14). Starch quite plentiful. Pedicularis rosea Wulf. (Scrophulariacece.) Dry rooi.— Grains as in the preceding. Size about 7fi. With these are found compound grains (see type 14). Starch quite plentiful in some parts of the plant. Pedicularis acaulis Scop. (Scrophulariacece.) Dry root. — Grains spherical to oval; the larger ones with a small central cavity. Size about 8/x. Occasionally the size is about 11^; the grains then have a swollen appearance and a large cavity. Among the above are compound grains of few mostly equal components (see type 14). Primula calycina Dub. (Prinmlacece.) Dry root. — Grains spherical to elliptical; the larger ones have a small cavity at the thicker end instead of a hilum, one-half and one-third eccentric. Size about 15;u. These grains appear to belong to the eccentric-conical type. Among the above are some compound grains of few almost equal components (see type 14). Starch plentiful. The starch in the root-stock is less plentiful, and is contained in thick-walled cells. The grains are smaller (about 7ju) , rounded to oval ; the larger ones have a small cavity. Primula officinalis Jacq. (Prinmlacece.) Dry root-stock. — Grains rounded to oval, the larger with a small cavity. Size about 7n. Among the above are some compound grains of 2 to 4 com- ponents, and also separated-grains. Starch not plentiful; cells thick-walled. Soldanella alpina Linn. (Primulaceoe.) Dry root-stock. — Grains spherical, oval, frequently somewhat irregular, occasionally with a small cavity. Size about Q^i. Also some compound grains of few components of equal size (see type 14). Starch plentiful in thick-walled, porous cells. Dry roots of the same plant: Grains spherical or oval-spherical; with a small central cavity, frequently with several short radial fissures. Size about 13^. Also some compound grains (see type 14). Glaux marilima Linn. (Primulacece.) Dry creeping stems.- — Grains spherical or spherical-oval; frequently with a small cavity from which short fissures may radiate. Size about 12m. Also, there are some compound grains composed of a few mostly equal components, as in type 14. Starch plentiful; cell-walls not noticealjly thick. Lysimachia vulgaris. (Prinmlacece.) Dry root-stock. — Grains spherical or rounded-oval; frequently with a central cavity and a few short radial fissiu-es. Size about 15^. Among these are some compound grains of few equal or unequal components (see type 14). Pyrola rotundifolia Linn. (Ericacece.) Dry root-stock. — Grains rounded to rounded-oval, occasionally somewhat angular or irregular; many with a small cavity. Size 7 to 9ai. There are some compound grains of few nearly equal components (see type 14). Astrantia major Linn. (Umhelliferw.) Dry root-stock.- — Grains rounded or rounded-oval, often an- gular or irregular, many with a small cavity. Size about 7^- There are several compound grains and many separated-grains. Buplcurum longifolium Linn. (Umhelliferos.) Dry root-stock. — Grains rounded or oval, often irregu- lar. Size about 7^. There are some compound grains of few components. Starch not plentiful. Meum athamanticu7n Jacq.; Radix mei. (Umbelliferce.) Dry root. — Grains rounded to lanceolate, occasionally somewhat angular or irregular; the broader ones compressed to one-half and over of their width. Size about 8^. The small grains are rounded and isodiametric ; the large ones tabulate or rod-shaped. There were no indications of compound forms as in the roots of other Umbclliferae. Some of the apparently simple grains may be separated-grains. Gaya simplex Gaud. (Malvacecc.) Dry root-stock. — Grains rounded to oval, often angular or irregu- lar; many with a small cavity. Size 6 to 8ai. There are some compound grains and many separated-grains. TYPE 10. GRAINS SIMPLE, STRUCTURE OBSCURE. 239 Levisticum officinale Koch. (Umbellifcrcc.) Dry root. — Grains spherical, occasionally somewhat anpilar ; no lamellae ; with a small central cavity. Size 12 to 18//. There are separated-grains, as in tjTJc 1-1. Archangelica officinalis Koffm. {Unibelliferce.) Dry root. — Grains rounded. Size about 8^. Very little starch. Imperatoria ostnithium Linn. {Vmhelliferce.) Dry root-stock. — Grains oval, conical, oblong, occa- sionally somewhat curved, rarely slightly constricted in the middle; one-third to one-half as thick as long; occasionally almost as thick as long; without fissures. Size about 14/j, thick- ness about 6^. Hilum most likely toward the narrow aspect. Pastinaca sativa Linn. {Umhellifcrm.) Dry root. — According to Payen (Ann. Sc. Nat., 1838, ii, p. 28; pi. 4, fig l), the grains are rounded, and compressed to about half their width. Size 7.5^. Cornns suecica Linn. (Cornacece.) Dry root-stock. — Grains spherical to oval; the larger ones with a small cavity. Size 7 to 9//. There are compound grains of few almost equal components. (See type 14.) Seduvi fabaria Koch. (Crassulacea;.) Dry root-stock. ^GraiTis rounded, rounded-triangular, oval, conical, spindle-shaped, lanceolate, at times somewhat curved and more or less irregular; some with 1 or even 2 and 3 prominent solid angles and papillary processes; one-third to as broad as long; the broad ones slightly compressed; no lamellae; many with a longitudinal slit. Length about 17yLt, width 12/^. Among the above are some compound grains of few components which arise through division of the hilum, as well as from the segmentation of the solid angles or the processes. Mittella ■pentandra Linn.; Drwmnondia mittelloides DC. (SaxifragacecB.) Dry root-stock. — Grains oval, conical, rod-shaped, spindle-shaped, sometimes triangular, frequently somewhat curved, mostly more or less irregular; one-fourth to almost as broad as long; the broad ones compressed; some with a cavity or a slit. Size about 14/i. Doublets with unequal halves. Starch quite plentiful. Mittella diphylla Linn. {Saxifragacece.) Dry root-stock. — Grains rounded, oval, oblong, conical, rounded-triangular, and reniform; frequently more or less irregular; two-fifths to as broad as long; the broad ones compressed to about one-third of their width. Size about 15ai. Starch quite plentiful. Ranmicidus pyrenoBus Lirm. {Ranunculacece.) Dry-root. — Grains rounded, rounded-triangular, reni- form, oval, or shortened-conical ; frequently more or less irregular, compressed ; from the nar- row aspect a longitudinal slit is frequently observed; no lamellae. Size about 14/:. Ficaria raminculoides Moench. var. from Algiers; Ratmncidus ficaria Linn. (Rammculacew.) Dry thickened roots. — Grains rounded, triangular ^'ith rounded corners, and quadrangular, oval, conical, frequently more or less irregular; the broad ones compressed to one-half and more of their width; from the narrow aspect a longitudinal slit is observed. Size 24/i. Among the above are some doublets. The grains seem to belong to the eccentric-conical (type 7), or cuneiform type (type 8). The grains of indigenous plants show a distinct eccentric cunei- form or compressed structure. (See type 8.) Podyphyllum peltatum Linn. {Berberidacece.) Dry root-stock. — Grains spherical to oval. Size about 5fi. Some compound grains of many components and separated-grains. (See type 16.) Sanguinaria canadensis Linn. {Papaveracem.) Dry root-stock. — Grains rounded, oval, conical, or frequently irregular; occasionally with a cavity. Size about 16/^. There are some compound grains of few unequal components (see type 15). The starch undoubtedly belongs to an eccentric type. Papaver orientate Linn. (Papaveracece.) Fresh root-stock. — Grains rounded to oblong; one-half to as thick as long. Size 8n. Dentaria bulbifera Linn. {Brassicacew; Cruciferce.) Dry scales of the small axillary bulbs. — Grains oval, conical, or rounded-triangular; broad ones compressed; some with a longitudinal slit. Size about 14/i. The grains appear to belong to the eccentric-conical type. Cells entirely filled with starch-grains. Cochlearia armoracia Linn. (Brassicacece; Cruciferoe.) Fresh root. — Grains spherical to oblong; one- third to just as broad as long; occasionally slightly compressed; many with a rounded and elongated or a slit-like cavity. Size 11 to 14;i. 240 DIFFERENTIATION AND SPECIFICITY OF STARCHES. Parnassia palustris Linn. (Saxifragacew.) Dry root-slock. — Grains rounded to elongated-oval, and conical; frequently somewhat irregular; with a cavity, a slit or some fissures. Size about 18;u. Type imdetermined. The hilum is eccentric, and often appears to be at the narrow aspect. Viola paluslris Linn. {Violaccw.) Dry root-stock. — Grains spherical or rounded-oval. Size about 8(1. There are some compound grains of few equal or unequal components. (See type 15.) Cereus variabilis Pfeiff.; Cereus quadrangularis Hort. {Cactacem.) Fresh pith of the stem. — Grains rounded or oval, rarely conical; usually irregular and with either protruding rounded or blunt angles; occasionally projecting on one side into a sharp angle; half to as broad as long; broad ones slightly compressed; one-third to as thick as long; lamellae distinct, mostly more or less irregular, spiral-like, crooked, or noncontinuous in the radius of greatest thickness (in the latter case two to five systems of lamellse may arise) ; hilum more or less eccentric, in some instances about one-fiftieth. Length about lOO/i, thickness GOfx. Among the above are some semi-compound grains as in type 11 and compound ones of few unequal components, type 15. Many much smaller, simple grains and numerous compound and separated-grains were found in the pith of 1.5 inch sprouts, while the separated-grains rarely occurred in older stems. Eccentric structure could occasionally be noticed in the separated-grains. These may possibly have developed from those mentioned above with irregular lamellae, as well as from the semi-compound and compound grains of the mature pith. The starch-grains in the base of the sprouts were decidedly larger than those at the summit. Those in the paren- chyma of the cortex, which on the whole are very scarce, resemble those in the pith. Cereus martianus Zuccar. {Cactacew.) Fresh pith of the stein. — Grains rounded, occasionally some- what irregular or polyhedral as a result of pressure, sometimes just as thick as broad, some- times compressed into a lenticular form to about two-fifths of their width; few lamellse; hilum central or toward one margin. Size about 35ju. The surface of many of the grains is reticulated, probably a form of decomposition. Among the above are some semi-compound grains with 2 and 3 hila, and also some compound forms. Some cells in the parenchyma of the pith are filled with starch. Cereus erinaceus Haw.; Echinocactus erinaceus. (Cactacece.) Stems. — According to Payen (Ann. Sc. Nat., 1838, ii, p. 18), the grains are usually rounded and irregular; lamellae concentric and with sinuate radial fissures (fentes sinuese). Size about 75fi. These grains were observed in an old exotic stem, and were very scarce. In the pith of a cultivated plant also examined by Payen (loc. cit., p. 27; plate 4, fig. e) he found spherical or somewhat irregular grains; lamellae and hilum indistinct. Size about 12yii. There are some doublets. Cereus peruvianus Haw.; Cactus peruvianus. (Cactacem.) — According to Payen {loc. cit., p. 24; pi. 6, fig. 21), the grains are usually spherical or ellipsoidal, sometimes irregular; with a few dis- tinct lamellte; hilum frequently visible, about one-fifth eccentric. Size about 30/i. Among these are many compound grains of 2 to 4 components. They frequently have lamellae and an eccentric hilum. Cereus flagelliformis Mill.; Cactus flagelliformis. (Cactacece.) — According to Payen (loc. cit., p. 26, pi. 4, fig. d), the grains are irregular; many with upper surface sinuate; compressed to about one-third the width. Size about 15yu. Among these are indistinctly compound grains con- sisting of few components. Cereus serpentinus Lagasc; Cactus serpentinus. (Cadacece.) — According to Payen (loc. cit., p. 28, pi. 4, fig. m), the grains are round, sometimes imeven; lamellae and hilum usually invisible. Size about 7.5/j; also some doublets. Cereus monstrosus DC.; Cactus monstrosus. (Cactaceoe.) — According to Payen (loc. cit., p. 28; pi. 4, fig. n), very few grains are rounded; hilum and lamellae mostly invisible. Size about 6n. Mamillaria discolor Haw. (Cactaceoe.) — According to Payen (loc. cit., p. 27, pi. 4, fig. j), the grains are rounded, sometimes uneven; hilum and lamellae invisible. Size 8m. Some doublets are found among the simple grains. Rhipsalis funalis Salm. (Cactacece.) Fresh pith and parenchyma of the cortex. — Grains rounded or rounded-oval, frequently somewhat angular or irregular; slightly compressed, or almost to the middle of their width; occasionally with small delicate, concentric lamellae; from the narrow aspect frequently a longitudinal slit is observed. Size about 22/j, among these are observed some compound grains of small 2 to 4 components. Starch originated in the chlo- rophyl grains. TYPE 10. GRAINS SIMPLE, STRUCTURE OBSCURE. 241 Oijuntia brasiliensis Haw.; Cactus brasilicnsis. {Caclaceoe.) Stems. — According to Payen (Ann. So. Nat., 1838, p. 25, pi. 4, fig. b), the grains are irregular, with elevations and depressions; lanielhr and liiiuni indistinct. Size alioiit 20yu; some grains end in a lateral curved hook, probably a partly dccoini)osed form. Among these are some comjiound grains of few com- ponents. Opuntia curassavica Mill. (Cadacew.) — According to Paj^en {loc. cit., p. 27; pi. 4, fig. h), the grains are rounded or oblong, somewhat sinuate; lamellae and hilum invisible. Size about 10m. Some doublets and triplets are also observed. Opuntia tuna Mill.; Cactus opuvtin tuna. (Cactaceoe.) — According to Payen {loc. cit., p. 27, pi. 4, fig. f), the grains are spherical, sometimes rather irregular; lamella? and hilum invisible. Size about lOfi. Some doublets are also observed. Opuntia ficus indica Mill. — According to Payen {loc. cit., p. 27, pi. 4, fig. g), the grains are similar to the preceding, only somewhat smaller and less numerous. Pereskia grandiflora Haw.; Cactus pereskia grandiflora. {Cactacece.) Pith. — According to Payen {loc. cit., p. 25, pi. 4, fig. a), the grains are rounded, usually irregular and angular; with a few distinct lamellae and more or less eccentric hilum. Size about 22.5^. Many separated- grains are also observed. Porttdaca megalantha Steud. {Portulacacece.) Dry root. — Grains spherical, or rounded-oval; occa- sionally with a central cavitj'. Size about 14^. Also some compound grains of few equal components are observed (see type 14). Ullucus tuberosa Lozano. {Chenopodiacece.) Tubers. — According to Leon Soubeiran (Jom-n. Pharm., 1854, XXV, p. 99), the smaller grains are oval or spherical, larger ones elongated and slightly curved, some indistinctly triangular; the oblong grains have distinct lamellae and hilum. Size 20 to 50 and 60^. Saponaria officinalis Linn. {Caryophyllaceoe.) Dry root. — Grains rounded or acute-angled; the latter evidentlj^ separated-grains. Size 7 to 8^. Poor in starch. Althcea rosea Cav. {Malvacece.) Fresh root. — Grains rounded or oval, rarely oblong; not at all or slightly compressed. Size 7 to 9ju. Poor in starch. Adansonia digitata Linn. {Malvacece.) Dry fruit pulp. — Grains rounded-oval to elliptical; curved on one side and almost straight on the other; two-thirds to as broad as long; the broad ones compressed; lamellie none, or very delicate; usually wdth a slit-like cavity from which deli- cate transverse fissures proceed; cavity is very distinct from the narrow aspect and very indistinct from the broad aspect. Size about IS/u. The starch resembles the centric oval Androsceiim officinale All. {Hypericacew.) Fresh root. — Grains rounded to oval, often irregular; compressed to one-fourth their width; frequently sunken or with a cavity in the middle. Size 14 to 18^. Hypericum elodes Linn. {Hyperiacacece.) Dry creeping root-stock. — Grains rounded to oval, usually shghtly compressed; from the narrow aspect a longitudinal slit is usually observed. Size about 13^. Among these are found compound grains of few equal or unequal components (see type 14). Byrsonima crassifolia DC. {Malpighiacece.) — Grains spherical or oval; once to almost twice as long as thick, lamellae and hilum none; rarely a cavity in place of the hilum, about one- fourth eccentric; the two ends apparently alike in thickness. Length about 18ju, thickness about ISyti. Among these are found compound grains of few, mostly equal components (see type 14). Euphorbia cyparissias Linn. {Euphorbiaceoe.) Fresh root-stock. — Grains rounded, oval, conical. Size about 10/j. Separated-grains with 1 to 4 pressure facets. Size 2 to 8^. Only a few perfect compound grains were seen within the cells. They consist of 2 to about 6 and 8 chiefly unequal components. Cretan eluteria Swartz. Cascarilla bark. {Euphorbiacem.) Dry bark. — Grains rounded or oval, occasionally somewhat angular or irregular; many are hollow. Size about 12fi. Poor in starch. Galipea officinalis Hancock; Cortex angusturce verus. {Rutacece.) Dry bark. — Grains spherical or angular with rounded angles; many with a cavity which is frequently large. Size about 12/j. Poor in starch. 16 242 DIFFERENTIATION AND SPECIFICITY OF STARCHES. Dictamnus aJhus Linn. (Ridaccw.) Dry root. — Grains rounded or angular. Size about S/x. Many of them apparently separated-grains. Whole compound grains are seldom observed, and when present they consist of 2 to 8 almost equal components. Poor in starch. Guaiacum officinale Linn. ; Cortex gnniaci. {Zygophyllacece.) Dry bark. — Grains rounded, triangular with rounded angles, conical or elongated-oval, occasionally somewhat irregular; many with a small central cavity. Size about lOyu. Poor in starch. MyriophyUum verticiUatum Linn. {Haloragacece.) Base of dry rooted stems. — Grains spherical or rounded-oval; the larger ones with a cavity. Size about 10/:;. Among these, compound grains of few, nearly equal components are observed (see type 15). Trapa natans Linn. {Onagracece.) Dry rooted stolons. — Grains rounded or rounded-oval; the larger ones slightly compressed with a longitudinal slit from the narrow aspect. Size about 12/^. The starch resembles the eccentric-lenticular type. Doublets and triplets are also present. Punica granatum Linn. ; Cortex radicis granati. {Lythracece.) Dry cortex of the roots. — Grains rounded to elongated-oval, sometimes triangular and often slightly irregular. Size about 9^. Poor in starch. Potentilla tormentilla Linn.; Tormentilla erecta Linn.; Radix tormentillw. (Rosacea;.) Dry root-stock. — Grains rounded to oblong, frequently triangular with rounded angles, usually irregular; more or less compressed; from the narrow aspect a longitudinal slit is observed; occasionally somewhat thickened on the convex border, and thinned and squared on the opposite one. Size about 17/x. The grains seem to be affected by slight heat; from the broad as well as from the narrow aspect the margin is frequently bent with knob-like protuberances. The cells are packed with starch. Spirwa filipendula Linn. (Rosacece.) Dry root-stock. — Grains rounded-triangular, oval, conical, spindle-shaped; often more or less irregular; some with a cavity. Size about 12ju. Also com- poimd grains of few equal or unequal components are ol^served. (See type 15.) Ononis spi7iosa; Radix onomdis. (Leguminosce.) Dry root. — Grains roimded, often somewhat angular, rarely oval; many with a smaller or somewhat larger central cavity. Size 7;u. Also compound grains of few components are observed. Some of the apparently simple ones may be sepa- rated-grains. Trifolium alpinuni Linn. {Leguminosce.) Dry root-stock. — Grains rounded to oblong, conical, and elongated spindle-shaped; three or more times as long as thick. Length about 10^- Trifolium montanum Linn. {Leguminosce.) Dry root-stock. — Grains oval, conical, spindle-shaped; frequently linear. Four or more times as long as thick. Length 11^- Trifolium badium Schreb. {Legu7ninos(F.) Dry root-stock. — Grains rounded, oval, conical; aljout twice as long as broad; the broad ones compressed, some with a small cavity. Length about 10^. Doublets and triplets are also observed. All three species of Trifolium are poor in starch. Glycyrrhiza. {Leguminosce.) Dry root. — Grains usually spherical; sometimes elongated oval, or ovate-conical, frequently irregular; a few grains have a small central cavity. Size about lOyu. Also some compound grains of 2 to 4 components, and some separated-grains are observed. Starch is found in the pith, cortex, and wood. Phaca alpina Jacq. {Papilionacece.) Dry root. — Grains rounded to oval. Size about 7^. Some compound grains of 2 to 4 components are also observed. Size 9,u. Poor in starch. Astragalus incanus Linn. {Leguminosce.) Dry root-stock. — Grains rounded or oval. Size about 13/i. Also compound grains of few mostly equal components are observed (see type 14). Not much starch in the pith cells, the walls of which are very thick, lamellate, and gelatinous. Chara. {Characece.) — The spores contain two kinds of grains: (1) large ones rounded and com- pressed (see type 2) ; (2) smaller grains without lamellae or hilum. In sha])e the latter cor- respond to the starch-grains in the seeds of Leguminosce and thus seem to approach type 3. They differ, liowever, in being thinned and knife-like at the concave border. This is always observed in grains at a certain stage of growth, and may later more or less completely dis- appear. Nageli regards the large and small grains as being distinct, inasmuch as when viewed on end the former are always spindle-shaped or elliptical, never conical or pear-shaped. A small gas-bubble is noted in many of the small, usually homogeneous grains, after they have been subjected to dry heat (210° C.) and examined in alcohol. This bubble indicates the position of the hilum, and corresponds to the middle of the curved slit which ijcfore treatment was visible. Eccentricity amounts therefore from one-half to one-third and slightly over. TYPE 10. GRAINS SIMPLE, STRUCTURE OBSCURE. 243 Chara fcrtida A. Briiuii. Fnsh spores. — Grains usually triangular, rarely quadrangular, oval, reni- form or irregular; compressed to one-half or more of their width; one border, straight or concave, and thinned into a more or less knife-like edge; the other strongly convex and thickened, .'^o that when vioweil on end the grain appears conical or pear-shaped; wthout lanielkc or hilum; from the broad aspect a delicate curved slit is observed, from the center of which several short fissures may radiate. Size about 20/u. Cham hispida Linn. Dry spores. — Grains rounded; oval, frequently reniform or triangular with roundcnl angles; usually less dense at the straight or concave border; the broad ones com- pressed to half tlu'ir wiilt h ; from the narrow aspect a distinct slit, and from the broad longi- tudinal aspect a delicate slit is observed; the latter, in the reniform and triangular grains, is curved toward the convex margin. Size 16 to 25,u. Chara aspera Willd. Dry spores. — Grains usually triangular or reniform, sometimes oval or irreg- ular; scarcely half to almost as broad as long; shortened to two-fifths of their length, one side of the margin usually strongly convex and thickened, the other, slightly convex, straight, or concave and thinned to a knife-like edge; occasionally from the broad aspect with a very delicate curved and somewhat eccentric slit; from the narrow aspect a somewhat more marked, straight median slit. Size about 16/i. Chara alopeciiroides Willd. Dry spores. — Grains reniform, or 3, rarely 4 to 5, angles; usually almost as broad as long; compressed to one-half or over their width; thicker on one edge than on the other. Size about 22fi. Among these there are smaller grains, which have one thick- ened convex and one sharpened more concave edge, as in Chara aspera. The following species of Chara have grains similar to those already described above, namely: Chara baiieri A. Braun, size about 25jli; Chara barhata Meyen, size about 20^; Chara fragilis Desv., size about 17/^; Chara conlraria A. Braun., size about 20/;; Chara gymnophylla A. Braun, size about 21/^; Chara coronata Zig., size about 18^. These were all examined in the dry state. Isoetes lacustris Linn. {Isoetacew.) Dry gyninospores. — Grains spherical, rarely somewhat irregular. Size about 8yu. Starch not plentiful, embedded among oil-bodies and plastids. Pinus sylvestris Linn. (Coniferm.) Dry pollen. — Grains oval to pear-shaped. Size about ifi. Poor in starch and rich in oil. Avena puhescens Linn. {Graminacece.) Fresh pollen. — Grains rounded, oval, pear-shaped. Size about 4.5At. Starch plentiful. Bromus mollis Linn. {Graminacece.) Fresh pollen. — Grains as in preceding. Size about 6ju. Starch plentiful. Alpinia nutans Rose; Globba nutans. (Zingiberacece.) Pollen. — According to Payen (Ann. Sc. Nat., 1838, p. 26, pi. 5, fig. 4), the grains are oblong; .sometimes curved, about 3 times as long as broad. Size about 15/i. Starch plentiful in the large pollen-grains; lacking in the smaller ones. Naias major Roth. (Naiadacece.) Pollen. — According to Payen {loc. cit., p. 28, pi. 5, figs. 5 and 6) and Fritsche (Ueber den Pollen, Taf. iii, fig. 5), the grains are oval to almost cylindrical, more or less curved, and about 2 to 3 times as long as broad. Size about 7.5yu. Starch plen- tiful. Ruppia maritima Linn. {Naiadacem.) Pollen. — According to Payen (loc. cit., p. 27, pi. 3, fig. 25), the grains are rounded or oval, more or less irregular, occasionally cylindrical with rounded ends, more or less crooked. Size about 11^. Starch plentiful. Syri7iga vulgaris Linn. (Oleaceoe.) Fresh pollen. — Grains rounded. Size hardly 3m. Starch less plentiful than oil. Veronica chamwdrys Linn. {Scrophulariacew.) Fresh pollen. — Grains rounded, oval, pear-shaped. Size about 6^. Starch plentiful. Ranunculus bulbosus Linn. {Ranuncidacece.) Fresh pollen. — Grains rounded. Size at most 3^. Viola cornuta Linn. (Violacew.) Fresh pollen. — Grains rounded, oval, pear-shaped. Size about 4/i. Starch plentiful. Geranium molle Linn. {Geraniacem.) Fresh pollen. — Grains more or less rounded. Size about 3^. Starch quite plentiful. Geranium pratense Linn. {Geraniacece.) Fresh pollen. — Grains rounded, oval, pear-shaped. Size about 5m. Similar grains are found in the cell-lumen and in the cell-tissues. 244 DIFFERENTIATION AND SPECIFICITY OF STARCHES. Encephnhrtns spiralis Lehm. (Cycadacew.) Dry seed. — Grains rounded to oval, sometimes triangu- lar to rcniform, frequently more or less irregular; the broadest ones compressed to about one-third the width; without lamellae and hilum; a longitudinal slit which is quite dis- tinct from the narrow aspect, and rarely visible or less distinct from the broad one; Size 15 to 20fi. Among the above are some compound grains of few components. Many of the grains are affected by heat. The starch-grains in the embryo belong to the eccentric- conical type. Brachypodium pinnatiim Beauv.; Festuca pinnata Moench.; Bromus rupestris Host. (Graminacew.) Dry seed. — Grains rounded to oblong, frequently conical, triangular, or reniform, usually more or less irregular; 1 to 4 times as long as broad; the broad ones compressed; some grains are hollow; from the narrow aspect a longitudinal slit is observed in the compressed forms. Length about ll/x, width 8^. Among the above some compound grains of 2 to 3 components and some separated-grains are observed. Starch as in Bromus. Boissierra bromoides Hochst. {Graminacece.) Dry seed. — Grains rounded and rounded-triangular to conical and oblong; the broad ones compressed. Size about 1^. Starch as in Bromus. Boissierra bromoides Hochst., var. pappophorum pumilio Trin. (GraminacecB.) Dry seed. — Grains oval to elongate-lanceolate; one-third to two-thirds as thick as long. Length about 6^., thick- ness about 2.5m. Bromus. {Graminacem.) Seed. — The grains are simple, usually without lamellae or hilum. Struc- ture rarely distinct, if so always centric, but sometimes more spherical, sometimes more rounded-lenticular, sometimes oval or lanceolate, or even almost terete and compressed. Individual distinctly compound grains consisting of 2 to 3 components rarely occur. In many species, however, there are forms in varying numbers which are somewhat constricted in the middle and may be doublets. Bromus madritcnsis Linn. Dry seed. — Grains spherical or rounded-oval, occasionally somewhat irregular; many slightlj' compressed, lamellae concentric, delicate and numerous; instead of the hilum a small cavity is observed, which is either rarely spherical and occasionally elon- gated, or frequently flattened; from this cavity a few or numerous fissures radiate; from the narrow aspect a distinct median cleft is very often observed. Bromus polystachus DC. Unripe seeds. — Grains simple, as in the preceding. Bromus maximus Desf. Dry seed. — Grains as in Bromus madritensis, usually somewhat smaller in diameter. Bromus gussonii Parlat. Dry seed. — Grains rounded-oval, rounded-reniform; triangular, very fre- quently somewhat contracted in the middle and thus quadrangular; slightly compressed, up to about half their width; lamella; invisible, or concentric and delicate; from the narrow aspect a longitudinal slit is frequently observed, from the broad aspect occasionally with some radial fissures. Size about 35/i. The grains resemble those of the Hordeaccxe, differing, however, in their more angular (never circular) forms as well as in the lack of small grains. Grains of almost the same size are only found in one cell. Bromus rigidus Roth. Dry seed. — Grains rounded, reniform, 3 to 4 angles; occasionally somewhat irregular, and frequently somewhat contracted in the middle; slightly compressed up to about half their width; lamellae concentric and delicate, or indistinct; cavity often slit-like, from which a few short radial fissures may emerge. Size about 33 to 40/x. Starch as in Bromus g^issojii. Large starch-grains of seeds from botanical gardens in Berlin (1855) are 24 to 29fi, and those from the Paris gardens 20 to 24/;. Otherwise they agree with those above described. Bromus xcolgcnsis Jacq. Dry seed. — Grains rounded, or rounded-oval, more rarely oval, rounded pear-shaped, or triangular; two-thirds to as broad as long; not at all or slightly compressed with single distinct, concentric lamellae; instead of the central hilum either a rounded or oblong cavity, or rarely a longitudinal slit is observed. Size about Slfi. Also some compound grains of 2 to 4 components are present. Bromus squarrosus Linn. Dry seed. — Grains rounded to oval, occasionally somewhat irregular; half to as broad as long; the broad ones compressed; occasionally with a central cavity. Length about llyu. Bromus arvensis Linn. Dry seed. — Grains rounded to oval, slightly compressed; many with a cav- ity. Length about 17/i. TYPE 10. GRAINS SIMPLE, STRUCTURE OBSCURE. 245 Bromus rubens Linn. Dry seed. — Grains rounded, rounded-reniform, oval, frequently somewhat angular, the broad ones compressed to about half their width. Length about 13 to 17/u. Bromus crcctus Huds. Dry seed. — (trains rounded, rounded-reniform, oval, shortened-eonical, usually irregiilar ; the broad ones ('ompressed to about half their width; from the narrow aspect a longitudinal slit is frequently observed. Length about 13p. Bromus sterilis Linn. Dry seed. — Grains rounded and rounded-triangular to oval, frequently some- what angular or irregular; usually compressed to about one-half; with a central cavity or median slit from which fissures occasionally radiate. Size about 36/ii. Broryms tectorum Linn. Dry seed. — Grains rounded to oval, sometimes triangular and quadrangular with rounded angles, frequently somewhat irregular; usually strongly compressed; from the narrow aspect a longitudinal slit is observed. Length about 23m- Bromus adoensis Hochst. Dry seed. — Grains rounded, triangular to quadrangular with rounded angles, oval, shortened-eonical; usually more or less irregular; the broad ones compressed to about one-half and over of their width. Length about ISft. Bromus ciliatus Linn. Dry seed. — Grains rounded to oval, the broad ones compressed to half their width; a central cavity is occasionally observed. Length about 14/i. Bromus brachystachys Hornung. Dry seed. — Grains rounded, oval, shortened-eonical; occasionally somewhat irregular; the broad ones compressed to about half their width; instead of a central hilum, a cavity is present which is rounded from the broad aspect and cleft-like from the narrow aspect. Length about 14yu. Bromus caucasiciis Fisch. Dry seed. — Grains rounded, oval, shortened-eonical, usually irregular; the broad ones compressed to about one-third of their width. Size about 12/i. Bromus laxus Hornem. Dry seed. — Grains rounded to oval-oblong, very oft«n irregular and angular, and sometimes constricted in the middle; the broad ones compressed. Length about 14yu. Bromus aleutensisTrm . Dry seed. — Grains rounded-oval, oval-oblong, elongated-triangular; some- times reniform or quath-angular, frequently somewhat irregular; half to almost as broad as long; the broad ones compressed to half their width. Length about 17 to 20ju, width aliout 14ju. Among the above some compound grains of 2 to 3 equal components are oliserved. Br(nmis longiflorus Willd. Dry seed. — Grains rounded to oval-oblong, often somewhat irregular and angular and frequently constricted in the middle; the broad ones compressed to a little more than half their width. Length 10 to 12/i. A central cavity is sometimes seen in each half of the constricted gi-ains, demonstrating that they are doublets without a visible line of division. Bromus purgans Linn. Dry seed. — Grains rounded, rounded-triangular to oblong, ovoid, and con- ical; occasionally constricted in the middle; the broad ones compressed to one-third of their width; from the narrow aspect a longitudinal slit is frequently observed. Length about 12^. Bromus brizceformis F. & M. Dj-y seed. — Grains oval, oblong, conical; occasionally slightly com- pressed. Length about 6^. Bromus lanceolatus Roth. Dry seed. — Grains oblong, rarely rounded. Length about Qfi. Bromus velutinus Schrad., var. hordeaceus Gmel.; Bromus secalinus var. hordeaceus. Dry seed. — Grains rounded-oval to olilong; one-half to three-fourths as broad as long, the broad ones slightly compressed, some have a central cavity. Length 10 to 11^. Bromus patulus Mert. & Koch. Dry seed. — Grains rounded-oval to oblong, sometimes slightly irregular, two-fifths to four-fifths as broad as long; the broad ones slightly compressed; many with a central cavity. Length about 12fi. Bromus schraderi Kunth; Ceratochloa pendula Schrad. Dry seed. — Grains rounded, reniform, oval, oblong, frequently irregular; compressed, many with a central cavity. Length about 20ix. Bromus unioloides Willd.; Ceratochloa uuioloides DC. Dry seed. — Grains as in the preceding species, the broad ones compressed to one-half and more of their wiilth; many with a longitudinal slit-like cavity. Bromus cornrnutatus Schrad. Dry seed. — Grains rounded to oblong; frequently somewhat angular or irregular; two-fifths to almost as broad as long; the broad ones rather strongly compressed. Length about 13/i. Bromus inermis Poll. Dry seed. — Grains rounded-oval to oblong; broad ones compressed to iialf their width. Length about 10^. 246 DIFFERENTIATION AND SPECIFICITY OF STARCHES. Bromus canadensis Michx. Dry seed. — Grains rounded, oval, oval spindle-shaped, conical; the broad ones compressed to half their width. Length about 8/i. Bromus pendulinus Schrad. Dry seed. — Grains rounded-oval to oblong and conical; sometimes con- stricted in the middle; frequently irregular; compressed to about one-half and one-third their width. Length about 21ju. Bromus arduennensis Kunth. Dry seed. — Grains rounded to oblong, occasionally triangular and quadrangular with blunt angles; the broad ones compressed to about one-third their width; from the narrow aspect the central cavity is slit-like. Length about I'Z/x. Bromus asper Murr. Dry seed. — Grains roundetl-oval to elongated rod-shaped and conical; usually twice as long as broad. Leng-th 5 to 7yu. Bromus secalinus Linn. Dry seed. — Grains oval to elongated-lanceolate; one-third to two-thirds as broad as long. Length 6 to 7n. Bromus diandrus Curt. Hort. berol., 1855. Dry seed. — Grains rounded-oval and rounded-triangular to rod-shaped and elongated-conical; the broad ones compressed to one-third their width; narrow ones about 3 times as long as broad and almost terete; many somewhat irregular, and several more or less constricted. Length aliout 14/^. Bromus mollis Linn. Dry seed. — Grains rarely rounded-oval, usually oblong or lanceolate. Length about 4 to 5n, width 1.3 to l.Sfi. Bro7nus divaricaius Rohde. Dry seed. — Grains rounded-oval to lanceolate, frequently irregular and ang-ular, often triangular; one-fourth to almost as broad as long; the broad ones strongly compressed. Length about 15;u. Bromus vestitus Nees. Dry seed. — Grains oval to lanceolate and lanceolate-linear; occasionally slightly constricted in the middle; 1.33 to 3 and 4 times as long as broad; the broad ones compressed to half their width. Length about lO/u. Bromus confertus Biebrst. Dry seed. — Grains minute, rounded or oblong. Length 2.5^, width hardly 1.4/i. If the starch-grains are freed from the cells they frequently hang together in globular masses, but these are unquestionably not compound grains. At least not any are ever noted within the uninjured cell. Hordeacece {Trilicum, Agropyrum, Secale, Elymus, Hordeum, Mgilops, Braconnotia) . Seeds. — Besides the large grains which are related to the centric lenticular type, there are small grains in great numbers observed in the inner tissues, while only the small grains are found in the outer cells. They are spherical, rounded-angular, or even almost polyhedral, and are more or less markedly compressed. Size about lOyu. The grains are simple, although separated- grains are also found among them. Whole doublets and triplets are rarely observed. Cyperacew. — The dry seeds contain starch-grains which are rounded or oval, sometimes compressed into a lenticular form; many of them are angular or polyhedral and resemble separated- grains; the latter are, however, undoubtedly simple grains which are merely changed in form as a result of pressure. Plastids which resemble compound starch-grains and which become yellow or golden yellow when treated with iodine are often observed. Cyperus flavescens Linn. (Cyperacem.) Dry seed. — Grains rounded or oval, frequently rounded- angular. Size 8 to ll/i. Cyperus strigosus Linn. (Cyperacece.) Dry. — Grains rounded, oval, oval-conical, frequently some- what irregular and angular. Size 8 to 10/i. Compound plastids are also observed. Mariscus jacquini Humb., Kunth. (Cyperacece.) Dry seed. — Grains rounded or polygonal, occa- sionally irregular, strongly compressed. Size about 12/i, thickness 1 to 2yLi. Mariscus umbellatus Vahl. (Cyperacece.) Dry seed. — Grains as in the preceding. Kyllingia odorata Vahl. {Cyperacece.) Dry seed. — Grains rounded or rounded-oval, angular or poly- gonal, compressed to half or more of their width. Size about 7^, thickness about 3/i. Plas- tids are lacking. Heleocharis palustris R. Br. (Cyperacece.) Dry seed. — Grains rounded, usually angular. Size about 8m. Heleocharis ovala R. Br. (Cyperacece.) Dry seed. — Grains rounded or rounded-angular; and com- pressed to half their width. Size about 8 to lO/z. Scirpus mucronatus Linn. (Cyperacece.) Dry seed. — Grains polyhetlral, slightly compressed. Size about Sm- Sdrpus maritimus Linn. (Cyperacece.) Dry. — Grains spherical or slightly angular. Size about 5^. TYPE 10. GRAINS SIMPLE, STRUCTURE OBSCURE. 247 Isolepis setacea 1\. Br.; Scirpiis setaceus Linn. (Cyperaceoe.) Dry seed. — Grains rounded-angular, many irrc^gular, and sonic also polygonal; compressed into lenticular form, occasionally with large reticulated markings. Siz(> about 9^. Plastids scarce and irregular. Isolepis supina R. Br. ; Scirptis supiitus Linn. {Cyperacece.) Dry seed. — Grains polygonal, compressed, with square reticulations on the surface. Size about 10/j. Starch-grains are crowded in the cells. Plastids none or very few. Isolepis eckloniana Schrad.; Isolepis verruculosa Steud. (Cyperacece.) Dry seed. — Grains rounded- angular or polygonal, usuallj' irregular; compressed to one-third of their width. Size about 10;u. Plastids few and irregular. Isolepis holoschcenus Roeni. and Schult.; Scirpus holoschoenus Linn. {Cyperacece.) — Grains as in the preceding. Fimbristylis dichotoma Vahl. [Cyperacece.) Dry seed. — Grains rounded, oval, usually angular, often strongly compressed. Size about 10^. Plastids of irregular form are also present. Fimbristylis annua Roem. and Schult. (Cyperacece.) Dry seed. — Grains rounded or rounded-angular, and compressed. Size about 9/i. Fimbristylis brizoides Smith.; Frimbristylis laxa Vahl. (Cyperacece.) Dry seed. — Grains rounded or rounded-oval, frequently somewhat angular, rarely almost polyhedral; slightly or not at all compressed. Size 9 to 12/i. Eriophorum alpinum Linn. (Cyperacece.) Dry seed. — Grains rounded, oval, frequently angular. Size about 6//. Oil plentiful; plastids numerous. Erioplwrum vaginatum Linn. (Cyperacece.) Dry seed. — Grains as in the preceding species. Rhynchospora fusca Roem. and Schult. (Cyperacece.) Dry seed. — Grains rounded, usually angular, and often strongly compressed. Size about 15^. Some doublets and triplets are also observed. Plastids are irregular. Cladium mariscus R. Br. (Cyperacece.) Dry seed. — Grains rounded or angular. Size about 9/i. Some compound plastids are also observed. Chcetospora nigricans Kunth.; Schcenus nigricans Linn. (Cyperacece.) Dry seed. — Grains rounded. Size about 5ti. Some oil and plastids are also observed. Blysmus compressus Panz.; Schasnus compressus Pers. (Cyperacece.) Dry seed. — Grains spherical; rounded-triangular, oval, shortened-conical; occasionally somewhat angular, frequently slightly compressed. Size about 9/i. Scleria triglomerata Michx. ; Cladium triglomerata Nees. (Cyperacece.) Dry seed. — Grains rounded, rarely elongated-oval, frequently somewhat angular, the broad ones occasionally compressed. Size about 6 to lO^u. There is a great amount of oil and considerable starch in seeds, which are not fully ripe. Scleria bracteata Cav. (Cyperacece.) Dry seed. — Grains as in preceding, but somewhat less numerous; size not more than 7/j. Seeds not fully developed. Scleria microcarpa Nees. (Cyperacece.) Dry seed. — Grains rounded, rarely oval, usually more or less angular, the broad ones slightly compressed. Size 8^. Seeds are well developed, and contain much oil and starch, as well as numerous almost polyhedral plastids. Scleria ophryoscleria, species from Brazil. (Cyperacece.) Ripe seed. — Grains similar to the preceding. Size about 7^l. Starch wanting in the embryo. Scleria hispidula Hochst. (Cyperacece.) Dry seed. — Grains rounded. Size almost 5^. Compara- tively little oil and starch in unripe seeds. Carex pulicaris Linn. (Cyperacece.) Dry seed. — Grains rounded-angular to polyhedral; frequently compressed; many have a central cavity. Size about 16/ti. Carex arenaria Limi. (Cyperacece.) Dry seed. — Grains angular with rounded angles or polyhedral; the large ones compressed. Size about 7fi. Carex maxima Scop. (Cyperaceas.) Dry seed. — Grains rounded-angular; the large ones have a central cavity. Size about 6;u. Oil rather plentiful. Elyna spicata Schrad. (Cyperacece.) Dry seed. — Grains spherical or oval, frequently angular. Size about GfjL. Starch-grains and irregular plastids are crowded in the cells. Kobresia caricina Willd.; Elyna caricina Mert. and Koch. (Cyperacece.) Dry seed. — Grains spherical, rarely oval or shortened-conical, occasionally somewhat angular. Size about 6^. Flagellaria indica Limi. (Flagellariacece.) Dry seed. — Grains rounded or usuallj' polyhedral as a result of pressure. Size about 5^. Starch-grains fill the cells. 248 DIFFERENTIATION AND SPECIFICITY OF STARCHES. Lilium bulbiferum Linn. (Liliacca;.) Fresh endosperm or unripe seeds. — Grains rounded, rounded- triangular, oval, elliptical, shortened-conical; two-thirds (rarely one-half) to as broad as long. Length about lO/i, width T/x. Starch quite plentiful, little oil, and much protoplasm are found in the thick-walled cells. Sparganium nutans Limi. (Typhaceoe.) Dry seed endosperm. — Grains rounded, oval-elliptical, short- ened-conical, frequently somewhat irregular; three-fifths to as Inroad as long, the broad ones slightly compressed, many with a cavity. Size 7 to 8;u. Among the above are some doublets and triplets with parts almost equal. Quercus pedunculata Willd. (Cupuliferce.) Dry cotyledons. — Grains rounded or oval, sometimes irregularly angular with blunt angles; frequently slightly compressed; two-thirds to three- fourths as thick as broad; the hilum and very delicate lamellse are rarely distinct; eccentricity about one-fom-th. Size about 29fi. Compound grains of few mostly unequal components. (Type 15.) Quercus cerris Linn. {Cupuliferce.) Dry cotyledons. — Grains rounded, oval, shortened-conical, fre- quently more or less irregular; lamellae indistinct; instead of the hilum a small cavity toward the thicker end is frequently observed; eccentricity usually one-half, rarely one-third and one-foiu-th; frequently one or two slits with delicate, radiating fissures are found. Size about 26m. Isolated compound grains of few components also observed. Quercus ilidfolia Wangh. (Cuptdiferoe.) Dry cotyledons. — Grains as in both of the preceding species; the greatest number are oval. Size 17 to 18m. Castanea vesca Gart. (Cupuliferce.) Fi-esh and dry cotyledons. — Grains rounded, triangular and quadrangular with rounded angles, oval, shortened-conical, frequently irregTilar; half to as broad as long; the broad ones slightly compressed; lamellte are invisible, or very delicate; hilum in the fresh grains frequently indistinct. After desiccation, instead of the hilum, either a small cavity or occasionally a delicate triangular slit is observed, which is either central or toward the thicker margin. Size about 20m, rarely 27m. Many grains are trian- gular, thickened at the end in which the hilum is located, and thinned and squared at the opposite end. These belong to the cuneiform type (type 8). Among the above are some semi-compound grains, also some doublets and triplets, the components of which are usually unequal. Size of the separated-grains is about 10m. Fagus sylvatica Linn. var. pendula. (Cupuliferce.) Dry cotyledons. — Grains spherical, the larger ones occasionally somewhat irregular; many have a small central cavity. Size about 6m. Forestiera acuminata Poir.; Borya acuminata Willd. {Oleacece.) Dry cotyledons. — Grains rounded or oval, frequently angular or irregular. Size about 7m. Rather rich in starch, besides much oil. Only oil is found in the endosperm. Cinnamomum ceylanicum Nees. (Lauraceoe.) Dry seed endosperin. — Grains spherical to oval with single radial fissures. Size 15 to 18m. Also some compound grains of 2 to 4 equal compo- nents are observed. Oil and starch appear to be present in equal quantities. Apollonias canariensis Nees. (Lauracece.) Dry cotyledons. — Grains rounded, conical, oblong and elongated spindle-shaped, frequently somewhat irregular. Size about 7m. Agathophyllum aromalicum Willd. {Lauracece.) Dry cotyledons. — Grains spherical, rarelj' somewhat angular, frequently with a small central cavity. Size about 13m. Some compound grains of 2 to 4 equal components are also present. Hernandia sp. (Lauracew.) Dry cotyledons. — Grains rounded or angular. Size about 5m, rarely more. Some starch and much oil; seeds probably not entirely ripe. FlumbaginacecB. Seed endosperm. — The starch-grains (only those in ripe, dry seeds were examined) entirely fill the cells. They are rarely rounded, usually either blmatly angular, or more often polyhedral with sharp edges and angles, thus completely resembling separated-grains. No compound grains were noticed even within cells, so that in all probability they are all simple and have become flattened as result of pressiue; even the small ones (2m) sometimes show an angular structure. In all likelihood the grains belong to the centric-spherical or oval type. Armeria formosa Hort. (PlumbaginacecB.) Dry seed endosperm. — Grains polyhedral, usuallj' with sharp edges and angles, frequently irregular; two-fifths to as broad as long; frequently with a rounded or oblong cavity from which sometimes radial fissures emerge. Size 15 to 18m. TYPE 10. GKAINR SIMPLE, STRUCTURE OBSCURE. 249 Armeria alpina WilUl., v;ir. angustifolia. {Pluiiibaginaccct.) Dnj seed. — Grains rounded or oval, bluntly angular or sharply polyhedral; two-fifths to as thick as long; usually with a rounded or oblong cavity and frequently with radial fissures. Size about 8 to 21;u. Statics limonium Linn. {I'lumhaginacecc) Dry seed endosperm. — Grains rounded or oval, blunt- angular, or polyhedral, rarely with sharp edges and angles; frequently with a cavity or slit. Size about 3 to 18/i. Statice elata Fisch. {Flumbaginacece.) Dry seed. — Grains rounded to oblong, angular, or sharply polyhedral; one-third to as thick as long; either a rounded or an oblong cavity, or even a longitudinal slit, is frequently observed, from which fissures occasionally radiate. Many grains are slightly shrunken. Size about 27 to 32yu. Goniolimon exirnium Boiss. (Plumhaginacece.) Dry seed endosperm. — Grains isodiametric to t^vice as long as thick, polyhedral ynt\\ sharp angles and edges; frcciuently they have a central cavity; without either fissures or lamellae. Size about 28/i. Cells are thin-walled and filled with starch. Plumbago juicrantha Ledeb. (Plumbaginacece.) Dry seed endosperm. — Grains sharply polyhedral, with a small or a large cavity. Size about 8 to lO/i. Campanula sp. {Campamdacece.) Fresh unripe seed endosperm. — Grains rounded-oval or rounded- angidar. Size about 7(U. Also some separated-grains are present. Seeds very young, and contain considerable starch. Menodora sp. (Oleacece.) Dry seed tegment. — Grains rounded, rounded-angular, or polyhedral; with a large or a small cavity. Isolated compound grains are also present. Erycibe paniculata Roxb. {Leguminosce.) Dry cotyledons. — Grains spherical or rounded-oval, occa- sionally somewhat angular. Size about Tm- Eutoca viscida Benth. {Euphorbiacece.) Dry seeds. — Grains spherical; a small central cavity is found in the larger ones. Size about 6 to 8;u. A great deal of oil is found, and most of the seeds contain onlj^ oil. Starch in small quantities is probably present in the seed, which is not fully ripe. Phacelia congesta Hook. {Hydrophyllacece.) Dry seed endosperm. — Grains as in the preceding species. Digitalis lutea Linn. (Scrophidariacece.) Fresh, imripe fruit, placenta, and funiculus. — Grains roimded, or oval, and frequently angular. Size about lOix. Also compound grains of 2 to 8 equal com- ponents are observed. Starch plentiful; wanting in the ovules and seeds. Verbascuni schraderi Mey. (ScrophidariacecB.) Fresh placenta of unripe fruit. — Grains angidar, here probably mostly separated-grains. Size about 6 to Six. Starch very plentiful. Verbascum schraderi Mey. {Scrophidariacece). Fresh, unripe seed-coats. — Grains rounded or oval, usually angular. Size about 5^. Starch rather plentiful. Thunbergia fragrans Roxb. {Acanthacece.) Dry seeds. — Grains spherical, oval, or somewhat irregular. Size about lOju. Many seem to be compound. Some show colorless appendages which are not stained after treatment with iodine. The grains are poor in starch, but rich in oil; the former may disappear at maturity. Delphinum ajacis Linn. (Ranunculacece.) Fresh unripe seed coats. — Grains rounded, or angular with round angles. Size about 8yu. Starch rather plentiful. Chelidonium majus Limi. {Papaveracece.) Fresh seed coat. — Grains rounded. Size about 7 to 9;u. They are stained brown when treated with iodine. Chelidonium majus Linn. {Papaveracece. ) Fresh unripe seed coat. — Grains rounded. Size 5 to 6^- They are stained brown or violet when treated with iodine. Brassica napus Limi. {Brassicaceoe; Cruciferce.) Fresh unripe seeds. — Grains (in the embryo) rounded or oval, frequently somewhat angular; some are evidently separated-grains. Size about 5/u. Compound and separated-grains predominate in the seed coats (see type 14). Starch is plentiful in the perisperm, especially rich in the embryo, and mostly found in the cells of the seed-coats with the exception of the outermost layer, the walls of which soon thicken. The perisperm later entirelj- disappears, the seed-coats and the embryo lose their starch, since in the former the membranes are thickened and in the latter starch is replaced by oil. As the green seeds turn yellow the solution of the staixh takes place; in the brownish-yellow seeds it has entirely disappeared. In the earliest stage the embryo contains merely oil ; when it has entirely reijlaced the perisperm and thus completely filled up the cavity within the seed coats both oil ami starch are found, while in the latest stages oil alone is again observed. 250 DIFFERENTIATION AND SPECIFICITY OF STARCHES. Cucumis salivis Linn. {Cucurhitacece.) Fresh unripe cotyledons. — Grains rounded, rarely angular. Size about 5 to G/i. Starch rather plentiful. Vatica robusta Steud.; Shorea robusta Roxb. (Dipterocarpacea:.) Dry cotyledons. — Grains oval, elliptical, shortened-conical, rarely rounded-oval to oblong; occasionally somewhat angular, two-fifths to two-thirds as broad as long; the larger ones frequently have a cavity or a slit. Length about 11/^, thickness about Sfi. These grains may belong to the eccentric-conical type (tjTe 7). Thea bohea Linn. (Camelliacew.) Dry cotyledons. — Grains spherical or roundetl-oval, rarely oval; instead of a hilum there is a small central cavity with radial fissures; the cavity is usually central, rarely one-half to two-thirds eccentric. Size about 16 to 19yu. The starch probably belongs to the eccentric-conical tyj^e. Some compound grains of few unequal components are also observed as in type 14. Calophyllum lanceolatum Blume. (GuttifercB.) Dry cotyledons. — Grains spherical or slightlj' angular; majority with a small central cavity. Size about 10 to 12/i. Some compound grains of 2 to 3 equal components are also observed. Calophyllum tacainahaca Willd. {Guttiferce.) Dry cotyledons. — Grains mostly rounded. Size about 6^. Isolated larger, oval ones. Size about 13/u or more. Starch appears to be in the process of solution; oil very plentiful. Acer laurinum Hook. (Lapindacece.) Dry cotyledons. — Grains rounded or oval. Size 8 to lOyu. Many are distinctly separated-grains with one curve and 1 to 3 pressure facets, others appear to be simple. Many of them are probably in the process of solution, so that the stai-ch may disappear at maturity of the seed; oil plentiful. Acer pseudoplanatus Liim. (Lapindacece.) Fresh, unripe, still very green cotyledons. — Grains rounded, rarely oval, usually more or less angular. Size about 5 to 7ju. Starch rather plentiful. Banisteria sp. {Malpighiacece.) Dry cotyledons. — Grains rounded to almost polyhedral. Size about 5jx. The starch-grains lie singly or very few in the cells, which are filled with plastids of rounded, angular shape. Aleurites moluccana Willd. (Euphorbiacece.) Dry seeds. — Grains spherical, sUghtly angular, or poly- hedral; a small central cavity is found in the larger ones. Size about l/x. Aleurites sp. {Eiiphorbiacew.) Dry cotyledons. — Grains spherical to oval, sometimes slightly irregular or angular. Size about 11/j. Starch not very plentiful, and embedded between plastids. The albumen adhering to the seed coats was not examined. Ochna lucida Lam. (Ochnacem.) Dry cotyledons. — Grains rounded, oval, elliptical, many somewhat angular; half to as thick as long; the larger ones have a small cavity. Length about 5 to 6yu, thickness about ifi. The seeds are probably not fully developed. Protoplasm and rounded or angular plastids were found along with the starch. Fatty oil seems to be wanting. Ochna squarrosa Linn. {Ochnacece.) Dry cotyledons. — Grains rounded, oval, frequently angular; the bi'oad ones slightly compressed; the larger ones have a small central cavity. Size about 7m. Some oil is present. Ammannia latifuUa Linn. (Lythracece.) Dry seeds. — Grains spherical, or almost so, with small cen- tral cavity from which single fissures occasionally radiate. Size about 10^. Some doublets are also observed. Besides the oil the seeds contain varying amounts of starch, or some- times none. Ammannia vesicatoria Roxb.; Ammannia baccifera Linn. (Lythracem.) Dry seeds. — Grains rounded or oval, occasionally somewhat angular or irregular; the broad ones slightly or not at all compressed; with a small, almost central, cavity, and some single radial fissures. Size about 10 to 13/i. Some doublets and triplets are found among the simple grains. The seeds of both species contain much oil. Starch, which occurs in varying quantities, is proljably found in special cells. On pressing the seeds the oil comes out first, and later the starch. Galega biloba Sweet. {Leguminosw.) Dry cotyledons. — Grains rounded or oval, sometimes irregular. Size about 4 to 5^. Starch is not very plentiful, in addition much protoplasm and little oil is present. Arachis hypogcea Linn. {Leguminosw.) Dry cotyledons. — Grains spherical, with a small central cavitj'. Size about 12//. Much oil is also found. TYPE 11. GRAINS SEMI-COMPOUND. 251 Cuilandina boiidiic Linn. {Lcguminoscc.) Dry cotyledons. — Grains spherical, many with a small central cavity; in the larger ones single, delicate, short fissures are occasionally observed. Size a])0ut 7yu. Acacia inelanoxylon R. Br.; Acacia latifolia Desf. (Leguminosce.) Dry cotyledons. — Grains spherical or rounded-oval. Size about 4;u. Much oil and protoplasm are also found. Type 11. Grains Semi-compound. Several components are enveloped completely or merely upon one side by a common substance which is not pierced by partitions. This common substance, which belongs to the originally simple grain as well as the components, may or may not have lamellte. The comiionents usually arise by the splitting of the hilum, and develop into the type of the simple grains. More or less delicate lines of division are found between the components, which extend to the common lamellae. Some- times such partitions are wanting, and should there also be no lamelliB then the semi-compound structure is indicated only by the presence of several hila lying in a homogeneous mass. Chara stelligera Bauer. {Algoe.) Dry star-shaped bodies. — Grains rounded, oval, irregularly blunt- angular with 1 to 5 angles; usually isodiametric ; sometimes elongated to twice as long as thick and occasionally compressed to about two-thirds of their width; or slightly cunei- form; with 2 to 25 or even 40 hila which usually have the appearance of small cavities, and are sometimes embedded in an apparently homogeneous mass, and also appear singly or in a group surrounded by distinct lamella}. Lines of division between the components usually absent. Size about 70 to 85;u. Among the inclosed grains a central one frequently is found which far exceeds the others in size. This one is surrounded by distinct lamellae, is almost spherical in shape, and contains one or more hila near its center. Several short radial fissures are usually found in the interior of these larger grains. Single parts of many grains are cut off by delicate lines. This structure forms a transition to the true compound type, occurring amoi\g the semi-compound ones (see type 15). Marsilea pubescens Tenore. {Marsilacece.) Dry gymnospores. — Grains oval or oblong, sometimes slightly irregular; 2 to 4 times as long as broad; the broad ones compressed to three-fifths of their width; with 2, rarely 3 and 4, hila or components placed in one row; components appear oval from the broad aspect, and lanceolate or linear-lanceolate from the narrow aspect; the axes and the largest planes of which coincide with those of the whole grain; single fissures occasionally are observed which may be between the components, or may pass through the centers of the grains, and thus either coincide with the longitudinal axis or even cut it at right angles; with distinct lamellae in the common surrounding substance, and in the outer substance of the components. Length about 175;u, breadth 52/i, thickness 33/:i. Among the above some simple centric-oval types are found. Hyadnthus orientalis Linn. {Liliacece.) Fresh scales of the bulbs. — Grains rounded to oblong, usu- ally irregular, with more or less numerous protruding angles; two-fifths to as broad as long; very often slightly compressed and cuneiform, the shorter border thickened, the opposite lunger one with a sharp edge; lamelke are occasionally observed which are chiefly found toward the sharp edge; with 2 to 7 hila arranged in a single row along the thickened border; incomplete fissures are frequently seen between the hila; small components are sometimes found in the angles or at the sharp-edged margin. Length about 45|i, breadth about 35m. Among the above are found some simple, eccentric, cuneiform grains, and also compound grains with components arranged in 1 or 2 rows. In the young bulbs which are still inclosed within the scales of the older bulbs the grains are very small; in fact, they frequently show a Brownian movement. They are spherical, frequently somewhat irregular in shape, and without lamellae ; a broad, somewhat sharpened edge can be noticed in .some of the grains, and occasionally one of the corners is cut off. Size about 10 to 15^. Among the above many comjioimtl gi'ains of 2 to 8 components with delicate lines of division are observed. Last year's bultjs which have borne this year's leaves and buds contain at the base of the inner scales similar but larger compound grains. Separated-grains are wanting. The simple grains have developed still more irregularly, and by splitting of the hilum have changed into semi-compound ones, and by the breaking away of corners into compound grains of unequal components. Semi-compound grains were found almost exclusively in the scales 252 DIFFERENTIATION AND SPECIFICITY OF STARCHES. at the top of these biilljs. Complex compound grains and separatcd-gi'ains of the same do nut occur; tliey are either not formed at all or they fall to pieces and the separated-grains then look like the simple ones. Fissures usually radiating from the hilum are formed in grains in bulbs which have been allowed to dry over the winter, and the divisions between the components become wider. The grains in the outer, decayed scales of the bulbs are under- going disintegration, which usually begins where the partitions between the components touch or nearly touch the surface. Canna lanuginosa Rose; Canna lagunensis Lindl., etc. {Cannacece.) Fresh root-stock. — Grains simple, usually more or less mussel-shell-shaped; hilum very eccentric, lying in the upper, narrow, protruding end. In this species as well as in others the hilum is frequently split in two; 2, rarely 3 or 4, hila lying next to each other. In the longitudinal axis of the grains of some species a single and here and there a double row of inclosed components is formed. In the symmetrical grains this row is usually in the median line, while in the unsym- metrical ones, the grains being usually more or less curved, it is found toward the con- cave border. The number of inclosed grains varies from 2 to 12; either by transverse or longitudinal fission of these grains individual ones arise, in which lamellae become visible as they increase in size. Ficaria ranunculoides Moench.; Ranunculus ficaria Linn. (Rammculacea.) Dry thickened roots.— Grains rounded-cuneiform, rounded-triangular, frequently unequally quadrangular, some- times pentagonal, frequently somewhat irregular; about twice as broad as long; compressed to half their \vidth; in the triangular grains one of the rounded angles and in the quadrangular ones the naiTOW side is thickened; the latter with a row of 2 to 6 or more hila at right angles to the longitudinal axis; eccentricity one-fourth and one-sixth; the distal margin broad, with a thiimed knife-like edge, which terminates mostly on either side in rounded or some- what pointed angles; lamellae none or indistinct; instead of the hila, small cavities, fre- quently with single delicate, short, radiating fissures are found; delicate separating fissui-es between the components are rare. Size about 35ai. Among the above some doublets and triplets and simple eccentric-cuneiform grains are observed. Cereus variabilis Pfeiff. (Cactacece.) Fresh pith of the stem. — ^Grains rounded or oval, usually of irregular form and with lamellae; with 2 to 6 and 9 larger or smaller inclosed components, the latter when larger have lamellae, and are usually divided from one another liy parti- tions; lamellae eccentric; occasionally some of the larger components are also semi-compound. Size about 80jn. Among these are simple grains of uncertain structure. Type 12. Grains Compound, with Fused Components. The components are not surrounded by a common substance, nor separated from one another by lines of cleavage. At first glance the grains of this type bear a resemblance to certain semi- compound grains, but differ from them in not possessing a common surrounding substance. They also differ from the ordinary compound grains in the lack of lines of cleavage between the com- ponents; and for this reason they do not, as a rule, split into separated-grains. In Commelina and other species, compound grains arise by grains which were originally separate, and pressing upon one another as the result of further growth, so that at first grains in one cell of the same generation, and later those of older ones, fuse with one another, and at the same time the clefts between them disappear. The whole cell lumen is then filled with a uniform, reticulated, parenchymatous mass, in which the hollow space represents the soft, internal matter of the components, while the net- like framework corresponds to the coalesced dense external substance. Commelina ccelestis Willd. (Commelinacece.) — Compound grains polyhedral, due to pressure; 1 to 2.5 times as long as thick, of parenchymatous structure coiosisting of 2 to more th an 200 com- ponents, the latter usually fused with one another, rarely separated by delicate clefts, each with a large angular cavity. Separated-grains (which occasionally can be set free by press- ing) rounded-angular or polyhedral. Size of the com])ound grains about 40 to 55^, thick- :iess about SOfi. Size of the comi^onents 1.5 to G and 8/^. Some isolated simple spherical grains. Size about d/x. In semi-ripe seeds the compouml grains are sjjherieal to oblong (about 44yu) , consisting of apparently solid, equally dense components separated by delicate lines or distinct, narrow clefts. In youug seeds, where the endosperm has just begun to TYPE 12. GRAINS COMPOUND, WITH FUSED COMPONENTS. 253 develoiJ, the cells forming a 3 to 4 layered covering for the embryo-sacs are comijletcly filled with small starch-grains. The latter are usually spherical, some distinctly compoimd, and some only granular or almost homogeneous. Some diminutive separated or simple grains also occur. Commelinfi n^idicaulis Burm. {Commelinacea.) Dry ripe seeds. — Grains as in Commelina cmlestis Willd. In half-ripe seeds the compound grains are spherical-oval, elliptical or oblong. Size 28 to 34/i. The components of the same are solid, and consist of a uniformly dense sub- stance; they are separated by lines or by distinct clefts which contain water; in this stage the compound grains easily fall apart, thus giving rise to numerous obtuse-angular or poly- hedral separatod-grains. Tinnaiia fugax Scheidw. (CommcUnacece.) Dry endosperm. — Starch as in CommeUmi. Compound grains about once to twice as long as thick, consisting of 2 to more than 1700 components. Size about 5Qix, thickness about SG/lj. On pressing the grain, rounded or polyhedral separated- grains are set free, often still hanging together in irregular masses, and are usually 2 to Zix, rarely 4 and 5/j, in size. Some cells seem to be filled with a homogeneous mass. Cyanotis cristata Don.; Tradescanlia mrginica Linn.; Heterachtia pulchella Kze. (Commelinaceoe.) Dry endosperrn. — Grains as in Commelina. Zingiber officinale Rose; Amomuni zingiber Lirm. (Zingiberacew.) Dry seed endosperm. — The cells are densely filled with components which usually represent a uniform mass, showing no indication of the compound grain. Through the complete union of the components this mass appears either reticular parenchymatous by complete fusion of the components, since only the hollow spaces lying in the apparently homogeneous substance are visible, or granu- lar, in which case the components may be separated from one another. The compound grains, rounded or oval grains within the cell, can rarely be seen or set free. Size of com- ponents 1 to 5ii. The smaller components are solid, the larger ones are hollow. Amomxim cardamomum Linn.; Amomum javanicuni. (ZingiberacecB.) Dry seed endosperm. — The cells usually filled with a uniform parenchymatous network which often falls out of the sectioned cells in a single mass, and in which either indistinct or no divisions corresponding to compound grains may be noticed. This mass frequently has a granular appearance and easily falls apart by means of pressure into separated-grains, of which the smaller ones are frequently rounded and the larger ones polyhedral, both having a large or a small cavity. Size of the separated-grains 1.5 to 5.5ju. Amomum granum-paradisi Afzel; Elettaria cardamomum White; Cardamomum minus. {Zingibera- cecB.)— Grains as in preceding. Hedychium gardnerianum Wall. (Zingiberaccm.) Dry seed endosperm. — Compound grains, spherical, oval, rarely somewhat angular; slightly granular; containing about 8000 components. Size about 21«. Separated grains rounded. Size 0.7 to 'I/x. These grains are transition forms to ordinary compound grains of many components. Costus sp. (Zingiberacece.) Dry seed endosperm. — The separated-grains, which are 1 to 3 and 4ju in size, are packed in the cells as a dense uniform mass in which the structure of the compoimd gi'ains can rarely be recognized. This mass sometimes appears granular, but more often has a reticular parenchjonatous appearance, the meshes of which are frequently arranged in parallel rows. The endosperm cells are distinguished by numerous knob-like processes. Thalia dealbata Fras. (Zingiberaceoe.) Dry seed endosperm. — Compound grains, polyhedral as a result of pressure, \vith sharp angles and edges; frequently more or less irregular; about ti^ace as long as broad; with 2 to 12 rounded or angular, more or less distinct, cavities lying in a homogeneous mass, giving to the grain a parenchymatous appearance; hues of divisions are seldom noted between the cavities. Size of the compound grains about 20 to 25/i. The cavities in the outer denser part of the endosperm are small (1 to 2/t) and rounded; in the iimer part they are large (about 8m) and angular. Maranta sp. {Marantacece.) Dry seed endosperm.- — Compound grains consisting of 2 to 6, rarelj^ 8, almost equally large components, but usually of irregular structure and irregularly disposed. The components are more or less fused, with indistinct or no lines of cleavage, each component containing a rather large but frequently very indistinct cavity. Size of the compound grains 9 to 13m, and of the components 2 to o/i. These grains seem to come between those of Thalia 254 DIFFERENTIATION AND SPECIFICITY OF STARCHES. and those constituting type 14. Some simple tornloid or indented grains also occur similar to those in Maranta raniosissima (p. 229), but smaller (about 11 to 14/i), and only slightly compressed. Heliconia sp. (Musacece.) Dry seed endosperm. — Grains rounded-oval to oblong, sometimes curved, often irregular, frequently somewhat toruloid; one-fourth to almost as broad as long; the broad ones slightly compressed. Some appear to be simple grains; in the majority of them several (2 to 12 or more) cavities are noted. Length about 17/i, width about lOjit. These grains stand between those of the Maranta species and Thalia dealbata. Piper. {Piperacece.) Seed endosperm. — Besides the ordinary compound grains wth numerous components, there are some whose components are completely fused, as in type 16. Type 13. Grains Compound, in One or Two Rows. Components 3 to 11; form a compound grain; arranged in 1 or 2 rows separated by clefts, and at maturity falling away as separated-grains. These compound grains arise in two ways, either by the repeated splitting of the hilum of an originally simple grain, or by the fusion of a number of originally separate grains lying next to each other in an elongated chloroplast (as in Chara). In the latter case they may for some time partially or entirely retain their green color. Chara hispida Linn. {Algw.) Fresh cells of the internodes. — Compound grains oblong or rod-shaped; one-fifth to one-half as broad as long; usually slightly compressed; consisting of 2 to 8 almost equal components. Length about 34;^, ^vidth about 9/i. Components in two rows ^vith over- lapping ends; two-fifths to two-thirds as broad as long; without lamellte. Length 8 to 16^. The starch-grains which originate in the chloroplast are originally green and later become colorless. Hyacinthus orientalis Linn. (Liliacew.) Fresh scales of the bulbs. — Compound grains oval to lanceo- late; one-half to one-third as broad as long; consisting of 2 to 8 imequal comjjonents. Length about 35yu. Components are arranged in one or two rows; lamcUiB rare, delicate, usually with a small central hilum; also simple cuneiform and semi-compound grains are found. Carina lagunensis Lindl. {Cannacece.) Fresh root-stock. — Compound grains, elongated to linear, usually more or less curved; 3 to 8 times as long as broad; consisting of 3 to 11 usually un- equal components arranged in 1 and sometimes in 2 rows. Length about 7yu. Components with distinct or indistinct concentric lamellae, and either an almost central hilum or occa- sionally with several hila, which are undergoing longitudinal transverse fission. Size 3 to 18/i. Among these are .simple, eccentric compressed grains and semi-compound grains, as well as all the intermediate transitional stages between these two. Canna pedmiculata Sims. (Cannacece.) Fresh root-stock. — Compound grains oblong or lanceolate, usually irregular, frequently curved; 2 to 6 usually unequally divided grains often arranged in one row. Size of smallest components are 2 to 3 and of the largest 30/i.; in the latter the lamellae are sometimes distinct and the hilum very eccentric. Also simple, eccentrically compressed grains are found. Canna lanuginosa Bosc. (Cannacea;.) Fresh root-stock. — Compound grains oblong or linear-lanceo- late, usually irregular, frequently curved, occasionally compressed; 2 to 8 equal or unequal components arranged in one row. Among these simple and semi-compound grains are found. Canna coccinea Ait. (Cannacece.) Fresh root-stock. — Compound grains oblong or linear-lanceolate, usually irregular and frequently curved; 2 to 6 to 8 equal or unequal components arranged either in one or rarely in two rows. Simple grains and all transitions to the semi-compound ones are found. Fagopyrum esculentum; Polygonum fagopyrum Limi. (Graminaceoe.) Fresh and dry seeds. — Compound grains rounded, elongated, rod-shaped, sometimes curved or bent; frequently irregular; angular or lobulate; 1 to 8 times as long as broad; the narrow ones terete, the broad ones compressed; consisting of 2 to 15 components which are arranged in one or two rows, rarely in more rows, or even in a simple layer. Length about 38^^, width about 20ai. Components are homogeneous, or each have a central hilum. Size 4 to 10/x. Simple centric spherical grains are also observed. type 14. grains compound, of few components of equal size. 255 Type 14. Grains Compound, of Few Components of Equal Size. Two \o f(>ii or more components of about equal size, united into one compound grain, separated by fiissuro!<, and at comijlete maturity breaking away as sejjaratcd-grain.s with one curved surface and one or more pressure facets. The compound grains originate bj- the division of the hilum and by repeated division of the same, and pass over into the compound type with many components, as in tj'pe 16. These occur ran4y alone, but are frequently mingled either with simple grains or with compound ones belonging to type 15 which are formed by their angles or edges being cut off. Cycas circinalis Linn. (Cijcadacecc.) Pith. — According to Payen (Ann. Sc. Nat., 1838, ii, p. 18 pi., 6, figs. 4, 5), the compound grains consist of 2 to 8 and 10 equal and regular components. Separated-grains with one curved surface and one to seven pressure facets; with indistinct lamella?; hilum about one-fourth eccentric. Size about 45^. Commercial sago changes slightly in moist heat. Size of the starch-grains about TOju. Starch-grains in the base of the petiole similar to those in the pith of the stem. Coix lacryma Linn. {Grayninacece.) Dry roots. — Compound grains, rounded or oval; consisting of 2 to 4 and 8 almost equal components. Size about ll^i. Separated-grains 2.5 to 5m. Also some simple spherical grains are observed. Size about 7^. Starch quite plentiful in the cortex of the root. Panicum arenarium Brot.; Panicum repens Linn. {Graminacem.) Dry root-stock. — Compound grains, consisting of 2 to 4 equal, rarely unequal, components (only a few that have fallen to pieces are present). Size about 20 to 26^. Separated-grains rounded or oval; homogeneous or with a small cavity instead of a hilum; about one-fourth eccentric. Size about 13yu. Oplismenus colonus Hiiinb. and Kunth.; Panicum colonus Linn. (Graminacece.) Dry root-stock. — Separated-grains rounded or angular with rounded angles, usually compressed. Size about 6m . Some are distinctly .separated-grains -n-ith one curved surface and 1 to 6 pressure facets. The compound grains, of which none are any longer present, may have consisted of 2 to 10 components. Poor in starch; starch entirely wanting in the roots. Vilfa pungens; Sporobolus pungens Kunth. (Graminacece.) Dry stolons. — Grains compound, con- sisting of 2 to 10 and 12 almost equally large components. Size about 14^. Separated- grains 3 to 8m; the larger ones have a small cavity, occasionally with single delicate radial fissures. Cell-walls very thick, and with numerous indentations. Cynodon dactylon Pers. (Graminacece.) Dry stolons. — Grains compound, consisting of 2 to 7 and 9 equal, rarely unequally large components. Size about 12 to 16m. Separated-grains nearly rounded, 3 to 8m; the larger ones hollow. Andropogon muricatus Retz.; Anatherum muricatus Beauv.; Radix iwarancusoe. (Graminacem.) Dry roots. — Grains compound or 2 to 4 equally triangular or tetrahedral components. Size about 25m. Separated-grains conical, with blunt, slightly protruding ends; 0.75 to 1.33 times as broad as long; lamellae none; toward the narrow aspect instead of a hilum a small cavity is occasionally found with 1 to 2 radiating fissures; eccentricity about one-fourth. Length 7 to 13m, width 8 to Hm- Also a few simple, spherical grains are observed. The grains described by Schleiden (Grundzuge, 3 Aufl. i, 185, fig. 15) had been changed by heat. According to Berg (Pharmacognosie), Radix iwarancusce contains no starch. In numerous specimens obtained from various sources, Nageli always found starch plentiful in the pith, but none in the cortex. Sdrpus maritimus Linn. (Cyperacece.) Dry root-stock. — Compound grains spherical or oval, con- sisting of 2 to 10 usually equal and regularly disposed components. Size about 18m- Sep- arated grains mostly have a small central cavity from which, in the larger grains, several fissures radiate. Size 5 to 14m. Also some simple spherical grains, with small central cavity and single radial fissures. Size about 15m. Carex maxima Scop. (Cyperacece.) Dry root-stock. — Compound grains rounded, oval, usually more or less irregular; consisting of 2 to 10 and 14 usually equal components. Size about Qm- Size of the separated-grains 1.5 to 4m. Also some simple rounded and oval grains are found. Carex bicolor All. (Cyperacece.) Dry root-stock. — Compound grains rounded or oval, frequently somewhat irregular, consisting of 2 to 8 and 10 usually almost equal components. Size about 7m. Size of the separated-grains 1 to 3m- Simple grains of incomplete formation are also found. 256 DIFFERENTIATION AND SPECIFICITY OF STARCHES. Willdcmmnn terca Tluinih. {Rcsliovacccn.) Dry roots. — Soparated-grains with one cur'i'Cfl and 1 to 5 pressure facets, occasionally with a central cavity. Size al)out lOp. (^nly a few compound grains of 2 to 4 components arc still present. The greater ninnber have fallen ajiart. Judg- ing from these separatcd-grains they must have consisted of at least 8 components. Some simple spherical and rounded-oval grains are also found. Reslio incurvatus Thumb. (Restionacem.) Dry root-stock. — Grains compound spherical or oval, con- sisting of 2 to 10 and 20 unequal or chiefly equal components. Size about 14/li. Separ- ated-grains 2 to 8m; the larger ones have a small cavity. Luzula spadicea DC. {Jiincacece.) Dry root-stock. — Compound grains consisting of 2 to 8 and 12 usually almost equal components. Size about 13/j. Separated-grains 2 to 6^. Colckicum autumnale Liim. (Colchicacew.) Fresh tubers. — Grains compound, of 2 to 4 usually ecjual components, usually arranged in a tetrahedron, rarely 3 in a row. Size about 28/x. Sep- arated-grains without lamellce; with distinct hilum; occasionally, instead of the hilum, asmall cavity from which 2 fissures extend toward the edges. Size 3.8 to IS/x- Also a few simple spherical or flattened spherical grains are found. Colchiaini variegatum Linn.; Radix hermodaclxjli (Colchicacece.) Dry tubers. — Compound grains, spherical, oval, or oblong; 2 to 8 usually equal components. Size about GOaj. Separated- grains without lamellae, and with distinct fissures radiating toward the corners. Size about 30iU. Some simple spherical and spherically-oval grains are found. Monochoria plantaginea Kunth, PI. Ind. or ed. Hohenacker. (Pontederiacece.) Dry root-stock. — Compound grains, rounded or oval; consisting of 2 to 10 and 16 usually almost equal com- ponents. Size about 12yu. Separated-grains 2 to 8yu; the larger ones have a central cavity. Smilax sp. {Liliacece.) Dry sarsaparilla root. — Compound grains spherical, rarely oval or oblong; consisting of 2 to 8 usually equal comiionents, which are either arranged as 3 in a triangle (rarely in a row), 4 in a tetrahedron (rarely in a square), or 6 in a hexahedron. Size about 23fi. Separated-grains without lamellae; instead of a hilum usually a small almost central cavity is observed with frequently 2, rarely 4, radial fissm-es. Also simple, spherical grains with small central cavity and radial fissures are rarely observed. Size 16/u. Smilax china Linn. ; Radix chirm. {Liliacece.) Dry root-stock. — Compound grains more or less angular as a result of pressure, frequently polyhedral, consisting of 2 to 4, rarely 6, equal components. Size about 60a£. Separated-grains polyhedral, rarely with a few distinct lamcUiE; instead of a hilum a small central cavity is found, from which several fissures usually radiate. Size 15 to 30/i. Also simple grains with incomplete formation. Tacca pinnatifida Forst. {Taccacece.) Tahiti arroivroot. Root-stock. — According to Walpers (Bot. Zeit., 1851, p. 333), the compound grains probably consist of 2 to 6 components. Separated- grains with 1 to 4 pressure facets; indistinct lamellae; hilum eccentric, with one well-marked transverse fissure, as in Maranta arundinacea. Size as in the latter or somewhat smaller. According to Leon Soubeiran (Journ. Phar., 1854, xxv, 180) the grains are spherical, often cut ofT vertically to the longitudinal axis, elliptical, occasionally somewhat pear-shaped; some have stellate fissures in the place of the eccentric hilum. Size varies from 30 to 40^, some grains, however, are not over lOju. (Judging from the illustration the larger grains appear to be simple grains, not separated ones.) Meristostigma silenoides Dietr. (iridacece.) Dry bidbs. — Grains compound, consisting of 2 to 10 and 12 usually equal-sized components which are regularly arranged. Size about 42^. The larger separated-grains have a small cavity from which single short fissm-es may radiate; one-half eccentric in oval grains. Size 2 to 18^. Some simple spherical or oval grains are also observed. Size about 20/:. Gladiolus communis Linn. {Iridacece.) Dry bulbs. — Compound grains spherical or oval; 2 to 6 and 8 equal or somewhat unequal components which are usually regularly arranged. Size 7 to 28;u. Separated-grains, size 2.5 to IS/i. The larger ones mth a small cavity from which strongly marked fissures radiate toward the edges . Trichonema bulbocodium Ker. {Iridacece.) Dry bulbs. — Compound grains consisting of 2 to 10 and 12 usually equal components, which are regularly arranged. Size 28/^. Separated-grains, size 2 to 14/Li; the larger ones have a small cavity, instead of a hilum, from which occasionally single short radial fissures radiate toward the edges. Some simple spherical grains are also present. Size about 16^. TYPE 14. GRAINS COMPOUND, OF FEW COMPONENTS OF EQUAL SIZE. 257 Crocus sativus All. (Iridacecc.) Dry bulbs. — Compound grains, of 2 to 8 and 10 usually equal com- ponents, which are regularly arranged. Size aliout 38yu. Separated-gi'ains, size 2 to 18/i, the larger ones have a small cavity, instead of a hilum, from which long, delicate fissures radiate toward the edges. Anigosanthus ritfa Lahil. (HwmodoracecB.) Drij root-stocks. — Compound grains, of 2 to 4 and 6 usually equal components. Size about 18;u. Separated-grains of 3 to 10//; frequently with a small cavity. Billbcrgia amcena Lindl. (Bronieliaceas.) Dry root-stocks. — Compound grains of 2 to 4 usually equal, sometimes unequal, components. Separated-grains, size 5 to I5fi; instead of a hilum a small cavity is observed from which short, distinct fissures radiate; in the larger oval separated-grains the hilum is about one-third eccentric. Some simple eccentric-conical grains arc also found. Maranta iiidica Rose. (Cannacem). West India arrowroot. — According to Schleiden (Grundzuge, 3 Aufl., I, p. 185, fig. 13), the compound grains consist of 2 to 4 almost equal components. He also finds similar grains in other species of Maranta, and Miinter (Bot. Zeit., 1845, p. 203) records the same for Maranta bicolor Ker. {Calathea bicolor Steud.). According to Walpers (Bot. Zeit., 1851, p. 334), the grains of Maranta indica are either oval, drum-shaped, drum-shaped with inclined sides, or almost pear-shaped, as well as with 2 to 3 glandular- connate divisions. These grains, Nageli states, should not be confused with those of any other kind of arrowroot, as they must surely come from a plant different from those described by Schleiden and Miinter. Sparganium ramosum Huds. (Typhacew.) Dry root-stock. — Grains consisting of 2 to 6 and 12 usually equal components. Size about 17/j. Separated-grains usually with a small cavity; one-third eccentric in some oval forms. Size 3 to lO/x. Also isolated simple, oval grains, hilum about one-fourth eccentric, are found. Size about 13/i. Aponogeton. (Alismacew.) Tubers. — According to Schleiden (Grundzuge, 3 Aufl., i, p. 185) the com- pound grains consist of 2 to 4 almost equal-size components. Parietaria diffusa Mert and Koch. (Urticaceoe.) Dry root-stocks. — Grains consisting of 2 to 4 almost equal components. Size of the separated-grains 3 to 6^- Some simple grains, of undeveloped structure, are also found. Rheum undulatum Linn. (Polygonacece.) Fresh root. — Compound grains rounded or angular with rounded regular or irregular angles; 2 to 8 equal or unequal components which are often irregidarly arranged. Size about 25/i. Separated-grains obtuse-angled to sharply poly- hedral; lamella? rarely visible, very few (1 to 4), and delicate; hilum usually distinct, central, or slightly eccentric. Size 4 to 20^. Simple spherical grains rare. Rheum sp. {Polygonacece.) Rhubarb root. — Compound grains consisting of 2 to 5 components. Separated-grains -ivith obtuse or acute angles; lamelhe none; instead of a hilum a small cavity is found which is central or sUghtly eccentric, from which 2 to 5 short radial fissures pass out. Size 3 to 17/n. Some simple, almost round grains are also found. Rumex arif alius All. (Polygonacew.) Dry root-stock. — Compound grains rounded to elliptical; 2 to 8 equal or unequal components which are often regularly arranged. Size about 12/^. Separated- grains 2 to 6m; a small cavity is found in the larger ones. Also simple grains of undeveloped structure. Rumex tubo-osus Liim. (Polygonacece.) Dry root-stock and thickened roots. — Compound grains con- sisting of 2 to 6, rarely 10, equal or unequal components, sometimes several small components adhere to 1 or 2 larger ones. Size about 26^. Separated-grains ^\^th a small cavity from which single fissures radiate toward the edges; eccentricity about one-half. Size 3 to 14^ and 17/(. Swollen, hollow separated-grains, size 22yu. The grains in the thickened roots are generally somewhat larger than those in the root-stock. Boerhavia repens Linn. (Nyctaginacece.) Dry root. — Compound grains spherical to oblong and tabular; 2 to 10 and 14 usually equal components, which are regularly arranged. Size about 30«. Separated-grains usually isodiametric, with a small cavity from which long, delicate fissines radiate toward the comers; eccentricity about one-half. Size 3 to 20,u. Cinrmmomum ceylanicum Nees.; Cortex cinnamomi interior. (Lauracece.) Cortex. — Compound grains consisting of 2 to 4 almost equal components. Size about 14^. Numerous spherical and angular grains with rounded comers, the former being simple and the latter separated- grains. Size about 9ix. A small central cavity is found in the large grains. 17 25S DIITERENTIATION AND SPECIFICITY OF STARCHES. Aristolochia scrpcntaria Linn. {Arislolochiaccm.) Pith of dry root-stock. — Compound Rr.ains sjilicrical or oval-spherical, 2 to 8 usually equal components, arranged regularly in triangles, tetra- hedrons, hexahedrons, or octahedrons. Size about 18^. Separated-grains without lamell®, hilum, and fissures. Size 3 to 12^. Simple spherical homogeneous grains are also present. Size about 12//. Aristolochia pistolochia Linn. {Aristolochiacece.) Dry roots. — Compoimd grains spherical, rarely elliptical, consisting of 2 to 8 and 12 usually equal components which are regularly arranged. Size about 18yu. Separated-grains 3 to 11/u; the larger ones have a cavity. Some large sep- arated-grains swollen and apparently hollow; and some simple, rounded ones are also ob- served. Valeriana officinalis Linn. {Valerianacem.) Fresh root-stock and roots. — Compound grains consist- ing of 2 to 4 equal components which are arranged regularly in triangles or tetrahedrons. Separated-grains sometimes with a distinct central hilum. Size about 8/x. Also simple grains of incomplete formation are observed. Valeriana tvherosa Lirm. (Valerianacece.) Dry root-stock — Grains consisting of 2 to 4, rarely 6, almost equal components. Size 22/i. Separated-grains with a large or a small cavity from which fissures radiate toward the corners. Size 4 to 15ju. Also some simple grains M-ith structure undeveloped. Valeriana saliunca All. {V alerianacem.) Dry root-stock. — Compound grains sjjherical to oval and elliptical, with unequal sides, consisting of 2 to 6 and 10 equal or somewhat unequal com- ponents. Size about llju. Separated-grains 2 to S/j, the larger ones with a small cavity. Also simple grains with imperfect structure. Succisa 'pratensis Moench. (Dipsacaccce.) Dry root and root-stock. — Compound grains spherical to oval and elliptical; consisting of 2 to 10 and 16 components which may be equal, though more frequently unequal. Size about 20/i. Separated-grains have a small or large cavity, from which short, delicate, radial fissures very often pass out, usually almost central, rarely one-half eccentric. Size 2.5 to 10 and 14/j. Vinca minor Linn. (Apocynacece.) Dry root. — Grains consisting of 2 to 4 and 0 almost equal, rarely unequal, components. Size about 25^1. Separated-grains somewhat oblique; instead of a hilum usually a small central cavity is observed from which delicate fissures radiate. Size about 12/j. Simple, usually eccentric-conical grains are also found. Swertia perennis Linn. {Gentianacece.) Dry root-stock. — Grains consisting of 2 to 4 and G usually almost equal components. Size about 12/j. Separated-grains rarely have a small cavity. Size 3 to 7/i. Simple grains of incomplete formation are also observed. Tcucrium hyrcanicnm Linn. (Lahiataccw.) Fresh unripe pericarp. — Grains rountled, usually angular; size about Sfi; majority probably separated-grains. Some compound grains of 2 to 6 com- ponents are also found. Omphalodes verna Moench. {Boraginacem.) Dry root-stock. — Compound grains roimded or oval, consisting of 2 to 4 and 6 equal or rarely unequal components. Size about 9yu. Separated- grains have occasionally a small cavity. Size 2 to 5^. Simple grains of incomplete formation are also observed. Starch not very plentiful. Symphytxim tuberosum Linn. (Boraginacece.) Dry root-stock. — Grains consisting of 2 to 4 usually equal components. Separated-grains have a small central cavity from which sometimes several short, delicate radial fissures emerge. Size about 12yu. Simple, usually eccentric- conical, grains are also found. Convolvidits soldanclla Linn. (Convolvulacece.) Dry stolons. — Compound grains rounded or oval, consisting of 2 to 6 and 8 equal or imequal components. Size about 24/i. Separated-grains, size 3 to 14 and 18/i; instead of the hilum, the larger ones have a small cavity from which fissures radiate towards the corners; oval forms with eccentric hilum, almost one-half, arc rarely present. Simple grains of incomplete formation are also found. Convolvulus lincatus Linn. {Convolvidacece.) Dry root-stock. — Starch as in the preceding. Grains usually smaller, consisting of 2 to 10 components. Size about 20/j. Size of the separated- grains 2 to 13ai. Simple grains of imperfect formation are also observed. Convolvulus imperati Vahl.; Batatas liltoralis Chois. {Co^wolvulaceo'.) Dry stolons. — Compomid grains rounded or oval; consisting of 2 to 4 and G usually equal components. Size about 2Hix. TYPE 14. GRAINS COMPOUND, OF FEW COMPONENTS OF EQUAL SIZE. 259 Seixarated-grains with a small cavity from which fissures radiate toward the margin; in the oval forms the eccentricity is one-half. Size 5 to 16/i. Simple grains of incomplete forma- tion are also foimd. Batatas edulis Chois. {Convolvulacem.) Tubers. — According to Payen (Ann. Sc. Nat., 1838, pi. 6, figs. 15, 17), the compoimd grains consist of 2 to 6 and possil)ly 10 apparently almost equal components. Separated-grains, size about 45^, broadQy-conical, semi-spherical, or poly- hedral, with 1 curved surface, and 1 to 7 pressure facets; lamelliB distinct; hilum about one- fifth to one-sixth eccentric, at the narrow arched end. According to Criiger (Bot. Zeit., 1854, taf. II, fig. 4) some of the separated-grains are thickened and club-shaped at the hilum end and thinned at the distal end, whore the pressure facets are located. According to Leon Soubeiran (Journ. Pharm., 1854, xxv, 92), the grains vary from 10 to 20^ and 40 to 50^; the smallest ones are spherical or oval; those of the middle size are almost regularly poly- hedral, and the largest ones oval and elliptical. (Judging from the description, Nageli writes, it is quite possible that some simple forms occur among the compound grains.) Nolana prostrala Linn. (Nolanacece.) Dry roo^.— -Compound grains of 2 to 8 and 10 almost equal components. Size about IG/x. The separated-graias 3 to ll/(, usually with a central cavity. No simple grains are observed. Polemonium reptans Linn. (Polemoniacem.) Dry root-stock. — Compound grains of 2 to 8 almost equal, rarely unequal, components. Size about 10m. Size of the separated-grains 2 to 6^. Among these are simple grains of imperfect formation. PhysaKs alkekengi Liim. (Solanacew.) Dry stolons. — Compound grains rounded or oval, consisting of 2 to 8 and 12 almost equal, rarely unequal, components. Size about 18/i. Separated- grains, size 3 to 9m; the larger ones have a small cavity from which several delicate, short radial fissures rarely emerge, cavity central, rarely (in oval forms) about one-half eccentric. Starch plentiful. Simple grains are not present. Similar grains were found in the roots; the compound grains, however, had nearlj' all fallen apart into separated-grains. Spherical simple forms may perhaps occur in the roots. Solanuni nigrum Limi. (Solanacew.) Dry root. — Compoimd grains; rounded or oval, consisting of 2 to 8 and 12 almost equal components. Size about 20^. Separated-grains, size 4 to ll/x; the larger ones have a small cavity. Starch here and there is quite plentiful. Atropa belladonna Linn. (Solanacece.) Dry roots. — Compound grains; nearly all have fallen apart; few remaining ones consisting of 2 to 4 components. Separated-grains somewhat elongated; in the direction of the longitudinal axis about twice as long as broad; with 1 to 3 pressure facets at the broad end; lamellae invisible or indistinct; instead of the hilum, usuallj^ a small cavity is observed which is rarely central, usually nearer the narrow end, two-fifths to one- fourth eccentric. Simple grains of imperfect formation are also foimd. Scrophularia nodosa Limi. {Scropkulariaccw.) Dry root-stock. — Compound grains rounded to ellip- tical, frequently irregular, consisting of 2 to 4 and 8 almost equal or unequal components. Size about 12^. Separated-grains solid, size 2 to 6^- Also compound grains of imperfect formation are observed. Gratiola officinalis Lirm. (Scrophulariacece.) Dry root-stock. — Compound grains rounded to ellip- tical, frequently irregular, consisting of 2 to 4 and 8 almost equal or unequal components. Size about 9^- Separated-grains are solid, size 1.5 to 5^. Simple grains of incomplete for- mation. Starch rather scarce. Veronica austriaca Linn. {Scrophulariacece.) Dry roots. — Compound grains (many of them fallen apart) rounded or oval, consisting of 2 to 4 and 8 equal or unequal components. Size about 13m. Separated-grains, size 2 to 9m; the larger ones have a small central cavity; also simple grains of incomplete formation are present. Wulfenia carinthiaca Jacq. (Scrophulariacece.) Dry root-stock. — Compound grains rounded to oval, frequently irregular, consisting of 2 to 4 and 6 equal or unequal corajjonents. Size about 9m. Separated-grains are large and solid, size 1.5 to 6m. Simple grains of incomplete formation are also found. Starch rather scarce. Pedicularis barrelierii Rchb. (Scrophulariacece.) Dry root. — Compound grains rounded to oval, consisting of 2 to 4 and 7 almost equal components. Size about 14m. Separated-grains, size 2 to 6m; the larger ones have a small central cavity. Simple grains of incomplete for- mation are also found. 260 DIFFERENTIATION AND SPECIFICITY OF STARCHES. Pedicularis rosea Wulf. (Scrophulariacecc.) Dry root. — Compound grains (most of which have fallen to pieces) consist of 2 to 5 almost equal components. Separated-grains, size 2 to 5m; the larger ones have a central cavity. Simple grains of incomplete formation are also found. Pedicularis acaidis Scop. (Scrophulariacew.) Dry root. — Compoimd grains consisting of 2 to 6 usually almost equal components. Size about 14^. Separated-grains, size 2 to 6/^; the larger ones have a small cavity. Occasionally 10/i in size, then swollen up owing to a large cavity. Simple grains of imperfect formation are also observed. Orobanche sp. (Scrophulariacece or Orobanchacew.) Dry root-stock. — Compound grains consisting of 2 to 4 usually equal components. Size about 50ai. Separated-grains, lamellae none or not distinct; instead of the hilum a central or one-half eccentric cavity is observed from which well-marked fissures radiate to the margins. Size about 30ju. Also simple, usually eccentric- spherical grains are observed. Primula calycina Duby.; Primula glaucescens Moretti. {Primulacece.) Dry roots. — Compound grains of 2 to 4 usually equal components. Separated-grains, size 5 to lO^t; instead of a hilum a small central or one-half eccentric cavity is observed. Simple grains of incomplete forma- tion are also found. Soldanclla alpina Linn. {Primulacem.) Dry root-stock. — Compound grains of 2 to 4 almost equal, rarely somewhat unequal, components. Separated-grains, size about 27/^; the larger ones have a small cavity. Simple grains of incomplete formation are also found. The grains in the root are somewhat larger. Size of the separated-grains is about lO/z; they have a small cavity with isolated short fissures. Simple grains are also present. Glaux maritima Liim. {Primidaceoe.) Dry creeping stems. — Compound grains of 2 to 4 and 8 usually equal, sometimes unequal, components. Size about 15/i. Separated-grains, size 4 to lO/x; the larger ones have a small cavity from which fissures occasionally radiate. Simple grains of incomplete formation are also observed. Lysimachia vulgaris Linn. (Primulacem.) Dry root-stock. — Compound grains of 2 to 8 and 10 equal or unequal components. Size about 2lfi. Separated-grains, size 4 to 17fi; the larger ones have a small cavity from which fissiu-es radiate toward the corners. Simple grains of incomplete formation are also observed. Starch plentiful; cell-walls are not perceptibly thickened. Pyrola rotundifolia Linn. {Ericacecc.) Dry root-stocks.- — Compound grains of 2 to 4 and 6 usually equal components. Size about 15/x. Separated-grains, size 3 to Sfi; the larger ones have a small cavity. Simple grains of incomplete formation are also observed. Hydrocotylc vulgaris Linn. {Umbelliferw.) Dry rooted caulicles. — Compound grains of 2 to 4 and 10 usually equal, rarely unequal, components. Size 18 to 24/i. Separated-grains with 1 to 5 pressure facets; instead of a hilum, a small cavity is observed from which fissures radiate toward the corners. Size 4 to 17^; also isolated simple, spherical or ovate grains are found. Apium graveolens Linn., var. cult. {Umbelliferce.) Fresh root. — Separated-grains rounded or angular with rounded comers, with 1 curved surface and 1 to 5 pressure facets; neither hilum nor lamellse. Size about 12/u. The compound grains have fallen apart; only a few may be dis- tinguished within the cells, and these are oval or spherical and consist of 2 to 8 components. Petroselinuni sativum Hoffm.; Radix petroselini. {Umbelliferce.) Dry root. — Compound grains spherical to elongated-oval, consisting of 2 to 10 and 15, rarely 30, components. Size about 25^. Sep- arated-grains, size 2 to 10/:i; the majority have one curved surface, but several are bounded merely by pressm-e facets; the larger ones have a small central cavity, and several fissures turned toward the corners. Pimpinella saxifraga Linn.; Radix pimpinella alba. (Umbelliferce.) Dry roots. — Grains rounded or angular with rounded angles, frequently with one curved surface and 1 to 4 distinct pressure facets; solid and homogeneous. Size about 9n. The majority of these grains are undoubtedly separated-grains. However, simple grains also occur, which can be recognized by their spherical shape. Only a few compound grains are present. Levisticum officinale Koch. (Umbelliferce.) Dry root. — Separated-grains with one curved surface and 1 to 4 pressure facets; instead of a hilmn, usually a small central cavity is present from which isolated fissures diverge. Size about 12 to 18^. Grains are crowded in the cells; only a very few compound grains were noted. Some simple grains of incomplete formation are also found. TYPE 14. GRAINS COMPOUND, OF FEW COMPONENTS OF EQUAL SIZE. 261 Oslericum pahistre Bess. {Umhdliferce.) Dry root-stock. — Compound grains of 2 to 4 and 8 usually almost equal components. Size about 14/^. Separated-grains, size 3 to 8m; the larger ones have a small cavity. Some spherical and oval grains, size 11/i, are also observed. Peucedanum cervaria Lap. {Umbelliferce.) Dry root-stock. — Compound grains consisting of 2 to 4, rarely 6, usually almost equal components. Size about 20yu. Separated-grains, size 3 to 12/x; the larger ones have a small cavity. Cornus suecica Linn. (Cornacew.) Dry root-stock. — Compound grains consisting of 2 to 8 and 12 usually almost equal components. Size about 13//. Separated-grains, size 4 to 8/u; the larger ones have a small cavity. Also simple grains of incomplete formation are also observed. Ranunculus thora Linn. {Ranunculacece.) Dry roots. — Compound grains consisting of 2 to 6 and 8 equal or unequal comi)onents. Size about 21/i. Separated-grains, size 4 to 15ju; the larger ones have a small central cavity and several short fissures; one-half eccentric in oval forms. No simple grains are present. Ranunculus flammula Linn. {Ranunculacece.) Dry roots. — Compound grains consisting of 2 to 8 and 10 components; the majority of them have fallen apart. Separated-grains with one curved surface and 1 to 6 pressure facets. Size about lOju; the larger ones frequently have a small cavity. Starch plentiful. Ranunculus rutcejolius Limi. {Ranunculacece.) Dry roots. — Compound grains of 2 to 8 components, the majority fallen apart. Separated-grains, size 2 to T/t; the larger ones have a small cavity. Some apparently simple rounded grains are also observed. Not verj' rich in starch. Chelidonium majus Linn. {Papaveracece.) Fresh root. — Separated-grains with 1 curved surface and 1 to 7 pressure facets. Size 2 to lOyu. A few compound grains, consisting of 2 to 10 almost equal components are present. Corydalis solida Smith. {Fumariacece.) Dry tuberous root-stock. — Compound grains of 2 to 4 and 6, rarely 8, components which are mostly equal. Size about 32//. Separated-grains with a small central cavity, and several radial fissures running toward the corners. Size 4 to 18;u. No simple grains are found. Corydalis fabacea Pers. {Fumariacece.) Dry root-stock. — Grains as in the preceding. Size of sep- arated-grains 14 to 17/j. Cordyalis pumila Host. {Fumariacece.) Dry root-stock. — Grains as in the two preceding. Size of the separated-grains IG to 20//. Drosera rotundifolia Linn. {Droseracece.) Dry roots. — Compound grains rounded or oval, mostly of somewhat irregular shape, consisting of 2 to 5 and 8 components of different sizes, which generallj- are somewhat irregularly arranged. Size about 11/t. Separated-grains, size 2 to 8m; a cavdty is frequently observed. Starch quite plentiful. Viola cucullata Ait. {Violacew.) Dry root-stock. — Compomid grains of 2 to 4 and 6 equal or unequal components; size 18/;. Separated-grains, size 4 to 13/x; instead of a hilum a small almost central cavity is observed from which single fissures radiate; in oval forms the cavity is about one-half eccentric. Viola palustris Linn. {Violacece.) Dry root-stock.- — Compound grains of 2 to 6 and 9 equal or unequal components. Size about 12/i. Size of the separated-grains 2 to 6/i. Simple grains of in- complete formation are also present. Viola pinnata Linn. {Violacece.) Dry roots and root-stock. — Compound grains rounded or oval, consisting of 2 to 8 and 12 equal or unequal components. Size about 14/i. Separated-grains, size 2 to Sn; the larger ones have a cavity. Portulaca megalantha Steud. {Portulacacece.) Dry roots. — Compound grains consisting of 2 to 4 equal components. Separated-grains 5 to ll/i; frequently with a central cavity. Also simple grains of incomplete formation are found. Stellaria bulbosa Wulfen. {Caryophyllacecc.) Dry tubers. — Compound grains of 2 to 4 almost equal components, the majority have fallen apart. Separated-grains, size 3 to 8/i; occasionally a small cavity is observed, instead of a hilum, which may be about one-half eccentric. Also simple, eccentric-spherical grains are present. Melochia pyramidaia Linn. {Sterculiaceoe.) Dry roots. — Compound grains of 2 to 4 and 8 almost equal components, the majority fallen apart. Separated-grains, size 3 to S/x. Also simple, spherical, rounded-oval grains are found. 262 DIFFERENTIATION AND SPECIFICITY OF STARCHES. Triumfelia schimperi Hochst. {Tiliacecc.) Dry root. — Compound grtiins rounded or oval; consist- ing of 2 to 4 and 8 equal or unequal components, the majority fallen apart. Size about 15yu. Separated-grains, size 3 to 9m; the larger ones have a small cavity or a slit. Also some simple rounded grains are present. Canella alba Murray; Cortex canellce albce interior. (Canellaceoe.) Dry cortex. — Compound grainsof 2 to 4 and 5 mostly equal components. Size about 9/i. Size of separated-grains Gfi; also almost rounded simple grains, size 7ju, are observed. Hypericum elodes Linn. (Hypericacece.) Dry creeping root-stock. — Compound grains almost round or oval, consisting of 2 to 4 and 6 equal or imequal components. Size about 19;u. Separateil- grains, size 3 to IZ/i; most of them have a small cavity or a delicate cleft from which solitary short delicate fissiu'es may radiate. Simple grains of incomplete formation are also found. Byrsonima crassifolia DC. (Malpighiaceoe.) Cortex.- — Compound grains of 2 to 4 almost equal components. Size about 25/i. Separated-gi-ains, size about 13^; homogeneous, almost solid. Simple eccentric grains of incomplete formation are also observed. Krameria triandra R. P.; Ratanhiaivurzel krameriece. (Polygalacece.) Dry roots. — Compound grains of 2 to 4, rarely 6, almost equal components which are regularly arranged; rarely small com- ponents adhere to a large one either upon one side or upon the two opposite sides, elongated curved doublets are mostly found. Size about 35/i. Separated-grains rounded to oval; occasionally slightly compressed; usually homogeneous; rarely with indistinct lamellae; usually solid; rarely, instead of the hilum, a small cavity is located at the narrow arched end, about one-foiu-th eccentric, from which short radial fissures occasionally proceed. Simple eccentrically elongated grains are also found. Manihot utilissi7na Pohl.; Jatropha manihot Linn. (EupJiorbiacece.) Tapioca, cassava starch, Brazilian arrowroot. From the root. — Separated-grains semi-spherical to oval, occasionally 3 to 4, more rarely 5, angles with one curved surface and 1 to 5 and 7 pressure facets; the larger ones have one or several delicate lamelte; instead of the hilum, usually a small cavity about one-half eccentric is observed, frequently with single, short, radial fissures. Size 2.5 to 25;u. Compound grains have all disappeared; they evidently consisted of 2 to 8 partly equal or partly unequal components (several small components adhere to a single large one). Among the starch-grains above described are mingled simple grains which are oval, strongly com- pressed, and have delicate lamelke and eccentric hilum. Size 50 and 60/i. These resemble the grains of Curcuma zedoaria. They probably belong to Curcuma starch (Tikmehl). Other species of tapioca which are somewhat changed by means of heat: Separated-grains, size 17 to 50^i; lamellae rarely distinct; with a central or slightly eccentric, often large, cavity, from which two fissures pass out and embrace a conical substance or a dense, almost round, granule within the cavity. Only a few compound grains of 2 to 4 almost equal components are still present. Also a few simple grains, spherical or oval-spherical, with concentric lamellae, or with a central cavity, are observed. Pachysandra procumbens Michx. (Euphorbiacece.) Dry root-stock. — Compound grains of 2 to 4 almost equal components. Separated-grains, size 3 to llyu; the larger ones have a small cavity. Some simple grains, spherical and roimded-t)val, size about 12;u. This specimen is poor in starch. Epilobiurn hirsutum Linn. (Onagracece.) Dry stolo7is. — Compomid grains of 2 to 4 and 6 components which are frequently equal. Size about 21/^. Separated-grains, size 4 to 13fi; instead of the hilum, a small central or about one-half eccentric cavity is observed, from which delicate fissures radiate to the corners. Myriophyllum verticillatum Linn. (Haloragacece.) Base of dry rooted stem,s. — Compound grains of 2 to 4 almost equal components; size about 15^. Separated-grains, size 4 to 9yu; mostly with a cavity. Also simple giains of incomplete formation are observed. Pyrus -malus Linn. (Rosacece.) Fresh ptdp of the fruit. — Com])ouud grains of 2 to 4 usually equal components. Siz(! aljout lOyu. Comjioncnts with indistinct, small, almost central hilum. Size 4 to 11/i. Also a few simple spherical grains are present. Comarum paluslre Linn. {Rosacea;.) Dry root-stock. — Comjjoimd grains of 2 to 4 and 6 components which are usually almost equal. Size about lOyu. So])arated-grains, size 2 to G/x; the larger ones generally have a small cavity. Also simple grains, spherical or rounded-oval, are ob- served; size about 8>i. TYPE 14. GRAINS COMPOUND, OF FEW COMPONENTS OF EQUAL SIZE. 2G3 Potenlilla aurea Linn. (Rosaceoe.) Dnj root-stock. — Separated-grains, size 2 to 5ix; the smaller ones rounded, the larger ones more or less polyhedral, and frequently with a small cavity. Most of the compound grains have fallen apart; those still present are composed of 2 to 8 almost equal components. Some more complex ones seem to have been present. Simple rounded grains are also found. Alchetnilla alpina Linn. (Rosacew.) Dry roo^-stocA;.— Compound grains, almost round to elongated- oval, frequently somewhat irregular, consisting of 2 to 10 and 20 components which are almost equal. Size about 14//. Separated-grains, size 2 to 4 and 6/i ; occasionally with a small cavity. Some simple spherical and oval grains are also present. Geum xirhanum Linn.; Radix caryophyllacw. (Rosaceoe.) Dry root-stock. — Compound grains almost rounded or oval, rarely elongated-lanceolate, consisting of 2 to 8, rarely more than 8, com- ponents which are equal and regularly arranged (a few are in one row). Sizes about 10/:. Separated-grains almost round or angular with rounded comers; size 1.5 to 5/i. Some simple spherical grains are also observed, many of which appear at first sight to be simple prove by closer observation to be divided by delicate lines. Geum montanum Liim. (Rosaceoe.) Dry root. — Compound grains of 2 to 8 different-sized com- ponents. Size of separated-grains, 2 to 6//. Simple almost roimd or oval grains are also present. Crotolaria incana Linn. (Leguminosce.) Dry root. — Compound grains of 2 to 4 and 6 components which are generally equal. Size about 14/t. Separated-grains, size 2 to 7//; many with a small cavity. Also simple, almost round grains are found. Starch rather scarce. Astragalus incanus Linn. (Leguminosce.) Dry root-stock (rhizome). — Compound grains of 2 to 4 almost equal components. Size about 17/1. Size of the separated-grains 4 to lO/t. Simple grains of incomplete formation are also observed. Lathyrus pratensis Limi. (Leguminosce.) Dry roots. — Compound grains consisting of 2 to 8 com- ponents of different sizes. Size about IG/i. Separated-grains, size 2 to 9/:; occasionally with a small cavity. Also solitary simple, almost round grains are present. Starch plenti- ful. In the root-stock of the same plant the starch-grains are rather scarce. Size of the compound grains lO/i; of the separated-grains 1.7 to 6/t. Also simple round gi-ains are present. Lathyrus palustris Linn. (Leguminosce.) Dry root-stock. — Compound grains of 2 to 10 and 15 com- ponents of different sizes. Most of the compound grains have fallen apart. Size about 24/i. Separated-grains, size 3 to 14//; instead of the hilum a small cavity is usually observed, which in the oval forms is one-half and one-third eccentric. Starch plentiful, filling up the cells of the cortex. Zornia angustifolia Smith. (Leguminosce.) Dry root-stock.- — Compomad gi-ains of 2 to 10 and more components of different sizes; size about 14/i. Size of the sepai-ated-grains, 2 to 8//. Some simple grains, almost round or oval, are also foimd. Alysicarpus ferrugineus Steud. (Leguniinosoe.) Dry root-stock. — Compound grains of 2 to 10 and 15/i, almost equal components. Size about 16/i. Size of the separated-grains 2 to 7/t. A few simple rounded and oval grains are also present. Starch rather plentiful. Desmanthus virgatus Willd. (Leguminosa;.) Dry root. — Compound gi-ains of 2 to 4 and 6 components. The majority have fallen apart. Size about 18//. Size of the separated-grains 3 to 10//. Also simple spherical and oval grains are also foxmd. Starch not plentiful. Encephalartos spiralis Lehm. (Cycadacece.) Dry embryo.- — Compound grains of 2 to 4, rarely more (up to 8), components which are usually equal and regularly arranged. Separated-grains with 1 to 3 pressure facets, which are semi-spherical and also frequently somewhat oblique, or they may be oval and often cut off obhquely at one end; instead of the hilum a small cavity is observed from which short delicate fissiu-es radiate, one-half to one-third eccentric; usually one or two circular lamellie around the hilum, as well as several incomplete lamellae toward the distal end. Size of the separated-grains, 5 to 25 and 30/i. Simple eccentric- conical grains are also present. Ophirus alhiopicus Rujir. ; Ophirus papillosus Hochst. (Graminacem.) Dry endosperm. — Compound grains when lying loose in the cell are almost round or oval; when crowded they are angular or even polyhedral. They consist of 2 to 10 rarely 20 components. Size about 15/j. Sep- arated-grains, size 2 to 8//; usually polyheilral, sometimes with 1 curved surface and 1 to G 264 DIFFERENTIATION AND SPECIFICITY OP STARCHES. pressure facets; tho larger ones have a small cavity and frequently several radial fissiu-es. Sim- ple spherical and spherical-oval grains are also observed; size 13^. These apparently belong to the centric-spherical type. Some of the polyhedral grains may also be simple forms. Kottbaella arundinacece Hochst. (Graminaceoe.) Dry endosperm. — Separated-grains with one curved surface and 1 to 4 pressure facets; instead of the hilum a small cavity is usually observed from which single fissures sometimes radiate. Size 3 to 8n. Compound grains are less freely observed; they probably consist of not more than 8 to 10 components. Simple centric- spherical grains are also present. Rolthcella campestris Nutt. {Graminacece.) Dry endosperm. — Compound grains spherical or oval, consisting of 2 to 20 equal parts (few seen free). Separated-grains, size 2 to 10^; some with one curved surface and 1 to G pressure facets; others completely polyhedral; the larger ones usually with a small cavity and often with single radiating fissures. Also simple grains, spherical or spherically-oval; size about 12ju; the larger ones with a small central cavity and several radial fissiires, belonging to the centric-spherical type (type 1). Many of the poly- hedral gi'ains may also be simple. Lucwa colorata Hochst. {Graminacece.) Dry endosperm. — Sejjarated-grains with 1 curved siu-face and 1 to 5 pressure facets, hollow. Size 3 to WjJ-. Compound grains of not many components are very scarce; also simple hollow grains almost round, rarely oval, and somewhat com- pressed. Size about 13;u. The simple and the separated-grains (the former about one-half to one-third as munerous as the latter) look exactly alike; the central cavity in both is large, uneven, angular, jagged, or extends into short radial fissures. Wachendorfia Mrsuta Thumb. {Hcemodoracece.) Dry enrfosperm.— Compound grains (most of which have fallen apart) consist of 2 to 12 and more components, which are almost equal, and regularly or irregularly arranged. Separated-grains, size 3 to 20ju; rounded-angular or polyhedral, with large or small cavity which is frequently very conspicuous. The starch gives the impression of having been affected by heat. Hoeniodorum (a species from Swan River). {Hcemodoracece.) Dry endosperm. — Compound grains almost round to oblong and pear-shaped, frequently somewhat irregular, consisting of 2 to 10 almost equal components (frequently arranged 4 in a square, and sometimes in a row), rarely falling apart; lines of division sometimes indistinct or entirely wanting, so that the cavities of the components are seen, as in type 12. Size about 28^. Single components, size about 13ju; almost oval and have rather large, often slit-like cavity. Simple grains, size about 16ju; oval, and slightly compressed with a slit-like cavity. It is rather uncertain whether these belong to the eccentric or to the oval type. A constant analogy which the compound grains of few components holds with those of the seeds of Leguminosce is in support for the latter type. Barhacenia rogieri Hort. {Amaryllidacece.) Dry endosperm. — Separated-grains more or less poly- hedral, frequently with sharp angles and edges; isodiametric or twice as long as broad; with a central almost round or elongated cavity from which radial fissures diverge. Size 3 to 18/u, width about 14/i. Few compound grains are less seldom present; they consist of 2 to 8 components. Hohenbergia strobilacea Schult. fil. {Bromeliacece.) Dry endosperm. — Compound grains spherical or oval, sometimes somewhat irregular as a result of pressure, consisting of 2 to 5 almost equal components which are arranged either irregularly or regularly (3 in a triangle or 4 in a square or tetrahedron). Size about 15//. Separated-grains almost isodiametric; they frequently have a small central cavity, without fissures. Also simple centric-spherical grains are also found. BiUbergia zebrina Lindl. {Bromeliacece.) Dry endosperm. — Compound grains spherical-oval, rarely irregular; consisting of 2 to 6ju, rarely more, almost equal components which are regularly arranged. Size about 21yu. Separated-grains with rather sharp angles and margins; with a small central cavity from which single fissures occasionally proceed. Size 4 to lO/i. Simple centric-spherical grains are also present. Pitcairnia alhuccejolia Schrad.; Pitcairnia punicea Lindl. {Bromeliacece.) Dry endosperm. — Sejiar- ated-grains with 1 to 4 ]5ressure facets with rather sharp angles and margins; usually with a large cavity from which single ratlial fissures occasionally proceed. Size 4 to 12 and \&n. Compound grains few, consisting of 2 to 6 components. Simple centric-spherical grains are also present. TYPE 14. GRAINS COMPOUND, OF FEW COMPONENTS OF EQUAL SIZE. 265 Dyckin remoliflora Otto. (Bromeliacece.) Dry endosperm. — Compound grains spherical to oval, usually of almost regular shape, about twice as long as broad, consisting of 2 to 10 and 12 nearly equal components which are often regularly arranged. Length about 24;u, width about Ifl/u. Separatinl-grains with acute angles and margins, with a cavity at the hihun which is somewhat nearer the pressure facets and from which 1 to 3 fissures radiate toward the angles. Size about 14^. Isolated spherical or oval simple grains are also observed. Pistia (c.vensis Klotsch. (Aroidece.) Dry endosperm. — Compound grains spherical, oval, rarely shortened-conical ; consisting of 2 to 12 and 20 almost equal components. Size aljout 11 to 14ju. Se])arated-grains, size 2 to Gyu; the larger ones with a small central cavity. Pistia stratiotes Linn. (Aroidew.) Dry endosperm. — Separated-grains occasionally with a small central cavity; size 2 to 8/u. Compound grains few, consisting of 2 to 8 components. The starch-grains of both species of Pistia form a wall-like layer in the endosperm cells. The lumen is filled with a substance which takes a golden yellow stain when treated with iodine. Aponogeton distachyum Thumb. (Alismaceoe.) Dry endosperm. — Separated-grains with one curved surface and 1 to 3 pressure facets; usually with a small cavity. Size 2.5 to llyu. Compound grains (only a few intact ones present); size about ISyu; consisting of 2 to 5, rarely 10, compo- nents. According to Payen (Ann. Sc. Nat., 1838, ii, 25, pi. 6, fig. 23), the size of the sep- arated-grains is 22.5^; with eccentric hilum and occasionally with several fissures radiating toward the distal end. Corylus avellana Linn. {Corylacece.) Fresh immature cotyledons. — Compound grains consisting of 2 to 8 almost equal components, which are irregularly arranged; small and falling apart easily. After the compound grains have fallen to pieces the separated-grains which are angular with rounded angles increase considerably in size about 5/i. Simple almost round or oval grains are also found. Size about 6ju. Nectandra rodioei Schomburgh; Fructus bebeeru. (Thymelacece.) Dry cotyledons. — Compound grains of 2 to 4, rarely more, almost equal components which are arranged regularly. Separated- grains isodiametric to twice as long as broad; instead of the hilum a small cavitj' is observed from which 2 short fissures occasionally radiate toward the jjroximal end (the ones with pressure facets), eccentricity about one-third. Size 8 to 22yLi. A few rounded or shortened- conical simple grains are also present; the hilum is toward the thicker end. Therefore they belong to the eccentric-conical type (type 7). Laurus nobilis Linn. (Lauracece.) Dry cotyledons. — Compound grains almost round to elongated- oval, consisting of 2 to 6 almost equal components. Separated-grains with a slightly eccen- tric cavity from which several short fissures diverge; size about 15,u. Simple eccentric inverted-conical grains are also present. Scabiosa atropurpurea Linn. (Dipsacacece.) Fresh immature endosperm. — Grains rounded or rounded- angular; size about 4/i. The greater number of these are probably separated-grains; the smaller number are simple. Some larger compound grains consisting of 2 to 8 almost equal components are also observed. The starch forms a lining to the wall of the cells, which have transparent, colorless contents. Atropa belladonna Linn. (Solanacew.) Fresh immature seed-coats. — Grains spherical-oval and rounded-angular; size about 6 to 8^. Many of them are separated-grains. Compound grains consisting of 2 to 8 components are also present. jEgiceras majus Gart. {Myosinacece.) Dry cotyledons. — Compound grains consisting of 2 to 4, rarely 8, usually equal components which are arranged regularly. Separated-grains, size 4 to 18/^; with a small cavity from which fissures radiate toward the circumference of the pressure facets. Starch very plentiful. Lucuma rivicoa Gart. fil. (Sapotacece.) Dry cotyledons. — Compound grains consisting of 2 to 4, rarely 8, equal or unequal components. Size about 28;u. Separated-grains with 1 curved surface and 1 to 4 pressure facets. Siz(! 4 to 15ix; the larger ones have a small cavity and single radial fissures; eccentricity about one-half, rarely one-third. Isolated simjjle, rounded or shortened-conical grains are also present. Length about 20/x; wth a cavity toward the thicker end, from which 2 fissures radiate toward the narrowed end. Starch belongs to the eccentric-conical type (type 7). 266 DIFFERENTIATION AND SPECIFICITY OF STARCHES. Lucuma sp. (perhaps Luauna cainilo A. DC.) (Sapolacece.) Dry cotyledons. — Compound grains consisting of 2 to 4, occasionally more, components which arc usually equal. Size about 20ju. Separated-grains, with 1 to 3 and 4 pressure facets; size 3 to 12ix; the larger ones have a small cavity with single fissures radiating toward the angles; eccentricity about one-half. Lucuma sp. {Sapolacece.) Dry cotyledons. — Compound grains rounded to oblong, occasionally somewhat irregular, 2.5 times as long as broad, consisting of 2 to 12 and more almost equal components. Length about 23/^, thickness about 18/j. Separated-grains with one carved surface and several pressure facets; with one large or small cavity; size 2 to lO/i. Some sim- ple grains are also found. MyrisUca moschata Thumb. (Myristicacece.) Dry endosperm. — Compound grains rounded to elon- gated-oval, occasionally somewhat compressed, consisting of 2 to 10, rarely more, equal components which are usually arranged regularly. Size about 27fi. Separated-grains, size 4 to 15m; with a small central cavity from which single short fissures occasionally proceed. MyrisUca salicifolia Willd. {Myristicacece.) Dry seed. — Compound grains almost round to oljlong, consisting of 2 to 10 and 16 usually equal components which are arranged quite regularly. Size about 34^. Separated-grains frequently with a small central cavity from which short fissm'es radiate; size 5 to 18/i. Pwonia. {Ranunculacece.) Fresh unripe endosperm of various species. — Compound grains spherical or oval, consisting of 2 to 4 and 6, rarely 8, components which are usually equal. Size about 16^. Separated-grains size 3 to 12/i. Isolated spherical or oval simple grains are also present. Seeds have sprouted and are either still gi-een or have become red, and are entirely filled with starch. At complete maturity they contain no trace of starch, but a great deal of oil. Corydalis lutea DC. {Fumariacece.) Fresh placenta of immature fruit. — Separated-grains rounded- angular; size about 5/.(. Compound grains consisting of 2 to 8 components are also found. Starch quite plentiful. Brassica napus Linn. {Brassicacece or Cruciferce.) Fresh endosperm, cotyledons, and seed-coats of immature unripe seeds. — Separated-grains with 1 curved surface and 1 to 4 pressure facets; size 8 to 12;u. The few compound grains which have not fallen apart consist of 2 to 4 almost equal components. Simple grains of incomplete formation are also observed. Carolinea princeps Linn. {Malvacece.) Dry cotyledons. — Compound grains consisting of 2 to 6 com- ponents which sometimes are equal and regularly or often unequal and irregularly arranged; 1 large and 1 to 4 small components. Separated-grains with 1 curved surface and 1 to 4 pressure facets; with a small central cavity and single radial fissures. Size about 15 to 20;u; also simple eccentric-conical grains are also observed. Heritiera littoralis Ait. {Sterculiacece.) Dry cotyledons. — Compoimd grains spherical elongated- lanceolate, frequently somewhat irregular, 3 times as long as broad, consisting of 2 to 12 or more components which are almost equal and usually arranged regularly (rarely in a simple row). Size about 23/i. Separated-grains with a central cavity from which short fissiu-es radiate. Size 4 to lOju. Some simple spherical grains are also observed. Thcohroina cacao Linn. {Sterculiacece.) Dry cotyledons. — Separated-grains rounded-angular, fre- quently with 1 curved surface and 1 to 3 pressm-e facets. Size about 8fi. Compound grains consisting of 2 to 4, rarely more, almost equal components. The cells are filleil with oil in which the starch-grains are scattered. Thea bohea Linn. {Camelliacece.) Dry cotyledons. — Compound grains consisting of 2 to 4, rarely 8 components, which are usually equal and regularly arranged (of which 3 are usually arranged lineally, 3 and 4 in one plane). Size about 20 to 25/i. Separated-grains, size 4 to 9 and 12ju; instead of a hilum a small cavity is observed from which fissures usually radiate toward the sui'face of the pressure facets. Simple grains of incomplete formation are also found. Triphasia aurantiola Lorn'.; Limonia trifoliata Linn. {Rutacece.) Dry cotyledons. — Grains consist- ing of 2 to 4 and 8 comjionents of different sizes. Separated-grains usually with 1 to 3 pros- sure facets and a small cavity; size 3 to 8/i. A few simple rounded and rounded-oval grains, size about 9^, are also present. The seeds are quite conspicuous and seem to be mature. TIk; cells are filled with starch, and have little or no oil. Trichilia mia-antha Benth. {Mdiacece.) Dry cotyledons. — Compound grains rounded to oblong, frequently irreg\ilar, consisting of 2 to 8 and 12 components which are usually equal. Size about 15 to 20|i. Sepai'ated-grains, size 3 to 11/^; the larger ones have a small central cavity. TYPK 14. GRAINS COMPOXIND, OF FEW COMPONENTS OF EQUAL RIZR. 207 Trichilia sp. {Mdiacccv.) Dry. — Conii)oiiud f;;raiiis coiisistiiifj; of 2 io 8, rjiroly more, equal com- ponents which ju-e generally arranged regularly. Size about 18;u. Separated-gi-ains, size 4 to 12ju; a small cavity and single indistinct, short radial fissures are observed. Guarea trichilioides Linn. {Meliacew.) Dry cotyledons. — Compound grains rounded or oval, consist- ing of 2 to 12 and 20 mostly equal components. Size about 15/^. Separated-grains usually with 1 ciu'ved surface and 1 to 5 pressure facets; the larger ones with a small cavity. Size 2 to 8fi. Enjthro.rylon cohunhlnum Mart. (Linaccw.) Dry cotyledons. — Compound grains rounded or oval, as thick as Ijroad or somewhat compressed, consisting of 2 to 12 and 20 components which are usually almost equal. Size about 20 to 25/i. Separated-grains, size 3 to llju; with a small central cavity and single radiating fissures. The starch-grains in the scarce perisperm of the same plant are like those in the cotyledons, although the average compound ones consist of only 2 to 8 antl 10 comi)ou(nits, and are somewhat larger. The diameter of the separated- grains is about 14 and 17;u, and the cavity at the hilum is frequently one-half and one-third eccentric. Erythro.vylon (species from Brazil) . (Linacece. ) Dry seed. — In the large cotyledons the grains are com- pound, consisting of 2 to 8 components; the size of the separated-grains is about 14yu. Sep- arated-grains in the scarce perisperna are about 20/^. Erythro.rylon nitidum Mart. {Linacece.) Dry seed. — -The grains are similar to those in Erythroxylon columbmum. The comjjound grains in the large cotyledons consist of 2 to 10 and 16 com- ponents. Size 15/i. Size of the separated-grains about 6 to 7^. Perisperm almost entirely wanting. Erythroxylon mucronatuni Benth. {Linacece.) Dry seed. — Grains as in E. columbinum. Size of the separated-grains in the large cotyledons about Six, and in the very scant perisperm about 9^. Erythroxylon rufum Cav. {Linacece.) Dry seed. — Grains as in Erythroxylon columbinum. The size of the separated-gi-ains in the small cotyledons is about 10;u, in the plentiful perisperm about IdfjL. Compound grains in the latter consist of 2 to 8 and 12 components. Erythroxylon obtusum DC. {Linacece.) Dry seed. — Grains as in Erythroxylon columbinum. Size of the separated-grains in the abundant perisperm is about 15/^. Some simple spherical gi-ains, size about 15;u, are also found. Cupania tomentosa Swartz. {Sapindacece.) Dry cotyledons. — Compound grains consisting of 2 to 8 and 12 usually equal components which are arranged regularly. Size about I2fx. Size of the separated-grains 2 to 7n. Protium pubescens; Idea pubescens Benth. {Burseraceoe.) Dry cotyledons. — Separated-grains angular with rounded angles, or oval with 1 curved surface and 1 to 5 pressure facets; instead of a hilum a small cavity is observed, and in the larger ones also single radial fissures. Size 2 to 9/u. Compound grains (the most of which have fallen apart) consist of 2 to about 10 usually equal components. Simple grains, rounded, oval, and low cone-shaped, size about 10m, are also present; instead of the hilum a small cavity is found from which single fissiu-es radiate, about one-half eccentric. These grains probably belong to the eccentric-conical type. Alangium dccapelalum Lan. {Cornacece.) Dry endosperm. — Separated-grains with I curved surface and 1 to 4 pressure facets; usually with a small central cavity; size 2.5 to 8/i. Compound grains are no longer present; probably they consisted of 2 to 8 equal components. Alangium hexapetalum Lam. {Cornacece.) Dry endosperm. — Grains polyhedral, with a small central cavity; size 3 to %ix. They lie closely packed in the cells, which they completely fill. They are probably separated-grains. Nageli could not clearly distinguish any compound ones. Memecylon capense Eckl. Zeyh. {Melastomacece.) Dry cotyledons. — Compound grains consisting of 2 to 4 and more mostly equal, rarely unequal, components. Separated-grains, size 4 to 14ju, are usually oblique and uasymmetrical; they have a small cavity and single radial fis- sures. Also simple eccentric-conical grains are present. Memecylon amplexicaule Roxb. {Melastomacece.) Dry cotyledons. — Grains as in the preceding spe- cieS; but the separated-gi'ains are more symmetrical. Simple grains are also present. Pyrus malus Linn.; Pyrus communis Linn. {Rosacea;.) Fresh unripe cotyledons. — Grains roimded and angular with rountled corners ; size about 5/i. These arc s(ii)arated-grains, since in the still younger stages tnily small compound grains occur, consisting of 2 to (J and 8 almost equal components. 2G8 DIFFERENTIATION AND SPECIFICITY OF STARCHES. Ctcer, Pisum, Ervwn, Orolms, etc. (Leguminosa;.) Cotyledons. — Among the simple centric-oval grains are found compoiuid grains which consist of 2 to 4 usually equal components which are more or less regularly arranged. Sepai-ated-grains oval, rarely rounded, with 1 curved surface and 1 to 3 small pressure facets; lamellse none, or few and delicate; with an oblong compressed hilum; after drying, a distinct median slit is mostly visible from the narrow aspect, and radial fissures from the broad aspect. Size of the compound grains of Pisum sativum Limi. about 75jx. Size of the separated-grains about BQfi. Size of the separated- grains in Cicer arietinum Liim. about IQ/x. Amphicarpcea monoica Nutt. (Leguminosa;.) Dry cotyledons. — Compound grains consisting of 2 to 6 components which are usually almost equal. Size about 14yu. Separated-grains; size 3 to 10;u; the large and oval forms have instead of a hilum a small cavity about one-half eccentric. Also simple eccentric-conical grains are found. Type 15. Grains Compound, Components Few and Unequal. From 2 to 10 unequal components united into a compound grain; divided by fissures and at maturity falling apart into separated-grains which have one curved surface and one or more flat- tened pressure facets. These compound grains for the most part arise from the cutting off of angles and edges. For this reason they frequently consist of several small components piled upon one large one, or even of two or more large ones which have small ones at their line of union. They rarely occur alone, but usually are mingled with simple grains or compound grains of equal compo- nents, the latter being formed by division of the hilum and thus are related to type 14. They may have their origin by a number of simple grains in one chloroplast uniting into a compound one, the latter frequently, partially or entirely, retaining its green color. Chara foetida A. Braum. (Algce.) Fresh nodes of the stem [basal nodes of the whorled branches). — Com- pound grains spherical and oval to rod-shaped; about 4 times as long as thick; consisting of 2 to 12, rarely 20, unequal components; green at first, colorless later. Length about 34yu, thickness about 23ju. Separated-grains polyhedral, usually with 1 curved surface and 1 to 4 pressm-e facets; 1 to 2 times as long as broad; no lamellae; usually without a hilum. Size 5 to 17/i. Starch-grains arise within the chloroplasts. Pteridophyta. Green parts. — Compound grains spherical, arising within the chloroplasts and con- sisting sometimes of 2 to 5, sometimes 2 to 10 or more, components which are more or less equal, and usually lie in one plane, and are green to colorless. Occasionally with irregular accretions of fatty oil. Some simple grains are also present. Opuntia coccinellifera Mill. (Cactacece.) Green parts. — The simple grains are somewhat disk-shaped; size about 16//. The compound ones, size about 24/i, consist of 2 to 5 components; the size of a comiDonent is 3 to 12/i. Begonia sp. (Begoniaceoe.) Leaves. — Compound grains consisting of 1 to 3, rarely 16, components; size 24yu; part-grains 3 to 12^. Nerium oleander Linn. (Apocynacece.) Young leaves. — Simple grains almost roimded; size usually about G/x. Compound grains, of 2 to 4 and 5 components; size abut 9ai. Nephrolepis exallata Schott. {Polypodiaceoe.) Base of the frond. — Compound grains of 2 to 8 and more components; size 31^- Size of the components 3 to 16/i. Chara stelligera Bauer. (Algce.) Dry stellate bodies. — These compound grains, which approach the semi- compound type, are distinguished by having 1 to 30 to 70, and occasionally 100, small compo- nents adhering to one large rounded one. These contain one or usually several hila (or as a result of drying, small hollow spaces are found instead of the hila), and thus they become semi-com- poimd, frequently showing distinct concentric lamellte (around the mathematical center), and short radial fissures in the interior. The smaller components have a small central cavity, but neither lamelliE nor fissures. They are found either at one or the two opposite sides of the large component, and when 2 to 5 are joined they lie in one row, thus forming a border. When present in larger numbers, they lie in one flat plane and form a sort of shell which covers one-sixth to two-thirds of the circumference of the large component. This shell can often be loosened by means of pressure as well as by the expansion of the larger component, which in the center has from two to several lamella;. Size of the complete compound grain about 70 to SSyu; of the large components 55 and 70ai; and of the small adhering components 5 to 13 and 20/i. TYPE 15. GRAINS COMPOUND, COMPONENTS FEW AND UNEQUAL. 269 Isoeles lacitstris Linn. (Isoetacece.) Fresh stew!S.— Compound grains, very irregular, with protruding angles and short lobes; 1 to 3 times as long as broad; consisting of 2 to about 10 unequal, irregularly arranged components. Length about SSfi, thickness about 20yLi. Separated- grains irregular, usuallj' with sharp corners; homogeneous or with a small central hilum; occasionally with short, radial fissures. Size about 17/u. Simple grains of almost round or irregular shape arc also present. The starch-grains in the youngest part of the stem are also irregular, one margin being thickened and the opposite one broadened, thinned, and angular. They are clearly either simple grains, or verj' delicate lines may be recognized which indicate division. In somewhat older parts of the stem almost all of the grains show more or less distinct lines of separation, and therefore appear to be compound, while the oldest tissues contain compound grains, as well as numerous separated-grains, and also simple rounded forms (perhaps separated-grains which after having fallen apart have become rounded by further gro\\'th). The fissures of the innermost components are first indicated in the later stages. Mariscus elaius Vahl. (CyperacecE.) Dry root-slock. — Compound grains consisting of 2 to 4 and 6 usually unequal components. Size about II/jl. Separated-grains, size 2 to 6yu; frequently with a small cavity. Simple, spherical, oval, and low-conical grains are also found. Cyperits phymatodes INIhlbrg.; Cyperiis repens Ell. (Cyperacece.) Dry root-stock. — Compoimd grains consisting of 2 to 4 mostly imequal components. Size about 14/i. Separated-grains, size 2 to S/Ji; usually with a small cavity. Simple spherical, oval, and pear-shaped grains (resem- bling those of Cyperus esculentus) are also present. Abildgaardia monostachya Vahl. (Cyperacew.) Dry root-stock. — Compound grains consist of 2 to 4 and 6 equal and unequal components. Size about 10^. Separated-grains, size 2 to G/x; occasionally ■uith a small cavity. Simple grains of incomplete formation. Carex arenaria Linn. (Cyperacece.) Dry root-stock. — Compound grains consisting of 1 large almost round or roimded-oval and 1 to 2 small components. Size of the former 8yu, of the latter, 1.3 to 2//. Also simple grains of incomplete formation are present. Carex hirta Linn. {Cyperacece.) Dry root-stock. — Grains as in the preceding. Triglochin harrcUeri Lois. (Naiadacece.) Dry root-stock. — Compound grains consisting of 2 to 4 mostly unequal, rarely equal, components. Separat«d-grains, size 4 to 28^; the larger ones have a longitudinal axis from which sometimes single, lateral fissures proceed. Simple grains of incomplete formation. Veratrum album Linn. (Ldliaceoe.) Dry roots. — Compoimd grains of 2 to 4 and 5, occasionally equal, more frequently unequal, components. Size about 2.5yu. Separated-grains almost rounded or oval, \vith 1 to 3 pressure facets. Size 4 to 18/^. The larger ones have a small cavity from which two short fissures diverge; about one-third eccentric. Simple spherical or spher- ical-oval grains are observed which have instead of the hilum a small central, or about one- half eccentric, cavity. Size about ISai. The starch-grains in the root-stocks are similar to the above, but smaller and usually simple. Bulbocodium vernum Linn. {Liliacew.) Dry tubers. — Compound grains of 2, rarely 3 and 4, compo- nents, which are occasionally equal, more frequently unequal. Size about 8fi. Size of the separated-grains 1..5 to 6^. Simple grains of incomplete formation are also present. Gagea lutea Schult. (Liliaceoe.) Dry bulb. — Compound grains, of 1 large and 1 to 3 and 4 small components. Separated-grains, size 4 to 30/i; the larger ones almost round, with a small central cavity and radial fissures. Also simple grains of rudimentary formation are found. Scilla maritima Linn. {Liliacea.) Dry bulbs. — Compound gi-ains almost round to oblong, usuallj^ irregular, consisting of 2 to 15 xnd 20 components which arc usually either unequal, or occa- sionally equal, and arranged irregularly. Size about 21^^. Separated-grains, size 2tollM; frequently with a small cavity. Simple, round oval, and pear-shaped grains are also present. Size about 15fi. Narthedum ossifragum Huds. (Liliacece.) Dry root-stock. — Compound grains consisting of 2 to 4 and 8 components which are usually unequal, or sometimes equal, and arranged irregularly. Size about 10/j. Size of the separated-grains, 2 to 6^. Simple grains of rudimentary structure are also found. 270 DIFFERENTIATION AND SPECIFICITY OF STARCHES. Galanlhus nivalis Linn. (Amaryllidacca.) Fresh bulb. — Compound grains consisting mostly of 1 large and 1 to 8 small components. Size of the former about 25/i, of the latter 2 to 5/i. Simple eccentric-cuneiform grains are also observed. Small components (part-grains) are cut off from the angles and the distal margin of these cuneiform grains; these components when very numerous usually form one or two transverse rows; very rarely do they show an irregu- lar arrangement. Canna Infjimcvsis Lindl.; Canna pedunculata Sims; Canna sp.; Canna warszewiczii. (Cannacece.) Fresh roof-stock. — Compound grains consisting of one large and several (2 to 10) small com- ponents. The latter are usually cut away from the distal end, rarely from the lateral surfaces. Length of the large component about 117 ix, width about 75ju. Diameter of the small compo- nent 3 to 5fjL. Simple eccentrically compressed grains are also found. Arum 7naculatum Linn. {Aroidecc.) Fresh root-stock. — Compound grains roimd or oval, usually obtuse-angular and irregular; consisting of 2 to 10 components which are sometimes equal but usually unequal, and often irregularly arranged. Size about 23/u. Separated-grains with one curved surface and several pressure facets, usually with sharp corners; lamellae none; rarely with a distinct hilum, size of the more polyhedral separated-grains about 9ju, and of the semi-spherical or oval ones about 15/^. A few simple spherical grains are also present. Ariwi ternalum Thumb. {Arddece.) Fresh root-stock. — Compound grains consisting of 2 to 10 com- ponents which are mostly unequal, and often irregularly arranged. Separated-grains usually without lamella;, though occasionally a few delicate ones are present; hilum is generally visible and often with 2 to 3 short, radial fissures. Size of the separated-grains 5 to 2ifi. Also a few simple spherical grains are found. Typha minima Hoppe. (Typhacece.) Dry root-stock. — Compound grains consisting of 2 to 12 and 20 com- ponents which are occasionally equal, though more frequently imequal. Size about 15/i. Size of the separated-grains 1.6 to Syu. Isolated simple round grains are also present; size about 9;u. Sagus rumphii Willd. (Palmacew) and other palms. Sago from the pith of the stem. — According to Schleiden (Grundzuge, 3 Aufl., i, 186), the grains are compound by means of several small ones adhering to one large one. The sago flour which has not been changed by heat consists princijjally of separated-grains; length 65/i, width about 50/z; about 0.66 to 2. .5 times as long as broad; the narrow ones terete, the wider ones slightly compressed; instead of the hilum, a small cavity is observed, mostly with a cross-fissure, occasionally with radial fissures; about one-fourth to one-sixth eccentric; lamellae numerous, but delicate and indistinct. The larger separated-gi-ains have 1 to 3 pressure facets at the distal end, since these small components have been cut off at this end. Only a few com- pound grains are present; they consist of 1 large and 1 to 3 small components. A few simple gi-ains also occur, resemljling the larger separated-grains but without pressure facets; length about 70//, width about 55/^; roimded, conical, oblong, or triangular, broader at the distal end; two-thirds to almost 3 times as long as broad; the broader ones slightly compressed; from the narrow longitudinal aspect they are sometimes thicker, sometimes thinner, towards the distal end. All eccentric forms are represented among the simple and the separated-grains, namely, the inverted-conical, the conical, and the flattened forms ; but generally they are without pronounced features; the gi-eatest number, however, belong to the eccentric rod-shaped tj^pe {type 9). Peperomia monostachya R. P. (Piperaceoe.) Dry root-slock. — -Grains compound, of 2, 3, rarely 4, mostlj' imcqual components. Size about 21fi. Separated-grains, size 5 to 15ju; the larger ones have a small central cavity from which several radial fissures proceed. Polygonum aviculare Linn. (Polygonacece.) Dry root. — Compound grains consisting of 2 to 4 and 6 (•f)mponents which are occasionally equal, more frequently imequal, most of which have fallen apart. Size 10 to 13//. Size of the separated-grains at)f)ut 7^; usually with a small or liirge cavity. Also simple grains of incomplete formation are jircsent. Polygonum convolvulus Linn. {Polygonacece.) Dry root. — Compound gi-ains consisting of two to four components of different sizes. Size about llfi. Separated-grains, size about Qn. Also simple grains of incomplete formation are present. Rumex acetosa Linn. {Polygonacece.) Dry root-stock. — Compound grains round to elongated-oval and conical, consisting of 2 to 4, rarely 6, components which are occasionally equal, though more frequently unequal. Size about 12/u. Separated-grains, size 2 to S/x; the larger ones have a cavity. Simi:)le grains of incomplete formation are also observed. TYPE 15. GRAINS COMPOUND, COMPONENTS FEW AND UNEQUAL. 271 Rumcr murilinnts Linn. {Polygonacew.) Dry roots. — Compound grains consisting of 2 to 1 equal or unequal components; the majority have fallen apart. Size about lifi. Separate d-grains isodiamctric or oval; instead of the hilum a small cavity is oliserved from which fissures fvcqupntlj' radiate toward the angles; occasionally with a large cavity, eccentricity about one-half. Size about 10m- Simple grains of incomplete formation. Aristolochia longa Linn. (Aristolochiacea:.) Dry root. — Compound grains consisting of 2 to 8 com- ponents which are occasionally equal and regularly arranged, though frequently unequal and of irregular arrangement. Size about 20 to 25//. Separated-grains, size about 16;u;0.5 to 1.5 times as broad as long; lamelkie none, or indistinct; instead of a hilum a small cavity from which two fissures diverge towards the distal end, one-half to one-fourth eccentric; very often 1 to 4 small components are attached to a large one. If two large components are united, they have usually a depressed form and very frequently even an oljlique form, and one or several small components are then found adhering at the line of junction. Plantogo maritima Linn. {Plantaginacecc.) Dry root-stock. — Compound grains round, rarely oval, consisting of 2 to 4 and 6 usually' unequal components. Size about 10 to 14yu. Separated- grains, size 3 to 11/:; rarely with a small cavity. Simple grains with complete formation are also foimd. Plantago media Linn. Fresh root-stock. — Separated-grains with one ciu-ved surface and 1 to 3 pres- sure facets; size 2 to 7/t; nearly all the compoimd grains have fallen apart. Simple grains of incomplete formation are also observed. Diodea dasycephala Cham. (Rubiacece.) Dry root. — Compound grains consisting of 2 to 12 compo- nents which are occasionally equal, but more frequently unequal, in size. (Often 1 to 7 com- ponents are attached at one end of a large grain, or 1 to 6 are firmly fixed at the line of union of two large components.) Size about 24/t. Separated-grains, size 3 to 14/(; usually with a small cavaty from which in the larger ones delicate radial fissures diverge. Riclmrdsonia scabra Linn. (Rubiacece.) Dry root. — Grains consisting of 2 to 4, rarely 6, mostly unequal components (rarely 3 to 4 small components attached to one large one). Size about 21 to 28/1. Separated-grains, size 2 to 21/i; the larger ones are oval; and rarely a small cavity instead of a hilum, eccentricity about one-half. Also simple eccentric-conical grains are present. Cephwiis ipecacuanha Rich. (Rubiacece.) Gray variety; dry root. — Compound grains rounded, elon- gated-oval, consisting of 2 to 6 components, aliout one-half of the grains having them equal and rogulaiiy arranged (many in tetrahedrons), and the other half containing unequal and irregularly arranged components. In the latter case, 1 to 4 small components frequently adhere to one end of a large one, rarely with 1 to 3 small ones attached at the line of junction between two lai-gc components. Size about 14 to 16/i. Separated-grains isodiamctric or oval, with one curved siuface and 1 to 4 pressure facets; lamella; none; solid or with a small cavitj' instead of a hilum, eccentricity about one-half. Ipecacuanha root, brown species, variety 1. — Compound grains rounded or elongated-oval, consisting of 2 to 6 and 10 components which are either equal and regularly arranged or unequal and irregularly arranged (in the latter case 1 to 4 small components are attached to one large one, and sometimes 1 to 3 small ones at the junction between the two large components). Size about 15 to 20/1. Separated-grains isodiametric or oval, with one curved surface and 1 to 5 pressure facets; homogeneous, or with a few very indistinct lamellse; solid, or instead of a hilum a small cavity is observed from which very short radial fissures occasionally diverge; eccentricity about one-third. Ipecacuanha root, brown species, variety 2. — Compound grains spherical or round-oval, rarely oval, consisting of 2 to 12, seldom up to 30, components which are usually equal and regularly arranged (quadrangular, tetrahedrons, hexahedrons, and octahedrons). Size about 23/i. Separated-grains isodiametric, usually with 1 curved surface and 1 to 6 pressure facets; lamellffi none; generally with a small central cavity and several short radial fissures. Size 4 to 12/1. Carum bxdhocastanum Koch.; Bunium bulbocastanum Linn. (Umbelliferce.) Dry tubers. — Compoimd grains consisting of 2 to 4 and 6 components which are usually unequal, though rarely equal. Size about 28/i. Separated-grains, size 3 to 20/i; instead of the hilum a small cavity is found in the larger ones from which fissures radiate, eccentricity about one-half and one-third. 272 DIFFERENTIATION AND SPECIFICITY OF STARCHES. Also simple eccentric-conical grains are present. In the two examples which were examined the one from Zwei])rucken contained compound grains, size about 28/x, and sejjarated-grains, size 5 to 20«; on the one from Zermatt the size of the compound grains was about 18m, and of the separated-grains 3 to 13m. Cheer ophyllum bulbosum Linn. {Umhellijerw.) Fresh tubers. — Separated-grains, size 2 to 12 and 18^; frequently more or less irregular; almost twice as long as broad; with 1 to 2, rarely 3 to 4, pressure facets; hilum about one-third eccentric. Most of the compovmd grains have fallen apart; the majority of them consisted of few (2 to 5) unequal components. Numerous simple grains are found which are round, low-conical shaped, rarely cuneiform, and frequently irreg- ular. Size 12 to 18/x; hilum about one-fourth eccentric. Starch plentiful, entirely filling the cells. Driviys icirdcri Forster.; Cortex ivinteranus interior. {Magnoliacece.) Dry bark. — Compound grains spherical or oval, consisting of 2 to 5 and 8 usually unequal components. Size about 20^. Separated-grains (very numerous), size 2 to 14^; more or less angular, usually with sharp edges; without a cavity. Isolated simple spherical or oval grains are also present. Anemone ranunculoides Linn. (Ranunculacece.) Dry root-stock. — Compound grains consisting of 2 to 10 and 16 equal or unequal components. Size about 22^. Separated-grains, size 2.5 to 8m; the larger ones have small cavity, and sometimes also delicate radial fissures. Simple eccentric-conical grains are also observed. Ranunculus bulbosus Linn. (Ranunculacew.) Dry tubers. — Compound grains consisting of 2 to 4 and more components which are sometimes equal, more often unequal; size about 14m. Sepa- rated-grains, size 3 to 9m; the larger ones have a small cavity, and occasionally also single dehcate radial fissures. Simple eccentric-conical grains are also present. Hellebarus viridis Linn. (Ranunculacew.) Fresh root-stock. — Compound grains consisting of 2 to 8 equal or unequal components; most of them have fallen apart. Separated-grains with 1 curved surface and 1 to 6 pressure facets; no lamelliB; rarely with a distinct hilum. Size 3 to 14m; the more polyhech-al forms, about IOm- Simple spherical grains are found; size about 14m. Helleborus dumetoruni Waldst. Kit. (Ranuncukicew.) Dry root-stock. — Compound grains consisting of 2 to 10 sometimes equal, more frequently unequal, components. Separated-grains, size 2 to 8m; the larger ones with a small cavity. Aconilum anthora Linn. (Ranunculacece.) Dry napiform root. — Compound grains consisting of 2 to 8 equal or unequal components; most of them fallen apart. Separated-grains, size 2 to 10m; isodiametric; with one cm-ved surface and 1 to G pressure facets; the larger ones with a small central cavity, and also several short radial fissures. Isolated separated-grains are oval; size 17m ; cavity at the hilum somewhat eccentric. Simple eccentric-conical grains are also present. Aconitum napellus Linn. (Ranunculacece.) Dry tuberous root. — Compound grains consisting of 2 to 8 equal or imequal components. Separated-grains without or with very delicate lamellie; they have a small, slightly eccentric cavity, instead of a hilum. Size of the more polyhedral separated-grains about 12m, of the oval about 20m. Pwonia officinalis Retz. {Ranunculacece.) Fresh root-stock. — Compound grains consisting of 2 to 4 equal or unequal components. Separated-grains, lamella; none, or delicate and few (2 to 3) ; hilum not often visible; size about ISm- Simple eccentric-conical grains are also observed. Sanguinaria canadensis Linn. (Papaveracece.) Dry root-stock. — Compoimd grains consisting of 2 to 4 and 6 sometimes equal, more often unequal, components. Separated-grains, size 4 to 11m; the larger ones with a small cavity. Simple grains of incomplete formation are also found. Nymphwa alba Linn. {Nymphaacecc.) Fresh root and root-stock. — Separated-grains isodiametric or oval with one curved surface and I to 5 pressure facets; lamellse none; hilum frequently distinct, in the oval forms about one-third eccentric. Size 2. .5 to 12m. Few intact compound grains, consisting of 2 to 8 equal or unequal components. Simple eccentric-conical grains are also found. Bryonia dioica Jacq. {Cucurbitacew.) Dry root. — Compound grains rounded or oval, occasionally slightly compressed ; consisting of 2 to 8 components which are rarely almost equal, but more frequently imequal (in the former case the components are arranged in a layer, in the center of which one or two supernumerary ones are suj)crimposed ; in the latter case, I to 6 small components are attached to the end of one larger one). Size about 30m. Separated-grains, without lamellse, cavity or fissures; in the largest ones eccentric lamellie are very rarely visible; size 3 to 24m. TYPE 10. GRAINS COMPOUND, MANY COMPONENTS. 273 Cereus variabilis Pfeiff. {Cactacem.) Fresh pith of the stem. — Compound grains rounded or oval, usually irrcpdar, consistins of 2 to 12 and 20 components which are almost always unequal. Size 40 to (iOyu. Sc^paratcd-grains, size 4 to 30 and 50ju; the larger ones with eccentric and frequently irregidar lamellte. Occasionally semi-compound forms with 2 or more inclosed components are observed. Simple grains, as in type 10 of incomplete formation arc also present. Malva borealis Wallmaim. (Malvacece.) Dry root. — Comjiound grains consisting of 2 to 4 and 8 components which arc sometimes equal, though more often unequal; size about 13yu. Sep- aratetl-grains, size 3 to 9fi; the larger ones have a small cavity, and occasionally also several very short, radial fissures. Simple grains spherical and rounded-oval are observed. Starch not very plentiful. Gossypium indicum Linn. {Malvacece.) Dry root. — Compound grains consisting of 2 to 10, rarely to more than 20, components which are equal, though more frequently unequal; size about 24/i. Separated-grains, size 3 to 16yu; the larger ones have a small cavity, and occasionally also single short radial fiss\u-es. Circcea lutetiana Linn. (Onagracece.) Dry stolons. — Compound grains round to oblong, frequently irregular, consisting of 2 to 12 components, which are either equal or unequal, and usually irregularly arranged; size about 24yu. Separated-grains, size 3 to 10^; the larger ones have a small ca\'ity instead of the hilum; eccentricity about one-half. Spirwa filipendiila Linn. (Rosacea;.) Dry roots. — Compoimd grains roimded to elongated spindle- shaped, most of them more or less irregular, consisting of 2 to 8 and 12 equal or unequal com- ponents; size about 16ai, separated-grains; size 2 to 12/i; the larger ones with a small cavity. Simple grains of incomplete formation are also present. Orobus albus Linn. (Leguminosw.) Dry thickened roots. — Compound grains consisting of 2 to 4 and 6 mostly unequal components; size about KV- Separated-grains, size 3 to 12;u; isodiametric or oval; frequently they have a small cavity instead of a hilum, which in the larger oval forms is about one-third eccentric. Simple eccentric-conical grains are also present. Ruppia maritima Linn. (Naiadaceoe.) Dry seeds. — Compound grains consisting of 2 to 3, rarely 4 to 5, components which are occasionally equal though usually unequal (in the latter case, I to 2 small components are attached to a large one). Separated-grains have central, fre- quently irregular, cavity from which radial or irregidar fissures diverge; size about 21ij.. Simple grains, probably centric-oval, are also present. Quercus pedunculata Ekrh. {Cvpidiferw.) Fresh cotyledons. — Compound gi'ains consisting of 2 to 3 and 4 comi)onents which are sometimes regular, though more often irregular, components. Separated-gi-ains isodiametric or oval; occasionally with a distinct central hilum; size about IRyu. Simple grains of incomplete formation are also present. Castanospermum australe C\mn. {Cupiliferce.) Dry cotyledons. — Compound grains spherical or oval, consisting of 2 to 8, rarely^ more, usually unequal components. Size about 20 to 25/^. Separated-grains; size 3 to 15/^; with one curved surface and I to 5 pressure facets; the larger ones have a small cavity, and single fissm^es extending to the sm^face of the pressure facets. Type 16. Grains Compound, Many Components. From 20 to many thousand components united into a compoimd grain separated by slits (sometimes almost invisible), and at full maturity falling away into separated-grains, most of which are outUned on all sides by flattened pressure facets. These compound grains probably arise from repeated division, and usually consist of almost equal or exactly equal components. Occasionally they are not associated mth other types, but often are found either with isolated grains of few and equal components (which belong to type 14), or very rarely with a few simple grains. Sometimes transitions between types 16 and 14 occur, making it cloul:)tful to which group such grains belong. As the components increase in number there is usually a decrease in their size and in the width of the separating fissures, the two, however, do not always occur in like proportion. If the separating lines decrease in width more rapidly than the components, the grain appears netted or delicatelj- reticulated, but if the components decrease in size until they have become about l/x in diameter, the whole grain appears granulated, and if they are smaller than l/z it looks homogeneous. In both cases the fact that the starch-grains are compoimd can be recognized only after they^ have fallen apart. Occasionally at certain stages in the development of an organ, for example in ripening seeds, 18 274 DIFFERENTIATION AND SPECIFICITY OF STARCHES. only separated-griiins arc found since tlie compound grains had fallen apart already within the cells or during the preparation. In such cases some doubt may exist as to whether the polyhedral forms are simjjle grains changed liy pressure, or whether they are separatcd-grains. Such cases must be decided by the life-history, or if that is lacking, by their analogy to related genera and species. Good examples may be found in the seeds of Chenopodiacece, Amarantacew, and Caryophyllacew. Arundo donax Linn. {Graminacew.) Fresh root-stock.— Compound grains rounded, granular, con- taining about 200 components; size 11/x. Separated-grains rounded or rounded-angular, rarely polyhedral; the larger ones have a small central cavity; size 0.7 to 6^. Crocus vermis All. (Iridacece.) Fresh bulbs. — Compound grains spherical, elongated-oval, consist- ing of 2 to 20 equal components; size about 20/i. Separated-grains usually polyhechal with acute angles; without lamellifi or hilum; size 2 to 9ai. Cypripedium calceolus Linn. (Orchidacece.) Dry root-stock. — Compound grains spherical or oval, consisting of 10 to 20 and 40 components; size about lO/x. Separated-grains, size L5 to 5/j; usually rounded-angular, compressed to about one-half their width. Dorstenia brasiliensis Liim.; Radix conlrajervoe. (Artocarpacew.) Dry root-stock. — Compound grains spherical or oval-spherical, consisting of 2 to about 50 equal components; size about 14/i. Separated-grains, size 2.5 to G and 7m; rounded-angular or polyhedral; the larger ones have a small central cavity. Simple grains of incomplete formation are also present. Dorstenia contrajerva Linn. (Artocarpaceoe.) Dry root-stock. — Compound grains spherical or oval, consisting of 2 to 40 and 60 eqi:al or unequal components; size aljout 12 to 14/i. Separated- grains; size 1.8 to 7m; rounded-angular or polyhedral; the larger ones hollow. Simple spher- ical grains are also found. Chiococca racemosa Linn. (Caincaroot). (Rubiacece.) Dry root-stock. — Compound grains rounded to sometimes rather irregidar, granular or reticulate, consisting of about 70 components; size about 18m- Separated-grains size 2 to 6(u; rounded or polyhedral; the larger ones hollow. Isolated compound grains have an appearance similar to a delicate parenchyma, which is due to the fusing of the components and the cavities being separated from one another by homogeneous walls, similar to the starch-grains found in type 12. ChrysophyUum ghjcyphlceum Casar.; Cortex monesice. (Sapotacece.) Dry bark. — Compound grains spherical or oval, consisting of 2 to about 60 components; size about lO/j. Size of the sepa- rated grains 2 to 5ai (not easily separated). Simple spherical or oval grains are also present. Podophyllum peltatum Linn. (Berbcridacece.) Dry root-stock. — Compound grains reticulated or merely granular, consisting of 3 to 20 and more components; size about IO/jl. Separated- gi-ains, size 2 to 4ju; roinnled-angular. Simple grains of incomjiletc formation also present. It is difficult to decide wlu^ther many of the grains are simple or sejiarated grains. Epimcdimn alpinum Linn. {Berberidacece.) Dry root-stock. — Compound grains spherical or oval, sometimes slightly angular and irregidar, reticulated or granular, consisting of 3 to 500 and more components; size about 15^. Separated-grains size 1.5 to 7m ; the smaller ones rounded, larger ones polyhedral. Epimediunimacra7ithu7nLmd\. (Berberidacew.) Dry root-stock. — Compound grains spherical, oval, con- ical, rarely somewhat angular as a result of pressure, granular, consisting of 12 to more than 2000 components; size about 17//. Separated-grains, size 1 to 2m; round or rounded-angular. Ayenia piisilla Linn. (Stercidiacew.) Dry roots. — Compound grains rounded or oval, consisting of 2 to about 40 usually equal components; size about 17m. Separated-grains frequently have a small cavity; size 2 to 8m. Starch not plentiful. Tribulus terrestris Linn. {Zygophyllacem.) Dry root. — Separated-grains, size 3 to 8m; almost poly- hedral, sometimes with a small cavity. Only a few intact compound grains are present, consisting of 2 to 12 or more components, which are usually equal. Starch rather scarce. Oxalis stricta Linn. {Oxalidacem.) Dry tap-root. — Compound grains round or oval, consisting of 2 to 20 equal or unequal components; size about 15m. Separated-grains, size 3 to 9m; the larger ones have a small cavity instead of a hilum, which is central or about one-half eccentric. Ordbus tuberosus Linn. (Leguminosw.) Dry tubers of the root-stock. — Compoimd grains round or oval; consisting of 4 to 20 equal or unequal components; size about 13m. Separated-grains, size 1.5 to 7m; the larger ones have instead of a hilum a small central or about one-half eccentric cavity. TYPE If). GRAINS COMPOUND, MANY COMPONENTS. 275 Leersia oryzoides Swartz. {Graminacecv.) Dry endosperm. — Compound grains round or oval, fre- quently somewhat anp;uhir as a result of pressure, sometimes almost polyhedral, consisting of about (iOOO or more; components; size about 25 and 30/u. Separated-gi-ains rounded- angular, or i)olyhedral; the larger ones with a small cavity; size 2 to 7fi. Oryza sativa Linn, ((rrnminacew.) Dry endosperm. — Compound grains spherical or oval, consisting of 4 to 101) or mor(> almost cqiuil comjionents; size about 25/^. Separated-grains polyhedral; fretiuently with a rather large cavity; size 3.5 to 8/x. Hydropyrum esadenlum Linn. ; Zizania aqiialica Linn. {Grammacew.) Dry endosperm. — Compound grains, polyhedral as a result of pressure; with numerous components; size 25 and 28;u. Sep- arated-grains rounded-angular; size L5 to 4/u. Zinzania davulosa Michx. {Graminacece) probably belongs to this type. The compound grains consist of numerous components; they are isodiametric to oblong, and polyhedral as a result of pressure; size about 30/u. Separated-grains; size 2 to 5/i; the smaller ones rounded-angular, the larger ones polyhedral and hollow. Ehrharta panicea Smith. {Graminacecr.) Dry endosperm. — Compound grains spherical to elon- gated-oval, occasionally somewhat irregular, homogeneous, or reticulate and granular, con- sisting of more than 3000 components. Size about 40 and 45/x. Separated-grains poly- hedral, usually with sharp borders and angles; frequently hollow; size 2 to 6yu. Microlwna stipoides R. Br. {Graminacece.) Dry endosperm. — Compound grains almost round or oval, frequently' angidar as a result of pressure, numerous components; about 20/x. Sepa- rated-grains, size 2 to 5 and 7^; the smaller ones are rounded-angular, the larger ones poly- hedral, and have a cavity. Cormicopice cucullatum Linn. {Graminacece.) Dry endosperm. — Compound grains rounded to elon- gated-lanceolate, sometimes slightly irregular; two-fifths to as thick as long; reticular-granu- lated; size about 54/^. Separated-grains rounded-angular or polyhedral, usually wth a cavity; size L5 to 6/j. Phalaris canariensis Linn. (Graminacece.) Dry endosperm. — Compound grains spherical to oblong, rarely angular, half to as thick as long, reticulate-granulated, rarely almost homogeneous, consisting of 4 to 2000 components; length about 36ai, thickness about 25/j. Separated- grains rounded-angular or polyhedral; size 2 to B/x. Phalaris bulbosa Cav.; Phalaris cccndescens Desf. (Gramiiuicew.) Dry endosperm. — Starch as in the preceding. Compound grains consisting of 3000 and more components; size about 48/i. Antho.Tanthnm amarum Brot. {Graminacem.) Dry endosperm. — Compound grains spherical or oval, reticulate-granulated; size about 25yu. Separated-grains rounded-angular; size 1 to 4/i. Hierochloa borealis Roem. and Schult. {Graminacece.) Dry endosperm. — Compound grains spherical or oval, sometimes slightly irregular; consisting of 1000 or more components; size about 30 to 40jLi. Separated-gi-ains with rather sharp borders and angles; size 2 to 7^. Holcus lanalus Linn. {Graminacece.) Dry endosperm. — Compound grains round or oval, homo- geneous or distinctly granular, consisting of 1000 or more components; size about 14/i. Size of the separated-grains 1 to 2n. Bechmannia erucccformis Hochst. {Graminacece.) Dry endosperm. — Compound grains rounded-oval to oblong, finely granular; size about 28 to 35iU. Separated-grains round or polyhedral; size 1 to 3 and 4/a. Lygeum spartum Liim. {Graminacece.) Dry endosperm. — -Compoimd grains spherical to elongated- oval, reticulated-granular; size about 24>i. Separated-grains rounded angular or poly- hedral; size 1 to 4m . Milium effusum Linn. {Graminacece.) Dry endosperm.— Componnd grains spherical or oval, reticu- lated-granular; size about 24/i. Separated-grains rounded to polyhedral, frequently with a small cavity; size 1.5 to 5/i. Milium vernale Biebrst. Dry endosperm. — Compound grains spherical or oval; homogeneous, or granulated by delicate reticulations; size about 14yu. Separated-grains rounded-angular; size 1 to 3ju. Panicum commelincefolium Rudge. {Graminacece.) Dry endosperm. — Compound grains rounded or oval, frequently more or less angular as the result of pressure, granular; size about 12 to 15^1. Separated-grains roimded; size 1 to 2.5yLi. Starches of the seeds from Brazil and Guate- mala agree perfectly with those above described. 276 DIFFERENTIATION AND SPECIFICITY OF STARCHES. Urochloa deprcssa Steiid. (Graminacccc.) Dry endosperm. — Conijioiind grains isodiametric to oblong, more or less polyhedral as the result of pressure, reticulated-granular or granular; size about 20]U. Separated-grains round or rounded-angular; size 0.7 to 2.5^. ArundineUa nepalensis Trin. {Graminaeeee .) Dry endosperm.- — Compound grains spherical, or oval, or low cone-shaped (not angular) ; two-fifths to as thick as long; reticulated-gTauular or almost homogeneous; consisting of 4 to 1000 or more components; size about .30/i. Separated- grains size 2 to Gfi; the larger ones poljdiedral, and have a cavity. Dichelachne vulgaris Trin. (Graminacew.) Dry endosperm. — Compound grains rounded, elliptical, oblong, conical; delicately granular; consisting of 1000 or more components; size about 18;:. Separated-grains roinided; size 0..5 to 2,u. Urachnc parviflora Trin.; Piplatherum midtiflorum Beauv. {Graminacea.) Dry endosperm. — Com- pound grains round or oval, consisting of 6 to 1000 or more equal components; size about 18m. Separated-grains rounded-angular, the larger ones polyhedral, usually with a central cavity; size 2 to 7yu. Stipa papposa Nees. (Graminacece.) Dry endosperm. — Separated-grains rounded-angular, the larger ones polyhedral, and have a cavity; size 2 to 6 and 7ix. Of compound grains only fragments and a few complete ones remain. Stipa gigantea Lagasc. (Graminacew.) Dry endosperm. — Separated-grains rounded-angular, the larger ones with a small cavity ; size 1 to 4;u. Only a few compound grains are present. Simple centric-lenticular grains are also found. Stipa pennata Linn. (Graminacea;.) Dry endospcr7n. — Separatcd-gi'ains rounded-angular or poly- hedral; size 2 to 4yu. Compound grains are only observed within the cells. Lasiagrostis calamagrostis; Stipa calamagrostis Whlbrg. (Graminacece.) Dry endosperm. — Com- pound gi-ains round or oval, frequently slightly angular, as a result of pressure; reticulated- granular; falling apart easily; size about \5ti. Separated-grains usually polyhedral, fre- quently have a cavity; size 2 to 9 and ll;u. Grains which are probably simple are found among the above; size 9 to 11 and 13^; with central cavity and occasionally single rachal fissures. Aristida hystrix Linn. (Graminacece.) Dry endosperm. — Separated-grains, size 1..5 to 6/k; the smaller ones roimded, larger ones polyhecbal, and have a cavity. Compound grains embedded in protoplasm are only observed wthin the cells of thin sections of the tissue; they are rounded or rounded-angular; size about 15 to 20^. Ari.'stida amplissima Trin.; Aristida stipifnrmis Lam. (Graminacece.) Dry endosperm. — Separated- grains polyhedral, antl have a cavity; size 2 to lOyu. A few free compound grains which in thin sections can be distinctly seen within the protoplasm, which latter takes a yellow color when treated with iodine. Grains spherical, oval, or rounded-angular; reticulated; size about 28(U. Aristida plumosa Linn. (Graminacea;.) Dry endosperm. — Separated-grains ]3olyhedral; the larger ones have a cavit}'; size 2 to 8/i. Not many distinct compound grains; round, oval, or ellip- tical; reticulate, or reticulated-granular; size about 26^. Aristida funiculata Trin.; Aristida kotschyi Hochst. (Graminacea;.) Dry endosperm. — Separated- grains, size 1 to 5m; smaller ones rounded, larger ones polyhedral, and have a cavity. Compound grains few, round or oval. Nardus stricta Linn. (Graminacea;.) Dry endosperm. — Compound grains spherical or oval (not an- gular), three-fifths to as thick as long, and consisting of 4 to 300 components; size about 18/i. Separated-gi'ains rounded-angular, the larger ones polyhedral, and have a cavity; size 1.5 to dfjL. Mgopogon cenchroides Willd. (Graminacea;.) Dry endosperm. — Separated-grains polyhedral, the larger ones have a cavity; size 2.5 to 7fx. A few compound grains are free and some are indistinctly observed ^vithin the cells; round or o^'al and more or less angular as a result of pressure; size about 20^. JEgopogon multisetus Trin. (Graminacece.) Dry endosperm. — Starch as in the preceding. Size of the separated-grains 2 to 6^. Lycurus phalaroides Humb. Bonpl. (Graminacece.) Dry enrfosperm.— Separated-grains polyhedral, and have a cavity; size 2 to lO/i. Compound grains are observed indistinctly within the cells, and also a few free grains; isodiametric or oval, more or less polyhedral due to pressure; size about 25^. TYPE If). GRAINS COMPOUND, MANY COMPONENTS. 277 Phippsia algida R. Br. {Graminacece.) Dry endosperm. — Compound grain.s round or oval, more or less angular, as a result of pr(>ssure; reticulate or retieulated-angular. Size about 20 to 25^. Separated-grains rounded-angular or polj'hedral, the larger ones with a cavity; size 1.5 to 5;u. C oleanthus siiblUis Seidel; Schmidtea utriculosa Sternb. {Graminacece.) Dry endosperm. — Compound grains spherical to elongated-oval, more or less angular as a result of pressure; homogeneous or reticulated-granular; size about 24ju. Separated-grains rounded-angular or polyhedral, the larger ones with a cavity; size 2 to 5/1. Alopecurus geniculatus Linn. {Graminacece.) Dry endosperm. — Compoimd grains spherical or oval; delicately granular, or reticulated-granular; size about 2.5//. Separated-grains rounded- angular, rarely polyhetiral; size 1.5 to 4^:. Alopecurus pratensis Linn. {Graminacece.) Dry endosperm. — Compound grains spherical or oval, reticulated-granular; size about 26m, rarely 33yu. Separated-grains rounded-angular, or polyhedral; the larger ones with a cavity; size 1.5 to bix. Alopecurus alpinus Smith. {Graminacece.) Dry endosperm. — Starch as in the preceding species. Alopecurus utriculatus Schrad. {Graminacece.) Dry endosperm. — Starch as in Alopecurus pratensis. Size of the separated-grains about l/x, of compound grains about 33ju. Phleum asperum Vill. {Graminacece.) Dry endosperm. — Compound grains almost rounded or oval, sometimes slightly irregular, three-fifths to as thick as long, consisting of 2 to about 60 equal components; size about 38//. Separated-grains polyhedral, usually with sharp borders and angles, frequently with a cavity and single fissures; size 6 to 16//. Phleum tenue Sclu^ad.; Achnodonton hellardi Beauv. {Graminacece.) Dry endosperm. — Compoimd grains spherical or oval, frequently slightly angular and irregular, consisting of 4 to about 60 components; size about 40//. Separated-grains polyhedral, usually with sharp borders and angles, frequently with a cavity from which single radial fissures proceed; size 4 to 14//. Crypsis schcenoides Lam. {Graminacece.) Dry endosperm. — Compound gi'ains almost round elon- gated-oval, half to as thick as long, reticulated-granular or almost homogeneous; size about 30//. Separated-grains polyhedral, frequently with sharp margins and angles, the larger ones have a cavity; size 2.5 to 7//. Vilja coromandelina Beauv.; Sporobolus coromandelina Kunth. {Graminacece.) Dry endosperm. — Compound grains spherical or oval, three-fifths to as thick as long, reticulated-granular, consisting of 6 to 1000 and more components; size about 19//. Separated grains rounded- angular, or polyhedral; the larger ones usually with a small cavity; size 2 to 5//. Agrostis verlicUlata Vill. {Graminacece.) Dry endosperm. — Compound gi-ains round or oval reticu- lated-granular; size about 20/i. Separated-grains rounded to polyhedral; size 2 to 7//. Triachyrum longifolium Hochst. {Graminacece.) Dry endosperm. — Compound grains, isodiametric or oval, more or less polyhedi'al as a result of pressure, consisting of numerous components, most of them fallen apart; size about 21/i. Separated-grains polyhedral, the majority with a large or small cavity; size 3 to lO/i. Triachyrum cordofanuyn Hochst. {Graminacece.) Dry endosperm. — Separated-grains, size 1.5 to 5//; the smaller ones rounded-angular, the larger ones polyhedral, and have a cavity. Only a few intact compound grains are observed; these are reticulated and about 14// in size. Colpodium steveni Trin. {Graminacece.) Dry endosperm. — Compound grains round, oval, and low- cone-shaped, slightly angular or almost polyhedral, consisting of 400 and more components; size about 25//. Separated-grains 2.5 to 6/i, polyhedral, the larger ones frequently with a small cavity. Apera spica-venii Beauv.; Agrostis sp. Linn. {Graminacece.) Dry endosperm. — Compound grains round to oblong, two-fifths to as thick as long, granular or almost homogeneous; length about 25//, thickness about 15//. Separated-grains round or rounded-angular; size 0.7 to 2//. Muehlenbergia loilldenowii. {Graminacece.) Dry endosperm. — Compound grains spherical or oval, reticulateil-granular (most of them have fallen apart); size about 24//. Separated-grains, polyhedral; the larger ones have a cavity; size about 7//. Cinna racemosa Kunth.; Muehlenbergia glomerata Trin. {Graminacece.) Dry endosperm. — Separated- grains polyhedral, the larger ones have a cavity; size about 17//. Among the above there are a few spherical or oval compound grains wiiich have not fallen apart; size about 22/i. 278 DIFFERENTIATION AND SPECIFICITY OF STARCHES. Cinna arundinncea Linn. iClraminacca.) Dry endosperm. — Compound grains spherical or oval, frequently somewhat angular, consisting of numerous components, most of them fallen apart ; size about 20ai. Separated-grains rounded-angular to polyhedral ; the largerones have acavity ; size 2 to 5 and 7/j. Echinopogon ovatus Beauv. (GraminacecB.) Dry endosperm. — Compound grains round, oval, or low cone-shaped, more or less angular as a result of pressure, delicately reticulated; size about 20/u. Separated-grains rounded-angular to polyhedral; size 1 to 4/i. Lagurus ovatus Limi. (Granimacece.) Dry endosperm. — Com])ound grains round or oval, homogen- eous or finely granulated, consisting of 10 to over 6000 components; size about 20/i. Sep- arated-grains round or somewhat angular; size 1 to 3/i. Polypogon jnonspeliensis Desf . (Graminacea.) Dry endosperm. — Compomid grains oval, rarely round, one-half to almost as thick as long; finely granulated, frequently almost homogeneous, consisting of 30 to 1 000 components ; size about 18yii, angular. Separated-grains arealso present ; size 1 .5 to 4 /^ . Chwturus fasciculatus Link. {Graminacece.) Dry endosperm. — Separated-grains round or rounded- angular; size 1 to 3/i. Compound grains have all fallen apart. Gaslridium australe Beauv., var. muticum Spreng. (Graminaceoe.) Dry endosperm.. — Separated- grains polyhedral, usually with sharp angles and borders, with a cavity; size 3 to 12/i. Com- pound grains nearly all have fallen apart; those remaining are more or less angular or even exactly polyhedi'al as a result of pressure; size about 23/i or more. Perotis latifoUa Ait. {Graminaceoe.) Dry endosperm. — Separated-grains, size 2 to 8yu; the smaller ones rounded-angular, the larger ones polyhedi-al, usually with a larger or a small cavity. The cells are crowded with separated-grains, while no compound gi-ains could be positively distinguished. Chamagrostis minima Borckh. ; Sturmia minima Hoppe. {Graminacece.) Dry endosperm. — Compound grains round or oval, three-fifths to as long as broad, usually finely reticulated, consisting of 8 to 500 components; size about 33yu. Separated-grains usually polyhedral, without a cavity; size 2 to 8 and 9/i. Calamagrostis sylvaiica DC. {Graminacece.) Dry endosperm. — Compoimd grains rounded or oval consisting of 2 to more than 300 components, which when few in number are often unequal ; size about 48/i. Separated-grains usually polyhedral; the larger ones with a central cavity; size 2 to 9;u. Calamagrostis icilldenowii Trin.; Deyeuxia retrofracta Kunth. {Giaminacece.) Dry endosperm. — Compound grains spherical or oval, consisting of numerous components; size about 16;u. Separated-giains angular or romided-angular; size L5 to 4/x. Ammophila arenaria Link.; Calamagrostis arenaria Roth. {Graminacece.) Dry endosperm. — Com- pound grains spherical to elongated-oval, two-fifths to as thick as long, reticulated-gran- ular; size about 50;u. Separated-grains romided-angular or polyhedral; size 1 to 4^. Arundo niawitanica Desf. {Graininacea.) Dry endosperm. — Separated-grains size 2.5 to S/i, polyhedral, the larger ones usually with a small angular cavity. A few intact compound grains are present, romided-oval, frequently angular, or even polyhedral, consisting of numerous components. Phragmiics communis Trin. {Graminacece.) Dry endosperm. — Separated-grains rounded or rounded- ang-ular; the larger ones rarely witli a distinct cavity; size 0.7 to 2.5fi. Only a few compound grains are present; small, romided, granular. Gyrwium cinereiim Humb. {Graminacece.) Dry endosperm. — Compound grains spherical to elon- gated-oval, frequently slightly angular, reticulated-gi-anular; size about 23/1. Separated- grains, size 2.5 to 8/^; the smaller ones rounded-angular, the larger ones polyhedral with sharp borders and angles, frequently with a small cavity. Gyneriuni argenteum Nees. {Graminacece.) Dry endosperm. — Separated-grains size 1 to 4/j, rounded- ang-ular to polyhedral. Only a few compound grains could be positively distinguished. Besides the starch, much oil and protoplasm are foimd in the cells. Pappophorum schimperianum Hochst., var. persicum. {Graminacea-.) Dry endosperm. — Compound grains spherical or oval (not at all angular), consisting of 2 to over 100 comi)onents; size about 25 and 30/i- Separated-grains polyhech-al, usually with a small central cavity, and sometimes also with single delicate radial fissm'es; size 3 to 10, rarely 12/^. The isolated sinijile grains, size about IG/x; the smaller ones si>herical or rounded-oval, the larger ones circular or roundeil-oval ami compressed to about half theii' wiilth. Tlie starch-grains which are embedded in jsrotoplasm are not closely packed in the cells. TYPE in. GRAINS COMPOUND, MANY COMPONENTS. 279 Pappophorum macrostachywn Nees. {Gramiiiacece.) Dry endosperm. — Separated-grains size 3 to 8 and 10m; polylicdral or with 1 curved KiirfaiH? and 1 to 5 pressure facets; usually with a large or a small cavity. Niigeli observed only a few compound grains, ami also spherical or oval simple grains. Poppophorum nigricans R. Br. (Graminacem.) Dry endosperm. — Compound grains spherical or oval, consisting of 2 to 10 or more equal comjjonents; size about 30 and 30^. Separated- grains size 5 to 20 antl 2,')/^; usually with one ciirvetl surface and 1 to 5 pressure facets; rarely entirely polyhedral (only outlined by j)ressure facets) ; with a central cavity antl radial fissm-es. Simple centric-spherical gi'ains are also present. The compound grains of Pappophorum nigricans properly belong to type 14. They are placed here because of their relationshij) to the other two species. Ptiloneilema plumosum Steud. ; Etdriana abyssinica R. Br. {Graminacece.) Dry endosperm. — Sep- arated-gi'ains polyhedral, frequently with a small cavity and also single delicate radial fis- sures; size 3 to 12/Li. The majority of the compound grains have fallen apart, the few remain- ing ones are polyhedral as the result of pressure. Echinaria capitata Desf. (Graminacew.) Dry endosperm. — Compound grains (few intact ones still present), spherical or oval, frequently somewhat angular; size about 34/i. Separated-grains polyhedral with sharp borders and angles; size 4 to 18ju. Ctenium elegans Kunth. {Graminacem.) Dry endosperm. — Separated-grains solid or with a central ang-ular cavity, frequently with 1 curved surface and 1 to 5 pressure facets, many of them entirely polyhedral. The majority of the compound grains have fallen apart, those remain- ing are rounded or oval, consisting of 2 to about 20 components; size about 12/i. Simple grains few, spherical, or rounded-oval, solid or with a small central cavity; size about 10/i. Ctenium chapadense Trin. (Graminacew.) Dry endosperm. — Separated-grains, size about 11/i; frequently jiolyhedral (only outlined l)y pressure facets); many of them have one curved surface and 3 to 5 pressure facets; the larger ones have an angular cavity from which short radial fissures proceed. Very few compound grains are present. Microchloa setacea R. Br. {Graminacew.) Dry endosperm. — Separated-grains polyhedral, the majority with a large or a small cavity; size 3 to lOyu. Only a few free compound grains are present; oval or almost polyhech-al; size about 20 to 2.5/i. Chloris petrwa Thumb.; Eustachys petrwa Desv. (Graminacece.) Dry endosperm. — Separated-grains polyhedral, the larger ones with a central cavity; size 2 to 9^- All the compound grains have fallen apart. Chloris submutica Humb. Bonp.; Eustachys submutica R. and S. (Graminacece.) Dry endosperm.— Compound grains spherical or oval, reticulated; size about 24/i. Separated-grains rounded- angular to polyhedral; rarely with a small cavity in the larger ones; size 2 to 8yu. Ctenopsis pectinella Taris.; Festuca Nop. Del. {Graminacece.) Dry endosperm. — Separated-grains rounded or rounded-angular; size 0.8 to 3//, a few compound grains are still found, granular; size about 14yu. Eleiisi7ie caracana Gai-t. {Graminacew.) Dry endosperm. — Separated-grains polyhedral, only outlined by pressure facets or with one curved surface and several pressure facets, with sharj) nuirgins and angles; the larger ones have a small central cavity from which several fissures frequently radiate toward the angles; size 4 to 13^. The greater number of the compound grains have fallen apart; the remaining ones are spherical or oval, about 30/i in length, and consist of 2 to 50 and 100 equal components, which when few in number (4 to 10) show a regular arrangement. Dactyloclenium wgyptiacum Willd. {Graminacew.) Dry endosperm. — Compound grains spherical or oval (not at all angular), granular or reticulated-granular; size about 15 to 21/i. Separated- grains rounded-angular to polyhedi'al; the larger ones with a small cavity; size 1.5 to 5.5/i. Cynodon daclylon Pers.; C. linearis Willd. {Graminacew.) Dry endosperm. — Se]5arated-grains poly- hedral, usually with a small or large angiilar cavity; size 2.5 to 8/^. A few free compound grains are observed, oval or nearly polyhedral; size about 20^. Chondrosium sj). {Graminacew.) Dry endosperm. — Sei)arated-gTaius rounded-angular to poly- hedral; size 2 to l'^^l. Onlj' a few compound grains are ob.served. Sparlina cynusuroides Willd. {Graminacew.) Dry endosperm. — Separated-grains j)olyhedraI, several with a small cavity. Only a few intact compound grains still present, consisting of 2 to 12 components. 280 DIFFERENTIATION AND SPECIFICITY OF STARCHES. Euiriann oligostachya Kiinth.; Athcropogon oligostochya Niitt. (Granmwcece.) Dnj endosperm. — Separjited-grains polyliodral, the larger ones with an angular cavity; size 2 to S/n. (Jnly a few compound grains are obser\'ed, consisting of 2 to over 300 components, spherical, oval, or, as the result of pressure, angular; size about 20ai. Melanocenchrys royleana Nees.; Pommereulla royleana Steud. (Graminacece.) Dry endosperm. — Compoimd grains spherical or oval, frequently angular by reason of pressui'e, consisting of 2 to 200 or more com]5onents; size about 20 to 25ju. Most of the separated-grains are poly- hedi-al, many with 1 curved surface and several pressure facets; the larger ones with a small central cavity, also occasionally with single radial fissures; size 3 to 12,u. Corynephorus canescens Beauv. ; Aira canescens lAim. (Graminacece.) Dry endosperm. — Compound grains spherical or oval, granular or homogeneous; size about 13/j. Separated-grains rounded; size 0.7 to 1.5/i. Deschampia jxincea Beauv.; Aiva juncea Vill. {Gi-aminacem.) Dry endosperm. — Compound grains spherical or oval, frequently angular as a result of pressure, reticulated-granular or granular; size about 24^. Separated-grains size 2 to Tyu ; the smaller ones rounded-angular, the larger ones polyhedral, and frequently with a small cavity. Deschampia ccespitosa Beauv.; Aiva cccspitosa Linn. {Graminacece.) Dry endosperm.- — Starch as in the preceding. Size of compound grains about 30/j, of separated-grains 2 to S/i. Deschampia pulchella Trin., var. tenovei; Aiva pulchella Willd., var. tenorei Guss. {GraminacecE.) Dry endosperm. — Separated-grains rounded-angular; size 2 to 6/i. Airopsis globosa Desv.; Aiva globosa Thore. {Graminacco!.) Dry endosperm. — Compound grains spherical, oval, pyriform, occasionally slightly angulai- by reason of pressure; more or less distinctly granular, or often homogeneous; size about 12yu. Separated-grains round or rounded angular; size 1 to 3/i. Airopsis agrostidea DC; Aiva agrostidea Loise. (Graminacece.) Dry endosperm. — Starch as in the preceding species. Trisetum argenteum Reen and Schult. (Graminacece.) Dry endosperm. — Compound grains spherical or oval, occasionally angular tlu-ough pressure; delicately reticulated or granular-reticular to homogeneous; consisting of 3 to 100 components; size about 14fx. Separated-grains rountl, more or less angular; size 1.5 to 5/x. Trisetum neglectum Willd. (Graminacece.) Dry endosperm. — Starch as in the preceding species. Size of the compound grains 13;u, of separated-grains 1.5 to 4/i. Avena orientalis Schreb.; Avena hirsuta Roth; Avena brevis Roth. (Graminacece.) Fresh and dry endosperm. — Compound grains almost round to oblong, frequently somewhat angular through pressm-e; consisting of 2 to 300 equal or somewhat imequal components; size about B/j.. Sep- arated-grains polyhedral with rather sharp margins; if fresh they have a small central indis- tinct hilum, when dry, they frequently have a small cavity instead of the hilum; size 7 to 12^. Simple spherical grains are found among the above described. Gaudinia Jragilis Beauv.; Avena fragilis Linn. (Graminacece.) Dry endosperm. — Separated-grains rounded-angular to polyhedral; the larger grains \vith a small central cavity; size 2 to G, rarely 1 to 8ju. Compound grains few, spherical or oval, consisting of 5 to 200 almost equal components; size about 21yu. Arrhenatherum elatius Mert. & Koch.; Avena elatius Linn. (Graminacece.) Dry endo-^perm. — Com- pound grains spherical, oval, or pjTiform; consisting of 4 to 400 or more components; size about 30/i. Separated-grains polyhedral; size 2 to lOyu. Eriachne ampla Nees; Airopsis ampla Nees. (Graminacece.) Dry endosperm. — Separated-grains, size 2 to 6 and 7fi; the smaller ones rounded-angular, the larger ones polyhedral with a central cavity. A few free compound grains observed; round or oval, frequently angular; size about 21ai. Eriachne microphylla Nees. Dry endosperm. — Compound grains rountled or oval, more or less angular by pressure, granulated; size about 15/:. Separated-grains 1 to 4/i ; rounded or rounded-angular. Trislachya barbata Nees. (Graminacca;.) Dry endosperm. — Separatetl-grains polyhedral; the larger ones occasionally with a small central cavity; size 1.5 to 5 and G/z. A few free comjiound grains are observed, only perceived indistinctly within the cells; lounded or oval, frequently angular by reason of pressure; size about 20/i. Isolated simple spherical grains and many separated-grains with one curved surface and one or several pressure facets are also present. TYPE 10. GRAINS COMPOUND, MANY COMPONENTS. 281 Danlhonia pwrincialis T>C. (Craminacca;.) Dr;/ endosperm.— Compound grains spherical to cIoti- gated-oval, sonii'timcs irronnilar; rcticulated-graimlar; size about '38ix. Separated-grains usually polyhedral; the larger one.s with a small cavity; size 2 to l/x. Danthonia koesllini Hochst. (Graminacece.) Dry endosperm. — Separated-grains, size 1.5 to 5 and 6n; the smaller ones round, the larger ones jiulyhedral and with a central cavity. Few compound grains observed; rouiul or oval, frequentlj' angular; size about 20(U. Uralepis aristulata Nutt. {Graminacece.) Dry endosperm. — Separated-grains rounded angular to polyhedral: size 2 to 8 and lO/i; the larger ones with a small central cavity and often ^\^th single radial fissures. Only a few comi)ountl grains of polyhedral form can positively be distinguished. Triodia decumbens Beauv.; Danthonia deciimbens DC. {Graminacece.) Dry endosperm.- — Compound grains rounded-angular or polyhedral; size about 20 and 25yu. Separated-gi-ains polyhedral; the large ones have a small cavity, also frequently radial fissures; size 2.5 to 8/i. Poa nemoralis Linn. {Graminacece.) Dry endosperm. — Compound grains spherical or oval (not at all angular), consisting of about 1000 components; size about 30 to 36yu. Separated-grains usually polyhechal, frequently more or less distinctly compressed, without a cavity; size 2 to 8/i. Eragrostis abyssinica Linlt. {Graminacew.) Dry endosperm. — The compound grains have become polyhedral as the result of pressure, mostly with acute margins and angles; consisting of 20 to 500 and more components; size 33/n. Separated-grains polyhechal with acute margins and angles, rarely with a small central cavity; size 2 to 12^. Brizopyrum siculum Link.; Fesluca unioloides Kunth. {Graminacece.) Dry endosperm. — Compound grains almost rounded-oval, oblong, frequently irregular, homogeneous or finely granulated; size about 16/u. Separated-grains round or rounded-angular; size about 1 to 2 or rarely 3/i. Brizopyrum acutiflorum Nees. {Graminacew.) Dry endosperm. — Separated-grains polyhedral, and have a cavitj'; size 2 to 7yu. Only very few compound grains are observed. Briza triloba Nees.; Calotheca triloba Beauv.; Chascolytrmn triloba Nees. {Graminacece.) Dry endo- sperm.— Compound grains spherical or oval-spherical, frequently somewhat angular, finely reticulated or reticulated-granular, sometimes almost homogeneous, consisting of 4 to 400 components; size about 16;u. Separated-grains round or angular; size L5 to 4/i. Briza geniculata Thumb. {Graminacece.) Dry endosperm. — Separated-gi-ains rounded or angular; size 2 to In] nearly all the compound gi-ains have fallen apart. Glyceria nervata Trin.; Glycerin michauxii Kunth. {Graminacece.) Dry endosperm. — Separated- grains polyhedral, usually with acute margins and angles; occasionally with pressvu-e facets; the larger ones with a small central cavity; size 2 to 7 and 9m. Compound grains few, round or oval; size about 30/u. Glyceria distans Mert. & Koch.; Festucathalassica Kunth. {Graminacece.) Dry endosperm. — Compound grains spherical or oval, reticulated-granular or with distinct lines of separation; size about 28^. Separated-grains 2 to 7^; the majority are polyhedral, the larger ones have a cavity. Glyceria maritima Mert. and Koch. ; Festuca thalassica Kunth. {Graminacece.) Dry endosperm. — Com- pound grains rounded or oval; finely reticulated or with distinct separating lines; size about 25^. Separated-grains usually polyhedral, the larger ones have a cavity; size 2 to 7,u. Catabrosa aquatica Beauv.; Ghjceria aqucdica Presl. {Graminacea.) Dry endosperm. — Compound grains round to oblong and high cone-shaped, reticulated-granular; size about 28yu. Sepa- rated-grains, size L5 to 4^; the smaller ones round, the larger ones angular. Lophochlcena californica Nees. {Graminacece.) Dry endosperm. — Separated-grains size 1.7 to 5 and 6m; the smaller ones somewhat rounded-angular, larger ones polj'hedral and ^vith a cavity. Compound grains distinctly observed \\'ithin the cells, but only a few are found free (since they fall ajmrt as they emerge from the cells). They are rounded or oval, more or less angular as the result of pressm-e; size about lO/i. LophocUwna obtusiflora Trin. {Graminacece.) Dry endcsperm. — Separated-grains polyhedral; usu- ally with a consi)icuous more or less angular cavity; many also have radial fissures. No free compound grains are observed; ])olyhe(hal when found in the cells; size about 16 to 21m- Melica ciliata Linn. {Graminacece.) Dry endosperm. — Compound grains polyhedral by means of pressure, sometimes with acute, sometimes rounded angles and margins, consisting of 4 to over 500 components; size about 40m. Separated-grains polyhedral, usually with acute margin and angles; the larger ones frequently with a small central cavity; size 2 to 15m. 282 DIFFERENTIATION AND SPECIFICITY OF STARCHES. Molinia cwrvlcn Mocnch. {draminacca'.) Dry cndospam. — Soiiaratod-grains rouiuled-anftiilar or polj'licdral, with coniptiiatively large cavity; size 2 to 5yu. Cuinixjuuil grains have nearly all fallen apart. Koeleria laxa Link. [Graminacfce .) Dry endosperm. — Compound grains spherical, oval or elliptical (not at all angular), three-fifths to as thick as long, granulated, consisting of 5 to over 2500 components; size about 15yu. Separated-grains round or slightly angular; size 1 to 3(U. Schismus marginatus Beauv. (Graminacew.) Dry endosperm. — Compound giains spherical or oval, frequently somewhat irregular, reticulated-granular; size about 21/i. Separated-grains angu- lar, mostly polyhedral, frequently have a cavity; size 2 to G/x. Sesleria eloiigala Host. {Graminacca;.) Dry endosperm. — Compound grains, spherical or oval, fre- quently irregular, reticulated-granular; size 24 to 28/i. Separatetl-grains, rounded-angular, or mostly acute-polyhedral, very frequently have a cavity; size 2 to 6 and 7yu. Cynosurus echinatus Limi.; Chrysurus echinatus Beauv. {Graminacece.) Dry endosperm. — Com- jiound grains spherical to oblong and conical, frequently irregular, two-fifths to just as thick as long, granular or homogeneous, consisting of 10 to over 13,000 almost equal com- ponents; length about 42 to 51;u, thickness about 30 to 36//. Separated-grains, round or rounded-angular; size 0.7 to 2, rarely 3//. Lamarkia aurea Moench. (Graminacece.) Dry endosperm. — Compound grains rounded-oval to ob- long, half to almost as thick as long, homogeneous or finely granular, consisting of 30 to over 8000 components; length about 14//. Separated-grains, rounded; size 0.5 to 1.5yu. Harpachne schimperi Hochst. {Graniinaceve.) Dry endosperin. — Separated-grains size 1.7 to 8//; the smaller ones rounded-angular, the larger ones polyhedral, and usually with a central cavity which is rounded-angular, or provided with single radial fissures. The compound grains are indistinctly observed within the cells, and only a few are found free. Ectrosia leporina R. Br. {Graminacece.) Dry endosperm. — Separated-grains polyhedral; size 2 to 10//; the larger ones have an angular cavity, also occasionally with several radial fissures. The compound grains are only indistinctly observed within the cells; more or less polyhedral; size about 24//. A few free forms are also found. Elytrophorus articulalus Beauv. (Graminacece.) Dry endosperm. — Compound grains round or oval, bj' means of pressure they become more or less polyhedral, consisting of over 300 compo- nents; size aljout 25 to 30//. Separated-grains polyhedi-al, with a small roinid or a large ang-ular cavity, occasionally with several radial fissures; size 2 to 10//. Festuca fascicularis Lura.; Diplachne fascicularis Beauv. (Graminacece.) Dry endosperm. — Compound grains, rounded-oval, elliptical, oblong, conical; most of them by reason of pressure become angular or even polyhedral; reticulated or almost homogeneous. Size about 18 to 22//. Separated-grains usually acute-polyhedral; the larger ones with a central cavity. Size 2 to 8 and 9//. The grains above described came from seeds of cultivated specimens; in other seeds (of North American plants) the sizes of the compound gi-ains were aliout 30 to 40/i and of the separated-grains 2.5 to 7//. Festuca Jlavescens Bellard; Festuca varia Haenk; Festuca var. jlavescens. (Graminacece.) Dry endo- sperm.— Separated-grains, size 2 to 8//; the smaller ones rounded-angular, larger ones acute polyhedral, and frequently have a cavity. No free compound grains are observed. The rather thick-walled endosperm cells separate easily; they are fiUetl with polyhedral part- grains which frequently lie in groups. Festuca arundinaeea Schreb.; Festuca elatior Linn. (Graminacece.) Dry endosperm. — Compound grains, spherical or oval, reticulated, consisting of 2 to over 1000 components. Size about 35/1. Separated-grains, size 2 to 8/i; the smaller ones rounded-angular, the larger ones acute- polyhedral; simple grains few, almost round or oval; size about 9//. Festuca diversifolia Balansa. (Graminacece.) Dry endosperm. — Compound grains, round or oval, reticulated; size about 20//. Separated-grains polyhedral, the larger ones sometimes hollow; size 2 to 7/1. Festuca ehdior Linn.; Festuca prcdcnsis Iluds. (Graminacece.) Dry endosperm. — Cornpountl grains almost round or elliptical, reticulated-granular or with distinct ilividing lines, consisting of 0 to over 2000 components. Size aljout 21//. Separated-grains, size 1.5 to 7//; the larger ones polyhedral with a central cavity. In other specimens Niigeli found the size of compound grains to be 28//, and of the separated-grains 1 to 4 and 5;/. TYPE If). GRAINS COMPOUND, MANY COMPONENTS. 283 Fcduca sylvatica Vill.; Fcstuca calamaria. Smith. {Graminacca'.) Dry aidosperm. — Compound grains spherical or oviil, reticulated. Size about 25;u. Separated-grains, size 1.5 to 7/i; the smaller ones rounded-angular, the larger ones acute polyhedral, and often have a cavitj\ Festuca spadicea Linn. {Graminaceo'.) Dry endosperm. — Separated-grains rounded-angular or frequently acute polyhedral, and with a central cavity. Size 2 to 6 and In. Compound grains few. Festuca speclabilis Jan. (Graminacecr.) Dry endosperm. — Compound grains rounded to oblong, reticulated-granular, or with distinct lines of division. Size aliout 21ju. Separated-grains, size 1.5 to 5 and 7n; the larger ones acute polyhedral, with a (central cavity. Festuca heterophylla Haenkts. {Graminaceo;.) Dry endosperm. — Compound grains round or oval, reticulated-granular or homogeneous. Size about 25|i. Separated-grains mostly polyhech-al; size 2 to Qix. Festuca elegans Boiss. {Graminaceo:.) Dry endosperm.- — Compound grains rounded or oval, reticu- lated-granular. Size about 17^. Separated-grains roimded-angular, or mostly acute-poty- hedral; size 2 to 6^. Festuca fenas Lagasc. {Graminacece.) Dry endosperm. — Compound grains spherical or oval, reticu- lated-granular. Size about ZO/x. Separated-grains rounded-angular or mostly acute-poly- hedral. Size 2 to 6m. Festuca procumbens Kunth. {Graminacece.) Dry e«dosp«-?M.— Compound grains spherical or oval, reticulated-granular. Size Z&ix. Separated-grains rounded angular or the majority poly- hedral; with a central cavity. Festuca salzmamii Boiss. {Graminaceo.) Dry endosperm. — Compound grains round or oval, fre- quently a result of pressure becoming somewhat angular, reticulated. Size about 30/j. Sep- arated-gTains size 2 to G/u; the smaller ones rounded-angular, the larger ones polyhedral. Festuca pitmila Vill. {Graminacece.) Dry endosperm. — Compound grains round or oval, reticulated (most of them have fallen apart). Size about IS/j. Separated-grains 2 to 5 and 6^; the smaller ones rounded-angular, the larger ones polyhedral, and occasionally have a cavity. Festuca petrcea Guthn. {Graminacece.) Dry endosperm. — Compound grains roimd, oval, elliptical, frequently angular as result of pressure, reticulated. Separated-grains rounded-angular or polyhedral, sometimes have a cavit}^ Size 2 to 5 and 6//. Festuca lachenalii Spema.; Festuca poa Kunth. {Graminacece.) Dry endosperm. — Compound gi-ains roimded or oval, reticulated-granular. Size 36ai. Separated-grains rounded-angular or polyhedi-al, usually with a central cavity. Size 1.5 to 6ju. Festuca nutans Willd. {Graminacece.) Dry endosperm. — Separated-grains rounded-angular or poly- hedral. Size 2 to Sju. Very few compound grains observed free, while in the cells they are distinctly perceived. Festuca urvilleana Stend. {Graminacece.) Dry endosperm. — Compound grains roimded, elliptical, elongated-oval, conical, finely reticulated. Size about 35/i. Separated-grains rounded-angular or polyhetkal. Size 1.5 to 5/^. Festuca corealis Mert. et Koch. {Graminacece.) Dry endosperm. — Compound grains spherical or oval; reticulated, reticulated-granular or homogeneous. Size about 28/:. Separated-grains rounded-angular to acute polyhedral. Size 1.5 to 5ju. Festuca gigantea Vill. ; Brcmius giganleus Linn. {Graminacece.) Dry endosperm. — Compound grains spherical or oval, reticulated-granular or finely granular. Size about 24 to 30/i. Separated- grains rounded-angular to acute polyhedral; many are hollow. Size 1.5 to 5yu. Festuca rubra Linn. {Graminacece.) Dry endosperm. — Compound grains round, reticulated-granular. Size about 20//. Separated-grains rounded-angular or polyhedral; the larger ones are hollow. Size 1.5 to 5/1. Festuca pseudo-esJcia Boiss. {Grnminaccce.) Dry endosperm. — Compound grains roimd or oval, reticulate (most of them fallen apart). Size about 20/^. Separated-grains, size 1.5 to 5/i; the smaller ones rounded-angidar, larger ones polyhedral. Festuca vaginata W.ald. Kit. {Graminacece.) Dry endosperm. — Compound grains round or oval, occasionally somewhat irregular, lujmogeneous, or reticulated-granular. Size about 25/i. Separated-grains rounded angular to acute polyhedral. Size 1.5 to 5/x. 284 DIFFERENTIATION AND SPECIFICITY OF STARCHES. Festuca stuartina Stcud. {Grami nacecc .) Dry endosperm. — Compound grains round, oval, pj'riform, frequently somewhat angular as result of jjressure, reticulated. Size about 15//. Separatod- grains, size 1.5 to 5ai; the smaller ones rounded-angular, the larger ones polyhetlral, and have a cavity. Simple grains are also observed which are rounded, oval, or pyriform; the larger ones hollow. Size 6 to 7ju. Festuca hromoides Linn. {GraminaceeB.) Dry endosperm. — Compound grains round to elongated- elliptical and conical, granular or reticulated-granular. Size about 28/li. Separated-grains, size 1 .5 to 5/i ; the smaller ones rounded, the larger ones polyhedral, frequently with a central cavity. Festuca alopecurus Schousb. {Graminacece.) Dry endosperm. — Compound grains almost round to elongated-oval antl conical, occasionally by means of pressure, somewhat angailar, reticulated. Size about 17/i. Separated-gi'ains rounded-angular or polyhedral. Size 1.7 to 5ju. Festuca mariiima DC. {Graminacece.) Dry endosperm. — Compound grains round, pyriform, ellip- tical oblong, occasionally somewhat angular as result of pressure, reticulated. Size about 40/i. Separated-grains rounded-angular to polyhedral. Size 1.5 to 5/i. Festuca dumetorum Linn; Festuca rubra var. Linn. {Graminacece.) Dry endosperm. — Compound grains usually oval or elliptical, frequently somewhat angular, granular or reticulated- gi-anular. Size about 31yu. Separated-grains, size 1 to 4 and 5^; the smaller ones rounded, the larger ones polyhedral with a central cavity. Festuca abyssinica Hochst. {Graminacece.) Dry endosperm. — Compound grains rounded, oval, elliptical, conical, frequently somewhat angular as result of pressure. Size about 17yu. Sep- arated-grains rounded-angular or polyhedral; size 1.5 to 4ju. Simple grains (or separated- grains which as result of continued growth have become rounded off) ; rounded or oval are also present. Size about 5yu. Festuca unighimis Soland. ; Vtdpia memhranacea Link. {Graminacece.) Dry endosperm. — Compound grains rounded or oval; reticulated-granular. Size about 23/i. Separated-gi-ains round to polyhedral, the larger ones with a small cavity. Size 1 to 4/i. Festuca nigrescens Lam.; Festuca helerophylla var. Haenke. {Graminacew.) Dry endosperm. — Com- pound gi-ains spherical or oval, homogeneous or gi-anular. Size about 16ju. Separated-grains rounded to polyhedral ; size 1 to ifj.. Festuca glauca Lam. ; Festuca ovina var. {Graminacece.) Dry endosperm. — Compound grains spherical, oval, or conical; half to as thick as long, homogeneous or granular, consisting of 20 to over 8000 components. Size about 18;U, rarely 25^. Separated-grains rounded to polyhech-al. Size 1 to 3 and 4/i. Festuca trijlora Desf. {Graminacece.) Dry endosperm. — Compound grains round or oval, reticulated or reticulated-granular. Size about 17yu. Separated-grains, size 1 to 3 and 4//; the smaller ones rounded, the larger ones polyhedral. Festuca myurus Limi. {Graminacece.) Dry endosperm. — Compound grains rounded, oval, or ellip- tical, angular through pressure, reticulated or granular. Size about 15 to 20,u. Separated- graiiis, size 1 to 3.5/u; the small ones rounded, the larger ones polyhech-al. Festuca broteri Boiss. and Reut. {Graminacece.) Dry endosperm. — Compound grains rounded, oval, elliptical, conical, finely reticulated. Size about 14yu. Separated-grains, size 1 to 3jL(, rounded- angular. Festuca cynusuroides Desf. {Graminacece.) Dry endosperm. — Compound grains round or elliptical, granular or almost homogeneous. Size about 18;u. Separated-grains rounded angular. Size 0.7 to 3m. Festuca tenella Willd. {Graminacece.) Dry endosperm. — Compound grains rounded-oval to elongated- elliptical, granular. Size about 26//. Separated-grains rounded-angular; size 1 to 3/i. Festuca sabulicola Dufour {Graminacece.) Dry endosperm. — Compound grains rounded to elongated- elliptical, reticulated-granular. Size about 20^. Separated-grains rounded-angular; size 1 to 3(U. Festuca lolium Balansa. {Graminacece.) Dry endosperm. — Compound grains round or oval, reticu- lated-granular or granular. Size a])out 17ju. Separated-grains roinided or rounded-angular. Size 0.7 to 3^. Festuca rutlbcelloides Kunth. {Graminacece.) Dry endosperm. — Compound grains rounded, oval, ellijitical, occasionally somewhat angular by means of pressure, finely reticulated or almost homogeneous. Size about 10^. Separated-grains rounded-angular; size 1 to 3/i. TYPE 16. GRAINS COMPOUND, MANY COMPONENTS. 285 Fcstuca rigida Kuuth.; Sclcrockloa ri(jida Panz. {Graminacece.) Dry endosperm. — Compound grains round to oblong usually somewhat irregular, half to almost as thick as long, granular or almost homogeneous, consisting of over 2000 components. Size aliout 27/u. Scparated- grains almost roimd or rounded-angular; size 1 to 3/u. Festuca diraricata Desf. {Gravrrnacem.) Dry endospervi. — Compound grains almost round or ellip- tical, granular. Size about 21(U. Soparated-grains rounded-angiilar; size 0.7 to 2 and 3^. Festuca tenuiflora Schrad. {Graminacece.) Dry endosperm. — Compound grains almost round or oval, frequently somewhat irregular, homogeneous or finely granular. Size about 20/^. Separated-grains rounded-angular; size 1 to 2, rarely 3^. Festuca delicatida Lagasc. ; Vulpia dclicaluln Link. {Graminacem.) Dry endosperm. — Coznpound grains rounded-oval to oblong, occasionally somewhat irregular, two-fifths to almost as broad as long, granvilar or almost homogeneous, containing over 3000 components. Size about 16 to 22ji:. Separated-graius round or rounded angular. Size 0.7 to 2.bii. Festuca ciliata Link. {Graminacece.) Dry endosperm. — Compound grains rounded, elliptical, oblong, conical, frequently angular by means of pressure, finely reticulate or almost homogeneous. Size about 16.«. Only a few separated-grains rounded or rounded-angular are present. Size 0.7 to 2.5m. Festuca geniculata Willd. {Graminacca:.) Dry endosperm. — Compound grains round or oval, fre- quently angular, also gi-anular. Size about 18//. Separated-gi-ains rounded, size 0.8 to 2/i. Festuca memphitica Boiss. {Graminacece.) Dry endosperm. — Compound grains spherical, oval, elliptical, homogeneous or finely granulated. Size about 15//. Separated-grains rounded, size 0.8 to 2/1. Festuca mocrophylla Hochst. {Graminacece.) Dry endosperm.— ComY,o\i\\A gi-ains round to oblong, frequently angular as a result of pressure, reticulated-granular or granular. Separated- grains rounded. Size 0.7 to 2^. Bromus littoralis Hort. {Graminacew.) Dry endosperm.— Corn-pound grains spherical or oval, reticu- lated-granular, finely granular or homogeneous. Size about 17//. Separated-grains rounded- angular or acute polyhedral, size 1.5 to 5//. This starch is related to the genus Festuca. Diarrhena americana Beauv. {Graminacem.) Dry endosperm. — Compound grains spherical or oval, reticulated. Size about 24/t. Separated-grains polyhedral with acute angles and margins; larger ones have a cavity. Size 2 to 6/:. Loliurn canadense Michx. {Graminacew.) Dry endosperm.— Compound grains rounded or oval, half to as thick as long; granular, consisting of 20 to over 4000 components. Length 40 to 45/t; thickness about 30//. Separated-grains rounded-angular to acute polyhedral. Size 1.5 to 6//. Loliurn temulentum Liim., var. speciosum Link. {Graminacece.) Dry endosperm.— Compound grains spherical, oval, or somewhat irregular, homogeneous, or reticulated-granular. Size about 42//. Separated-gi-ains polyhedral; the larger ones have a small cavity. Size 1.5 to 6/i. Psilurus nardoides Trin. {Graminacew.) Dry e7idosper7n.— Compound grains rounded to oblong, frequently irregularly angular, half to almost as thick as long, granular to homogeneous, consisting of over 2000 components. Size about 15//. Separated-grains round or rounded- angular. Size 0.5 to 4//. Lepiurus incurvalus Trin.; Ophiurus incurvatus Beauv. {Graminacew.) Dry endosperm. — Compound grains rounded-oval to oblong (not at all angular), two-thirds to almost as thick as long; slightly granular or almost homogeneous, consisting of 20 to 8000 components. Length about 33//; thickness about 18//. Separated-grains round. Size 1 to 2//. Lepturus filiformis Trin.; Ophiurus fdiformis Roem and Schult. {Graminacew.) Dry endosperm. — Compound grains rounded-oval to oblong, usually more or less angular by means of pressure, homogeneous, rarely finely granular. Size about 16//. Separated-grains round. Size 0.5 to 2/i. Centrolepis fascicidaris Labil.; Desvauxia biUardieri R. Br. {Centrolepidew.) Dry endosperm. — Compound grains spherical or oval, consisting of 2 to over 16 equal components. Size about 18//. Separated-grains semi-spherical with one or more pressure facets, or polyhedral wth one curved surface, or exactly polyhecbal (only bounded by pressure facets), usually with a large or small cavity and several radial fissures. Aphelia cyperoides R. Br. {Centrolepidew.) Dry endosperm.— Separated-grains angular or poly- hedral, more or less irregular, usually compressed. Size 2 to 9//. Compoimd grains could not be positively distinguished. 286 DIFFERENTIATION AND SPECIFICITY OF STARCHES. Reslio fcrruginosus Link. (Restiacecc.) Dry endosperm .■ — Compound grains spherical to elongated-oval, many somewhat angular as a result of pressure, reticulate, containing about 1000 components. Size about 30^. Separated-grains polyhedral; size 2 to 7//; the larger ones frequently have a small cavity and several delicate radial fissures. Anarthria prolifera R. Br. (Restiacece.) Dry endosperm.- — Separated-grains polyhedral, sometimes with a very small cavity and delicate radial fissures. Size 1.5 to 7^. Very few free comjiound grains are observed, though they are frequentlj' distinctly perceived in the cells; almost round or oval, by means of pressure more or less polyhedi-al; size about 20/^. The whole cell lumen sometimes appears to be cuneiform filled with separated-grains. Pcepalanthus caulescens Kunth. {Eriocaulacere.) Dry endosperm.- — Compound grains almost round to elongated-oval, frequently polyhedral, reticulated-granular, or almost homogeneous, con- sisting of about 1000 components. Size 15 to 22/u. Separated-grains polyhedral, the larger ones with a central cavity. Size 1.5 to 8/:. Pmpalanthus frigidus Mart. (Eriocaulacece.) Dry endosperm. — Starch as in the preceding species. Compound grains containing over 1000 components. Size about 25 to 30//. Separated- grains, size 2 to lOyu. Xyris operculata Labill. (Xyridacece.) Dry endosperm. — Compound grains spherical or oval, usually more or less polyhedral by reason of pressure, reticulate, containing over 300 components. Size about 30/i. Separated-grains polyhedral; the larger ones sometimes have a small central cavity, and also single delicate radial fissures. Size 2.5 to 10 and 12/i. Xyris semifuscata. (Xyridacece.) Dry endosperm. — Starch as in the preceding. Compound grains spherical to elongated-conical; containing 400 or more components. Size about 35yu. Size of the separated-grains 2 to 8 and lOfi. Mayaai vandellii Schott and Endl. {Xyridacece.) Dry endosperm. — Separated-grains, rounded-angular to acute polyhedral ; many have a very small central cavity and single delicate radial fissures. Size 2 to 8 and lOju. The endosperm cells are entirely filled with these polyhedral grains; only indistinct groups may be recognized within these cells, which probably correspond to compound grains. Mayaca michauxii Schott and Endl. (Xyridacece.) Dry endosperm. — Starch as in the preceding. Sepa- rated-grains polyhedral, usually with a small or large cavity and radial fissures. Size 2.5 to I2fj,. Arum orientale Biebrst. (Aroidece.) Dry endosperm. — Compound grains rounded or oval, three- fifths to as thick as long, consisting of 4 to 1400 components. Length about 36 and 48/i, thickness al)Out ZQix. Separated-gi-ains polyhedral, usually with acute margins, edges and angles; the larger ones have a central cavity and a few short radial fissures. Size 2 to Wfi. Zanledeschia (cthiopica Spreng. ; Kichardia wthiopica Kunth. (Aroidecr.) Dry endosperm. — Compound grains spherical to elongated-oval and conical, consisting of 2 to over 400 components. Size about 28m. Separated-grains polyhedral, frequently with a small cavity. Size 2 to 8 and 12yii. Simple isolated rounded or elongatcd-oval grains are also ])resent. Typha tenuifolia Humb. Bonp. (Typhacew.) Dry endosperm. — Compound grains roimiled to oblong- oval, frequently somewhat angular or irregular, reticulated-granular. Size about 15^. Sep- arated-grains more or less polyhedral. Size 2 to 5yu. Poor in starch. Piper nigrum Linn. (Piperacece.) Dry endosperm. — Compound grains spherical or oval, frequently polyhedral as result of pressure, reticulated-granular, containing over 4000 components. Size about 33^. Separated-grains rounded to polyhedral. Size 1 to 4/i. Piper cubeba Linn. fil. (Piperacece.) Dry endosperm. — Compound grains spherical or oval, fre- quently by means of pressure polyhedral, reticulate, or with distinct lines of separation, containing over 600 components (most of them have fallen apart). Size about 32/i. Sepa- rated-grains polyhedral, rarely with a small central cavity. Size 3 to lOju. The endosperm cells of Piper nigrum and cubeba are closely packed with starch. When making sections, a few compound grains fall out along with numerous separated-grains. The comjjound grains are a little larger and not so crowded in the innermost less compact tissues of the seed; they are packed in the more external dense tissue, the divisions between the compound grains frequently being indistinctly observed, while in the outermost cells the compound grains consisting of coalesced components form a contin\ious imiform mass having a reticu- lated or parenchymatous appearance (with dense septa and apparently hollow alveoli) as in Amomum, Coinmelina, etc. (See type 12.) TYPE l(i. CHAINS COMPOUND, MANY COMPONENTS. 287 Patotnorplw sidwfolia Miq.; Hekeria sidcefolia Kunth. (Pipcracece.) Dry endosperm. — Conipouiul grains spherical or oval, occasionally by moans of pressure somewhat angular, homogeneous or distinctly granular. Size about 24/i. Separate d-gi-ains rounded and rounded-angular. Size 1 to .3ju. In other seeds nearly all the comjiound grains have fallen apart. Size of the rounded and iiolyhedral separated-grains L.'S to .5/i. Peperomia luaculosa Hook. (Pipcracece.) Dry endosperm. — Compound grains spherical or oval, occasionally somewhat angular as result of pressure, almost homogeneous to distinctly gi-anular. Size about 22^1. Separated-grains rounded or rounded-angular. Size 0.7 to .3^. Airiplex hortcnsis Linn. (Chenopodiaccw.) Dry endosperm. — Compound grains rounded or oval, half to almost as thick as long, finely granidar or almost homogeneous, containing over 15,000 components. Length about 45ju; thickness about 2.5^. Separated-grains rounded. Size 0.5 to l.Sju. The compound grains within the endosperm cells are frequently embedded in a finely granulated mass of separated-grains. Airiplex hastatn Linn., var. ealotheca Rafn. (Chenopodiacecp.) Dry endosperm. — The endosperm cells are filled vnih round separated-grains. Size 0.5 to Ijk. Axyris amarantoides Linn. [Chenopodiacea .) Dry endosperm. — The endosperm cells are filled with a finely granular starchj^ mass of large separated-grains. Size 0.5 to l/n. Acnida iuberculata Moq. {Chenopodiacece.} Dry endo.sperrn. — The endosperm cells are packed with roimd or rounded-angular separated-grains sometimes hanging together in short rows. Size 1 to 3fi. Spinacia glabra Mill.; Spinacia inermis Moench. (Chenopodiaeece.) Dry endosperm. — Compound grains rounded-oval, elliptical, conical, and lanceolate, one-fifth to almost as thick as long, sometimes polyhedral by means of pressure, also granular, rarely almost homogeneous, consisting of 20 to 30,000 components. Length about 60 and lOO^u, thickness about 20 to 41(1. Separated-grains rounded. Size 0.5 to 2.5yu. Panderia pilosa Fisch. and Mey. (Chenopodiacece.) Dry endosperm. — Compound grains roimded to oblong, occasionally somewhat angular, finely gi-anulated or homogeneous. Size about 20 to 24^. Separated-grains, romid. Size 0.5 to l.oix. Blitum capitaturn Linn. {Chenopodiacece.) Dry endosperm. — Separated-grains round, size 0.7 to 2/i, rarely 3^. Very few compound grains are present. Atnbrina grarcolcns Moq. (Chenopodiacea:.) Dry endosperm. — Compound grains siiherical or oval, occasionally somewhat angular, granular. Size about 23 to 30ai. Separated-grains rounded. Size 0.7 to 2.5//. Beta oricnlalis HejTi. (Chenopodiacea.) Dry endosperm. — Compoimd grains roimded, elliptical, oblong. Size about 30/x. Separated-grains rounded or somewhat angular. Size 1 to 3/i. Beta vulgaris lAwn. (Chenopodiacece.) Dry endosperm. — According to Payen (Ann. So. Nat., 1838, II, p. 28, plate 4), the separated-grains are rounded. Size about 4fi. Echinopsilon hyssopifolium Moq. (Chenopodiaccw.) Dry endosperm. — Separated-grains rounded. Size 1 to 2yu. No compound grains are observed. Kochia scoparia Schrad. (Chenopodiacece.) Dry endosperm. — Separated-grains rounded. Size 0.5 to 1/i. No compound grains are present. Cyclolepis platyphylla Moq. (Chenopodiacea.) Dry endosperm. — Separated-grains rounded. Size 0.5 to about Ifi. No compoimd grains are observed. Teloxys aristaia Moq.; Chenopodium arisiatum Limi. (Chenopodiacece.) Dry endosperm. — The endo- sperm cells are entirely filled with almost rounded or rounded-angular separated-grains, frequently lying in rows. Size 1 to 2 and 3^. Chenopodium ([idnoa Willd. (Chenopodiacece.) Dry endosperm. — Compound grains rounded, oval, oblong, conical, one-third to almost as thick as long, not often slightly angular bj- means of pressure; sometimes pointed at one end, more rarely at both ends; finely granular or homo- geneous; consisting of over 14,000 components. Length about 40/t, rarely 54//; thickness about 23//, rarely 28/i. Separated-grains round. Size 0.7 to 2/i. Acroglochin chenopodioides Schrad.; Acroglochin persicarioides Moq.; Lecanocarpus cauliflorus Nees. (Chenopodiacece.) Dry endosperm. — Compoimd grains rounded to elongated-oval, sometimes slightly angular, granulated or reticulated-granular. Size about 36//. Separated-grains rounded; the largest ones rounded-angular. Size 0.8 to rarely 3/*. 288 DIFFERENTIATION AND SPECIFICITY OF STARCHES. Hablitzin tamnoidcs Biebrst. (Chcnopodiaccw.) Dry e7idospcnn. — Compound grains spherical to elongated-lanceolate, frequently somewhat angular; one-third to as thick as long, granular. Size about 30/i. Separated-grains rounded. Size 0.7 to scarcely 3/1. Basella alba Linn.; Basella ramosa Jacq. {Chenopodiacew.) Dry endosperm. — Compound grains rounded or oval. Size about 2.5yu. Separated-grains round to polyhedral. Size 2 to 6/i. Starch found only in the endosperm. Schoberia cornicidata Aleyer. (Chenopodiacece.) Dry endosperm. — Compound grains spherical to lanceolate, one-seventh to as thick as long; with blunt or pointed ends; frequently by means of pressure more or less polyhedral, granular or homogeneous, consisting of over 10,000 com- ponents. Length about 50yu, thickness about 14/n. Separated-grains rounded. Size 0.7 to 2ju. Schoberia .so/.s/i Meyer. (Chenopodiacew.) Dry endosperyn. — Separated-grains round. Size 1 to 2^. Monolepis chenopodioides Moq. {Chenopodiacew.) Dry endosperm. — Compound grains round, oval, oblong, two-fifths to almost as thick as long, rarely somewhat angular as result of pressure, finely granulated or homogeneous, consisting of over 12,000 components. Length about 38 to 45^1, thickness about 25yLi. Separated-grains rounded. Size 0.5 to 1.5 and 2ix. Corispermum hyssopifoUum Linn. (Chenopodiacece.) Dry endospenn. — Compound grains rounded or oval, sometimes oblong, two-fifths to almost as thick as long, rarely somewhat angular by means of pressure, granular or almost homogeneous, consisting of over 9000 components. Length about 25 to 31/i, thickness about 18^. Separated-grains almost round. Size 0.5 to 1.5 and 2ju. Corispermum marshallii Steven. {Chenopodiacew.) Dry endosperm. — Compound grains spherical or oval, granular, or almost homogeneous. Size about 2Qjx. Separated-grains rounded. Size 0.7 to 1.5m. Achyranthcs argenlea Lam. {A7narantacew.) Dry endosperm. — According to Leon Soubeiran (Journ. Pharm., 1854, xxv, 9G), the endos]')erm cells, which are easily sejDarated from one another, contain rounded or oval grains with one dark central spot (probably separated-grains). Size 5/1. Achyranthes fruiicosa Lam. {Amarantacem.) Dry endosperm. — According to Leon Soubeiran (Journ. Pharm., 1854, xxv, 96), starch as in the preceding, except that the grains have fissures radiating from the center. Li the embryo, single rounded, rather large starch-grains may also be present. Albcrsia bliinm Kunth; Amaranlhus blitum Linn. {Amaraniacew.) Dry endosperm. — The endosperm colls arc filled with roimd(xl separated-grains. Size 1 to 2/i. Alternanthcra j)aronychioidcs St. Hil. {Amarantaccw.) Dry endosperm. — Compoimcl grains roimded or oval, occasionally somewhat angular, granular or homogeneous. Size about 27/t. Sepa- rated-grains rounded. Size 0.5 to 2/i. Amaranlus bullatus Besser.; A^narantus sanguineus Linn.; Ainaranhis frujnentaccus Roxb. {Amaran- tacew.) Dry endosperm. — The cells, which easily separated from one another, are filled with separated-gi'ains. Size scarcely l/x. No compound grains were observed. Amblyogyne polygonoides Rafn. {Amaraniacew.) Dry endosperm. — Cells closely filled with sepa- rated-grains, frequently arranged in rows. Size scarcely l/i. No compound grains were observed. Celosia crislala Linn. {Amaraniacew.) Dry endosperm. — Compound grains spherical or oval, finely granular. Size 16/i. Separated-gi-ains rounded or roimded-angidar. Size 1 to 2.5/:. Chamissoa albida Mart. {Amaraniacew.) Dry endosperm. — Compound grains mostly spherical, rarely oval, consisting of from 6 to over 200 almost equal components. Size 17/i. Sepa- rated-grains nearly round to almost polyhedral, rarely with a small cavity. Size 1 to 4/i. Desmochwla patula Roem & Schult. ; Achyranthus paiula Linn. fil. {Amaraniacew.) Dry endosperm. — Compound grains spherical, oval, oblong (not at all angular), half to as thick as long, finely granular, consisting of over 6000 components. Size about 25/i. Separated-grains roimd. Size 0.7 to 1.5ai. Euxolus emarginalus A. Br. and Bouch6. {Amaraniacew.) Dry endosperm. — Separated-grains rounded or rounded-angular. Size 1 to 3/j. Compound grains no longer present. Euxolus caudalus Moq. {Amaraniacew.) Dry endosperm. — According to Leon Soubeiran (Journ. Pharm., 1854, xxv, 99), the starch-grains in the cells which are easily separated are oval or spherical, without lamellse or hilum. Size 5/i. (Probably homogeneous compound grains.) TYPE IG. GRAINS COMPOUND, MANY COMPONENTS. 289 Hoplothcca floridana Nutt.; Frmlichia Jloridana Moq. (Amarantacece.) Dry endosperm. — Compound grains s]iliorical, oval, oblonp; (nut at all angular) ; two-fifths to as thick as long, finely granular. Length about 2;^; thicknoss about lOyu. Soi)aratcd-grains rounded. Size 0.7 to l.Hfi. Iloplotheca texana A. Bramn.; Frcelichia gracilis Moq. (Ainarantacew.) Dry endosperm. — Size of poparated-grains hardly Iju. No compound grains are present. Gomphrena deawibens Jacq. {A?})ara7>tacecc.) Dry endosperm. — Size of separated-grains hardly over 1^. No compoimd grains are present. Iresine nervosa Hort. (Amarantacece.) Dry endosperm. — Separated-grains round. Size 0.5 to l/x. No compound gi-ains are present. Polycnemvm majus A. Braun. [Amarantacece.) Dry endosperm. — Separated-grains entirely filling the cells, arranged in rows. Size 1 to 1.5ju. No compound grains are visible. Pupalia prostrata Mart. (A7nara7itacece.) Dry endosperm. — Compound grains spherical to elongated- oval, sometimes angular, two-fifths to as thick as long, finely granular or homogeneous, consisting of 25,000 components. Size 30/^, rarely 40^. Separated-grains round. Size 0.5 to scarcely over l/n. Scleropus amarantoides Schrad. (Amarantacece.) Dry endosperm. — Separated-grains round or rounded- angular, closely packed in the cells, and frequently arranged in rows. Size 1 to 3yit. Teleianthera polygonoides Moq., var. braehiata. (Amarantacece.) Dry endosperm. — Compoimd grains spherical or oval, finely granulated, size about 30/^. Separated-grains round, size 0.5 to about 2n. Ahronia arenaria Horh. (Nyctaginacece.) Dry endosperm. — Compound grains rounded to elongated- oval (completely filling the cells) ; granulated or reticulated-granular, consisting of over 4000 components. Size about 20 and 25^^. Separated-grains rounded 1 to 2/i. Mirabilis jalappa Liim. (Nijctaginacece.) Dry endosperm. — Separated-gi-ains rounded or slightly angiilar, completely filling the cells. Size 0.7 to 2ju. Compound grains are not visible. Mirabilis longiflora Linn. Dry endosperm. — Separated-grains rounded. Size 1 to 2;u. No com- pound grains are present. Oxybaphiis cervantesii Lagasc. (Nyctaginaceoe.) Dry endosperm. — Compound grains spherical, granular, consisting of 30 to over 1,000 equal components. Size about 15,u. Separated- grains round or rounded-angular. Size 1 to 3/i. Allionia ovata Pm-sh. ; Calxyhymenia panieulata Desf . (Nyctaginacece.) Dry endosperm. — Separated- grains rounded. Size 0.7 to 1.5;u. Few compound grains remain (partly in the cells, partly free), rounded or oval. Size about SBfx, finely granulated, containing over 24,000 compo- nents; most of the cells are packed merely with tiny separated-grains arranged in rows. Allionia nyctaginea Michx. (Nyctaginacece.) Fresh and dry endosperm. — Separated-grains almost round to rounded-angular. Size 1 to 2 and 2.5^. They completely fill the endosperm cells. No compoimd grains could be distinctly observed. Allionia incarnata Lirm. (Nyctaginaeeoe.) Dry endosperm. — Separated-grains round. Size 0.7 to about 2// ; cells are crowded with them ; no compound grains occur. Bougainmllca spectabilis Willd. (Nyctaginacca:.) Dry endosperm. — Separated-grains rounded or somewhat angular. Size 0.7 to 2jx. No compound grains are present. Avicennia tomentosa Lirm. (Verbenacece.) Dry cotyledons.- — Compound grains mostly rounded, rarely oval or conical, frequently somewhat irregular and angular, consisting of 2 to over 20 components which are almost equal, rarely unequal. Size about 14 and IS/x. The large separated-grains have a small central cavity. Size 2 to 8/i. Starch plentiful. Nymphoca rubra Roxb. (Nyjnphceaeece.) Dry endosperm. — Compound grains by means of pressure polyhedral, with rather sharp margins and angles, reticulate, consisting of 8 to over 200 com- ponents. Size about 23;u. Separated-grains round, rounded angular to acute polyhedral; tlie larger ones with a central cavity. Size 2 to 10^. Nymphma dentata Th. et Schum. (Nymphceacew.) Dry endosperm. — Starch as in the preceding species. Nymphcea coerulea Savign. (Nymphoeacece.) Dry endosperm. — Starch as in Nymphcea rubra. Sepa- rated-grains mostly polyhedral, without a cavity. Size 2 to 7ix. Nuphar lutcum Smith. (Nymphceacece.) Dry endosperm. — Compound grains spherical or rounded- oval, slightly angular as result of pressiu-e, finely granulated, consisting of over 10,000 com- ponents. Size 25 to 30/i. Separated-grains rounded or rounded-angular. Size hardly 1 to dfi. 19 290 DIFFERENTIATION AND SPECIFICITY OF STARCHES. Barchnja nblongn Wall. {Ny7nphceaceoc.) Dry endosperm. — Compound grains rounded or oval, usually polyhedral as a result of pressure, reticulate, consisting of 20 to over 800 compo- nents. Size about 26ju. Separated-grains 2 to 8/i in size; the smaller ones rounded-angular, the larger ones polyhedral and with a central cavity. Frmikenia jmlvendenta Andr. {Frankeniacece.) Dry seeds. — Separated-grains almost rounded, frequently clinging to each other like flakes of snow. Size 0.7 to 2/i. No compound grains could be positively distinguished. The starch is probably found in the endosperm. Mesembryanthemum ■pinnatifidmn Linn. fil. {Ficoidece or Mcsembryacew.) Dry endosperm. — Com- pom^d grains sjjherical to oblong, one-third to as thick as long, frequently angular or even polyhedral, as result of pressure, granular or homogeneous, containing over 5000 compo- nents. Length about 36;u; thickness about 20/x. Separated-gi'ains rounded-angular. Size 1 to 3m. Tetragonia expansa Ait. (Ficoidece or Mesemhryacece.) Dry endosperm. — Compound grains oval, elliptical, or conical, one-fourth to three-fom-ths as thick as long; polyhedral as result of pressure, reticulated-granular or almost homogeneous, consisting of over 4000 components. Length about 60m, thickness about 27^. Separated-grains rounded-angular to polyhedral. Size 1.5 to 5fi. Also isolated simple spherical grains, size about 12^, are observed, as well as compound grains of few components, which increase in size as the grains are less compound (in doublets about lO/i). Aizoon hispanicum Linn. (Ficoidece or Mesemhryacece) Dry endosper^n. — Separated-grains rounded to polyhedral. Size 1 to 3 and 4^. Completely filling the endosperm cells, in which groups of polyhedral grains which resemble compound grains may occasionally be seen. Size about 16^. Aizoon canaricnsc Linn. (Ficoidece or Mesemhryacece.) Dry endosperm. — Separated-grains round or rounded-angular. Size 0.5 to hardly 2^. Completely filling the endosperm cells. No com- pound grains were observed. Trianthema vumogynum Linn. (Ficoidece or Mesemhryacece.) Dry endosperm. — Compound grains, rounded. Size 10 to 12^. Separated-grains rounded to polyhedral, the larger ones have a central cavity. Size 1 to 5ai. Portulaca megalantha Steud.; Portvlaca grandiflora Hook. (Portulacacece.) Dry endosperm. — Sep- arated-grains rounded-angular to acute-polyhedral. Size 1.5 to 6/^. A few small compound grains and broken fragments of larger ones are also present. Talimmi patens Willd. (Porhdacacea'.) Dry endosperm. — Compound grains roimded or oval, two- thirds to as thick as long, reticulated-granular, consisting of 4 to over 600 components. Size about ISju. Separated-grains rounded-angular to almost jiolj'hedral. Size 1.5 to 4/i. Calandrim'a co7nprc.' I I 1 1 1 1 p-| — I 1 1 1 — u. 1 1— _j ^ > > ss D O 522 ou;^ QOO a: w 3 2 a KE K a a E g z u^ ^ a. K a. t~ X SEEESESE gp 2 £ £ ii 2 2 u O as*/ ESSE £225 u b q w 1. z a g n t- a. 0- ESS £22 ciciri < Z 5 <; J a 5 S : -- ::j 3 : li- a < t < < < J -J _l -1 - fr- pi p z s s s „HX 35 eg? tZQZ Z.XZ 33D3DDOO :ct:n oooooo O O c j^; «; ^ ^ 5 : a a S: e: a a & a a 6 a a s| Ess i^ ESS ; sS SSSSSSd gggg SaS ■ J O u Q. u C < < < 3 o o o £ tf h: tf ^gi So g < < < a a Q o o o (i: a: o: 33 3 :^:i 2 as PS « Kg a £ £ o ^ tofcoitt B ■< 66o66655 SSl^SSSdd 33333333 =^3 o o aatiHuaaa aaQQOPQQ THE DEGREES OF POLARIZATION OF VARIOUS STARCHES. 315 < PL, 3 p 1 CO o I > ~~| I ___j__ U. 1 1 r-i 1 1 J 1 -I > ] 1 b^SS Uj w It! w i^ t &a.a.a.a.a,gg SSS §3 . S y < « hi tti a CCS S i ^ ^3 (A (/I D P) ^^ u u K K P P 23 2gl as a D w E U tj -J >i ^ o U (-> (A o O H u ; o Oh I < X O s 5 3 < 2JIS S o o tA 3 SEE Ess < (U o EEE i s :^ ::! s 2 £ S2? 32 5 5 o o si, r u :< IS SES S ^^ ... m eacSB: a a: a x a. a. - g^5E2233or: •Jo 3< < P f- o 353 o o S5 > > 33 O O II 99 a: a: o o >! 2 < 3 oa Q, u u. ui t/i ui in PP P E3 O O O O Q Q 5 Q ^ ^ .J ^ o u o o ijl ■i EC d > > < < < 3 S < < < 3 ai o O O O 13 H tJ 2 E «;«-•• SS " £ ? Ill 33333: SSS oooooo „„ ;<< 223232 22 e< O E EC < i l.- E * >• J. « 2 ioE S i«E S . 9 w < 33 3 O O 316 DIFFERENTIATION AND SPECIFICITY OF STARCHES. Q > TTTI n I I Q I — 1 1 1 1 — . . , . . 1-| — ,,,,.. I , r-ri — n 1^. J _1 I 11 > -I- r " _ L L B to 09 3 8 T3 O ►-I g I PQ X O 5S > > < b X s w a a fi. CK a. H X < I/) .|S^ ^ ^ a £ a. 5S£ ;6: BSSESEEEg =5 = S ESE 533 rid: 3S£ o o o u u u o o o U O U < s ^ '-' £ J 2 g E D a. ^ -J to c/; in 1/1 VI 1/) w D » t> O D 3 H H H h- ^- (- □: a: Li: cf o: a O O O O O O B £ X X £ X u u u o u u o o o o o o o o u u u u 52 5; iS^gg adeS953o2S -Ht-hHHHHHHt- < 5oo: o o u a. w u <_» DQQ QQQCQaQQ OOO OOOOOOOO 3; < < < - 3 :ii .-53 Q _^ > ~ ~~ " Q 1— I p-| 1— rn r — n r u. 1 1 1 1— I _i — — . 1 . ^ _i _ _ _ I 3 3 1 a Q : a; a ) (J u i3S . •-*■ . ■^ to S w S « S t: o ^ mmooOOHo EcsExsxxx u o < < ; z z, : < < < A 3] n .r. Z^^ THE lOUINE REACTIONS OF VARIOUS STARCHES. 317 ca Ph o CO S c> o 05 o o Q . , I I , > I I I I Q 1 1 TTTTTTl 1 1 1 1 1 1 1 n U. 1 r--| 1 — I J 1 1 1 1 r _1 I I > ■ ~r fcg S U .OH. d 6 ■it* fc Lu ,.; !« 5 c < < < K K E w w u < < < E £ E a -i <■<■<< < < < < U>> > > > > fr-t-H EEEEEKE 5S5 U V w . 5 S *£ Ln LJ < < IS 3 14 bl CSS P D 5 ^ ^ P ^e 5 s a H Pi all is 9 8 S3 0. D I as L U M S E E 5 g S Ei CIS 3 Q > II 0 — 1 J h- u. C-3 _ _ _ __ _ _ — — _ - -- _ - _ - -- _ — - j - gs .5 5 u. _; .. < K > u 2 00 w IS EEE •J O IP 3 iiii ^s >^ s 3s i2 EE < o 531 ^ m < 00 ££ 00 5 gS « « o 33:j33»3 -'•^^S^S ;■ E o o o E3 r 5 S m m e n 2^2 X XX ^ ~) CO <* ,» U U U U (J u c a oL X a u u- 1 :; £ E Z E E U K K a oe oe ci: a m 5 . ». - E < ; A ^ < r^ c d X E U 2§ E E O O > > &. P. o S2 w 2 " ° S E «« 1 - a o ■ a < J I o. u u. Cj 3 3 J 0000 c o 5 5 - O != ■^ £ x2E E E E 000 < < < > > 2 t- H H = = ^ !j O O a t- <-» O O O u H o K IX c£ u O a£ O (J o w o. u E E E E E E 000000 K o: a a « a a 2 £ 2 2 2 £ E S l- t- COCO > > < < O (-» ss ai Ik bb o j2 S BOB §5 O D 3 U U U 000 a cB " u u u a fj 318 DIFFERENTIATION AND SPECIFICITY OF STARCHES. a < Q > ~" D ll. r |-pT 1 1 1 1 1 1— PI _1 J o e aa n sig a? 5H3g ->->■< X X X o Q !^ K ■ «t 5 o- -5 ,->. uj QQQOaQQQ 00000000 ixfKKcsixasuix 11 2 iiS o n > Q u. _j . 1 > 1 1 1 1 1 5i i 2 E E E E E 2 K K a: 000 000 000 a. a. a. 000 a oi 01 ^ < O Q Q w u ^ K S 35 - = §1 E E D 3 aa 4 i e i it Si. SaSS! s o a < I o^ PS o E- < 5 : o 3 1 H S C < Pb O H H tf) * D 3 ;ss £ E:: o 00 3 3 3 3 5 :s :s ii y y ^ y S X a H 3 uj 6 5 D D S ~J ^ ; s s t e£ d SSS5SS >>>>>> S E E E S E eepi 25 || So > > E S a o Eg o o o o < < e EE & & & & o. cu tf B E ie 53 S E bI 0: K > -* •* 7) Ul o o d 6 c o ■ 3 S a. a: H to to tf) « 10 p o o D D u u u o o 0: tf el cS b: o-o o- c o S u t s ESS 15 3 3 33 X X (J o 3 3 3 3 3 3 X X X X X a tj O (J O O tj <<<■<'<< 2; K K E K ! a a i THE GENTIAN VIOLET REACTIONS OF VARIOUS STARCHES. 319 S e v> 5 S B Ah Q > Q 1 PJ-, U. 1 _j — _I as •pa e •2 s i d K K o a , > Q _ 1 u. 1 1 1 [— 1 1 1 1 — r-i 1 1 I I I [ _j 1 — I 1 TTl 1 II II 1 1 1 n J ^_^ 1 1 l-I-l !-■ >r^TT ' r TT I7i._irrr._i ii iii ii •J S z « OS O oo|pi|-|B - m £ O am _-,-^h- DD3 D3:^C'p:3 P. Q W < 5 Se o > ^ > > > - BE aiiii mill III 53c > > o o >■ a£ < _i ;; SSSS a ^ O O O O - H o o o o 3 Sep z; K £ o o o is Sri j < < <;< 223223 aa s 5S5250 o u o ta H ■-> O O O o H O tt « C£ M O tf U t-f U M o> o Z. z: K Z Z. K o o o o o o §5 u a. 9S" EC O E Sog ^ K a: o o Si Sii !» u^ tf K O O p D O 5^ s ^ tt tf cd O ^ U U U K U 320 DIFFERENTIATION AND SPECIFICITY OF STARCHES. 60 ?! s Q > Q U. 1 J I n 1 1 I CO i , Sbog isd sssss BEEEB ddriridcidciddciri ssssss D D g g g g H O- ^ ESS g p D d5d 3 H 55 ooooooooo SS OOOOOOOOO . 5 D a 5; e > . "E|ao gSBH kEuSE QOZo 88S88 SBS3 3^ 2^5 O U B SSI g g 0000 O Q c 00c n Q a 500 S S2 £ 0. P H >1 J fti I S 1 1 .± III , , . 1 1 1 iUhiUi ^S3 i= = S2-l?' = « ?>* (GOL (WH TA(N TA(C ATA ( ATA( RATA RATA RATA SEE ;5S53££335 P !=» B EVER EVER IHDU INDD INDE INDE SACC SACC SACC 000 a a: a 000 ?SSSSSS?5 S S S £g£ ES£E£S£SE 000 uiuuuuuwuu sas > > n si 3 >>>>> E E E S E E g g g P p p : 0-3 ■ H to !/) K < g s i a s : o u 0 ( » 5 > ; 3S a, 0. U K »■ w S £ BBSS OS •; Is S E D O Eg < ec a: z •>« 3 ■< > > ggg: — ""SSiS mmougS — o 06066633 5555 0. 0. o> 0. a a eeees o ©• o- o- o- w z s 50^ a u C SEE D D 3 a o; M < ■< < S< 3 U. U. U. ffc t XX U. U. U. U. L THE SAFRANIN REACTIONS OF VARIOUS STARCHES. 321 e S s CO s 1^ a ^ , "* I — I I Q TTTn 1 [1. 1 1 1 1 — I— -I — I 1 1 1 r-| r r _i ^ 1 . 1 1 r-~, 1 p-| 1 _i , , , , r-, > n X ^5 i W M S «E < 3<° ^32! ^5 a I I 1 I < D 3 J 2 O O in 3 £E£ 9Ea Id^ K a e o. 2 Si sa o s W U bl a. a c afesi c: x a lg Si g3 5; 11 = 3 3s> = lis « t> g oe tt o > > og < <<> iliililigiiiiislll' liiyl Ll fa] UJ say agSS§§3 O a. ea -^-^-.t- crt t/i «i w w tn (« D O O S Op ~^ ig fa. &. 11? i<^ In KEZEEEEBl S > J U U 3 O O E -4 = s o P u u u « u 21 322 DIFFEKBNTIATION AND SPECIFICITY OF STARCHES. o e> I Oh o b CD in h h- > D D ^o o s tJ o ^ as E D •< < -J z: 3 5 Q w K a; ™ - n — t! r* SSs as 2; 3S B5 S SESSEP £5BB2c Q£ Of o o oooooooc : p p p ooooooooo otnOQDuooa< 333 CC SSEE in If) P p D 3 ^3 << <:^ Z g O O O O TT OO O O KM a a I X ouoouuuvu Hh-e-i- I ^ 3 p 3 ^ D s : j J o 6 o ■>•>->•>- i- tf CC DC D-. D m in M Hi y W W S S U o o o < HrfuzEuSE u o -t s o o ;E3 ESS u u o •< ■< < ^8 U EEEEEEEE ■p. o 9 Q £ Q O K o w a < < a H ij -J a a Q Q O O OO ct a a oi. w te a ^ ^ H ! tj Q o [- < "< [ 3oS^»hSo -lUO(jQ.'/)<_'0 QDQQQQQa OOOOOOOO 2gi o " CD U) s o p 1 o m IB b 5i U) b - _j h si ass? 2 < o> O bl ,A V^ Si S3 !2 _r i o s X S s u (J £ 35a S S t-* .lis !j OS d: tt < < < < §3 a. S3 95 ss 0 y w 0000 23 3 9 0000 £ £ z a 83 = 3S|5P2 z e z Z Z-K •e -< -e (-• o u : a: a£ a: tt c ■< -t < < . > > > > t <: < -< -c z z z z z z - -; a z z z g 5 5 K Z t < < < E J 22 E l/i 0 "* « z 0 z 0 Sj32 X ssSI 1: < < ■< ^- f- H H Z Z Z K < < -a; << U UJ Ul M a a: ffi a z < ■a:t:<<<-i: _. WHJ w w u w 00 W W W w K Z Z Z 0000 23S2 < < -r < 5353 EESSES 35 ZEES ilsE H ^z ^ £ z z z ;^ 5555 ^ a o VI * " £ H «! f a C G S < < u S2 b ^3 Z Z 22 K O 2 5 6? J z o c H u u a 0 b CO e . s , 'n \- - - ^ - - (0 <0 1 ■ - -r in b J J in J _ ^ _ J _ 2 o o to 3 SEE H O 2 3 liil sa sss 3 3 » o£ a: u a a. a. ill is ^3 K Z 22 ^^ I Si* o o 3 - ssag 5 9 > > 2 ./ a CO „ o p. m 2Sg t> a. < o ^2 ET oi i^ o . -« -J ■< > 3 co-Old Z Z K N o o o < !5d !55 32 O (J > > 23: !SS □: a: z z u w 3 u o ^ ce a; (K a: cK c <-<■<-«-< ■ z E z. z z : C en X K » f < < < < :; K Z Z Z SSgaggggggaga 2<; ^5 gg p ri 00 ^„,„ tf) o ii maou. ^ 5 :i:S S333 <2 Sq 0000 o t 5§ = Id III 333333 9S9 SSoggg o u 22 aa Si aa SdS O H S5 N a o z Sag u oe z H U tf) -J 06 55 po3 00. 3 VI . ^ ^ » w < SS ">> -J -J lA fl f> 00 L> D tJ a a o uo 000 q: « a: 55 324 DIFFERENTIATION AND SPECIFICITY OF STARCHES. PS ■< P-, CO 3 o s o a; 2 in o 10 o LI o in n o 10 CM o CI g in 1 1 1 j_ 1 ,1 _ E o ii O. H X m BE : SSS3SS5S m =? z < r: a. ce. a. t£. ti. IX lA a. ac OOOOOOOOO uuooouuuu OOCOOOOOO <:<;^.<-<<<<^ uuuuuuuuu 3S-S ||Sd s - H H Ht- H K D Z H SEES D O P D ^^^^ COCO o o o o a K s a; K K K E o: b: a: ix OO O O 5 -i 2 P o '-' -■-^« w in a E - u. E E w * a - u- w 2 2 t 3 Q Q < O U O ssssss B52BSS E E*E E E E O O O c o o o: DC a: a: K a: tc a a a E a a: K tt K u o M M tJ W M b sit !i| O CO « Hi ;e3: ?; D O S ;se : o o 5 U U !P9 "Sr!-< -^SS-iieKSo ££k i !3S S3S 3 Q O ; o o O ££ESEEES I ^ .J u o aasassass Q QQaQQQQQ O OOOOOOOO s 2 i° 3 -^ " E E S 33, o O THE CHLORAL HYDRATE-IODINE REACTIONS OF VARIOUS STARCHES. 325 •pa s u o "C a a^ o pt, o s (0 n (D in in o in in 5 - CO n -■ CM CM - - . . . . . J _ _ _ _ _ _ _ _ _ _ 1 1 Eg S i sss -I 3 D 3 H S D < U U fcoS > « < SEE o s a a, X a. z3s a< lis 3 is S S2 si ^s 2 3 3 "> K W HH >»! ,/,«, 35 sa £2 ii aa ss ss _J CO < 3 iiis BES g b B y « «'«!« 5 4'< ^ w - S o o 3 = ^ 113 S PC O OEgS 3°ot > o-oc ;* E z z i O O C s5 ts =S;5 c ex. a u < < < < S K E B EBS SZEEZ „ > > > < H :; s g a a a s a H S s r, if\ m i/t i/i ui VI tn in i/i zgaggggaggs -j p o >. 2 < > > E52 £ D u^ X ^ E ^5 sa 3333 3 3 o o o o 333 SEE o oo < < < > > 2 I- 1- 1- i2 = '^ u o !-> o e t-" O O O u *- o cc Qc c: U O K u u o y; 0. u 333333 K K a E K C OO o o o o > > < < PI . §9iS is 8§ U lit. < < SS22^22 ^S 326 DIFFERENTIATION AND SPECIFICITY OF STARCHES. s pS "8 a < o i o t- IlLii^i 1 1 1 1 h 1 1 1 1 1 il 1 h lllll II 1 Ih , 1 1 1 1 1 1 _ D 5 -Jl □ E 5 K iuQ e CO - >^ J 9 ^ >• E Sri ESESEEEEKS SriSriSririridri ads z a; £ ^3a 3 ssSgg OOOOOOOOO aaaouuofi id 2222Saa3 OOOOOOOO uuuuuuuuu a ti. 0, o, 0. 0. H H H B D B c 3 0; 6. C 0.0.0^0.0.0,0.0. O. 0. 3SS o o o O Q Q O O O K ?, K o o o t) u E S E S I/) If I s p p b 33 <<^^ E £ O O O O - S o o o o w M a: K KX XX (- H H H lA ub 2 E ^ ^ V ^ O Q < O ssas HB22 E 3 a K o o o o OS IX CS CC X X X X ad z o 11 m « M Sid £ = s 5| ^ ^ m t^ ^ ^u)< kEoEE qOeuj oo K t^ Q "^ "- is i R d t: o ti " ■^ ^ O, K E - -3 o S < ■< < 3 S H a I 3 o o I 3 Pg L> O. 1. ■« < • Q Q I O O C 32 PL, W omoogogg K a E E £ . THE CHROMIC ACID REACTIONS OF VARIOUS STARCHES. 327 e S 1 09 s I I I . . ■ . I . . I 1 I I cuo. a SEE 9gg£2 ISg§ o§8g3 '^m E£g^£ 5555 EiS O (0 g a gg 3 3 o u 9 S D g y o e «; Zl = = - a £ S S ? a. £ E S H o S < < g u u O X O 2 •" o> o f- in , 1 ,,, 1 1 1 _ ± 1 _ 1 _ J .a < < < > aOoiooo< !-■)-• OOOO K K a K a K o o o o o o ?;:< 222222 kS ■ S . (^ n o: 2 Og « D §^^: o a K 5 ui -J a "^ t » p D ^ J o o o So. K a a; o >. u u u K u 328 DIFFERENTIATION AND SPECIFICITY OF STARCHES. Eh I O I 05 2 3 < w o 2: t q J t: o u < >j 2 w 0, z o « < p H 3 o -J juoua.cnuo < <<<:<<<<< aQOOOQaQQ o 00000000 a> AAmnnnmn ii§ ° 3S ^33 0 0 (0 ;« a. 0 10 ts in > S E E S P P o o SS ill In o u S 5 2 JQ 2 P •< I- I 3 t^ S £ 2 pSppp <<<< oo'oo'c' uouu r- K z z ti < < -^ < r; s 5 s lis ^ e KJ K ^ P J -' ^ O < < 3 p A. O t E K ■ ■£ P ^ ^ ** lI ?■ C6 a S s £ fc: o ^ EroSo fea « oamuu59!::o !< ffiKXffiKKKa \ < <,<.<.'r.<<<.< 9 KEKZ: = fc£K 2 3 !S2 I W U U M M ^ Uj i 3 C Q 3 & o o THE PYROGALLIC ACID REACTIONS OP VARIOUS STARCHES. 329 s e 8 p5 o 'S S (55 a. ■< K o < < — O O K 352 ss '-'S3" « < o> O w w 1« is SS |5 SiS < < :53 W Ifl 3 D Q t w E K m I- < < < E U U LA M < < < < o [J O o o < L^ ku (b tC U O o o o o E S (X Q£ s o: S £ S K B Q Q Q < C ca n ta s o o 3 o K E E E £2 O Q J u a Z K K E C t; E E E E U U U U U O (^ g|||| 2223 < < 5 > 3 5 !. l^ /-\ r\ t DC 5 Ij > ^ > S ^S S O O O < < < < ~ <<<<<< E u O e; E O E o «i ta. m -> lU u ui u E E E E O O O O ESSE .S2 C D I t EE CSS u u nJ ;« >• o o u u 1 . , 1 . 1 . 1 1 1 h II 11 1 1 ,,,, 1 ll . . 1 cc u, a. ai cfiotu:»»K O O O 5 a, a: DC U O »• u g> u ii: u 330 DIFFERENTIATION AND SPECIFICITY' OF STARCHES. « Ah s a o « Oh s > ss » D »a: a: Hhi^ sss ririii eg a 5 S S i : Sri 3 X ^ W SKSE D P P P ^353 2^ p Q. :: ,ESg .5 ssgggf m ^ Z S -1 _ DDDQPO:i>»D <<<<<<■ ^' t^ ^ 35S 3 o; a: 500 5 22 It, U (J a, o: a oL X oL ct 0000000 X » z a a X ni u o u u o o o 0000000 (J O O U Lf U tj ciaOQuoOo.^ E] & a c s n a ^ a tj ^33 2£S gals See oozG aa 0.H0 J ^^^ QQQQQQCia t^SS 00000000 Aa%£ osoamianaacqn £ Q O o ?"■ ti O < J o « 5 U O U i d t: 1^ ! w a, z 1 H -2 o . w u o : < < < 1 Q o a ) o o o - -2 "0.1- sli a ss - 0 10 0 u> 0 0 in 0 1 CM 10 0 u 1 S w S 5 THE FERRIC CHLORIDE REACTIONS OF VARIOUS STARCHES. 331 « PL, o s in o 10 o in in O 10 o 10 o n . ll 1 1 :nl 1 1 1 1 , 1 II CO o ^1 l2 5 S t is I g d b u < < 5 2 = Q c. p w £ 2 S >- W M ^ (^ 5 > Q, m ■< < < E d d •< < • ^j w 5 s ^ -5 u E e « a oi 5 " •< ■< < 5 5 > > O Q - . . . a « (J o o o o < (b uH ui ik '-* i^ u. u. u. u. _ _ O O O O S H ci a c: a 2 5 Ld w M u) x s: ^< 2 b: S5 1 bi£. O H -; u < Q eg CO « < < > > acai at'. ss a: a: -_, „ X U U cc <<<«;->S =gSililll |22SS2 W f- 2 5 1 2>S> ! <<<-(! w w w w : a: X 3 K . H- t- t- t- ' <■<<<> U O O O ( << < SEE >->■ >- 000 gs ^^i; ^ i^ w g g ^ g "* K K E S O d _ A E J -J WW g>- >• w w g K Z Z K K < < K Z. o o EE Z K U (J O t« ^ S •- ^- s S3 ^ 2 « * ^ o ^i 2i 9 ft. p w E E P •t < K .4 >J < a H 3 S ec »^ « PL, in to 0 in in 0 10 10 t 0 n in 0 (*) n M in 0 10 1 1 sSis W H H H . i ^ I^S SEE a: u: Qc O E 5 5 ^ z z 22 ■ p 3 S» is ■ ^o gi ; S w o o g2S g gg 9.^SSS2 X A Bi ta aU o PPD3S93 ODDPOP pld Sis pe«o . .<< > z z o o S E 22 O o o o E K e-H P D P D 0000 S553 < < < < S215S gag 8g K K K H ? 000 - ^ w ii KOz <2So O o t- O 0: w 3 tC U (A c. u K K Z Z 0000 Ip 22 §5 U 0. p w < & O O E fl. K X a O ^ o <^ o cc u 332 DIFFERENTIATION AND SPECIFICITi' OF STARCHES. OS (15 sa 50 o P -; g Q ;i c ^0 S CQ 10 s S -J I 2|^ 553 . u u u M (A V> 0001 o Q a 1 000; K z: z I O O O ' k2S 1 U O U I i ESES I P P D P i -i-i-i-i '0000 i O o o o s: :c s s 1 H H H H I £ S Z K > u. o: ci: oi ,0000 oi i/ 2 O P ■ z E; : O O ( a. ai I X X t bgs O K O 000 ESS eee f. K K 000 ak. a. Hi < ffl m n iii >>> H H H ; e z K : c « a: « « a sss SEP e|g <; oa ^ ^ ; hSuzEoEE Qos55y;;j£-__-- ; o:< > > p. 0. i s ^s da§dSg 35 Oi Lil w" < "^ ■ 5 ^5 ""-Jui^E b2b5 oiojaitKOca ^i«o-* >>>>>> gjgo EEEESE gs2s iggili 5* £ E! a; X X S I- t- I- H < < < • fey &y si CO u u . !j p 3 5 a z « 32? , 2S?d 5 3 5! g _g5 a: P M •7 p » A. 0. 0. A. Ih A. ESS p p p u: IX A 2 i 353S > > p- > 2 2 fea < S 22 ^ > a s a 2; : ^ ^ n -J tu H in < < 3 < 3 a"2 2 u (J u 2 2 2 000600; X X : THE PURDY'S solution REACTIONS OF VARIOUS STARCHES. 333 s o 5 s o e s o CO 5n o 5 to o (0 in 10 n L> -J 't - 1 ^ n CJ IN 1 1 1 . Ill 1 , 1 1 _ ! ^o^-j 5555555555 1 1 i c « O C u u (-> oc u 334 DIFFERENTIATION AND SPECIFICITY OF STARCHES. THE MEAN TEMPERATURES OF GELATINIZATION OF VARIOUS STARCHES. The importance that lias been attached to differences in the temperatures of gelatiniza- tion of starches, together witli the fact of the exceedingly contradictory records by different investigators, and the reliability of the method used in this research, together with the value of this means alone of differentiating starches, has suggested that a table stating the mean temperature records recorded in this investigation would be found of special con- venience and value. The Mean Temperatures of Gelalinization of Various Starches. Zea mays var. everta (Gulden Queen) 6.3.25° Zea mays var. everta (White Rice) 63.25° Zea mays var. indurata (North Dakota) 68.00° Zea mays var. indurata (Compton's Early) 68.50° Zea mays var. indentata (Early Learning) 66.50° Zea mays var. indentata (Hickory King) 66.75° Zea mays var. saccharata (Stowell's Evergreen).. 66.85° Zea mays var. saccharata (Black Mexican) 65.00° Zea mays var. saccharata (Golden Bantam) 66.50° Andropogon sorghum var. (White Kaffr Corn).. . 68.00° Andropogon sorghum var. (Yellow Branching Sorghum) 70.00° Andropogon sorghum var. (Shallu) 66.90° Panicum crus-galli var 75.25° Oryza saliva var 74.75° Triticum sativum var. vulgare 64.00° Triticum sativum var. dicoccum 62.90° Secale cereale var. (Mammoth Winter) 61.00° Secale cereale var. (Spring) 62.00° Hordcum sativum var. (Champion) 61.00° Avena saliva var. (Clydestlale) 63.00° Arrhenatherum elatius var 60.25° Vicia sativa 72.50° Vicia villosa 66.00° Vicia faba 65.00° Vicia fulgens 72.00° Vicia gerardi 71.00° Phaseolus vulgaris var. (Red Kidney Bean) 74.50° Phaseolus lunatus var. (Henderson's Bush Lima) . 79.75° Dolichos lablab 79.50° Mucuna pruriens 74.00° Lens esculenta 72.25° Lathyrus oiloratus var. shahzada 68.50° Lathyrus sylvestris 58.00° Lathyrus latifolius var. albus 63.00° Lathyrus magellanicus var. albus 69.00° Pisum sativum var. (Eugenie, Yellow) 73.60° Pisum sativmn var. (Eugenie, Green) 74.50° Pisum sativum var. (Thomas Laxton) 73.25° Pisum sativimi var. (Electric Extra Early) 70.00° Pisum sativum var. (Mammoth Gray Seeded) . . . 70.50° Pisum sativum var. (Large White Marrowfat). . . 68.50° Wistaria chinensis 71.50° Arachia hypoga?a 66.50° Polygonum fagopyrum var. (American) 63.50° Polygonum fagopyrum var. (Japanese) 64.75° Quercus alba 60.75° Quercus muhlenbergi 63.75° Quercus jirinus 64.75° Quercus rubra 63.00° Quercus texana 64.25° Castanca americana 59.25° Castanea sativa var. nunibo 60.00° Castanea sativa var 6i.00° Castanea jnimila 59.75° jEsculuR hippocastanum 69.00° Arum palffistimun 70.50° Ai-um cornutum 70.00° Arum italicum 72.00° Arisiema triphyllum 70.00° Dracunculus vulgaris 64.25° Richardia clliotiana 76.00° Richardia africana 75.00° I Richardia albo-maculata 77 I Dieffenbachia scguine var. nobilis (Pith) 69 j Dieffenbacliia seguine var. nobilis (Cortex) 69 ; Dieffenbachia seguine var. maculata (Pith) 70 > Dieffenbachia seguine var. maculata (Cortex) 70 i Dieffenbachia seguine var. irrorata (Pith) 69 Dieffenbachia seguine var. irrorata (Cortex) 68. Dieffenbachia illustris (Pith) 69 Dieffenbachia illustris (Cortex) 70, Lilium candidum 61 lilium longitlorum var. giganteum 62 Lilium longiflorum var. eximium 65 Lilium parryi 52 Lilium rubellum 63 Lilium philadelphicum 63 Lilium tigrinum var. splcndens 62 Lilium henryi 56 Lilium auratum 58 Lilium speciosum var. album 64 Lilium martagon 59 Lihum superbum 61 Lilium tenuifolium 57 Lilium pardalinum 63 Lilium puberulum 62 Fritillaria nieleagi'is 59 Fritillaria p3'renaica 63 Fritillaria pudica 49, Fritillaria aurea 02, Fritillaria armena 66, Fritillaria imperialis var. aurora 68 Fritillaria liliacea 55 Fritillaria rccurva 58 Calochortus albus 53 Calochortus mawcanus var. major 59 Calochortus bcnfliami 01 Calochortus lilacinus 59 Calochortus nitidus 54 Calochortus howcllii 50 Calochortus leichtlinii 63 Calochortus luteus var. oculalus 58, Calochortus splendens 55, Tulipa hageri 56, Tulipa sylvestris 57 Tuhpa greigi 50 Tulipa billietiana 55 Tulipa didieri 50 Tulipa didieri var. mauriana 50 Tulipa didieri var. fransoniana 54 Tulipa clusiana 57 Tulipa clusiana var. persica 54 Tulipa oculuB-solis 53 Tulipa praecox 55 Tulipa australis 51 Scilla sibirica 66 Scilla peruviana 65 Scilla bifolia 64 Chionodo^ia lucilia; 60 Chionodoxa tmolusi 62 Chionodoxa sardensis 61 Puschkinia scilloides 56 Puschkinia scilloides var. libanotioa 56 Ornithogalum nutans 61 Ornithogalum umbellatum 56 50° 00° 00° 50° 60° .00° .75° 60° .00° .65° .55° .60° ,60° .96° 90° 05° .00° 60° .35° .10° 60° 30° 30° .15° 60° .15° .60° .75° .60° .40° 10° .70° 00° 50° 50° .00° 76° .50° 00° 10° ,10° .90° 15° 30° .25° ,75° .95° .85° 10° 75° 75° 80° 16° .00° .50° .76° .00° .35° .10° .35° 70° .25° ,76° MEAN TEMPERATURES OF GELATINIZATION OF VARIOUS STARCHES. 335 The Mean Temperatures of Gelalinization of Various Starches. — Continued. OriiitIiug:iIiini niirboncnsc (pyrainidalc) 55.50° OrniUioRalum thyrsoidea var. aureum 06.00° Erythroiiinin deiis-cania 52.05° Krydironiuin dcns-canis var. grandiflorum 54.00° Erythroniuiu americanum 53.45° Eryllironium grandiflorum 53.00° Eiythrouium citrinuin 50.45° Erythrouium californieum 57.90° llyacinthus orientalis var. alba suporbissima 67.00° Ilyacinthus orientalis var. albulus (Wlrte) 09.00° Hyadnthus orientalis var. albulus (Italian) 69.50° Galtonia candicans 58.60° Rluscari botryoides 71.00° Mu.scari paradoxum 73.00° Muscari micranthum 69.75° Muscari conicum 73.00° Muscari conimutatum 72.00° Muscari racemosum 74.75° Muscari compactum 70.00° Muscari coniosum 68.00° Brodiica pcduncularis 08.50° Brodia^a Lxioides var. eplendens 61.75° Brodia;a Candida 02.00° Brodiaja lactca 04.50° Brodiipa la.xa 61.00° Brodiii'a coccinea 70.00° Brodiira grandiflora 65.90° Briidia'a californica 67.30° Brodia'a purdyi 07.00° Brodiica stellaris 08.10° Bi'odia;a capitata 05.57° Brodia;a congesta 09.25° Triteleia uniflora 09.45° Lachcnalia pendula 00.65° Lachenalia tricolor var. luteola 73.10° Convallaria inajalis 01.00° Trillium grandiflorum 68.00° Trillium ovatum 50.75° Trillium sessile var. californieum 55.00° Colchicum parkinsoni 01.25° Amaryllis belladonna major 74.00° Hippeastrum vittatum 72.00° Ilippeastrum cquestre 72.50° Hippeastrum aulicum var. robustum 71.75° Vallota purpurea 74.25° Crinum fimbriatulum 70.00° Crinum americanum 77.00° Zepliyranthe.s Candida 73.00° Zephyrantlies rosea 70.00° Sprckelia formosissiraa 77.00° Ha;manthus katherina; 80.00° Hymenocallis undulata 71.50° Hymenocallis calathina 68.50° Leucoium vernum 71.50° Leucoium a;stivum 72.50° Galanthus nivalis 06.50° Galanthus elwesii 70.00° Alstrcemeria ligtu 58.50° Alstro-meria brasiliensis 57.75° Alstrctmeria aurantiaca (aurea) 54.75° Stcrnbergia lutea 75.50° Narcissus horsficldii 74.00° Narcissus maximus 75.25° Narcissus bulbocodium 72.00° Narcissus bulbocodium var. conspicua 72.75° Narcissus bulbocodium var. monophyllus 74.00° Narcissus incomparabills 76.25° Narcissus odorus 74.50° Narci.ssus poeticus 73.25° Narcissus biflorus 75.00° Narcissus jonquilla 76.00° Narcissus jonquilla rugulosus 75.25° Narcissus jonquilla campernelli rugulosus 76.50° Narcissus tazetta var. orientalis 77.00° Tacca pinnatifida 63.00° Iris florentina 71.50° Iris pallida speciiisa 74.50° Iris pumila var. cyanca 73.25° Iris bismarckiana 72.00° Iris iberica 73.10° Iris xiphiuni var. grand tresorier 65.50° Iris xiphium var. wilhdinine 06.50° Iris .xipliium var. lusitanica 67.00° Iris tingiana 65.50° Iris reticulata 65.25° Iris histrio 65.50° Iris alata 66.50° Iris caucasica 04.50° Mora;a tristis 70.00° Homcria coUina 72.50° Tigridia pavonia var. grandiflora alba 01.00° Tigridia jiavonia var. conchiflora 60.00° Gladiolus byzantinus 78.50° Gladiolus primulinus 80.50° Gladiolus cardinalis (Blushing Bride) 70.50° Gladiolus floribundus 81.00° Watsonia humilia 75.50° Watsonia iridifolia var. o'brieni 74.50° Watsonia meriana 70.25° Tritonia crocata 82.50° Tritonia crocata var. lilacina 81.75° Tritonia crocata var. rosea 81.50° Tritonia securigcra 81 .00° Tritonia pottsii 72.50° Tritonia crocosmaflora 73.00° Freesia refracta var. allia 76.75° Freesia refracta var. leichtlinii 77.25° Antholyza croeosmoides 74.00° Antholyza paniculata 72.00° Crocus susianus (Cloth-of-Gold) 70.00° Crocus versicolor (Cloth-of-.Silver) 75.00° Crocus var. (Baron von Brunovv) 74.00° Romulea rosea var. speciosa 73.50° Cepylla herberti 55.00° Marica gracilis 70.00° Gelasine azurea 73.50° Sparaxis grandiflora alba 71.50° Sparaxis var. (Albertine) 72.50° Ixia speciosa 84.00° Ixia viridiflora 83.00° Ixia var. (Emma) 84.50° Babiana var. (Violacea) 80.00° Babiaua var. (Anthraction) 82.00° Musa cavendishii 07.75° Musa cavendishii (green fruit) 07.50° Musa sapientum 08.00° Musa ensete 03.00° Zingiber oflicinalc 73.85° Zingiber ofticinale var. Jamaica, No. 1 82.25° Zingiber oflicinale var. Jamaica, No. 2 85.90° Zingiber oflicinale var. Cochin 90.00° Hedychium coronariura 72.30° Hedychium gardnerianum 71.15° Curcuma longa 82.50° Curcuma petiolata 82.50° Canna warszewiczii 09.00° Canna roscoeana 09.50° Canna musa?folia 70.50° Canna edulis 67.50° Canna var. (Konigen Charlotte) 65.25° Canna var. (Pnesident Carnot) 65.50° Canna var. (L. E. Bally) 09.00° Canna var. (Mrs. Kate Grey) 02. .50° Canna var. (.Jean Tissot) 70.00° Canna var. (J. D. Eiscle) 62.50° Maranta arundinacea 77.00° Maranta arundinacea var. No. 1 72.50° Maranta arundinacea var. No. 2 72.00° I Maranta massangeana 69.00° I Maranta leuconeura 74.00° 336 DIFFERENTIATION AND SPECIFICITY OF STARCHES. The Mean Temperaliires of Gelatinization of Various Starches. — Continued. Maranta musaica 88.50° ' Acta!a spicata var. rubra 56.50° Calathea lietzei 78.50° I Cimicifuga racemosa 59.50° Calathoa vittata 83.25° I Eranthis hyemalis 51.50° Calathea wiotiana 85.50° ; Ranunculus bulbosus 56.00° Calathea vandenhcckei 77.75° Ranunculus ficaria 64.00° Stromanthe sanguinea 82.75° ' Adonis amurensis 64.50° Nymphffia alba 69.00° Cochlearia armoracia 62.50° Nymphsa marliacea var. albida 68.00° ! Jatropha curcaa 59.50° Nymphaea marUacca var. carnea 66.50° Maniiot utilissima 62.00" Nympha;a gladstoniana 70.00° I Cyclamen repanduni 55.50° Nymphaea odorata 67.50° Cyclamen coum 56.50° Nymphaea odorata var. rosea 67.00° ' Solanum tuberosum 66.00° Nelumbo nucifera 58.75° i Batatas edulis 74.00° Nclumbo lutea 60.00° \ Gesneria tubiflora 64.50° Anemone apennina 53.50° Gloxinia var 69.25° Anemone fulgcns 60.75* i Trianosperma ficifolia 59.00° Anemone blanda 63.25° ! Cycas revoluta 73.50° Anemone japonica 63.00° ! Cycas circinalis 72.00° Aconitum napellus 52.75° | Dioon edule 72.50° Actaea alba 56.00° Zamia integrifolia 76.50° Especial attention may be directed to the figures of the foregoing table for compari- son with those recorded by other investigators, and also to certain features regarding the range of temperature of gelatinization of the starches as a whole. The records of Lipp- mann (page 175), from the time of their publication, over 50 years ago, have had a very wide publication and generally ha^'c been accepted as standards of reference, but the methods used by him and succeeding workers are crude and liable to lead to very fallacious results. Since Lippmann's studies, little has been published on the temperatures of gelat- inization, notwithstanding that much literature has accumulated that treats of the properties of starch, and the inherent importance of the meanings of differences of such temperatures in relation to the starches of different plant sources. This seeming neglect is doubtless to be accounted for in the very variable results that must have been obtained by various experimenters who have felt that their records were unworthy of publication. Whymper (page 175), for instance, found with barley, maize, rye, potato, rice, wheat, and tapioca starches that the values differed in most cases from previous figures and varied with the maturity and size of the grains. Such discrepancies are illustrated in the case of rice starch, the temperatures by Lippmann, Lintner, and Dafert being 61.25°, 80°, and 73°, respectively; of barley, Lippmann recording 62.5°, and Lintner 80°; of corn starch, Lippmann giving 62.5° and Lintner 75°; and of wheat starch, Lippmann recording 67.5° and Lintner 75° to 80°. In the preceding table the temperatures given as determined by the method used in this research are for the same starches, rice 74.75°, barley 61°, corn 63.25° to 68.5°, and wheat 62.9° to 64° — the last two varying according to the variety. The lowest temperatures recorded were among the Liliacece (among members of Tulipece — Calochortus, Tulipa, and Erythronium) in which in six instances the figures ranged between 49.60° and 50.95°; while the other extreme, where the temperature exceeded 85°, was noted among the Zingiberacew (Zingiber) and Marantacecc {Maranta and Calathea), the highest being 90° — a range of over 40°, which is certainly remarkable. Of the total number of starches studied, 0.3 per cent had a temperature of gelatinization in the forties, 20 per cent in the fifties, 39 per cent in the sixties, 34 per cent in the seventies, and some- what more than 6 per cent from eighty to ninety. Nearly 75 per cent had a temperature of gelatinization between 60° and 80°, and about 65 per cent between 60° and 75°. CHAPTER VIIL GENERAL APPLICATIONS OF THE RESULTS OF THIS RESEARCH. SPECIFICITY AND CONSTANCY OF THE STEREOCHEMIC CHARACTERS OF STARCHES IN RELATION TO GENER.\ AND SPECIES. Reference has so frequently been made to the relations of the reactions in both speci- ficity and constancy to genera and species that it would be almost a matter of superero- gation to do more at this point than to lay down the dictum that as starch-producing plants become modified in their botanical characters the properties of the complex syn- thetized substances, such as starch, are accordingly modified, and hence that every genus and species has a form of starch which is specific and constant in relation to that genus or species. Evidence in support of this is found throughout the work. In the investigations on the crystallography of hemoglobins, referred to in the Preface, it was found that these substances exhibit differences which are in specific relationship to genera and species; and, moreover, that the results of a few preliminary studies with plant proteins and certain other groups of corresponding vital substances justified the announcement of the belief that the remarkable zoological distinctions shown by the hemoglobins would be found to be presented by other complex organic metabolites. Furthermore, it v/as made clear that not only does there exist these specific stereochemic differences, but also that there are definite gradations of these differences which correspond with the positions occupied by animals in relation to one another as is recognized by the systematic zoologist. Upon such a basis, one may assume upon logical grounds that corresponding gradations will be demonstrated to exist between starches and plants. It follows, as a corollary, that in the existence of individualities of corresponding stereoisomers in relation to genus and species we ha-\-e a strictly scientific basis for the classification of plants and animals. APPLICATIONS OF THE RESULTS OF THIS RESEARCH TO PHARMACOGNOSTICS, COMMERCE, AND TECHNICAL PURSUITS. The grains of the starches of vegetable di-ugs vary considerably in size, shape, and other characteristics, as is instanced in the classification of starches by Kraemer (page 71). It is well knowm that differences merely in the histological characters of starches are often of value in suggesting or practically determining the plant source, but it is also clear, as has been frequently pointed out, that the histological method may be absolutely misleading. If, therefore, there be coupled with this method such others as have been employed in this research, especially in the examination of storage starches (as those of seeds, tubers, rhizomes, corms, etc.), the identification and the determination of substitution and adul- teration of drugs that contain starch are usually rendered quite easy. Commercial starches probably number about thirty, including among them particularly those of corn, rice, and white potato, together with the relatively many kinds that are marketed solely or chiefly as "ar^o^\Toot." The first three starches are, as is well known, used extensively in various technical industries and in the arts, and as articles of food, while the arrowToots are sold almost entu-ely as articles of diet, especially for the use of the sick. The recognition of corn, rice, and white potato starches, respectively, and the detection of adulterations, substitutions, and gross impurities is a matter of common 22 337 338 DIFFERENTIATION AND SPECIFICITY OF STARCHES. experience and usually of simple microscopic jirocediire. Witli tlie arrowToots one has not to deal with a single kind of starch but with a number of kinds from many different botanical sources. The commercial term "arrowroot" long since lost its original signifi- cance in so far as it applies to the derivation of the starch from Maranta arundinacea or its varieties and forms or closely allied species. The supposed especial dietetic value of this form of starch reported by Hughes in 1751 naturally brought arrowroot into prominence as a peculiarly valuable form of food, since which time a number of so-called, or false, arrow- roots obtained from a most diverse variety of plants have been introduced into trade. In another part of this report (Part II, The Starches of Marantacew) it is pointed out that the arrowroots of commerce are derived from species, varieties, and agricultural forms which represent twelve families, seven orders, and three classes. It will be observed from the results of this research that the recognition of the source of the arrowroot, true or false, by means of the methods employed is a simple matter. In many technical pursuits the employment of starch enters, to a more or less important degree, into various operations, particularly those of certain textile industries. Especial reference was made on pages 176 and 177 to the differences in the characters of starch-pastes and pseudo-solutions obtained from starches of different kinds, and to the notable varia- tions in penetrability and stiffening strength exhibited by boiled starches from different plant sources when prepared under the same conditions. From the records of this investi- gation it is manifest that these properties, which are of such importance in the manu- facture of many fabrics and in laundering, are variable to marked degrees in different starches, so that a given starch may be entirely satisfactory for a certain purpose whereas another may be very poor or almost if not entirely useless. Applications of the results of this investigation to such conditions are strikingly apparent. For instance, if a starch from a given species be found somewhat deficient in stiffening strength, such deficiency may be absent from the starch of another species, variety, form, or hybrid of the same genus. Moreover, as the starches of the parents are represented in the hybrid in modified form it would seem obvious that desirable properties are obtainable by intelligent experi- mentation along the lines carried out by the plant-breeder. Inasmuch, however, as a form of starch that is lacking in certain properties that may be essential in certain textile industries or in the arts may likely be rendered into a suitable state by simple and inexpen- sive procedures, as instanced in the method of Bellmas (page 176), the kind of starch will in all probability be found to be of insignificance in comparison with the cost of the raw material. Nevertheless, the principles underlying the natural protluction of special forms of starch by cultivation, selection, in-breeding, cross-breeding, and hybridization have a broad applicabilitj^ in connection with many plant products which have a wide economic value, especially as regards medicinal substances, notwithstanding the very important production of the latter synthetically during very recent years. APPLICATIONS OF PRINCIPLES TO PHARMACODYNAMICS. The cultivation of medicinal plants, which as yet has been carried on to but a very limited extent, has shown that in the case of certain members of Ritbiacecc and Enjthroxy- laccw cultivation has increased the yield of medicinal substances, whereas in other plants, especially those of Solanacece, the change from the wild to the cultivated state has had the opposite effect. The author is not familiar with any instance where an attempt has been made to qualitatively modify medicinal substances by any of the methods in use by the plant-breeder, but there are many known cases where certain toxic properties common to a genus are modified both cjuantitatively and qualitati\-ely in the difTerent members of the genus. To what extent the qualitative variations are to be attributed to different proportions of the active constituents or to stereoisomeric modifications is as yet wholly speculative. It would seem, however, inasmuch as the various gross characteristics of APPLICATIONS OF PRINCIPLES TO PHARMACODYNAMICS. 339 plants are expressions of molecular peculiarities, that such peculiarities should be shown in individual substances as well as in the cellular aggregates that constitute the basic data of the systematic" botanist; and, moreover, inasmuch as it has been foimd in these researches that the starches arc modified in relationship to botanical peculiarities, and hemoglobins in relation to zoological peculiarities (and incidentally the same relationships in the case of glycogens and certain other complex synthetic metabolites), that there is adequate justification for assinning that corresponding modifications will be found to exist in alka- loids. It would therefore seem improbable that such complex organic substances as hyos- cyamine, strychnine, and other alkaloids, which in each case have been obtained from different species, and in some instances from different genera and even different families and orders, exist in single forms, but rather that each substance exists in as many stereoiso- meric forms as there are different botanical sources of origin. For instance, hyoscyamine has been obtained from Hyoscyamus niger, Alropa belladonna, Datura slramonimn, Scopolia carniolica, Duboisia myoporoidcs, Anisodus luridus, and Lachica saliva, and also from other species of some of the genera represented. Here are seven genera that belong to two families {Solanaccw and Compositw) that are so distantly related as to be assigned to dif- ferent orders {Polemoniales and Valerianalcs). That stereochemic differences have not been recorded, or that subtle physiological difTerences may not have been observed in so \'irulent a poison, is by no means evidence in disproof of the hypothesis of different stereo- isomeric forms. In fact, our knowledge of the chemical constitution of alkaloids is gener- ally very meager, and is not only being added to but accepted statements are being modi- fied or entirely discarded. A few years ago hyoscyamine and atropine were given different molecular fornuda^, but now they are recognized as stereoisomers; pitiu'ine (CioHjoN;) from Duboisia hopiooodii was believed to be identical with nicotine (CioHi,,N.,), but is now recog- nized as being a distinct individual; and the statement is frequently found that a given plant species contains poisons which resemble those of another of the same or of a related genus. Such instances might be considerably nmltiplied in order to show not only the superficiality of our knowledge of the chemistry of the toxic constituents of plants, but in justification of the hypothesis of the existence of such substances in stcreoisomeric forms. Moreover, as instanced in this memoir, and recorded in much subsequent in\'estigation as yet unpublished, hybridization has the effect, in both plants and animals, of causing definitely distinctive differences in the offspring in both gross and molecular properties, so that, as regards especially the latter, the starch or hemoglobin, etc., can easily be dis- tinguished from the corresponding substance of the parents. It would follow from the foregoing, as a corollary, that given medicinal substances may, by hybridization, be modi- fied in their physiological properties because of stereochemic changes. • In the Introduction (page 8 et seq.) attention is directed to the extremely important fact that slight stereochemic differences may be manifested in more or less marked varia- tions in physiological actions, as have been recorded in the nicotines, cocaines, hyoscines, hyoscyamines, and other organic substances; and, moreover, that under conditions similar to or identical with those which exist in plants there may occur a transmutation of one stereoisomeric form into another. In supplementation of what is there stated, it may be of value to refer to certain facts of much significance: Alropa belladonna yields both atropine and hyoscyamine, the former being the racemic form of the latter, while the latter is Isevo-rotatory (dextro-hyoscyamine is an artificial base). It was a baffling problem for years why the proportion of these two substances obtained from given specimens of belladonna varied with differences in the laboratory methods used in their preparation, and it was finally discovered that there occurred a ready transmutation of the hyoscyamine (lajvo- rotatory) into atropine (racemic) in the presence of fusion, a weak alcoholic solution of sodium hydroxide, ammonia, etc. Another instance of especial interest is recorded in the researches of Lanterer with the active principles of Duboisia. He found that the old 340 DIFFERENTIATION AND SPECIFICITY OF STARCHES. leaves of D. myoporoidcs yield hyoscyamine (C17H23NO3), but the young leaves scopolamine (Ci7H;,N04), alkaloids that are closely related chemically and physiologically. Whether or not the difference in the product of the old and new leaves is owing to a chemical trans- formation of the scopolamine into hyoscyamine, or to a change in the protoplasmic mechan- ism incident to development, or to some other obscure condition, is entirely speculative, yet this instance (coupled with the foregoing, and to which others might be added) is highly suggestive of the plasticity of such substances and of protoplasm, and hence of the exceedingly important results that may be expected in the modifications of stereoisomers by hybridization. GENERAL APPLICATIONS OF THIS RESEARCH IN SYSTEMATIC BOTANY. It follows, as a corollary from the foregoing data, that upon the basis of intramolec- ular and intermolecular peculiarities of starches, plants can be classified by means not only of the modifications that are observed in their metabolites, as are manifested in the gross macroscopic forms, such as the leaves, flowers, roots, sexual organs, etc., but also by means of the molecular characters of products which are passive, non-structural constituents of the plant. It will be observed that while the findings of this research are in general in correspondence with the recognized classifications of the systematic botanist obtained by means of macroscopic and microscopic methods, they also are in harmony with the shifting of species from one genus to another, and in the remodeling of classes, famiUes, etc., that is continually going on ; and that by the aid of the peculiarities of such metabolites as starch it is manifest that a logical, practically permanent, and scientific classification of plants can be established. For instance, Galtonia candicans was for years known as Hya- cinthus candicans and referred to the genus Hyacinthus. It is clear from the reaction curve that its assignment to another genus is correct, and judging from the peculiarities of the reaction curves it stands between the hyacinths and the members of the tulip tribe. Many plants known years ago as species of Amanjllis have since been moved to other genera, as for instance (among plants represented in this research) Hippcastrum cquestre and H. auHcum, Sprekclia formosissima, Sternbergia lutea, Vallola purpurea, and Zephyranihes Can- dida, all of which have been or are recorded as being corresponding species of Amaryllis. It will be seen, by examining plates 51 to 54 and charts 241 to 246, inclusive, and 252, that while all of the starches have in common the same gross fundamental histological characters, the reaction curves show quite clearly that not one could be mistaken for a species of Amaryllis. Likewise, some of the Crinums have been designated Amaryllis, but it will be noted by an examination of charts 241 and 242 that the generic pecidiar- ities of Crinum are quite distinct from those of Amaryllis. Likewise, Tritonia and Mont- bretia have been confused: Tritonia pottsii was named by Baker Montbretia pottsii; T. crocosynceflora is a hybrid of T. pottsii X pollen of Crocosmia aurea, also known as T. aurea Pappe. Comparing the forms of the grains of the six specimens of Tritonia studied in this research, it ^vill be noted that the grains of T. crocata and T. securigera have in common certain features wliich distinguish them from T. pottsii and its hybrid, which in turn have some characteristics in common. An examination of the reaction-curves will show this same grouping very strikingly, the curves of the first and those of the last corresponding. The differences are so marked that it would seem that T. pottsii is misclassed. T. securigera has been known as Gladiolus securiger, but while the starch-grains have general character- istics in common with those of Gladiolus a comparison of the reaction-ciu'ves indicates clearly that its grouping as a Tritonia is correct. Again, there has been a shifting of Tritonias and Ixias, and here too, wliile the gi'ains belong to the same class so far as their shapes are concerned, the reaction-curves are dis- tinctly unlike, so much so that there should not be any difficulty in distinguishing the members which strictly belong to one genus from those of the other. Tritonias and Free- APPLICATIONS TO SYSTEMATIC BOTANY. 341 sias have also been confounded, and, in fact, among the earliest illustrations of Freesia are pictures of Tritonia refracta and T. odorala. While the starch-grams of these genera ai-e so ahke in shape that it would be difficult or impossible to distinguish one from the other with certainty, the reaction-curves differ distinctly, and those of Freesias bear much closer resemblances to T. pottsii and its hybrid than to T. crocata and T. securigera. Finally, the systematic classification of plants is of an arbitrary character, as is evi- dent in the large number of classifications that have been used frum time to time, in the continual changes that are going on at the present time, and in the recognition that there is no system at present that is uni\'ersally accepted or which is regarded as having more than a tentative value. In view of these facts, coupled with the specificities of starches which have been shown to exist in relation to species and genera, and the concordance of these specificities with the classification of the botanist whenever the assignment of the latter has been accepted as being absolutely or reasonably permanent, it would seem to follow, as a corollary, that where doubt exists as to classification that is based upon botan- ical characters, such doubt can be confirmed or set aside and revisions suggested by the reaction method employed in this investigation. It has been conclusively shown that all members of a genus have a certain tj'^pe of reaction, and that each may be distinguished from the others by variations of the reaction-curves; that in the case of all genera of a given family the genera when closely related show striking resemblances in the types of their reaction-curves; and that the farther the members of a genus or of a family are separated, the greater the departures from the corresponding given types of curves that may be taken to be typical of the genus or family. If, therefore, in a given recognized genus we find among its members one, for instance, that has a reaction-curve that does not correspond with those of the others, it may be taken for granted that it either belongs to another genus or is a hybrid resulting from a cross with a member of another genus, or some aberrant form, etc.; and again, if the reaction-curves of different genera assigned to a family do not correspond in general characters, it may be likewise concluded that one or more of the genera are inisclassified. The systematic botanist, by the conventional methods of research, is continually finding e\'idence of misclassification, as is evident in the continual shifting of species and genera in the rebuilcUng of broken-up famifies, and in the breaking-up of families by assign- ing certain genera to other families or establishing entirely new fanfifies. Thus we find in very recent years that ConvaUaria, Trillium, and Asparagus have not, by some botanists, been included among the Liliacece, but have been set apart as a distinct family, Conval- lariacece; and Aletris has been transferred to Hoemodoraceoe. Musacece, Zingiberacea, Can- naccw, and Marantaccce are by some grouped in one family, but by others as separate families. We are therefore deafing not with a stable classification, but one that is most unstable. If therefore, as stated, we find in the reaction-cm-ves evidence of misclassification, it may be taken for granted that we have the strongest kind of evidence to suggest modifications. For instance, it will be seen that reaction-curves of Lilium, Fritillaria, Calochortus, Tulipa, Scilla, Chionodoxa, Ornithogalimi, Puschkinia, and Enjthronium are all in accord with a given type; that in Hyacinthiis and Muscari there is a marked modification of tliis type wliich is manifested chiefly in the much lower temperature of gelatinization and the low reactive in- tensity with Purdy's solution; that in Galtonia there is another kind of mocUfication, and of a character which suggests that Galtonia stands, as it were, between Hyacinthus and Muscari on the one hand and the fii'st group (which constitutes almost wholly the tulip tribe) on the other; that in Brodimi and Triteleia there are types of curves wliich bear close resemblances to each other, but which cfiffer distinctly from those of the other groups noted. According to tliis data, without going into further detail, all of wliich is in accord with the data of the botanist, it would seem that there is a logical basis for a separation of the Liliacew into a number of families, each of which can be distinguished by the sum of the botanical and 342 DIFFERENTIATION AND SPECIFICITY OF STARCHES. reaction peculiarities. Upon such a basis a family might be formed including Brodicea, and certain other genera of the Liliacece that have a recognized very close relationship. Similarly, it seems manifest that the family IridacecB, as now constituted, contains a number of genera which should be assigned to other families; as it is, it certainly seems, from the reaction-curves and forms of the starch-grains, to be a very heterogeneous group. If we take the reaction-curve of Iris as being the family type, it will be observed that, as in the case of Liliaceo', the reaction-curves of some of the genera correspond, as for instance Tigridia, while those of others differ so markedly as to suggest misclassification. Thus, for instance, the curves of Morcea and Homeria differ so markedly from the Iris type as to suggest that the genera properly belong to another family. Homeria is stated to be closely allied to Tigridia, but there are certainly well-defined differences in their starches which do not confirm this view. Then, again, Gladiolus, Watsonia, Tritonia, and Antholijza have forms of starch-grains and reaction-curves wliich correspond, but which do not harmonize with the forms and curves of Iris; and Marica and Gelasine differ so from each other, and also from Iris, in the same respects that they could not be identified as members of the same family. In other words, it would seem that the family Iridacew contains a number of genera wliich strictly speaking could with much appropriateness be assigned to other families. These are but few of the many instances suggested bj^ the results of this preliminary study, which indicate that by means of such methods as have been pursued in this research the classification of plants may be revolutionized and placed upon a lasting basis. (See Prefatory Notes in Part II.) SUMMARY AND CONCLUSIONS. (1) Every starch and every mature starch-grain is a mixture of different forms of the starch-substance. (2) Owing to the heterogeneity of starches and the crudities of the methods employed in the differentiation of different starches the results of this investigation are to be regarded as being for the most part of a gross quantitative and cjualitative character. (3) The temperatures of gelatinization of starches are as specific in relationship to the chemical composition and constitution of starches as are the melting-points of various isomers in their distinction, and the method employed in this research is approximately exact. (4) With each agent the starches exhibit a wide range of reaction-intensity, with a tendency to a close correspondence between varieties of a species, closely related species, and closely related genera, respectively. (5) The characters of the modified forms of starch in relation to each genus and each species are specific and constant. (6) In pharmacognostics, commerce, and technical pursuits the methods employed in this research must obviously have great value in the identification of vegetable drugs and starches, and in the detection of adulterations and substitutions, and in showing the way of providing starches and other plant products which have such special properties as may be demanded in certain textile industries, etc. (7) It is probable that by hybridization and other procedures of the plant-breeder the properties of medicinal substances can be modified by the changes in stereoisomeric forms. (8) Stereochemic peculiarities of starches, hemoglobins, and similar complex synthetic metabolites constitute a siriclly scientific basis for the classification of all forms of life. Note.— At the ends of Chapters II, III and IV (pages 79, 160 and 195, respectively) there will be found sum- maries and conclusions relating to the subject-matter of those chapters. INDEX OF STARCHES. Abildgaardia monostachya Vahl. Root-stock, 233, 269 Abroma angustum Linn. Seed, 220 Abronia arenaria Hook. Seed, 289 Acacia farnesiana Willd. Seed, 251 latifolia Desf. Seed, 251 melanoxylon R. Br. Seed, 251 Acantlius mollis Linn. Seed, 203 Acer laurinum Hook. Seed, 250 pseudoplalanus Linn. Seed, 250 Achimcnes alba Hort. Root-stock, 213 hirsuta. (See Looheria hirsuta.) tubitiora Nichol. Root-stock, 231 Achnodonton bellardia Ueauv. Seed, 277 Achyranthes argentea Lam. Seed, 288 fruticosa Lam. Seed, 288 patula Linn. f. Seed, 288 Acnida tuberculata Moq. Seed, 287 Aconitum anthora Linn. Root, 217, 272 napellus Linn. Root, 72, 272, 859 pyramidale Mill. (See A. napellus.) tauricum Jacq. (See -A. napellus.) Acorus calamus Linn. Root-stock, 72, 235 Acroglochin chenopodioides Schi'd. Seed, 287 persicarioides Moq. Seed, 287 Actsea alba Willd. Root, 861 rubra Bigl. Root. (See A. spicata var. rubra.) rubra Willd. Root. (See A. spicata var. rubra.) spicata var. rubra Ait. Root, 862 Actinocarpus damasoniuni Sm. Seed, 206 Adarasia scilloides Willd. (See Puschkinia scilloides.) Adansonia digitata Linn. Pulp of Fruit, 241 Adenogramma galioides Fzl. Seed, 291 Atliantum capillus-veneris Linn. Root-stock, 232 Adonis araurensis Regel and Radde. Root-stock, 870 Adoxa, 50, 67, 68, 70, 71 moschatellina Linn. Root-stock, 187, 217 .iEgiceras majus Gart. Seed, 265 ^gilops. Seed, 246 iEgilops caudata Linn. Seed, 206 truincialis Linn. Seed, 206 iEgopogon cenchroides Willd. Seed, 276 multisetus Trin. Seed, 276 jEscuIus hippocastanum Linn. Seed, 181, 183, 186, 220, 438 Agathophyllum aromaticum Willd. Seed, 248 Agropyrum. Seed, 246 Aeropyrum cristatum R. S. Seed, 205 rigidumR. S. Seed, 205 Agrostenima coronaria Linn. Seed,J292 Agrostis spica-venti Linn. Seed, 277 verticillata Vill. Seed, 277 Aira agrostidea Loisl. Seed, 280 ca-spitosa Linn. Seed, 280 canescens Linn. Seed, 280 globosa Thore. Seed, 280 juncea Vill. Seed, 280 pulchella Willd. Seerl, 280 tenorei Guss. Seed, 280 Airopsis agrostidea Loisl. Seed, 280 arapla Nees. Seed, 280 globosa Desv. Seed, 280 Aizoon canariense Linn. Seed, 290 hispanicum Linn. Seed, 290 1 Alangium decapetalura Lam. Seed, 267 hexapetalum Lam. Seed, 267 Albersia blitum Kth. Seed, 288 Alchemilla alpina Linn. Root-stock, 263 Aleurites moluccana Willd. Seed, 250 sp. Seed, 250 Alisma plantago Linn. Root-stock, 234 ranunculoides Linn. Seed, 206 Alismacese, 197 Allionia incarnata Linn. Seed, 289 nyctaginea Mich. Seed, 289 ovata Pursh. Seed, 289 Alocasia putzeysi Hort. Rhizomes, 472 Alopecurus alpinus Sm. Seed, 277 geniculatus Linn. Seed, 277 pratensis Linn., 277 utriculatus Schr., 277 Alpinia galanga Sw. Root-stock, 230 nutans Rose. Pollen, 243 Alstrceraeria aurantiaca (aurea) Don. Root-stock, 661 brasiliensis Spreng. Root-stock, 660 ligtu Linn. Root-stock, 659, 815 Alternanthera paronychoides St. H. Seed, 288 Althtea officinalis Linn. Root, 218 rosea Cav. Root, 241 Althenia filiformia Petit. Seed, 209 .Alysicarpus ferrugineus Steud. Root-stock, 263 Aniarantus blitum Linn. Seed, 288 buUatus Bess. Seed, 288 frumentaceus Roxb. Seed, 288 sanguineus Linn. Seed, 288 Amaryllidacese, 625-685 .\maryllis belladonna major Linn. Bulb, 625 formosissima Linn. Bulb. (See Sprekelia formossisima.) Amblyogyne polygonoides Rafn. Seed, 288 Ambrina graveolens Moq. Seed, 287 Ammannia baccifera Linn. Seed, 250 latifolia Linn. Seed, 250 vesicatoria Rxb. Seed, 250 Ammophila arenaria Lk. Seed, 278 Amonium cardamomum Linn. Seed, 253 granum-paradisi Afz. Seed, 253 javanicum. Seed, 253 zingiber Linn. Seed, 253 Amorphophallus rivieri Dur. Tuber, 472 Ampelodesmos tenax Lk. Seed, 200 Amphicarpiea monoica Nutt. Seed, 221, 268 Ampliicarpum purshii Kunth. Seed, 199 Amyris sylvatica Jacq. Seed, 221 sp. Seed, 221 Anacardium occidentale Linn. Seed, 210 Anarthria prolifera R. Br. Seed, 286 Anatherum ivvarancusa;. Root, 66, 255 muricatum Beauv. Root, 255 Andropogon aciculatus Retz. Seed, 201 argenteus DC. Seed, 201 cernuus Rxb. Seed, 201 contortus Linn. Seed, 201 cymbarius Ilochst. Seed. 202 cymbarius I, 03 Forestiera acuminata Poir. Seed, 248 Frangida, 71 Frankenia pulverulenta Linn. Seed, 290 Freesia refracta var. alba Ilort. Bulb, 194, 736 leiehtlinii Hort. Bulb, 738 Fritillaria. Bulb, 65 Fritillaria armena Boiss. Bulb, 504 aurea Schott. Bulb, 503 inferialis var. aurora Hort. Bulb, 505 latifolia Willd. (See F. aui'ea.) lutea Miller. (See F. aurea.) lihaeea Lindl. Bulb, 506 meleagris Linn. Bulb, 189, 222, 496 persica Linn. (Plate 27, figs. 101, 162.) pudiea Spreng. Bulb, 499 pyrenaica Linn. Bulb, 498 recurva Benth. Bulb, 507 Frcelichia floridana Moq. Seed, 289 gracilis Moq. Seed, 289 Gagea lutea Schult. Bulb, 234, 269 stenopetala Rchb. Bulb, 234 Galanthus elwesii Hook. Bulb, 656 nivalis Linn. Bulb, 188, 223, 270, 655 plicatus Bbrst. Bulb, 235 Galega biloba Sweet. Seed, 250 Galipca officinalis Hanc. Bark, 241 Galtonia candicans Decne. Bulb, 574 Gastridium au.strale Be.auv. Seed, 278 muticum Spr. Seed, 278 G.audinia fragilis Beauv. Seed, 280 Gaya sini])lex Gd. Root-stock, 238 Gelasine aziu-ea Hort. Root-stock, 754 Gelsomium, 72 Geranium, 72 lividuni Her. Root-stock, 218 16] INDEX OF STARCHES. Geranium molle Linn. PoUon, '2-V,i pluuum Linn. Kool,-s(,ock, 21S piatensc Linn. I'ollcn, 2i:i sj'lvaticum Liun. Root-stock, 21!) Gcsnprarox, SS4-8SS Gosncria tubiflora Hort. Root-stock, 880 Gcsneriana. Bulb, 188 Gcum montanum Linn. Root-stock, 263 urbanum Linn. Root-stock, 26.3 Gladiolus byzantinus Miller. Corra, 715 cardinalis (Blushing Bride) Hort. Conn, 717 communis Linn. Conn, 256 floribundus Jacq. Conn, 718 primullnas Hort. Corm, 716 Glaux maritima Linn. Stem, 238, 260 Globba marantina Linn. Seed, 228 nutans. Pollen, 585 Globularia pilulifera. Androspore, 208 Gloriosa superba Linn., 201 Gloxinia hirsuta Lindl. Tuber, 216 speciosa Lodd. Tuber, 216 tubiflora Hook. Tuber, 231 var. Hort. Tuber, 888 Glutinosa, 53 Glyceria aquatica Presl. Seed, 281 distans M. K. Seed, 281 maritima M. K. Seed, 281 michauxii Kth. Seed, 281 nervata Trin. Seed, 281 Glycyrrhiza. Root, 72, 242 Gora]ilirena decumbens Jacq. Seed, 289 Goniolimon exiraium Boiss. Seed, 249 Gossypium. Bark, 72 Gossypium indicum Lam. Root, 273 Graminacea;. Seeds, 343 Granatum, 72 Gratiola officinalis Linn. Root-stock, 238, 259 Guaiacum officinale Linn. Bark, 242 Giiarea trichilioides Linn. Seed, 267 Guilandina bonduc Linn. Seed, 251 Guthnickia atrosanguinea Reg. Root-stock, 231 Gymnopogon foliosus Nees. Seed, 200 Gymnothrix cenchroides R. S. Seed, 200 Gyncrium argenteuni Nees. Seed, 278 cinereum Humb. Seed, 278 Gypsophila altissima Linn Seed, 292 Hablitzia tamnoides Bbrst. Seed, 287 Hiemanthus katherina; Baker. Bulb, 644 ILemodorum sp. Seed, 264 ilarpachne schimperi Hochst. Seed, 282 Heckeria sidaefolia Kth. Seed, 287 Hedychium coronarium Koenig. Root-stock, 787 flavescens Car. Root-stock, 224 gardnerianum Wall. Seed, 253 Root-stock, 230 Roscoc. Root-stock, 788 hirsutum. Root-stock, 230 Helleboracea;, 859-864 Helleborus hyemalis Linn. (See Eranthis hyemaUs.) Heleocharis ovata R. Br. (Eleocharis). Seed, 246 palustris R. Br. (Eleocharis). Seed, 246 Helianthemura a!gyptiacum Mill. Seed, 210 Heliconia sp. Seed, 254 Helleborus dumetorum W. K. Root-stock, 272 viridis Linn. Root-stock, 272 Helopius annulatus Nees. Seed, 200 Hemarthria fasciculata Kth. Seed, 201 Hemerocallis fulva Linn. Root-stock, 229 Heritiera littoralis Ait. Seed, 266 Hermannia althea;folia Linn. Seed, 220 nemorosa Eckl. Seed, 220 Hernandia. Seed, 248 Herniaria glabra Linn. Seed, 291 Hesperoscordum hyacinthum Hort. (Sec B. lactea.) lactum Hort. (See Brodia;a lactea.) Heterachtia pulchella Kze. Seed, 253 Heteranthelium pilifcrum Hchst. Seed, 205 Heteranthera limosa Vahl. Seed, 208 llctcropogon contortus R. S. Seed, 201 llierocliloa borealis R. S. Seed, 275 llimantoglossum hircinuin Rich. Bulb, 230 Hippeastrum aulicum var. robustum Hort. Bulb, 630 equestre Herb. Bulb, 629 vittatum Herb. Bulb, 627 Hohenbergia strobilacea Schult. f. Seed, 202, 264 Holcus lanatus Linn. Seed, 275 Homeria coUina Vent. Bulb, 710 Hookcra californica Greene. (See Brodia;a californica.) coronaria Sali.sb. (See Brodiase grandiflora.) Hoplotheca floridana Nutt. Seed, 289 texana A. Br. Seed, 289 Hordeacea!. Seed, 205-246 Hordeum. Seed, 65, 69, 166, 169, 176, 177, 178, 179, 180, 187, 189, 190, 191, 192, 193, 194, 246 Ilordcum bulbosum Linn. Seed, 206 distichum Linn. Seed, 179 himalayeu-sc Ritt. Seed, 206 sativum var. (Champion) Hort. Seed, 372 vulgare Linn. Seed, 206 Hyacinthus. Bulb, 50, 68 H. botryoides Hort. (See Muscari botryoides.) H. candicans Baker. Bulb. (See Galtonia candicans.) H. comosus Hort. Bulb. (See Muscari comosum.) H. compactus Hort. Bulb. (See Muscari compactum.) H. orientalis Linn. Bulb, 188, 222, 251, 254 II. orientalis var. alba superbissima Hort. Bulb, 570 H. orientalis albulus (white) Hort. Bulb, 571 H. orientalis albulus (Italian )Hort. Bulb, 573 H. racemosvisHort. Bulb. (See Muscari racemosum.) Hydrastis, 71 Hydrocharis morsus-rana; Linn. Seed, 209 Hydrocotyle vulgaris Linn. Root, 260 Hydrodictyon, 54 Hydrophyllum virginicum Linn. Root-stock, 227 Hydropyrum esculentum Lk. Seed, 275 Hymenocallis calathina Nicols. Bulb, 648 undulata Herb. Bulb, 646 Hj'pericum coris Linn. Root-stock, 241 elodes Linn. Root, 262 Hypogynium campestrc Nees. Seed, 201 Icica pubescens Bnth. Seed, 267 Illecebrum verticillatum Linn. Seed, 291 Imperata anmdinacea Cyr. Seed, 202 Imperatoria ostruthiuni Linn. Root-stock, 239 Ipecacuanha, 72, 271 Ijiomoea batatus. (Sec Batatus edulis.) purga Schlcht. Tuber, 213 turpethum R. Br. Root, 237 Iresine nervosa Hort. Seed, 289 Iridaceae, 688-670 Iris alata Poir. Bulbous rhizome, 702 bismarckiana Hort. Rhizome, 693 caucasica Hoffm. Bulbous rhizome, 703 florentina Linn. Rhizome, 65, 72, 230, 688 germanica Linn. Rhizome, 36, 50, 67, 69, 70 histrio Reichb. Bulbous rhizome, 700 iberica Hoffm. Rhizome, 694 pallida Lam. Rhizome, 36, 230 speciosa Hort. Rhizome, 690 pumila var. cyanea Hort. Rhizome, 691 reticulata M. Bieh. Bulbous rhizome, 699 sambucina Linn. Rhizome, 230 tingitana Boiss and Rent. Bulbous rhizome, 698 xiphium var. Grand Tresorier Hort. Bulbous rhizome, 695 xiphium var. Lusitanica Hort. Bulbous rhizome, 697 Wilhelmine Hort. Bulbous rhizome, 696 Isachne australis R. Br. Seed, 199 Ismene calathina Herb. (See Hymenocallis calathina.) [7] INDEX OF STARCHES. Isoetes lacustris Linn. Gymnospore, 243 Stem, 269 Isolepis eckloniana Sehrd. Seed, 247 holoschoenus R. S. Seed, 247 setacea R. Br. Seed, 247 supina R. Br. Seed, 247 vcrruculosa Steud. Seed, 247 Isoloma vestitum Bnth. Root-stock, 227 Ixia epeciosa Andr. Bulb, 760 var. (Emma) Hort. Bulb, 763 viridiflora Lam. Bulb, 762 Jambosa vulgaris DC. Seed, 221 Jatropha curcas (Manihot) Linn. Root, 66, 262, 874 Juncus. Seed, 206 Juncus acutifloru.s Ehrh. Seed, 207 balticus Deth. Seed, 207 Root-stock, 234 bulbosus Linn. Seed, 207 Root-stock, 234 compressus Jacq. Seed, 207 Root-stock, 234 effusus Linn. Seed, 207 glaucus Ehrh. Seed, 207 sylvaticus Reich. Seed, 207 var. macrocephalus. Seed, 207 tenuis, 41 Kobresia caricina Willd. Seed, 247 Koehia scoparia Schrd. Seed, 287 Kceleria laxa Lk. Seed, 282 Kramcria triandra R. P. Root, 72, 231, 262 Krokeria edulis. Seed, 210 Kyllingia odorata Vahl. Seed, 246 Lablab vulgaris Sav. Seed, 212 Lachenaha fcndula Ait. Bulb, 610 tricolor var. luteola Hort. Bulb, 611 Lachnanthcs tinctoria EU. Seed, 209 Lagurus ovatus Linn. Seed, 278 Lamarkia aui'ea Mnch. Seed, 282 Lappago racemosa Willd. Seed, 200 Lasiagrostis calamagrostis Lk. Seed, 276 Lathrxa. Seed, 188, 189 Lathra^a squamaria Linn. Root-stock, 65, 216 Lathyrus, 72, 174 aphaca Linn. Seed, 211 latifolius var. albus Hort. Seed, 398 magellanicus var. albus Hort. Seed, 399 ni.ssolia Linn. Seed, 211 odoratus var. shahzada Hort. Seed, 395 palustris Linn. Root-stock, 263 pratensis Linn. Root-stock, 263 sativus Linn. Seed, 211 sylvestris Linn. Seed, 397 Laurus exaltatus Sieb. Seed, 214 nobilis Linn. Seed, 214, 265 Lecanocarpus cauliflorus Nees. Seed, 287 Lechia thymifolia Michx. Seed, 210 Leersia oryzoides Sw. Seed, 275 Leguminosa;, 378^17 Lens esculenta var. Hort. Seed, 174, 176, 193, 393 Lepigonum medium Wahlbg. Seed, 291 Leptandra, 71 Lepturus filiformis Trin, Seed, 285 incurvatus Trin. Seed, 285 Lepyrodiclis holosteoides Fzl. (Arenaria). Seed, 291 Leucoium ajstivum Linn. Bulb, 621 vernum Linn. Bulb, 223, 653 Levisticum officinale Koch. Root, 239, 200 Lileae subulata H. B. Seed, 206 Liliacea>, 40, 65, 474-615 Lilium auratum Lindl. Bulb, 485 autumnale Hort. Bulb. (See L. superbum.) bloomeriammi Kell. Bulb. (SeeL.pardalinum.) bulbiferum Linn. Bulb, 222 Seed, 248 californicum Domb. Bulb. (SeeL.pardalinum.) Hort. Bulb. (SeeL. puberulum.) 18 Lilium candidum Linn. Bulb, 188, 222, 474 Seed, 188, 189 dalhansoni Hort. Bulb. (See L. martagon.) dalmaticum Vis. Bulb. (See L. martagon.) eximium Nichol. Bulb. (See L. longifolium var. eximium.) harrisii Carr. Bulb (See L. longifolium var. eximium.) humboldtii Roez. and Leicht. Bulb. (SeeL. pardalinum.) henryi Baker. Bulb, 484. longifolium var. eximium Nichol. Bulb, 478 giganteum Hort. Bulb, 476 martagon Linn. Bulb, 488 pardalinum Kellogg. Bulb, 171, 492 parryi Wats. Bulb, 479 philadelphicum Linn. Bulb, 481 puberulum Duclu-. Bulb, 493 pumilum Hort. Bulb. (See L. tenuifolium.) rubellum Baker. Bulb, 480 spceiosum var. album Hort. Bulb, 487 superbum Linn. Bulb, 489 tenuifolium Fisch. Bulb, 491 tigrinum var. siilcndens Leicht. Bulb, 483 wa-sliingtonianum Hort. Bulb. (See L. parda- linum.) Linieum glomeratum E. Z. Seed, 292 Liranocharis pimnieri Rich. Seed, 208 Liraonia trifoliata Linn. Seed, 266 Liuiuii raphanthus, 189 Locheria hirsuta Reg. (Achimenes). Root-stock, 227 Lolium canadense Michx. Seed, 285 speciosum Link. Seed, 285 temulcntum Linn. Seed, 285 Lojihochla^na californica Nees. Seed, 281 obtusiflora Trin. Seed, 281 Lopholepis omithocephala Dcsne. Seed, 200 Loranthus europaius Linn. Seed, 219 Lotus edulis Linn. Seed, 210 Luca;a colorata Hchst. Seed, 264 Lucuma cainito, A. DC. Seed, 266 rivicoa Gart. f. Seed, 265 sp. Seed, 266 Luzula. Seed, 20(3 Luzulu forsteri Desv. Seed, 207 multiflora Lejeun. Seed, 207 nivea DC. Seed, 207 spadicca DC. Root-stock, 234, 256 Lychnis cceli-rosa Desr. Seed, 292 coronaria Desr. Seed, 292 dioica Linn. Seed, 292 \'cspertina Sibth. Seed, 292 Lycurus phalaroides H. B. Seed, 276 Lygeum spartum Linn. Seed, 275 Lysimachia vulgaris Linn. Root-stock, 238, 260 Macis, 72, 176 Maizilla stolonifera Schlcht. Seed, 199 Mavaca michauxii, Sch. E. Seed, 286 vandellii Sch. E. Seed, 286 Malva borealis Wallm. Root, 273 Mamillaria cUscolor. Stem, 240 Mammea americana Linn. Seed, 232 Mangifera sp. Seed, 210 Manihot aipi Pohl. (See M. palmata var. aipi.) carthagenesis. Root,, 815 palmata var. aipi Miill. Root, 814 utilissima Pohl. Root, 814 Secd,72,170, 176, 190, 191, 262, 876 Manisuris granulans Sw. Seed, 201 Maranta, 27, 65, 66, 70, 169, 174, 176, 177, 180, 189, 190, 191, 193, 194 arundinaccaLinn. RootStock, 194, 224, 813, 814, 815 var. Hort., 818, 820 bicolor Ker. Root-stock, 257 INDEX OF STARCHES. Maranta indica Rose. Root-stock, 257, 813 kerehovcana E. Morr. Root-stock. (SeeM. Icuconeura.) kcrchovei Hort. Root-stock. (Sec M. Icu- coneura.) Icuconeura E. Morr. Root-stock, 824 massangpana E. Morr. Root-stock, 822 musaica Hort. Root-stock, 826 ramosissima Wall. Seed, 229 sanguinoa Hort. (See Stronianthe sanguinca.) sp. Seed, 255 Marantacea;, 813-839 Marattia. Leaves, 66 Marica gracilis Herb. Rhizome, 753 Mariscus ehitus Vahl. Root-stotk, 269 jaoquini H. K. Seed, 246 umbellatus Vahl. Seed, 246 Marsilea diffusa, 38 pubescens Ten. Androsporc, 208 Gymiiosiiore, 208, 251 Mayaca michauxii Schott and Endl. Seed, 286 vandellii Schott and Endl. Seed, 286 Melanocenchrjs royleana Nees. Seed, 280 Melhania didvnia E. Z. Seed, 220 erytliroxylon R. Br. Seed, 220 Melica ciliata Linn. Seed, 281 Melochia corchorifolia Linn. Seed, 220 p\Tamidata Linn. Seed, 220 Root, 261 Memecylon amplexicaule Rxb. Seed, 221, 267 capense E. Z. Seed, 221, 267 Menispermum palmatum Lam. Root, 217 Menodora. Seed, 249 Meristostigma silenoides Dictr. Bulb, tuber, 256 Mesembryanthemum pinnatifiiluni Linn. Seed, 290 Meum athamanticum Jacq. Root, 238 Mezereum, 72 Microchloa setacea R. Br. Seed, 279 Microla>na stipoides R. Br. Seed, 275 Microtea maypiu'ensis Don. Seed, 292 Milium effusum Linn. Seed, 275 vernale Bbrst. Seed, 275 MirabiUs jalapa Lmn. Seed, 181, 185, 289 longiflora Linn. Seed, 289 Mitella diphylla Linn. Root-stock, 239 pentandra Linn. Root-stock, 239 Moacurra sp. Seed, 293 Moliuia coerulea Mnch. Seed, 282 Root, 233 MoUugo cerviana Ser. Seed, 290 verticillata Linn. Seed, 290 Monochoria plantaginea Kth. Root-stock, 256 Monocosmia corrigioloides Fzl. Seed, 290 Monolepis chenopodioides Moq. Seed, 288 Montbretia pottsii Baker. Bulb. (SeeTritoniapottsii.) Montia minor Gmel. Seed, 290 Moraea coUina Thumb. (See Homeria collina.) tristis Ker. Corm, 708 Mougeotia gracilis Kutz. Seed, 293 Mucuna pruriens DC. Seed, 214, 391 urens DC. Seed, 214 Muehlenbergia glomerata Trin. Seed, 277 willdonowii Trin. Seed, 277 Musa, 176, 193, 194 cavendishii Lamb. Green fruit, 773 ensete Gmel. Stalk, 771 paradisiaca Linn. Fruit, 231, 814 sapicntum Linn. Stalk, 814 Musaceae, 771-778 Muscari botryoides Mill. Bulb, 222, 577 commutum Guss. Bulb, 582 comosum Mill. Bulb, 586 compactum Baker. Bulb, .585 conicum Baker. Bulb, 591 micranthvmi Baker. Bulb, 580 paradoxum C. Koch. Bulb, 579 Muscari racemosum Mill. Bulb, 583 Myriophyllum verticill:itum Linn. Stalk, 242, 262 Myristica inosihala 'I'lihg. Seed, 72. 266 saliril'olia Willd. Seed, 266 Nicgclia zebrina Reg. Root-stock, 216 Naias major Roth. (N. marina Liim.) Pollen, 243 Seed, 209 Narcissus biflorus Curt. Bulb, 674 bulbocodium Linn. Bulb, 669 var. conspieua Hort. Bulb, 669 monophvllus Baker. Bulb, 670 " clusii Dunal. Bulb. (See N. bulbocodium var. mono])hyllus.) horstieldii Burb. Bulb, 665 incomparabilis Mill. Bulb, 671 jonquilla Linn. Bulb, 675 campernelli var. rugulosus Hort. Bulb, 678 var. rugulosus Hort Bulb, 677 maximus Hort. Bulb, t)67 monophyllus Moore. Bulb. (See N. bulbo- codium var. monophyllus.) odorus Linn. Bulb, 672 poeticus Linn. Bulb, 188, 235, 673 tazetta var. orient alis Hort. Bulb, 679 Nardus stricta Linn. Seed, 276 Narthecium ossifragum Huds. Root-stock, 234, 269 Nectandra rodi;ci Schomb. Seed, 265 Nelumbo lutea Willd. Root-stock, 850 Seed, 210 nucifera Gasrtn. Root-stock, 849 Seed, 209 8i)eciosa Willd. Seed, 209 Root-stock, 815 Neottia nidus-avis, 189 Nepenthes destillatoria Liim. Seed, 203 Nephrolepsis cxaltata Schott. Base of frond, 268 Nerium oleander Linn. Leaves, 268 Niphsea oblonga Lindl. Root-stock, 213 Nitella batrachosperma. Spores, 204 exilis A. Br. Spores, 204 fasciculata A. Br. Spores, 204 flabellata Kiitz. Spores, 204 flexilis Ag. Spores, 204 gracilis Ag. Spores, 204 hyalina Kiitz. Spores, 204 syncarpa Kiitz. Spores, 204 tenuissima Kiitz. Spores, 26, 204 translucens Pers. Spores, 204 Nolana prostrata. Root, 259 Nuphar luteum Sm. Seed, 289 Nympha;a alba Linn. Root-stock, 218, 272, 840 cajrulea Savign. Seed, 289 dentata Th. and Schura. Seed, 289 gladstoniana Tricker. Root-stock, 844 marliacea var. albida Hort. Root-stock, 842 carnea Hort. Root-stock, 843 odorata Ait. Root-stock, 845 var. rosea Pursh. Root-stock, 846 rubra Roxb. Seed, 289 Nympha;aceffi, 840-852 Ochna lucida Lam. Seed, 250 squarrosa Linn. Seed, 250 ffidogonium echinospermum .\. lir. Spore, 204 landsboroughii Kiitz. !^porc, 204 vesicatum Lk. Spore, 204 Olfersia undulata Presl. Root-stock, 232 Olyra paniculata Sw. Seed, 199 Omphalodes verna Mnch. Root-stock, 237, 258 Onobrychis caput-galli Lam. Seed, 211 sativa Lam. Seed, 211 Ononis spinosa Linn. Root, 242 Ophioglossum vulgatum Linn. Spore, 233 Ophiurus a;thiopicus Rupr. Seed, 263 filiformis R. S. Seed, 285 [9] INDEX OF STARCHES. Ophiurus incurvatus Bcauv. Seed, 285 papillosus Hchst. Seed, 263 Ophryoscleria sp. Seed, 247 Oplismenus colonus H. K. Seed, 199 Root-stock, 255 frumentaceus Kth. Root-stock, 199 Opuntia brasiliensis Haw. Stem, 241 coccinellifera Mill. Green parts, 268 curassavica Mill. Stem, 241 ficus-indica Mill. Stem, 241 tuna Mill. Stem, 241 Orchis globosa Linn. Tuber, 235 latifolia Linn. Tuber, 235 m.ascula Linn. Tuber, 235 militaris Linn. Tuber, 235 Orcodajjline exaltata Nees. Seed, 214 Ornithogalum narbonense (pyramidale) Hort. Bulb, 557 nutans Linn. Bulb, 555 thyrsoides var. aureum Ait. Bulb, 558 umbellatum Linn. Bulb, 215, 556 Orobanche. Seed, 188, 189 Orobanchc procera Koch. Root-stock, 216 sp. Root-stock, 216, 260 Orobus. Seed, 268 Orobus albus Linn. Root, 219, 273 lathyroides Linn. Seed, 211 niger Linn. Seed, 211 tuberosus Linn. Tuber, 274 Orthoclada laxa Beauv. Seed, 201 Oryza sativa Linn. Seed, 53, 58, 05, 70, 166, 168, 169, 170, 176, 177, 178, ISO, 181, 185, 186, 189, 190, 191, 192, 194, 275 var. Hort. Seed, 363 (^stf'ricum palustre Bess. Root-stock, 261 ( )i I c'lia alismoides Pers. Seed, 208 Uxalis acetosclla Linn. Root-stock, 219 crenata Jaeq. Tuber, 228 lasiandra Zucc. Young bulb, 228 ortgiesi, 69 pentaphylla Sims. Young bulb, 228 stricta Linn. Root, 274 Oxybaplius cervantesii Lag. Seed, 289 (Ixyria digyna Camp. Seed, 202 Pachysandra procumbens Michx. Root-stock, 262 Pxonia. Seed, 266 Pajonia officinalis Retz. Ront-i^lock, 217, 272 Piepalanthus caulescens Kth. Seed, 286 frigidus Mart. Seed, 286 Pancratium calathinum Ker. (See Hymcnocallis cal- athina.) Panderia pilosa F. M. Seed, 287 Panicum. Seed, 59, 176, 178, 185, 190 Panicum acuminatum Sw. Seed, 200 alopecuroideum Schrb. Seed, 199 antipodum Spr. Seed, 199 arenarium Brot. Root-stock, 255 colonum Linn. Seed, 199 Root-stock, 255 commelina;folium Rdg. Seed, 275 crus-galli var. Hort. Seed, 362 frumentaccus Rxb. Seed, 199 glaucua Linn. Seed, 199 hoffmannseggii, R. S. Seed, 200 italica Linn. Seed, 199 miliaceum Linn. Seed, 200 repcns Linn. Root-stock, 255 tonsum Stcud. Seed, 200 Paniculacea;, 878-891 Papaver, 189 orientale Linn. Root-stock, 239 Papilionacese, 203 Pappophorum macrostachyum Nees. Seed, 279 nigricans R. Br. Seed, 200, 279 persicum. Seed, 278 pumilio Trin. Seed, 244 I Pappophorum scliimperianum Hchst. Seed, 278 Pareira, 72 Pari 193 194 Parietaria diffusa M. K. Root-stock, 236, 257 Paris quadrifolia Linn. Root-stock, 215 Paruassia palustris Linn. Root-stock, 240 Paspalum complanatum Nees. Seed, 199 dilatatuni Poir. Seed, 199 platycaule Poir. Seed, 199 stoloniferum Bosc. Seed, 199 Pastinaca sativa Linn. Root, 239 Patomorphe sidaefolia Miq. Seed, 287 Pedicularis acaulis Scop. Root, 238, 259 barreUerii Rchb. Root, 238, 259 rosea Wulf. Root, 238, 259 Peganum harmala Linn. Seed, 221 Pelargonium, 43 Pellionia daveauana, 43, 44, 45, 48, 49, 50, 51, 68, 70, 71 Peltandra virginica. Root-stock, 472 PenicUlaria pluckenetii. Seed, 200 spioata Willd. Seed, 200 Pennisetum cenchroides Rich. Seed, 200 longistylum Hchst. Seed, 200 typhoideum Rich. Seed, 200 Pejieromia maculosa Hook. Seed, 287 monostachya R. P. Root-stock, 270 Pereskia giandiflora Haw. Pith, 241 Perotis latifolia Ait. Seed, 278 Petiveria aUiacea Linn. Seed, 292 Petroselinum sativum Hoffm. Root, 260 Pcucedanum cervaria Lap. Root-stock, 261 Phaca alpina Jacq. Root, 242 Phacelia congesta Hook. Seed, 249 Phaius, 33, 41, 188 bicolor Lindl. (See P. wallichii.) grandiflorus Rcichb. f. (See P. wallichii.) Pseudotuber, 37, 38, 70, 189, 223 grandifolius Lour. Seed, 515. (See Bletia.) wallichii Lindl. Pseudotuber. Plate 103, figs. 617, 618. (See P. grandiflorus.) Phalaris bulbosa Cav. Seed, 275 canariensis Linn. Seed, 275 ccerulescens Dsf. Seed, 275 Phallus, 63 Pharnaceum verticillatura Spr. Seed, 290 Pharu.s scaber H. B. Seed, 199 Phascolus aureus Ham. Seed, 211 lunatus var. Hort. Seed, 387 midtiflorus Lam. Seed, 181, 188, 211 saponaceus Sav. Seed, 211 vulgaris Linn. Seed, 38, 65, 176, 178, 185, 186, 211 var. Hort. Seed, 386 Pliilodendron grandifolium Schott. Root-stock, 22, 45, 231 Philydrura lanuginosum Gart. Seed, 208 Phippsia .algida R. Br. Seed, 277 Phipsalis, 26 Phleum asperum Vill. Seed, 277 tenue Schi-d. Seed, 277 Phragmites communis Trin. Seed, 278 Physalis alkekengi lyinn. Stolon, 259 Phvsostigma, 72 Phytolacca esculenta V. H. Seed, 72, 292 Phytophysa treubii, 64 Pilularia globulifera Linn. Gymnospore, 205 minuta Dur. Gymnospore, 205 Pimenta, 72 Pimpinella saxifraga Linn. Root, 260 Pinus sylvestris Lum. Pollen, 243 Seed, 38 Piper. Seed, 254 Piper cubeba Linn. Seed, 286 nigrum Linn. Seed, 71, 176, 286 10] INDEX OF STARCHES. Piper sidiEfolia. Seed, 287 Piptatberutn multiflorum Bcivuv. Seed, 276 Pircunia latbenia Moq. SceSccci, 202 Sagina apctala Linn. Seed, 291 Sagittaria sagittrfolia Linn. Seed, 200 Sagus rumphii Willd. Stem, 06, 166, 167, 171, 176, 190, 191, 270 Salsola soda Linn. Seed, 207 Sanguinaria canadensis Linn . Root-stock, 72, 239, 272 SapLndaceK, 438-139 Saponaria officinalis Linn. Root, 241 Cell-sap, 62 persica Mey. Seed, 292 vaccaria Linn. Seed, 292 Saprolegniacea;, 63, 64 Sarsaparilla, 72 Sassafras, 72 Saurunis cernuus Linn. Root-stock, 213 Saxifraga granulata Ijinn. Bulbils, 65, 228 Scabiosa atropurpurea Linn. Seed, 265 Scheeria mexicana Seem. Root-stock, 227 Scheuchzeria palustris Linn. Seed, 206 Root-stock, 234 Scliismus marginatus Beauv. Seed, 282 Schmidtia utriculosa Stbg. Seed, 277 Schoberia corniculata Mey. Seed, 288 salsa Mey. Seed, 288 Schoenus compressus Pers. Seed, 247 mucronatus Linn. Root-stock, 214 nigricans Linn. Seed, 247 Sciadocalyx warszewiczii Reg. Root-stock, 227 Scilla amcenavar. prsecoxDon. Bulb. (SeeS.sibirica.) autumnalis Linn. Bulb, 222 bifolia Linn. Bulb, 543 ciliaris Hort. Bulb. (See S. peruviana.) clusii Pare. Bulb. (See S. peruviana.) maritima Linn. Bulb, 269 peruviana Linn. Bulb, 215, 542 sibirica Andr. Bulb, 540 Scirpus holoschoenus Linn. Seed, 247 maritimus Linn. Seed, 246 Root-stock, 255 mucronatus Linn. Seed, 246 pungens Vahl. Root-stock, 234 rothii Hoppe. Root-stock, 234 setaoeus Linn. Seed, 246 supinus Linn. Seed, 246 (riqueter Linn. Root-stock, 233 Scitaminae, 65 Scleranthus perennis Linn. Seed, 291 Scleria bracteata Cav. Seed, 247 hispidula Hchst. Seed, 247 microcarpa Nees. Seed, 247 ophryoscleria sp. Seed, 247 triglomerata Miohx. Seed, 247 Sclerochloa rigida Panz. Seed, 285 Scleropus amarantoides Schrd. Seed, 289 Scolopendrium officinarum vSw. Root-stock, 233 Scrophularia nodosa Linn. Root-stock, 237, 259 Secale cereale Linn. Seed, 65, 69, 174, 176, 178, 180, 187, 191, 193, 194, 206, 246 var. (Mammoth Winter) Hort. Seed, 368 (Spring) Hort. Seed, 370 Sedum fabaria Kocli. Root-stock, 239 Seemannia ternifolia Reg. Root-stock, 231 Serpentaria, 72 Sesieria elongata Host. Seed, 282 Setaria flava Kth. Seed, 199 glauca Beauv. Seed, 199 italica Beauv. Seed, 199 Shorea robusta Rxb. Seed, 250 Silene ambigua Camb. Seed, 292 conoidea Linn. Seed, 292 Sinapsis, 189 Sinningia, 888 Siphoneae, 63 Smilax china Linn. Iloot-stock, 234, 257 ornata. Root-stock, 64, 66 sp. Root, 257 Solanticea;, 880, 882 Solanum nigrum Linn. Root, 259 tuberosum Linn. Tuber, 19, 20, 22, 26, 41, 46, 55, 56, 59, 67, 68, 70, 72, 166, 167, 168, 169, 170, 172, 174, 176, 177, 178, 179, 180, 181, 182, 185, 186, 189, 190, 191, 192, 193, 194, 195, 213, 882 Soldanella alpina Linn. Root-stock, 238, 260 Sorghum cemuum Willd. Seed, 201 vulgare Pers. Seed, 69, 201 Sparaxis grandiflora alba Hort. Bulb, 756 var. (Albertine) Hort. Bulb, 758 Sparganium natans Linn. Seed, 248 ramosum Huds. Root-stock, 257 Spartina cynosuroides Willd. Seed, 279 Spergula arvensis Linn. Seed, 291 Spergularia salina Prsl. Seed, 291 Spinacia glabra Mill. Seed, 287 inermis Much. Seed, 287 Spigelia, 71 SpirEca filipendula Linn. Root, 242, 273 Spirogyra jugalis Kutz. Seed, 293 orthospii-a var. spiralis Nag. Seed, 293 Sporobolus coromandelinus Kth. Seed, 277 pungens Kth. Stolon, 255 Sprekelia formosissima Herb. Bulb, 642 Statice elata Fisch. Seed, 249 limonium Linn. Seed, 249 Stellaria bulbosa Wulf. Tuber, 218, 261 Sterculiacese, 219 Stembergia lutea Ker. Bulb, 235, 664 Stillingia, 72 Stipa calamagrostis Whlbg. Seed, 276 gigantea Lag. Seed, 205, 276 papposa Nees. Seed, 276 pennata Linn. Seed, 276 Stratiotes aloides Linn. Seed, 208 Stromanthe sanguinea Sender. Root-stock, 837 Strophanthus, 72 Sturmia minima Hoppe. Seed, 278 Succisa pratensis Mnch. Root-stock, 258 Sumbul, 72 Swertia perennis Linn. Root-stock, 237, 258 Symphoricarpus racemosus, 33 Symphytum bulbosum Schmp. Tuber, 215 tuberosum Linn. Root-stock, 216, 258 Syringa vulgaris Linn. Pollen, 243 Syzygium giiineense DC. Seed, 221 Tacca oceanica. Rhizome, 815 pinnatifida Linn. Rhizome, 256, 686, 815 Taccacese, 684-685 TaUnum patens Willd. Seed, 290 Tamus communis Linn. Root-stock, 229 Teleianthera polygonoides Moq. Seed, 289 Telephium imperati Linn. Seed, 291 Teloxys aristata Moq. Seed, 287 Tetragonia expansa Ait. Seed, 290 Teucrium hyrcanicum Linn. Fruit, 258 Thalia dealbata Fras. Seed, 253 Thea bohea Linn. Seed, 250, 266 Theobroma cacao Linn. Seed, 266 Thunbergia fragrans R,xb. Seed, 249 Tigridia pavonia var conchiflora Hort. Bulb, 713 grand, alba Hort. Bulb, 711 Tiimantia fugax Scheidw. Seed, 253 Tofieldia calyculata ^Vhlbg. Root-stock, 234 Tonka, 72 Tormentilla erecta Linn. Root-stock, 242 Tradeseantia dicolor. Leaves, 181 virginica Linn. Seed, 253 Tragacantha, 72 TragopjTum lanceolatum Bies. Seed, 203 Trapa natans Linn. Seed, 207 12] INDEX OF STARCHES. Trapa natans Linn. Stolon, 212 Trevirania lungillora Keg;. Hoot-stock, 227 Tiiachyruni cordofanuni llc-hst. .Seed, 277 lougifolium Hclist. Seed, 277 Trianosperma fieifolia. Hoot, S90 Trianthema monogynuiu Linn. Seed, 290 Tribuhis terrestris Linu. Root, 274 Trichilia micrantha Benth. Seed, 2G0 sp. Seed, 267 Tricliohena tonsa Neea. Seed, 200 Tiiehoiienia bulbocodium Ker. J3ulb, 256 'rriloliuni aliiinum Linn. Hoot-stoek, 242 badium Sclirb. Hoot-stock, 242 montanuni Linn. Root-stock, 242 Triglochin barrelieri Lois. Seed, 206 Root-stock, 215, 209 maritimum Linn. Root-stock, 234 Trillium erythrocarpum Ilort. Root-stock. (See T. grandiflorum.) grandiflorum Salisb. Root-stock, 618 ovatum Pursh. Root-stock, 619 rhomboideum Mchx. Root-stock, 215 sessile var. calif ornicum Wats. Root-stock, 620 var. giganteum Torr. Root-stock. (See T. sessile var. californicum.) Tiiodia decumbens Beauv. Seed, 281 Triphasia aurantiola Lour. Seed, 266 Tripsacum hermaphroditum Linn. f. Seed, 200 Trisetum argenteum R. S. Seed, 280 neglectum Willd. Seed, 280 Tristachya barbata Nees. Seed, 280 Triteleia iniiflora Linn. Bulb, 608 Triticuni. Seed, 46, 55, 50, 58, 59, 65, 09, 72, 166, 167, 168, 169, 170, 171, 172, 174, 176, 177, 178, 180, 185, 186, 187, 189, 191, 193, 194 Triticum amyleum Ser. Seed, 205 cristatum Sclirb. Seed, 205 dicoccum Sclirnk. Seed, 205 inonococcum Linn. Seed, 205 rigidum Schrd. Seed, 205 sativTim var. dicoccum ITort. Seed, 362 vulgareHort. Seed, 181,184,364 turgidum Linn. Seed, 205 Tritouia crocata Ker-Gawl. Bulb, 726 var. lilacina Hort. Bulb, 727 rosea Hort. Bulb, 729 crocosm^flora Lemoine. Bulb, 733 pottsii Benth. Bulb, 732 securigera Ker-Gawl, 730 Triumfetta schiuiperi Hchst. Root, 262 Tuda!a picta Dcsn. Root-stock, 227 regelii Heer. Root-stock, 227 Tulipa, 65, 188 australis Linn. Bulb, 537 billietiana Jord. and Four. Bulb, 528 clusiana Vent. Bulb, 532 var. persica Hort. Bulb, 533 didieri Jord. Bulb, 529 var. fransoniana Ilort. Bulb, 531 mauriana Jord. Bulb, 530 florentina Hort. Bulb. (See T. sylvestris.) var. ordorata Hort. Bulb. (See T. sylvestris.) gesneriana Linn. Bulb, 222 greigi Regel. Bulb, 527 hageri Held. Bulb, 524 OGulus-solis St. Aman. Bulb, 534 prajcox Tenore. Bulb, 535 sylvestris Linn. Bulb, 222, 526 Tunica saxifraga Koch. Seed, 292 Tyi)ha minima lloppe. Root-stock, 270 teuuifolia H. B. Seed, 286 Ullucus tuberosus Loz. Tuber, 241 Ulinus, 72 Umbilicus pendulinus DC. Root-stock, 217 Uniola latifolia Mchx. Seed, 201 Urachne parviflora Tiin. Seed, 276 Uralepis aristulata Nutt. Seed, 281 Urochloa depre.ssa Steud. Seed, 276 Vaccaria vulgaris Host. Seed, 292 Valeriana oHicinalis Linn. Root-stock, 72, 258 saliunca All. Root-stock, 237, 258 tuberosa Linn. Root-stock, 237, 258 Vallisneria spiralis Linn. Root-stock, 230 Vallota pm-purea Herb. Bulb, 633 Vanilla planifoha, 31 Vatica robusta Steud. Seed, 250 Vaucheria tuberosa A. Br. Tuber, 232 Veratrum album Linn. Root-stock, 234, 269 viride. Root-stock, 72 Verbascum sohraderi Mey. Placenta, 249 Seed, 249 Veronica austriaca. Root, 238, 259 chamoedrys Linn. Pollen, 243 Viburnum opulus, 71 prunifolium, 71 Vicia calcarata Dsf. Seed, 211 faba Linn. Seed, 69, 211, 381 fulgcns Batt. Seed, 382 gerardi Vill. Seed, 384 narbonensis. Seed, 42 sativa Linn. Seed, 211, 378 villosa Roth. Seed, 380 Vigna glabra Sav. Seed, 212 Vilfa coromandelina Beauv. Seed, 277 pungens Beauv. Stolon, 255 Vinca minor Linn. Stolon, 215 Root, 258 Viola cornuta Linn. Pollen, 243 cucullata Ait. Root-stock, 261 palustris Linn. Root-stock, 240, 261 pinnata Linn. Root-stock, 261 Viscum album Linn. Seed, 219 Visenia tomentosa R. P. Seed, 220 Vulpia delicatula Lk. Seed, 284, 285 membranacea Lk. Seed, 284 Wachendorfia hirsuta Thbg. Seed, 264 Waltheria indica Linn. Seed, 220 Watsonia alba Hort. Bulb. (See W. iridifolia var. o'brieni.) ardernei Hort. Bulb. (See W. iridifolia var. o'brieni.) humilis Mill. Bulb, 721 iridifolia var. alba W. Rob. (See W. iridi- folia var. o'brieni.) iridifolia var. o'brieni N. E. Br. Bulb, 722 meriana Mill. Bulb, 723 var. alba Hort. Bulb. (See W. iridifolia var. o'brieni.) o'brieni Mast. Bulb. (See W. iridifolia var. o'brieni.) Willdenowia teres Thbg. Root, 256 Wistaria chinensis Linn. Seed, 413 Wulfenia carinthiaca Jacq. Root-stock, 238, 259 Xanthoxylum, 72 Xyris operculata Lab. Seed, 286 Bemifuscata. Seed, 286 Zamia integrifolia Ait. Root-stock, 815, 898 Zannichellia pedicellata Fr. Seed, 209 Zantedeschia ajthiopica Spr. Seed, 286 Zea mays Linn. Seed, 31, 39, 40, 41, 55, 56, 65, 69, 72, 166, 167, 168, 169, 170, 171, 172, 174, 176, 177, 178, ISO, 186, 187, 189, 190, 191, 192, 193, 194, 199 var. everta (Golden (Jueen) Hort. Seed, 343 (While rice) Ilort. Seed, 344 indentata (Karly Learning) Hort. Seed, 348 (Hickory King) Hort. Seed, 349 indurata(Compton'BEarly) Hort. Seed, 347 131 INDEX OF STARCHES. Zea mays var. indurata (North Dakota) Hort. Seed, 346 saccharata (Black Mexican) Hort. Seed, 351 (Golden Bantam) Hort. Seed, 352 (Stowell's Evergreen) Hort. Seed, 350 Zephyranthes Candida Herb. Bulb, 639 rosea Lindl. Bulb, 640 Zingiberaceae, 65, 176, 779-795 Zingiber officinale Rose. Seed, 223, 253 [14] Zingiber officinale Hort. Root-stock, 70, 72 var. cochin Hort. Root-stock, 784 Jamaica Hort. Root-stock, 781, 782 Zizania aquatica Linn. Seed, 275 clavulosa Mchx. Seed, 275 Zornia angustifoUa Sm. Root, 263 Zostera marina Linn. Seed, 209 nana Roth. Root-stock, 215 Zoysia tenuifolia Willd. Seed, 202 Zygnema cruciatum Ag. Seed, 293 Zygnomaceae, 293 INDEX OF AUTHORS. Abderhalden and Rona, 153 Albertoni, 9 Appiana (with Menozzi), 9 Arabel, 110 Archbold, 167 Armstrong, 149, 151 Armstrong and Horton, 98 Asboth, 107 Auld (with Henry), 149 Bach, 157 Bayer, 157 Baker, 135, 147 Baker (with Ling), 117, 130, 131, 145, 146 Baranetzky, 178, 181, 189 Bau, 144 Baudrimont, 91 Bechamp, 89, 104, 109, 121 Bellmas, 176 Belzung, 40 Berg, 223, 226, 227, 255 Berge, 99, 106, 128, 131 Bernard, 63, 128, 170 Berthelot, 158 Berthelot and Gaudechon, 158 Berthold, 63 Berzelius, 85, 156 Bevan (with Cross), 6 Beyerinck, 63 Bial, 128 Binz, 43 Biot, 175 Biot and Persoz, 85, 93, 120 Bioz, 88 Bischoff, 230 Bloemendal, 53, 170 Blondeau, 169 Blondeau de CaroUea, 88, 121 Blumer, 105 Bokomy, 159 Bokomy (with Loew), 159 Boubier, 54 Boucliardt and Dandraa, 88 Bouillac and Giustiaua, 159 Bourquelot, 98, 141 Bourquelot (with H^rriay), 98 Brion, 8 Brown and Heron, 17, 41, 98, 102, 103, 111, 123, 139, 141, 178 Brown and Millar, 11, 134, 147, 152 Brown and Morris, 64, 94, 105, 115, 124, 126, 141, 142, 145, 146 Brown, Morris, and Miliar, 13, 145, 193 Brucke, 84, 93, 94, 122 Bruckner, 39 Brunner, 87, 140 Buchner (with Meisenheimer), 98 Billow, 117, 130, 137 Bumcke and Wolfenstein, 6 Bums (with Golenkin), 63 Buscalioni, 41 Butlerow, 157 Butschli, 45, 46, 79 Butyozin, 190 Castaro, 120 Caventeau, 18, 85 Clodounsky and Sulc, 131, 145 Cohen and Jahn, 157 Conrad (with Werner), 9 Convisart (with Nifipce Saint-Victor), 90 Coombes, 112 Cremer, 63 Cioss, 105 Cross and Bevan, 6 Cruger, 22, 89, 230, 231, 259 Cuisenier, 141 Cashing, 9 Czepeck, 159 Dafert, 175 Dandras (with Bouchardt), 88 Darbishire, 73 Day, 58, 166, ISO, 189, 190, 192 Dean (with Henderson), 153 Defren (with Rolfe), 16, 152 Delffs, 27, 91 Demoussy, 169 Denniston, 56, 64, 296 De Saussaure, 18, 84, 140 De Vries, 40 Dierssen, 148, 152 Dioscorides, 18 Dodel, 43 Dollfus and Scheurer, 107 Dragendorff, 28, 92 Dubosc, 167, 171 Dubrunfaut, 10, 84, 138, 177 Duclaux, 98 Dufour, 62 Dull (with Lintner), 98, 100, 102, 116, 129, 130, 143, 144, 151, 152 Duroy, 91 Eberdt, 42 Efifront, 126, 142, 150, 152, 168, 179, 192 Ehrlich (with Einborn), 9 Einborn and Ehrlich, 9 Emmerhng, 149 Emslander and Freulich, 167 Erlenmeyer, 157 Errata, 63 Euler, 158, 159 Ewart, 62 Famintzin, 76 Fehling, 88, 89, 121 Fernbach, 168, 179, 180 Fembach and Hubert, 168 Fernbach and Wolff, 103, 107, HI, 138, 149, 192 Fenton, 158 Findlay, 83 Fischer, 4, 12, 51, 55, 79, 96, 98, 142, 158, 167 Fischer (with Lindner), 98 Flourens, 127, 142 Fluckiger, 28, 91, 813 Ford, 103, 118, 168, 169, 192 Ford and Guthrie;, 148, 170, 193 Fouard, 6, 83, 103, 104 Franz, 108 Franz (with Zulkowflki), 127 [15] INDEX OF AUTHORS. Fresinius, 90, 91 Fritzsche, 19, 04, 86, 87, 102, 111, 224, 226, 230, 243 Furstenbeig, 121 Gastine, 58, 176 Gatin-Gruzewska and Maquenne, 106 Gaudechon (with Berthelot), 158 Geduld, 142, 143 Geinsbergen, 6 Geronamous (with Rolfe), 148, 152 Gibson and Titherly, 159 Gniehu, 3 Goldsehniidt, 76 Golenkin and Burns, 63 Grafe, 159 Grafe and Viesor, 159 Green, 103, 171 Gregory, 72 Grierson, 191 Griessmayer, 41, 84, 93, 94, 115, 121, 142 Gris, 185 Gruber (with Musculus), 94, 95, 114, 139 Grtiss, 146 Gruters, 119, 148, 153 Gu6rin-Varry, 18, 19, 86, 177 Guibourt, 18, 85, 101 Guistiana (with Bouillac), 159 Guthrie (with Ford,) 148, 170, 193 Guye and Walden, 5 Haddock (with Rolfe), 148, 152 Hale, 118, 135 Halske (with Siemens), 106 Hamburger, 145 Hammarsten, 178, 190 Hanausek, 180, 813, 814 Hanofsky, 127 Hansen, 63, 79 Hansen (with Hem-iques), 153 Harting, 23, 229 Hartwig, 106 Harz, 57, 119, 137, 174 Hasse (with Windisch), 99 Hefelmann and Schuiitz-Diimont, 132 Henderson, 151 Henderson and Dean, 153 Heuriques and Hansen, 153 Henry and Auld, 149 Henaen, 128 Heron (with Brown), 17, 41, 98, 102, 103, 111, 139, 141 Heruaaey, 64 Herzfeld, 114, 124 Hiepe, 116, 130, 144 Hill, 149, 156 Hoffmann and Philippe, 167 Hofmeister, 96 Horton (with Armstrong), 98 Hubert (with Fernbach), 168 Hughes, 813 Ihl, 172 Ishizuka, 8 Jacobson, 98 Jacquelin, 87, 114, 140 Jahn (with Cohen), 157 Jalowetz, 144, 145 Jessen, 26, 28, 29, 90, 92, 93 Johnson, 146, 150, 151 Kabsch, 28, 92, 310 Kahlenberg and True, 8 Kahlinowsky, 88, 121 Kantorowirz, 1 10 Kattein (with Hoilcwald), 46, 109, 111 Kemper, 91, 92 Kimpelin, 159 Kinhoff, 18, S4, 93, 138, 191 Klebs, 63 Knappe, 6 Knop, 28, 91 [16] Konig, 166 Konig, Siiieckmann, and Olig, 158 Kouigsberger, 43 Kopke (with StoUe), 111 Krabbe, 186 &aemer, 55, 71, 172 Kraut, 91 Kuhnemann, 146 Kiilz and Vogel, 144 Kiitzing, 20 Lagerheim, 171 Lake, 105 Lauga, 124 Laui'ent, 63 Le Bel, 3 Leeuwenhoek, 18, 64, 84 Leitner, 108, 127 Lenz, 174 Leuchs, 85 Levberg, 190 Liebig, 88 Lindet, 192 Ling, 118, 137, 146, 147, 168, 177, 179, 192 Ling and Baker, 117, 130, 131, 145, 146 Lintner, 104, 131, 141, 142, 143, 144, 175, 178 Lintnerand Diill, 98, 100, 102, 116, 129, 130, 143, 144, 151, 152 Lippmann, 91, 175 Lob, 158 Loew, 158 Loew and Bokorny, 159 Loewi, 153 Low, 93 Magendie, 128 Maly, 157 Maquenne, 102, 112, 113, 178, 180 Maquenne and Roux, 46, 58, 99, 112, 149 Maquenne, Fernbach, and Wolff, 111 Maquenne (with Gatin-Gruzewska), 106 Marker (with Schulze), 93 Martens, 176 Mathieu de Donibasle, 191 Mayen, 20, 64 Mayer, 8, 9, 158 Mayer (with Neuborg), 8 Meisenheimer, 98 Melsens^ 23, 90 Menozzi and Appiana, 9 Meyer, 17, 35, 36, 37, 41, 47, 67, 77, 102, 115, 117, 130, 166, 169, 172, 179 Miahle, 88 Miescher, 7 Mikosch, 40, 78 Millar (with Brown), 118, 134, 142 Millar (with Brown and Morris), 13, 145 Mittelmeier (with Scheibler), 127, 142 Mohi-, 76 Moore and Roaf, 97 Moreau, 119, 136, 137, 148 Mori, 159 Morris, 144 Morris (with Brown), 64, 94, 105, 115, 124, 126, 141, 145, 146, 147, 152 Morris (with Brown and Millar), 13, 145 Morris and Wells, 116 Mulder, 20, 87, 90, 121, 150 Munche, 144 Mtinter, 20, 71, 224, 257 Musculus, 30, 91, 92, 93, 97, 98, 114, 121 Musculus and Gruber, 94, 95, 114, 122, 139 Muter, 71, 176 Nagano, 8 Nageli, C, 1, 23, 34, 60, 06, 75, 77, 78, 90, 92, 814, 815, 167, 177, 185, 197, 814, 815 Nageli, W., 30, 84, 115, 122, 170 Nasse, 39, 121, 122, 139 INDEX OF AUTHORS. Nastukoff, 6 Neuberg and Mayrr, S Neuborg and \N oldgeinuth, 8 Neumann (with Parow), 170 Neunieister, 95 Newconibe, 04 Niepce de Saint-Victor and Convisart, 90 Nossian, 91, 167 Noyes, 137, 148 Olig (with Kouig and Spieckmanii), 158 Ost, 117, 130, 145 Ostwald, 79, 90 O'SulUvan, 84, 93, 94, 99, 103, 122, 124, 13S, 140, 148, 178, 190 Ott, 175 Pariera, 813, 814, 815 Parow, 170 Pai'ow and Neiunann, 170 Pasteur, 3, 7, 11 Pauli, 90, 97 Pavy, 89 Pavy and Bywaters, 63 Payen, 64, 87, 90, 91, 92, 121, 107, 199, 211, 213, 217, 222, 223, 227, 228, 235, 239, 240, 241, 243, 255, 259 Payen and Persoz, 19, 85, 88, 93, 97, 120 Pellet 91 Persoz (with Biot), 85, 93 Persoz (with Payen), 19, 85, 88, 93, 97 Petit, 99, 102, 131, 134, 135, 138, 140, 151 Pfeffer, 53 Pliilip, 93 PhiUppe (with Hoffmann), 107 Piutti, 9 Plander and Ravenna, 159 Poggendorff, 17, 84, 87 Pohl, 8, 91 PoUacci, 157, 159 Potter, 40 Pottevin, 99, 102, 118, 132, 133, 134, 140, 147, 177, 179 Poulsson, 9 Pregl, 108, 135 Priestly (with Usher), 157, 158, 159 Pringsheim, 63 Prior, 117, 130, 144 Puriewitsch, 46 Raspail, 18, 84, 85, 223 Ravenna (with Plander), 159 Reichard, 109, 120, 138 Reichert, 156, 177, 194 Reinke, 143, 159 Reinsch, 23, 90 Reissek, 20, 88 Renard, 158 Reychler, 191 Reymond, 137 Rhcinfeld, 119 Roaf (witii Moore), 97 Robert, 51 Rodenwald and Kattein, 46, 79, 109, 111 Roessing, 149 Rohmann, 128 Rolfe and Defren, 151, 152 Rolfe and Geronaraoua, 148, 152 Rolfe and Haddock, 148, 152 Rona (with Abderhalden), 153 Rosanoff, 5 Roux, 46, 112, 149, 192 Roux (with Maquenne), 40, 48, 99, 112, 149 Roux (with Wolfl), lOS Saare, 109 Sacharow, 95, 180 Sachs, 28, 29, 185 Sachsse, 138 Salomon, 111, 115, 124, 140, 160, 167, 190 Schacht, 230 Schardinger, 112 Scheibler and Mittclmeier, 127, 142 Scheurer (with Dellfus), 107 Schiff, 128 Schifferer, 110, 129, 143 Schimper, 30, 31, 33, 63, 70 Schleiden, 20, 21, 64, 84, 88, 222, 223, 224, 226, 231, 255, 257 Schmerber, 107 Sclimidt and Tiemann, 9 Schmitz, 33, 63 Schmitz-Dumont (with Hefelmann), 132 Schonbein, 91 Schryer, 159 Schulze, 64 88, 124, 140 Schulze and Marker, 93 Schumann, 99, 126 SchUtzenberger, 93 Schwann, 85 Schwarz, 88 Schwarzer, 93 Seegen, 139 Shubert, 39 Sieben, 141 Siemens and Halske, 106 Siemens and Witt, 107 Skraup, 83, 106, 153 Soubeiran, 213, 223, 224, 241, 256, 259, 288 Southy, 123 Soxlet, 140, 150, 167, 169 Spieckmann (with Konig and Olig), 158 Sqmre, 123 Stolle and Kopke, 111 Stone, 191 Strasbirrger, 37, 53 St. Jentys, 46, 58 Sulc (with CWodounsky), 131, 145 Syniewski, 51, 90, 102, 109, 110, 118, 132, 135, 136, 146, 147, 151 Tanret, 106, 138 Tebb, 153 Thompson, 108 Tiemann (with Schmidt), 9 Timberlake, 54 Titherly (with Gibson), 159 ToUens, 109, 143 Tr(Sboux, 159 Trommer, 88 True (with Kahlenberg), 8 Ulhk, 107 Uh-ich, 145 Linger, 20 Usher and Priestly, 157, 158, 159 Van Bosse, 64 Van Laer, 135, 152 Van't HolT, 3, 4, 5, 156 Van Tieghem, 63 Vaquolin, 18, 84, 120 Vieser (with Grafe), 159 ViUiers, 128, 143 Vines, 34 Virneisel, 105 Vogel, 120 Vogel (with Kulz), 144 Von Allihn, 140 Von Hohnel, 813 Von Lang, 70 Von Mering, 139, 141 Von Mohl, 27, 84, 90 Von Wittich, 128 Wacker, 138 Widden (with Guye), 5 Walla ton, 3 Walpers, 21, 223, 256, 257 Weber van Bosse, 04 NN'ederhake, 62 :i7i INDEX OF AUTHORS. Weisner (with Weiss), 92 Weiss and Weisner, 92 Weldon, 73 Wells (with Morris), 116 Welwart, 106 Werner and Conrad, 9 Whymper, 175 Wiclce, 27, 90 Wickstrom, 813 Wiesner, 814 Windisch and Hasse, 99 Winton, 72 WisUcenus, 3 Witt (with Siemens), 107 Wolff, 90, 100 Wolff and Houx, 108 Wolff (with Fernbach), 103, 107, 111, 138, 149 Wohler, 5 Wohlgemuth (with Neuberg), 8 Wolfenstein (with Bumcke), 6 Wortmaun, 186, 189 Wotherspoon, 105 Wroblewski, 109, 110, 117 Young, 131 Zimmermann, 42, 62 Zopf, 63, 64 Zulkowski, 99, 108, 127 Zulkowski and Franz, 127 [18] INDEX OF PHOTOMICROGRAPHS. (Magnititation 300 diameters.) Aconitum napellus, pi. 95, figs. 567 and 568. Actsea alba, pi. 95, fig. 569. spicata var. rubra, pi. 95, fig. 570. Adonis amurensis, pi. 96, figs. 575 and 576. iEsculus hippocastanum, pi. 13, figs. 75 and 76. Alocasia putzeysi, pi. 15, fig. 89. Alstroemeria aurantiaca (aurea), pi. 57, figs. 341 and 342. brasiliensia, pi. 57, figs. 339 and 340 ligtu, pi. 57, figs. 337 and 338. Amaryllis belladonna major, pi. 51, figs. 303 and 304. Amorphophallus rivieri, pi. 15, fig. 90. Andropogon sorghum var., pi. 1, figs. 3 and 4. Anemone apennina, pi. 94, figs. 561 and 562. blanda, pi. 94, fig. 564. fulgens, pi. 94, fig. 563. japonica, pi. 95, figs. 565 and 566. Antholyza crocosmoidea, pi. 73, figs. 437 and 438. paniculata, pi. 74, figs. 439 and 440. Arachis hypogoea, pi. 10, figs. 57 and 58. Arisaema triphyllum, pi. 14, figs. 83 and 84. Arrhenatherum elatius var., pi. 3, figs. 17 and 18. Arum comutum, pi. 14, figs. 79 and 80. italicum, pi. 14, figs. 81 and 82. palsstinum, pi. 13, figs. 77 and 78. Avena aativa var., pi. 3, figs. 15 and 16. Babiana var. (Athraction), pi. 78, figs. 467 and 468. (Violaeea), pi. 78, figs. 465 and 466. Batatas edulia, pi. 100, figs. 597 and 598. Brodisa californica, pi. 47, figs. 279 and 280. Candida, pi. 45, figs. 269 and 270. capitata, pi. 48, figs. 285 and 286. coccinea, pi. 46, figs. 275 and 276. congesta, pi. 48, figs. 287 and 288. grandiflora, pi. 47, figs. 277 and 278. ixioidea var. splendens, pi. 45, figs. 267 and 268. lactea, pi. 46, figs. 271 and 272. laxa. pi. 47, figs. 273 and 274. peduncularis, pi. 45, figs. 265 and 266. purdyi, pi. 47, figs. 281 and 282. stellaris, pi. 48, figs. 283 and 284. Calathea lietzii, pi. 90, figs. 535 and 536. vandenheckei, pi. 91, figs. 541 and 542. vittata. pi. 90, figs. 537 and 538. wiotiana, pi. 90, figs. 539 and .540. Calochortus albus, pi. 28, figa. 163 and 164. benthami, pi. 28, figs. 167 and 168. howellii, pi. 29, figa. 173 and 174. leichtlinii, pi. 30, figs. 175 and 176. lilacinus, pi. 29, figs. 169 and 170. luteus var. oculatus, pi. 30, figs. 177 and 178. maweanus var. major, pi. 28, figs. 165, 166. nitidua, pi. 29, figa. 171 and 172. aplendena, pi. 30, figa. 179 and 180. Canna edulis, pi. 83, figa. 497 and 498. mussefolia, pi. 83, figs. 495 and 496. roscoeana, pi. 83, figs. 493 and 494. warszewiczii, pi. 82, figs. 491 and 492. var. (J. D. Eisele), pi. 85, figs. 509 and 510. (Jean Tissot), i)l. 85, fig.s. 507 and 508. (Konigen Charlotte), ])1. 84, figs. 499 and 500. (L. E. Bally), pi. 84, figs. 503 and 504. (Mrs. Kate Grev), pi. 85, figs. 505 and ,506. (President Carnot), pi. 84, figs. 501 and 502. Castanea americana, pi. 12, figs. 69 and 70. pumila, pi. 13, figs 73 and 74. sativa var. numbo, pi. 12, figs. 71 and 72. Chionodoxa lucilise, pi. 36, figs. 211 and 212. aardensia, pi. 36, figs. 215 and 216. tmolusi, pi. 36. figs. 213 and 214. Cimicifuga racemosa, pi. 96, figs. 571 and 572. Cochlearia armoracia, pi. 97, figs. 581 and 582. Colchicimi parkinsoni, pi. 51, figs. 301 and 302. Convallaria majalis, pi. 50, figa. 295 and 296. Crinum americanum, pi. 53, figa. 315 and 316. fimbriatulum, pi. 53, figa. 313 and 314. Crocus var. (Baron von Brunow), pi. 75, figa. 445, 446. versicolor, pi. 74, figa. 443 and 444. Busianus, pi. 74, figa. 441 and 442. Curcuma longa, pi. 82, figa. 487 and 488. petiolata, pi. 82, figs. 489 and 490. Cycas circinalis, pi. 101, figs. 605 and 606. revoluta, pi. 101, figs. 603 and 604. Cyclamen cilicum, pi. 99, figs. 593 and 594. coum, pi. 99, figs. 591 and 592. repandum, pi. 99, figs. 589 and 590. Cypella herberti, pi. 75. figs. 449 and 450. Dieffenbachia illustria (cortex), pi. 19, figs. Ill and 112. (pith), pi. 19. figs. 109 and 110. D. segiiine var. irrorata (cortex), pi. 18, figs. 107, 108. (pith), pi. 18, figa. 105, 106. maculata (cortex), pi. 18, figa. 103, 104. (pith), pi. 17, figs 101, 102. nobilia (cortex), pi. 17, figs. 99. 100. (pith), pi. 17, figs. 97, 98. Dioon ed\ile, pi. 102, figs. 607 and 608. Dolichos lablab, pi. 6, figs. 35 and 36. Dracunculus vulgaris, pi. 15, figa. 85 and 86. Eranthia hyemalia, pi. 96, figs. 573 and 574. Erythronium americanum, pi. 39, figs. 233 and 234. califomicum, pi. 40, figs 239 and 240. citrinum, pi. 40, figs. 237 and 238. dens-canis, pi. 39, figs. 229 and 230. var. grandiflora, pi. 39, figa. 231 and 232. Freesia refracta var. alba, pi. 73, figa. 433 and 434. leichtlinii, pi. 73, figa. 435 and 436. Fritillaria armena, pi. 20, figs. 153 and 154. aurea, pi. 26, figs 151 and 152. imperialis var. aurora, pi. 26, figs. 155 and 156. liliacea, pi. 27, figs. 157 and 158. meleagris, pi. 25, figs. 145 and 146. persica, pi. 27, figa. 161 and 102. pudica, pi. 25, figa. 149 and 150. pyrenaica, pi. 25, figs. 147, 148. recurva, pi. 27, figa. 159 and 160. GalanthuB elweaii, pi. 56, figa. 335 and 336. nivalis, pi. 56, figs. 333 and 334. Galtonia candicnns, pi. 42, figa. 247 and 248. Geleaine azurea, pi. 76, fig.s. 453 and 454. Gesneria tubiflora, pi. 100, figs. 599 and 600. Gla 2") and 2<). Virid fuli/rns. 27 and 2S. I'lriV/ (icmnti. 2!) anil ;{(). M iicitiiii iiniririis PLATE 6 ^cr^^ & .^:*:- V-:-. ■>•••,■;'-.,•, . m^ L M « ■ ::^ ■'^:^' Vi--v<^> 36 31 anil 32. I'/iiixi'nlus rnlijaris var. 33 and 34. I'liiisrahis liiiidliis var. 3r> and 3(1. Ihilichns tiililali. PLATE 7 %>• V "ii: - . / • • ' 1 • ' ' »> X# 37 and 38. Lcj/s exculcntn var. 39 and 40. Latln/rux ntlnrnlun var. shnhziida. 41 and 42. Liithyrux ayhteslris. PLATE 8 ^"-1% 43 and 44. Lalln/rus latifolius var. albiis. 4o and 46. Lalhijrus magellanicus var. nlhus. 47 and 48. Pisum sativum var. (Eugenie). PLATE 9 ^ Mi -.*4'' ^"^' •*:— ' o MM*— Vv-: Jr >.r 34 49 and 50. Pisum sativum var. (Thomas Laxton). 51 and 52. Pisum sativum var. (Electric Extra Early). 53 and 54. Pisum sativum var. (Mammoth Grey Seeded). PLATE 10 oo and of). FHsurn sativum var. (Large ^\■hitL' Marrowfat.). 57 and 58. Arachis hypogcea var. 59 and 60. Polygonum fagopyrum var. PLATE 11 it. - .i^'^'^r^^ ^'*-ir>.**'v; 66 01 and ()2. Qiiemis nlba. (hi anil t'i4. Qiicrcnti miihlcnheryi. ti') and (iti. Qiiirciit' priiiux. A 0 '$% S) • % ^3 L^#& ■ / w '^^--C'^r'^* fL^. PLATE 12 IS* ^ ^"Wc^nC/ 71 ^ ^G o o 72 67. Qnercus rubra. ()8. Quvrcus texana. 69 and 70. Caslanea americana. 71 and 72. Castanea satim var. iiiimbo. PLATE 13 5 -^^^T' o:. ^^W %> V • 78 73 and 74. Castanea pumila. 7') and 7(). .^sculus hippocaslanum. 77 and 7S. Arum pala-xliiiiiin. PLATE 14 mt^i^m 80 t 79 and SO. Arum cornulum. 81 and 82. Arum ilalicum. 83 and 84. Ariswma triphyllum. PLATE 15 6 " > * A* 85 and 86. Drarunculus vulgaris. 87 and 88. PcUaiidra undulata. 89. Atocasia putzcysi. 90. Amorphophallus rivicri. PLATE 16 J .'^O ^^^^^y^iSk^m .0^ -^.o? p :>*%^. ,-j5l?LSiS*',^*^-%«'i iSi-.^^ 93 '""C 'S& 'i^ V v>5i»^ 91 and 92. Richardia elliotiana. 93. Richardia africana. 94. Richardia solfatarre. 95 and 96. Richardia albo-maculata. PLATE 17 97 and 98. Dieffenbachia seguine var. iwbilix (pith). 99 and 100. Dieffenbachia seguine var. nobilis (cortex). 101 and 102. Dieffenbachia seguine var. maculata (pith). PLATE 18 103 and 104. Dieffenhachia seguitie var. niaculiita (cortex). 10.5 and 106. Diiffctihachia seguine var. irroraln (pith). 107 and 108. Dieffenhachia seguine var. irrorala (cortex). PLATE 19 im :iM(l 111). Diiffcnhachin illiislris (pith). Ill and 112. Diijfdiliiirliiii illi(slris (cortex). l\'-i and 114. Etloci of chloral hydrate-iodine on the starcli of Diiffoilmchia seguine var. maculata (pith), partial reaction. PLATE 20 115 and IIG. Liliuni candulum. 117 uiul lis. Lilium longijlorum var. giganleum. 119 and 120. Lilium longiflorum var. eximium. PLATE 21 3RS>> 121 and 122. Litium parri/i. 123 and 124. Lilium rubellum. 125 and 126. LUium. philadelphicum. PLATE 22 127 anfl 128. Lilium tigrinnm var. xplcndens. 129 and 130. Lilium heiirifi. 131 and 132. Lilium auralnm. PLATE 23 133 aiul 134. Lilium speciosnm var. album. 135 and 136. Lilium martagon. 137 and 138. Lilium superbum. PLATE 24 139 anil 140. Liliuni tenuifoliiim. 141 and 142. Lilium pardalmum. 143 and 144. Lilium puberulum. PLATE 25 lOri 145 and 146. Frililhria mekoijris. 147 and 148. Fnlillaria pyrenaica. 149 and 150. Fritillaria pudica. PLATE 26 151 and 152. Frilillaria aurea. 153 and 154. Frililtaria armciid. 155 and 156. Frilillaria imj>criiili--< \'ar. uitrora. PLATE 27 157 and 158. Frilillaria liliacea. 159 and KiO. Frilillaria recumi. 161 anrl 162. Frilillaria perxica. PLATE 28 1(1:5 and 104. Calnchnrlun alhii.'^. K').") and ItJO. ('iriiriiiiifi. 209 and 211). SriUn hifolia. PLATE 36 § 5* 211 and 212. Chionmlnxa lucilicp. 213 and 214. Chionoduxa tmolusi. 21.5 and 21(). Chioiioddxa xardetiais. PLATE 37 217 unil '21S. I'lisrhlHiiia scilloities. 21i) and 220. I'lixrhkiiiin xcUlniilfx • 221 and 222. Oniillioijiilum iiultuis. lihitnolirn. PLATE 38 223 and 224. Ornithngnlum umhcllnliim. 225 and 226. OmUhognhim narbonaisc {jujramidak). 227 and 228. Ornilhognlum Ihyrsoides var. aureum. PLATE 39 220 ;iiul 2:iO. Eri/lhrnnium (Irrix-riuiif;. 2'.',\ and 2:i2. Ernthroiiiuin ihnx-cdiiis vur. (jrai;diflorum. 233 aud 231. Enjthroiduin amcricanuin. PLATE 40 238 239 ^ 235 ami 2:56. Erylhronium ynindijlorum. 2:57 and 2;i8. ETylhronium citrinum. 239 antl 240. Erylhrurdum californirum. PLATE 41 241 anil 'J42. Hi/nriiilhiin orietitalis var. iiUm aiijHTlnxsima. 2-i.'5 and 244, II ijaciidhus oricrilalis var. albulus (white). 24.5 aud 240. Hyacinlhus oritntalis var. albulus (Italian) PLATE 42 "^CJ Q *% ^If' 250 247 and 248. (lallouia candicans. 249 and 250. Mii.'^mri bolryoides. 251 and 252. Muscari paraduxum. PLATE 43 ^ — i » • J — - 253 and 254. M iixcari mirranlhum. 255 and 'J5t'). Miisairi roniniin. 257 and 25S. .l/M.scort commulalum. PLATE 44 259 ami 260. Muxcnri rnrcmosum. 201 anil 2t>2. Muxrari r(iiiii>aclii)H. 2t):5 ami 2(14. Miisciiri rnninxinil. PLATE 45 >S.Vi 266 2tl.") and 2l")ri. Brniliirn pedunriihiriit. '2li7 iuiil 2(iS. liriiiliiiii ixioiili'^ var. xploiiliiis 2l)'.l and 271). Br and 27)1. lirinliiin mrciitcii. PLATE 47 ^"^ £^ /**v.- *•• « * 282 277 anrl 27S. Broiliira grnmlijlora. 27!) and 280. Brodia-a cnlifornica. 281 and 282. Brodiwa iiunlyi. PLATE 48 283 and 284. Brudura slcllans. 285 and 286. Broditrn capilala. 287 and 288. Brodiwa congesla. PLATE 49 '•• »!»v 290 289 and 290. Trilcleia uniflorn. 291 and 292. Lachenalia pendula. 293 and 294. Lachenalia tricolor var. luteola. PLATE 50 .'> 297 ^$t^f^' 2',to ami 296. CnnviiUdrin iiinjiilU. 2'J7 and 208. Trillium (iniiiiiiHnrum. 209. Trillium ovaium. 300. Trillium sessile var. californiciim. PLATE 51 301 and 302. Colchicum parkinsoni. 303 and 304. Amaryllis belUuiona major. 305 and 300. Vallnla purpurea. PLATE 52 ^ V 307 and 30S. Hippea.ilrurn vittatum. 309 aiiil 310. Hippeastriim equeslrc. 311 :iiiil 312. Hippenslrmn ntdiciuii var. m1>uf:tuin. PLATE 53 ^^^.,*\ ^^••f*: 313 and 314. Criuwii fimhrUilidum . 315 and 316. Criniim aiiicricaniiDi. 317 and 318. Sprekdid foriiwuissima. PLATE 54 319 and 320. Zcphyrnnihcs Candida. 321 anil 322. Zijthiiriiiithcn rosea. 323 and 324. HaiiKinllinx kdlhcriniv. PLATE 55 325 and 326. H ipnenocallis undulnta. 327 and 328. H ymenocallis calalhiiia. 320 and 330. Leucoium astumm. PLATE 56 331 and 332. Leucnium vernum. 333 and 334. GnJaiilhuf: nivalU. 335 and 336. Galanlhus elwcsii. PLATE 57 3;i7 and 338. Alstrcemeria ligta. 339 and 340. Alstrcemeria brasiliensis. 341 and 342. Alslraimeria aurantica (aurea). PLATE 68 '>\1, « ^ 4»' t 344 343 and 344. Slcrnhergia luten. 345 and 34(i. Xarcissufi horsfieldii. 347 and 348. .Varcissus maximus. PLATE 59 ^11 ^;. 354 349 and 350. Xarcissus biilbocndium. 3.51 and 352. Xarcissus bulhocodium var. cotispicua. 353 and 354. Nardssiis bulhocodium var. monophi/lln PLATE 60 ^» ' '>'-:/#i- f *^vv» ,^. r#v 358 355 and 356. Narcissus incomparahilis. 357 and 358. Narcissus odonis. 359 and 360. Narcissus poelicus. PLATE 61 301 and 362. Nnrcissns hiflorus. 363 and 304. Narci^sitH joiiquilla. 365 and 366. Narciscus jonquilla \:\v. ntgulostiK. PLATE 62 367 and 368. Xarcixsu.i jonquilln var. campernclK riifjuhsiis 369 and 370. XarcissuK tazelta var. nrientalis. 371 anil 372. Turm jiinridtifida. :li>S^- PLATE 63 3/4 373 and 374. Irix florentinn. 375 and 370. Iris pallida speciosa. 377 and 378. Iris pumila var. cyanea. f^^^ PLATE 64 3V9 and 380. Iris hisniarckiana. 381 and 382. Iris iberiai. 383 and 384. Iris xiphium var. ynind tresorier. PLATE 65 385 and 386. Irix xiphium var. wilhdmine. 387 and 388. Iris xiphium var. lusitanica. 389 and 390. Iris tingitana. PLATE 66 396 391 and 392. irU nikulala. 393 and 394. IrU hUtrio. 395 and 396. Iris alata. PLATE 67 '^^iT^ 397 and 398. Iris caucasica. 399 and 400. Monea tristis. 401 and 402. Homeria collina. PLATE 68 403 and 404. Tigridia pavonia var. graiidiflora alba. 405 and 406. Tigridia paronia var. conchiftora. 407 and 408. Gladiolus byzantinus. PLATE 69 -^^9 400 and 410. Gladiolus primtdimt.i. 411 and 412. Gladiolus cardiiiali.-< (Blusliing Bride). 413 anfl 414. GlndioluK florilinndux. iS-i PLATE 70 415 -^^^ Vj ''.?''•?> ^s^&^/i o r 4i5 and 416. Watsonia humilist. 417 and 418. Walsonia iridijolia var. o'brieni. 4 lit and 420. Watscmia meriana. PLATE 71 421 and 422. Trilonia crocata. 423 and 424. Trilonia crocnin var. Hlacina. 425 and 426. Trilonia crocata var. rosea. PLATE 72 W'M V 5? \ ■ - 430 427 and 428. Trilonia securiyera. 429 and 430. Trilonia poltsii. 4.31 and 432. Tritonia crocosmwflora . PLATE 73 3:^ € «o roo UW "? • "• ^ ?:^*^% 8>!t,, 435 "^ " Q V>^ 433 and 434. Fnrsin nfrnrld \ar. (ilha. 435 and 430. Frcc.sin ref facta viiv. hichUitiii. 437 and 438. Anlholijza crocosmoides. PLATE 74 _Qv. 441 439 and 440. Atdholijza paniculata. 441 and 442. Crocus susianus. 443 and 444. Crocus versicolor. PLATE 75 Jff^'^ 445 and 446. Crocus var. (Baron \-on Bnmow). 447 and 448. Romulea rosea var. specinm. 449 and 450. Cypdla herberli. PLATE 76 ?fo,. 451 and 452. Marica gracilis. 45:5 and 4.54. Gelasine azurea. 4.55 and 45t>. Spnrnxis grnmlijlord iilliii. PLATE 77 n ••'■:! !»!.''■ •••-V;' 4h() 457 and 458. Sparaxis var. {alberline). 459 and 460. /xmi speciosa. 461 and 462. /xia viridiflora. PLATE 78 W '^ >» ^O~,,^0_o^^^ 'Mam s ^ 463 and 464. Ixin var. (Emma). 46.') and 466. Bahinna var. (violacea). 467 and 468. liiOiinna var. (athrnciion). PLATE 79 y%««« /Ti (^ 469 and 470. Musa cavendishii (root-stalk). 471 and 472. Musa cavendishii (green fruit). 473 and 474. Musa sapicnlum. PLATE 80 s^. 47.5 and 47(j. Musa ensele. ill and 478, Zingiber officinale. 479 and 480. Zingiber officinale var. Jamaica. PLATE 81 481 and 482. Zingiber officinale var. cochin. 483 and 484. Hedychium coronarium. 485 and 486. Hedychium gardnerianuni. &^' PLATE 82 .. ^o V^. f^' 4<)0 487 and 488. Curcuma longa. 489 and 490. Curcuma petiolalu. 491 and 492. Canna warszcwiczii. PLATE 83 404 493 and 494. Canna roscoeana. 495 and 490. Canna muHa-folia. 497 and 498. Canna eduli-i. PLATE 84 400 and 500. Carina var. (Konigen Charlotte). 501 and 502. Canna var. (Prosidont Carnot). 503 and 504. Canna var. (L. E. Bally) PLATE as 505 and 506. Ctintia var. (Mrs. Kate Grey). 507 and .508. Carina var. (Jean Tissot). 509 and 510. Carina var. (J. D. Eisele). PLATE 86 511 to olli. Effects of chloral hvdrato-iodino on Canna warszemczii. PLATE 87 --— V s^ ^ 321 522 517 aiul .lis. EfTccls of rliliiial liydratc-ioeliiu; on the starch of Solanum tulierosum. 519 to 522. Effects of chroinif acid on Carina warszewiczii. PLATE 88 52.? and 524. Maranta arundinacea. 525 and 526. Maranta arundinacea var. No. 1. 527 and 528. Maranta arundinacea var. No. 2. PLATE 89 i." o 0= o t ~ o & b \ 0 © cP, •-m .% '» 529 and o30. Maranla massangeana. 531. Marnnla huconeura. 532. Maranla ijoveniana. 533 and 534. Maraida musaica. ^^1^^ T^j^. PLATE 90 #?% 1^M\ ij^i") and .")30. C'liliillicd licliei. o;J7 and .")3S. ('(itntlua mttnUi. 53'J and .540. ('iitnlltm uiiiliamt. PLATE 91 541 aiul o42. Calathea vandenheckei. 543 atifl 544. Siromdiithc saiiguinea. 545. XymphiFd alba. 546. Xymphii'a mcxicana. PLATE 92 .•■. •*>- U -r5t 548 " tt ;; ., . ,S 550 547 anil .j4.s. Xi/iii/jhea marliacea var. nlbidn. .")4'.) ami ')')(). Xi/iiipluea marliacea var. rarneii. .").")l unci "j.'j'J. Xijiiiiiliirii ijlatlsldiiiniiii. PLATE 93 i'^^S'- 4C°=^?o°c^ 553 and 55-t. Xi/ntphaa odnrata. 555 and 556. Xymphcea oitonitu var. rosea. 557 and 558. Ndumbo nurifera. PLATE 94 ,-*>«:^» - ■5, ^OS' H 561 Q oV; ^^ €U?^ oi^;. :£:i/) :^^ ^ ^ ^^' ®.;g/' • w.'^'^ Sft^ =^ '5k'^='^^-C'.A< off ^ 563 ft* "cP.**' ''.^ ..")1) and r,m. olU anil .-,02. )(j:5. Xelumbo liilca. A iieiNonc (ipenuino. A iieinonc fulgens. Ancmoiif })l(iiiila. PLATE 95 Ve "' 565 .^ ^^ fem-?* •r r- .a' oor^oo ,H^ V^/l '^.fxO Cvi cfii^' 565 and 506. Anemone japonica. 567 and 568. Aconilum napellns. 569. .-Ictea n7?)a. 570. Ackea spicala var. rubra. PLATE 96 571 V*»^ ^Ji".^ 573 X®^"^" OP, Piiii ppf-- 571 and .")72. Cimlcijuga racemoaa. i)7o :uiil r)7-l. KranlhU hyemalix. ")75 :iiiil .j7(i. .tr/o(ti6' (iinurcnsU. PLATE 97 ,OOo 577 rio /^'¥ •^•^-.^,^^/i- \^li ^'Jl * c^-*^ '/ , ./ o 581 ^ ^^ V-y ^ -3i c "(77 and 57S. lianuiiculiiJy hiilhosus. ■>''.) and ")V^o ■>^i!^ .j^' k^ ♦^N t^ ^^H.x -Nnl ^ ov ^'^ h'"*^^^^>^^-'^ H.Wk^t '^^^wil ■^sX ^ ..V^N \ . ^ x*^