STARCH-GRAIN AS A SPHEROCRYSTAL. 77 



consist of radially arranged crystalloids; that the crystallizetl starch-substance has the 

 formula CcHjoOs; that there are probably several isomers; that various conditions can be 

 citetl to ]irove that starch-crystals always appear in aggregate bundles and never singly; anti 

 that the conditions which account for the occurrence of aggregates are difficult solubility, 

 with low power of crystallization, and the viscosity of the solution in which crystallization 

 takes place, and that a.ny one of these factors may in some cases suffice. 



The study of the starch-grain as a spherocrystal was taken up by Meyer in 1881 

 (loc. cit.) and later in 1895 (loc. cit.), and he gives the name trichiles to the hypothetical 

 crystal elements named by Niigeli miceUcc. Meyer states that the hypothesis that starch- 

 grains are spherocrystals of a carbohydrate furnishes the simplest explanation of the 

 lamellated structure; of the growth of the grain by external accretion; of the later origin 

 of the outer layers; and of the relatively low density of the innermost parts of the mature 

 grains. If the grains are spherocrystals it must be assumed, he states, that they grow 

 like other crystals and show phenomena which characterize spherocrystals. If sphero- 

 crystals of a carbohydrate, such as sugar, be caused to form, and if conditions attending 

 crystallization are altered periodically by placing the preparation at a window where the 

 sun warms the solution periodically, concentric, lamellated spherocrystals are formed; 

 but if crystallization take place under constant conditions the crystals show an absence 

 of lamella?. The hilum of the artificial spherocrystals of sugar was found to be usually 

 less dense than the rest of the crystals, and that it incloses "mother substance," rarely 

 air. Application of these facts to the starch-grain shows, he states, that most starch-grains, 

 even in young stages, have a small, relatively soft center, and that since the plant is sub- 

 jected to periodic changes in external conditions the grains must be built up of layers of 

 alternating density. In examining crystals of carbohydrates he notes the occurrence of 

 all transitions between spherical-crystal groups and typical spherocrystals. If in the pro- 

 duction of crystals they are produced from impure solutions of carbohydrates, they have 

 an apparently homogeneous appearance, in which, as in most intact starch-grains, no radial 

 striations can be percei\'ed. Such spherocrystals are formed from carbohydrates ha^'ing 

 a high molecular weight, such as amylodextrin and inulin. In the production of crystals 

 of amylodextrin, as long as the solutions contain 20 per cent of amylose and dextrin, 

 spherocrystals form very readily on evaporation or congelation, or by slow cooling of a 

 concentrated solution. 



Typical spherocrystals are, writes Meyer, composed of very thin elongated needle- 

 like or thread-like crystal units, which, as before stated, he terms irichites. In all sjjhero- 

 crystals the trichites are assumed to be joined in tufts. Concentric lamellation is stated 

 to be a very constant phenomenon in typical spherocrystals. This structural peculiarity, 

 he notes, is observed in spherocrystals of minerals and in many artificial crystals of inor- 

 ganic substances, and that it has long been observed in the spherocrystals of carbohy- 

 drates, in which the lamellae may appear in the apparently structureless forms and in the 

 coarsely radial thread-like forms. The lamella? of spherites of amylodextrin are stated 

 to be produced by alternate layers of trichite tufts which vary in their crystalline structure. 

 Such spherites are described as consisting of a hilum or center with radial trichites, and 

 around the hilum layers of trichite tufts. The layers are differentiated by the varying 

 thickness of the trichites ; by the varying lengths of the trichites ; by the compact or open 

 arrangement of the trichites; and by the extent of branching of the trichite tufts. The 

 trichites are always most easily separated in a radial direction, and spherocrystals are 

 always porous. Starch-grains behave so similarly to spherocrystals of other substances 

 that Meyer holds that one is justified in designating them spherocrystals of amylose; and 

 likewise he would speak of spherocrystals of a-amylose, of /3-amylose, and of amylodextrin. 

 WTiether or not amylodextrin crystallizes with amylose into a mixed crystal, or whether 

 three kinds of trichites mix, he postpones for fm-ther investigation. 



