76 DIFFERENTIATION AND SPECIFICITY OF STARCHES. 



As the molecule grows, even if the form were originally spherical, it becomes elongated, 

 polyhedral, or prismatic. The peculiarity of form depends upon the position of the mole- 

 cule in relation to other molecules; thus the outer and inner sides, if turned towards another 

 molecule, become flattened and pointed if directed towards the interstices. The larger 

 the molecule grows the greater the deviation from the original form, both because the 

 movement of the fluid encourages the deposition of new material at the sides turned 

 towards the interstices and because here the resistance to the water envelopes is diminished. 



NiigeU gives in line di-awings schematic representations of the form and arrangement 

 of the molecules. In one is shown a hypothetical section of a superficial molecular layer 

 showing among developed molecules or micellae young and as yet spherical molecules, 

 and the molecules with their envelopes and interstitial canals shown in a tangential direc- 

 tion. In the second figure he represents a part of a hypothetical radial section showing the 

 molecules lying beside one another in horizontal rows belonging to the same molecular layer. 



The structm'al likeness of the starch-grain to spherocrystals of inorganic substances 

 was first strikingly pointed out by Famintzin (Heidelberger Jahrbiicher der Literatur, 

 1SG9, Lxii, 226), who fomid starch-like forms of crystals of calcium carbonate. He records 

 the parallelisms in the manner of growth, in the lamellation, in the differences in solubility 

 of the inner and outer parts, and in the formation of compound grains and crystals. Erosion 

 phenomena which bear a striking resemblance to those occurring in starch-grains have 

 been observed by Goldschmidt (Zeit. f. Krystallographie u. Mineral., 1904, xxxviii, 1) 

 in spherocrystals of calcium carbonate. 



Schimper {loc. cit.) supported the hypothesis of the starch-grain being a crystalline 

 body upon the basis of the cohesive and optical properties wliich distinguish amorphous 

 from crystalline structures, and he studied with some care the analogies between the starch- 

 grain and ordinary spherocrystals. He observed that the grains break more easily trans- 

 versely than parallel to the lamella?, and that such a peculiarity has not been noted in 

 amorphous bodies, because the absence of regularity of arrangement of the molecules 

 makes them split irregularly when crushed. A filamentous crystalline aggregate when 

 crushed, lie observes, separates in lines parallel to the bundles, because the force which 

 binds the latter is more easily overcome than the cohesion of the molecules of the individual 

 crystals. The striated nature of the broken faces shows, he states, how the fillers are 

 separated, and that the starch-grains exhibit this same phenomenon, thus resembling 

 radio-fibrous crystalline aggregates and differing from amorphous bocUes. The polariscopic 

 characteristics he found to agree entirely with the cohesive properties, and like them to 

 be owing to the crystalline nature of the starch-grain. Viewed in polarized light the "inter- 

 ference figure" is such as starch-grains should exhibit if they are composed of bundles of 

 crystalline uniaxial or rhomboidal elements, and would split up transversely to the lamella*. 



Schimper states that Baily (Philosophical Magazine, 187G) and also von Lang (Ann. 

 d. Physik. u. Chemie, cxxiii; Carl's Reportorium, iii) reached this conclusion; and that 

 Mohr's (Botanische Zeitung, 1858) assertion that the arms of the interference cross always 

 run perpendicular to the lamella? holds good only for a regular centric or spherical structure, 

 as they often intersect eccentric grains at a very sharp angle. In regular centric spherical 

 structures, as well as in the axes of eccentric ones, the doubly refractive elements are straight 

 and extinguish the light in its entire length; the lateral parts of eccentric grains behave 

 differently, as is shown by the bundles and the fissures, which described a curve, and which 

 according to tliis satisfy the conditions for extinguishing the light in more or less of their 

 length. These characteristics, he writes, like those of cohesion, can be attributed only to 

 the fact that starch-grains consist of crystalline bundles which run parallel to the lamella;. 

 Starch-grains, he notes, differ from ordinary spherical crystals in their property of swelling, 

 and he proposes to distinguish spherocrystals and all crystalline bodies which possess 

 this property as crystalloids. Schimper states that one may conclude that starch-granules 



