474 
AMERICAN JOURNAL OF BOTANY 
[Vol. 9. 
which is thin on the larger particles and thick on the smaller particles. 
They may consist of either of the two substances, granulose or cellulose, 
which Nageli considered to be the two components of starch. The denser 
layers of the grain are composed of the larger particles while the less dense 
or softer layers contain only the smaller particles with the thick water-shell. 
He supports this construction by long discussions of the forces which 
must be exerted to account for various starch reactions with water. In 
several places he said these particles are not crystals, but in 1881 (29) he 
stated that the micellae are of a crystalline nature. 
About thirty years after Nageli's theory appeared, Schimper (30) 
Undertook to determine whether the starch grain is crystalline or amorphous. 
He decided that the cohesive ^nd the optical properties would differentiate 
these two states. From his work with polarized light he concluded that 
the grains must consist of fibrous crystals arranged at right angles to the 
concentric layers; and that they differ from the ordinary spherocrystal 
through their ability to swell in water. 
In 1895, Arthur Meyer (3) in his large monograph, " Untersuchungen 
iiber die Starkekorner," accepted, to a great extent, the conclusions of 
Schimper and elaborated the spherocrystal concept' on of the starch grain. 
He used the definition (3, pp. 101-107) of the spherocrystal which Nageli 
(31) and Rosenbusch (32) proposed — that it is a more or less globular body 
composed of radially arranged needle-like crystals. Fox Talbot (33) in 
1836, Sir David Brewster (34) in 1853, Rosenbusch (32) in 1885, and 
others had studied the optical properties of globular crystal aggregates in 
which they could see the needle-like crystals arranged radially from a 
central point. The dark cross produced by polarized light seemed to be 
the same in them as that formed in other globular bodies in which the 
individual crystals were invisible. The latter, then, they reasoned, con- 
sisted likewise of invisible radially arranged crystals. In addition to this, 
they found spherical forms which presented gradual transitions from those 
with visible crystals to those in which the crystals were not visible. 
Basing the conclusion primarily on this work, Meyer (3, pp. 1 16-129) 
felt justified in assuming that the starch granule is a spherocrystal. The 
invisible needle-like crystals he called trichites. He decided that the con- 
centric layers of the grain are visible because the trichites differ in size and 
number in the different layers. By comparing the properties of inulin and 
amylodextrin spherocrystals with those of the starch grain he arrived at 
the following conclusions: 
1. In the starch grain radial lines of weakness occur just as they do in 
the spherocrystals, where they are due to radially arranged trichites. 
2. Optically the spherical grains behave exactly like the spherocrystal, 
while the eccentric forms behave as though they were built up of trichites 
placed at right angles to the concentric layers of the grain. Very small 
broken pieces also behave as though composed of trichites. 
