504 WALTER H. BUCHER 
concentrated solution of this salt, as was noted accidentally, 
exhibits all the characteristics of an emulsion colloid; for instance, 
a noticeable change of viscosity with change in temperature, 
stability in the presence of salts with polyvalent ions, etc. When 
dried at ordinary room temperatures, large spherocrystals are 
formed, perfectly spherical in shape when formed freely suspended 
in the heavy liquid, hemispherical in shape when formed in contact 
with the walls of the vessel or with the surface of the liquid. So 
far I did not succeed in preparing thin sections of these bodies, 
owing to the low melting-point of the salt (42° C.). In very thin 
layers, however, spread out on a slide, the growth is seen to take 
place in one plane only, yielding layers thin enough for optical 
study. This, combined with a macroscopic observation of the 
bases of the hemispheres, gave the following results: 
Spherites formed relatively rapidly during normal evaporation 
appeared structureless on cross-section with the center distinctly 
darker but not sharply separated from the rest. In other cases 
there was a sharply defined light center, and sometimes an outer 
lighter layer, distinct from the rest. Some spherites seemed to have 
formed through the union of several macroscopic drops, judging 
from an indistinct pattern shown in cross-section. When formed 
on a slide, growing in one plane only, they always exhibited a radial 
crystalline structure, the individual fibers being visible under the 
microscope. Superimposed on this radial structure there appeared 
in some of the spherites, on the same slide with the others, concentric 
lines outlining separate crystalline layers, which, however, all con- 
sisted of the same material. 
Spherites formed during extremely slow evaporation (due to a 
crust formed on the liquid) showed a concentric structure of delicate 
layers of different color, especially in their outer part, representing 
in every respect true odlites. In several cases ten and more layers 
were counted in a radius of 2 mm. 
These experiments illustrate well the principles of Schade: 
(x) the spherical shape of spherites is due to the tendency of the 
droplets, forming during the separation of the dispersed phase of an 
emulsoid, to coalesce; (2) the difference between spherites of 
« The solid crust which, in some experiments, formed on the viscous liquid con- 
taining the growing spherites was, in every case, ruptured upon further loss of water, 
