Denniston—Growth and Organization of the Starch Grain. 689 
the same time the layers at the anterior end of the grain begin 
to push out. 
When once this swelling of the anterior part of the grain 
commences, it proceeds rapidly (Fig. 45). The crack at the 
hilum continues to enlarge, forming a large cavity in this por¬ 
tion of the grain (Figs. 46, 47). The size of the anterior 
part of the grain soon considerably exceeds that of the poste¬ 
rior part. When the anterior part of the grain has swollen 
considerably, a peculiar invagination takes place beginning at 
the outside at a point near the hilum (Figs 48-50). This 
infolding appears to relieve the tension caused by the formation 
of a large internal cavity. The posterior portion of the grain 
is the last to swell, and if the grain is stained by the triple 
stain this unswollen portion takes a faint violet color and the 
swollen portion an orange color. Undoubtedly this swelling of 
the grains is caused by the absorption of water in the layers. 
The parts immediately surrounding the hilum take up the 
water most readily and are the first to swell. The outer layers 
of the grain are thinnest at the anterior end of the grain, and 
it is at this end that the stretching is greatest. The crack 
which begins at the hilum extends through the central part 
of the grain toward the posterior end and evidently follows 
the direction of the most readily absorptive material. The 
layers at the posterior end of the grain are less easily penetra¬ 
ble by the water and maintain their form for a longer time 
than the inner portions. 
Krabbe is of the opinion that the streaming motions which 
take place when a crystal of alum is brought into contact with 
a solvent also play an important part in the solution of starch 
grains by diastase. 
More recently, Goldschmidt (12, p. 656) obtained corrosion 
figures on spheres of calcium carbonate subjected to- the ac¬ 
tion of strong acids, which in their earliest stages show a 
strong resemblance to those formed in the wheat starch grain as 
the result of diastase action. The figures which appear in 
the calcite crystals take the form at first of hemispherical 
hollows; these Goldschmidt attempts to explain by the supposi¬ 
tion of the presence of molecular streams directed toward the 
