Denniston—Growth and Organization of the Starch Grain. 669 
talline grains. Radial crystal needles were formed by Bis¬ 
marck brown, chrysodin, malachite green, brilliant green and 
thionin. Fischer says that it was plainly to be seen that the 
precipitates were not in the denser substance of the grain, but 
in the watery zones. The periphery always remains unaf¬ 
fected. 
This peripheral layer of the starch grain has already been 
described in a preliminary paper (5), and is discussed further 
below; its staining capacity, form and constancy seem to indi¬ 
cate that it is different in composition from the layers inside. 
The following experiment gives strong evidence that the 
Violet stain passes through the orange layer readily but is not 
absorbed by it. In microtome sections, as already described, 
the staining of a large eccentric Ganna grain may be watched 
under the microscope by allowing a solution of gentian violet 
to run under the cover. The layers inside this peripheral 
layer begin to absorb the stain at once, but the outer layer is 
not at all affected. The violet stain passes through the outer 
layer without being fixed. Of the layers inside, some are 
stained a deep violet, others take up only a small amount of 
stain and appear pale violet in color. This phenomenon is 
quite inconsistent with the view held by Salter that the outer 
layer is merely denser starch. 
Rarely, in large Ganna starch grains, we find the peripher¬ 
al orange-stained layer followed toward the inside by a narrow 
dark violet layer, broadest at the posterior end (see PL 
XXXIX, Fig. 3,6). Next toward the inside of the grain 
there is a layer which is stained in some cases orange and ini 
others pale violet. This layer is fairly broad and is contin¬ 
uous around the hilum. The remaining layers of the grain, 
with the exception of those immediately surrounding the hilum, 
are incomplete. 
Oftentimes in the same material, a dark crescent-shaped 
line appears in the middle of a broad orange peripheral layer 
and at the posterior end of the grain (see Fig 38). The 
orange material between this line and the inner violet layers 
usually shows a faint violet color though still predominantly 
orange. This appears to be the beginning of a violet layer, 
