DEVELOPMENT OF SEED IN POLYGONACEAE 465 
the endosperm is almost exhausted of its reserve materials, then the 
protoplasm becomes clear and transparent, the nucleus having dis- 
appeared. 
Torrey (22), working on the cytological changes accompanying 
the secretion of diastase in the epidermal cells of the scutellum in 
maize, finds granules appearing in the cytoplasm during germination 
very similar to the condition found by Brown and Morris in the 
diastase secreting cells in the barley. In both these cases the nucleus 
finally disappears when the cell has exhausted itself after long con- 
tinued secretion. In Rumex the writer was unable to follow the earlier 
changes undergone by the nucleus; however, it was noted that the 
nuclei lose their staining power and finally disappear as the secreting 
cells become exhausted. It is, nevertheless, quite probable that the 
nuclei do play an important role in the production of the ferment. 
Torrey gives a very interesting interpretation of the origin of these 
granules in the epithelial cells of the scutellum in maize. At the 
beginning of germination the nuclei of these cells contain dark staining 
granules, which are ejected into the cytoplasm through minute breaks 
in the nuclear membrane. These granules at first spread through the 
cell, but later collect at the end of the cell next to the endosperm, where 
they are ultimately dissolved. In Rumex the granules are distributed 
quite uniformly through the cytoplasm of the secreting cells and no 
such origin as Torrey describes is evident. He also notes an enlarged 
condition of the secreting cells similar to that which occurs in Rumex. 
He evidently considers this enlargement to be due merely to osmotic 
pressure without an increase of cell wall structure as the writer believes 
to be the case in Rumex. He suggests (p. 430) that the osmotic 
activity is very likely set up by some substance secreted in the cell 
at this time. The protein nature of diastase prevents it from being 
as osmotically active as other less complex substances, as for instance 
organic salts or acids, which when present in the cell exercise a strong 
osmotic attraction. Nevertheless, the enlargement of the cell occurs 
when the nucleus has become completely filled with these granules. 
The formation of some organic acid during the great metabolic activity 
of the cell may be a possible explanation of this phenomenon. This 
hypothesis is further supported by the fact that diastase, to be espe- 
cially effective as a ferment in physiological experiments, must be 
dissolved in a slightly acid liquid. 
In contrast to the condition found in the plant kingdom, Mathews 
