ACCOMPANYING THE SECRETION OF DIASTASE 429 
A sufficient number of stages in germinating barley have been 
studied to make it clear that we have the same series of phe- 
nomena here during the secretion of diastase as in maize. The 
granules arise in the nucleus, they find their way into the cyto- 
plasm, and immediately after their disappearance from the cell we 
note the action of an enzyme on the endosperm. The collecting 
of the diastase granules at the tip of the cell nearest the endo- 
sperm is especially noticeable in barley (/. 8). 
Although the iron haematoxylin stain is one of the most use- 
ful ones for cytological purposes, there is a common source of 
error in its employment, which should be carefully guarded 
against. It has a tendency to stain in a very deceptive way any 
precipitates that may be present in the cell. Such precipitates are 
easily mistaken for natural metabolic products. Accordingly 
Auerbach’s method was used as a precautionary measure. It is 
interesting to note that those stages of the nuclei which stain 
an intense black with iron haematoxylin, also stain very deeply 
with methyl green; whereas the nuclei with granule-filled cyto- 
plasm stain very lightly with methyl green. This is another proof 
that matter is ejected from the nuclei at the same time that gran- 
ules appear in the cytoplasm. 
3. Comparative—Very few examples of secreting cells in 
plants have been studied with any degree of care. The results of 
Huie and Schniewind-Thies, accordingly, are of especial inter- 
€st in this connection. Miss Lily Huie (96) found that during 
secretion in Drosera the basophile cytoplasm was depleted. The 
basophile chromatin increased in the meantime greatly in extent, 
and, after ejection from the nucleus, gave rise to new basophile 
cytoplasm. After long-continued secretion the basophile cyto- 
plasm was entirely replaced by a scanty eosinophile cytoplasm. tt 
was not determined whether the nuclear, or nucleolar chromatin, 
was the primary product of the metabolism. In diastase-secreting 
cells the cytoplasm is not so directly concerned in the formation 
the secretory product as appears to be the case in Drosera. 
The only evidence of the possible activity of the cytoplasm in the 
former is the increase in size of some of the granules scat- 
tered through it. Schniewind-Thies (97), who has made a very 
thorough study of the nectar cells ina number of plants, finds that, 
