1172 Stozvard—A my toe/as tic Secretory Capacities of the 
The evidence afforded by these experiments, positive in character and 
extremely well defined, leaves little doubt that the embryo secretes an 
active cellulose-dissolving enzyme. 
When we turn to the control sections, these experiments being con¬ 
ducted contemporaneously and under identical conditions, the contrast 
is remarkable ; absolutely no visible sign of change is demonstrable in them. 
In order to more closely examine the cell walls, many of the sections were 
divested of their starch by treatment with diluted saliva ; microscopical 
examinations of these sections and similarly treated sections, for purposes of 
control, taken from endosperms of seeds which had been simply steeped, 
failed to show any evident sign of alteration, the two series presenting 
identical appearances. 
These results show with reasonable clearness that the changes observed 
were induced by the embryo. 
Similar series of changes to the foregoing are induced by the aleurone 
layer, but there is, as already pointed out, this important difference: they 
are more long drawn out, i. e. the action is slower. The important fact to 
be noted, already to some extent foreshadowed by the evidence afforded by 
the experiments described in Section IV a, is that, if the miniature ‘ culture ’ 
experiment is sufficiently prolonged (10-21 days), corrosion of starch gramdes 
identical in every recognizable respect with that induced by and hitherto attri¬ 
buted solely to the amylase secreted by the columnar epithelium takes place . 
The type of starch erosion observed by Brown and Escombe (loc. cit., 
p. 13 ), which they designate f sub-aleuronic represents, so far as my observa¬ 
tions go, one of the earlier phases of the phenomenon. During the final 
stages of the experiments (21-28 days or longer) the phenomenon becomes 
quite general. The cytohydrolysis produced by the aleurone-layer section 
presents during its various stages features which closely parallel those 
described for the embryo. 
During the course of the work described in this section the idea 
frequently suggested itself that the dissolution and digestion of both cell 
walls and starch granules were induced not by the direct action of the cyto- 
clastic enzymes secreted by the epithelial and aleurone-layer cells, but 
possibly indirectly by the action of substances other than secretions, i. e. 
substances, not enzymatic in nature, on the pre-existent enzymes in the 
tissues, which pass during the prolonged course of these experiments 
from these secretory tissues into the subjacent inner endosperm tissue, 
and possibly activate the pre-existent enzymes present in them. 
The experimental evidence adduced in Section VI did not favour this 
suggestion. The control experiments in those just described were not 
regarded as quite conclusive, for the reason that no means had been taken 
to annihilate the enzymes pre-existent in the thin section of endosperm 
employed ; it must, however, be admitted that the controls as thus con- 
