92 BULLETIN OF THE 
of fact they are depressed (Pl. V. fig. 28; Pl. VI. fig. 30), and, there- 
fore, can be accounted for more reasonably on the supposition of a centri- 
petal force. It has already been indicated that these areas are formed 
during the contraction of the vitellus ; surface views and sections com- 
bined show that the very plastic protoplasm of the blastema is moulded 
to the surface of the peripheral yolk corpuscles, to which it at first 
forms caps, all of the caps being joined at their margins. Owing to mu- 
tual pressure these subsequently appear as polygonal areas. The cause, 
then, producing this surface phenomenon is mechanical, and depends 
upon the contraction of the protoplasm of the egg. But under what 
influence does this contraction transpire? As the phenomenon takes 
place during the stage which is characterized by the existence of the first 
sesmentation-nucleus, it is more than probable that the cause is resident 
in this central nucleus, which exerts its attractive influence on all the 
protoplasm of the egg, but finds its external manifestation at this period 
principally in the blastema. Thus, it is probable that the same cause 
which produces in these eggs contraction of the vitellus, also induces the 
division of the blastema into areas. 
The manner in which the protoplasm acts upon the yolk in the assimi- 
lation of its substance is clearly indicated in the blastema stage. That 
portion of the surface of the corpuscles which is in contact with the pro- 
toplasm of the blastema appears deeply eroded (Pl. VI. figs. 30-33), and 
the fine fragments into which the detached yolk substance is broken 
gradually merge into the still more finely granular protoplasm. A some- 
what similar fate overtakes the yolk corpuscles in the vicinity of the 
central nuclei (Pl. V. figs. 28, 29; Pl. VI. fig. 34). Here, however, 
there is a very gradual transition from the larger corpuscles to the much 
smaller ones which immediately surround the finely granular protoplasm 
of these central cells, —a process of fragmentation appears to precede the 
erosion, and thereby a much greater surface of yolk substance is exposed 
to the action of the protoplasm. In the former case there is no total 
fragmentation of the yolk corpuscles, and the erosion proceeds from one 
side only, leaving the opposite side with a sharp, more or less even out- 
line. The preliminary fragmentation of the yolk in the vicinity of the 
rapidly proliferating cells is, without question, correlated with the rapid 
growth of the latter. 
Claparéde (’62), Barrois (’78), and Balfour (’80), have each given 
explanations of the reversion of the embryo. Balfour’s is the simplest. 
