F G 
These facts demonstrate, as I believe, that the development of 
an isolated blastomere of the 2-celled stage of Amphioxus is not, 
as in the frog, at first a half-development afterwards completed by a 
process of regeneration (“post-generation”, Roux). The isolated 
blastomere develops as a unit, not as a half-unit; and 
the cells to which it gives rise cannot be individually 
identified with those of a normal embryo-half. The 
development is transformed from the beginning. In 
this respect Amphioxus differs both from the frog (Roux) and 
from the sea-urchin (DrisschH), in both of which the development at 
first agrees with that of a normal embryo-half and only later gives 
rise to a perfect embryo by a process which may provisionally be 
called “regeneration”. It is an important fact however that the “re- 
generation” takes place at a much earlier period in Echinus than 
in Rana, and it is perhaps possible to regard the case of A mphi- 
oxus as due to a throwing buck of the regenerative process to a 
