664 JULIAN $8. HUXLEY 
a large number of buds practically simultaneously by segmenta- 
tion of the very long post-abdomen, with subsequent differentia- 
tion of each segment to form a whole zooid, is apparently 
responsible for this. After dedifferentiation of the primary 
zooid has proceeded a certain way, either death supervenes 
or else the post-abdomen, released from subordination now that 
the dominant region is thus adversely affected, manifests its 
independence by producing new individuals. Once these new 
individuals start to develop they become dominant. The non- 
recovery of the partially-dedifferentiated original zooids may 
be ascribed to this, or to greater susceptibility. In spite of 
this absence of the power to redifferentiate the process of dedif- 
ferentiation is very similar to the early stages of the same 
process in Clavellina. or such behaviour there is ample 
evidence as regards numerous forms reproducing asexually in 
the work of Child and his pupils (Child, 19156). We may thus 
say that, under the conditions which prevail in the colony, 
or in pieces of it, in Amaroucium, complete dedifferentiation 
of single zooids is not possible. The colony or piece regarded 
as a whole, however, may be said to undergo dedifferentiation 
followed by redifferentiation. 
Oozoites.—These had the advantage over blastozoites 
that they could be obtained singly. They were got by allowing 
larvae to metamorphose in the laboratory. They could be 
induced to dedifferentiate either by lack of change of water, 
or, after a longer period, by starvation. The process was very 
similar to that in the blastozoites, with the exception that the 
formation of buds was never observed. ‘This latter fact is 
undoubtedly to be correlated with the small relative size 
of the post-abdomen and the small absolute size of the whole 
organism. 
Here, too, dedifferentiation never got beyond a stage in 
which a sausage-shape was assumed (fig. 27, Text-fig. 1). 
The complete opacity and the spheroidal shape of the final 
stages of the process in Clavellina were not observed ; neither 
did I succeed in obtaining redifferentiation. 
On the whole, dedifferentiation in oozoites went a little further 
