434 WILSON. [Vo. III. 
and the ventral chain are, therefore, at first not connected with 
each other. 
This explanation is extremely ingenious, but I must point out 
the fact that the separate origin of the cephalic ganglia and the 
ventral chains in the foetal types is by no means an established 
fact, and that, even if it were, it would be capable of a much 
simpler interpretation. Hatschek has always maintained the di- 
rect and primary connection of the apical plate with the ventral 
nerve-chain, and this view is supported by Vejdovsky’s observa- 
tions on Rhyuchelmts and my own on Lumbricus. In Lumbricus 
the cephalic foundations are from the first in continuity with 
those of the ventral chain, though the cesophageal commissures 
are represented only by rows of neural cells, which are still fused 
with the general ectoblast, and would scarcely be recognizable 
were it not for their connection with the neuroblasts. In later 
stages the commissures lag somewhat behind the other parts, 
and may thus easily be overlooked. It appears, therefore, as 
Hatschek has pointed out (No. 20, p. 72), that the entire ques- 
tion as to the separate or common origin of the cephalic and 
trunk ganglia relates simply to the period at which the ceso- 
phageal commissures are differentiated; and the possibility of 
secondary acceleration or retardation in these structures cer- 
tainly cannot be denied in view of the fact that in many other 
metameric animals, entire somites in the middle region may lag 
far behind other parts, or even be suppressed for a long time — 
as for instance in the Decapod zoéa. 
To sum up, it appears from the foregoing discussion that the 
only certain embryological ground for maintaining the contrast 
between the head and trunk lies in the fact that the head-cavity 
is unpaired, while the trunk-cavities are paired; and this I 
believe to be a real, though not a fundamental distinction. The 
fact that the foetal and larval types agree in this respect indi- 
cates its primitive character, and this conclusion is, moreover, so 
strongly supported by the facts of comparative embryology that 
no one would have called it in question but for Kleinenberg’s 
account of its development in Lumobricus. 
All the evidence seems to show, therefore, not only that the 
cephalic cavity is unpaired, but that this is an ancestral feature 
of the annelid body. But the evidence appears to me to indicate, 
furthermore, that the head-cavity is to be regarded as homo- 
