No.3.) ZHE EMBRYOLOGY OF THE EARTHWORM. 417 
question, however, appears to be largely one of terms only, for 
both the neuroblasts and the neural cells form at this period 
part of the general ectoblast (Figs. 53, 73), and all the cells 
are constantly undergoing division. The fact that the neuro- 
blasts, when first distinguishable, lie some distance behind the 
mouth, seems to indicate, however, that the cephalic ganglia and 
the cesophageal commissures are first differentiated zz sztu 
directly out of the ectoblast, and that the neuroblasts are after- 
wards differentiated as special centres of apical growth as the 
embryo begins to elongate. 
Soon after their union the cephalic ganglia project into the 
prostomial cavity, and become surrounded by a mesoblastic in- 
vestment. As the stomodzum grows inward, these ganglia, as 
well as the commissures extending around the cesophagus, are 
carried backwards until they lie in the third somite, which is 
their permanent position (cf Figs. 50, 51, 79, 81). 
The ventral chain arises entirely by concrescence of the two 
neural cell-rows, a process which has been accurately described 
by Kleinenberg. I have sought in vain for any indication of the 
participation of a median element (‘ Mittelstrang”’) in the forma- 
tion of the nerve-cord. The lateral nerves are formed as out- 
growths from the ganglionic cord. 
Concrescence of the neural rows begins in the first somite im- 
mediately behind the lower lip, where a slight ganglionic swell- 
ing is formed. A second more distinct double swelling is found 
in the second somite, a third in the third somite, and so on through 
the body. As the cerebral ganglia travel backwards in the man- 
ner described above, the ganglia of the first three somites fuse 
completely together to form the ganglionic mass known as the 
suboesophageal ganglion, which would therefore seem to repre- 
sent three pairs of ganglia. Itis, however, possible that the orig- 
inal anterior enlargement in the first somite is not morphologi- 
cally to be regarded as a ganglion. 
A careful study of cross-sections of the neural cords along 
the region of concrescence brings out the fact that the gangli- 
onic regions are the first to unite, and that they can be distin- 
guished as enlargements of the neural cords before concrescence 
begins; z.e. behind the point of union. Figures 90 and 91, PI. 
XXIL., illustrate this point. The former, which is taken through 
the region of the commissures, shows the neural cords perfectly 
