196 
PROGRESS OF MIctlOSCOPICAL SCIENCE. 
accordance with Hcnsen’s * observations on the development of the 
posterior roots of the spinal nerves in the rabbit. 
Opposite the centre of each protovertebra the external epiblast 
grows downwards as a small conical process on either side of the 
spinal cord and in close contact with it. These processes were 
mistaken by His | for the commencements of the spinal nerves, but 
are clearly seen to have no connection whatever with the nerve- 
rudiments. His is the only previous observer who assigns an epi- 
blastic, instead of a mesoblastic, origin to the nerves in the chick ; 
he, however, derives them directly from the external epiblast, while, 
according to the description just given, they really arise from the in- 
voluted epiblast of the neural canal. 
From their mode of origin the cranial and the anterior spinal 
nerves will be seen to bo all connected together at first by a longi- 
tudinal commissure of spherical cells, while the two nerves of each 
pair, whether cranial or spinal, are also connected together across the 
top of the neural canal. 
The attachment of the nerve, whether cranial or spinal, is at first 
to the extreme summit of the neural canal. Shortly after their 
appearance the attachments shift slightly outwards, and in the case of 
the spinal nerve, become much more slender. The shifting is 
believed to be apparent rather than real, and to be caused, as first 
suggested by Balfour, by rapid growth of the cells at the summit of 
the canal, which has the effect of separating the roots of the two sides 
from one another and forcing them apart. 
Though the proximal part of the nerve-root becomes thus more 
slender in the spinal nerves, the distal part enlarges considerably, and 
grows down as an oval mass (the spinal ganglion ) between the spinal 
cord and the protovertebra'. At this period the most prominent part 
is situated opposite the interval between two protovertebrae. 
During the third day a great change occurs in the point of attach- 
ment, which is now considerably lower down, in the position occupied 
by the root in the adult. The nerve is now attached, not by its apex, 
but by a small process growing out from its side, and projects con- 
siderably above the point of attachment. Owing to the surrounding 
mcsoblast this stage is very difficult to investigate; but the appearance 
strongly suggests that the original attachment of the nerve to the 
summit of the card is lost, and a new one acquired lower down, and 
that tbe projection of the nerve above the point of attachment, which 
becomes inconspicuous very shortly afterwards, is a remnant of tho 
original attachment. 
The anterior roots of the spinal nerves arise later than the 
posterior, and have not been observed earlier than the latter part of 
the third day. They appear as small outgrowths from the lower part 
of the sides of the spinal cord, and from the first occupy tho position 
held by them in tho adult. This position is indicated before the 
actual appearance of the roots by a slight convergence of the cells at 
the outer part of the cord. The anterior roots are very slender, and 
* ‘ Zeitschrift f. Anatomie u. Entwickelungsgeschiclite,’ 187G, Btl. i. 
t ' Die erste Anlage ties Wil'belthierleibes.’ 
