xxvi Journal of Comparative Neurology. 



nally were added to the paleocranium from the vertebral column 

 (neocranium ) are in the ontogeny secondarily again separated by an 

 articulation and added to the vertebral column. Another case is that 

 of the epibranchial and hypobranchial musculature. There is embry- 

 ological evidence in both of these cases that we have derivatives of the 

 somites here and not a visceral musculature, and the present research 

 shows that both are innervated wholly from the cervical plexus and 

 not from the vagus, as several previous authors have stated. But in 

 both cases an extended comparison of many species shows that the 

 number of myomeres does not correspond with the number of 

 nerves which supply them nor do the myomeres correspond with 

 the gill arches. As to the cause of this dysmetamerism, it is 

 suggested that probably migrations of the pectoral fin constitute a 

 disturbing factor, while the discrepancies between the muscles and 

 their arches is due to the fact that these two systems of muscles having 

 been derived from the somites their relation to the branchial arches 

 is secondary throughout. In fact the existing branchial arrangments 

 must be interpreted as the resultants of two opposing forces which 

 strove for the mastery very early in the phylogenetic history of the 

 vertebrates, the backward (caudal ) growth of the branchial apparatus 

 and the forward (cephalic) growth of the somites of the spinal trunk 

 musculature. 



Among the selachians there is evidence of the absorption into the 

 head of the metameres corresponding to those of the occipital system 

 of nerves; but the nerves themselves have not in the selachians been 

 fused with the proper palingenetic nerves of the head ( branchial 

 nerves); on the other hand this system is always very distinct from 

 that of the vagus. Furthurmore the occipital nerves give evidence of 

 a progressive reduction from before backward, as indicated by the 

 fact that their number is greater in the lower members of the group 

 ( NotidanidK ) and becomes progressively less up to the higher 

 (Rajidae) and by the fact that they diminish in size as we pass cephal- 

 ad. This reduction is correlated with that of the corresponding body 

 segments, the sensory elements having disappeared first. 



In a few cases (e. g.,Notidanidce) the last occipital nerve retains 

 its dorsal root and ganglion in the adult ; in other cases one or more- 

 dorsal roots have a transitory appearance in the embryo. The degen- 

 eration of the dors;'.l sensory roots of the occipital and occipito-spinal 

 nerves is explained as due the progressive encroachment of these seg- 

 ments into the domain of the vagus, so that they have been replaced by 

 the rami laterales of the X and IX nerves. The principle seems to be 



