Evolution of Animals 435 



and mesencephalon. Only detailed comparative study can 

 again accurately determine. In distribution and function 

 two pairs of lateral nerves that arise from the ventral ganglia 

 correspond well with the 9th and 10th or the glossopharyngeal 

 and vagus nerves of vertebrates. 



The point of origin, invariable superficial position, and 

 median dorsal course of the dorsal nerve cord are all highly 

 instructive in relation to the vertebrata. Arising from the 

 middle part of the dorsal commissure, it would follow that, were 

 the dorsal ganglia completely fused, this dorsal nerve cord 

 might become the leading tract for conveyance of stimuli 

 lengthwise through the body of the animal. 



But we would consider that the earlier position taken up 

 by Hubrecht, and by his predecessor Harting, namely that 

 the lateral nerves of nemerteans, and their anterior or ventral 

 cephalic ganglionic expansions, have gradually converged and 

 enlarged dorsally to form medullary and spinal cord constitu- 

 ents, has greatly more in favor of it than Hubrecht's later 

 view {136: 131) that the dorsal nerves, by gradual growth 

 and backward extension, gave rise to the medullary-spinal 

 system. 



For it is to be observed that the two divergent groups of 

 the annelid-arthropod series and of the nemertean-chordate 

 series start in common amongst primitive rhabdocoel forms 

 with a strong ventral nerve pair, a feebler lateral pair, and a 

 delicate dorsal pair. This becomes a very strong ventral 

 pair, a small lateral pair, and an attenuate or absorbed dorsal 

 system in annelid-arthropods. In Nemertinea on the other 

 hand the ventral pair becomes small, the lateral pair specially 

 strong, and the dorsal system reduced to a small median med- 

 ullary or dorsal nerve. With gradual absorption of the pro- 

 boscis into the oral cavity, and increasing solidification of the 

 sheath-notochord, the ventral ganglia and their nerves would 

 undergo condensation behind and above the front of the evolv- 

 ing notochord, to form the medulla and spinal cord. 



So, if we compare the diagram of the brain of Eupolia with 

 that which we would regard as a protochordate type (Figs. 

 18a, 18c), an exact progressive similarity can be traced. For 

 by conversion of the ventral commissure into the Pons Varolii, 

 as the proboscis sheath became closed anteriorly and func- 

 tioned as a notochord, the ventral lobes would rise upward 

 and backward, so as to be below and behind the dorsal centers, 

 and thus constitute the medulla; the formation dorsally of a 

 correlating commissure above the notochord would give rise 

 to the cerebellum, that is still a rudimentary crest in Myxine; 



