130 COMPARATIVE ANATOMY. 



of the ganglionic chain are widely separated from each other an- 

 teriorly. In the Sabellidfe and Hermellida? the chords are closer 

 together, and indeed in the anterior portion of the venti'al nerve- 

 chord the transverse commissures are much shorter than in the 

 posterior. Finally, the Terebellidfe close the series, for in them the 

 transverse commissures between the ganglia are distinct in the 

 posterior portion only, while in the anterior portion the ganglia of 

 either side are completely fused. 



With regard to the ganglia, the perfection and greater develop- 

 ment of the cerebral ganglia in Hirudinea and Annelides must be 

 mentioned as contrasting with the lower Vermes. The two halves 

 are very seldom united into a single mass; in Enchytrteus, 

 this condition appears to be due to degeneration. A breaking-up 

 into distinct lobate segments, of .which the Nemertina present a 

 simple case, may give rise to a great diversity in form. The lobes 

 frequently have the form of rounded projections, at times almost 

 pedunculated. The cerebral ganglia are then complexes of smaller 

 ganglia. 



In the ganglia of the ventral chord also, remarkable differen- 

 tiations may arise, partly from an increase in size, and partly from 

 concrescence. In the Hirudinea, the first ganglion is generally very 

 large, and always larger than the rest; it corresponds to a lai'ge 

 number of separate ganglia united with one another, as may be seen 

 from the segments which compose it, as well as from the nerve- 

 branches which arise fi"om it. A similar condition is found at 

 the end of the ventral chord, where the larger ganglion present, 

 and innervating the sucker, is produced by the concrescence of 

 several primitive ganglia (seven in Clepsiue), which corresponds 

 to as many metameres as form the sucker. This phaenomenon of the 

 approximation (by shortening of the longitudinal commissures) of 

 separate ganglia obtains also in the Scolcina, but here the inde- 

 pendence of the parts is often clearly recognisable, owing to the 

 presence of separate transverse commissures. The Hermellid^e 

 among the Chjetopoda are an example of this, for in them the first 

 seven ganglia on each side are in direct contact with one another. 

 The length of the commissures, as well as the number of ganglia, is 

 directly connected with metamerism. In the Lumbricida3 with 

 small rings they are very close together, so that the whole ventral 

 chord presents a compact series of swellings and constrictions. The 

 ganglia in Clymene and Cirratulus are still closer together. 



Owing to this close association of the ganglionic structures of 

 the ventral chord, it has been supposed to be analogous to the spinal 

 chord of the Vertebrata. And the ventral ganglionic chain has there- 

 fore been called the " ventral medulla." Although we may allow 

 the analogy, there is no reason at all for supposing that there is any 

 homology. Position, development, and structure forbid such a 

 supposition. As to structure, let it be here noted that the ganglion 

 cells in the ventral chord are in the periphery of the ganglia, 

 while the inner part is essentially occupied by fibrous bands. 



