148 COMPARATIVE ANATOMY. 



the body (embedded within the muscular layers), in others (CErstedia) 

 they approach one another ventrally, and are distinguished by swell- 

 ings at the points where nerve-branches are given off. This is an 

 anticipation of the future development of ventral ganglia, 

 the elements of which are already pi'esent in the longitudinal trunks. 

 The ventral approximation of the longitudinal trunks shows us 

 how the central system got its ventral position, which becomes 

 further developed by the formation of ganglia. But the ventral 

 approximation of the longitudinal trunks, by surrounding the oeso- 

 phagus, leads to the formation of an oesophageal nerve-ring, 

 when the ventral longitudinal trunks meet. It is an open question 

 whether the oesophageal ring of the Bryozoa, and of the Nemathel- 

 minthes arose in such a way as this, and we would only just remark 

 that even if they had a similar origin it does not follow that there 

 is any connection between them and the oesophageal nerve-ring of 

 other worms. For in the Nemertina we see that the origin of this 

 arrangement is due to the two peripheral longitudinal trunks, which 

 are not present in the other case. 



In the Nemertina the oesophageal ring is not closed ; in the 

 Annulata it is closed by transverse connections between the primi- 

 tive longitudinal trunks. These have gained a central significa- 

 tion, owing to the large number of ganglion cells in them ; the 

 longitudinal trunks now appear to commence as commissures, 

 which connect the primitive dorsal nerve-centre (cerebrum) above 

 the oesophagus with the ventral one, which is formed from the 

 longitudinal trunks. 



§ H8. 



The cerebral ganglionic mass is not always present in the 

 Gephyrea. It is well marked in Sipunculus and Sternaspis, and in 

 the former it is divided into two parts. In Bonellia and Priapulus, 

 however, only fibrous elements surround the oesophagus, so that in 

 comparison with the other two a change has occurred, in which the 

 central elements must be supposed to have become degenerated, or 

 to have taken on a ventral position. The consequent arrangement 

 corresponds to a great development of the commissure, which 

 was formerly present between the two halves of the superior ganglia. 

 Instead of the two ventral longitudinal trunks there is a single 

 nerve-chord, in which the fusion of two trunks is only a matter of 

 inference. This ventral chord generally lies within the coelom, but 

 in some it is placed outside the muscular layer, just below the 

 integument (Priapulus). 



As a rule, there are no collections of ganglion cells into special 

 swellings, or expressions of metamerism ; it is only in Echiurus that 

 they are present, and in it they are but feebly developed ; in other 

 cases (Sipunculus, Sternaspis) there is a terminal thickening, of the 

 chord which gives off fine filaments. 



The ventral chord gives off filaments on either side, which are 

 frequently irregular in origin ; they are the peripheral nerves. 



