GENERAL CONSIDERATIONS. 173 



Comiiuriiig the labio-l)uccal systems of Prnneoincniu lunraiiensis and a 

 Chiton {Trachydermon raymouiii) it is seen (hat in the Solenogastres the con- 

 nectives attaching the su])ra(hilai-, buceal, and labial systems witli the brain 

 are of great length; in the Chitons thej^ are very short. In the Chitons the 

 buccal ganglia are clearly differentiated; in the Solenogastres they are fused 

 with the labial. These homologies have been treated in another paper (Heath 

 '05), and offer, so far as 1 can see, no serious difficulty. 



Nerves from the labio-buccal ganglia have been seen in both of the divisions 

 of the Solenogastres to pass into the pharynx which they doubtless innervate. 

 In Chaetoderma erudita they have been traced as far as the end of the pharynx. 

 In the Chaetodermatidae practically all of the nerves innervating the posterior 

 end of the body arise on the suprarectal commissure or in close proximity to it. 

 In Limifossor the gills are innervated by two pairs of branches from the com- 

 missure antl in Chaetoderma attenuata, C. erudita and probably others the same 

 is true. 



While the facts discussed in the foregoing paragraphs appear to justify the 

 conchision that the Solenogastres are most closely related to the Chitons, they 

 do not as certainly settle the cjuestion as to which group has retained the 

 greater number of ancestral characters. The condition of the coelom in the 

 first named division appears to be very primitive and probably palingenetic; 

 and, generally speaking, the musculature is more simple, and this is true to 

 some extent of the digestive tract, though these may have been secondarily 

 modified. The absence of a shell and well-developed foot, the relatively simple 

 condition of the circulatory apparatus, the concentration of the nervous system, 

 and, in the Neomeniina, the high degree of development of the coelomoducts 

 point more clearly to modifications of a more primitive type. Without enter- 

 ing into further detail it would appear that, with the data now available, the 

 Chitons are to be considered the more primitive, in fact the most primitive of 

 all mollu.scs. 



