Evolution of x\nimals 395 



restricted form by an aquatic involution, or by an aerial lung. 

 Both groups are lacustrine or terrestrial, and include about 

 one-third of living forms of such habitats. 



From a common ancestry TNith the aquatic pulmonates 

 the^ patellid, nudibranch, and related marine forms seem to 

 have been derived, and while often retaining a soft body, or 

 one with spicules, m others there was developed a shell from 

 the shell gland that has hereditarily been derived in common 

 with the cement gland of rotifers, and the egg-cement gland 

 of some annelids. 



The gradual development of ctenidia for breathing, as an 

 advance on simpler mantle respiration, has occurred most 

 extensively in marine forms, and this, along with various impor- 

 tant details of alimentary, nervous, renal, and reproductive 

 structure, suggests that they are derivative types from lacus- 

 trine or river inhabitants. Such however need not militate 

 against the view that some of these last may have migrated 

 back into rivers, lakes, and other inland localities. Varied 

 evidence favors such a conclusion, but it by no means indicates 

 that molluscan hfe mainly originated even along littoral areas. 



The Scaphopoda are all marine, but, though showdng affini- 

 ties with the Gastropoda, their simplified structure makes it 

 difficult to determine their relationship and primitive geo- 

 graphic area. Their first appearance and rarity in the mid- 

 silurian, and their climax of abundance in the cretaceous age, 

 help little in the elucidation of the group. Further observa- 

 tions on them are made in a later chapter (p. 5^7). 



The Lamellibranchiata as a living group seems to favor 

 strongly a marine origin, while the greater number of the 

 genera are marine, only about thirty, or one-seventh of the 

 entire number being estu.arine or fresh-water. It seems to 

 be a more modified and recent one than the two last, even 

 though concentration of the nervous ganglia does not reach 

 to so marked a degree. Many, probably all, of the fresh- 

 water genera have been derived from a marine ancestry, while 

 the most primitive division — the Protobranchia — are all marine 

 so far as known, alike in the fossil and the living sj)ecies. But 



