AFFINITIES OF CHIM^ROIDS II5 



derived from its conditions as ancestral. The dentition " 

 of Chimaeroids alone is so remarkable that no direct proc- 

 ess of differentiation could convert it into the structures of 

 lun2:-fish or Ganoid. A number of archaic features draw 

 fishes together in the lines of their descent, but they can- 

 not be interpreted as linking the Chimaeroids with the 

 Dipnoans, or the Dipnoans with the Chimaeroids. Auto- 

 stylism, often adduced to ally these groups, differs widely 

 in its characters in each (p. 254) : and the apparent similar- 

 ities in dental plates and membrane bones are closely 

 paralleled by the sharks. The diphycercal tail of the 

 Chimaeroid can be made no standard of comparison, since 

 it is evidently a secondary structure, arising within the • 

 limits of the group, as it may well have done among 

 sharks (Pleuracanthus) or Teleostomes (Polypterus, eel). 



If the sum of the general characters of Chimaeroids be 

 considered, their affinities would clearly be to the most 

 ancient sharks. Their structures are not so widely at vari- 

 ance with those of Elasmobranchs that they cannot rea- 

 sonably be derived from their more generalized conditions 

 in vertebral characters, cranium, mandible, girdles, fins, 

 membrane bones, gills. Absence of swim-bladder is again 

 strikingly shark-like. Like the ancient sharks, they have 

 been well adapted for survival by evolving but few special- 

 ized structures {e.g. dentition, gills). Their ventral clasp- 

 ing organs separate them clearly from the Dipnoans. 

 Until the discovery of Harriotta the frontal clasping spine ^\, 

 remained as one of the most distinctive features of Chi- 

 maeroids ; its high degree of specialization in Liassic times 

 is alone significant of the antiquity of their descent. 



