ig20.] S. Iv. HoRA : Indian Homalopteridae. 213 



torrents has called for the reduction of the bladder, but it has not 

 gone so far as in the two families mentioned above. In P. ienta- 

 ciilatus the air-bladder is quite normal. In Psilorhynchns sp. the 

 posterior chamber is reduced while the anterior one shows lateral 

 expansions. In some individuals of P. haliioya (Cat. No. 940 ) the 

 posterior chamber is pushed out of its place and comes to lie 

 close to the anterior one, partly covered by it, while an extreme 

 phase is reached in other individuals of the same species (No. 

 1098), in which the bladder is much reduced but still retains the 

 essentials of the normal form, the posterior chamber being very 

 small while the anterior one is covered by a thick fibrous coat. 

 Even the vertebral elements near the bladder show slight modi- 

 fications. 



When dealing with forms especially adapted to a particular 

 environment it is very difficult to distinguish homologous from 

 analogous characters and there is always a probability that fishes 

 coming from an absolutel}' different stock have been similarly 

 modified in response to a particular environment. In a former 

 paper by Dr. Annan dale ' and the author it was pointed out that 

 two absolutely differe nt stocks, Schizothoracinae and Salmonidae, 

 have come to have a close superficial resemblance to one another, 

 owing to their life in rapid running streams which necessitates 

 migrating up stream at certain periods in their lives. Numerous 

 other such instances could of course be adduced. 



Day in his Fishes of India points out that the genus Psi- 

 lorhynchits is intermediate between Hoinaloptera and Discognathus. 

 We have alread}^ seen that Homaloptera is closely allied to the 

 Cobitidae, whereas Psilorhynchiis belongs to the Cyprinidae. It 

 will be worth while to discuss its relations with Discoo^nat/ius at 

 this stage. I agree with Dr. Annandale* that Psilorhynchus is not 

 a primitive form of Discognathus ^ but that both of these genera 

 have been evolved from a primitive form like Crosso.hilus or Laheo 

 and show a parallel evolution. The specialisation in the former 

 is due to its life in mountain streams, while that in the latter is 

 due primarily to its peculiar mode of feeding The forms like 

 Garra nasutus and D. hlandjordii, which in all probability have 

 come to live secondarily in hill streams, the muscles of the chest 

 are modified like those found in Psilorhynchus, and thus true con- 

 vergence is established. 



The following has been the probable course of evolution in 

 these different genera, so far as it can be known in the present 

 state of our knowledge : — 



Taking Laheo as a central type of Cyprinidae I believe that 

 the evolution of a form like Crossochilus has taken place ; from this 

 evolution proceeded along two lines which ran more or less parallel 

 to one another. One of these lines culminated in Psilorhynchus 



' Annandale and Hora, Rec. Ind. Mils. XVIII, p. l68. 

 S Annandale, Rec. hid. Mas. XVI, pp. 113 — 116. 



