DEC. 28, 1916. FISHES OF PANAMA MEEK AND HILDEBRAND. 233 



The two representatives of the family Gobiida taken here by us are 

 widely distributed on the Pacific slope of tropical America. 



GENERAL CONCLUSIONS WITH RESPECT TO DISTRIBUTION. 



We regard Panama as a very remote center of distribution and 

 believe that the fresh water fishes have mostly migrated to this region 

 within comparatively recent geologic times. 



The fish fauna of Panama is essentially that of South America, and 

 most of the forms seem to have entered from that direction. The 

 families Loricariidce, Pygidiidce, Cyclopidce, Callichihyidas, as well as several 

 of the representatives of the families Silurida, Characida, Gymnotidce 

 and Cichlida known from South America reach their northern limit of 

 distribution in Panama. The extreme likeness of the fauna of the 

 Pacific slope of Panama and that of the Atlantic slope of Colombia, 

 and in particular the close similarity of the fishes of the Rio Tuyra and 

 the Rio Atrato strongly indicate that the last gap between these two 

 streams was closed within comparatively recent times. As previously 

 stated the watershed between the upper tributaries of these streams is 

 still very low and the natives continue to drag their canoes from the 

 head waters of one stream to those of the other. 



Several species of the family Loricariida, as well as some of the 

 South American forms of the families Silurida, Characidcs and Cichlida, 

 have found their way into the Rio Chagres. These seem to us to have 

 reached this river before the last gap between the Rio Chagres and the 

 streams of middle Panama was closed, rather than to have migrated 

 directly from the Colombian streams. If this is the case, the last gap 

 between the Chagres and one of the Pacific slope streams opposing it 

 must have been closed earlier than the last passage between the Rio Tuyra 

 and the Rio Atrato, for the similarity between the fishes of the two slopes 

 in the region of the Canal Zone is not nearly so great as it is farther east- 

 ward (Rio Atrato Rio Tuyra). It then is very probable that only 

 the earliest migrants succeeded in reaching the Rio Chagres Basin. 



The representatives of the family Poeciliidce may have entered 

 Panama from Central America, as the majority of the species resemble 

 Central American forms more than they do South American. The 

 fresh water representatives of the family Mugilidce have evidently 

 migrated to this region from the north and reach the southern limit 

 of their range in the Rio Chagres and the Rio Bayano. 



The fresh water species of the family Gobiida are more numerous 

 in Panama than elsewhere. Most of them still frequent brackish water 

 and it is probable that they evolved from marine shore forms, which 

 have worked their way up streams and into fresh water. 



