148 PROCEEDINGS OF THE NATIONAL MUSEUM vol. no 



handle. There is no evidence in existing samples of intergradation 

 between mexicanus and townsendii. 



The geographic ranges of the two forms overlap in northern Sonora; 

 a sample from El Tigre Mountains contains typical individuals of 

 both forms, and Burt (1938, p. 26) may have had both species in 

 material from S^ric. It is true that populations of P. townsendii 

 become smaller in body size and darker in color toward the Mexican 

 border, but there is no accompanying increase in the frequency of 

 the accessory cusp of I^, reduction in the number of interfemoral 

 cross-ribs, reduction in contrast between bases and tips of dorsal 

 hairs, disproportionate decrease in the size of the auditory bullae, 

 the auricles and tragi, or the average length of the maxillary tooth 

 row, increase in the average depth of the brain case, or depression of 

 the rostrum. In other words, populations of P. tovmsendii, neighbor- 

 ing P. mexicanus on the north, show a convergence of some morpholog- 

 ical characters toward P. mexicanus, perhaps in response to similar 

 environmental stimulae. Lack of convergence in other characters, 

 however, indicates a different genetic structure in the two forms. 



Dice (1940) used a concept of ecological subspecies to explain the 

 discontinuous distribution of dark-colored subspecies on isolated 

 mountain peaks surrounded by lowlands inhabitied by pale-colored 

 subspecies. Dice supposed that the several isolated dark-colored 

 populations might have originated in situ as independent derivations 

 from the pale-colored stock, presumably in response to similar environ- 

 mental conditions. According to this hypothesis, some subspecies 

 may be ecological units, rather than units of similar heredity. 



Were it not for the mixed samples (Sierra de Valparaiso and Santa 

 Rosa) and absence of intergradation between the two forms, the 

 concept of ecological subspecies could be an adequate explanation for 

 the distribution of the Mexican bats of the subgenus Corynorhinus. 

 It could be assumed that the alternative forms developed wherever 

 certain ecological conditions prevailed. Under this hypothesis, inde- 

 pendent development of the ^^ mexicanus^ ^ -type might be expected 

 wherever relatively cool, moist conditions of environment occur, and 

 the "australis" -type would be expected under conditions of greater 

 warmth and aridity. The distribution of both forms then would be 

 largely independent of geography and would be related primarily to 

 ecology. The mLxed samples could be accounted for as chance min- 

 gling of the two forms during the nonbreeding season as a consequence 

 of the interfingering of humid and arid environments. Existing 

 samples appear to be too numerous and too well distributed, however, 

 to have failed to show evidence of intergradation, if such exists. 



It thus appears that the hypothesis of conspecificity is in no wise 

 applicable to the Mexican populations of the subgenus Corynorhinus. 



