1883.] Allen on Geographical Variation in Size amotig Birds. 8 1 



The various families of North American Mammals were re- 

 viewed in illustration of these propositions, which they seemed 

 to abundantly support. The only species showing a marked 

 increase in size southward were members of the genera Felis^ 

 Procyon^ and Sczurzis. The first- and last-named have their 

 centre of development and largest species in Southern Mexico 

 and southward, while Procyon belongs, with the exception of a 

 single species, to a family wholly tropical. 



As birds, with exceedingly few exceptions, are migratory at the 

 far North, and are thus able to escape the wintry severity of the Arc- 

 tic and Subarctic regions, they are less likely to present the double 

 decadence in size implied in the first proposition above-quoted, 

 and really shown in some mammals. Yet the cases of decrease 

 in size northward among the Anatidce referred to by Mr. Ridg- 

 way in the last number of the Bulletin (Vol. VIII, p. 63) may 

 fall under this head or may be viewed as only coincidences.* 

 The few marked cases among North American birds of increase 

 in size southward seem to occur among certain genera of Oscines 

 which are either for the most part tropical or belong to tropical 

 groups. Perhaps the most striking case is that of Catherpes 

 mexicanus, with its two small northern races, conspersus and 

 pttncHilatus. Again Thryothoriis ludovicianus has a large 

 southern race {miamensis) in Florida. The genus Thryothortis 

 is mainly tropical in its distribution, and belongs, like Catherpes^ 

 to a subfamily chiefly represented in the American tropics. 



The yellow-throated, black-masked section of the genus Geoth- 

 lypis^ a group also mainly tropical, affords several illustrations 

 in point. G. ti'ichas^ the only species of the section having a 

 wide range in North America, is the smallest of the group, with 

 a large race in Mexico and another large race in the Bahamas. 

 G. poliocephala is represented by a small race in Mexico and a 

 larger one in Central America. G. cequinoctialis is represented 

 by a large race in equatorial America and a smaller one in 

 Brazil. 



In Pyranga, the only North American genus of a tropical 

 American family, P. cestiva of the United States is represented 

 by a larger race {cooper i) in Mexico. P. saira, of equatorial 



* So far as the distribution and breeding ranges of Fulix marila and F. affinis are 

 known to me they seem to hardly fall into the category here cited. 



