BIRDS OF NORTH AND MIDDLE AMERICA. 337 



as long as outer toe but much stouter; claws usually moderately 

 developed or rather small, moderately to strongly curved, acute, that 

 of the hallux, however, sometimes as long as or longer than the digit. 



The above diagnosis, while by no means complete, is quite sufficient 

 to enable one to recognize a member of this extensive and exceed- 

 ingly varied family. The more typical Tyrannidse, especially the 

 smaller North and Middle American species, resemble, superficially, 

 the true Flycatchers (family Muscicapida) of the Old World, both 

 in general appearance and in habits; but the latter are true Oscines, 

 more nearly related to the Thrushes than to any other group of birds. 

 In South America, where the Tyrannidge are most numerous and 

 deversified, remarkable teleological reproductions or mimicries of 

 other oscinine groups have been developed. Some are similar 

 in general appearance and habits to the Kinglets (Regulinae) and 

 Titmice (Paridse); others are terrestrial and closely resemble Larks, 

 Stone-chats, "Wagtails, or Pipits; again, there are swallow-like, aerial 

 forms with deeply forficate tails; others strikingly Tody-like in appear- 

 ance; some closely resemble the Thrush-like genus Myadestes, others 

 recall Vireos; in fact there are few oscinine t^'pes which are not 

 more or less closely imitated in the group. 



More than five hundred and fifty species and subspecies of Tyran- 

 nidae are known, referable to more than eighty genera; and, as Doctor 

 Sclater remarks," " many more no doubt remain to be discovered, 

 as the remote portions of South America come to be investigated." 



Probably no other group of birds is more difficult to study; for, 

 among tlie very large number of species there are, in contrast with 

 the CotingidiB, comparatively few highly differentiated forms, 

 much the greater number being birds of dull coloration, with very 

 slightly modified structural characters; and even when the latter 

 appear distinct enough to form the basis of supergeneric subdivi- 

 sions of the family it is found on examination of others that, with 

 a few exceptions, the most extreme types grade almost imperceptibly 

 one into another. It is by no means unlikely tliat, as Doctor Sclater 

 has said, "future anatomical study on the lines pursued b}' Garrod 

 and Forbes may eventually succeed in giving us a better system 

 than anything that can be obtained from the study of the dried 

 skins;" and I am also of the opinion tliat such study will not only 

 resvdt in a satisfactory arrangement of the genera in accordance 

 with their co-relationships but will develop the necessity of dividing 

 the group into two or more, one of wliich' may include certain genera 



1 Cat. Birds Brit. Mu.s.. xiv. 2. 

 11422— VOL 4—07 22 



