1909 ] Taylor, Hybridization in Hummingbirds. 293 



A moment's consideration is sufficient to convince one that both in 

 separating the genera Trochilus, Selasphorus and Calypte, and in 

 distinguishing between the species within each respective genus, 

 characters of practically equal rank have been used. For example, 

 compare the kind of characters used by Ridgway (Report U. S. 

 Nat. Mus., 1890, p. 340) in diagnosing the genus Selasphorus, 

 with those which separate alleni from rufus within the genus. 

 Feather coloration generally, the form of the outer primary, and 

 the form and coloration of the tail-feathers are perhaps the most 

 important of the "generic" characters. Upon turning to the 

 species descriptions it becomes evident at once that characters of 

 the same importance have been made use of. The clearest distinc- 

 tion between rufus and alleni, apparently consists in a difference in 

 form of tail-feather, rufus possessing a notch on the inner web of 

 the next to the middle pair of rectrices which alleni does not have. 

 At the same time the "generic" characters of the most weight con- 

 cern the form of the feathers of tail and wing. It would seemingly 

 be as reasonable to put Selasphorus alleni in one genus and Selas- 

 phorus rufus in another, as to split up Trochilus on the basis of 

 characters of no more weight than those separating these two species. 



There are differences between Calypte, Selasphorus and the 

 present Trochilus to be sure, but are the differences of such impor- 

 tance that these subdivisions should be accorded generic rank? 

 The relatively frequent occurrence of hybrids would seem to sup- 

 port a negative answer to this question. 



So far, there have been recorded from California hybrids of 

 Calypte anna + Trochilus alexandri, Calypte anna + Selasphorus 

 alleni, Trochilus alexandri + Calypte costa>, and Selasphorus 

 rufus + Stellula calliope (Thayer & Bangs, /. c). It will be 

 observed that the Trochiline hybrids occur only between species 

 whose ranges overlap or adjoin, isolation of habitat therefore 

 being perhaps the main reason why other combinations do not 

 appear. 



