16 



BULLETIN 86, UNITED STATES NATIONAL MUSEUM. 



1 New subspecies ; see p. 57. 



=i See p. 71. 



» See p. 98. 



* New subspecies ; see p. 100. 



" New subspecies ; see p. 109. 

 * New subspecies ; see p. 116. 

 ' New subspecies ; see p. 118. 



Recognition of subspecies. — All but two of the forms here treated 

 as subspecies are found to intergrade with some one or even several 

 other forms, either by gradual mergence across an intermediate geo- 

 graphic area, or if isolated on islands or elsewhere, then by individual 

 variation, and in the first case not infrequently both geographically 

 and individually, so that each species is made up of a complete chain 

 of connected races. The two exceptions mentioned are Chordeiles 

 rupestris xyostictus^ and Chordeiles rupestris zdleucus,^ both being 

 little-known birds of probably limited distribution, the characters 

 of which are such as reasonably presuppose their subspecific, rather 

 than specific, relationship, for which reason we have here assigned 

 them such a rank. 



The problem of the number of subspecies to be recognized is, as 

 in the case of similar wide ranging and plastic animals, a somewhat 

 difficult one, particularly in Chordeiles virginianus and Chordeiles 

 acutipennis. The great individual variation seriously complicates 

 the situation, for however few may be the subspecific divisions of 

 particularly these two species, such subspecies will more or less 

 overlap those of one or more others, hence the characters can in no 

 case be more than average. There are, notwithstanding this exces- 

 sive and troublesome individual variation, well-marked geographic 



8 See p. 116. 

 » See p. 118. 



