'fgii J Todd, Bahaman Species of Geothlypis. 245 



they arrived on the island as one species, in what manner could 

 natural selection so act as to produce two distinct species on one 

 small rocky island, without hills, rivers, or any pronounced geo- 

 graphical features?" 



The point here raised has been quite fully discussed by Prof. 

 Dean C. Worcester in his paper on the distribution of Philippine 

 birds. 1 Prof. Worcester doubts (from negative evidence) if two 

 closely allied species would fuse under such circumstances, but 

 thinks that either they would both continue to exist, or one would 

 exterminate the other. Instances of such coincident distribution 

 are not so numerous, however, as to allay the suspicion that we 

 may not be dealing here with a case of this kind. Mr. Bonhote 

 has advanced the theory that the observed differences are due 

 entirely to age, the duller-colored individuals being birds in first 

 nuptial plumage, while the brighter ones, are in second or later 

 nuptial plumage, basing his arguments mainly on a specimen in 

 his collection apparently showing the transition. Through Mr. 

 Bonhote's courtesy this specimen, No. 728 of his collection, is now 

 before me. It is an individual just completing (September 9) 

 the postnuptial moult, retaining only the feathers of the flanks and 

 tibia 3 , which in color differ little from those of G. rostrata. Other- 

 wise this specimen is typical of G. maynardi, being unusually richly 

 colored below, owing of course to its fresh condition. 



I should not be disposed to accept Mr. Bonhote's conclusions 

 on the evidence of a single specimen of somewhat dubious char- 

 acter, especially in view of the fact that his theory finds no support 

 by analogy when certain other species of this genus are examined, 

 but there are other considerations not to be overlooked. The 

 series of these birds which I have studied shows that in the 

 dull-colored examples the plumage, particularly the remiges and 

 rectrices, is more worn and faded than in bright-colored individuals 

 taken at the same season. This is especially marked in the speci- 

 mens from Abaco and Great Bahama, and fully accounts for the 

 average difference in wing and tail measurements between the 

 two series. Such a condition is precisely what obtains in the case 

 of many species whose first nuptial plumage is merely the first 

 winter plumage plus wear, to which the feathers of this stage are 



i Proc. U. S. Nat. Mus., XX, 1898, 600-617. 



