244 Todd. Bahaman Species of Geothlypis. LApril 



all these characters the present form is approached by some speci- 

 mens of the maynardi-tanneri type, to which it is obviously allied. 

 The female, too, is more richly colored than in the other forms, 

 being much brighter olive (as bright as the male) above, and 

 brighter and more extensively yellow below, while the supercil- 

 iaries are distinctly yellow. The alleged differences in proportions, 

 and in the width of the black frontal band, seem inconsequential 

 upon comparison. 



It appears, therefore, that on every island except Eleuthera 

 where the large Yellow-throats occur two styles of bird are found, 

 one of duller, the other of brighter, colors. What relation do these 

 two types bear to each other, and to their respective inter-island 

 variants? Are the five forms which are susceptible of definition 

 full species, geographical variations of one or more specific types, 

 or different plumages of such subspecies? Obviously the two forms 

 from any one island cannot be regarded as conspecific without 

 violating the definition of a subspecies as a "geographic race," 

 inasmuch as both occur together in an unusually restricted area. 

 We have seen that G. rostrata and G. incompta on the one hand, 

 and G. maynardi, G. tanneri, and G. coryi on the other, represent 

 respectively the dull and the bright birds on the various islands. 

 W 7 e might therefore be justified in arranging the forms as follows: 



1. Geothlypis rostrata rostrata. New Providence, Andros. 



2. Geothlypis rostrata incompta. Abaco, Great Bahama, etc. 



3. Geothlypis tanneri maynardi. New Providence, Andros. 



4. Geothlypis tanneri tanneri. Abaco, Great Bahama, etc. 



5. Geothlypis tanneri coryi. Eleuthera. 



So far as the actual characters of the forms go such an arrange- 

 ment is natural enough, but there still remains the question why 

 two so closely related specific types should thus occur together, 

 occupying the same habitat, and still preserve their distinctive 

 characters. Would not such an association of the two forms con- 

 stantly tend to interbreeding between them to a greater or less 

 extent, resulting ultimately in but one form? Mr. Bonhote, in 

 the paper before referred to, presents the question thus: "Sup- 

 posing that they reached the island [of New Providence] as two 

 separate forms, they would be bound to approximate and merge 

 together; or supposing, which is almost certainly the case, that 



