65^) BULLETIN 50. UNITED STATES NATIONAL MUSEUM. 



extends in winter to Cape St. Lucas, but is mainly resident, at least 

 in the more southern parts of its range. 



(6) Geothlyj)h trichas siniwsa. — A small form (slightly smaller even 

 than G. t. trichas)^ apparently confined to the salt marshes al)Out San 

 Francisco Bay. The coloration is essentially identical with that of 

 G. t. arizela^ but slightly darker, especially on the upper parts and 

 tlaiiks. 



(7) GcotJdypk trichas iiiodt.sta. — The resident form of western Mex- 

 ico (States of Sinaloa, Jalisco, and Colima and Territory of Tepic'). 

 This form is practically identical with G. t. arizela in coloration and 

 general dimensions, but has the bill much larger (about the same size 

 as in G. t. occidentalu)^ and would hardly be recognized by name were 

 it not for the fact that it has a definite range far removed from that 

 of G. t. arizda, the mucli more brightly colored G. t. occldentalis com- 

 ing between the two. 



(8) GeotliJyph tricluis mdanoj^s. — This is the subspecies of eastern 

 Mexico, and is evidently an ofi'shootfrom G. t. occidc/tta/f's, from which 

 it differs in larger size, in having the under parts entirely rich yellow, 

 and the white post-facial ]>and ]>roader and somewhat tinged with 3^el- 

 low on its lateral branches. Some specimens from Chihuahua, north 

 ern Mexico, are clearly intermediate between this very strongly marked 

 subspecies and G. t. oreldentalis. 



Geotdihjph helduKj!^ of the Cape St. Lucas district, and the several 

 forms peculiar to the Bahama Islands are evidently derived from the 

 same ''original stock"" as the forms of G. tricJias^ but being no longer 

 connected by intermediates may be considered specifically' distinct. 

 All these are much larger than any of the G. trichas group. G. heldingi 

 has the post-facial band entirel}^ yellow, a character shared by a much 

 smaller species of eastern Mexico {G. Jiavot^cJata)^ two of the Bahama 

 forms ( G. coryi and G. flavlda) having this marking also yellow except 

 across the crown, being thus intermediate so far as this character is 

 concerned, the other Bahama forms having the post-facial band gray 

 or grayish white, as in the G. trichas group. 



The Bahaman insular forms present a puzzling problem, which can 

 not })e solved until mau}^ more specimens have been secured, especially 

 from the islands of Great Bahama,^ Abaco, Andros, and Eleuthera. 

 Three easily recognizable forms unquestionably inha])it the small 

 island of New Providence, while two are certainly inhabitants of 

 Abaco; therefore they can not be subspecies of one form. One of 

 the three New Providence forms ( G. rostrata) is very different from 

 any inhabiting the other islands; the other two^ resemble, respectively, 



^A specimen (female) from Tehuacan, Pueblo, is doubtfully referred to the same 

 form; also one from Ortiz, Sonora. 



* I have not been able to examine a specimen from the island of Great Bahama. 



* These second and third New Providence forms are G.flavida, representing G. coryi, 

 and G. maynardi, representing G. tanneri. 



