386 bulletin: museum of comparative zoology. 



smaller than T. h. marchii. The average of 14 males, and 9 females 

 from Grenada and Grenadine, 4 males and 2 females from St. Vincent, 

 and 6 males and 4 females from Barbados, — 39 specimens in all are : — 

 wing 51.3 (48-52.5); tail 38.2 (36-40); exposed culmen 9.1 (8.5- 

 9.5); depth of bill 7.4 (7-8); tarsus 17.3 (15.5-18). While on the 

 other hand a series of 39 specimens, 26 males and 13 females from 

 Jamaica, representing T. b. marchii, the average is : — wing 52.4 

 (50-54.5); tail 40.6 (38.5-42.5); exposed culmen 8.9 (8.5-9.5); 

 depth of bill 6.8 (6.5-7.5); tarsus 17.1 (16.5-17.5). 



In coloration it differs widely from T. h. marchii in that the dark 

 area of the breast is not sharply divided, but merges gradually into 

 the white of the belly-region with often scattered spots in blotches of 

 darker color, extending posteriorly along the sides of the belly and 

 encroaching considerably into the white area. 



It is similar to T. b. omissa but has a longer tail and different color- 

 ation. Clark (Proc. Bost. soc. nat. hist., 32, p. 286) considers both 

 the Grassquit from Grenada and St. Vincent as referable to T. b. 

 omissa. The series of skins before me, however, show that grass- 

 quits from Grenada, the Gernadines, Barbados, and St. Vincent differ 

 from those of the other islands of the Lesser Antilles by having more 

 white on the belly. The tails of the former birds are relatively longer 

 averaging 38.2 against 37.1 of those of the latter. These characters 

 are rather constant throughout and are sufficient, I believe, to dis- 

 tinguish separate geographical races. 



I have examined a series of twenty-seven skins from other islands 

 of the Lesser Antilles but fail to find any characters upon which to 

 separate the bird occurring on the islands from St. Lucia to Porto Rico 

 from the mainland specimens. It seems anomalous to find T. b. omissa 

 in Tobago and Venezuela'and then skipping Grenada, the Grenadines, 

 and St. Vincent occurring again on the other islands of the Lesser 

 Antilles. But until a larger series of skins can be examined it is 

 perhaps premature to cite this as a case of convergent evolution. 



Ridgway (Bull. 50, U. S. N. M., 1901, pt. 1, p. 539) records two speci- 

 mens of T. b. omissa taken in Cuba. Since there are no other Cuban 

 records I incjuired into their authenticity. Mr. C. B. Cory, of the Field 

 Museum, informed me by letter that the birds were given to him by 

 Gundlach and Cory supposed that they came from Cuba. Like so 

 many other of Gundlach's birds they were probably collected during 

 one of Gundlach's three trips to Porto Rico. The Grassquit is not so 

 abundant on Guadeloupe as in the northern Lesser Antilles. On 

 Guadeloupe it is confined to the lowlands where it prefers the hot 



