an example from Bahia in Brazil shows the characteristic hoary forehead and whitish 

 sides of neck which are among the chief features of P. hesperia. A male in the 

 second year from Washington, killed on the 30th of May (Henshaw collection), also 

 closely approaches P. hesperia in the above-named particulars. Both these specimens, 

 however, are true P. purpurea, as they are not so white on the throat, are more dingy 

 on the breast and abdomen, and have the long under tail-coverts very distinctly centred 

 with brown ; this last character seems to be a very well-marked one for distinguishing 

 P. purpurea. 



Besides the typical specimens from the Sierra de la Laguna, Mr. Brewster says that 

 he has seen others in Mr. Batchelder's collection from the Ojai Valley, in California, and 

 Mr. Xantus's birds from Cape St. Lucas are also P. hesperia, as is shown by a specimen 

 in the British Museum. 



It is extremely difficult for us to determine the exact range of this western race of 

 Purple Martin owing to lack of specimens, but we have some doubts whether the 

 distinctive characters of P. hesperia will be upheld by future observers. Judging from the 

 small series at our disposal, P. purpurea is at certain seasons very similar to P. hesperia, 

 and the latter can be nothing but a western race, which, indeed, is all the status that 

 Mr. Brewster claims for it. The British Museum possesses a young male from Big 

 Trees, obtained by Mr. Porrer, which is certainly referable to P. hesperia, but amongst 

 the series in the Museum there are several birds which seem to connect the two races 

 entirely. 



Mr. L. Belding, in his paper on the Birds of the Pacific District, claims for the 

 western race a range reaching even into British Columbia, but we believe that the 

 Purple Martin from several of his more northern localities must be P. purpurea. This, 

 however, is a question which the American ornithologists alone can settle. We subjoin 

 Mr. Belding's note : — 



"A dozen or more of both sexes were temporarily sojourning at San Diego, April 28, 

 durin 0, a cool rain-storm. It does not appear to breed on the coast about San Diego 

 (L. P.). 



" Poway, twelve miles from the coast. Pirst seen May 1, 1881, a few only ; common 

 in the spring of 1883 (Blaisdell). 



" Little Santa Maria Valley, April 1, 1881, one only (Emerson). 



" Julian, April 4, 1881 (A T . S. Goss). 



" San Bernardino. Bare summer resident in the mountains ; rare migrant in the 

 Valley (F. Stephens). 



" Santa Cruz, common [Joseph Skirm). 



" San Jose. Pirst seen May 3, 1881, two or three ; they did not remain. Arrived 

 April 9, 1885 (A. L. Parkhurst). 



" Contra Costa County. Bare summer resident ( TV. JE. Bryant). 



" Olema. Pirst seen May 8, 1881 ; breeds (A. M. Incjersoll). 



" Stockton. Common summer resident here and in many localities in Central 

 California below fir-forest, where it is very rare (L. P.). 



