FRIXGILLIDJ-: — THE FINCHES. 43 



streaks of mixed brownish and du?i<y. Bend of wing edged with light yellow. Wino-^ 

 2.55 ; tail, 2.80 ; bill, .28 from nostril and .23 deep ; tarsus, .68 ; middle toe, .55. 



Young. (45,277, Laredo, Texas, June 28.) Very similar, but with a few drop-shaped 

 streaks of dark brown on the jugulum and along sides. The feathers above have a more 

 appreciable terminal border of buff. 



Hab. Rio Grande region of Southern Middle Province; Kansas, breeding (Allen). 

 San Antonio, Texas, summer (Dresser, Ibis, 1865, 489 ; eggs) ; ? Orizaba, temp. reg. (Scir. 

 M. B. S. I, 551). 



In the Birds of North America, the specimen characterized on p. 637 of the 



present work as mstivalis, var. arizoncr; was referred to P. cassini, those 

 specimens which are here retained as such being considered as in quite im- 

 mature phimage. A more recent examination of additional material, how- 

 ever, has compelled us to change our view. In consequence of the similarity 

 of the specimen in question to ccstivalis, as noted in the article referred to 

 above, the general acceptation of the name cassini has been that of a term 

 designating a variety of the common species ; but we have as the result of the 

 investigation in question found it necessary to retain under the head of 

 " cassini " only the typical specimens from the Eio Grande region, and refer 

 the supposed aberrant specimen to astivalis. In this Los Nogales specimen 

 we find existing such differences in proportions and colors as are sufficient to 

 waiTant our bestowing upon it a new name, and establishing it as the Middle 

 Province race of ccstivalis, in this way connecting the South Atlantic and 

 Mexican races (var. a'stivalis and var. hottcrii) by a more similar ibrm than 

 the P. cassini, which must be set apart as an independent form, — in all 

 probability a good species. Several facts are favorable to this view. First, 

 we have of the P. cassini specimens which are beyond question in perfect 

 adult plumage, and others which are undoubtedly immature ; they differ 

 from each other only in such respects as would be expected, and agree 

 substantially in other characters, by which they are distinguished from the 

 different styles of ccstivalis. Secondly, the region' to be filled by a peculiar 

 race of ccstivalis is represented by the var. arizonm, which is undoubtedly 

 referrible to that species ; thus we have in one province these two different 

 forms, which therefore are probably distinct. 



The present bird is hardly less distinct from the races of ccstivalis than is 

 ruficeps ; and we would be as willing to consider all the definable forms pre- 

 sented in the synopsis as varieties of a single species, as to refer the present 

 bird to ccstivalis. 



Habits. This Finch, in its general appearance, as well as in respect to 

 habits, nesting, and eggs, is quite similar to Bachman's Finch. It was first 

 met with by Dr. Woodhouse, in the expedition to the Zuni Biver, when he 

 found it in Western Texas. He shot it on the prairies near San Antonio, on 

 the 25th of April, 1851, mistaking it for Passerculus savanna, which, in its 

 habits, it seemed to him very much to resemble, but upon examination it was 

 found to be totally distinct. 



Dr. Heermann afterwards, being at Comanche Springs in Texas, had his 



