NATURAL SCIENCES OF PHILADELPHIA. 387 



variations to be as follows. The difference in total length of body is rather 

 more than -50 of an inch. The average size is rather less than 550 inches. 

 The difference in length of wing from the carpal joint is somewhat over -25 

 of an inch. In length of tail the differences are about the same. The feet are 

 constant, both absolutely and relatively, the variations being scarcely appre- 

 ciable. The claws, however, vary quite notably in length and amount of cur- 

 vature. To express the difference in the feet of this species and the A. Una- 

 rms, relatively without measurement, we have said that in the former the tar- 

 sus exceeds the middle toe and claw, and that in the latter the toes are 

 equal ; but this variation in the size of the claws may cause the fact not to hold 

 good in all cases. The bills vary quite notablj' in size and color, they being 

 sometimes nearly as bright yellow as in Unarms ; but they are usually almost 

 entirely dusky, and they are never so acutely poiuted and so much compressed 

 as in that species, preserving their regular conical shape without much varia- 

 tion. The plumuli vary in color, from dingy whitish to dusky, but are usually 

 of the former color, and are always heavy and full. The upper parts are 

 usually as given in the description ; lighter than in Unarms, the conspicuous 

 yellowish or buff of that species fading into whitish more or less pure. Some- 

 times, however, the upper parts are quite like Unarms ; in other specimens 

 the dusky centres of the feathers become so broad and conspicuous as to 

 give the prevailing color of the upper parts, causing the specimens to approach 

 A. fuse esc ens in this respect only. In these cases, however, the light edges of 

 the feathers, though so narrow, are nearly pure white, and the rump is very 

 broadly pure white, entirely without streaks, forming a marked contrast. 

 The edgings of the wings and tail do not vary notably, being always nearly 

 identical with those of Unarms, but perhaps a little purer. The rump in adult 

 birds is pure white, with a rosy tinge, without spots or streaks ; but in imma- 

 ture specimens it is frequently marked with dusky, though never so thickly as 

 in Unarms. The dusky streaks on the sides are usually very sparse, and 

 though varying in number and intensity, never become so thick and dark as in 

 Unarms. These streaks in immature and female birds frequently extend as 

 slight touches quite across the breast. In adult males the breast is immacu- 

 late, with a light rosy tint. In the fullest plumaged males the rosy is always 

 several shades lighter than in the males of Unarms of the same age. 



Comparison with allied species, A. Unarms being most closely related, in the 

 foregoing descriptions the comparisons have been made with that species. 

 The points of difference may be summed up as follows : The smaller, more 

 conic, less compressed, darker colored bill, with its very heavy and dense 

 nasal plumuli ; the different character of the streaks above ; the white imma- 

 culate rump ; the paucity of the streaks on the sides ; the much lighter rosy 

 tinge of the breast and abdomen ; the smaller, every way weaker feet, with 

 their much shorter toes. 



Some specimens the largest and lightest colored resemble A. canescens in 

 general appearance, having the same conic bill, heavy plumuli, white rumps, 

 sparsely streaked sides, &c. They may, however, be readily distinguished by 

 the great discrepancy in the size of the feet and claws, those parts in canes- 

 cens being as much larger and stronger in Unarms as they are smaller and 

 weaker than in that species in exilipcs. 



The species requires no comparison with A.fuscescens, still less so with A. 

 rostratus, the differences from both of those species being sufficiently obvious. 



The very large, bright yellow bill, with the short plumuli of A. IlolloUi, 

 at once separates that species. The smaller size, more acute bill, less dense 

 plumuli, general rufous tinge and shorter tail of A. rufescens will serve to dis- 

 tinguish it. Both these species, moreover, have the same character of feet as 

 has the A. Unarms. 



Discussion of synonymy. The only name which we have met with which can 

 be referred to this species, is the Fringilla, or Linaria borcalis Temm., of Au- 



1861] 



