NATURAL SCIENCES OP PHILADELPHIA. 375 



the genu?, and not existing in any other, though there is seen an approach to 

 it in Cannabina. 



In the type of the genus the bill is exceedingly acute and much compressed, 

 the lateral outlines even concave. From this character of bill, there is seen 

 through A. Ilolbolli and fuscescens, a gradual transition to the A. rostratus, 

 where the bill is much larger, less compressed and acute, and more full and 

 turgid. The plumuli are also considerably shorter and more scant; and the 

 whole appearance of the bill much that of a Cannabina or even of a Carpodacus. 

 The wings are very constant throughout the genus, differing scarcely appre- 

 ciably in length or pointedness in the different species, though the proportions 

 of the primaries vary considerably in the same species. Exactly the reverse, 

 however, is the case with regard to the feet ; i. e. the characters differ con- 

 siderably in the different species, but always are quite constant in each. Thus 

 in linarius, ru/escens, Ilolbolli, fuscescens and rostratus the proportions of tarsus 

 and toes are much the same, the difference in absolute length being only pro- 

 portional to the size of the birds. In exilipes the feet are smaller and weaker, 

 and the toes shorter, the difference being specially noticeable in the middle 

 toe, which, with the claw, is shorter than the tarsus. In canescens, on the other 

 hand, the feet are larger and stronger, even more so than is proportional to 

 the greater size of the bird ; but the toes are excessively short, so much so 

 that even the unusually long claw does not make the middle toe equal to the 

 tarsus. The tail differs but slightly, if at all, in the different species ; for 

 though A. canescens is spoken of by authors as having a comparatively longer 

 tail than has linarius, the difference seems hardly more than is proportional to 

 the greater size of the bird. 



With respect to color, the species present a remarkable similarity, not only in 

 the pattern of coloration, but also in the tints. In all, there is found the crim- 

 son pileum, which varies in size, and in the intensity of the color, with sex and 

 age. It has sometimes a peculiar coppery or brazen reflection, very different 

 from the usual deep crimson tint. The breast, with the sides of the head and 

 body to some distance, as well as the rump, are tinged with rosy or carmine. 

 In the examination of a great number of specimens I have noticed a fact that I 

 have seen nowhere stated. It is that the depth and intensity of the color on 

 the breast and rump is in direct proportion to the lightness or darkness of the 

 general colors of the bird. Thus, in rostratus and fuscescens, the rosy ou the 

 breast becomes so bright as nearly to equal in intensity the crimson of the 

 crown. Canescens and exilipes present the other extreme, the rosy of the breast 

 being very light, scarcely more intense than that on the rump. Linarius, and 

 rufescens and Ilolbolli are intermediate between the two extremes in this respect. 

 The females of every age, and the very young males, either want entirely this 

 rosy on the breast and rump, or else show but very slight traces of it. All the 

 species are streaked above ; the feathers having very dark centres and light 

 borders; but in fuscescens and rostratus these borders are so narrow that the 

 parts appear almost tiniformly dusky. In canescens and exilipes these streaks 

 disappear on the rump, leaving that part pure white ; in the other species of the 

 genus the rump is thickly streaked. In all, the sides of the body are more or 

 less streaked with dusky ; and here the same rule holds good as with regard 

 to the rosy of the breast. In the darkest colored species A. fuscescens and 

 rostratus the streaks are most numerous, darkest, and most distinctly defined : 

 they become less numerous and distinct in linarius and exilipes, and are some- 

 times almost wanting in canescens. In the females these streaks extend quite 

 across the breast. 



It will thus be seen that the species of the genus are all very closely related ; 

 the characters, when taken from the colors, being chiefly those of intensity, 

 and when based upon form, being found in the varying combination of several 

 features. .4. rostratus, indeed, differs from the others in the possession of a 

 much larger and more turgid bill ; but as the other characters agree strictly 

 with the type, and especially as the transition from one extreme to the other, 



1861.] 



