ARKANSAW SISKIN. 
129 
and it is so similar to the European, that it might, with 
a much greater degree of propriety, be considered as a 
variety, than those regarded as such by authors. They 
can, however, be easily distinguished by the following 
comparative characters : All the under parts of the 
Arkansaw siskin are bright yellow, whilst the corre- 
sponding parts of the European siskin are tinged with 
greenish, the throat being black, and the belly, vent, 
and flanks, whitish, spotted longitudinally with black ; 
the margins and spots of the wing and tail-feathers are 
white in our bird, and yellow in the European siskin ; 
the white spots on the tail of the Arkansaw siskin are 
confined to the three outer feathers, whilst, in the 
foreign bird, all the feathers, excepting the two middle 
ones, are marked with yellow ; the bill of our species 
is also a little shorter, less compressed, and less acumi- 
nated ; finally, we may notice another trifling difference, 
which consists in the proportional length of the pri- 
maries, the four first being nearly equal in the American 
bird, and the three first only in the European, the 
fourth being almost a quarter of an inch shorter. The 
other approximate species, Fringilla magellanica , Vieill. 
considered by Gmelin and Latham as a variety of the 
European siskin, is readily distinguishable by having 
the head entirely black. 
Though the Mexican siskin ( Fringilla mexicana , 
Gmel.) may prove to be the female of our bird, or the 
male in an imperfect state of plumage, (and, from the 
locality, we should possibly have referred it to that 
name, had the classification of it fallen to our lot,) yet, 
as nothing positive can be drawn from so unessential 
an indication as that of the Mexican siskin, we have no 
hesitation in following the same course with Say, who 
Considers it as entirely new, and have retained his 
elegant name of Fringilla psaltria, It is very possible 
that not only the Fringilla mexicana , but also the 
Black Mexican siskin, ( Fringilla catotol , Gmel.) may be 
the same bird as our Fringilla psaltria ; but how can 
we determine, from the vague descriptions that have 
been given of those species ? They are equally appli- 
VOL. iv. i 5 
