244 U. S. P. R. R. EXP. AND SURVEYS—ZOOLOGY—GENERAL REPORT. 
maximum size, as S. vulpinus, cinereus, carolinensis, &c.; and that having S. hudsonius as its 
type, characterized by smaller size, shorter and thinner tail, and shorter muzzle. This short- 
ness of the muzzle is strikingly shown in the skull, which is very broad and full, much more 
convex longitudinally and transversely, and with the nasal bones one fourth the length of the 
skull instead of one third. 
The determination of the species of squirrels of North America has always been a matter of 
great difficulty, owing to many different reasons. The species themselves exhibit an unusual 
tendency to run into varieties of color, among which red, gray, and black are the predominating 
ones, with all possible intermediate shades ; these varieties are sometimes more or less constant 
in particular localities, sometimes changing with every litter. I am not aware that there is 
any material difference of color at different seasons or ages in the same animal. 
Another source of perplexity to the naturalist is the alteration in average size with the lati- 
tude. Many of our animals become smaller as we proceed southward, until, on the sea coast of 
Georgia, Florida, and the Gulf they reach their minimum. This is very strikingly seen in the 
common deer, which on the sea islands of Georgia is so small as to be readily lifted and thrown 
across a horse with perfect ease by a man of ordinary strength. 
It is in the Sciwridae that, next to the deer, we find this law to prevail most decidedly. 
Nearly all the species of extensive north and south range will be found, on careful examination, 
to substantiate this position. 
A similar variation in color to that of the squirrels is seen in the foxes and wolves, most 
strikingly in the former. It is now well known that the red, cross, and black foxes are identical 
in species, the same litter frequently embracing all the colors, and that some of these varie- 
ties again are more or less permanent, while the more boreal the locality the greater tendency 
to black. This is the case also with the squirrels, where the smaller species assume the black 
pelage to the greatest extent in the more northern portions of the United States. 
As a general rule it may be stated that where a squirrel exhibits any annulation of the fur on 
the throat or belly it is a variety of some species, typical specimens of which have the under 
parts either uniformly white or reddish to the roots, which, however, are sometimes plumbeous. 
In every such instance that has come under my examination I have had no difficulty in tracing 
it to its proper type. Such annulation is usually accompanied by a duskier color of the pelage. 
The tendency to annulation below is strongest in the squirrels of the Mississippi Valley, and 
applies both to gray and fox-colored species. No such instances of annulation have yet come 
under my notice among the squirrels west of the Rocky Mountains. 
As a general rule the bones of the entire skeleton of the fox squirrels, or those with rusty 
bellies, are red, while the white bellied varieties have them white. 
After carefully examining a large number of American squirrels I have reluctantly come to 
the conclusion that very many of the species now borne on our scientific records must be 
dropped, or reduced to the position of temporary or local varieties. I hope to be relieved from 
the imputation of wanton and unnecessary reduction in the number of species when I state that 
the species already published as new by myself have in nearly every case shared the fate of 
others older in date. The nominal species have usually been based on the varying size or dif- 
ferent colors of different specimens. The dusky varieties and those with the hairs of the under 
parts annulated have in nearly every instance been raised to the rank of a distinct species. 
In the following pages it will be seen that I recognize only twelve species of squirrels as 
satisfactorily proved to belong to the United States, and should S. limitis and castanonotus prove 
— 
