218 Bulletin American Museum of Natural History. [Vol. XXXIV, 
from each of a dozen or more different localities. The individual variation 
in coloration and size has already been sufficiently indicated. 
The form of the nasals in most mammals is usually considered an im- 
portant diagnostic feature, but the range of individual variation in most 
species of squirrels (and particularly in the hof'manni group) is so great 
that the nasals afford no sure basis for the discrimination of local forms. 
In average specimens of hoffmanni the nasals are squarely truncate pos- 
teriorly and terminate practically on a line with the fronto-premaxillary 
sutures, but they often end a little in front of this line, and sometimes extend 
behind it. In other specimens they terminate far in front of it, with the 
posterior border slightly or deeply v-shaped, or even nearly square. In 
extreme cases the nasals terminate one fifth to one fourth of their length 
anterior to the fronto-premaxillary suture. 
Formerly hoffmanni, as shown by the synonymy given above, was con- 
sidered to be inseparable from @stuans, or merely a subspecies of it. While 
resembling estuans in a general way in coloration, it is a much larger species, 
with a broader and relatively shorter tail, and is wholly separated from it 
geographically by the griseogena group. 
The type locality of Sciurus estwans var. hoffmanni was not definitely 
indicated in the original description, being given as “Costa Rica.” Since 
two forms of the hof/manni group occur in Costa Rica, chiriquensis of Bangs 
ranging from Chiriqui into the low coast districts of Costa Rica, it will be 
a convenience to restrict hoffmanni to the higher altitudes, or to the areas 
above 3000 feet, and to consider the type locality as San José, in the Volcan 
de Irazti district. This is warranted by the fact that Dr. Hoffmann, who 
collected the type, is known to have collected for a considerable period at 
San José during his explorations in Costa Rica. 
As thus restricted, specimens of hof'manni from Costa Rica are indis- 
tinguishable from representatives of the hoffmanni group from the upper 
Rio Cauca valley, as recognized by Nelson (J. c.) in 1891 and by me (I. c.) 
in 1912. Specimens since received show that true hoffmanni ranges not 
only through the southern Western Andes of Colombia but also into the 
southern part of the Central Colombian Andes, and thence southward in 
the Andean region through central Ecuador to the northern border of Peru. 
Series of specimens from widely separated localities in this large area are 
_ practically indistinguishable, although it may be possible later to add slightly 
differentiated local forms to those here recognized, from points in the 
general area not represented by the material at present available. 
ee ey eae ene ee ce a een ee eeu nem aN ee 
1Cabanis, Uebersicht der im Berliner Museum. befindlichen Vogel von Costa Rica. 
Journ. fiir Orn., 1860, 1861, 1862, passim. 
