MAMMALS OF NORTHERN COLOMBIA—HERSHKOVITZ 129 
be questioned, at least with respect to elegans Lund, but unless the 
types, if still in existence, show anything to the contrary, these 
conclusions must be accepted. On the basis of the original description, 
Echimys macrourus Jentink cannot be identified with Proechimys. 
The type is a skin only from Surinam with the following measurements: 
Head and body, 221; tail, 320; ear, 25; hind foot,41 mm. The tail is 
too long, both actually and in proportion to combined head and 
body length, for Proechimys as known; the hind foot is too small for 
any Proechimys of comparable head and body length. These measure- 
ments may have been taken from the dried skin; hence they are not 
really comparable. Tate (1939, p. 180) admitted the possibility 
of macrourus being a form of Echimys armatus. This opinion is 
probably correct. 
Leaving to one side the nomenclatorial problems that may arise 
once the exact status of each of such forms as dimidiatus and vacillator 
is entirely clarified, we find the subgenus Proechimys to consist of four 
recognizably distinct species. These are iheringi, canicollis, hendeet, 
and the composite P. guyannensis. To these are added two more 
species, one described as new, which form a group distinguishable by 
the enamel pattern of the molariform teeth. 
CHARACTERS OF THE SPECIES 
It has been found that external characters are not wholly reliable 
in distinguishing the species of typical Proechimys from one another. 
Differences in coloration are evident when comparing representatives 
of two or more species from the same or nearby localities. In other 
localities the differences between the same species may be reversed! 
The character of the spines is important in distinguishing super- 
specific categories of the Echimyinae, but in Proechimys (sensu stricto) 
it is at best only of some slight relative diagnostic value. Size 
differences between the species cannot easily be demonstrated. 
There may be some evidence of gradients in size within any one species, 
as well as some proportional differences, particularly in the length of 
the tail to the combined head and body length. Taken as a whole, 
however, no significant differences in size exist between any of the 
species. Most cranial characters, which are absolute for distinguish- 
ing two or more species from each other when these species are from 
the same or nearby localities, break down or reverse themselves when 
applied to representatives of these same species from widely separated 
localities. Of all the characters studied and of those described as 
diagnostic, only three or four seem to be of some value. In canicollis 
the walls of the mesopterygoid fossa are considerably fenestrated, more 
so than in the other species where the tendency is for no fenestration 
at all. In hendeei, dimidiatus, and iheringi the palatal notch extends 
anteriorly beyond the plane of M®’; in the other species it does not 
