130 PROCEEDINGS OF THE NATIONAL MUSEUM VOL. 97 
extend beyond M*. It is possible that proportional differences in 
cranial parts exist between the species, but this cannot be demon- 
strated on the basis of present material and would be most difficult to 
show with any amount of specimens. The individual and local 
variation among these rats is so great that it is practically impossible 
to select a sufficient number of truly comparable skulls for determining 
proportional differences. 
The enamel pattern of the molariform teeth offers a reliable means 
for classifying the species of Proechimys, but even here the variations 
frequently introduce difficulties into any attempt to interpret cor- 
rectly the nature of the structures concerned. Nevertheless, the 
species of typical Proechimys, and annectant forms, may be classified 
primarily according to the following formulae, which show the number 
of outer enamel folds in each of the upper molariform teeth and the 
inner folds of the lower teeth: 
2-2-2-2 , 3-2-2-2 
A. 5-9-9-9 *° 3-995 
5 eS Sg 0 ie ately at lpia A tc UI St feeb jel Sell tweed ye P. canicollis 
mi Spmeytabore bate ein poe welt be ha ord Trinomys (subg.) 
3—3-3-2 3-3-4-4 
B. 3-9-9-9 © £37373 
1. Palatal notch extending forward beyond anterior plane of M3 
a. Zygoma broad, outer surface of jugal nearly plane, the ridge almost 
DORNER OM Ec Ps SST ee Se eae Bee 1 eee iheringi 
eee normal, outer surface of juga] crossed by a well-developed 
MOU Hcp crite, et hale! Mee hye P. hendeei (and dimidiatus, type) 
2. Palatal notch not extending forward as far as posterior plane of 
MEG 0 01 Foeaait (hy Ail ee Sia fh P. guyannensis (composite) 
tO. 
Corse seat doraeriiets- ae araprocias- 2+" see Se Quadruplicatus group 
(P. ignotus, P. quadruplicatus, n. sp.) 
p, =**4 Hopl ( 
0 ay ey ee rn a eee oplomys (genus) 
Four folds, as in the Quadruplicatus group, appear to be the maxi- 
mum if not the original number in the upper molariform teeth. This 
has been pointed out by Winge (1888, p. 86) in his description of the 
teeth of P. guyannensis from Lagéa Santa. In this species a fourth 
fold may be evident on the crown of the last two upper molars. It 
usually appears as a complication or branch of the third fold. With 
wear, the point of union between the two folds may be erased, with the 
result that the fourth fold becomes an enamel island. On the other 
hand, in some populations of guyannensis there is a tendency for the 
loss of even the third fold of the last upper molar and in each of the 
lower molars. In canicollis this is carried to an extreme. Here the 
third fold is found in the lower premolar only in certain populations. 
Throughout the genus the general trend is toward a greater simplifi- 
