128 PROCEEDINGS OF THE NATIONAL MUSEUM VOL. 97 
of the subgenus Proechimys presented by Ellerman (1940, pp. 115- 
' 122) m his monumental treatise of the rodents of the world shows a 
sharp reduction in the number of species. In typical Proechimys, 
Ellerman listed six species in the ‘“cayennensis [=guyannensis] 
group” and one, P. rtheringi, in the ‘“iheringi group.” The subgenus 
Trinomys consisted of albispinus and setosus. Reasons for making the 
above divisions of the genus were given. The discussion here is 
limited mainly to the species of the guyannensis group and to the 
species of the subgenus Proechimys listed by Ellerman as not seen 
and not allocated to group. 
The six species of Ellerman’s “‘cayennensis group” are guyannensis, 
vacillator, hendeei, rattinus, canicollis, and dimidiatus. The species 
- guyannensis, as constituted by Ellerman, will doubtlessly prove to be 
composite. He admitted the possibility of error in listing under 
guyannensis the Proechimys described from Central America, in this 
following the tendency among authors to assign these forms to a 
single species. However, there are at least two kinds of spiny rats 
common to Central America and northwestern South America. To 
determine what names are applicable to them would require a more 
careful examination of the type specimens in the British Museum 
than the present author has been able to make. It remains a moot 
point which ones, if any, of the described Central American forms are 
indeed referable to P. guyannensis. Under the circumstances it is 
best to accept provisionally Ellerman’s classification and to continue 
treating P. guyannensis as a composite species. Possibilities for its 
ultimate subdivision into natural entities will be pointed out in 
following discussions. P. canicollis differs in important characters 
and appears to be annectant between P. guyannensis and the subgenus 
Trinomys. P. vacillator may prove to be a race of canicollis. P. 
hendeei is very nearly related to guyannensis but distinct specifically. 
The type of P. rattinus, a skull only, is that of a race of guyannensis. 
P. dimidiatus is, according to the writer’s notes, an immature indi- 
vidual and presents no special peculiarities other than the very deep 
incision of the posterior palatal notch, on a plane with M?. It is 
certainly a member of the group, but it cannot now be determined 
whether it is a local form of the common species or whether it replaces 
the name of a more recently described species, possibly hendeei. 
Of the 12 species listed by Ellerman as not seen and not allocated 
to group, the present author has been able to examine ochraceus, 
poliopus, oconnelli, steerei, boimensis, and kermiti. These are all 
referable to guyannensis and are discussed under that specific heading. 
P. leucomystax Miranda Ribeiro, as described and figured, is also 
a representative of the common species. According to Thomas 
(1921, p. 141), myosurus, leptosoma, cinnamomeus, elegans, and 
fuliginosa are all synonyms of setosus, a Trinomys. This opinion may 
