12 
DESCRIPTION OF PSEUDOMYS NOVAE-HOLLANDIAE 
Teeth. — The cingular cusp of hermannshurgensis actually 
varies to a considerable degree (see pi. II). in some 
specimens, shoAvs well-marked convexity of contour at the 
base of the cusp, whilst in others this is almost absent. The 
teeth of novae-hollandiae resemble those of hermmmsburgen- 
sis in shape, and in tilting of laminae. The cinglar cusp is 
variable, and is indistinguishable in its variation from that of 
the latter species. 
Colour of pelage . — Though hermannsburgensis is generally 
warmer in tone than novae-hollandiae the colours intergrade, 
and a specimen of the fomier sxDecies from Barrow Creek, 
Central Australia, is a perfect match for the average tone 
of the latter. 
Conclusions. 
On the above facts it will be seen that there is no good 
reason for separating the Central and South Australian forms 
of hermannsburgensis. The aiiproximation of their aA^erage 
measurements, together Avith the variation found in each, 
makes it impossible to differentiate betAveen them. 
It is also clear that novae-hollandiae and hermannsburgen- 
sis are very closely allied. Length of tail is the only feature 
in Avhich they differ consistently. Other measurements and 
colour intergrade so completely that there Avould seem to be 
little reason for separating them, except difference in length 
of tail and possibly geographical distribution. 
The teeth of novae-hollandiae agree more closely with the 
characters described for Leggadina than with those for 
Gyomys, and there is no doubt that it must be assigned to that 
subgenus. 
Finlayson (4) has already expressed a doubt as to the con- 
stancy of Thomas’s subgeneric characters, and suggested that 
his (Thomas’s) division was intended as a tentatWe step only. 
Whether the latter suggestion is, or is not, correct, the varia- 
bility of many skull and teeth characters must be admitted. 
Owing to the fact that considerable numbers of specimens are 
needed to explore fully the variation of cranial and dental 
characters, it is probable that no single institution has the 
necessary material for a complete revision of the genus. It is 
therefore urged that museums possessing extensive series, 
even of a single species, should publish a critical survey, so 
that the whole can later be correlated and a much-needed 
revision undertaken on reliable evidence. Until such time 
