980 ON A ZOOLOGICAL EXPEDITION TO MADAGASCAR. [ Dee. 1, 
follows :—Cricetus is a terminal form amongst its congeners, some- 
what connected with the Malagasy Nesomyine by means of the 
miocene Cricetodon, from which it is probably directly derived. 
I have lastly to consider the affinities of the Malagasy Rodentia 
with the American Hesperomyes, urged by Peters with regard to 
the molars of his genus Wesomys. The resemblance is certainly 
striking between the pattern of the Mesomys-molars and of some 
of the bunodont Hesperomyes, and this resemblance extends alike 
to the form of the skull in both groups (conformation of the 
boundaries of the infraorbital foramen, small size of the tympanics, 
&e.). There exists more agreement between these two groups 
than between them and Cricetus and its Old-World allies. On the 
other hand, the Malagasy Rodents present unmistakable family 
features of their own, which all point in one direction, stamping 
them as lower, more primitive Muride than the Hesperomyes. The 
two anterior, and in some cases all three, molars are more like 
each other in size and pattern, although in this respect the 
difference is much less striking between the American Hesperomyes 
and the Malagasy Nesomyes than between the first and the Murine. 
Similar remarks apply to the skull: the infraorbital foramen 
is larger in the Malagasy mice, the malar bone always stronger 
developed and extending farther forwards and upwards towards 
the lachrymal &e. 
Zittel is of opinion that all the Myomorpha of South America 
are recent, having immigrated from the north towards the end of 
the Diluvium?. The reasons for supposing this are, that so far 
only hystricomorphous Rodents have been met with in the older 
formations, the (Patagonian and) Santa Cruz beds. Here it must 
be borne in mind that the presence of a premolar and the pattern 
of the molars approaching the Hystricomorpha are not sufficient 
criteria for assigning to these latter several small Rodents of the 
Santa Cruz beds, of which the only parts preserved are the teeth. 
On the contrary, if there exist forerunners of the Muride in the 
Santa Cruz beds, they are likely to have possessed premolars. If 
the Rhizomyes and the Malagasy Brachywromys possessed pre- 
molars—and there is strong evidence that this was recently actually 
the case—we would be inclined, without knowing more of them 
than their molar series, to assign them to the Hystricomorpha. 
As matters stand at present, it must be admitted that all 
appearances speak against the ancient domicile of the Hesperomyes 
in South America, whereas we have in the miocene of North 
America such forms as Zwmys and others, which might be con- 
sidered to be the ancestors of the Hesperomyes. Moreover, Crice- 
todon of the European Miocene is more closely related to the 
Hesperomyes than to Cricetus. 
There is therefore at present not sufficient evidence of a direct 
relationship between the Malagasy Rodents and the western 
Hesperomyes, although it seems to me difficult to explain their 
affinities as a result of mere convergence. 
1 Handbuch der Palaeontologie, 1, iv. p. 556. 
