1897.] MALAGASY GENITS BEACHTUUOMTS. 719 



The light which the investigation of the Malagasy Eodentia has 

 thrown on their relatives outside the Island has been somewhat 

 unexpected to me. Prom what Peters had noted concerning the 

 affinities of one of the genera ' and from other considerations, I 

 was prepared to meet with the nearest and perhaps the only close 

 affinities amongst the American Hesperomyincv. These affinities 

 certainly exist, and I have endeavoured to put them in their true 

 light. However, other affinities, apart from those just mentioned, 

 are very remarkable. 



The genus Brachytarsomys, which, as stated on a former occasion ^ 

 stands somewhat apart from the other Malagasy Eodents, proves 

 to be a forerunner of the Microfirue. It is, however, certainly not 

 a member of the genus Microtus, nor of any of the other genera 

 included in the subfamily ; it cannot even, in my opinion, be 

 placed within this subfamily, for it lacks the specializations which 

 characterize the latter. Apart from the molars being not only 

 rooted, but even perfectly brachyodont, neither the last upper nor 

 the first lower molar show any additional increase to the normal 

 Muridine form ; the skull, too, differs from the Microtine cranium 

 in all the characters, which in these are the direct outcome of 

 the increased vertical size of the molars and the adaptation to a 

 subteri'anean life. But otherwise the teeth as well as the 

 cranium (size and shape of the jugal, form of the rostrum, of the 

 outer wall of the infraorbital foramen and of the foramen itself, 

 general conformation of the upper region of the skull and its 

 crests) are precisely such as we might expect them to have been in 

 the forerunners of the Microtina'. 



Next as to the genus Nesomys. The large size and breadth of the 

 foramina incisiva, and, what is still more to the point, the lai-ge 

 size of the infraorbital foramen, and the strong development of 

 the jugal — which characters this genus shares with most of the 

 other Malagasy Eodents — show it to be a very low member of 

 the Muridce, approaching the Dipodidce. The two anterior molars, 

 agreeing in size and general form with each other, tell the same 

 tale. The intimate structure of the molars, as compared with the 

 Mesperomyince and the Murina, might induce us to consider 

 Nesomys as a connectmg-link between these two groups. But the 

 relationship to them will be more rightly expressed by considering 

 it to be ancestral to both ; especially if we bear in mind that the 

 characters of both the cranium and teeth are less specialized than 

 in the two subfamihes mentioned. 



The present paper deals chiefly with a third genus, BracTiyuromys. 

 Its affinities with some fossorial Eodenis, viz. Tacliyoryctes from 

 Abyssinia, Rhizomys from the Oriental Eegion, Spalacc and Siph- 

 neus from the Palaearctic, have been fully discussed, and as one 

 of the results these four genera are classed amongst the lowest 

 Muridce. Eetirement under the earth and adaptation to fossorial 

 habits have done for these four genera what isolation has done for 



1 Sitzungsber. Ges. naturf. Freunde Berlin, Oct. 18, 1870, pp. 54, 55. See 

 also P. Z. S. 1896, p. 978. 

 » P. Z. S. 1896, p. 979. 



47* 



