1897.J MALAGASY GBJS^US BEACHYUliOMYS. 703 



3. Sijjhneus. — Gerrit Miller has recently expressed the opinion 

 that it may eventually prove necessary to unite the " Siphneinoe," 

 i. e. Siphneus and Ellohias, with the Allcrotince '. This corresponds 

 fairly with Alphonse Milne-Edwards's views, who has considered 

 the ^^Siphnes" to be '• des Arvicoles anormaux,"' O. Thomas 

 maintains the subfamily for SipAneus alone, since according to 

 his view its " diiferenees, both external, cranial, and dental, are 

 clearly sufficient to demand separate subfamily rank."^ 



It is to be remarked that Winge has treated the question of the 

 relationship of Siphneus exhaustively many years ago. Whilst 

 uniting Ellobius with the Microt'uue, he considers Siphneus to be a 

 lower type, and accordingly places it w"ith the Cricetince; its 

 resemblance with the Voles resting solely on their having open 

 roots to the molars, which otherwise are not different from the 

 "Hesjjeromys-tjpe." He points out that no Vole has such a smaU 

 m. 1 inf., with not more than the usual five loops. Winge showa 

 besides, that Siphneus lacks the powerful crest in the wall of the 

 temporal fossa, which gives the characteristic feature to the 

 Microtine skull, whereas the temporal muscles are inserted on the 

 surface of the skull in the same manner as in the Criceti ; and that 

 the basioccipital region aud the centrum ot" the hasisphenoid are 

 broader, the tympanic bones smaller, than in the Voles ^. In 

 the more simple structure of the molars he sees an indication of 

 closer relationship with " Cricetulus.'^ The shape of the outer wall ' 

 of the infraorbital canal is said to be about as in the Hamsters ; 

 likewise the zygoma and the crests on the cranium, only slightly 

 stronger ; and equally the flattening and forward inclination of the 

 occiput, the only difference being that these characters too are more 

 strongly developed than in the Hamsters ^. 



It seems to me that there is little to object to AVinge's view 

 of the question ; I would even go a little farther still. In his 

 arrangement of the Muridas^, Winge opposes his JRhizomyini 

 (i. e. Cricetodon, Eumys, Rhizomys) to the rest of tlie Muridfc, m. 1 

 in the former being only slightly larger, in the latter considerably 

 larger than m. 2. 



As regards this character, Siphneus certainly belongs to the 

 former group, with more right than Cricetodon and Eumys, which 

 both herein are scarcely different from the Hesperomyes. In 

 pattern and size the two anterior molars of Siphneus agree as 

 much with each other as they do in SpuJax, Tachyoryctes, and 



' Gerrit S. Miller, " Genera and Sub^ienera of Voles and Lemmings'' (North- 

 Ainerican Fauna, No. 12, p. 8, footnote 3). Washington, 1896. 



^ H. Milne-Edwards et Alph. Milne-Edwards, ' Eecberch. p. servir a I'Hist. 

 nat. des Mammiferes etc.,' t. i. Texte. pp. 76-79, Paris, 1868-74. 



^ O. Thomas, "On the Genera of Rodents" (P. Z. S. p. 1021 footnote 1). 

 London, 1896. 



'' This does not hold good, however, with regard to all the genera of 

 Microtina, e. g. Fiber and Ellobius. 



' H. Winge, " Om grseske Pattedyr " (Videnskab. Meddel. fra d. Natiirh. 

 Foren. Kjobenhavn for Aaret 1881, pp. 47-49. Kjobenhavn, 1882 ; see also 

 H. Winge, ' Gnavere fra Lagoa Santa,' pp. 124, 125, 126). 



^ ' Gnavere fra Lagoa Santa,' p. 125. 



46* 



