978 DB. 0. 1. ronsYTH majob on a [Dec. 1, 



the frontal joins the maxillary, thus separating the lachrymal and 

 nasal ; and, besides, there is such a variability in the size and 

 mutual connections of the bones in this part of the skull in young 

 specimens, and, to a certain extent, in adult ones as well (four 

 nasals, obliteration of the Jachrymo-frontal suture, &c.), that we 

 cannot here enter into more particulars. 



My purpose \\'as to show that, in respect of the above characters 

 also the Alalagasy IJippojiotaml are intermediate between If. siva- 

 lensis and //. ampJiihws, and appear to be in close relationship 

 with both. Occasionally young specimens are hexaprotodont, as 

 the Siwalik forms. 



I think that, from what I have stated, we are fairly entitled to 

 surmise that the Hippopotami entered Africa at a time when they 

 were still in possession of all the characters of the Siwalik species, 

 and that they crossed over to Madagascar when they had reached' a 

 condition intermediate between //. siva'knsis and H. cmrphibius. In 

 this condition they persisted in Madagascar, whilst on the neigh- 

 bouring continent they progressed (or retrogressed) farther in the 

 same direction. It is a curious circumstance that the ITippopo- 

 tannis major from the Upper Pliocene of Italy has gone beyond 

 //. amphibius in the same specialization ; this may have had 

 something to do with its earlier extinction. 



EODBNTIA. 



But little attention has hitherto been paid to the Eodent fauna 

 of Madagascar. Although four or five more or less nominal genera 

 had been founded, it has been surmised that the Malagasy Rodentia 

 have immigrated in recent times and are not even specifically 

 Madagascar genera. This supjiosition rests on the assumption tliat 

 the Eodents are, as a rule, jjassively wandering (AVallace, Zittel), 

 and was made in spite of Peters having long ago pointed out that 

 Nesomys, the first known Malagasy Rodent, resembles the American 

 Ilesperomyes in the conformation of the enamel and in the pro- 

 portions of its molars '. 



My collections contain some hundred specimens of Eodentia, 

 belonging to five genera and eight species, five or six of the 

 species and two genera being new, besides two new genera found 

 in a fossil condition. This material I have begun to \^'ork out, 

 and, although my investigations are far from being completed, I 

 do not thinlc that the following conclusions will hereafter have to 

 be modified in their main points. 



The great majority of Malagasy Rodents at present known, viz. 

 the genera Nesomys, Ilallomys, Gymnuromys, EHurus, Brachy- 



• " Eine neue Gattung der Miirinen auB Madagascar, welchein dem Zahnbau 

 sich am niiebsten den Hesperomys der westlicben Hemisphiire anBcliliesst, und 

 so ein neues Beispiel von der geographisch so merkwiirdigen Verwandtsehaft 



der Fauna TOn Madagascar niit der von Amcrika liefert Die Backziihne 



?il?' '" 'I'l'*'' Schmelz bildung und Proportion iibnlicb deuen von Hesperomys." 

 (Sitzuugsber. Qes. naturf. Freunde Berlin, Oct, 18, 1870, pp. 54, 55.) 



