10 ART. f..— S. YOSHIWAKA AND J. IWASAKI. 



2.9'"' in rlic lonijor und shorter ;ix(js (jn tlie musticatinir .surture. 

 Moreover tlie pillars are not arranged in three transverse rows, and 

 tliere is no sign of either its anterior or its posterior face having* been 

 in «'ontaet with another tootli. The outer surface of one of the two 

 longitudinal halves containing three tubercles is convex, but that of 

 the other half is almost fiat. The Avearinii' is rather slii«ht and the 

 dentiiie is exposed onlv in a small round section, Tlie two-rooted 

 «•haracter of the base is distinctly shown in the tooth examined. 

 Comparing the form of the present tooth Avith that of the above 

 described specimens, there is among these not one which quite agrees 

 with it in the ai-rangement of the column-like tubercles. In the size 

 of the crown the tooth somewhat approaches the second premolar of 

 the other specimens, but there exists a great différence in the number 

 of tubercles. Therefore it is not certain whether we have before us a 

 premolar or a molar, a milk tooth or a permanent one. 



PHYLOGENETIC VIEW OF THE PRESENT SPECIMENS. 



Our specimens decidedly belorig to a new genus. Among the 

 Proboscidea there are at present two families, viz.. Dinotheridaî and 

 Elephantida?. The former contains <jn!y one genus called D'uiothe- 

 riuiii, wliich is found in the upper and middle Miocene (or even in 

 the 1 Miocene?) of Europe and the East Indies. The animals which 

 are commonly known as elephants are all included in the latter 

 family which comprises tliree genera Mastodon. Stegodon and Elephas. 

 The chief characters Avhich distinguish those families fi'om one another 

 seem to me to be the following : 



(1) In the Dinotheridaî a ])air r)f incisors appear only in the 

 lower jaw, aiid are directed vertically downwards in consequence of the 

 abru])t downward flexure of the front part of the mandible, while in 



