MURID2E— ARVICOLIN.E— EVOTOMYS EDTILUS GAPPERI. 147 



As usual in Arvicolince, the upper incisors are broader than they are 

 deep, ungrooved, and yellow on their faces. As in all Arvicolince, except 

 Myodes and Synaptomys, the under incisors run past the last molar, and reach 

 up the ramus half-way to the condyle itself. 



The molar series are not quite parallel, diverging a little from before 

 backward. Spite of their rootedncss, in which they resemble*' the teeth 

 of Murince, they are essentially constructed upon the plan of ArvicolincT, 

 being truly prismatic, with acute salient and reentrant angles, and Hat crowns 

 composed of triangles of dentine surrounded by enamel walls — these walls 

 meeting in several instances along the middle line of the tooth, and isolating 

 dentine islands, in other instances allowing contiguous dentine islands to 

 become continuous. Details of the molar crowns are as follow: — 



The front upper molar presents little or nothing characteristic, having the 

 form constantly preserved throughout the subfamily. There is an anterior 

 closed triangle or semilune, then an interior closed triangle, then an exterior 

 one, then an interior one, then a postero-external one — five in all. The 

 middle upper molar is the same essentially, but with only four alternating tri- 

 angles, of which the first after the anterior one is external, the second inter- 

 nal, the last postero-external. This is much as in the Pedomys and Pitymys 

 section of Arvicola, and not as in the riparian section, where the last triangle 

 develops a snag or spur from its inner face, making five in all, two of them 

 internal. The back upper molar is the most peculiar and characteristic of 

 all; indeed, it seems to be the case throughout Arvicolincv that this tooth is 

 diagnostic of the genera and subgenera; the sculpture of its crown certainly 

 differs more than that of any other tooth. In the present case, the tooth is 

 remarkable, first, for being absolutely longer antero-posteriorly than either of 

 the other upper ones, which is not the case in Arvicola. In general, it comes 

 nearest to the riparius type of Arvicola, having really the posterior crescentic 

 loop and two distinct external triangles, so characteristic of riparius (instead 

 of a simple posterior trefoil and one exterior triangle, as in Pedomys and 

 Pitymys and Chilotus); but the details are even more complicated than in 

 riparius. We have, fii-st, the anterior loop quite across the tooth in front, 

 then comes the first exterior triangle, then the first interior triangle; then 

 all the rest is the "plication of the immense posterior crescent, thus: the 

 crescent has its back, which is long and nearly straight, to the outside of the 

 tooth, but it throws off a spur anteriorly, forming the second exterior triangle, 



