148 MONOGRAPHS OF NORTH AMERICAN RODENTIA. 



which may be quite closed, or its area confluent with the general island of 

 1 he crescent ; In 'hi ml, the crescent loops broadly, forming the end of the tooth, 

 then throws out a spur on the postero-internal corner of the tooth, then in 

 front of this is thrown into a large loop, which makes the second interior 

 triangle, not closed, however, but continuous with the general area of the 

 crescent. Thus there are in all, on the hack upper molar, three exterior 

 salieueies and tour interior saliencies. The precise details of this tooth vary 

 a little with individuals, but the pattern, as just described, we have never 

 found effaced or even obscure; it is, therefore, highly diagnostic. 



The molar crowns of Arvicolce in general appear to have central enamel, 

 or interior folds and ridges separated from the general enamel wall that 

 enfolds the teeth, but this appearance is deceptive; there is but the single 

 enveloping sheet of enamel around the whole tooth, which is so deeply 

 indented or folded in at the reentrant angles of the several prisms that the 

 enamel sheet of opposite sides meets and fuses along a central line, often no 

 wider than a single sheet of enamel, producing the appearance just mentioned. 

 Now, in Erotomys, the enamel of opposite sides, in the upper jaw, meets at 

 various places, but the fusion is not complete; either the two sheets are 

 apparent where they touch each other, or else the imperfect fusion results in 

 a wall the composition of which is evident by its being broader or thicker 

 than a single sheet of enamel is anywhere. And in the under series, which 

 we now come to examine, the enamel walls are still more distinct, revealing 

 their true relations; they never quite fuse, and, even where they press upon 

 each other most closely, we can discern two distinct folds, and thus trace the 

 single enveloping sheet of enamel, in and out, in its various plications, all 

 around the tooth. 



The posterior lower molar affords nothing diagnostic, being, as in 

 Arvicola, composed simply of an anterior, a middle, and a posterior spheri- 

 cal triangle, each one of these reaching quite across the tooth, and thus lying 

 directly one after the other; but a singular tiling is, that the middle lower 

 molar copies the same pattern. In our Arvicolce proper, this middle lower 

 molar has an anterior triangle, succeeded by alternating lateral triangles; hut 

 in this genus the lateral triangles are opposite instead of alternate, which liict, 

 together with the lack of a median lengthwise line of enamel, throws the 

 two lateral ones into one that reaches quite across the tooth. It is surprising, 

 in this case, with essentially the same pattern, such a little variation as this 



