MURIDEZ—ARVICOLINE—ARVICOLA RIPARIUS. 159 
enamel, the lobe is cut off from the rest of the trefoil, and transformed into 
an actual closed external triangle, of which, therefore, the tooth may show 
indifferently two or three. A step further in modification is this:—the trefoil, 
after being partitioned off as just explained, may develop an external lobe on 
its midleaf, and so remain in effect a trefoil, giving in all five external salien- 
cies, viz: one from the posterior loop of the tooth; two from the two closed 
triangles proper; one from the additional closed triangle; and one from the 
supplementary lobe of the trefoil* We have not noticed the same thing 
with the inner leaflet of the trefoil, but presume it may occur. 
The upper molars reverse the pattern of the lower; their transverse loop 
is in front instead of behind, and the lateral triangles alternate from before 
backward instead of from behind forward; and the last upper molar ends 
behind, just as the first lower molar begins in front, with a formation not 
seen on any other teeth. The first upper molar, like the last under molar, 
is very constant throughout the whole genus; the last upper and first under 
are our diagnostic teeth for the several subgenera, though the middle ones, 
both upper and under, furnish collateral characters. 
The front upper molar has in front a transverse loop, succeeded by two 
internal and two external alternating lateral closed triangles; the first of these 
being internal, the last postero-external. The saliencies and reéntrances on 
both sides are sharp; the saliencies, both external and internal, are three in 
number, counting a corner of the anterior loop in each enumeration. 
The middle upper molar likewise consists of an anterior loop succeeded 
by alternating lateral closed triangles, but here the first is an exterior instead 
of an interior one. The anterior ellipse usually sits a little oblique, its con- 
vexity looking inward as well as forward. Regularly there are only two 
perfectly-closed lateral triangles on this tooth; first an exterior, then an inte- 
rior. for the last one is as much posterior as interior, and not often regularly 
triangular. It may be called an exterior triangle, appendaged with a 
posterior lobe or spur representing an imperfect additional internal triangle. 
* This is our view of the formation of a variable number of lateral closed triangles in this tooth ; 
it is simply a higher degree of complexity in the folds of the anterior trefoil. In Pedomys and Pitymys, 
the reverse, of less complexity, is seen. Here the median zigzag of enamel does not run so far forward ; 
the anterior triangles are not fairly closed up, and their openness throws their dentine islets into one 
diamond-shaped area tkat is usually also continuous with the interior of the trefoil itself. This is 
nearly as good a character for distinguishing Pedomys or Pitymys from any style of 4. riparius as that 
afforded by the back upper molar itself; though it is sometimes obscure, we have never seen the anterior 
lateral triangles fairly open in riparius, nor completely closed in the other two subgenera. Chilotus, 
however, is like riparius in this respect. 
