MURID2E— ARVICOLIN.E— ARVICOLA RIPARIUS. 1(51 



but which, especially in connection with the characters of the other teeth, 

 proved decisive. The second noticeable modification of the crescent affects 

 the posterior part of its convexity, which may develop a little lobe or spur 

 simulating a third external triangle; then, as a necessary result, the curve is 

 sinuate instead of regularly convex, having a little concavity in one part. 

 The third principal modification is in respect of the posterior horn of the 

 crescent. This is generally a simple lobe finishing the tooth behind; but it 

 may, as one of endless slight changes, produce a sort of curved spur or hook. 

 When this fold is pretty long, narrow, and curved, it looks like a little addi- 

 tional crescent lying in the belly of the main one. It is hardly necessary to 

 add that none of these modifications have the slightest significance beyond 

 indices of individual variability; and we should not have entered into the 

 tedious minutiae except to bring this fact prominently forward, and to char- 

 acterize the pattern of this important tooth in all its phases. 



We shall open the discussion of A. riparius with the examination of 

 nineteen typical specimens from the locality of the original riparius Ord. 

 These were all collected at the same time and place, at a season of the year 

 when they are necessarily full-grown, and were all prepared by the same 

 person, in the same style of skin. Whatever differences, therefore, may be 

 found are to he attributed to purely individual variation, of which we thus 

 obtain a perfectly reliable equation, free from any disturbing functions, such 

 as age, season, climate, locality, &c. Whatever variation in size and propor- 

 tion may be found becomes a still purer index, if possible, from the tact that 

 not one of these specimens is distinguishable from any one of the other 

 eighteen in color; for, as far as color is concerned, they all might have 

 belonged to the same litter. In respect of the measurements, we should 

 premise that, as great care was taken in preparing the skins, probably no one 

 of them differs in total length 25 from the original dimension in life, and 

 that, if anything, the lengths as given are a trifle under life-size, from shrink- 

 age in drying; and, secondly, that the loss in drying of the feet and tail may 

 be estimated at about 5 per cent. 

 11 M 



