MUKID^E— AltVICOLINvE— ARCTIC ARVICOLiE. 189 



When we began to look up A. townsendi, we anticipated no difficulty in 

 making it out specifically distinct from ordinary riparius; but even with very 

 little material to work up, we are fairly drawn to the opposite conclusion. 

 In fact, the only tangible difference we can make out is that townsendii is 

 larger, with a longer tail on an average; and even this is not constant, for 

 several, out of our few specimens, exceed average riparius but little, or not 

 at all, and come well within the limits of riparius. Still these specimens 

 might have grown a little larger, and the average of the series stand at the 

 extreme limit of riparius, while their tails and ears exceed this limit. We 

 therefore feel justified in retaining the name townsendi as expressive of a 

 geographical differentiation, occurring in a particular locality, to the apparent 

 exclusion of the usual United States exponent of the subgenus. It is not at 

 all a reliable species, and one of its strongest features — length of tail — 

 reminds us forcibly of the case of Hesperomys "boylii" from the same local- 

 ities; while the Columbia River series affords direct jxissage into riparius. 



We have taken some pains to inform ourselves respecting the variation. 

 in size of the common large European species of the genus Arvicola amphi- 

 bius, the well-known water-rat; and we find that all the variability we claim 

 for riparius, and even for the Arctic forms of that species (as we shall pres- 

 ently see), sinks into insignificance beside the variations known to, and 

 admitted by, all the better-informed writers in the case of A. amphibius. 



The propriety of sinking A. townsendii to a mere variety of riparius 

 will appear in still stronger light when we have dealt with Arctic animals of 

 this subgenus full seven inches long. We take up this question next. 



Discussion of the Arctic Arvicola of the riparius type. 



The United States variations of A. riparius are disposed of easily in 

 comparison with the intricacies of the Arctic material, respecting which some 

 general observations will be offered. 



Although we have handled a far larger number of these animals than 

 has ever before been examined by all other investigators put together, yet, 

 perhaps for the very reason that we have seen so many specimens, we are 

 still unprepared to make identifications without reserve. Nor can we deter- 

 mine with certainty all of Richardson's supposed species, after careful study 

 of his accounts, as well as of Audubon's and Bachman's supplementary 

 notices, and although we have been favored by Dr. Sclater with additional 



