MDIUDAE— SIOMODONTES— HESPBUOMYS LEDCOPUS. 



53 



occupies from one-third to nearly or quite one-half of the circumference — 

 g(;nerally about two-fifths. Hut, in many cases, the line of separation is 

 «)l)scure, and tlien tlic tail is simply paler below than above. Even some other- 

 wise typical Massachusetts specimens show this last condition. 



The variation in absolute and relative length of the tail is greater than in 

 any other dimension. It ranges from obviously longer than the head and 

 body to about equal to the body alone. This, it should be observed, is inde- 

 pendent of locality, and exclusive of what we have admitted further on as a 

 variety nonorietms. Mr. J. A. Allen, who has paid jiarticular attention to the 

 variability of feral animals, finds that in Massachusetts specimens alone the 

 proportion of tail to trunk may be as 1.18 : 1.00, or as 0.69 : 1.00 — a difference 

 of over fifty per cent, of the mean ; and that the number of caudal vertebral 

 tiienhsclves ranges from twenty-four to thirty or more. But this ceases to be 

 remarliable when we recollect that it is purely a matter of what has been aptly 

 calle'i "vegetative repetition". It seems to be a well-nigh universal law that 

 those parts or organs that are the least specialized, — i. c, those of which several 

 have the same or corresponding character and function, — are liable to be pro- 

 duced with a iiigh degree of irregularity as regards their number; and the 

 more such there are the wider are the limits of variation apt to be. In this 

 species, one of our longest-tailed rodents, the law is perfectly illustrated. 



Since none of the nominal sjjecies that have been erected upon the vari- 

 ations of leucopus depend in any way upon cranial characters; and since the 

 valid species of Vesj)erimus likewise offer no noticeable cranial features beyond 

 slight differences in dimensions (greater in californicus and less in michigan- 

 ensis for example), the skulls of all our North American Hesperomys, except 

 those of the subgenera Onychomys and Oryzomys, may be most conveniently 

 examined collectively. 



