MAMMALS OF PENNSYLVANIA AND NEW JERSEY. II3 



the sands margins of streams where they had been caught. From their strik- 

 ing appearance when found dead in a trap, I do not hesitate to place them 

 in the highest rank among all our east American mammals for exquisite color- 

 ation, grace of form and elegance of manners. No doubt their habits in some 

 respects may be predicted from what we know of the commoner meadow 

 jumping-mice, whose hunting grounds in the far north to some extent overlap 

 those of their more refined and aristocratic cousins. Nevertheless the swamp- 

 mud and tangled-grass dweller has little in common with one who seems to 

 choose the night season only for his rambles along the clean shingle and sand 

 of the little trout brooks, creeping and leaping about from boulder to boulder 

 and diving among the ferns and deep beds of moss at each whisper of alarm. 

 In the daytime I have never seen them move about nor caught any in my 

 traps, and conclude they are more nocturnal than Z. hudsonius. ^^Napceoza- 

 pus " seems never to be found out of close reach of a stream. If you fail to 

 catch him there, it is useless to look in other parts of the forests for him. 

 As he is attracted by raisins and oatmeal, there is little doubt of the rather 

 omnivorous character of his diet. Having been unable to trap this animal in 

 any given season as late as Z. hudsonius in the neighboring marshes, I con- 

 clude that it begins to hibernate sooner than that animal. The peculiar, and 

 in most cases conspicuous, white tip to the tail of the woodland jumping- 

 mouse cannot but excite our curiosity, especially when it is remembered that 

 Z. hudsonius does not show it. Owing to its variable amount, ^ of an inch 

 to I ^ inches, it seems to show a condition of change, not improbably toward 

 a future atrophy of the terminal vertebrae and shortening of a member which 

 some change in environment has shown to be uselessly or dangerously long. 



Description of species. — There is only one American species of this sub- 

 genus known, divided into three races, typical insignis of the Canadian zone, 

 roanensis oi the Great Smoky Mts., and abietorum of the Hudsonian zone. 

 None of these are known to reach farther west than Lake Superior. They 

 are characteristic of the east Canadian and AUeghanian territories. Any 

 NapcKozapus can be distinguished from all other American jumping mice by 

 larger size, longer ears, fulvous or buff yellow coloration, especially on sides 

 of head, fore neck, and on ears, with much less admixture of the black hairs 

 seen in other species. It is never suffused with clay or fulvous on white 

 underparts as in Zapus, and lacks the upper premolar, a small peg-like func- 

 tionless tooth characteristic of Zapus, the absence of which in Napceozapus is 

 the only reason for its subgeneric separation from our other jumping mice. 

 I am unable to reconcile the elevation of this subgenus to full generic rank 

 with the fundamental rules of classification. 



