g6 MAMMALS OF PENNSYLVANIA AND NEW JERSEY. 



of the folding of the enamel in the second and third anterior triangles of the 

 first upper molar. In nearly all the southern N. J. specimens of rhoadsi these 

 triangles are not complete but are confluent ; the molar triangles are also 

 more flattened and wider, in relation to the length of the teeth, than in gapperi 

 from northern latitudes. But these characters have exceptions, a skull from 

 the Walkill valley being exactly of this rhoadsi pattern and a topotype of 

 rhoadsi from May's Landing having the angles closed as in gapperi. As re- 

 gards size of skull, in the southern animal it does not grow as long as, but is 

 relatively a little wider than, gapperi. Specimens from Sussex Co., N. J., are 

 intermediates in this respect, however. From the anatomical standpoint, 

 therefore, we lack the necessary evidence to consider gapperi and rhoadsi 

 distinct species. Geographically they have been supposed to be widely sepa- 

 rated, but, as I have previously shown, the gap which originally seemed to 

 exist has been narrowed by absolute research so much that our botanical 

 knowledge of the rest of the country seems likely to span the remainder. 



In this connection, I would remark that the suggestion made by Mr. Miller, 

 regarding the relationships of N. J. wood voles to the species carolinensis 

 inhabiting the balsam belts of the higher southern Alleghenies of North Caro- 

 lina, is most pertinent. My series of 22 specimens from Summit Mills, south- 

 em Somerset Co., Pa., shows remarkably close approach to all the distin- 

 guishing characters of size given by Bailey (Proc. Biol. Society, Washn., 1897, 

 p. 130) to separate carolinensis specifically from gapperi. In regard to color, 

 the Somerset Co. specimens are more like gapperi. The largest adults of 

 these measure as follows, in millimeters (Bailey's similar measurements of 

 carolinensis following each in brackets): Total length, 148 (149); tail verte- 

 brae, 41 (44); hind foot, 20 (20.2); ear, 11; basal length of skull, 22.8 

 (23.5); nasals, 7.8 (7.5); zygomatic breadth, 14.5 (14.4); alveolar length 

 of upper molar series, 5.5 (6). Specimens from West Virginia and from 

 Cambria Co., Pa., complete the links in an unbroken chain which connect 

 gapperi of northern Pa. and New York with his larger and darker kinsman, 

 e. g., carolinensis, of the Great Smoky Mountains, in much the same manner 

 as has already been pointed out in my remarks on Miller's deer mouse and 

 the Cloudland deer mouse of the same regions. Taking Canadian gapperi as 

 the parent stock, we thus have two darker and larger off-shoots projected 

 southward into the austral zone, one by virtue of the moist, boreal, lowland 

 climate of cedar swamps, the other on account of a similar environment re- 

 sulting from an elevation of 6,000 feet above the sea, amid the moisture-laden 

 winds of both sea and land. 



Measurements of species. — These body measurements apply alike to gapperi 

 and g. rhoadsi. Total length, 140 mm. (5^ in.); tail vertebrae, 40 (ij^); 

 hind foot, 19 (f); ear, from crown, 11 (y^). In rhoadsi the skull is not 

 materially larger than gapperi but the tooth rows are about ^ of a millimeter 



