386 BULLETIN 56, UNITED STATES NATIONAL MUSEUM. 



face is wood brown, thickly lined with black. The woolly tuft at 

 the anterior base of the ear is whitish, the dark spot on its outer surface 

 distinct. The pelage is very long and dense, the tail well haired and 

 sharply bicolored. An old female (No. U\U, U.S.N.M. coll. Inter- 

 national Boundary Commission), taken on the San Pedro River, Octo- 

 ber 16, 1892, is in short summer pelage, except on the muzzle, where 

 the winter coat is coming in. The under surface, ears, feet, and tail 

 are much more scantily coated with hair than in winter, and the 

 color of the upper surface is drab. The new hair on the muzzle is 

 more yellowish. Some specimens in summer pelage are more yellow. 

 An immature specimen (No. 58757, U.S.N.M., coll. International 

 Boundary Commission), taken at the town of Santa Cruz, Sonora, 

 October 21, 1893, is like the type. It is gray, tinged with drab, 

 whitish below. Mammae, P.], A.&, I.|=3 pairs. (For skull and 

 teeth see fig. 70.) 



Dr. J. A. Allen has already called attention to the dark coloration 

 of specimens from San Bernardino ranch, at the head of the Yaqui 

 River system; and our specimens from that locality are also darker 

 than usual, with the exception of No. 21327, which is curiously albi- 

 nistic. In the vicinity of Monument No. 77 of the Mexican Line, on 

 the San Bernardino River, these mice were semi-aquatic, living 

 amongst the aquatic vegetation beside the stream. Such individuals 

 were usually dark, but on the adjacent high ground nearly typical 

 examples of P. s. sonoriensis were trapped. While this dark form is 

 recognizably different from true sonoriensis, I have preferred for the 

 present, as in the case of the peculiar phase of P. eremicus from the 

 same region, to wait until the Yaqui Basin Tract can be further 

 explored, rather than name its peculiar forms from specimens from 

 its terminal twigs, without being able to assign definite geographic 

 ranges to the subspecies. 



Fifteen years ago, when treating of the western short-tailed species 

 of Peromyscus, then commonly styled the "sonoriensis group,"" I 

 referred the forms arcticus, nebrascensis, texanus, b sonoriensis, and 

 deserticola, as subspecies, to Hesperomys leucopus Rafinesque. I now 

 regard this course as illogical, because their intergradation with P. 

 leucopus was merely assumed to be a fact, and not proven. I am con- 

 vinced that our trinomial system of nomenclature will only impede 

 our progress, and come to be a hindrance, unless the rules upon which 

 the system is founded be so rigidly adhered to as to make it an accu- 



a Because Dr. Elliott Coues, in his monograph of the North American Muridae, published 

 in 1877, had grouped around Le Conte's "Hesperomys sonoriensis" not only those of the 

 forms here under consideration that were then known, but also several additional species 

 which are totally distinct from them, treating the whole assemblage of species and geo- 

 graphical races under the head of "Hesperomys leucopus sonoriensis." 



& This erroneously, as it has been shown me by Mr. W. H. Osgood to be a member of 

 the Peromyscus leucopus group. 



