the Mohave Desert, Thomomys perpallidus [= bottae] mohavensis, but failed to mention the specimens recorded by Bailey, and thus their subspecific identity is in doubt. We find that T. b. mohavensis differs from T. b. perpes in more pallid color (light yellowish as opposed to dark rufescent) larger size, larger and more angular skull, angular (as opposed to more evenly bowed) zygomatic arches, larger and deeper audital bullae, narrower interpterygoid space, and proportionately greater mastoidal breadth. In external measurements, size and angularity of skull, width of interpterygoid space and angularity of the zygomatic arch, the specimens from Lone Willow Spring seem to be intermediate between the two subspecies, but perhaps show more resemblance to T. b. mohavensis. Otherwise, the specimens closely resemble T. b. mohavensis to which they are here referred. The specimens provide a northern marginal record of occurrence for that subspecies.

Other specimens recorded as T. b. perpes by Bailey (loc. cit.) from Grapevine Ranch, California, have also not been mentioned in later publications although, from a geographic standpoint, they might be better referred to either Thomomys bottae pascalis or T. b. mohavensis. Comparison of specimens of T. b. mohavensis and T. b. pascalis from various localities show T. b. pascalis to be larger (including the skull), darker, and to possess a more nearly vertical occipital plane, wider-spread but less angular zygomatic arches, less inflated tympanic bullae, wider braincase (which consequently appears to be less inflated), proportionately longer and slenderer rostrum, and broader nasals distally. Cranially, T. b. pascalis differs from T. b. perpes in essentially the same ways, but to an event greater degree. In color, T. b. pascalis differs from T. b. perpes in being duller, less rufescent.

The series of four specimens, in the U. S. Biological Surveys Collection, from Grapevine Ranch clearly are not referable to T. b. perpes. They do, however, agree with T. b. mohavensis in all essential particulars except that in two of the four specimens the braincase is wider and the nasals are wider distally. This width is evidence of intergradation with T. b. pascalis. Seemingly, then, they are best referred to Thomomys bottae mohavensis.

Thomomys bottae muralis Goldman

When Goldman (Jour. Washington Acad. Sci., 26(3):112, March 15, 1936) described and named this pocket gopher from Arizona, he arranged it as a full species and stated that there is no evidence of intergradation with other named kinds. We have examined the holotype and three topotypes (202579-202582 BS) and compared them with specimens of other kinds of pocket gophers occurring in northern and central Arizona. The muralis gopher is a depauperate form clearly belonging to the bottae group. The characters which Goldman (loc. cit.) set forth as distinguishing muralis from other named kinds are readily apparent and, like Goldman, we see no evidence of intergradation. Nevertheless, the characters which serve to identify the race are, in a general way, those commonly found in populations of depauperate individuals of Thomomys bottae and T. talpoides. The small size, delicate structure, well-inflated braincase, short premaxillary tongues, and strongly recurved upper incisors, often appear in populations existing in inhospitable areas of shallow, unstable soils. For this reason we feel that the relationships of this population are best shown by arranging muralis as a subspecies of Thomomys bottae; the name should stand as Thomomys bottae muralis Goldman.

As far as known, T. b. muralis is completely isolated from other populations of pocket gophers by uninhabitable eroding cliffs. The animals have been found only on isolated terraces in the lower end of Prospect Valley (itself a lateral pocket) within the Grand Canyon of the Colorado River, Hualpai Indian Reservation, Arizona. Consequently it is unlikely that intergradation with other populations could exist at the present time.

In short, in arranging muralis as a subspecies of Thomomys bottae, we are influenced, not by the demonstration of intergradation, but by the degree of morphological differentiation of the population and the probable reasons therefor.

Thomomys bottae mutabilis Goldman

Goldman (Jour. Washington Acad. Sci., 28:342, July 15, 1938) named the subspecies Thomomys bottae pinalensis on the basis of only one specimen, an immature female (245709 BS) from Oak Flat, five miles east of Superior, Pinal Mountains, Arizona. Examination shows it to be indistinguishable in characters of taxonomic importance (coloration, external measurements, shape of skull and size of skull) from specimens of T. b. mutabilis of comparable sex and age. No. 245709 is well within the limits of individual variation of T. b. mutabilis as is shown by the several specimens (all in the U. S. Biological Surveys Collection) as follow: Nos. 214118, 214670 (topotypes from Camp Verde, Arizona), 212707 (Chiricahua Ranch, 20 mi. E Calva), 208635 (H-bar Ranch, 20 mi. S Payson), and 215762 (Turkey Creek). Therefore, the name Thomomys bottae pinalensis is here arranged as a synonym of the earlier name, Thomomys bottae mutabilis Goldman (Proc. Biol. Soc. Washington, 46:75, April 27, 1933), the type locality of which is Camp Verde, Yavapai County, Arizona.

Thomomys bottae patulus Goldman

When Goldman (Jour. Washington Acad. Sci., 26:113, March 15, 1936) named the subspecies Thomomys bottae desitus, he assigned to it (op. cit.:114) 10 specimens obtained at Wickenburg, Maricopa County, Arizona. He did not mention specimens from Wickenburg when he subsequently named the subspecies Thomomys bottae patulus (Jour. Washington Acad. Sci., 28:341, July 15, 1938) and stated that T. b. patulus was known only from the type locality in the "bottomland along [the] Hassayampa River, two miles below Wickenburg." Examination in 1950 of specimens referable to T. b. patulus in the U. S. Biological Surveys Collection shows all of them, including the holotype, to be labeled "Wickenburg." The 10 specimens from Wickenburg reported by Goldman in 1936 as T. b. desitus were included by him among the 16 (actually 17, one being a skull only) upon which he based his description of T. b. patulus in 1938. Examination of the field catalogues of 3 of the 4 collectors who obtained the specimens discloses that only the 7 specimens obtained last were recorded as occurring in the Hassayampa River bottoms; the first 10 were recorded only as from "Wickenburg." Briefly, only one subspecies, T. b. patulus, is present in the area, and Goldman in 1938 seems to have thought that the two localities were actually the same, and that "2 miles below Wickenburg" was the more precise designation.

Thomomys bottae providentialis Grinnell

We have examined a specimen, No. 26120/33526, from 12-Mile Spring, California, in the U. S. Biological Surveys Collection, which Bailey (N. Amer. Fauna, 39:73, November 15, 1945) referred to the subspecies Thomomys perpallidus [= aureus] perpes. We find the specimen to be referable to the later named Thomomys bottae providentialis on the basis of smaller ear, more massive, more ridged and angular skull, greater interorbital breadth, deeper and thicker rostrum, less globular bullae, and U-shaped rather than V-shaped interpterygoid space. Therefore, 12-Mile Spring is the northernmost locality of occurrence of the subspecies T. b. providentialis.