space. The slightly larger size of these specimens from Wapinitia suggests intergradation with M. m. montanus. The tympanic bullae in the specimens from Wapinitia seem to be smaller than in specimens of canicaudus, nanus, or montanus.

Because of the intergradation described above between Microtus montanus nanus and M. canicaudus, the latter should stand as Microtus montanus canicaudus.

Bailey (N. Amer. Fauna, 55:206, August 29, 1936) recorded canicaudus from Warm Springs in the Deschutes Valley of Oregon and from the state of Washington. Other authors also have recorded canicaudus from the state of Washington. Our examination of specimens leads us to conclude, as did Dalquest (Univ. Kansas Publ., Mus. Nat. Hist., 2:348, 349, April 9, 1948), that canicaudus does not occur in Washington. The reported occurrence of M. canicaudus at Warm Springs, Deschutes Valley, Oregon, seems to be the result of an error in identification. The specimens concerned seem to be two Microtus longicaudus mordax (Nos. 207060 and 207082 U.S.N.M.). They are labeled as collected at "Warm Springs (Mill Cr.—20 Mi. W of)". Bailey's (op. cit., fig. 46, p. 209) map showing the distribution in Oregon of Microtus mordax mordax [=Microtus longicaudus mordax] has a locality-dot at Warm Springs itself. Bailey seems to have erred; he should have placed this dot 20 miles farther west, we think. When preparing his map (op. cit., fig. 43, p. 205) showing the geographic distribution of Microtus canicaudus, Bailey seems to have misidentified these same two specimens as M. canicaudus, and for them placed a locality-dot on his map 20 miles east (instead of west) of Warm Springs. In brief, Bailey probably did not see any specimens of canicaudus or specimens of any other subspecies of Microtus montanus from Warm Springs.

Transmitted February 15, 1951.

23–7438