Perognathus artus. These identifications were reasonable in the light of knowledge available in 1941, but in the light of information presently available from the now more abundant material it is clear that all 10 of the specimens are P. g. artus. Examination (by Hall) of the specimens reveals that the differences relied upon by Burt and Hooper to differentiate the two species are well within the range of individual variation. For example, the variation (5.3 to 5.6 mm.) in width of the supraoccipital is less than in each of some other series of specimens of equal age of P. g. artus from other localities.

Also, there is geographic variation in the mice here assigned to the subspecies P. g. artus; skulls are smaller in the northern part of the geographic range and become gradually larger toward the south. In five adults from the northern part (Batopilas 3, and 26 mi. NE Choix 2) the mean of 12.6 of the mastoidal breadth of the skull is significantly smaller than the corresponding mean of 13.3 in 21 adults from the southern part (32 mi. SSE Culiacán 14, and El Dorado 7). The pelage of individuals from one and a half miles southwest of Tocuina is notably dark both above and below; the venter is dusky rather than white. We suppose that the darker color is a response to a dark-colored substrate—lava and soils derived from dark lava.

We have not seen any specimens that are intergrades between P. goldmani and Perognathus intermedius (subspecies intermedius or lithophilus), nor between P. goldmani and Perognathus nelsoni (subspecies nelsoni or canescens), nor between P. intermedius and P. nelsoni. Collecting and studying specimens from geographically appropriate places to test for intergradation between these three species would be worthwhile as a means of improving our knowledge of the taxonomy of these mice.

Transmitted October 1, 1959.

28-1243