From _richardsonii_, _alascensis_ differs in that the skull of the male averages smaller in every measurement taken and is 28 per cent lighter. Relative to the basilar length, the orbitonasal length is more and the braincase is shallower as measured at the anterior end of the basioccipital. The four adult females seen of _alascensis_ are more variable than those of _richardsonii_ and average smaller in some measurements and larger in others but give no proof of any consistent difference. From _haidarum_, _alascensis_ differs in that the rostrum and entire preorbital part of the skull is actually as well as relatively much smaller in both sexes. In males of _alascensis_ the length of the skull, and other cranial measurements of length, is more. In males, the mastoid breadth and zygomatic breadth are about the same as in _haidarum_, as also is the weight. M1 is larger but m1 and P4 are smaller. In females the anteroposterior extent of the inner moiety of M1 and length of tympanic bulla are about the same in the two subspecies but all other cranial and dental measurements in _alascensis_ are less. It is 29 per cent lighter. The difference in the preorbital region is of about the same degree as in the males. Comparisons of the skull with those of _arctica_, _salva_, _initis_, _celenda_, and _seclusa_ are made in the accounts of those subspecies. _Remarks._--The relatively few specimens known of this race seem always to have been referred to in the literature by the name _alascensis_ and the nomenclatural history is therefore simple. The original materials were obtained by the collector Clark P. Streator and the additional series of skeletons, one with skin, from Windham were procured by Stanton Price, a resident there. The subspecies is well differentiated from both _arctica_ and _richardsonii_. Although actual intergrades are lacking between _alascensis_ and the two races just mentioned I have no doubt that intergradation occurs with _richardsonii_ and think it probably does also with _arctica_. The assignment of the three females from Mitkof Island, Zarembo Island, and Loring on Revillagigedo Island, is tentative because each is so young as not to show diagnostic cranial characters. The two other specimens from Revillagigedo Island (Carroll Inlet), labeled as males, are in white winter pelage. Only one, no. 136358, a subadult, is accompanied by a skull. The small size of each specimen, and its cranial characters which are intermediate between those of males and females of _alascensis_ of the adjacent mainland, indicate the existence of a distinct race of weasel on Revillagigedo Island. On the chance that the one specimen with a skull is a dwarf, or is wrongly sexed as seems improbable, the population is tentatively referred to _alascensis_. [Illustration: FIG. 26. Map showing known occurrences and probable geographic ranges of the subspecies of _Mustela erminea_ in southeastern Alaska.]