Remarks.—The ten specimens studied by the writer fall into two groups of six larger individuals and four smaller. Upon comparing these with each sex of the three species of American Recent weasels, frenata, erminea and rixosa, it is seen that size, and to some degree shape, rule out of consideration both sexes of rixosa and also males of frenata. Thus we are left with females of frenata and males and females of erminea. So far as size is concerned, it can be assumed that the larger specimens are females of frenata and that the smaller are males of erminea. This assumption has in its favor also, the fact that the postglenoidal length of the skull accords with that in Recent specimens. The difference in this regard in Recent animals is that the postglenoidal length of the skull, expressed as a percentage of the total (condylobasal) length of the skull, amounts to:
| in frenata | in erminea |
| ♂ ordinarily less than 46 | ♂ ordinarily more than 46 |
| ♀ less than 47 | ♀ more than 48 |
In the fossils the percentage for the larger skulls is 46; for the smaller skulls it is 48.
It may be that the ten fossil skulls are six female frenata and four male erminea but I think not. In the first place a skull of different shape, seemingly of the frenata stock, is known from the deposit and it is almost certain that two subspecies of the same species would not occur at the same place at the same time. It is possible, of course, that parts of the deposits were laid down at times so far apart that a shift in geographic range of two subspecies had occurred. This one skull, seemingly of the frenata stock, is the type of Putorius gracilis Brown (see p. 404) and was regarded as the only known specimen of gracilis. Regardless of the specific identity of this one specimen named gracilis, the chances of obtaining otherwise from a deposit, like that in Conard Fissure, six females of frenata and four males of erminea without a male frenata or a female of erminea coming to light are so slight as strongly to incline me to the view that the six larger specimens are males of the same species to which the 4 smaller specimens belong. By either this interpretation, or the one initially considered (of female frenata and male erminea), the animals from the fissure are at least subspecifically distinct from any American Recent weasel. Furthermore, by this latter interpretation each sex of this weasel, angustidens, is intermediate between the frenata and erminea stocks in the feature of postglenoidal length which feature, at any place where the two Recent species occur together, serves to distinguish one from the other. In the northernmost subspecies of erminea (arctica for example) the postglenoidal length in some males is no longer than in males of frenata. Considering general size, angustidens agrees better with erminea than with frenata and this circumstance has influenced me to place angustidens as a subspecies of erminea.
Today, erminea is not known to occur nearer Conard Fissure than northern Iowa, more than 400 miles to the northward. In comparison with the race there, bangsi, males of angustidens are of approximately the same size but in the shorter distance between the glenoid fossa and anterior margin of the tympanic bulla, and also in the lesser postglenoidal length of the skull, angustidens resembles the northernmost American subspecies of erminea. Females of angustidens differ more from any living weasel than the males do. The females are much larger than those of bangsi, and among living American races of erminea most closely resemble intergrades between arctica and richardsonii which intergrades are found approximately 1700 miles to the north of Conard Fissure. In females, the preorbital part of the skull in M. e. arctica is broader and in M. e. richardsonii narrower than in angustidens. If it seems strange that females of angustidens resemble one subspecies whereas males, in size, resemble another subspecies almost a thousand miles distant, it should be remembered that the degree of sexual dimorphism varies much from one subspecies to another in the Recent animals. An example is furnished by Mustela erminea fallenda and Mustela erminea invicta.
The assemblage of mammals from Conard Fissure includes several species of boreal predilections which, like Mustela erminea, now occur only much farther north than Arkansas. At one time the edge of the sheet of ice was only about 200 miles north of Arkansas. It may be significant that the cranial characters of the female ermine from the Fissure, and qualitative cranial characters of males from there, are most nearly approximated among Recent weasels by those which live along the southern edge of the frozen tundra.
In view of what has been said, the possibility should be considered that the distinctive cranial features of angustidens may be the result of evolutionary change in time as well as of geographic variation resulting from horizontal placement.