Remarks.—One of the most noteworthy of the several unique characters of this large, tropical weasel is the longitudinal, median, abdominal band. The species exhibits the minimum degree of development of certain features that become progressively less apparent as one proceeds southward from Central America. The relative uniformity of the coloration of the upper parts (reduction in intensity of black color on the muzzle and tip of the tail) and the reduction of the tympanic bullae are two cases in point. Viewed dorsally the general outline of the skull is most nearly matched by that of the skull of Mustela frenata meridana from Venezuela or that of M. f. helleri from Perú. However, the resemblance is not close. The tympanic bullae, although unique among American weasels, are more like those of M. f. meridana from Venezuela than like those of any other kind. The great postorbital width (relatively less in M. africana than in several South American subspecies of Mustela frenata) and small angular process of the mandible are characters, in varying degrees, also common to all South American weasels. Structurally M. africana clearly is more nearly like other subspecies of M. frenata from South America than it is like any species or subspecies from North America.
Mustela africana is the most primitive of the American weasels. The distinctive cranial and dental characters, excepting the reduction in number of premolars, are of a primitive nature. For example, the relatively wide postorbital region, the large braincase that is inflated anteriorly, and the flattened, tympanic bullae, are points of resemblance to the holarctic Mustela erminea, which species is regarded as nearest the original stem form; also the mentioned characters correspond to ontogenetic stages passed through by other weasels. Mostly on these accounts, one is led to look upon M. africana as a migrant from North America. It may have become isolated from its original stock, by a water barrier in the Central American region, for a length of time sufficient to permit of a degree of differentiation to develop between it and the North American weasels which prevented crossbreeding with the frenata stock when that stock, at a later time, reached South America. This assumption is suggested only by evidence from the Recent specimens. No remains of true weasels (subgenus Mustela) have been recorded from deposits in South America older than the Recent period. The alternate possibility, that M. africana intergrades with some race of M. frenata in western or northern South America, has been considered and regarded as highly improbable.
Cabrera (1940:15) has made the distinctive structural characters of Mustela africana basis of the generic name Grammogale to include the one species africana. I am inclined to accord Grammogale only subgeneric rank.
It is possibly significant that Mustela africana is intermediate in several respects between Lyncodon and typical Mustela. The median, longitudinal, abdominal band of the same color as the upper parts in M. africana and the relative uniformity of the coloration of its upper parts might be considered as an intermediate stage between the dark, bicolored (black muzzle and tail tip and brown body) upper parts and light-colored underparts of the North American weasels on the one hand and the light, unicolored upper parts and dark-colored underparts of the Patagonian weasel (Lyncodon) on the other hand. The number of premolars,
2-3
---,
2
is also intermediate between the numbers
3 2
- and -
3 2
of the North American Mustela and the Patagonian Lyncodon, respectively. The American Mustela and the Patagonian Lyncodon, respectively. The more medially, as opposed to anteriorly, directed medial cusp of P4 (characteristic of approximately half of the specimens examined), and the structure of the skull in general of M. africana also seem to be morphologically intermediate between those parts of Mustela and Lyncodon.
On the chance that Lyncodon is closely enough related to Mustela, to be included in a group with Mustela rather than in a group with Grisonella, it is worth noting that Lyncodon lujanensis Ameghino (1889:324, 325), from the villa of Lujan and at the city of Córdoba, at each place in the Pampean [= Pleistocene] formation of Argentine (see also Cabrera, 1928:263) is the first and only fossil form of this group recorded from the whole of South America. Actually, however, Lyncodon seems to me to be as nearly related to Grisonella, if not more so, than to Mustela. If Lyncodon is more closely related to Grisonella and Grison than to Mustela, then the above remarked intermediacy in characters of M. africana has more of interest as a tendency to parallelism than it has of phylogenetic import. Appraisal of phylogenetic relationships would require appraisal of the ancestral stem forms of the Grison stock and the Mustela stock. None of either is known from deposits of the Pliocene, the period of time immediately preceding the Pleistocene.