M. f. goldmani in typical form occurs in high mountains of the Upper Tropical Life-zone and is most closely related to M. f. frenata and M. f. macrophonius. The altitude at which the two specimens were taken, twenty miles southeast of Teopisca in Chiapas, is not known. Merriam (1896:28) states that the type specimen was obtained at "about 8200 feet." The specimen taken by Stirton in Salvador comes from 8000 feet and the one obtained by Barber in Guatemala from 9500 feet. The specimen from Dueñas, the skin alone of a young animal, is not instructive.
As regards size, goldmani is larger than the immediately adjacent subspecies from the Lower Tropical Life-zone but is smaller than M. f. leucoparia or macrophonius. As compared with M. f. frenata, goldmani is longer, has an actually as well as relatively shorter tail, and a much longer hind foot.
The most outstanding difference in externals from frenata is the naked foot soles.
Molting probably takes place twice each year although actual proof of this is lacking. In number 133254 from twenty miles southeast of Teopisca, taken on May 12, the molt is well advanced. Another specimen from the same place still retains the winter coat.
In color, goldmani is much darker than frenata, has less extensive white facial markings, longer black tip on tail, more restricted color of underparts, and lacks the extension of color of the underparts onto the hind feet.
Of the adult males from the high mountains, the type specimen from Chiapas is lightest, and the one from Salvador is darkest. This progressively darker color to the southward probably is geographic variation.
In total length and relative and actual length of tail, the specimen from Salvador is the smallest of the five adult males from the higher mountains. In addition to its darker color and smaller size, no. 12523 from Salvador shows certain distinctive cranial characters. The zygomatic breadth is less than, rather than more than, or equal to, the distance between the condylar foramen and M1 or than that between the anterior palatine foramen and the anterior margin of the tympanic bulla. This difference appears to be correlated with geographic position, since no. 15953 from Guatemala has the three distances about equal and therefore is intermediate in this respect between the specimen from Salvador and those from Chiapas, in which the zygomatic breadth is greater than the other two measurements. Also in the greater depth of the skull and smaller size of the teeth, this specimen from Salvador approaches the subspecies of the Lower Tropical Life-zone. It has, however, the longest, highest and widest tympanic bullae of any of the five specimens. The amount of ventral convexity of the squamosal in front of each tympanic bulla appears not to be greater than in the other specimens.
As indicative of intergradation with perotae, leucoparia and possibly frenata, there is the specimen from Cerro San Felipe, Oaxaca. The degree of restriction of the color of the underparts is intermediate between that of goldmani and leucoparia. The same is true as regards the amount of projection from the braincase of the anterior margins of the tympanic bullae. The squamosal immediately anterior to each tympanic bulla is flat instead of ventrally convex as in goldmani or ventrally concave as in leucoparia and frenata. In accordance with the custom adopted in this paper of referring every specimen to some one subspecies, this specimen from Cerro San Felipe is referred to Mustela frenata perotae, to the description of which it most nearly answers.