SYSTEMATIC ACCOUNTS OF SPECIES AND SUBSPECIES (continued 9)

Color.—As described in Mustela frenata goldmani except that: Back near (n) Argus Brown, or Carbon Brown, tone 4 of Oberthür and Dauthenay, pl. 342. Underparts Ochraceous-Buff. Least width of color of underparts, in four males, young, subadult and adult, 24 (extremes 22-26) per cent of greatest width of color of upper parts; the corresponding per cent in one female is 32; black tip of tail, in two subadult males, averaging 29 (extremes, 28-30) per cent of length of tail-vertebrae; corresponding per cent in one female, 36.

Skull and teeth.—Male (based on type specimen, one adult topotype [?] and one subadult from San Rafel del Norte): See measurements and plates 25-27, 30. As described in Mustela frenata perda except that: Weight, 4.2 grams (estimated for adults); basilar length 45.0 (44.8-45.5); interorbital breadth more or less than distance between foramen opticum and anterior margin of tympanic bulla; anterior margin of tympanic bulla as far posterior to foramen ovale as width of four to five upper incisors; length of tympanic bulla not less than length of lower molar and premolar tooth-row; anterior margin of masseteric fossa below anterior margin of m2 or posterior to that tooth.

Female: Skull unknown.

Comparison of the skull of the male with that of M. f. costaricensis is made in the account of that subspecies. As compared with that of M. f. perda, which it most closely resembles, the skull of the male has a narrower, shorter rostrum, lesser interorbital breadth, lesser mastoid and zygomatic breadths and slightly shallower braincase, measured at anterior margin of basioccipital. The tympanic bullae are slightly less projected, at their anterior margins, from the braincase and the squamosal, directly anterior to each, is a little more convex ventrally. The skull of M. f. nicaraguae is, then, slightly shorter than that of M. f. perda and relatively narrower.

Remarks.—When naming this form, Allen (1916:100) characterized it as "Similar to M. tropicalis tropicalis but general coloration much darker and the white face markings somewhat reduced in area." In the sentence preceding the one quoted, Putorius tropicalis perdus was placed as a synonym of Putorius tropicalis tropicalis. M. f. nicaraguae and M. f. perda are nearly alike in color and color pattern but differ in cranial characters. M. f. perda and M. f. tropicalis are widely different in color and more especially in color pattern but differ only slightly in cranial characters. The aggregate difference between perda and nicaraguae is less than that between perda and tropicalis. All three are lowland forms and each is smaller than the adjacent highland forms, namely, M. f. goldmani, macrophonius, perotae and frenata.

The weasels from Honduras definitely are not typical of nicaraguae as it is known from the specimens from Nicaragua itself. The specimens from the state of Tegucigalpa, Honduras, are larger. Some are darker than topotypical nicaraguae. The dorsal outline of the skull is more nearly flat (less convex) in some. In these and several other differential features studied, the average of specimens from Tegucigalpa is intermediate toward goldmani, but everything considered the animals seem best placed with nicaraguae rather than with goldmani or perda, to which latter also, they show some resemblance. With better material from Nicaragua and additional specimens from Salvador (here referred to goldmani) a restudy of all the material now referred to the three races named would be profitable. Aims of this restudy might be to determine if a highland race additional to goldmani should be recognized and if the lowland races perda and nicaraguae differ from one another in the way that the existing specimens indicate.

In the five males from Matagalpa, the narrow white band in front of each ear is confluent with the color of the underparts on one side only in one specimen and on both sides in two specimens. None of these bands is confluent with the white patch between the eyes. A dark spot at the angle of the mouth is present on one side in one specimen. The corresponding area is dark colored in all other specimens but not separated from the color of the upper parts. In the specimen from San Rafel del Norte the white bands are not confluent with the color of the underparts. The female from Mambacho has the mentioned bands confluent with the color of the underparts. This female approaches M. f. costaricensis in the dark color of the upper parts but has more extensive white facial markings than some specimens from much farther north. Like a female seen of M. f. costaricensis, this one has a "frosted" nape but the white hairs on the back of the neck are less numerous than in the female of M. f. costaricensis.

M. f. nicaraguae in typical form, then, is thought of as a small, lowland, tropical subspecies only slightly differentiated from M. f. perda. By reason of its intermediate characters, it constitutes a link between the lowland forms, and the larger animals called M. f. goldmani and M. f. costaricensis.

None of the four skulls from Nicaragua shows signs of infestation of the frontal sinuses by parasites.

Specimens examined.—Total number, 16, listed by localities from north to south. Specimens are in the American Museum of Natural History, unless otherwise indicated.

Honduras: Alto Cantoral, 2; Cerro Grande La Paz, 1. La Flor Archaga, 1[75]; Comayagüela, 1[75]; vicinity of Tegucigalpa, 2; no locality more definite than Honduras, 1[4].

Nicaragua: San Rafel del Norte, 1; Matagalpa, 6; Ma[o]mbacho, 1.

Mustela frenata costaricensis Goldman

Long-tailed Weasel

Plates 25, 26, 27, 28, 29 and 30

Mustela costaricensis Goldman, Proc. Biol. Soc. Washington, 25:9, January 23, 1912.

Mustela brasiliensis, Gray, Ann. and Mag. Nat. Hist., 14(ser. 4):374, 1874.

Putorius (Gale) brasiliensis frenatus, Coues, Fur-bearing animals, p. 142, 1877 (part).

Putorius affinis, Merriam, N. Amer. Fauna, 11:31, June 30, 1896 (part).

Mustela affinis costaricensis, Allen, Bull. Amer. Mus. Nat. Hist., 35:101, April 28, 1916; Lönnberg, Arkiv för Zool., 14(no. 4):16, 1921.

Mustela frenata costaricensis, Hall, Carnegie Instit. Washington Publ. 473:109, November 20, 1936.

Type.—Male, young, skull and skin; no. 13770/37149, U. S. Nat. Mus.; San José, Costa Rica; obtained by C. H. Van Patten.

The skull (plates 28-30) is complete and unbroken. All teeth are present and unworn. The skin apparently has been remade. It lacks the distal two-thirds of the tail. The head is somewhat shrunken. The color is possibly faded but if so only to a slight degree. Otherwise, the skin is in good condition. The orange color of the underparts is so intense as to suggest that the full, adult pelage has not been acquired. No white markings are present on the face. There is no sex mark on the label attached to the skin but the size and proportions of the skull and the scrotal pouch on the skin prove that the specimen is a male. The presence of sutures on the dorsal face of the rostrum and the short, wide, and low sagittal crest show the specimen to be young.

Range.—Costa Rica. Altitudinal and zonal range unknown. See figure 29 on page 221.

Characters for ready recognition.—Differs from M. f. panamensis in lighter color of upper parts (tone 2 rather than tone 4 of Reddish Black of Oberthür and Dauthenay, pl. 344) and longitudinally flat interorbital region of skull; from M. f. nicaraguae in darker color of upper parts (of Oberthür and Dauthenay, tone 2 of pl. 344 rather than tone 4 of pl. 342) and greater width (more than 7) of tympanic bulla.

Description.—Size.—Male: No collector's measurements available of fully grown animals. Estimated measurements of adult males: Total length, 470; length of tail, 165; length of hind foot (taken from dried skins of 3 adults), 52 (50-52). Tail estimated to average 55 per cent as long as head and body. Length of hind foot more or less than (about equal to) basal length.

Female: A subadult or adult, from the Candelaria Mountains, and a subadult from Irazú, measure, respectively: Total length, 370, 385; length of tail, 130, 150; length of hind foot, 40, 31. Tail 59 per cent as long as head and body. Length of hind foot probably about equal to basal length.

The estimated differences in external measurements of the two sexes are: Total length, 92; length of tail, 25; length of hind foot, 16 (probably average difference is less).

Externals.—As described in M. f. panamensis (figure 21) except that foot soles are slightly more hairy.

Color.—As described in Mustela frenata panamensis except that: back is near Reddish Black, tone 2 of Oberthür and Dauthenay, pl. 344; chin, lips, and throat white or whitish; remainder of underparts near (c) Ochraceous-Buff; color of underparts rarely extending distally onto toes of forefeet. Least width of color of underparts, in eleven specimens, averaging 23 (10-36) per cent of greatest width of color of upper parts; black tip of tail, in six specimens, averaging 36 (31-38) per cent of length of tail-vertebrae.

Skull and teeth.—Male (based on 2 adults, no. 3.2.1.6. from vicinity of San José and no. 11408, U. S. Nat. Mus., from "Costa Rica"): See measurements and plates 25-30; weight, 5.9 grams; basilar length 49 +; zygomatic breadth more than distance between condylar foramen and M1 or than between anterior palatine foramen and anterior margin of tympanic bulla; mastoid breadth less than postpalatal length; postorbital breadth in undiseased skulls less than length of upper premolars (less than distance between posterior borders of P2 and P4) and less than width of basioccipital measured from medial margin of one foramen lacerum posterior to its opposite; interorbital breadth more or less than distance between foramen opticum and anterior margin of tympanic bulla; breadth of rostrum more or less (about equal to) length of tympanic bulla; least width of palate less than length of P4; anterior margin of tympanic bulla as far posterior to foramen ovale as width of 5 upper incisors; height of tympanic bulla less than distance from its anterior margin to foramen ovale; tympanic bulla longer or shorter than (about equal to) lower molar and premolar tooth-row and longer or shorter than (about equal to) rostrum; anterior margin of masseteric fossa directly below posterior border of m2.

Female: Skull of adult unknown.

Comparison of the skull of the male with that of M. f. panamensis has been made in the account of that subspecies. As compared with that of M. f. nicaraguae the skull of M. f. costaricensis is heavier and in every measurement taken is larger. The skull is generally more massive and it follows that most measurements of depth and width are greater in relation to the basilar length as well as actually greater. The individual teeth are larger and the tympanic bullae wider and at their anterior ends are more projected from the braincase. Indeed the skull is more like that of M. f. goldmani than like that of M. f. nicaraguae.

Remarks.—The half dozen ill-prepared skins, with partial skulls inside, of this form in the United States National Museum long were referred either to Mustela brasiliensis or Mustela affinis. It was not until 1912 when Goldman studied these specimens that the distinctive characters of the Costa Rican weasel were recognized and made the basis of the name costaricensis.

M. f. costaricensis is well differentiated from M. f. nicaraguae and M. f. goldmani which occur to the northward and from M. f. panamensis which occurs to the southward and is a large, heavy-skulled, dark-colored animal with white facial markings restricted or absent. In the type specimen and the female from the Candelaria Mountains the white facial markings are only narrow facial bars or a few white hairs, but in the young male from Cervantes there is a well developed bar 6 millimeters wide on each side of the face and a separate nasofrontal spot, 10 x 12 mm. The young female from Cachí has a V-shaped frontonasal spot, on the right side of the face a white bar 5 mm. wide and 17 mm. long connected with the color of the underparts, and on the left side a white spot in front of the ear and another between the ear and eye. White facial markings were not recorded in the other specimens. The color of the upper parts is only a little less dark than those of M. f. panamensis. Owing to the numerous white hairs on the dorsal side of the neck, the nape of the female from the Candelaria Mountains has a frosted appearance not present in other specimens.

M. f. costaricensis is a large animal and among its geographic neighbors is approached in size only by a specimen of panamensis from Boquete, Panamá. Also the young male from Cervantes suggests panamensis in the less flattened interorbital region, but even so is more like costaricensis. The small size of two young males, one from Navarro and the other from the vicinity of San José, is suggestive of M. f. nicaraguae. However, the large size of most of the specimens and the configuration of the skull are more as in M. f. goldmani than in M. f. nicaraguae and thus suggest that the known specimens are of high mountain subspecies. The long black tip of the tail is another point of resemblance to M. f. goldmani, the high mountain subspecies to the north. Perhaps in the lowlands of Costa Rica, there are weasels of another subspecies.

Of the eight skulls examined for malformation of the frontal sinuses, each of the two adults and two subadults shows signs of having the frontal sinuses infested with parasites.

Specimens examined.—Total number, 14, listed by localities from north to south.

Costa Rica: Irazú (Frasu or Irasu on label), 3000 M., 1[4]; Cervantes, 1[2]; San José, 1[91]; vicinity of San José, 2[7]; Azahar Cartago, 1[78]; Tucurrique, 1[7]; Cachí, 1[7]; El Muñco [= Muñeco?] (Río Nivarro [= Navarro?]), 4000 ft., 10 mi. S Cartago, Caribbean Slope, 1[76]; Navarro, 1[91]; Candelaria Mts., 1[75]; no locality more definite than Costa Rica, 3[91].

Mustela frenata panamensis Hall

Long-tailed Weasel

Plates 1, 25, 26, 27, 28, 29 and 30

Mustela frenata panamensis Hall, Proc. Biol. Soc. Washington, 45:139, September 9, 1932; Hall, Carnegie Instit. Washington Publ. 473:109, November 20, 1936.

Mustela brasiliensis, Alston, Biol. Cent. Amer., Mammalia, p. 78, 1879.

Putorius affinis, Bangs, Bull. Mus. Comp. Zoöl., 39:49, April, 1902; Hollister, Proc. Biol. Soc. Washington, 28:143, July 10, 1914.

Mustela affinis, Goldman, Proc. Biol. Soc. Washington, 25:10, January 23, 1912; Hollister, Proc. Biol. Soc. Washington, 28:143, July 10, 1914.

Mustela affinis costaricensis, Allen, Bull. Amer. Mus. Nat. Hist., 35:101, April 28, 1916; Goldman, Smithsonian Miscel. Col., 69 (no. 5): 161, 1920.

Type.—Female, subadult, skull and skin; no. 170970, U. S. Nat. Mus., Biol. Surv. Coll.; Río Indio, Canal Zone, near Gatún, Panamá; February 17, 1911; obtained by E. A. Goldman; original no. 20897.

The skull is complete and unbroken. The left lower incisor is broken off but all the other teeth are present and entire. The skin is well made and seems to be in faded, worn, first, adult pelage.

Range.—Sea level (type locality) to 5800 feet (Boquete, see Bangs [1902:49]); Upper Tropical and Lower Tropical life-zones of Panamá. See figure 29 on page 221.

Characters for ready recognition.—Differs from both M. f. meridana and M. f. costaricensis in darker tone (tone 4 of Oberthür and Dauthenay, pl. 344) of color of upper parts and in convex dorsal outline of skull (Compare figures of mentioned subspecies on plates 25-27).

Description.—Size.—Male: Two adults from Boquete in the Museum of Comparative Zoölogy, nos. 10112 and 10113, measure, respectively, as follows: Total length, 480 and 400; length of tail, 170 and 143; length of hind foot, 52 and 43. Hind feet of two other adult males measure 46 on dried skins. Tail, in two specimens mentioned above, is 55 and 56 per cent as long as head and body. Length of hind foot, in each of three adults, slightly longer than basal length. Corresponding measurements of no. 178970 from Mt. Pirre are: 422, 164, 50. Tail 64 per cent (same per cent as in young male, no. 137514 from Boquete) as long as head and body, and hind foot longer than basal length.

Female: An adult and a young from Chiriquí, nos. 18434 and 18435 (Acad. Nat. Sci. Philadelphia), measure, respectively: Total length, 372, 389; length of tail, 138, 144; length of hind foot, 42, 41. The type specimen measures: Total length, 408; length of tail, 159; length of hind foot, 46.5. Tail 64 per cent as long as head and body, and hind foot longer than basal length.

The average differences in external measurements of the two sexes from the vicinity of Boquete are: Total length, 59; length of tail, 15; length of hind foot, 6.

Externals.—Longest facial vibrissae black and extending beyond posterior border of ear; carpal vibrissae wholly, or in part, black and extending as far as hypothenar pad; hairiness of foot-soles as shown in figure 21.

Color.—Usually, posterior fourth of each upper lip and sometimes few hairs in front of ear, white; sides and top of head and neck posteriorly to, or behind, shoulders, black; dark areas at angles of mouth confluent with color of upper parts; tip of tail, black; remainder of upper parts near (n) Bay of Ridgway and Reddish Black, tone 4, pl. 344 of Oberthür and Dauthenay; chin and lips, whitish; remainder of underparts Warm Buff or near (16´ c) Ochraceous-Buff; near (12) Salmon-Orange in juveniles and small young; color of underparts extending distally on posterior sides of forelegs to wrists, but not to soles, and on hind legs to or slightly below knees. Least width of color of underparts, in seven specimens, averaging 18 (extremes 11-28) per cent of greatest width of color of upper parts; black tip of tail, in five adults and subadults, averaging 45 (extremes 41-50) per cent of length of tail-vertebrae.

Skull and teeth.—Male (based on three adults from Boquete): See measurements and plates 25-30; weight, 5 (4.5-5.4) grams; basilar length, 45.2 (42.8-48.3); zygomatic breadth more or less than distance between condylar foramen and M1 or than between anterior palatine foramen and anterior margin of tympanic bulla; mastoid breadth less than postpalatal length; postorbital breadth more than length of upper premolars and more than width of basioccipital measured from medial margin of one foramen lacerum posterior to its opposite; interorbital breadth not less than distance between foramen opticum and anterior margin of tympanic bulla; breadth of rostrum approximately same (more or less than) length of tympanic bulla; least width of palate less than length of P4; anterior margin of tympanic bulla as far posterior to foramen ovale as width of 4-1/2 to 5-1/2 upper incisors; height of tympanic bulla less than distance from its anterior margin to foramen ovale; length of tympanic bulla more or slightly less than (approximately equal to) length of lower molar and premolar tooth-row or length of rostrum; anterior margin of masseteric fossa directly below posterior fourth of m2.

Female (based on subadult, type specimen and one adult from Siola): See measurements; weight, 3.3 and 2.1 grams; basilar length, 41.3 and 39.3; zygomatic breadth more than distance between condylar foramen and M1 and more or less than (about equal to) that between anterior palatine foramen and anterior margin of tympanic bulla; postorbital breadth more than combined length of upper premolars or than width of basioccipital measured from medial margin of one foramen lacerum posterior to its opposite; least width of palate more than length of P4 (less in the adult); anterior margin of tympanic bulla as far posterior to foramen ovale as width of five upper incisors; height of tympanic bulla less than (about half) distance from its anterior margin to foramen ovale; length of tympanic bulla less than length of lower molar and premolar tooth-row or than rostrum.

The skull of the one adult female from Chiriquí is 58 per cent lighter than the average of the two adult males.

The skull of the male of M. f. panamensis as compared with that of M. f. meridana, is heavier and averages larger in nearly every measurement taken. Relative to basilar length, tooth-rows, orbitonasal length, interorbital breadth and zygomatic breadth averaging narrower. Mastoid breadth always narrower. Tympanic bullae longer, narrower, and usually slightly less protruded. P4 and m1 larger. Dorsal outline of skull, viewed laterally, more convex. Postorbital breadth actually and relatively greater. Postorbital processes, mastoid processes, and sagittal crest not so well developed. Differences between skulls of females, in so far as known, similar to those described between males.

As compared with M. f. costaricensis, M. f. panamensis has a lighter skull averaging smaller in every measurement taken except interorbital breadth, which is greater. Relative to basilar length, width of rostrum, interorbital breadth and depth of skull at plane of upper molars, less. Tympanic bullae shorter, narrower, less protruded. P4, M1, and m1 larger. Dorsal outline of skull, viewed laterally, more convex. Postorbital breadth relatively and actually greater. Postorbital processes, mastoid processes, sagittal crest and lambdoidal crest less developed. No skull of an adult female of M. f. costaricensis is available for comparison.

Remarks.—This subspecies had not been recognized by previous workers because specimens from Panamá were supposed to be Mustela affinis Gray up until 1916, when Allen (1916:100) restricted the type locality of M. affinis to Bogotá, Colombia. At that time Allen referred specimens from Panamá to Mustela affinis costaricensis, and Goldman (1920:161) followed Allen.

The specimens examined show much variation. Part of this is geographic variation. For instance the specimens from Boquete approach M. f. costaricensis in size more than do those from farther south. Too few adult females have been seen to ascertain the amount of secondary sexual variation. Bangs (1902:49) suggested that the sex of no. 10113 was wrongly recorded and that it was not really a male. If so, this would reduce the range of apparent variation in size of males from Boquete by half and bring it into accord with the amount normally existing in adult males from one locality. No. 10113 is adult but the skin shows no mammae which would prove it to be a female instead of a pigmy male. Although even smaller than 10113, the type specimen is so much larger than females of M. f. meridana that I have wondered if it is correctly sexed. However, the fact that it was sexed by E. A. Goldman, a collector of wide experience, lessens the possibility that a mistake was made.

The color of the underparts is more restricted in panamensis than in any other subspecies of the species. Excluding the specimen from Mt. Pirre, the least width of color of the underparts averages 16 (extremes 6-24) per cent of greatest width of the color of the upper parts. This feature, together with the black color, imparts an appearance to the Panamá weasel that is strikingly like that of a mink. M. f. panamensis is one of the two blackest weasels; M. f. aureoventris is the other. Each of these subspecies occurs in a region of heavy rainfall and there clearly is a positive correlation between high humidity and intensity of color. The black tip of the tail, as regards extent, here reaches the maximum attained among Central and South American weasels. The foot soles are less hairy than in any other member of the subgenus Mustela. The tympanic bullae are lower and less inflated than in any other subspecies of the species.

Adequate specimens from central and southern Panamá may reveal the existence of one or more additional subspecies since animals from each of the three localities now represented differ from those from the other two and some of these differences are correlated with geographic position. However, specimens from all three localities agree in several features. For example all of them have the dorsal outline of the skull highly convex, transversely, and, more especially, longitudinally. In this respect they are sharply differentiated from any other American weasel. Nevertheless, M. f. panamensis is clearly a link between the North and South American subspecies and panamensis intergrades with the adjacent subspecies. The large size of the skull and teeth and the slightly more ventrally projected tympanic bullae of no. 10112 from Boquete approach features seen in M. f. costaricensis. The smaller size of skull and teeth of no. 178970 from Mt. Pirre are points of resemblance to M. f. meridana.

The type specimen was selected from a region where M. f. panamensis is thought to have its distinctive characters well developed. The specimen is not adult and, therefore, does not show as many differential characters as does a nontypical adult from Boquete. Nevertheless, the majority of the above mentioned differential characters are shown by the type specimen and an adult from the same place would, it is judged, show all the differential characters better than would an adult from Boquete.

Of the 11 skulls examined, 6 show no signs of having had the frontal sinuses infested with parasites.

Specimens examined.—Total number, 19, listed by localities from north to south and unless otherwise indicated in the United States National Museum.

Panamá: Boquete, 10 (3[75], 1[8], 1[2], 3[4], 1[7]); Río Gariche [é], 5300 ft., 1[1]; Siola, 1[1]; Chiriquí, 1[7]; Río Indio, near Gatún, 1; Mt. Pirre, 3 (2[1]); Calovebora, 1[7] (locality not found, possibly misspelling of Calovébora); no locality more definite than Panamá, 1[4].

Mustela frenata meridana Hollister

Long-tailed Weasel

Plates 25, 26, 27, 37, 38 and 39

Mustela meridana Hollister, Proc. Biol. Soc. Washington, 28:143, July 10, 1914.

Putorius affinis, Robinson and Lyon, Proc. U. S. Nat. Mus., 24:147, October, 1901; Allen, Bull. Amer. Mus. Nat. Hist., 30:256, December 2, 1911.

Mustela affinis, Osgood, Field Mus. Nat. Hist. Publ. 155, zoöl. ser. 10:61, January 10, 1912.

Putorius macrurus, Allen, Bull. Amer. Mus. Nat. Hist., 31:92, April 19, 1912.

Mustela affinis affinis, Allen, Bull. Amer. Mus. Nat. Hist., 35:100, April 28, 1916 (part).

Mustela affinis costaricensis, Allen, Bull. Amer. Mus. Nat. Hist., 35:101, April 28, 1916 (part).

Mustela frenata meridana, Hall, Carnegie Instit. Washington Publ. 473:110, November 20, 1936; Hall, Physis, 16:175, 1939.

Type.—Male, subadult, skull and skin; no. 123341, U. S. Nat. Mus., 1630 meters elevation, Montes de Mérida, near Mérida, Venezuela; August 14, 1903; obtained by S. Briceno.

The skull (plates 25 and 26) lacks the right exoccipital condyle and posterior half of the right zygomatic arch. The teeth all are present, unworn and entire. The skin is well made and complete.

Range.—Near sea level (San Julián) to 8500 feet (Montes de Culata, Mérida, Venezuela), and 9000 feet (Santa Elena, Colombia). Temperate to Subtropical life-zones of Venezuela and northern and western Colombia. See figure 29 on page 221.

Characters for ready recognition.—Differs from Mustela africana stolzmanni by absence of median, longitudinal, abdominal stripe of same color as upper parts, presence of p2 and two roots, rather than one root, on P2; from M. frenata panamensis in lighter color of upper parts (tone 3 rather than tone 4, pl. 344, Reddish Black, Oberthür and Dauthenay), flat rather than convex dorsal outline of skull immediately behind postorbital processes (see pl. 27); from M. f. affinis, in males, by lesser average breadth and length of skull and greater actual and relative size (see measurements) of facial part of skull; from M. f. aureoventris, in males, by lighter upper parts (tone 3 rather than tone 4, pl. 344, Reddish Black, Oberthür and Dauthenay) and by smaller skull and teeth (basilar length less than 45, length of m1 less than 6.3, width of M1 less than 4.8, outside length of P4 less than 5.7).

Description.—Size.—Male: Average and extreme measurements of topotypes (as recorded by collectors on labels, and so uniform as to show them not to be accurate to within more than 5 mm.) are as follows: Total length, 434 (410-460); length of tail, 164 (150-180); length of hind foot, 50 (no variation in collectors' measurements). Tail averages 61 per cent as long as head and body. Length of hind foot more than basal length. Corresponding measurements of no. 22191, a young male from Mérida, measured by Osgood or Conover, are 439, 165, 54. The adult male no. 18703, from Páramo de Tama (eastern boundary of Venezuela) has the following measurements written on the label by Osgood: 404, 150, 47.

Female: Average and extreme measurements of topotypes (as recorded by collectors on labels and so uniform as to show them not to be accurate to within more than 5 mm.) are as follows: Total length, 347 (320-370); length of tail, 128 (120-130); length of hind foot, 40 (no variation in collectors' measurements). Tail averages 57 per cent as long as head and body. Length of hind foot more than basal length. Two females, adult no. 11034 and young no. 11033 from Cincinnati, Santa Marta, Colombia, measured by M. A. Carriker, Jr., measure, respectively, as follows: 371, 330; 140, 140; 38, 36. No. 14463, adult, from Río Zapata, Colombia, measured (by J. H. Batty), 315, 138, 39. No. 32182, adult, from Mira Flores, Cauca, Colombia, measured (by W. B. Richardson), 375, 150, 43.

The average differences in external measurements of the two sexes, at Mérida, are: Total length, 87; length of tail, 36; length of hind foot, 10.

Externals.—Longest facial vibrissae black (few rarely white) and extending beyond ear; carpal vibrissae colored like underparts or upper parts, and not extending beyond apical pad of fifth digit; hairiness of foot soles slightly greater than shown in figure 21.

Color.—As described in Mustela frenata panamensis except that: Posterior fourth of each upper lip rarely, and small spot in front of ear usually, white; black of head proper not extending back of ears and grading insensibly into color of upper parts; anterior half of upper parts of adults "frosted" with numerous white hairs (tick bites?), upper parts near (n) Bay or tone 2 of Reddish Black (pl. 344, Oberthür and Dauthenay) or tone 3 in freshest, unfaded pelage. Least width of color of underparts (in ten males from Mérida) 20 (17-23) per cent of greatest width of color of upper parts. Black tip of tail, in same series, 60 to 75 mm. long, thus longer than hind foot and 41 (40-44) per cent as long as tail-vertebrae.

Skull and teeth.—Male (based on type specimen and seven topotypes, five adults and three subadults): See measurements and plates 25-27; weight, 4.1 (3.8-4.3) grams; basilar length, 43.6 (42.3-44.3); zygomatic breadth more than distance between condylar foramen and M1 or than between anterior palatine foramen and anterior margin of tympanic bulla; mastoid breadth less than postpalatal length; postorbital breadth greater than length of upper premolars or than width of basioccipital measured from medial margin of one foramen lacerum posterior to its opposite; interorbital breadth not less than distance between foramen opticum and anterior margin of tympanic bulla; breadth of rostrum greater than length of tympanic bulla; least width of palate greater than length of P4; anterior margin of tympanic bulla as far posterior to foramen ovale as width of 4 to 5 upper incisors; height of tympanic bulla less than distance from its anterior margin to foramen ovale; length of tympanic bulla more or less than (approximately equal to) alveolar length of lower molar and premolar tooth-row and shorter than rostrum; anterior margin of masseteric fossa posterior to m2 and confined to posterior third (34 per cent average, 32 minimum, 37 maximum) of mandible.

Female (based on four adult topotypes): See measurements and plates 37-39; weight (no. 143665), 2.3 grams; basilar length 37.2 (36.3-38.2); zygomatic breadth more or less than distance between condylar foramen and M1 or than distance between anterior palatine foramen and anterior margin of tympanic bulla; postorbital breadth (sinuses badly infested with parasites) more than length of upper premolars or width of basioccipital measured from medial margin of one foramen lacerum posterior to its opposite; least width of palate more than length of P4; tympanic bulla as far posterior to foramen ovale as combined width of 4 to 5 upper incisors; height of tympanic bulla less than (one half to three fourths of) distance from its anterior margin to foramen ovale; length of tympanic bulla more than length of lower molar and premolar tooth-row.

The skull of the female is 44 per cent lighter than that of the average male.

Comparisons of the skull with those of M. f. panamensis and affinis have been made in the accounts of those subspecies. As compared with the skull of the male of M. f. aureoventris, that of meridana averages smaller in every measurement taken. Indeed, none of the skulls of meridana equals that of aureoventris in basilar length, length of tooth-rows, length of tympanic bulla, depth of skull at anterior margin of basioccipital or at posterior margin of upper molars, or measurements of teeth. Relative to the basilar length, most of the measurements are greater in meridana. Exceptions are the relative length of the tooth-rows, and the two measurements of depth of the skull which average less.

Remarks.—In 1914 when Hollister named this weasel he compared it with M. f. affinis and most of the differential characters which he ascribed to meridana were merely "more than" or "less than" in affinis. In affinis, Hollister included specimens from Chiriquí, Panamá, and the coast of Venezuela. The specimens from these three places were referred by Allen (1916:101) to Mustela affinis costaricensis, and he restricted (op. cit.:100) the type locality of Mustela affinis affinis to Bogotá, Colombia, and synonymized Mustela meridana with M. a. affinis. Hollister probably would not have named meridana had he had specimens from Bogotá for comparison and had he regarded them as topotypes of affinis for the difference is slight. Nevertheless, within the large geographic range of M. f. meridana there is some geographic variation. There is more of such variation in the color of the pelage than in shape and size of the skull. The specimen from San Julián is darker than the average and in this respect approaches true panamensis. San Julián is situated at a relatively low elevation on the coast of Venezuela.

M. f. meridana so closely resembles M. f. affinis that the writer has no quarrel with anyone who would synonymize meridana. However, as represented by topotypes, the two races unquestionably are, on the average, different, and specimens from the southeastern part of the range of affinis probably are individually distinguishable from topotypes of meridana.

Variation in the skulls of the series from Mérida is relatively small. This applies to both males and females. The external measurements recorded by native collectors are not accurate to within more than five millimeters but, considering this, variation in external measurements also seems to be slight. The difference in size of the two sexes appears to be uniformly greater than in weasels from Central America. The twenty-six topotypes show that the color and color pattern are relatively uniform. All are of nearly the same tone except juveniles or young which are, as in the case of panamensis, much brighter colored on the underparts. Also, the young have darker-colored upper parts. The adults, without exception, have numerous white hairs scattered over the back of the head, neck and between the shoulders. I have no trustworthy evidence to support the suggestion that these white hairs are the results of tick bites or that they are caused by other parasites which damage the hair follicles. The white facial markings vary relatively little in the 45 specimens carefully examined in this regard. Also, the variation in color pattern of the two sides of the head is small. Indeed, within rather narrow limits, the color of the two sides of the head is the same in every specimen except two. In these two the white spots anterior to the ears are confluent with the color of the underparts. Only one specimen, no. 21342, has a white spot between the eyes and this spot is small. Ten of the twenty-six specimens have a definite white spot or band in front of each ear. Two specimens have such a spot on one side only. The dark spots at the angles of the mouth are present on two sides in three specimens and on one side only in three others. The mentioned spots are, then, present nine out of a possible fifty-two times. When the spots are absent, dark color usually is present in the required area but is confluent with the color of the upper parts.

A young male from San Julián, Robinson and Lyon (1901:147) state ". . . was shot . . . as it ran over some bowlders in a ravine. Its eyes shone with the same greenish light as do the eyes of our common weasel, and it emitted the same strong odor." No. 14463, Am. Mus. Nat. Hist., from Río Zapata, Colombia, according to data on the label, was "taken in timber belt in valley in balk hills" and the native name is Cosonebi. Two specimens taken on the Páramo de Tama, head of Tachira River, Venezuela and Colombia are commented on by Osgood (1912:61) as follows: "One . . . was caught in a steel trap baited with birds and set by the side of a rushing mountain stream. . . . The other was shot in midday as it came prowling about our 'house' in the clearing. . . ."

Of the thirty-three skulls before me, twelve have the frontal sinuses malformed by parasites. These twelve include most of the adults for few of the subadults and fewer of the young show pathologic conditions in the frontal region.

Note on localities.—Several of the localities in Colombia mentioned in "Specimens examined" are described and located by Chapman (1917:640-656, pl. 41) in his "Distribution of Bird-life in Colombia." Place names for Colombia on labels, not found on any map, or duplicated names of which I can not certainly select one, are Río Barrotow, Río Oscuro, Río Zapata, Río Japata, Guasca and El Baldro. Sonson may or may not be the town of that name situated some eighty miles northwest of Bogotá and on the east flank of the Central Andes west of the Magdalena River on the drainage of the Cauca River. In Venezuela most of the specimens from Mérida are labeled 1630 meters, Montes de Mérida. San Julián is some seven miles east of La Guaira (see Robinson and Lyon, 1901:136). San Esteban is located a little way back from the coast between Puerto Cabello and Valencia. Páramo de Tama is on the Venezuelan-Colombian border near the source of the Tachira River (see Osgood, 1912:35). Mt. Duida is shown as at 3° 30´ N and 65° 40´ W by Chapman (1931:13) and Mt. Auyán-tepui as near 5° 15´ N and 62° 50´ W by Chapman (1937:760).

Specimens examined.—Total number, 78, arranged by localities from north to south and unless otherwise indicated in the British Museum of Natural History.

Venezuela: San Julián, 1[91]; Carácas, 2; Galipare, Cerro del Avila, 6500 feet, 1; San Esteban, 1[2]; Mérida, 45 (10[91], 14[2], 10[4], 2[60], 2[14], 1[78]); Páramo de Tama, 1[60].

Colombia: Páramo de Tama, 1[60]; Cincinnati, 3[9]; Valdiva, 3800 ft., 1; Medellín, 2; 7200 ft., Barro Blanco, 1[2]; Santa Elena, 9000 ft., 1[2]; Santa Elena, 1[2]; Sonson, 2 (1[91], 1[2]); Mt. Auyan-tepuy, 1[2]; Pueblo Rico, 5200 ft., 1[91]; Mira Flores, 1[2]; Jerico, near Cauca River, 1; Tornel, 20 mi. NE Quitichao, 1; Mt. Duida, 1[2]; El Tambo, Cauca, 1[78]; El Baldro, 1[2]; Río Japata, 2[2]; Río Zapata, 4500 ft., 1; Río Oscuro, 3300 ft., 1; Río Barrotow, 3300 ft., 1; Guasca, 1[75]; no locality more definite than Colombia, 1.

Mustela frenata affinis Gray

Long-tailed Weasel

Plate 30

Mustela affinis Gray, Ann. and Mag. Nat. Hist., 14(ser. 4):375, 1874.

Putorius (Gale) brasiliensis frenatus, Coues, Fur-bearing animals, p. 142, 1877 (part).

Putorius affinis, Merriam, N. Amer. Fauna, 11:31, June 30, 1896.

Mustela affinis affinis, Allen, Bull. Amer. Mus. Nat. Hist., 35:100, April 28, 1916; Allen, Bull. Amer. Mus. Nat. Hist., 35:220, May 31, 1916.

Mustela frenata affinis, Hall, Carnegie Instit. Washington Publ. 473:110, November 20, 1936; Hall, Physis, 16:175, 1939.

Type.—Male, adult, skull with skin; no. 54.1.11.3 (skull originally numbered 195d, later 54.6.3.4), Brit. Mus. Nat. Hist.; Colombia [given as new Granada in original description]; purchased from Mr. S. Stevens. Type locality restricted by Allen (1916:99) to Bogotá, Colombia.

The skin is in a good state of preservation and has been made over into a conventional study specimen from a mount on exhibition. Exposure to light when mounted probably accounts for the faded color. The skull (plate 30) lacks the middle 9 mm. of the right zygomatic arch, occiput, basioccipital and posterior two-thirds of the left tympanic bulla. The teeth all are present and entire.

Range.—Four thousand six hundred feet (Quetame) to 9154 feet (El Carmen), Tropical to Temperate life-zones of eastern Andes of Colombia. See figure 29 on page 221.

Characters for ready recognition.—Differs from Mustela africana stolzmanni by absence of median, longitudinal, abdominal stripe of same color as upper parts, by presence of p2 and by two roots rather than one root on P2; from M. frenata meridana, in case of males, by, on average, greater breadth and length of skull and lesser actual and relative size (see measurements) of facial part of skull; from M. f. aureoventris by lighter-colored upper parts (tone 2 rather than tone 4, pl. 344, Reddish Black of Oberthür and Dauthenay); from M. f. macrura by darker color (Reddish Black, tone 2, pl. 344, Ober. and Dauth., rather than Chocolate, tone 3, pl. 343, Ober. and Dauth.).

Description.—Size.—Male: Measurements in life, estimated from dried skins, are: Total length, 455; length of tail, 175; length of hind foot, 52. Proportions of parts supposedly as described in Mustela frenata meridana.

Female: Estimates from two dried skins: Total length, 365; length of tail, 135; length of hind foot, 43. Proportions of parts supposedly as described in Mustela frenata meridana.

The estimated differences in external measurements of the two sexes are: Total length, 90; length of tail, 40; length of hind foot, 9.

Externals.—As described in Mustela frenata meridana.

Color.—As described in Mustela frenata panamensis except that: posterior fourth of each upper lip and spot in front of each ear white in approximately half of the specimens; black of head proper not extending back of ears and grading insensibly into color of upper parts; upper parts near (n) Bay, or tone 2 of Reddish Black (pl. 344, Oberthür and Dauthenay). Least width of color of underparts (in five males from vicinity of Bogotá) 24 (15-29) per cent of greatest width of color of upper parts. Black tip of tail, in same series, 60 to 75 mm. long, thus longer than hind foot and averaging 38 per cent as long as tail-vertebrae.

Skull and teeth.—Male (based on three adults and two subadult topotypes): See measurements and plate 30. As described in Mustela frenata meridana except that: Weight, 4.5 grams (estimated); basilar length 45.8±; interorbital breadth not greater than distance between foramen opticum and anterior margin of tympanic bulla (type as in meridana where interorbital breadth is more than distance between foramen opticum and anterior margin of tympanic bulla); least width of palate not less than length of P4; masseteric fossa confined to posterior two-fifths (38 to 40 per cent; average 39 per cent) of mandible and not extended anteriorly to middle of m2.

Female: No adults examined.

As compared with M. f. meridana the skull of the male is larger, to the average amount of 2.2 mm. in basilar length and 1.2 mm. in zygomatic breadth of adults; length of tooth-rows and mastoid breadth average greater but relatively less; breadth of rostrum, interorbital breadth and orbitonasal length average actually and relatively less. Thus the skull of affinis is longer and broader, but the facial region is actually, as well as relatively, smaller. As compared with the skull of the male of M. f. aureoventris, that of M. f. affinis is about the same in basilar length. However, in no specimen of affinis are the measurements of length of tooth-rows or breadth of rostrum, actually, or relatively, as great as in aureoventris. The same is true of all measurements taken of M1, P4 and m1. The specimens from the vicinity of Quito and north of there, although referred to macrura, are nearly as dark as typical affinis, approach affinis in cranial characters, and indicate intergradation of affinis with macrura.

Remarks.—Mustela affinis was named by John Edward Gray in 1874 (p. 375) on the basis of a specimen from New Granada. Although usually synonymized with Mustela brasiliensis by later authors until 1896 when Merriam (1896:31) applied the name to weasels from Costa Rica, nearly all the South American and several of the Central American weasels have, at one time or another, had Gray's name, affinis, applied to them. Gray, in 1865 (p. 115) when giving measurements of Mustela aureoventris, probably mentioned the specimen, that later became the holotype. In 1916 (p. 98) Allen restricted the type locality to Bogotá, Colombia. Allen's action was a necessary procedure in clearing up the systematics of South American weasels and was based on good grounds. As set forth by Allen (loc. cit.), and more in detail by Chapman (1917:642), Bogotá has long been the shipping point for Colombian vertebrate specimens, many of which were obtained in the mountains to the east. Allen (1916A:220) quotes Thomas as saying that the type specimen was purchased from Stevens at about the same time that a number of Colombian birds were purchased from the same dealer. Also, specimens from Bogotá agree with Gray's description of the type specimen.

Mustela frenata affinis, as here defined, constitutes one of the several slight geographic variants met with, on the sides of, and between, the three north and south mountain chains of Colombia. The others are lumped under the name Mustela frenata meridana. M. f. affinis, in common with specimens from the northern part of the range of macrura has large teeth. Weasels of all of the region from Quito to Bogotá have large teeth. To the north there is the smaller-toothed meridana and to the south the smaller-toothed macrura grading into the still smaller-toothed agilis, and boliviensis.

Two skins, without corresponding skulls, from Caqueta are lighter colored than any others of affinis; possibly the skins are faded by exposure to light. Since they probably come from an elevation of less than 1000 feet in the Amazonian region, they may pertain to another subspecies.

Complete, unbroken, skulls of affinis are needed to ascertain the degree to which affinis and meridana differ in cranial features. The several specimens from the immediate region of Bogotá show well the color and the color pattern but lack collectors' measurements.

None of the ten skulls examined shows malformation of the frontal region due to infestation of the frontal sinuses by parasites. Possibly three of the four adults were infested, although not severely.

Specimens examined.—Total number, 27, arranged by localities from north to south and unless otherwise indicated in the United States National Museum.

Colombia: El Carmen, 1[2]; W. Cundinamarca, 1[7]; Muzzo [= Muzo?], 1[4]; Bogotá, 1; Castillo, near Bogotá, 1[7]; Fambrias, near Bogotá, 1[75]; Bogotá district, 1[2]; Choachí, 9 (1[75], 2[7], 1[84]); Páramo de Choachí, 2 (1[2], 1[84]); Laguna del Verjón (= City of Bogotá), 1[75]; Quetame, 2[2]; Fusagasuga, 1; Caqueta, 2[2]; no locality more definite than Colombia, 3 (1[7]).

Mustela frenata aureoventris Gray

Long-tailed Weasel

Plates 27, 28 and 29

Mustela aureoventris Gray, Proc. Zoöl. Soc. London, 1864:55, pl. 8, 1864; Gray, Proc. Zoöl. Soc. London, 1865:115, 1865.

Putorius (Gale) brasiliensis var. aequatorialis Coues, Fur-bearing animals, p. 142, 1877, part? ("merely as a substitute for Gray's [supposedly] preoccupied name," that is, aureoventris).

Mustela affinis costaricensis, Allen, Bull. Amer. Mus. Nat. Hist., 35:101, April 28, 1916 (part).

Mustela macrura, Lönnberg, Arkiv för Zool., 14 (no. 4):11, 1921 (part ?).

Mustela frenata aureoventris, Hall, Carnegie Instit. Washington Publ. 473:110, November 20, 1936; Hall, Physis, 16:175, 1939.

Type.—Probably female, juvenile, skull with skin, no. 64.6.6.3 (formerly 1432a), British Mus. Nat. Hist.; probably Subtropical Life-zone of western Ecuador (locality given as Quito, probably because received from that place).

The skin, once exhibited as a mount, has lost some hair from the back and other parts of the body and is not suitable for remaking into a conventional study specimen. The skull lacks the occiput, basioccipital, premaxillae, upper incisors, two of the lower incisors, all of the canines, premolars 2/2 on both sides, right P3, left p3, and has the left jugal mesially defective. The premolars present are not all fully emerged.

Range.—Pacific coastal regions of Ecuador and Colombia; Subtropical and Tropical life-zones. See figure 29 on page 221.

Characters for ready recognition.—Differs from Mustela africana stolzmanni by absence of median, longitudinal, abdominal stripe of same color as upper parts, by presence of p2 and by two rather than one root on P2; from Mustela frenata macrura by Reddish Black, tone 4, plate 344 rather than Chocolate, tone 3, pl. 343 (of Oberthür and Dauthenay), or slightly darker color of upper parts; from M. f. affinis and M. f. meridana by darker color (tone 4 rather than tone 2, Reddish Black of Ober. and Dauth.) of upper parts and larger size of teeth (M1 with length more than 2.4 and breadth more than 4.7; P4 with outside length more than 5.9; length of m1 more than 6.2).

Description.—Unless otherwise stated, information concerning this subspecies is derived from the one referred specimen available, a young male, no. 34677, Amer. Mus. Nat. Hist.

Size.—Male: Total length, 470; length of tail, 160; length of hind foot, 50. Tail 51 per cent as long as head and body.

Female: Not known.

Externals.—Longest facial vibrissae black and reaching beyond ear. Carpal vibrissae reaching to or beyond apical pad of fifth digit; hairiness of foot soles slightly less than shown in figure 20.

Color.—Sides and top of head and neck posteriorly to shoulders black; white facial markings represented by only five white hairs anterior to right ear, one anterior to left ear and three far back on forehead; dark areas at angles of mouth confluent with color of upper parts; tip of tail black; remainder of upper parts near (n) Bay or Reddish Black, tone 4 of Oberthür and Dauthenay, pl. 344; chin whitish; remainder of underparts Warm Buff, deep orange in juvenile, type specimen, according to Gray (1864, pl. 8); color of underparts extending distally on posterior sides of forelegs to wrists but not reaching foot soles and on hind legs to or slightly below knees. Least width of color of underparts equal to 15 per cent of greatest width of color of upper parts. Black tip of tail equal to 27 per cent of length of tail-vertebrae.

In color, no. 34677 is, to me as it was to Allen (1916:101), indistinguishable from the darkest specimens (nos. 178970 and 10112) of M. f. panamensis. Therefore, M. f. aureoventris is one of the two darkest subspecies of weasels.

Skull and teeth.—Male: See measurements and plates 27-29; weight, 4.3 grams; basilar length, 45.8; zygomatic breadth approximately equal to distance between condylar foramen and M1 and to distance between anterior palatine foramen and anterior margin of tympanic bulla; mastoid breadth less than postpalatal length; postorbital breadth more than length of upper premolars and greater than width of basioccipital measured from medial margin of one foramen lacerum posterior to its opposite; interorbital breadth greater than distance between foramen opticum and anterior margin of tympanic bulla; breadth of rostrum less (at least in young specimen) than length of tympanic bulla; least width of palate seldom if ever greater than length of P4; anterior margin of tympanic bulla as far posterior to foramen ovale as width of three (including I3) upper incisors; height of tympanic bulla not greater than distance from its anterior margin to foramen ovale; length of tympanic bulla more than length of lower molar and premolar tooth-row and shorter than orbitonasal length; anterior margin of masseteric fossa below anterior half of m2.

Skulls of males of M. f. aureoventris, and Mustela frenata macrura from the vicinity of Quito so closely resemble one another as not to be distinguished with the material now available, although the teeth of aureoventris are larger. Comparisons of the skulls of males with those of M. f. meridana and affinis, which are readily distinguishable from those of aureoventris, have been made in the accounts of those subspecies.

Female: Skull of adult unknown.

Remarks.—This subspecies of the Tropical Life-zone, or at least the Subtropical Life-zone, of Ecuador, in certain cranial characters resembles Mustela frenata macrura of the Temperate Life-zone. The two differ markedly in color. Nevertheless, a large number of the specimens collected in Ecuador are intermediate in color as well as in zonal distribution.

The type specimen is young or a juvenile. The measurements of no. 34677 from Gualea indicate an animal similar in size to M. f. affinis. Gray (1864:55) states that the type specimen measures "Length of body and head 6 inches, of tail 4-1/2 inches." The plate (pl. 8) accompanying Gray's original description (loc. cit.) is marked one-half natural size and represents the animal as having a head and body length of eight and one-half inches. One year later Gray (1865:115) gives the measurements of this species as "Length of body and head 12, tail 8 inches." Since he had at this time another specimen, larger than the type specimen (which specimen later, probably, became the type of Mustela affinis Gray), the larger measurements probably were taken from it.

Geographically, and as regards cranial characters, Mustela frenata aureoventris is most closely related to M. f. affinis and to the northern section of M. f. macrura, but in color to M. f. panamensis. M. f. aureoventris and M. f. panamensis are the two darkest-colored subspecies and each occurs in a region of extremely heavy rainfall. There is a skin only, no. 32620, Amer. Mus. Nat. Hist., from Munchique, obtained on June 1, 1911, which is appreciably darker than specimens of M. f. affinis in corresponding pelage and is intermediate between M. f. affinis and M. f. aureoventris in color as it is geographically. The specimen measures 495, 202, 52.

The name Mustela aureoventris Gray has been regarded by most authors as preoccupied by Mustela auriventer Hodgson (1841:909). However, the writer is not of this opinion and agrees with Thomas (1920:224) that "The name aureoventris is not invalidated by the auriventer of Hodgson, as, apart from 'one-letterist' differences, its first half comes from the adjective aureus, while Hodgson's name is based on the substantive aurum, so that not only the spellings but the derivations are different." The spelling of Gray's name should be aureoventris for this is the spelling in the original description which in pagination precedes the colored plate of the animal that is labeled Mustela aureoventris. Putorius brasiliensis var. aequatorialis Coues (1877:142) is the only name known to the writer that has been proposed as a substitute for Mustela aureoventris Gray.

Thomas (1920:224) treats Mustela macrura Taczanowski as a synonym of Mustela aureoventris Gray. Allen (1916:101) also treats the two names as applying to the same kind of weasel but regards aureoventris as preoccupied and therefore uses the name macrura. Taczanowski's original description (1874:311) and plate of Mustela macrura indicate an animal that is lighter colored than M. f. affinis. Gray's original description (1864:55) and plate of aureoventris indicate an animal that is darker colored than M. f. affinis. Indeed Gray (1865:115) in speaking of the type of aureoventris as compared with an adult from New Granada [= Colombia] that probably later became the type specimen of Mustela affinis, states: "The young from Quito is much darker than the adult;. . . ." Comparison of the plates accompanying the original descriptions of aureoventris and macrura well illustrate the difference stated in the written descriptions. My examination of the type specimens of M. macrura and M. f. aureoventris shows them to have been fairly accurately portrayed in the plates accompanying the original descriptions. Accordingly the two names are used for the two kinds of animals which appear, however, to be only subspecifically distinct.

Comparison of Gray's plate (1864, pl. 8) with the available specimens from South America indicates that the name aureoventris is based on an individual that is lighter colored than no. 34677 Amer. Mus. Nat. Hist., from Gualea, Ecuador, but on one which resembles no. 34677 more than it does the lighter-colored specimens from the Temperate Zone of Ecuador and northern Perú. Because Quito, Ecuador, is in the Temperate Life-zone and because the available specimens from this zone in Ecuador and northern Perú are distinctly lighter colored than Gray's plate representing the type of aureoventris shows this specimen to be, it is judged to have come from an altitude lower than that of Quito (9350 feet, according to Chapman, 1926:717); probably it came from the Subtropical Life-zone of Ecuador. Indeed Gray (1864:55) did not say that the specimen was collected or obtained at Quito but that it was ". . . received from Quito. . . ." Chapman (1926:717) has pointed out that Quito, since 1846 has been the distributing point for bird skins which specimens ". . . come from the vicinity of the city, from the 'Napo' region on the Amazonian slopes of the Andes, and from Nanegal, Gualea, and other localities on the Pacific side rarely below the Subtropical Zone." It is also pointed out that only some of the specimens are labeled with their approximate place of capture and that even then these localities cannot be accepted as definite; they indicate mainly whether the specimen is from the eastern or western side of the Andes.

The above mentioned considerations and information gained by study of the specimens cause me to think that the type is an intergrade tending toward the lighter-colored Mustela f. macrura of the Temperate Zone although sufficiently dark to be referred to the dark subspecies represented by no. 34677 Amer. Mus. Nat. Hist., from Gualea, Ecuador.

The skull of no. 34677 shows no infestation of the frontal sinuses by parasites.

Specimens examined.—Total number, 3, as follows:

Ecuador: Gualea, 1, Amer. Mus. Nat. Hist.

Colombia: 8325 ft., Munchique, 1, Amer. Mus. Nat. Hist. In the British Museum of Natural History, the type, (1).

Mustela frenata helleri Hall

Long-tailed Weasel

Plates 27, 28 and 29

Mustela frenata helleri Hall, Proc. Biol. Soc. Washington, 48:143, August 22, 1935; Hall, Carnegie Instit. Washington Publ. 473:110, November 20, 1936; Hall, Physis, 16:175, 1939.

Type.—Male, adult, skull and skin; no. 24133, Field Mus. Nat. Hist.; 3000 feet, Hacienda San Antonio, Río Chinchao, Perú; August 22, 1922. Obtained by Edmund Heller. Original no. 6589.

The skull (plates 27-29) is complete and unbroken. The teeth all are present, entire and but slightly worn. The skin is well made, unfaded, and in good condition.

Range.—Three thousand feet (type locality) to 6700 feet (Ambo), Tropical and Subtropical life-zones of eastern Perú. See figure 29 on page 221.

Characters for ready recognition.—Differs from Mustela africana stolzmanni by absence of median, longitudinal, abdominal stripe of same color as upper parts, presence of p2 and two roots rather than one root on P2; from Mustela frenata macrura by darker color (Carbon Brown, tone 3, pl. 342 rather than Chocolate, tone 3, pl. 343, Oberthür and Dauthenay) of upper parts.

Description.—Size.—Male: Measurements of the type specimen and topotype, no. 24132, are, respectively, as follows: Total length, 382, 418; length of tail, 152, 164; length of hind foot, 52, 48. Tail 66 and 65 per cent as long as head and body. Hind foot more than basal length.

Female: Measurements of two referred females, no. 24134 from Ambo and no. 24136 from Huanuco, are, respectively, as follows: Total length, 328 and 303; length of tail, 118 and 103; length of hind foot, 39 and 38.5. Tail 56 and 51 per cent as long as head and body. Hind foot shorter than basal length.

The average differences in external measurements of the two sexes are: Total length, 85; length of tail, 49; length of hind foot, 11.

Externals.—Longest facial vibrissae black and extending beyond ear; carpal vibrissae same color as upper parts and extending to apical pad of fifth digit; hairiness of foot-soles as shown in figure 20.

Color.—Rarely a few white hairs anterior to each ear; posterior fifth of each upper lip white; top of head, posteriorly to slightly behind ears, black, grading into color of upper parts of body; dark spots at angles of mouth absent; tip of tail black; remainder of upper parts near (n) Argus Brown and Carbon Brown, tone 3 (pl. 342, Oberthür and Dauthenay); chin whitish; remainder of underparts Warm Buff; color of underparts extends distally on posterior sides of forelegs to wrists but not reaching foot-soles and on hind legs to slightly below knees. Least width of color of underparts 24 per cent of greatest width of color of upper parts in each of two males and 19 to 30 per cent in three females. Black tip of tail longer than hind foot and averaging 40 (39-42) per cent of length of tail-vertebrae.

Skull and teeth.—Male (based on type specimen and adult no. 24132): See measurements and plates 27-29. As described in Mustela frenata macrura except that: Weight, 4.5 (4.2 and 4.8); basilar length, 44.6 (44.0-45.3); zygomatic breadth more than distance between condylar foramen and M1 or than between anterior palatine foramen and anterior margin of tympanic bulla; breadth of rostrum more or less than (approximately equal to) length of tympanic bulla; height of tympanic bulla less than distance from its anterior margin to foramen ovale; length of tympanic bulla less than length of rostrum; anterior margin of masseteric fossa posterior to m2 by length of that tooth.

Female (based on nos. 24134 to 24136): See measurements. As described in Mustela frenata macrura except that: Weight, 1.7 (1.5-1.9) grams; basilar length, 36.5 (35.3-38.1); zygomatic breadth less than distance between anterior palatine foramen and anterior margin of tympanic bulla; least width of palate more or less than (approximately equal to) outside length of P4; length of tympanic bulla less than length of rostrum.

The skull of the female averages 62 per cent lighter than that of the male.

The skull of the male is generally large and heavy as are the teeth. Comparison with macrura is made in the account of that subspecies. From males of affinis those of helleri differ in: skull shorter; breadth of rostrum and interorbital breadth actually and relatively greater.

Remarks.—The five specimens examined of this subspecies were taken by Edmund Heller for the Field Museum of Natural History in 1922 and 1923. It is to honor his contributions to mammalogy that the subspecies is named helleri. No. 24135 is the specimen carried as a pet for some time by Mr. and Mrs. Heller and of which Mrs. Heller (1924:481) has given an account.

This subspecies is insufficiently known, especially as to geographic range; probably it occupies a considerable range in the Tropical Life-zone along the eastern base of the Andes. The three females, two from Ambo and one from Huanuco, come from a much higher altitude than do the two males and the climate is said to be arid at Ambo and Huanuco. The skulls of the females are 62 per cent lighter and correspondingly smaller in measurements, than those of males. This difference is more than that found in any other South American weasel and it may be that the females are of a subspecies other than helleri.

The type specimen has a broad skull with major proportions strikingly like those of Mustela stolzmanni. Possibly the similar climatic conditions under which the two live have left their impress in similar fashion in this part of each of the two species. The teeth, tympanic bullae, and certain other parts of the skull are, however, so differently proportioned as to show that the skulls represent two species. The referred male has a much longer skull than the type specimen and the relative proportions of breadth and depth of the two skulls differ widely. Judging from large series of weasels examined from localities outside the range of M. f. helleri, the two skulls probably represent almost the maximum of individual variation occurring in one subspecies.

The dark color is as might be expected since helleri inhabits the humid Tropical Zone.

None of the five skulls shows signs of having had the frontal sinuses infested by parasites.

Specimens examined.—Total number, 5, all in the Field Museum of Natural History.

Perú: 3500 ft., Hacienda Buena Vista, Río Chinchao, 1; 3000 ft., Hacienda San Antonio, Río Chinchao, 1; Huanuco, 1; Ambo, 2.

Mustela frenata agilis Tschudi

Long-tailed Weasel

Plates 27, 28, 29, 39 and 40

Mustela agilis Tschudi, Fauna Peruana, p. 110, 1844; Gray, Proc. Zoöl. Soc. London, 1865:113, 1865; Taczanowski, Proc. Zoöl. Soc. London, 1874:311, 1874; Taczanowski, Proc. Zoöl. Soc. London, 1881:648, 1881; Allen, Bull. Amer. Mus. Nat. Hist., 35:104; April 28, 1916; Thomas, Proc. U. S. Nat. Mus., 58-224, 1920.

Mustela macrura, Allen, Bull. Amer. Mus. Nat. Hist., 35:103, April 28, 1916.

Mustela frenata agilis, Hall, Carnegie Instit. Washington Publ. 473:110, November 20, 1936; Hall, Physis, 16:176, 1939.

Type.—No type specimen, or type locality more restricted than cold, barren highlands of the Cordillera [referring to Perú] designated.

Range.—High, barren Cordillera of Perú (see Tschudi, orig. descr.); as here restricted, Temperate Life-zone and higher in western Andes and intermountain valleys of Perú. See figure 29 on page 221.

Characters for ready recognition.—Differs from Mustela frenata macrura by lighter color (Chocolate, tone 2 rather than 3, pl. 343, Oberthür and Dauthenay) of upper parts; length of upper tooth-rows, in females, less than 13; inside length of P4 more than 4.6; from M. f. aureoventris by smaller teeth (maximum size just given for agilis); from M. f. boliviensis by lighter color, upper parts being Chocolate, tone 2, pl. 343, rather than tone 4 or darker of Carbon Brown, pl. 342 (Oberthür and Dauthenay).

Description.—Size.—Male: The stuffed skin of an adult, from Lima, measures: Total length, 460; length of tail, 125; length of hind foot, 45.7. A skin alone from Huarochirí has a body, as now stuffed, 277 mm. long. The tail is missing and the bones of the hind feet have been removed.

Female: The mounted specimen, no. 565, Mus. Polonais d'Hist. Nat., yields measurements, taken by me, as follows: Total length, 250; length of tail, 75; length of hind foot, 32.5. The female, no. 21147, from Macate, measures, 300, 102, 34.

Externals.—Longest facial vibrissae, either dark-or light-colored and extending beyond ear; carpal vibrissae either dark-or light-colored and extending to apical pad of fifth digit; hairiness of foot soles as shown in figure 20.

Color.—Tschudi's description of the color is, in substance, as follows: Head, back and tail reddish gray; base of hair gray, followed by broader grayish-yellow ring and then reddish-brown tip; nose simply dark brown or upper lips edged with white; throat, breast, belly and higher parts of inner sides of extremities whitish gray, at times wholly gray, bases of hairs always gray; feet darker than body, almost chestnut brown; tail darker on tip than at base; ears externally dark brown, internally whitish.

No. 565 possibly somewhat faded from exposure to light, has all the upper parts near (14´ j) Ochraceous-Tawny or Cinnamon, and tone 4 of Oberthür and Dauthenay, plate 323; posterior half of each upper lip white; no other white facial markings present; dark spot at each angle of mouth (one spot confluent with color of upper parts); tip of tail probably black (tip missing); underparts white, belly probably originally with slight tinge of yellow or allied color; color of underparts extending distally on forelegs to feet and onto upper sides of toes and on hind legs to just above heels. Least width of color of underparts equal to about one-fourth of greatest width of color of upper parts.

No. 21147, subadult, from Macate, has a white band confluent with the underparts extending anterodorsally anterior to each ear and the posterior third of each upper lip white. Top of head near (n) Mars Brown, and Carbon Brown, tone 3 (pl. 342, Ober. and Dauth.); tip of tail black; remainder of upper parts near (16" j) Tawny-Olive, and Chocolate (tone 2, of pl. 343 of Ober. and Dauth.) or Raw Umber (tone 3 of pl. 301 of Ober. and Dauth.); anterior half of underparts, including posterior sides of forelegs and antipalmar faces of forefeet, white; remainder of underparts tinged with Warm Buff and extended on posterior legs almost to ankles.

No. 8.1.10.1., male adult, from Lima, is also light colored, and as described in no. 21147, except that left side of head has a white spot rather than bar; posterior eighth of each upper lip white; white frontonasal spot present, 11 x 11 mm.; antipalmar faces of forefeet spotted with brown color of upper parts; color of underparts extending distally on hind legs along medial side of foot to point halfway between heel and tip of inner toe.

No. 13257 from Huarochirí in color and color pattern closely resembles no. 21147. It differs from no. 21147 in slightly lighter color of upper parts, entirely white underparts, less extension of color of underparts onto forefeet, few white hairs instead of white band in front of each ear; color of underparts more restricted.

In each of the four specimens, the least width of the underparts, expressed as a percentage of the upper parts, is as follows: no. 13257, 11 per cent; no. 21147, 29 per cent; no. 565, 31 per cent; no. 8.1.10.1., nineteen per cent.

Skull and teeth.—Male (based on no. 8.1.10.1.): See measurements and plates 27-29. As described in Mustela frenata macrura except that: Weight 4.1 grams; basilar length, 42.5; zygomatic breadth more than distance between anterior palatine foramen and anterior margin of tympanic bulla; mastoid breadth less than postpalatal length; tympanic bullae shorter than rostrum.

Female (based on no. 21147): See measurements and plates 39 and 40. As described in Mustela frenata macrura except that: Weight (no. 21147, subadult), 1.5 grams; basilar length, 35.2; least width of palate less than outside length of P4; tympanic bulla as far posterior to foramen ovale as combined width of five upper incisors; no. 565 answers to the same description but differs from no. 21147 in greater basilar length and larger tympanic bullae which are slightly more projected, at their anterior margins, from the braincase.

To judge from the skull of the female from Macate and the skull of the male from Lima, the skull and teeth of agilis are smaller than in any other South American subspecies of Mustela frenata, except M. f. boliviensis.

Remarks.—Tschudi almost certainly used the name Mustela agilis in a composite sense. His statement (see quoted matter below) about the marked variation in color of this species, as represented by the skins carried by the Indian women as purses, indicates that the forms here designated as Mustela macrura, M. helleri and possibly others additional to the one here called agilis were included by him under the name Mustela agilis. Taczanowski took account of Mustela agilis when he described other species from Perú. Allen (1916:104) and Thomas (1920:224) were not convinced that Mustela agilis and Mustela macrura were distinct species or subspecies.

Search on August 28, 1937, in the Musée d'Histoire Naturelle, at Neuchatel, Switzerland, by Mr. Théodore Delachaux, assistant there, and the writer, revealed no trace of weasels from Tschudi's collection, although some other specimens of mammals that he figured in the "Fauna Peruana" are preserved in that Museum. Not only were the collections of specimens examined but the new catalogue and old catalogue of mammals were vainly searched for mention of weasels deposited by Tschudi. Later, at the British Museum of Natural History, on p, 105 of a personal notebook, of the late Mr. Oldfield Thomas, record was found of his fruitless search for the same specimens of Mustela in May, 1902, at Neuchatel.

Although Tschudi certainly used the name Mustela agilis in a composite sense, as subspecies are at present understood, his description most nearly applies to the light-colored animals from western Perú—the lightest colored of any South American weasels seen. They are of approximately the same color as North American subspecies inhabiting semiarid regions, for example Mustela frenata longicauda of the Great Plains.

Another, but in my opinion less weighty, justification for applying Tschudi's name agilis to these light-colored weasels of western Perú is that by one line of reasoning, Taczanowski in naming macrura (jelskii is a synonym of it) from farther eastward in Perú, and that Hall in naming helleri from still farther eastward, and boliviensis to the southeastward, geographically restricted the application of the name agilis. Hall's action did this because he recognized geographic variation and employed the subspecies concept. Taczanowski, however, proposed his name macrura for a kind of animal which he indicated was specifically (as opposed to subspecifically) distinct from agilis and his account (1881:649) of jelskii indicates that he thought Mustela agilis Tschudi might occur in the same place as the animals which he named as new kinds. Thus, we can not credit Taczanowski with intent to restrict the name agilis geographically, even though later authors may choose to rule that his naming of macrura in effect did so restrict the application of Mustela agilis Tschudi.

The equivalents in millimeters given by Allen (1916:104) for Tschudi's measurements of 9 to 10 inches entire length, and tails of 4 inches to 4 inches and 4 lines, apparently are based on the London scale in use today. If Tschudi employed the Rhine scale also of eight lines to the inch, but one which has the foot longer by an amount of 20 millimeters, or the Leipzig scale in which the foot is 22 millimeters shorter than the London foot, the measurements recorded by Tschudi differ in one direction or the other from those computed by Allen. However, knowledge of which scale Tschudi employed would not help much, if any, in more precise application of the name agilis because he does not indicate whether his measurements are of male or female animals; animals of the two sexes of the same subspecies differ more in external measurements than animals of the same sex of different subspecies of Peruvian weasels.

Specimen no. 565, in the Polish Museum of Natural History, without definite locality, is provisionally referred to this subspecies. The specimen is intermediate in several respects between the female from Macate and the one of macrura from Cutervo.

Tschudi (1844:111-112) has given the following account: "Lebensweise und geographische Verbreitung. Das peruanische Wiesel lebt auf den kalten, öden Hochebenen der Cordillera an sonnigen Steinhaufen und Felsen gewöhnlich in Gesellschaft von 8-12 Stücken. Diese Thierchen sind so ausserordentlich behende und scheu, dass bei dem leisesten Geräusche die ganze Schaar mit Blitzesschnelle verschwindet. Es ist uns auch nie gelungen, eines derselben zu erlegen. Die Indianer aber verstehen es, dieselben lebendig einzufangen und zu zähmen. Ein sehr zahmes sahen wir bei einer uns befreundeten Dame in Tarma; gegen alle Fremden biss es mit Wuth und liess sich nicht anfassen, während es sich von seiner Herrin Alles gefallen liess; sie öffnete ihm den Mund und steckte ihm den Finger hinein, ohne dass es eine böse Miene dazu machte, während es bei der geringsten Bewegung, die wir machten, es zu ergreifen, grimmig auf uns lossprang. Wenn es eingeschüchtert wurde, versteckte es sich in den Busen seiner Gebieterin und kroch ihr bald nachher zum Aermel heraus. An den Wänden und Meublen kletterte es mit grosser Behendigkeit und schlüpfte durch so kleine Ritzen und Löcher, dass wir fast an der Möglichkeit dieses Hindurchdringens gezweifelt haben würden, wenn wir es nicht selbst mit angesehen hätten. Wenn es unartig war, wurde es mit einer Schnur an seinem kleinen Halsbande festgebunden; dadurch vermehrte sich sein Zorn, so dass es zuweilen gegen die Dame auffuhr. Mehrmals verschwand es während 8-10 Tagen und kam dann plötzlich wieder zum Vorschein. Seine Nahrung bestand in Gemüse und Fleisch, besonders aber liebte es Zuckerbrod in Milch aufgeweicht; einmal machte es sich an einen Kanarienvogel, den es auch tödtete. Es erhielt seine Strafe und verschwand dann für immer. Die Indianer sollen dieses Wiesel zum Fange der Viscacha abrichten (davon weiter unten). Sie nennen es Comadreja, auch Ardilla. ([footnote] Ardilla ist spanisch und heisst Eichhörnchen. Mit diesem Namen werden sehr verschiedene Thiere bezeichnet; ausser dem Sc. variabilis und der Galictis agilis auch noch mehrere Nager und einige Didelphysarten.) Die Indianerinnen verfertigen sich aus dem kleinen Felle Geldbeutel. Des Sonntags trifft man unter den vielen tausend Punaindianerinnen die nach den grossen Dörfern der Sierra kommen, um ihre Einkäufe zu machen, kaum ein halbes Dutzend, die nicht solche Börsen mit sich führten, und dann kann man auch die verschiedensten Farbennuancen, die bei dieser Species vorkommen, beobachten."

None of the three skulls referred to this subspecies shows infestation of the frontal sinuses by parasites.

Specimens examined.—Total number, 4.

Perú: Macate, 1 (Field Mus. Nat. Hist.); Huarochirí, 1 (Mus. Comp. Zool.); Lima, 1 (British Mus. Nat. Hist.); no locality more definite than Perú, 1 (Mus. Polonais d'Hist. Nat.).

Mustela frenata macrura Taczanowski

Long-tailed Weasel

Plates 1, 27, 28, 29, 30, 37, 38, 39 and 40

Mustela macrura Taczanowski, Proc. Zoöl. Soc. London, 1874:311, pl. 48, May 19, 1874; ibid., 1881:647, May 17, 1881; ibid., 835, November 15, 1881; Lönnberg, Arkiv för Zool., 8 (no. 1):21, 1913 (?); Hollister, Proc. Biol. Soc. Washington, 28:143, July 10, 1914; Allen, Bull. Amer. Mus. Nat. Hist., 35:101, April 28, 1916; Lönnberg, Arkiv för Zool., 14 (no. 4):11, 1921.

Putorius (Gale) braziliensis frenatus, Coues, Fur-bearing animals, p. 142, 1877.

Mustela jelskii Taczanowski, Proc. Zoöl. Soc. London, 1881:647, May 17, 1881.

Mustela affinis, Lönnberg, Arkiv för Zool., 8 (no. 1):21, July 12, 1913.

Mustela aureoventris, Thomas, Proc. U. S. Nat. Mus., 58:224, 1920.

Mustela frenata macrura, Hall, Carnegie Instit. Washington Publ. 473:110, November 20, 1936; Hall, Physis, 16:176, 1939.

Type.—Male, adult, mounted skin, with skull separate; no. 561, Mus. Polonais d'Hist. Nat. (Warsaw, Poland); Lake Junín, central Perú; 1873; obtained by M. Jelski.

The skull (plates 27-29, 30), mounted with the skin but removed by me for study, lacks the right jugal, the basisphenoid, the basioccipital and parts of each exoccipital bearing the exoccipital condyles. The right tympanic bulla, although detached from the skull, is preserved separately. The teeth all are present and entire. The skin is fairly-well mounted, in a good state of preservation, and shows no fading due to exposure to light.

Range.—Altitudinally, 3200 (Guainche) to at least 12000 feet (Pichincha); Upper Subtropical and Temperate life-zones of central Perú and Ecuador north from the states of Apurimac and Cuzco, Perú, to San Antonio, northern Ecuador. See figure 29 on page 221.

Characters for ready recognition.—Differs from Mustela africana stolzmanni by absence of median, longitudinal, abdominal stripe of same color as upper parts; presence of p2 and two roots rather than one root on P2; from Mustela frenata helleri, M. f. affinis and M. f. aureoventris by lighter color of upper parts which are Chocolate tone 3, pl. 343, Oberthür and Dauthenay, whereas, with reference to the same color standard, the colors are: in helleri, Carbon Brown, tone 3, pl. 342; in affinis, Reddish Black, tone 2, pl. 344; in aureoventris, Reddish Black, tone 4, pl. 344; from M. f. agilis by darker color (Chocolate, tone 3 rather than 2, pl. 343, Oberthür and Dauthenay) of upper parts, length of upper tooth-rows, in females, more than 13, inside length of P4 more than 4.6; from M. f. boliviensis by lighter color of upper parts which are as above rather than tone 4 of Carbon Brown, pl. 342 of Oberthür and Dauthenay, and larger size (in males, hind foot more than 45 and m1 more than 5.6).

Description.—Size.—Male (measurements as recorded by Taczanowski in the original description, for two specimens, type and topotype, with correction of the length of tail of his "female" [= male]): Total length, 420, 415; length of tail, 150, 145; length of hind foot, 51, 51. An adult from Yana Mayo, Río Tarma, was measured by Hendee as 394, 134. Hind foot relaxed measures, 47. Tail 55 per cent as long as head and body. Length of hind foot more than basal length.

Female (based on measurements given by Taczanowski (1881:647) of no. 564): Total length, 323; length of tail, 120; length of hind foot, 37. Tail 59 per cent as long as head and body. Length of hind foot approximately equal to basal length.

Differences in external measurements of the two sexes are: Total length, 87; length of tail, 23; length of hind foot, 13.

Externals.—Longest facial vibrissae extending beyond ear; carpal vibrissae color of either upper parts or underparts; hairiness of foot-soles as shown in figure 20.

Color.—(Based on specimens from Cutervo and south thereof). Rarely few white hairs between eyes and in front of ears; top of head posteriorly to slightly behind eyes, near (n) Chestnut-Brown (Ridgway) and Carbon Brown, tone 2 or darker (pl. 342, Oberthür and Dauthenay); posterior half of upper lip rarely white; dark spots at angles of mouth absent; tip of tail black; remainder of upper parts near (l) Russet (Ridgway) and Chocolate, tone 3 (pl. 343, Ober. and Dauth.); underparts white or whitish on medial sides of forelegs, otherwise cream color with tinge of Ochraceous-Buff; color of underparts extended distally on posterior sides of forelegs to just below elbow (in type specimen) or onto forefeet (in specimen from Yana Mayo) and on medial sides of hind legs to points between knees and ankles. Least width of color of underparts averages (in six skins) 17 (14-21) per cent of greatest width of color of upper parts. Black tip of tail longer than hind foot and averaging 36 (32-49) per cent of length of tail-vertebrae.

Skull and teeth.—Male (based on type specimen and no. 562): See measurements and plates 27-30; weight, not known; basilar length, 43.2 (40.8 and 45.5); zygomatic breadth more or less than distance between condylar foramen and M1 and more than that between anterior palatine foramen and anterior margin of tympanic bulla; mastoid breadth more or less than postpalatal length; postorbital breadth more than length of upper premolars and greater than width of basioccipital, measured from medial margin of one foramen lacerum posterior to its opposite; interorbital breadth more than distance between foramen opticum and anterior margin of tympanic bulla; breadth of rostrum less than length of tympanic bulla; least width of palate more than inside length of P4; anterior margin of tympanic bulla as far posterior to foramen ovale as width of 4 (including I3) upper incisors; height of tympanic bulla not more than distance from its anterior margin to foramen ovale; length of tympanic bulla more than length of lower molar and premolar tooth-row and longer or shorter than rostrum; anterior margin of masseteric fossa below or behind m2.

Female (based on no. 564, from Cutervo, Perú, type specimen of Mustela jelskii Taczanowski): See measurements and plates 37-40; weight, not known; basilar length, 38±; zygomatic breadth less than distance between condylar foramen and M1 and not greater than that between anterior palatine foramen and anterior margin of tympanic bulla; postorbital breadth more than alveolar length of upper premolars and (probably) more than width of basioccipital measured from medial margin of one foramen lacerum posterior to its opposite; least width of palate more than inside length of P4; tympanic bulla as far posterior to foramen ovale as width of at least 5-1/2 upper incisors; height of tympanic bulla less than distance from its anterior margin to foramen ovale; length of tympanic bulla more than length of lower molar and premolar tooth-row and longer or shorter than rostrum.

As compared with that of helleri, the skull of the male of macrura from Junín southward has a lesser mastoid breadth, notably smaller teeth, and a flatter skull which averages lighter throughout. The skulls of females available indicate that the skull and teeth are larger than in agilis.

Remarks.—Seven years after Taczanowski named this subspecies, he applied the name jelskii to a female taken farther north than the original examples of macrura. As indicated in synonymy, various other names have been applied to animals included by the present author in this subspecies.

Mustela frenata macrura intergrades with M. f. affinis as shown by practically all the referred specimens from north of Junín. As one proceeds northward the color of the weasels becomes progressively darker and the teeth become larger until the conditions found in affinis are met with near the northern border of Ecuador. From the material available it appears that the light-colored upper parts found in macrura characterize weasels of, at least, the Temperate Zone, from Marcapata, Perú, to near Quito, Ecuador. West of the range of macrura there exists the still lighter-colored subspecies, M. f. agilis. Immediately adjacent on the north, east, and south, darker-colored weasels occur. So far as color is concerned, the geographic range of the subspecies M. f. macrura is not difficult to define. However, the small size of the teeth characterizes only that part of this light-colored subspecies from Junín southward including the subspecies boliviensis at the southern extremity of the range of the species. From Cutervo northward the light-colored weasels of the Temperate Zone have teeth similar in size to those of the darker, more northern affinis. To designate the slightly larger-toothed, light-colored animals from Ecuador as a subspecies distinct from affinis and macrura is one solution but at present it seems best to refer all of these light-colored animals to macrura.

The type specimen and topotype no. 562 differ more in the amount of inflation of the tympanic bullae than adult males of comparable ages from a given locality usually do. In other respects, the differences between the two skulls are not greater than those ordinarily found in specimens from the same locality. No. 562 has the tympanic bullae greatly, relative to the other South American weasels, inflated posteriorly. Otherwise, the bullae agree with those of the type specimen.

Specimens from southwestern Ecuador, average large, and include the largest specimens of the species Mustela frenata seen from South America. A subadult male, no. 61406, in the American Museum of Natural History, is the largest. Its external measurements are 482, 191, 56. The basilar length of the skull is 48.2 and the zygomatic breadth is 30.3. Although not so large as this specimen, the corresponding measurements of specimens from Alamor, El Chiral, and even from as far away as Sigsig also are distinctly large.

The skull of the female from Ollantaytambo and that of the male from Marcapata have teeth equally as small as do the specimens from Lake Junín.

The skin alone, no. 194328, from Ollantaytambo has the color of the underparts extended over the entire upper sides of the forefeet. The male from Marcapata has less of this color on the forefeet and is in this respect intermediate between the specimens from Lake Junín and the one from Ollantaytambo.

In size of teeth the female, type specimen of M. jelskii, from Cutervo, shows an approach to the larger-toothed weasels of the northern part of the range of macrura.

The specimens in the Riksmuseum from the vicinity of Quito, Ecuador, have been rather fully described by Lönnberg (1921:11-17) and need little comment here, except to say that they show, as he suggested, that the weasel of the Temperate Zone of Ecuador is an intermediate link between M. f. macrura and M. f. affinis.

The adult female and juvenal male labeled as from Ambato have little left of the skulls except some of the teeth and the assignment of the specimens to the subspecies macrura is made mainly on geographic grounds. These two specimens probably are part of the shipment of birds and mammals of which Chapman (1926:703) speaks as follows: "A small collection of native-made skins purchased by the American Museum from a commission merchant in New York City as from 'Ambato' proved to be from the eastern slope of the Andes." Another skin in the same Museum, labeled by a native collector as from "Baeza arriba" [= above Baeza] is so dark colored and has the color of the underparts so much restricted, as to suggest that it belongs to the race aureoventris. Possibly, therefore, it was taken not at Baeza, Ecuador, which I find to the eastward of Quito at 77° 55' W and O° 25' S, but at some place of the same name on the Pacific Slope, unless the locality has been altogether wrongly recorded on the label. If the specimen was taken near the Baeza above referred to, then it gives evidence of an unnamed race of Mustela on the eastern slope of the Andes, characterized by its dark color. Unfortunately the specimen is young and its skull therefore offers insufficient basis for the judging of its subspecific relationships.

Other specimens, in the British Museum of Natural History, recorded as taken "near Quito" and here tentatively listed under macrura, mostly, include specimens so dark colored as to lead me to think they came from country, lower than Quito, adjacent to the range of aureoventris.

Nematodes taken from the right frontal sinus of no. 562 from Junín proved to belong to the superfamily Oxyuriodea according to Professor W. B. Herms and Mr. O. L. Williams, who have independently identified them. Because these worms had been dried fifty-five years in the mounted specimen and were later boiled in cleaning the skull, a more accurate determination was impossible and whether or not they pertain to the same species found in North American weasels cannot be said. Of 18 adult skulls examined for this type of infestation, 13 were found affected as judged by the evident malformation of the frontal region.

Specimens examined.—Total number, 74, arranged by localities from north to south and unless otherwise indicated in the American Museum of Natural History.

Ecuador: Ibarra, 6600 ft., and 7500 ft., 2[7]; San Antonio, 8000-8500 ft., 5 mi. N Quito, 4 (2[7], 2[78]); Nono, 10000 ft., 1; Mindo, 1[78]; Zambiza, 8000-8100 ft., NE Quito, 4 (2[78], 2[95]); Carapungo, 8500 ft., NE Quito, 1[78]; Panecillo, 10000 ft., near Quito, 2[78]; Guapulo, 8800 ft., 3 mi. E of Quito, 1[78]; Pichincha, 10500 ft., and 12000 ft., 2 (1[78], 1[95]); San Ignacio, 11500 ft., Pichincha, 1; Santa Rosa, 9600 ft., Río Pita, 2; near Santa Rosa, 9000 ft., 1; Río San Rafel, 9000 ft., 1; N side Quito, 9000 ft., 1[78]; Quito, 1[4]; near Quito, 5[7]; Nára Papallacta, 11000 ft., 1[78]; below Papallacta, 9000 ft., 1[78]; Chillo Valley, 1[78]; "Hacienda Hda," 10000 ft., Pintag, Valencia, 1; Baeza arriba, 1; Ambato, 2; San Francisco, 8000 ft., E of Ambato, 1; Chunchí, Pagma Forest, 6400 ft., 1[1]; Canar, 2600 M., 1[7]; Malletura, 7600 ft., 1; Contrayerbas, 11000 ft., 1; Sisig, 8500 ft., 3[7]; El Chiral, 1; Almor, 1; Guainche, 3200 ft., 1; no locality more definite than Ecuador, 4[95]; "Received from Quito," 1[7]; Quisaya, 6000 ft. (locality not found), 1[7]; La Carolina (locality not found), 1[78].

Perú: La Lejía, 1; Huancabamba, 4 (2[75]); Cutervo, 9000 ft., 1[73]; Condechacha, 7000 ft., Río Utcubamba, 1[7]; San Pedro, 8600-9400 ft., S of Chachapoyas, 1; Celendín, 1[7]; Junín, 2[73]; Yana Mayo, Río Tarma, 1[7]; Ollantaytambo, 9000 ft., 3 (1[7], 2[91]); Ocabamba, 1[7]; Anta Cuzco, 3400 and 3500 M., 2[4]; Marcapata, 1[91].

Mustela frenata boliviensis Hall

Long-tailed Weasel

Plates 28, 29 and 30

Mustela frenata boliviensis Hall, Proc. Biol. Soc. Washington, 51:67, March 18, 1938.

Mustela frenata macrura, Hall, Carnegie Instit. Washington Publ. 473:110, November 20, 1936; Hall, Physis, 16:176, 1939 (part).

Type.—Male, adult, skull and skin; no. 72587, Amer. Mus. Nat. Hist.; Nequejahuira, 8000 feet, Bolivia; May 19, 1926; obtained by G. H. H. Tate; original no. 4135 (see plates 28-30).

Range.—As now known 8000 to 9500 feet in the Andes from Limbaní, Perú, south to Nequejahuira, Bolivia; upper Subtropical and Temperate life-zones. See figure 29 on page 221.

Characters for ready recognition.—Differs from Mustela africana stolzmanni by absence of median, longitudinal, abdominal stripe of same color as upper parts; presence of p2 and two roots rather than one root on P2; from Mustela frenata macrura by darker color of upper parts (tone 4 or darker of Carbon Brown, pl. 342 rather than tone 3 of Chocolate, pl. 343, Oberthür and Dauthenay) and lesser size (in males hind foot less than 45 and m1 less than 5.6); from Mustela frenata agilis by darker color of upper parts (as given above rather than tone 2 of Chocolate, pl. 343, of Oberthür and Dauthenay).

Description.—Size.—Male: The type and two young specimens from Limbaní, Perú, measure respectively, as follows: Total length, 383, 368, 304; length of tail, 140, 132, 115; length of hind foot, 43, 44, 41. Tail 55 per cent as long as head and body. Length of hind foot approximately equal to basal length.

Female: Unknown.

Externals.—As described in Mustela frenata macrura.

Color.—Top of head blackish posteriorly to behind ears; upper lips same color as upper parts of head; dark area at angle of mouth not separated from upper parts as a distinct spot; tip of tail black; remainder of upper parts near (n) Mars Brown of Ridgway and tone 4 or darker of Carbon Brown (pl. 342, Oberthür and Dauthenay); underparts Cream-Colored with strong wash of Ochraceous-Buff; whitish on insides of forelegs to just below elbow; color of underparts extended distally on forelegs over ankles onto antipalmar faces of inner toes, and on hind legs to knees. Least width of color of underparts averages 15 (11-19) per cent of greatest width of color of upper parts. Black tip of tail in type longer than hind foot and amounting to 36 per cent of length of tail-vertebrae.

Skull and teeth.—Male (based on the type): See measurements and plates 28-30. As described in Mustela frenata macrura except that: Weight, 2.8 grams; basilar length, 41.6; zygomatic breadth less than distance between anterior palatine foramen and anterior margin of tympanic bulla; anterior margin of tympanic bulla as far posterior to foramen ovale as width of 5 upper incisors.

Female: Skull unknown.

Remarks.—Apparently the first specimens of this race to find their way into a zoölogical collection were the two young males taken on February 17, 1904, at Limbaní, by Geo. Ockenden [sic.].

M. f. boliviensis is smaller than any other South American weasel except possibly agilis. Better material of the two races probably will show even agilis to be larger.

Early in my study of Mustela after examination of the one young specimen, from Limbaní, in the United States National Museum, an account of this race was drawn up, but the account was discarded for want of satisfactory material and the animal was referred to macrura. Then, in 1937, when the two other specimens were studied, the race was formally characterized as different from previously recognized kinds.

The collector has noted on the labels of the two young from Limbaní that they were shot in the afternoon when running together beneath bushes. The frontal sinuses of the type are malformed as a result of infestation by parasites.

Specimens examined.—Total number, 3, as follows:

Perú: Carabaya, Limbaní, 2 (one in U. S. Nat. Mus. and one in Berlin Zool. Mus.).

Bolivia: Nequejahuira, 1 Amer. Mus. Nat. Hist.

Mustela frenata (?) gracilis (Brown)

Plates 39 and 40

Putorius gracilis Brown, Mem. Amer. Mus. Nat. Hist., 9(pt. 4):182, pl. 17, 1908.

Mustela gracilis, Hay, Iowa Geol. Surv. Bull., 23:32, 1914; Hay, Carnegie Instit. Washington Publ. 322A:252, October 15, 1924; Hay, Carnegie Instit. Washington Publ. 390(vol. 2):528, 1930.

Mustela frenata gracilis, Hall, Carnegie Instit. Washington Publ. 473:112, November 20, 1936.

Type.—Adult skull without lower jaws, probably of a female, no. 12431, Amer. Mus. Nat. Hist.; from Conard Fissure, four miles west of Willcockson, Newton County, Arkansas; obtained sometime in the period 1903 to 1905 inclusive. (See plates 39 and 40.)

Range.—Known only from the Pleistocene deposit in Conard Fissure, at the type locality in northern Arkansas.

Description.—Skull. Probably female (based on the type): See measurements and plates 39 and 40; weight unknown; basilar length, 38.1; least width of palate less than greatest length of P4; tympanic bulla as far posterior to foramen ovale as width of 3 to 5 upper incisors; height of tympanic bulla less than distance from its anterior margin to foramen ovale; length of tympanic bulla less than length of rostrum.