SYSTEMATIC ACCOUNTS OF SPECIES AND SUBSPECIES (continued 5)

Skull and teeth.—Male (based on ten adults from Douglas County, Kansas): See measurements and plates 16-18; weight, 3.7 (3.3-4.2) grams; basilar length, 44.8 (43.8-46.0); zygomatic breadth more or less (less in 80 per cent) than distance between condylar foramen and M1 or than between anterior palatine foramen and anterior end of tympanic bulla (less in 70 per cent); mastoid breadth more or less (less in 80 per cent) than postpalatal length; postorbital breadth less than length of upper premolars and, except in one specimen, more than width of basioccipital measured from medial margin of one foramen lacerum posterior to its opposite; interorbital breadth more or less (less in 70 per cent) than distance between foramen opticum and anterior margin of tympanic bulla; breadth of rostrum less than length of tympanic bulla; least width of palate less than outside length of P4 (except in one specimen); anterior margin of tympanic bulla as far posterior to foramen ovale as width of 2-1/2 to 5 upper incisors; height of tympanic bulla more (except in one specimen) 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 (except in one specimen) than rostrum; anterior margin of masseteric fossa behind or just below posterior eighth of m1.

Female (based on 5 adults and subadults from Douglas County, Kansas): See measurements and plates 31-33; weight, 2.2 (2.0-2.4) grams; basilar length, 38.9 (37.6-40.7); zygomatic breadth less than distance between condylar foramen and M1 or than between anterior palatine foramen and anterior margin of tympanic bulla; postorbital breadth less than length of upper premolars and except in one specimen, more than width of basioccipital measured from medial margin of one foramen lacerum posterior to its opposite; 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 more or 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 than rostrum.

The skull of the female averages 41 per cent lighter than that of the male.

Compared with the skull of M. f. noveboracensis from Massachusetts, that of the male of primulina, in dorsal view, is seen to be shorter anteriorly to the postorbital processes and to have a more marked postorbital constriction. In lateral view the dorsal outline of the skull of primulina is less concave in the postorbital region. In ventral view the skull of primulina is seen to be wider across the mastoid processes and zygomatic arches but the most pronounced difference is in the tympanic bullae. In noveboracensis each bulla is scooped out on the anterior part of the medial face and appears to be narrower anteriorly than posteriorly whereas in primulina the anterior part of the medial face is not scooped out but is moderately inflated and the bulla appears to be of uniform breadth anteriorly and posteriorly. By actual measurement the breadth of the bulla averages 59 per cent of its length in primulina but only 50 per cent in noveboracensis. Other respects in which the skull of the male of primulina differs from that of noveboracensis are as follows: Linear measurements of teeth more; relative to the basilar length, the length of the tooth-rows averages more, whereas the interorbital breadth and orbitonasal length are less.

When skulls of females are compared, each of the differences mentioned above is found to apply, except that the degree of difference is in some parts greater, for example, in the tympanic bullae. In primulina, the bulla is in general like that of the male noveboracensis, whereas in the female noveboracensis it is less inflated, especially anteromedially, shorter, relatively narrower, and in ventral view projects little or none below the squamosal floor of the braincase. The breadth of the bulla averages 51 per cent of its length in primulina but only 47 per cent in noveboracensis. The bullae project below the basioccipital on the average, for a distance of 2.9 millimeters in female primulina and only 2.3 millimeters in female noveboracensis. In primulina the temporal ridges are well developed and fuse to form a low sagittal crest, but in noveboracensis the ridges are absent. Also, in primulina the mastoid processes project farther laterally beyond the braincase. The skull of female noveboracensis is much lighter than that of primulina. Average weights of the two are 1.7 and 2.2 grams. The skulls of females of primulina and noveboracensis differ more than do the skulls of males.

Compared with the skull of spadix, that of the male, and the female, of primulina averages smaller in every part measured. Expressed in percentages of the basilar length, the two depth measurements of the skulls are not significantly different, but, excluding the measurements of the bullae and teeth, the other cranial measurements are less. The main difference in relative proportions is in the tympanic bullae which average only a half millimeter shorter in males of primulina and one and one-tenth millimeters shorter in females. The bullae are, therefore, relative to the basilar length, longer in primulina. The skull of primulina, then, differs from that of spadix mainly in smaller size and relatively longer tympanic bullae, especially in males.

Compared with the skull of M. f. longicauda, that of both sexes of primulina averages smaller in every part measured, except in males where the length of the tympanic bulla, and breadth and length of M1 are the same or slightly larger in primulina. Relative to the basilar length, the length of the tympanic bullae, and in females only, the depth measurements are greater in primulina but all the others, in both sexes, are less. These ratios reflect the relative narrowness of the skull of primulina. Upon direct comparison the narrowness is especially noticeable in the interorbital region, mastoid region, tympanic bullae, and across the zygomata.

Compared with the skull of M. f. neomexicana that of both sexes of primulina averages smaller in every part measured. Excepting the measurements of the teeth, most of the other measurements are constantly larger. Relative to the basilar length, the length of tooth-rows and length of tympanic bulla are more, but excepting the depth measurements, the others are less. Still other differences are, in primulina, less well-developed sagittal crest, anterolateral corner of bulla rounded rather than "square," and in males a transversely convex rather than flat interorbital region.

Compared with M. f. frenata and M. f. texensis, the skulls of males of primulina differ in being smaller in every part measured but relative to the basilar length, have longer tooth-rows, a lesser zygomatic breadth and are less constricted interorbitally.

Compared with the skull of M. f. olivacea, those of both sexes of primulina average smaller in every part measured, have shallower (dorsoventrally) tympanic bullae, a lower sagittal crest and slightly weaker postorbital processes on the frontals. Relative to the basilar length, the several cranial measurements are about the same.

Comparison of the skull with that of M. f. arthuri has been made in the account of that subspecies.

Remarks.—The first specimens of this race known to have been preserved in study collections are one in the United States National Museum, taken at Bridge, Carroll County, Missouri, many years ago by J. Burbage, and less than a dozen specimens preserved before 1900 from eastern Kansas in the University of Kansas Museum of Natural History. In 1913 Hartley H. T. Jackson bestowed a name on this animal on the basis of two specimens taken by him in southwestern Missouri. Later, through the efforts of Charles D. Bunker, and his associates at the University of Kansas, nearly 100 specimens were saved from eastern Kansas, principally from Douglas County. In the course of the present study, Lawrence V. Compton obtained a topotype for the California Museum of Vertebrate Zoölogy, and with the assistance of Mr. B. G. Roberts, a good series of specimens from Boone County, Arkansas, was preserved in the same museum. In the early years of the 20th Century, the late B. H. Bailey at Coe College, Iowa, collected specimens from that state. The specimens from these several sources suffice to give a relatively clear idea of the characters of this subspecies.

Mustela frenata primulina is closely related to M. f. noveboracensis, from which, on the average, it differs in the lighter color of the upper parts of the summer coat, in the more intense coloring of the underparts, and in certain cranial features pointed out above. In the southern part of its range, however, noveboracensis has the underparts only a little less intensely colored with yellow than primulina. Also, the skull of the one topotype from 7-1/2 miles southeast of Carthage, a subadult male in brown, winter pelage, is almost exactly intermediate between that of noveboracensis from Massachusetts and primulina of Douglas County, Kansas, and Boone County, Arkansas. M. f. primulina often has the underparts white in winter, as does this topotype which agrees with the average of noveboracensis in small size of teeth and narrowness across the mastoid processes and zygomatic arches. However, it agrees with primulina in shape and relative size of the rostrum. It is almost exactly intermediate in shape and width of the tympanic bullae.

Three other males, but no females, all in winter pelage, are available from eastern Missouri. Of the two from Silex, Lincoln County, one is nearer noveboracensis and the other nearer primulina on the basis of cranial characters. The third specimen, from four miles south of Lesterville, so far as I can determine by examination of individual cranial characters and tabulation of results, is exactly intermediate. Final decision on the proper allocation of specimens from the parts of Missouri represented can best be made when skulls of females are available. From the fact that the skull of the female referred to noveboracensis from Golconda, Illinois, shows almost as many characters of primulina as of noveboracensis, it is judged that females from as far west as Silex and Lesterville, Missouri, will show even more characters of primulina and so be referable to that form. If this supposition be correct, the present reference of the almost exactly intermediate males, from eastern Missouri, will stand; otherwise, it may not.

Additional intergrades with noveboracensis are available from eastern Iowa. Of five specimens from Hillsboro, Iowa, two males and a female have tympanic bullae like those of primulina but the other two males have bullae like those of noveboracensis. The female is smaller than primulina and in this small size and in general configuration of the skull, viewed dorsally, is more nearly like noveboracensis. As a whole, the population averages almost exactly intermediate. The same is true of 3 males and one female from Muscatine. The subadult male from Keosaqua, to my eye, resembles noveboracensis in the greater length of the skull anteriorly to the postorbital processes, and in the relative narrowness across the mastoidal region, but otherwise is more like that of primulina. Two males and one female from Tipton, although in each instance variously intermediate, are as a whole nearer primulina, No. 2865, Coe College, male adult, from Cedar Rapids, has characters of the three races, spadix, noveboracensis and primulina. In the skull, the width suggests spadix, the narrow mastoid region, noveboracensis, and the tympanic bullae are as in spadix or primulina. One male, no. 12, Coe College, from Dubuque, is as narrow across the mastoid region as is noveboracensis although the bullae are well inflated as in primulina. The skull, without corresponding skin, of a female, no. 140a, Iowa State College, from Green's Island, also resembles noveboracensis in narrowness of the mastoidal region, and in small size of skull, but in larger teeth, broader tympanic bullae, and sagittal crest is referable to primulina. Of two females from Vinton, one adult is typical of primulina but the other, a subadult, is practically indistinguishable from female noveboracensis, from Ann Arbor, Michigan. Three males from Vinton agree well with primulina except that the interorbital region is wider than average and thereby suggests spadix or noveboracensis. An adult female from New Hartford also is typical of primulina except for the broader interorbital region. Three males from Fayette are typical of primulina.

Other specimens from Iowa are intergrades with spadix, or if not with spadix, with the animal of northwestern Iowa which in some ways combines the characters of longicauda and spadix. For example, no. 2665, Coe College, an adult male from Davenport, has the anterior part of the skull (all that is preserved) heavily ridged as in spadix and in addition, the underparts are marked with the shade of reddish displayed by topotypes of spadix and with some yellowish as seen in longicauda. The color pattern, however, is as in primulina. A young male, no. C-51, Iowa State College, from Kelley, Story County, has anteriorly truncate bullae as are more frequently found in the longicauda-spadix stock of northwestern Iowa, than in primulina. In other respects, the animal, in so far as can be judged from the broken skull, agrees with primulina as it certainly does in color, color pattern, and external measurements. An adult male, no. 499a, Iowa State College, from 2 miles east of Ledges St. Park, in Boone County, in short body, size of teeth, and size of skull, in so far as the broken parts can be measured, resembles primulina more closely than it does any other subspecies. The long tail, long hind foot, wide extent of the light-colored underparts, and extension of the color of the underparts onto the hind feet are more as in spadix. Other intergrades with spadix from Iowa are mentioned in the account of spadix.

The specimen from Swartz, Louisiana, suggests intergradation with arthuri in that the anteromedial part of the tympanic bulla is less inflated than in typical primulina.

Intergrades with longicauda are available from Riley and Pratt counties, Kansas. No. 7182, Univ. Kans., subadult male in winter pelage, from near Winkler, has a skull of larger size as in longicauda with which race it seems to agree in large size of body, tail and hind foot, although the collector's measurements are lacking. Color pattern and relative proportions of the skull throughout are as in primulina. The young male, no. 3495, Univ. Kans., from Pratt, Kansas, agrees in external measurements and large size of skull with longicauda, but has the color and color pattern precisely as in primulina. The teeth are smaller as in primulina. Immaturity prevents judging of its relationships on the basis of relative proportions of the skull.

The two specimens, skins only, available from Oklahoma, are provisionally referred to primulina. These are remarkable for the restriction of the color of the underparts and for the intensity of the yellow coloration of the underparts. The specimen from Norman has the color of the underparts entirely absent from the hind legs and not extending posteriorly to the penis. On the chest and lower throat, large spots of color of the upper parts are present and the yellow area of the underparts on the belly is narrower than in any other specimen of primulina examined. The specimen from 8 miles northwest of Stillwater has the color of the underparts only a little less restricted although this color does extend over the inguinal region almost to the knees. The skin of the posterior part of the body of a weasel is available from 10 miles south of Sulphur Springs, Texas. It, likewise, is only provisionally referred to primulina. The coloration is about as in the specimens from Oklahoma but the distribution of the color of the underparts cannot be made out.

The dark color of the upper parts occurs far westward in animals which otherwise display characters of longicauda. Among these intergrades, the larger size of longicauda generally is combined with this dark color. This geographic behavior of the dark color of the upper parts is analogous to the condition described in M. f. spadix. Stated in another way, the dark color of the upper parts is the character, of the eastern animal, last to disappear as one goes westward across the Mississippi Valley toward the range of longicauda which is a subspecies of markedly different size, shape of skull, and coloration.

Only two of 29 specimens from Kansas show infestation of the frontal sinuses. All four of the specimens from Missouri have the frontal sinuses malformed as do 9 of the 14 from Arkansas examined in this respect.

An adult female from Boone County, Iowa, bears the date May 9, 1938, and the annotation by T. G. Scott, "Fox-killed."

Specimens examined.—Total number, 131, arranged alphabetically by states and from north to south by counties in each state. Except as otherwise indicated, specimens are in the University of Kansas, Museum of Natural History.

Arkansas. Boone County: 3 mi. E Bergman, 4[74]; 3 mi. SE Bergman, 1[74]; 3 mi. S Bergman, 1[74]; 3 mi. SW Bergman, 1[74]; 4 mi. SE Bergman, 2[74]; 5 mi. SE Bergman, 1[74]; 4-1/2 mi. SE Bergman, 3[74]; 5 mi. SE Bergman, 1[74]; 5 mi. S Bergman, 2[74]; 5 mi. SW Bergman, 2[74]. Washington County: Fayetteville, 1[96]. Crawford County: 10 mi. S Winslow, 1. Sebastian County: Fort Smith, 1[91].

Iowa. Fayette County: Fayette, 3[12]. Dubuque County: Dubuque, 1[12]; Green's Island, 1[65]. Butler County: New Hartford, 1[12]. Hardin County: Union, 1[65]. Benton County: Vinton, 5[12]. Linn County: Cedar Rapids, 1[12]. Boone County: Worth Township, Sec. 21, 1[65]; 2 mi. E Ledges St. Park, 1[65]. Story County: Kelley, 1[65]. Cedar County: Tipton, 3[12]. Scott County: Davenport, 2[12]. Muscatine County: Muscatine, 4[12]. Henry County: Hillsboro, 5[91]. Van Buren County: Keosaqua, 1[65]; no locality more definite than county, 1[50]. Taylor County, 1.

Kansas. Riley County: near Winkler, 1. Pottawatomie County: Onaga, 1[83]. Atchison County: Doniphan Lake, 1; 5 mi. NE Muscotah, 1; no locality more definite than county, 1. Douglas County: Lawrence, 8; 6 mi. NW Lawrence, 1; 1-1/2 mi. W Lawrence, 1; 6 mi. S Lawrence, 1; 7 to 7-1/2 mi. SW Lawrence, 14; 10 mi. W Lawrence, 1; Clinton, 4; Baldwin, 1; no locality more than county, 29 (2[74]). Woodson County: 1-1/2 mi. S Neosho Falls, 1[59]. Greenwood County: 8 mi. SW Toronto, 2. Pratt County: Pratt, 1.

Louisiana. Quachita Parish: Swartz, 1[71].

Missouri. Carroll County: Bridge Creek, 1[91]. Lincoln County: Silex, 1[74]; 1 mi. E Silex, 1[74]. Reynolds County: 4 mi. S Lesterville, 1[74]. Jasper County: 5 mi. NE Avilla, 1[91]; 7-1/2 mi. SE Carthage, 1[74].

Oklahoma. Payne County: 8 mi. NW Stillwater, 1[82]. Cleveland County: Norman, 1[100].

Texas. Hopkins County: 10 mi. S Sulphur Springs, 1[43].

Mustela frenata arthuri Hall

Long-tailed Weasel

Plates 16, 17 and 18

Mustela noveboracensis arthuri Hall, Proc. Biol. Soc. Washington, 40:193, December 2, 1927.

Mustela frenata arthuri Hall, Carnegie Instit. Washington Publ. 473:105, November 20, 1936.

Type.—Male, subadult, skull and skin; no. 37515, Mus. Vert. Zoöl.; Remy, St. James Parish, Louisiana; December 15, 1926; obtained by Stanley C. Arthur.

The skin is stuffed and well preserved. The skull (plates 16-18) is unbroken. The teeth all are present and entire. The presence of a well-developed scrotal pouch shows the specimen to be a male. Contrary to what was stated in the original description the specimen was taken in 1926 and not in 1925.

Range.—Lower Austral Life-zone of southeastern Texas, Louisiana, and into Mississippi. See figure 29 on page 221.

Characters for ready recognition (of males).—Differs from M. f. olivacea in smaller size (adult males with hind foot and basilar length less than 45), depth of skull at anterior margin of basioccipital, ignoring sagittal crest, amounting to more than 63 per cent of mastoid breadth, and greater convexity of dorsal outline of skull in longitudinal axis (see pls. 16-18); from M. f. noveboracensis, in males, by zygomatic breadth not less than distance between anterior palatine foramen and anterior margin of tympanic bulla and by convex dorsal outline of skull in longitudinal axis; from M. f. primulina by evenly spreading zygomatic arches, lesser inflation of tympanic bullae anteromedially than posteromedially, and convex dorsal outline of skull in longitudinal axis; from M. f. texensis and M. f. frenata by absence of white facial markings and postorbital breadth more than distance between posterior borders of P4 and P2.

Description.—Size.—Male: The type, a subadult male, measures (inches and quarter fractions thereof, transposed into millimeters) as follows: Total length, 390; length of tail, 113; length of hind foot, 44. Tail is 41 per cent as long as head and body. Length of hind foot less than basal length.

Typical female unknown.

Externals.—Longest facial vibrissae black, or dark brown (both colors in the type) and extending beyond ear; carpal vibrissae same color as underparts and extending to within 3.5 millimeters of apical pad of fifth digit. Hairiness of foot-soles in type slightly less than shown in figure 20.

Color.—Upper parts in summer tone 4 of Burnt Umber of Oberthür and Dauthenay, pl. 304; underparts as described in M. f. olivacea. In winter, upper parts (based on type) near (1) Brussels Brown or grayer than tone 4 of Burnt Umber of Oberthür and Dauthenay, pl. 304, darker on top of head from nose to a line connecting posterior margins of ears. Chin and posterior third of each upper lip white. Remainder of underparts white with wash of Warm Buff. Tip of tail black. Color of underparts extends distally on posterior sides of forelegs over toes but represented on antipalmar faces of feet by only a few scattered hairs. Color of underparts extends distally on medial sides of hind limbs only to knees. Least width of color of underparts amounting to 15 per cent of greatest width of color of upper parts. Black tip of tail 50 mm. long; thus longer than hind foot and 44 per cent as long as tail-vertebrae.

Skull and teeth.—Male (based on type and 2 subadults): See measurements and plates 16-18. As described in M. f. noveboracensis except that: Weight, 4.0 (3.7-4.3) grams; basilar length, 43.5 (43.3-43.6); zygomatic breadth not less than distance between anterior palatine foramen and anterior margin of tympanic bulla; postorbital breadth more than length of upper premolars; interorbital breadth more than distance between foramen opticum and anterior margin of tympanic bulla; least width of palate more or less than length of P4; tympanic bulla longer than rostrum.

Female: Typical skull unknown. The skull from 12 miles east of Eagle Lake, Texas, lacks the convexity in the dorsal longitudinal axis and the skull agrees with those of larger individuals of primulina except that the anteromedial faces of the tympanic bullae are less inflated, and the mastoid and zygomatic breadths are greater than in any female seen of primulina. Probably this greater breadth is the result of intergradation with M. f. frenata to the westward.

Compared with the skull of M. f. olivacea that of arthuri differs as follows: Averaging smaller in every part measured; basilar length 5 mm. less; by weight a fourth lighter; relative to basilar length, interorbital breadth greater and zygomatic and especially mastoid breadth less; dorsal outline of skull more convex in longitudinal axis; tympanic bullae narrower and less inflated especially on anteromedial faces. Compared with the skull of noveboracensis that of arthuri has the zygomatic breadth equal to or exceeding the distance from the anterior palatine foramen to the anterior margin of the tympanic bulla, whereas the zygomatic breadth is less than this distance in noveboracensis. Also, in arthuri, the rostrum is relatively shorter, the braincase is more inflated anteriorly, the zygomatic arches are more uniformly spreading, and the dorsal outline of the skull is distinctly convex, both transversely and longitudinally, whereas it is transversely more nearly flat in noveboracensis and longitudinally is concave in the interorbital region.

Compared with M. f. primulina, arthuri has narrower bullae, which are much less inflated on their anteromedial faces, a less marked postorbital constriction, a braincase which is narrower across the mastoid region and broader anteriorly, and a skull, which, in longitudinal axis, has the dorsal outline markedly more convex.

Compared with the skull of M. f. texensis that of arthuri is smaller in every part measured; length one-fifth less; one-half as heavy; postorbital constriction less marked; braincase relatively narrower posteriorly and tympanic bullae less inflated especially anteromedially. Compared with the skull of M. f. frenata that of arthuri is smaller in every part measured; basilar length 6 mm. less; a third lighter; postorbital constriction less marked; relative to the basilar length the rostrum is broader, longer and deeper; the zygomatic expanse and breadth of the braincase across the mastoids is less; the dorsal profile of the skull is more convex in longitudinal axis; zygomata evenly spreading rather than abruptly protruding from skull posteriorly; tympanic bullae less inflated anteromedially.

Remarks.—In 1926, Stanley C. Arthur, then Director of the Division of Wild Life, for the Louisiana State Department of Conservation, obtained specimens of this weasel. Some of them were mounted and the remainder were placed in the collections of the United States National Museum and the Museum of Vertebrate Zoölogy. In 1938 to 1940 George H. Lowery saved specimens from Baton Rouge, which showed the color of the summer pelage and revealed that the size of males was more than was indicated by the original materials. In 1940 and 1941 Rollin H. Baker obtained specimens from eastern Texas which greatly extended the known geographic range.

In addition to the localities represented by specimens examined, Arthur (1928:117) has recorded specimens from Greensburg, St. Helena Parish; Braithwaite, Plaquemines Parish; Geismar, Assumption Parish; Laurel Hill, West Feliciana Parish; French Settlement, Livingston Parish; and Kentwood, Tangipahoa Parish. All these localities lie within the eastern half of southern Louisiana. A skin-only, no. 38902, Mus. Vert. Zoöl., obtained from a fur buyer by Stanley C. Arthur, was taken in Mississippi "south of Jackson." Possibly it is of the subspecies arthuri.

Intergradation with M. f. olivacea is indicated by a specimen from Mobile County, Alabama, commented on in the account of olivacea. Intergradation with primulina is indicated by the shape of the anteromedial part of the bullae of the specimen from Swartz, Louisiana, that is referred to primulina. The lack of specimens from the northern two-thirds of Mississippi and from western Tennessee, prevents any definite statement as to the limits of range of arthuri in those areas. In comparison with animals from the type locality, the slightly larger size of the adult male from Baton Rouge, and the still larger size of the adult male of primulina from Swartz, Louisiana, suggests that the olivacea "influence" may extend farther west in the latitude of northern Louisiana than anywhere else.

None of the skulls examined shows malformation of the frontal sinuses such as results from infestation by parasites in some races. Arthur (1928:115) speaks of the ". . . cut-over swamp land, where the tupelo and cypress have been removed, . . ." as constituting suitable habitat for this animal.

Specimens examined.—Total number, 13, as follows:

Texas. Colorado County: 12 mi. N Eagle Lake, 1[43]; 5 mi. W Eagle Lake, 1[43]; 3 mi. S Garwood, 1[43].

Louisiana. East Baton Rouge Parish: Baton Rouge, 4[71]. Livingston Parish: Springville, 1[74]. Saint James Parish: Convent, 1[91]; Remy, 2 (1[74], 1[45]). Assumption Parish: near Lake Verret, 1[45].

Mississippi. Harrison County: Saucier, 1[71].

Mustela frenata olivacea Howell

Long-tailed Weasel

Plates 16, 17, 18, 31, 32 and 33

Mustela peninsulae olivacea Howell, Proc. Biol. Soc. Washington, 26:139, May 21, 1913.

Mustela frenata olivacea, Hall, Carnegie Instit. Washington Publ. 473:104, November 20, 1936.

Type.—Male, adult, skull and skin; no. 180802, U. S. Nat. Mus., Biol. Surveys Coll.; Autaugaville, Autauga County, Alabama; December 22, 1912; obtained by L. S. Golsan.

The skull (plates 16-18), although cracked at two places in the interorbital region, is in one piece and not warped out of shape. The teeth all are present and entire. The skin is exceptionally well made and in perfect condition except for the extreme tip of the tail which is broken off.

Range.—Lower and Upper Austral life-zones in eastern Mississippi, Alabama, Georgia, South Carolina, and northern Florida. See figure 29 on page 221.

Characters for ready recognition.—Differs from M. f. peninsulae in finer, softer pelage and shorter (less than 15.8 in ad. ♀) tympanic bullae; from M. f. noveboracensis, in adult males by wider tympanic bulla which is more than rather than less than 8.5, in adult females by total length which is more than rather than less than 345, and by mastoid breadth which is more than rather than less than distance between articular faces of exoccipital condyle and glenoid fossa; from M. f. arthuri in larger size (adult males with hind foot and basilar length each more than 45); depth of skull at anterior margin of basioccipital, ignoring sagittal crest, amounting to less than 63 per cent of mastoid breadth, and lesser convexity of dorsal outline of skull in longitudinal axis (See pls. 16-18).

Description.—Size.—Male and female: External measurements of adults are available as follows:

Catalogue No. Sex Locality Total length Length of tail Length of hind foot

47165 ♂ Box Springs, Talbot Co., Georgia 454 160 48

47166 ♂ Box Springs, Talbot Co., Georgia 435 147 47

47167 ♂ Box Springs, Talbot Co., Georgia 422 145 45

41023 ♂ Thomas Co., Georgia 443 140 47

41025 ♂ Grady Co., Georgia 395 142 47

223880 ♂ Okefinokee Swamp, Georgia 416 145 49

198 ♂ Okefinokee Swamp, Georgia 425 140 48

Average 7 ♂ 427 146 47

49385 ♀ Gainesville, Alachua Co., Florida 396 124 45 [not typical]

41024 ♀ Thomas Co., Georgia 380 125 41

51527 ♀ Talbot Co., Georgia 376 128 43

The length of the hind foot averages less than the basal length in both males and females. The tail averages 52 per cent as long as the head and body in males and 51 per cent in females. Average differences in measurements of the two sexes are: Total length, 49; length of tail, 19; length of hind foot, 5. An adult male, no. 41023, and an adult female, no. 41024, each taken in February, 1929, on the Sinkola Plantation, Thomas County, Georgia, weighed 15 ounces (425 grams) and 7 ounces (198 grams) respectively according to Charles O. Handley.

Externals.—As described in Mustela frenata noveboracensis, except that hairiness of foot-soles slightly less than shown in figure 19.

Color.—Upper parts, in summer, near tone 4 of Burnt Umber of Oberthür and Dauthenay, pl. 304. In winter lighter, between tones 3 and 4 of Raw Umber of Oberthür and Dauthenay, pl. 301. Dark spot at each angle of mouth present or absent. Underparts ranging from Massicot Yellow to Cream Buff except on chin and upper lips which are white. Tip of tail black. Upper parts of uniform color. Color of underparts extends distally on posterior sides of forelegs over antipalmar faces of toes and on medial sides of hind limbs to ankles. Least width of color of underparts averaging, in a series of five males from Talbot Co., Georgia, 29 (extremes 24-34) per cent of greatest width of color of upper parts. Black tip of tail in same series, averaging 65 (extremes 60-70) mm. long, thus longer than hind foot and averaging 43 per cent of length of tail-vertebrae.

The spot at the angle of the mouth is absent in one-third of the specimens examined. The upper lips are white in specimens from the southern part of the range of olivacea but in the northern part of the range of the subspecies the upper lips are dark colored as in noveboracensis.

Skull and teeth.—Male (based on 5 adults from Talbot Co., Georgia): See measurements and plates 16-18; weight, 5.3 (5.0-6.4) grams; basilar length, 48.3 (45.8-50.1); zygomatic breadth more or less (usually less) than distance between condylar foramen and M1 and more or less (usually more) than distance between anterior palatine foramen and anterior margin of tympanic bulla; mastoid breadth more or less than (averaging about equal to) postpalatal length; postorbital breadth less than length of upper premolars and more or less than (about equal to) width of basioccipital measured from medial margin of one foramen lacerum posterior to its opposite; interorbital breadth more or less than (about equal to) distance between foramen opticum and anterior margin of tympanic bulla; breadth of rostrum less than length of tympanic bulla; least width of palate less than greatest length of P4; anterior margin of tympanic bulla as far posterior to foramen ovale as width of 3 to 5 upper incisors; height of tympanic bulla not 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 than rostrum (one exception); anterior margin of masseteric fossa below posterior half of m2.

Female (based on 2 adults from Thomas Co., Ga., and one from Talbot Co., Ga.): See measurements and plates 31-33; weight, 3.8 (3.5-4.0) grams; basilar length, 43.4 (42.7-44.0); zygomatic breadth less than distance between condylar foramen and M1 or than between anterior palatine foramen and anterior margin of tympanic bulla; postorbital breadth less than length of upper premolars and more or less (usually more) than width of basioccipital measured from medial margin of one foramen lacerum posterior to its opposite; least width of palate less than greatest length of P4; tympanic bulla as far posterior to foramen ovale as width of 3 to 4 (including I3) upper incisors; height of tympanic bulla not less (usually 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 than rostrum.

The skull of the female averages 28 per cent lighter than that of the male.

Compared with the skull of M. f. peninsulae, of which only one good skull, and that a female, is available, that of M. f. olivacea averages smaller and has relatively and actually smaller and less inflated bullae. As compared with the skull of M. f. noveboracensis, that of olivacea in the case of males is larger in every part measured and relative to the basilar length is broader across the zygomatic arches and mastoids. However, the rostrum and interorbital region are relatively narrower. The orbitonasal length is relatively less. The tympanic bullae are broader and more inflated. The same differences hold as between females of noveboracensis and olivacea. Indeed, the females of these two races differ more than do the males. Additional, selected differential cranial characters in the females are, in olivacea, as follows: Weight averaging 3.8 grams rather than 1.7 grams; braincase with, rather than without, sagittal crest; anterior border of tympanic bulla separated from foramen ovale by breadth of less than, rather than breadth of more than, 4 upper incisors (including I3); height of tympanic bulla not less than, rather than less than, distance from its anterior margin to foramen ovale; squamosal bone, between anterior margin of tympanic bulla and foramen ovale, ventrally concave rather than ventrally convex. Comparisons of the skulls with those of M. f. arthuri and M. f. primulina are made in the accounts of those subspecies.

Remarks.—Excepting two young specimens from South Carolina in the Charleston Museum, no specimens of this race of large weasel seem to have been preserved until Arthur H. Howell, in the course of his study of the mammals of Alabama, procured specimens on which his name, olivacea, was based. Later, Francis Harper obtained three instructive specimens from Okefinokee Swamp. Really adequate material, for the localities represented, owes its preservation to the alertness of Charles O. Handley, when he resided at Thomasville, Georgia, and to Hallie E. Fuller of Geneva, Talbot Co., Georgia.

The distinctness of M. f. olivacea from M. f. peninsulae is not satisfactorily established due to inadequate material of peninsulae. Differences shown by the specimens seen indicate that, as compared with olivacea, peninsulae is larger, has transversely wider light-colored underparts which possess more yellow, and a larger skull with more inflated tympanic bullae. In each of these characters, olivacea is intermediate between noveboracensis on the north and peninsulae on the south. The question arises, therefore, whether the animals here recognized under the name olivacea really constitute a recognizable subspecies or instead are only representatives of a subspecies which reaches its extreme development in Florida. In the latter event, the name peninsulae would apply to all. Examination of more material from Florida, especially from the southern half of Florida, will be necessary to answer this question.

This large weasel of the southeastern United States is remarkably different from noveboracensis. Indeed, were it not for actual intergrades such as the two from Fort Payne, Alabama, and York, South Carolina, which are described in the account of M. f. noveboracensis, and the six specimens from northwestern Alabama, which are referred to olivacea, the systematist, I believe, would have little or no hesitancy in designating the two as distinct species, especially on the basis of differences to be seen in the skull.

Not only are the two forms structurally more different than usually is the case but between two geographically, adjacent subspecies of the same species of mammal, but the belt where intergradation occurs appears to be narrow. Nevertheless, when material of the two races is laid out in geographic order, and examined in mass, certain features are seen to undergo gradual change as a person's eye travels from specimens from, say, the center of the range of noveboracensis to specimens from southern localities adjoining the territory occupied by olivacea. One of these features subject to gradual change is the color of the underparts. Beginning at the Adirondacks of New York where a large number of the specimens have white underparts, the underparts become more intensely yellowish southward through the range of noveboracensis into that of olivacea. Indeed, this progressive trend seems to continue right on southward through the range of olivacea into that of peninsulae. Turning in the opposite direction we find that the least width of the underparts decreases gradually northward toward the range of noveboracensis. There is, likewise, a decrease to the northward in length of the skull and relative, as well as actual, narrowing of the braincase and tympanic bullae. However, in least width of color of underparts and the mentioned cranial features, the trend stops relatively abruptly at the southern boundary of the geographic range of noveboracensis and does not continue on, northward, into the range of noveboracensis as is the case with the change in intensity of yellowness of the underparts.

Two males, in the United States National Museum, Biological Surveys Collection, from near Leighton, Alabama, no. 178386 from the Tennessee River nine miles north [of Leighton?] and no. 180240 from La Grange Mountain, although clearly referable to olivacea on the basis of cranial characters, show some approach to noveboracensis in lesser size of the skull and agree with noveboracensis in the narrowness of the color of the underparts. Also, these specimens, like others from the northern part of the range of olivacea, for instance, no 31.227, Charleston Museum, from Mayesville, South Carolina, have the color of the underparts extended only part way out on the hind limb toward the foot. In specimens of olivacea from the southern part of its range the color of the underparts is extended onto the hind feet and this trend reaches its extreme in peninsulae, specimens of which have the feet and larger parts of the limbs marked with the light color of the underparts.

An adult female, no. 32.32, Charleston Museum, although typical of olivacea in most respects, is nevertheless an intergrade. The teeth are as small as in some specimens of noveboracensis. The size of the skull is only slightly nearer that of olivacea than it is to that of noveboracensis. The proportions of the skull, however, are distinctly those of olivacea.

Five other specimens, from northwestern Alabama, namely two from eight miles north of Nauvoo, two from Shoal Creek, and one from White Creek, also show intergradation between noveboracensis and olivacea. The remarks concerning color and color pattern of the specimens from Leighton apply equally well to the five from northwestern Alabama. In cranial characters, no. 51658 from Shoal Creek is referable to olivacea, as also is no. 51677 from the same place, providing it is a female rather than a male as sexed by the collector. No. 57146 from White Creek also is referable to olivacea although the skull shows some approach to that of noveboracensis. Of the two males from near Nauvoo, no. 51652 is to me indistinguishable from noveboracensis, but no. 51653 does have some characters of olivacea, although on the whole, the latter, too, seems to be a little nearer noveboracensis than olivacea. However, because the mean of these seven specimens from northwestern Alabama is nearer olivacea than noveboracensis the former name may be applied.

Another specimen from "Souinlonie" Creek, Clark County, Mississippi, has the coloration and rostral configuration of primulina, narrow mastoidal breadth and smaller teeth of noveboracensis and skull of large size with "full" braincase as in olivacea. No. 235364, U. S. Nat. Mus., from the Mobile River at the "L. and N. RR. Crossing," Mobile County, Alabama, although definitely olivacea, shows approach to arthuri in that the dorsal outline of the skull is longitudinally more convex and the tympanic bullae are less inflated than in olivacea and in that the color of the underparts is almost exactly as in the type specimen of arthuri. The young specimen labeled as from "Silver Springs," Florida, has large tympanic bullae (17 mm. long) and several characters that show its relationship to peninsulae as that race is now understood. Because the sex is unknown the identification as olivacea is tentative and is made on the assumption that the specimen is a male. If it is instead a female, the animal is referable to peninsulae.

An adult, female specimen in the Charleston Museum, no. 27.239.1, taken at St. Matthews, South Carolina, on December 8, 1927, contained four embryos which averaged 19 mm. in length and 47.75 centigrams in weight. Another adult female, in the Charleston Museum, no. 32.32, taken on February 21, 1932, at the same place, has prominent mammae, and the collector has noted that two were slightly active.

Sixteen of twenty-nine adults examined show infestation of the frontal sinuses by parasites. However, in none is the malformation of the frontal region so great as frequently occurs in M. f. noveboracensis.

Specimens examined.—Total number, 52, arranged alphabetically by states and from north to south by counties in each state. Except as otherwise indicated specimens are in the University of California Museum of Vertebrate Zoölogy.

Alabama. Lawrence County: White Creek, 1; Little Sand Mt., Shoal Creek, 2. Winston County: 7-1/2 mi. N Nauvoo, 1; 8 mi. N Nauvoo, 1. Lauderdale County: near Leighton, 9 mi. N Tennessee River, 1[91]. Colbert County: Leighton, 1[91]. Autauga County: Autaugaville, 1[91]. Dale County: Midland City, 1[91]. Mobile County: Mobile River, 12 mi. NE Mobile, 1[91].

Florida. Alachua County: Gainesville, 4[61]. Marion County: "Silver Springs," 1.

Georgia. Spalding County, 1. Lamar County, 1. Talbot County: southwest part of county, 1; Box Springs, near Geneva, 3; Upatoie Creek, 1 mi. SW Box Springs, 2; 3 mi. SE Geneva, 1; 4 mi. W Geneva, 1; 5 mi. W Geneva, 1; 2 mi. E Geneva, 1. Chattahoochee County, 2. Grady County: Beachton, 3[91]; locality no more definite than county, 4. Thomas County: Sinkola Plantation, 2; locality no more definite than county, 2. Charlton County: 1/2 mi. E Chesser's Island, Okefinokee Swamp, 1[58]. County in question: Billy's Island, Okefinokee Swamp, 1[91]; Okefinokee Swamp, 1[58].

Mississippi. Clark County: Souinlonie Creek, 1.

South Carolina. Darlington County: Society Hill, 1[91]. Sumter County: Mayesville, 1[11]. Calhoun County: St. Matthews, 2[11]. Georgetown County: Sampit, 1[11]. Charleston County: Rantowles, 1[11]; 8 mi. N Charleston, 1[11]. Beaufort County: Yemassee, 1[2].

Mustela frenata peninsulae (Rhoads)

Long-tailed Weasel

Plates 16, 17 and 18

Putorius peninsulae Rhoads, Proc. Acad. Nat. Sci. Philadelphia, 1894:152, June 19, 1894; Bangs, Proc. Biol. Soc. Washington, 10:10, February 25, 1896.

Mustela peninsulae, Miller, U. S. Nat. Mus. Bull., 79:98, December 31, 1912.

Mustela p. peninsulae, Bailey, Bailey Mus. and Library Nat. Hist., 1(no. 5):1, December 1, 1930.

Mustela frenata peninsulae, Hall, Carnegie Instit. Washington Publ. 473:105, September 20, 1936.

Type.—Female, young, part skull and skin; no. 8515, Acad. Nat. Sci. Philadelphia; Hudson's, Pasco County [14 miles north of Tarpon Springs], Florida; before 1895; obtained by W. S. Dickinson.

The skull has been cut vertically in two at the plane of the glenoid fossae. These fossae and all the cranium posterior to them are missing. In addition to the part of the cranium anterior to the glenoid fossae, the lower jaws are preserved complete. The teeth all are present and entire. The prominent sutures on the rostrum and palate show the specimen to be young and its small size leaves but little doubt that the animal was a female. The light facial markings are more extensive than in any of the referred specimens. In the type these light facial markings consist of a median isolated spot immediately in front of the ears, a larger one on the nose, with an interrupted bar on each side extending posteroventrally in front of and anterior to the eye, a wider bar, on each side, extending anterodorsally between the ear and eye and finally an isolated spot at the anterior border of each ear. The skin is stuffed and in fair condition except that the vertebrae remain in the tail.

Range.—Austral and probably Tropical life-zones of Florida south of latitude 29°. See figure 29 on page 221.

Characters for ready recognition.—Differs from M. f. olivacea in coarser pelage and larger tympanic bullae.

Description.—Size.—Male: No external measurements available. Female: The type a young animal and no. 2379, an adult from Tarpon Springs, measure respectively as follows: Total length, 375, 378; length of tail, 100, 130; length of hind foot, 40, 44.5.

Externals.—As described in Mustela frenata noveboracensis except that hairiness of foot-soles as shown in figure 20.

Color.—Upper parts (in winter) near tone 3 of Burnt Umber of Oberthür and Dauthenay, pl. 304. Dark spot at each angle of mouth present or absent. Tip of tail black. Underparts Reed Yellow except on chin and usually on legs where white. Upper lips white entirely around. Upper parts of uniform color. Color of underparts extends distally on legs over both sides of feet and on front legs over wrists. Proximal part of tail slightly lighter below than above. Least width of color of underparts, in seven specimens, averaging 41 (extremes 31-52) per cent of greatest width of color of upper parts. Black tip of tail, in each of two females, 45 mm. long; thus slightly longer than hind foot and amounting to 36 per cent of length of tail-vertebrae.

The spot at the angle of the mouth is absent in four of the ten specimens and is present on both sides in the other six.

Skull and teeth.—Male (based on an adult from Apopka and the anterior part of an adult from Enterprise): See measurements and plates 16-18. As described in Mustela frenata olivacea except that: Weight, 7.0 grams; basilar length, 49.8.

Female (based on an adult from Tarpon Springs, Florida): See measurements. As described in Mustela frenata olivacea except that: Weight, 4.7 grams; basilar length, 44.2; zygomatic breadth more than distance between anterior palatine foramen and anterior margin of tympanic bulla.

In comparison with M. f. olivacea, the insufficient material of M. f. peninsulae suggests that its skull averages larger and has relatively as well as actually larger and more inflated tympanic bullae.

Remarks.—The first published mention of this weasel seems to have been the original description which appeared in 1894. This description was based on a single specimen sent to Samuel N. Rhoads by W. S. Dickinson, who, in the following year, procured another specimen at Tarpon Springs. So far as known only eight other specimens, as listed under "Specimens examined," have found their way into collections of study specimens.

H. H. Bailey (1930:1) credits the range of this subspecies as extending south "to the shores of Florida Bay and the Gulf of Mexico, where ever high ground occurs."

Evidence of intergradation between M. f. peninsulae and M. f. olivacea is provided by specimens of olivacea from Gainesville, Florida, and the Okefinokee Swamp, Georgia. These specimens, on the average, have the color of the underparts wider, the skull larger, and the tympanic bullae relatively larger than do specimens of olivacea from farther north. In these features, approach to M. f. peninsulae is shown.

Light facial markings occur in this subspecies. They are similar to those possessed by weasels which occur at the same latitude and under corresponding climatic conditions on the Pacific Coast. The type specimen and one from Tarpon Springs have white facial markings. Two of the three specimens from Apopka also show white facial markings, although in reduced amount. One of the four specimens of M. f. olivacea from Gainesville, Florida, has well-developed light (white) facial markings. Also of the four specimens of M. f. olivacea examined from Okefinokee Swamp, Georgia, one has prominent white facial markings. However, in it the pattern is so unusual as to suggest that it is an instance of partial albinism rather than an outcropping of a racial tendency, or a pattern of coloration induced by climatic factors.

None of the eight available skulls show any infestation of the frontal sinuses by parasites.

Specimens examined.—Total number, 10, arranged by counties from west to east.

Florida. Pasco County: Hudson's, 1[1]. Pinellas County: Tarpon Springs, 1[1]. Hernando County, 1[91]. Polk County: Auburndale, 1[91]; no locality more definite than county, 1[91]. Orange County: Apopka, 3[61]. Volusia County: Enterprise, 1[60]. Seminole County: Osceola, 1[2].

Mustela frenata spadix (Bangs)

Long-tailed Weasel

Plates 16, 17, 18, 31, 32 and 33

Putorius longicauda spadix Bangs, Proc. Biol. Soc. Washington, 10:8, February 25, 1896; Merriam, N. Amer. Fauna, 11:21, figs. 10, 11, June 30, 1896; Cory, Mamm. Illinois and Wisconsin, p. 374, 1912.

Mustela longicauda spadix, Miller, U. S. Nat. Mus. Bull., 79:98, December 31, 1912; Bailey, Journ. Mamm., 10:156, May 9, 1929.

Mustela longicauda, Johnson, Journ. Mamm., 11:439, November 11, 1930.

Mustela noveboracensis, Murie, Journ. Mamm., 16:321, November 15, 1935.

Mustela frenata spadix, Hall, Carnegie Instit. Washington Publ. 473:105, November 20, 1936.

Type.—Male, young, skull and skin; no. 3265/1786, Amer. Mus. Nat. Hist.; Fort Snelling, Hennepin County, Minnesota; June 25, 1889; obtained by Edgar A. Mearns; original no. 812.

The skull is complete although there are fractures on the top of the braincase, on the right side of the braincase and at the middle of the right zygomatic arch. The teeth all are present and entire. The skin, although overstuffed, is complete, well preserved, and in summer pelage.

Range.—Upper Austral and Transition life-zones of Minnesota, northern and western Iowa, southeastern North Dakota, eastern part of South Dakota, and northeastern Nebraska. See figure 29 on page 221.

Characters for ready recognition.—Differs from M. f. noveboracensis and M. f. primulina in that specimens of all ages have least width of color of underparts amounting to more than 41 per cent of greatest width of color of upper parts, and have light color of underparts extended onto hind foot rather than stopped short of ankle; adults with hind feet more than 50 in males and 40 in females; orbitonasal length more than 15.5 in males and 13.5 in females; length of tooth-rows more than 18.0 in males and 15.7 in females; mastoid breadth more than 25.5 in males and 22.0 in females. From M. f. longicauda by color darker than near (h) Clay Color, in males by a flattened occiput in which the depth of the skull, exclusive of the sagittal crest and taken at the anterior border of the basioccipital, amounts to less than 58 per cent of the mastoid breadth.

Description.—Size.—Male: Three adults from Elk River, Minnesota, yield average and extreme measurements as follows: Total length, 458 (444-467); length of tail, 154 (140-165); length of hind foot, 55 (52-59). Tail averages 51 per cent as long as head and body. Length of hind foot averages more than basal length. Corresponding measurements of three subadults from Madison, Minnesota, are as follows: 453 (438-469); 157 (152-165); 50 (47-51). Tail averages 53 per cent as long as head and body.

Female: Three adults from Elk River, Minnesota, yield average and extreme measurements as follows: Total length, 387 (380-391); length of tail, 131 (121-138); length of hind foot, 44 (43-46). Tail averages 51 per cent as long as head and body. Length of hind foot more or less than (approximately equal to) basal length. Corresponding measurements of two adults and one subadult from Madison, Minnesota, are as follows: 385 (379-396); 137 (119-159); 42 (38-44). Tail averages 55 per cent as long as head and body.

The average differences in external measurements of the two sexes from Elk River, are: Total length, 71; length of tail, 23; length of hind foot, 11. At Madison, corresponding differences are 68, 20, and 8. Two adult females from Elk River, Minnesota, weigh 205 and 210 grams.

Externals.—Longest facial vibrissae black, brown, or white (often all three colors in same specimen) and extending beyond ear; carpal vibrissae same color as underparts and extending to apical pad of fifth digit; hairiness of foot-soles (in summer pelage) as shown in figure 19.

Color.—Winter pelage all white except tip of tail. In southern part of range sometimes assumes a brown winter coat. Summer pelage with upper parts ranging from near (16´) Cinnamon Brown to Vandyke Brown. Chin and upper lips white. Remainder of underparts ranging from near (a) Olive Ocher to Ochraceous Buff and Pale Orange Yellow. Tip of tail at all times black. Upper parts of uniform color except for occasional slight darkening of nose. Color of underparts extends distally on posterior sides of forelegs over toes onto antipalmar faces of feet and ankles, on medial sides of hind limbs to ankle, over antiplantar faces of toes and distomedial fourth of each tarsus, and over proximal fifth to third of under side of tail. Least width of color of underparts averaging (in 3 specimens from Elk River) 54 (47-59) per cent of greatest width of color of upper parts. Black tip of tail averaging same length as hind foot and 28 per cent of length of tail-vertebrae. Save for the greater width of the light-colored underparts and relatively short black tip of the tail, both features of M. f. longicauda, spadix is variously intermediate, depending on locality, as between noveboracensis and longicauda.

Skull and teeth.—Male (based on 3 adults from Elk River, Minn.): See measurements and plates 16-18. As described in Mustela frenata longicauda except that: Weight, 5.6 (5.0-6.5); basilar length, 49.0 (48.7-49.2); zygomatic breadth sometimes less than distance between anterior palatine foramen and anterior margin of tympanic bulla; postorbital breadth more or less (about equal to) 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; anterior margin of tympanic bulla as far posterior to foramen ovale as width of 4 to 5 upper incisors; height of tympanic bulla more or less than distance from its anterior margin to foramen ovale; length of tympanic bulla less than length of rostrum; anterior margin of masseteric fossa below talonid of m1.

Female (based on 4 adults from Elk River, Minn.): See measurements and plates 31-33. As described in Mustela frenata longicauda except that: Weight, 3.5 (3.3-4.0) grams; basilar length, 42.9 (42.3-43.2); least width of palate more or less than greatest length of P4; tympanic bulla as far posterior to foramen ovale as width of 3 to 5 upper incisors.

The skull of the female averages 33 per cent lighter than that of the male.

Skulls of adult males of spadix from Elk River, Minnesota, as compared with those of longicauda from Alberta, are larger in every part measured. Relative to the basilar length these skulls of spadix are broader across the mastoid region, narrower across the zygomata, deeper through the plane of the postorbital processes, shallower through the braincase and have relatively shorter tympanic bullae. Whereas the tympanic bullae of longicauda are, on the average, approximately as long as the rostrum (orbitonasal length), in spadix the rostrum is longer than the bulla. Viewed posteriorly, the braincase of spadix is seen to be much shallower and wider than that of longicauda. Indeed, the depth of the braincase, measured at the anterior end of the basioccipital, amounts to only 56 per cent of the mastoid breadth in spadix as against 61 per cent in longicauda. The longer, waistlike, postorbital constriction, relatively smaller braincase, and especially the relatively narrower zygomatic expanse in spadix imparts to its skull a more slender appearance than has the skull of longicauda. These differences are not shown by the skulls of females. To be sure, spadix, in most of its cranial measurements, averages slightly larger, has a relatively shallower braincase and is relatively deeper through the postorbital processes, but these differences are so slight that inclusion of one more specimen, of slightly different proportions, in the average might cause the average measurements to read as they do in longicauda.

Compared with noveboracensis, from Massachusetts, adult skulls of spadix, taking sex into account, are larger in every part measured and are relatively as well as actually wider and deeper throughout. Also, in spadix: Sagittal and lambdoidal crests higher, especially in females; anterior margin of tympanic bulla projecting up sharply from squamosal; occiput more flattened in posterior view; tooth-rows relatively and actually longer but orbitonasal length relatively shorter; postorbital processes more robust; zygomatic arches widely bowed outward rather than evenly rounded; canines larger; squamosal less swollen ventrally, especially in females. Between noveboracensis and spadix, the differential cranial characters are greater in number and degree between females than between males. Comparison of the skull with that of M. f. primulina is made in discussion of that subspecies.

Remarks.—Edgar A. Mearns in 1889 and the early nineties took several specimens of this weasel and it was principally on these that Bangs in 1896 (p. 8) based his description. The best material, however, is that from Elk River, Minnesota, collected in later years by Bernard Bailey, and supplemented by one specimen taken in 1885 by Vernon Bailey and another by his sister Anna Bailey in 1891 at the same place.

Mustela frenata spadix has just one structural feature of a "unique" kind which serves to differentiate it from the geographically adjoining subspecies. This feature is large size. The other diagnostic characters ascribed to spadix are of an intermediate sort—intermediate as between two extremes, one found to the westward in longicauda and the other to the eastward in noveboracensis. For example, the dark-colored upper parts are merely darker than in longicauda and merely lighter than in noveboracensis. The color is not "different"; it is only "intermediate." Furthermore, each of the characters ascribed to spadix, including large size itself, undergoes change from one part of its geographic range to another; the characters are not constant over a wide area. Indeed, excepting the large size which remains relatively uniform over the northern two-thirds of the range, no two localities have been found from which the specimens can be said really to agree in characters.

By way of illustration, the coloration of the upper parts may be cited. Near the range of noveboracensis the average coloration of individuals from one locality is only a little lighter than in noveboracensis. Farther westward the average coloration is a little lighter and farther westward yet, toward the range of the extremely light colored longicauda, the average coloration is lighter still. Although all these animals are darker than longicauda and lighter than noveboracensis, those from the three places do not agree among themselves. Because of the lack of more than one character of a "unique" kind and because of the inconstancy, geographically, of other characters, and for that matter, lack of constancy geographically in combination of characters, the writer regards spadix as a barely recognizable subspecies.

Examination of the specimens of spadix shows that the individual variation in a single species is greater in a region of intergradation than it is some distance inside the borders of the geographic range of a well-marked subspecies. This is illustrated by three specimens of M. f. spadix in fresh summer pelage from the single locality, Elk River, Minnesota. In these, the color of the upper parts varies from a little darker than Cinnamon Brown to Vandyke Brown. At any one locality well within the range of longicauda, or noveboracensis, there is nowhere nearly so much variation in color, even in much larger series of specimens.

Study of the specimens here assigned to spadix reveals that some features regarded as of diagnostic value for one or the other of the two races, longicauda and noveboracensis, behave differently. For example, the dark coloration of the upper parts, which is characteristic of noveboracensis, manifests itself far westward within the range of spadix whereas the wider extent of the light-colored underparts, which is characteristic of longicauda, and the Olive Ocher, rather than Pale Orange Yellow, color of these underparts, are seen in varying degree all the way across the range of spadix. Thus, these animals are colored above like noveboracensis and below like longicauda, but not vice versa. In these animals, then, the longicauda type of underparts is dominant, in one sense of the word, over the noveboracensis type of underparts, and the noveboracensis type of upper parts is dominant over the longicauda type of upper parts. Each of these features is subject to actual intergradation and does not always behave as a "unit character," that is to say, one which is either present or absent. However, the noveboracensis type of upper parts is carried much farther west before being diluted than is the noveboracensis type of underparts. Indeed, within the range of noveboracensis itself, the broad extent of the longicauda type of underparts is manifest. This is, of course, near the western margin of the range of noveboracensis.

The large size of males of spadix, as exemplified by specimens from Elk River (see measurements on page 421), seems to be retained across the northern part of the range here assigned to the subspecies. This larger size than is found in longicauda from Alberta, is shown also by some specimens from eastern North Dakota which are assigned to longicauda. However, the average of these Dakotan specimens, all characters considered, is nearer to my concept of longicauda.

Inspection of the cranial measurements of spadix shows also that in addition to its large size it is distinguishable from any one of the geographically adjoining races by its relatively (to basilar length) greater, as well as actually greater, mastoidal breadth. This might be included with size as a unique character distinguishing spadix from longicauda and noveboracensis. However, it is not clear whether or not this greater mastoidal breadth is more than a function of the large size.

Excepting the greater mastoidal breadth and generally larger size of the skull, the cranial features distinguishing males of spadix from longicauda are features in which spadix shows approach to noveboracensis. This is true, in spadix, of the relatively longer (in comparison with longicauda) rostrum, relatively lesser zygomatic breadth, relatively shallower braincase measured at the anterior end of the basioccipital, and relatively deeper skull as measured at the posterior borders of the last upper molars. This same approach to noveboracensis already has been pointed out with respect to color of the upper parts and is evident also in the relative shortness of the tail which averages only 51 per cent of the length of the head and body rather than 55 per cent as in longicauda.

Because the longicauda type of animal previously has been regarded as specifically distinct from the noveboracensis type of animal, comment is offered below on selected specimens, referred to spadix, which are regarded as intergrades with noveboracensis or with other subspecies.

No. 8722, Univ. Wisconsin, adult male, in the white winter coat, from north central Itasca County, Minnesota, obviously has characters of M. f. spadix or longicauda that occur to the west and M. f. noveboracensis of the east. Selected outstanding characters of longicauda are its long tail, anteriorly truncate tympanic bullae and large teeth. Characters indicating its affinities with noveboracensis are smaller size of skull, general narrowness of skull, and relatively low tympanic bullae. The skull is intermediate as regards several individual structural features. For example, although long and narrow and in this feature more nearly approaching noveboracensis, the skull is wider than usual in that subspecies and thus approaches that of longicauda or spadix. The hind foot, in the dried state, measures 47 millimeters. This large hind foot, obviously long tail (the specimen lacks external measurements), and anteriorly truncate bullae constitute basis for here referring the specimen to spadix. However, the seemingly small size of the body and the narrow skull clearly show relationship to noveboracensis.

Specimens, referred to spadix, from northern Iowa, are instructive as showing what happens where the ranges of noveboracensis, primulina, spadix, and perhaps longicauda, meet. No. 47167, Univ. Mich. Mus. Zoöl., a nearly adult female, taken on November 22, 1915, at Island, Clay County, and in process of assuming a brown winter pelage, retains enough of the dark summer pelage to show that the color was slightly lighter than average for spadix. The color pattern, white lips, and extension of light color of the underparts onto the feet, agrees with spadix or longicauda as does also the long tooth-row. The overall length of the skull is intermediate between that of spadix and primulina. The proportions of the anterior part of the skull and of the tympanic bullae resemble those found in primulina. A subadult male skull only, no. 123846, American Museum of Natural History, from Webb, Clay County, shows approach to primulina in the narrowness of the rostrum.

A young male from Ruthven, Iowa, no. 48340, Univ. Michigan, has a large skull approaching in size that of spadix, has the longicauda-spadix type of light-colored underparts and color pattern, and is slightly darker above than true longicauda. Another subadult male in the white winter coat from Palo Alto County, no. 35756, Univ. Michigan, has a large skull, which shows approach to primulina in its narrowness anteriorly and in some other features. Although the tail is of moderate length, the body is large as in spadix or longicauda, and the length of the hind foot suggests spadix or longicauda.

A subadult male, no. 425a, Iowa State College, from Manson, Iowa, in brown winter pelage, agrees with primulina in the restriction of the area of the light color of the underparts and in less expanded zygomatic arches. The teeth are intermediate in size between those of noveboracensis and primulina on the one hand and those of spadix and longicauda on the other. In other respects it agrees with, or is more nearly like, spadix.

An adult female, no. 426a, Iowa State College, from Barnum, in the brown winter coat, agrees with primulina except that the orbitonasal length of the skull is more as in spadix and the presence of some light color on the lower part of the hind legs suggests spadix. The skull only, no. 440a, Iowa State College, labeled merely Webster County, Iowa, is almost a duplicate of no. 426a. A subadult male, no. 427a, Iowa State College, from Moorland, Iowa, only about six miles southeast of Barnum, likewise is indistinguishable from primulina except for having a white winter coat and in being relatively broad in the mastoidal region. Nevertheless, both of these animals are here referred to spadix because the average of specimens from this general area is nearer that of spadix. No. 497a, Iowa State College, an adult female in white winter pelage, from Ames, approaches primulina in the narrow rostrum and smaller teeth but otherwise approaches or even agrees with spadix.

Two adult males, without external measurements, from Pilot Mound, Iowa, have skulls quite like males of longicauda from Alberta. The only approach noted to eastern forms is the restricted color of the underparts on no. 2856, Coe College, which has a brown winter coat. The color of the underparts is not extended so far out on the feet as in longicauda. Also the tympanic bullae of this specimen are a trifle narrower. The other male, no. 2652, is in the white winter coat. The one female from the same place, no. 2660, Coe College, in brown winter pelage, has a skull notably unlike that of longicauda or spadix; the skull is narrower and practically indistinguishable from that of the largest female skull of primulina available from Lawrence, Kansas, save that the tooth-row is much longer. The color pattern also agrees with that of primulina or noveboracensis in that the color of the underparts extends only as far as the knee on the hind legs and is narrow on the belly. Nevertheless, another adult female, no. 120a from Amaqua Township, some 6 miles southwest of Pilot Mound, is in all respects typical of spadix. This is the more remarkable because another comparable specimen from less than 20 miles to the southwest in Worth Township is equally typical of primulina.

Two young females from Chester, Iowa, nos. 2656 and 2874/2873, Coe College, have skulls larger than those of corresponding age of primulina or noveboracensis. The color is as in spadix. The color pattern of the underparts also is as in spadix or longicauda except that the width of the area of light color on the belly is restricted somewhat although not so much as in noveboracensis or primulina. Of four males from the same place, also in the collection of Coe College, no. A2874 is a white skin only and does not provide diagnostic characters. The three other males, each in summer pelage, are marked and colored as are the two females from the same place except that male no. 2861 has the color of the underparts so much attenuated on the hind legs that it barely, uninterruptedly, extends to the feet. No. 2658 is young, or perhaps barely subadult. The skull is large and referable to spadix. The two adults, nos. 2861 and 2657, differ cranially from typical (Elk River, Minn.) spadix only in being slightly narrower across the mastoids and in having the bullae a little narrower. In these departures they show some approach to primulina and to noveboracensis. Another male, subadult, no. 2867, Coe College, from Decorah, which has acquired half of the white winter coat, agrees with the males from Chester except that the preorbital part of the skull is shortened much as in some specimens of primulina.

From Lansing, in extreme northeastern Iowa, a large subadult male, no. 2864, Coe College, of 453 mm. in total length and half through with acquiring the white winter coat, agrees with the males previously described from Chester except in having the palate narrower as in noveboracensis. The adult female available from Lansing, no. 2863/2862, Coe College, in white winter pelage except for the top of the head, although a large skin, has a skull smaller than that of any spadix or longicauda and of about the same size as that of no. 3838, Univ. Kansas Mus. Nat. Hist., of primulina, from Lawrence, Kansas, except that the skull of no. 2863/2862 is much narrower across the mastoids. This specimen, then, shows approach to noveboracensis in narrowness of the mastoidal region, to primulina in other respects and to spadix.

Many of these instructive specimens from Iowa, made available to the present writer by Mr. W. F. Kubichek, were brought together at the Coe College Museum by the late B. H. Bailey. Most of them were obtained from trappers who did not supply the conventional external measurements taken in the flesh. Even though these are lacking, the specimens clearly show that actual intergradation occurs where the ranges of M. f. longicauda, spadix, noveboracensis and primulina meet.

The dark color of the upper parts, restriction of the color of the underparts on the ankles with the result that the color reaches the toes in interrupted fashion, and large skull, of no. 18912 of the Museum of the University of South Dakota, from Roberts County, South Dakota, clearly place this specimen with spadix, rather than with longicauda. Likewise, male, subadult, no. 11376, Univ. South Dakota, from Clay County, South Dakota, is referable to spadix. Although without external measurements, the specimen obviously is large. The patch of summer pelage on its head and neck is darker than the summer pelage of longicauda, and the orbitonasal length is greater than the length of the tympanic bullae; all these features are characters of spadix. The adult male from Fort Sisseton, South Dakota, no. 188407, United States National Museum, figured by Merriam (1896, p. 20, figs. 7-9), is almost exactly intermediate between longicauda and spadix, although here referred to the latter.

Five specimens, nos. 147375, 147432, 147762, 148720 and 148721, U. S. Nat. Mus., including 3 skulls only from Beemer, Cuming County, Nebraska, are intergrades between M. f. longicauda, M. f. primulina and M. f. spadix. One skin is in white winter pelage and the other, a female, is in summer pelage which in coloration and color pattern agrees with that of spadix. External measurements of the male agree with those of longicauda. Measurements of the female agree with those of spadix except that the tail is shorter as in primulina. The skulls are as long as in longicauda but are more slender than in either longicauda or spadix although nearer the latter in this respect. In dorsal aspect, the skulls especially posteriorly to the orbital region, resemble primulina. All points considered, the animals seem best referred to spadix.

Although the degree of development of certain morphological features has been settled upon as indicative of the race spadix, some doubt remains as to where the western boundary of its range should be shown. This results from the fact that color has been taken into account as one diagnostic feature and this feature is lacking in the white winter specimens which, from the following places, are all that are available: Kittson County, Minnesota; Moorhead, Minnesota; Casselton and Valley City in North Dakota; Armour, South Dakota and Clay County, South Dakota. In summary, more specimens in the summer coat will be required to establish definitely the boundary between the ranges of longicauda and spadix.

Surber (1932:49) has referred to additional specimens of this weasel in the University of Minnesota Museum as from Winona, Hennepin and Isanti counties of that state.

At Elk River, Minnesota, B. Bailey (1929:156) found this species to be about half as abundant as Mustela cicognanii and that it is "more often found in the open timber and about the dry ridges and fields." Of seventeen adult or subadult skulls of this race from Minnesota, ten have obvious marks of infestation of the frontal sinuses. In no skull, however, has the infestation resulted in so much malformation, as occurs in noveboracensis.

Specimens examined.—Total number, 76, arranged alphabetically by states and from north to south by counties in each state.

Iowa. Lyon County: Granite, 1[65]. Howard County: Chester, 6[12]. Winneshiek County: Decorah, 1[12]; 8 mi. NE Ossian, 1[76]. Allamakee County: Lansing, 2[12]. Clay County: Island, 1[76]; Webb, 1[2]. Palo Alto County: Ruthven, 1[76]; no locality more definite than county, 1[76]. Calhoun County: Manson, 1[65]. Webster County: Barnum, 1[65]; Moorland, 1[65]; no locality more definite than county, 1[65]. Boone County: Pilot Mound, 3[12]; Amaqua Township, Sec. 19, 1[65]. Story County: Ames, 1[65].

Minnesota. Kittson County, 1[2]. Roseau County: 2-1/2 mi. SW Roseau, Jadis Township, 1[14]. Itasca County: T. 61N, R. 26W, 1[102]. Clay County: Moorhead, 2[9]. Atkin County: Atkin, 1[50]. Otter Tail County: Lake Lizzie, 1[9]; Parkers Prairie, 1[57]. Grant County: 3 mi. NW Barrett, 1[76]. Benton (now Mille Lacs?) County: Princeton, 1[91]. Sherburne County: Elk River, 14 (6[59], 4[14], 3[91], 1[74]). Hennepin County: Fort Snelling, 6 (5[2], 1[91]). Carver? County: Chaska, 1[60]. Lac qui Parle County: Madison, 5 (3[91], 2[1]); no locality more definite than county, 2 (1[68], 1[75]). Yellow Medicine County: Wood Lake, 1[2]. Blue Earth County: Rapidan, 1[64]. County in question: Moore Lake, 1[91].

Nebraska. Cuming County: Beemer, 5[91].

North Dakota. Cass County: Fargo, 1[91]; Casselton, 1[91]. Dickey County: Oakes, 1[91].

South Dakota. Roberts County, 1[102]. Marshall County: Fort Sisseton, 1[91]. Douglas County: Armour, 1[14]. Clay County, 1[102].

Mustela frenata longicauda Bonaparte

Long-tailed Weasel

Plates 16, 17, 18, 31, 32 and 33

Mustela longicauda Bonaparte, Charlesworth's Mag. Nat. Hist., 2:38, 1838.

Putorius longicauda, Baird, Mamm. N. Amer., p. 169, 1858; Coues, Fur-bearing animals, p. 136, 1877; Bangs, Proc. Biol. Soc. Washington, 10:7, figs. 1, 1a of pls. 1, 2 and 3, February 25, 1896; Merriam, N. Amer. Fauna, 11:19, pl. 3, figs. 3, 3a, 4, 4a, pl. 5, figs. 1, 1a, text figs. 7-9, June 30, 1896.

Mustela longicauda longicauda, Bailey, N. Amer. Fauna, 49:166, January 8, 1927.

Mustela frenata longicauda, Hall, Carnegie Instit. Washington Publ. 473:105, November 20, 1936; Hall, Canadian Field-Nat., 52:108, October, 1938.

Mustela frenata, Sowls, Journ. Mamm., 29:126, May 14, 1948.

Type.—Possibly not in existence. No. 43.3.3.3 [from Carlton House, Saskatchewan] in the British Museum of Natural History has been regarded by several zoölogists as the type. It is a subadult female, skull and skin, from North America. See the account of M. erminea cicognanii for reasons for and reasons against regarding this specimen as the holotype.

No. 43.3.3.3 from the collection of Dr. John Richardson is in the white winter coat and now (Sept. 24, 1937) is prepared as a study skin. Evidences of its once having been mounted are: holes in the soles of the hind feet for supporting-wires, large straight wire in the tail, folds in the skin of the now backward-projecting hind feet, and unevenness of the skin on the back resulting from straightening out the specimen. The tip of the tail and some skin from the middle of the belly are missing. Otherwise the skin is intact. The skull is that of an animal in its first year, lacks the zygomatic arch on each side, but otherwise is complete and unbroken. The teeth all are present and entire except that p2 on the right side is missing from its alveolus.

Range.—Transition and Upper Sonoran life-zones of the Great Plains, southward from central Alberta, Saskatchewan and southern Manitoba through eastern Montana, the Dakotas and Nebraska into southeastern Wyoming, northeastern Colorado and western Kansas. See figure 29 on page 221.

Characters for ready recognition.—Differs from M. f. primulina in near (h) Clay Color rather than Brussels Brown of upper parts, least width of color of underparts more than 40 per cent of greatest width of color of upper parts, color of underparts extended onto hind foot rather than stopped short of ankle, zygomatic breadth more than 28.8 in adult males and more than 24.1 in adult females; from M. f. spadix in lighter color being near (h) Clay Color, in males by deeper occiput in which the depth of the skull, exclusive of the sagittal crest and taken at the anterior border of the basioccipital amounts to more than 59 per cent of the mastoid breadth; from M. f. oribasus in near (h) Clay Color rather than near (14n) Brussels Brown color of the upper parts and in males by deeper occiput in which the depth of the skull, exclusive of the sagittal crest and taken at the anterior border of the basioccipital, amounts to more than 59 per cent of the mastoid breadth; from M. f. alleni in larger size, adult males having a total length of more than 400 millimeters, hind foot more than 45, basilar length more than 43.5, and females having a total length of more than 375 and basilar length not less than 40.0; from M. f. nevadensis in near (h) Clay Color rather than near (14n to 1) Brussels Brown of upper parts, basilar length more than 40 in females and averaging more than 45 in males; from M. f. neomexicana by near (h) Clay Color rather than Buckthorn Brown color of upper parts, absence of white and Argus Brown facial markings, and length of tooth-rows amounting to more than 37 per cent of basilar length.

Description.—Size.—Male: Five adults from Alberta yield average and extreme measurements as follows: Total length, 438 (418-473); length of tail, 158 (140-193); length of hind foot, 50 (46-54). Tail averages 56 per cent as long as head and body. Length of hind foot averaging more than basal length. Corresponding measurements of five adults and subadults from North Dakota are as follows: 465 (445-516); 164 (150-179); 51 (50-54). Tail averages 55 per cent as long as head and body.

Female: Six adults (Alberta, 4; Saskatchewan, 1; Manitoba, 1) yield average and extreme measurements as follows: Total length, 401 (383-425); length of tail, 145 (141-159); length of hind foot, 43 (41-44). Tail averages 57 per cent as long as head and body. Length of hind foot more or less than (approximately equal to) basal length.

The average differences in external measurements of the two sexes are: Total length, 37; length of tail, 13; length of hind foot, 7. General comparisons indicate that the Alberta-taken males may not attain so large a size as those from some other areas. Thus the differences in external measurements might be some greater elsewhere, say, in North Dakota.

Externals.—Longest facial vibrissae black, brown or white (often all three colors in same specimen) and extending beyond ear; carpal vibrissae same color as underparts and extending to apical pad of fifth digit; hairiness of foot-soles (in summer pelage) only slightly greater than shown in figure 20.

Color.—Winter pelage all white except tip of tail. Summer pelage with upper parts near (h) Clay Color or near tone 3 and 4 of Snuff Brown of Oberthür and Dauthenay, pl. 303. Chin and upper lips white. Remainder of underparts ranging from near (a) Olive Ocher to near (16´) Ochraceous Buff. Upper parts of uniform color except for occasional darkening of head in front of ears. Color of underparts extends distally on posterior sides of forelegs over toes onto antipalmar faces of feet and wrists, on medial sides of hind limbs to ankles over antiplantar faces of toes and distomedial third of each tarsus, and over proximal fourth to third of under side of tail. Least width of color of underparts averaging, in a series of 10 males from Alberta, 58 (45-60) per cent of greatest width of color of upper parts. Corresponding figures for 10 females from the same place are 57 (50-74). Black tip of tail in same series of males, most of which are in full summer pelage, averaging 43 (35-60) mm. long. Thus, averaging shorter than hind foot and 27 per cent of length of tail-vertebrae.

As compared with M. f. neomexicana, longicauda lacks the white facial markings, black ears, black forehead and nose, but otherwise is similarly colored. As compared with M. f. nevadensis, M. f. oribasus and M. f. spadix, each of color pattern similar to longicauda, selected differences of longicauda are its much lighter color, especially of the upper parts, with less conspicuous darkening on the nose. From M. f. primulina, longicauda differs in lighter color of upper parts, reddish rather than yellowish underparts, and light rather than dark-colored hind feet.

Skull and teeth.—Male (based on 5 adults from Alberta): See measurements and plates 16-18; weight, 4.7 (4.6-4.9) grams; basilar length, 46.0 (44.7-46.8); zygomatic breadth more than distance between condylar foramen and M1 or than between anterior palatine foramen and anterior margin of tympanic bulla; mastoid breadth more than postpalatal length; postorbital breadth less 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 greater than distance between foramen opticum and anterior margin of tympanic bulla; breadth of rostrum more or less (usually less) than length of tympanic bulla; least width of palate less than greatest length of P4; anterior margin of tympanic bulla as far posterior to foramen ovale as width of 3 to 4 (including I3) upper incisors; height of tympanic bulla 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 talonid of m1 or anterior half of m2.

Female (based on 5 adults: Alberta, 3; N. D., 1; Sask., 1.): See measurements and plates 31-33; weight, 3.1 (2.8-3.5) grams; basilar length, 42.3 (40.0-43.7); zygomatic breadth more or less (approximately equal to) than distance between condylar foramen and M1 or that between anterior palatine foramen and anterior margin of tympanic bulla; postorbital breadth less than length of upper premolars and more or less than width of basioccipital measured from medial margin of one foramen lacerum posterior to its opposite; least width of palate not more than greatest length of P4; tympanic bulla as far posterior to foramen ovale as width of 3 to 4 (including I3) upper incisors; height of tympanic bulla not less than distance from its anterior margin to foramen ovale; length of tympanic bulla more than length of lower molar and premolar tooth-rows and longer or shorter than rostrum.

The skull of the female averages 34 per cent lighter than that of the male.

Comparisons of the skull with those of M. f. primulina, M. f. spadix, M. f. oribasus, M. f. alleni, M. f. nevadensis, and M. f. neomexicana are made in accounts of those subspecies.

Remarks.—Richardson's (1829:47) account on which Bonaparte may be said to have based his name, records measurements in inches and lines which I transpose into millimeters as follows: Total length, 440 mm.; length of head and body, 305; length of tail (vertebrae), 135; length of tail (including fur), 164 mm. Specimen no. 43.3.3.3 in the British Museum, which has by some persons been regarded as the type, yields measurements as follows: Total length, 408 (which allows for 15 mm. loss of the fleshy part of the end of the tail); length of head and body, 272; length of tail (vertebrae), 136 (= 121 + 15); length of tail (including fur), 162 (142 + 20 mm. that appears to have been lost). Richardson's specimen would appear to have been of unusual proportions and to have been larger than no. 43.3.3.3. Some reasons for and reasons against regarding this specimen as the holotype are given in the account of M. erminea cicognanii.

The name longicauda was applied to practically all long-tailed weasels of the western United States at one time but as one after another of the geographic variants in the mountainous regions were designated as separable, the name longicauda came to be restricted to the light-colored, relatively large, animal of the Great Plains.

The intergradation of longicauda with spadix and oribasus has been commented on in the discussions of those subspecies. The larger size and darker color of specimens referred to longicauda from Devils Lake and Grafton, North Dakota, are features indicative of intergradation there with spadix. Two young females from Waterton Lake Park, Alberta, by their darker than average color, suggest intergradation with oribasus, as, for that matter, does the specimen from Waterton Lake [= Chief Mountain Lake, in Montana] itself, which, however, is even darker than the two specimens taken on the Canadian side of the line and hence is referred to oribasus. An adult female, no. 175586, U. S. Nat. Mus., from Moose Pass, Alberta, examined after the above was written, is larger than any other female seen of longicauda and in this respect may show approach to oribasus, which in the northern part of its range is of large size as judged by males from the Bowron Lake region.

One male, no. 8564, Nat. Mus. Canada, from Max Lake, Turtle Mountain, Manitoba, presents puzzling characters. The external measurements of 465, 170, and 57, are in keeping with the great length of the skull which has a basilar length of 48.8. The tooth-rows are 19.3 in length and the mastoid breadth, 25.4. The relative narrowness indicated by the mastoid breadth is maintained throughout the skull. The only other specimens relating to the Turtle Mountains that have been seen are two male, skins without measurements or corresponding skulls, nos. 38902 and 38903, Amer. Mus. Nat. Hist., labeled as from either "Stump Lake or Turtle Mts.," North Dakota. One of these, no. 38902, is much darker than the other. Possibly it is from the Turtle Mountains and the other, lighter-colored one, is from Stump Lake. Study of additional specimens from the Turtle Mountains might show the existence there of a distinct race.

Four specimens, in the collection of Myron Swenk, from Inland, Clay County, Nebraska, are instructive as showing how intergradation occurs between primulina and longicauda. A subadult male, no. 10, is intermediate in external measurements and in color but in each instance is nearer primulina. The same is true of the least width of the color of the underparts. The color of the underparts extends uninterruptedly over the hind legs to the toes as in longicauda, but is absent from the underside of the tail as in primulina. In the skull, the basilar length, breadth of bulla, and size of teeth are nearer longicauda, as are also the ratios to the basilar length of the length of tooth-rows, breadth of the rostrum, length of the tympanic bulla, and depth of the braincase at the anterior margin of the basioccipital. Ratios to the basilar length of the interorbital breadth, mastoid breadth, zygomatic breadth, and depth of the skull at the posterior borders of the upper molars are nearer to those of primulina. The relatively long rostrum, as represented by the orbitonasal length, is nearest to that of spadix. A young, almost subadult, female, no. 7, agrees with primulina in color, color pattern, and length of hind foot. The other external measurements are intermediate, but nearer those of primulina. Size of skull and teeth are as in longicauda. Relative proportions of parts of the skull are not diagnostic in specimens as young as this female. An adult female, skull only, no. 8, agrees with, or approaches nearer to, longicauda in size of skull and teeth and in relative proportion of every part studied. A juvenile, skull only, of questionable sex, no. 9, provides no diagnostic characters. On the basis of color, these specimens from Inland are distinctly nearer primulina. On the basis of cranial characters they are distinctly nearer longicauda. External measurements are intermediate and are a little nearer those of primulina. By placing the most weight on the cranial characters, the animals may be referred to longicauda. The same may be said of 2 skins, one skin with a skull, from Hastings, Nebraska. In each skin the color-pattern is as in primulina; in one the under side of the tail is nevertheless lighter-colored more as in longicauda and the skull, adult male 121651 American Museum of Natural History, approaches nearer to primulina in narrowness but has the large teeth of longicauda.

Intergradation with neomexicana is suggested by one specimen, no. 7936, Univ. Kans., from Thomas County, Kansas, which has well-developed white facial markings.

The specimen, no. 180, Kansas Agric. College, from Glasco, is mounted, of large size, in white winter pelage, and lacks external measurements. On the basis of its obvious large size, and a hind foot measurement of 49 millimeters obtained from the mounted skin, the animal is provisionally referred to longicauda rather than to primulina.

Putorius culbertsoni is a name now credited to Coues (1877:136). Although Coues probably intended only to indicate that Baird wrote this name on the labels of two specimens in the mammal collection of the Smithsonian Institution, Coues gave an "indication" of the application of the name by publishing at the same time the catalogue numbers of specimens whose labels bore the name and thus, in accordance with article 21 of the International Rules of Zoölogical Nomenclature, himself becomes the author of the name. Of the two specimens mentioned by Coues, only the first recorded by him, no. 4320 (with skull no. 37995, U. S. Nat. Mus.), can now be found.

Fortunately, the skull of this specimen labeled (see Lyon and Osgood, 1909:218) as taken at Fort Laramie, Wyoming, is well preserved. Its only defects are a fracture in the left zygomatic arch and the absence of parts of each of the first lower molars. In deciding on the subspecific application of the name Putorius culbertsoni Coues, the skull of the type must be principally relied upon, for there is available only one other specimen, a skin only (no. 12596, U. S. Nat. Mus.), from the same place, and it, like the type, is in white winter pelage and lacks flesh measurements.

The ranges as now known of three subspecies of Mustela frenata approach near to Fort Laramie. These are M. f. longicauda, M. f. alleni, and M. f. nevadensis. The skull of the type of culbertsoni is not typical of any one of the three mentioned races. The small size of its teeth and relative (to basilar length) shallowness of the frontal region of the skull through the postorbital processes of the frontal are as in nevadensis. The zygomatic arches are not so greatly expanded as in some specimens of longicauda and are more like the average for nevadensis or alleni, as also is the relatively (to basilar length) long orbitonasal length. However, each of these characters is subject to variation and alone is not surely diagnostic, especially toward the margin of the range of any one of the subspecies concerned. The same may be said of the relatively great breadth of the skull interorbitally—a feature typically found in longicauda. More important, in my estimation, is the large size of the skull; all parts measured (excepting the teeth, the depth at the posterior border of the last upper molars, the zygomatic breadth, and the depth of the tympanic bullae) equal or approach nearest to the average for males of longicauda of similar age.

The small size of alleni prevents its identification with culbertsoni. The question of application lies between nevadensis and longicauda. If the long-tailed weasel at Fort Laramie is found to be referable to the race earlier named longicauda, no change in current nomenclature will be effected. If, on the other hand, the long-tailed weasel from Fort Laramie is found to be referable to nevadensis this name will have to fall before the earlier proposed name culbertsoni. There is, however, a third possibility, namely, that the long-tailed weasel of the Transition and Upper Sonoran zones of southern Wyoming and northern Colorado, as for example, at Lay, Colorado, may represent a recognizable race characterized by size about as in longicauda, relative proportions of skull about as in nevadensis and coloration intermediate, to which the name culbertsoni may apply. For more detailed discussion of this possibility, see remarks under M. f. nevadensis.

Satisfactory application of the name Putorius culbertsoni Coues requires an adequate series of adult specimens, of both sexes in the summer coat with external measurements taken in the flesh, from the type locality and like material from elsewhere in southern Wyoming. On the evidence furnished by the skull of the type of culbertsoni, that name tentatively is placed in the synonomy of longicauda.

Only 2 of 25 adults examined for malformation of the frontal sinuses by parasites showed evidence of disease.

Specimens examined.—Total number, 138, arranged alphabetically by provinces and states and further by districts or counties from north to south except as otherwise indicated. Unless otherwise indicated specimens are in the collection of the United States National Museum.

Alberta. St. Albert, 1; S. Edmonton, 3; Islay, 4[77]; Battle River, south of Camrose, 1[77]; Daysland, 1[77]; Moose Pass, 1; Blindman River, 2 (1[75], 1[2]); Red Deer, 3 (2[2], 1[60]); Bearberry Creek near Sundre, 1[77]; Canad. Nat. Park, N.W. Territory, 1[60]; Red Deer River, Didsbury, 1; Canmore, 1; Calgary, 11 (6[60], 2[1], 1[86]); Red Deer River, 3[2]; Little Sandhill Creek, Red Deer River, 1[77]; Waterton Lake Park, 2[77]; Sweetgrass Hills, 1[77]; Alberta, 1[14].

Colorado. Yuma County: Wray 4 (1[88], 3[74]).