Systematic Accounts of Species and Subspecies (part 2)

On the testimony of a friend who had previously obtained several specimens for him, Swenk (1926:321) records the least weasel from Oshkosh, Garden County, Nebraska, which is a marginal record of occurrence to the southwest for M. r. campestris.

At an early stage in the study of American weasels the writer examined the specimens from Nebraska saved by Mr. Myron H. Swenk and recorded measurements of them. However, at the time of writing this account the specimens were not available for examination and the account of coloration is accordingly incomplete.

The large size, particularly the large external measurements, comprises the principal distinguishing character of this subspecies of the least weasel.

Of the four adults examined from Iowa and South Dakota one exhibits lesions such as result from infestation of the frontal sinuses by nematodes.

Specimens examined.—Total number, 21 as follows. Arranged alphabetically by states and by counties, from north to south in each state. Unless otherwise indicated, specimens are in the United States National Museum.

Iowa. Howard County: Chester, 1. Palo Alto County: Emmetsburg, 1. Kassuth County: Algona, 1. Clayton County: National, 1. Storey County: Nevada, 1. Wapello County: Ottumwa, 1. Henry County: Mount Pleasant, 1.

Nebraska. Holt County: Page, 1. Madison County: Norfolk 1. Cuming County: Beemer, 1. Hamilton County: Chapman, 1. Clay County: Inland to 1 mi. east thereof, 7.

South Dakota. Brown County: shore of Sand Lake, S. 15 T. 126N, R. 62W, 1. Day County: Waubay Migratory Waterfowl Refuge, 1. McCook County: Salem, 1.

Mustela Frenata Lichtenstein

Long-tailed Weasel

(Synonymy under subspecies)

Type.—Mustela frenata Lichtenstein, Darstellung neuer oder wenig bekannter Säugethiere, pl. 42 and corresponding text unpaged. 1832.

Range.—From southern Canada southward over all of the United States, México, Central America, Venezuela, and the republics of western South America to southern Perú and extreme northern Bolivia. All the life-zones from Alpine Arctic to Tropical are inhabited. In the extremely desert region of southeastern California and western Arizona the species is scarce or possibly absent although recovery of a skull (see under account of M. f. neomexicana) from near the center of this region at Potholes on the Colorado River, and a reported occurrence in the mountains of Baja California, México, indicate that a few individuals of the species live in favorable habitat even in this desert region.

Characters for ready recognition.—Differs from Mustela erminea, in regions where the two species occur together, by tail more than 44 per cent of length of head and body and by postglenoidal length of skull less than 46 per cent of condylobasal length in males and less than 48 per cent in females (see under characters of the species); from Mustela rixosa by presence of black pencil on tail, caudal vertebrae more than a fourth (2/5-3/4) of length of head and body, basilar length of skull more than 34 mm.; from Mustela africana by absence of thenar pad on forefoot, underparts without longitudinal, median, abdominal stripe of same color as upper parts, upper lips narrowly (rather than broadly) edged with color of underparts, longest facial vibrissae extending to or behind posterior margin of ear; presence of p2; more inflated (see pls. 23 and 30) tympanic bullae.

Characters of the species.—Size large: Total length 300 to 550 mm.; tail two-fifths to seven-tenths of length of head and body, with distinct black pencil at end; caudal vertebrae 19 to 23; skull with long precranial portion; postglenoidal length, expressed as a percentage of the condylobasal length, less than 47 in females and ordinarily less than 46 in males; upper parts brown; light-colored underparts, in summer pelage, tinged with buffy or yellowish and continuous from chin to inguinal region; some subspecies (southwestern United States, México, Central America, and Florida) with white or yellowish facial markings which do not occur in any other American species of the genus Mustela.

Geographic variation.—Forty-two subspecies are recognized, and the species is geographically more variable than any of the other 3 American species. Color, color-pattern especially on the head, relative proportions of the tail, hind feet, body including the head, and shape and size of the skull are the principal features in which geographic variation has been noted. The variation in the skull extends to the basicranial region (shape and size of tympanic bullae and related structures), interorbital region and preorbital region.

Natural History.—Habitat and Numbers.—As has already been remarked, the long-tailed weasel is absent from the extreme desert of the southwestern United States and northwestern México. Possibly the absence of water to drink is the limiting factor. In southern Nevada the finding of weasels only in places that were well watered, even though small rodents suitable as food for weasels were even more abundant in the surrounding desert, supports this possibility that the absence of water to drink is the limiting factor. Also at Berkeley, California, in early December of 1927 in the canyon at the head of Dwight Way and in the autumn and winter of 1928 in Strawberry Canyon on the campus of the University of California, I trapped extensively for this species in different habitats and obtained, in all, four individuals no one of which was farther than 10 feet from water. The lesser cruising range of the individual weasel than of, say, the coyote, probably explains why, in an arid region, for example Pahranagat Valley, Nevada, only the meadow mice and their riparian associates are preyed upon by the long-tailed weasel whereas the coyote preys upon these riparian rodents and also upon the kangaroo rats and other rodents which are so abundant in adjoining habitats that are devoid of water.

In areas where water is available every few hundred yards, no particular habitat seems to be avoided in summer providing there is food for the long-tailed weasel. In winter (January and March) there obviously was a choice of habitat, possibly occasioned by more abundant food or more satisfactory shelter, or both, in Centre County, Pennsylvania, where Glover (1943B) found the population density in the chestnut-oak habitat to be one weasel per 6.5 acres in areas of tree cuttings and slash and one weasel per 13.3 acres in the open forest. In the scrub oak-pitch pine forest type the population was one weasel per 26.4 acres in tree cuttings and slash and one weasel per 38.2 acres in the open forest. No weasel was found in an area of 9.6 acres comprising a wood lot, the edge of the forest, abandoned fence rows and an abandoned orchard. The two types of forest in which he did find weasels, 25 in all, comprised 381.6 acres. Glover's (op. cit.) data is the only precise information known to me on actual numbers of long-tailed weasels in a given area of any considerable size.

Fluctuations which I elsewhere (1946:57) have designated as multiannual fluctuations occur in this species but seemingly not in the degree that they do in Mustela erminea. This difference between the two species is to be expected because M. frenata does not range so far northward toward the polar regions as does M. erminea and populations of most kinds of animals in the polar, at least in the arctic, regions are subject to more extreme and more regular fluctuations than are kinds of animals in temperate or tropical regions. Indication of the means by which decrease in the weasel population is brought about is afforded by Osgood's (1935:156) observations around Rutland, Vermont. In the late winter of 1934, tracks indicated that weasels left their usual haunts and hunted cross lots, vainly trying to find food. Testing of the small mammal population in the spring and summer of 1934 showed that it was at low ebb. In the fall of 1934 mice and shrews were abundant again but weasels seemed to be entirely absent. The decrease in the population of weasels lagged behind the decrease in the population of the herbivorous prey as did the subsequent increase; this, of course, is the normal relation of carnivorous species of mammals and their prey, at least in and above the Transition Life-zone.

The average distance away from the central den which four weasels (sex unspecified) traveled in a single night at Ames, Iowa, was 312 feet; the maximum distance was 642 feet. These data were obtained in the winter of 1939 by Polderboer, Kuhn and Hendrickson (1941:115) who studied the tracks in the snow. In Manitoba, Criddle and Criddle (1925:143) noted that a female which lived in their basement often wandered more than half a mile away in search of food. In Michigan, Quick (1944:75) found the maximum distance traveled in one day (= night?) by a large male to be 3.43 miles although two miles was the average distance traveled by this individual. In 1942, from January 4 to March 4, in Centre County, Pennsylvania, Glover (1943B) studied tracks of 11 males and 10 females, in newly fallen snow, and ascertained that the distance traveled in a single night averaged 704 (60-2535) feet for the male and 346 (20-1420) feet for the female. The weasels in the open timber traveled farther per trip than those in the brushland and dense stands of trees.

Behavior

An adult female (now the holotype of Mustela frenata nevadensis) seen running across a field, and, I think, unaware of my presence, at every bound bent her back up so far that she reminded me of a measuring worm. For part of the time when running, the tail was held off the ground straight out behind, and then, for a while, inclined upward at an angle of about 45°. Another weasel that I saw in the daytime, and that I think was unaware of my presence, was bounding along among the Baccharis bushes on the south-facing slope of Dwight Way Canyon, Berkeley, California. This individual, at each bound, arched the back up so high as to remind me, again, of a measuring worm.

The long-tailed weasel is a land mammal and unlike its close relative, the mink, is seldom seen in the water. That it can swim, however, is attested by the capture of one while it was swimming across the Río Ramos in México (Davis, 1944:381). Also, Green (1936), in May, in Gratiot County, Michigan, saw a weasel, running with a Peromyscus in its mouth. The weasel dropped the mouse, entered the water and swam to a hole among stones.

More instances of climbing, than of swimming, have been reported in the literature for the long-tailed weasel. Seton (1929 (2):625) quotes William M. Graffius of Pennsylvania as having seen a weasel closely pursue a red squirrel nearly to the topmost branch of a large hemlock. When the squirrel loosed its hold and dropped into a stream, the weasel descended to the ground and caught and killed the squirrel when it emerged from the water. Pearce (1937:483), in central New York State, on July 29, 1931, watched a weasel chase a chipmunk up a black cherry tree ten inches in diameter, and noted that the first rush carried the weasel "straight up the trunk for approximately 10 feet, where it hesitated momentarily before continuing. Then, instead of climbing vertically, it made progress by traveling in short ascending spirals around the trunk, scarcely making 3 feet in height for each circuit of the tree. Upon reaching the limb by which the chipmunk escaped, the weasel followed out along this in the same spiral manner. This limb had a diameter of about 4 inches at its base and extended upward at an angle of perhaps 20 degrees above the horizontal . . . it made its way head first almost down to the ground, using the same spiral mode of progress, but at a leisurely pace. . . . While traveling down the side limb it appeared practically to wrap its sinuous body around the limb."

A male long-tailed weasel, from Colorado, which I kept captive was often fed freshly killed mice. These I thrust through one of the small openings in the wire mesh. The weasel quickly learned to seize any part of a mouse thus introduced and his tugging aided in getting the mouse into the cage. Occasionally a mouse too large to be got through the mesh had to be withdrawn. In such an instance, if the weasel had already had hold of the mouse, he would screech frightfully. I have heard no other vocal sounds from a weasel except a kind of purring.

The sense of smell apparently is well developed; at any rate it is keen enough to allow the weasel to follow the trail of an intended victim by the scent left by the latter. Murie's (1935:321-322) account, for example, of a weasel pursuing a snowshoe rabbit gives clear evidence that the weasel relied on scent in following the rabbit.

A captive male weasel obtained at Gainesville, Florida, stamped his hind feet when annoyed (Moore, 1945:259).

A male from Colorado that I kept for months in a cage at Lafayette, California, was several times found in a sleep so deep that he was awakened with difficulty. Seton (1929 (2):629-630) writes: "In my small menagerie, I have had half-a-dozen Weasels of the New York species. Their sleeping dens are arranged so as to be easily and silently opened. Several times I have lifted the lid to find the weasel in a deep sleep—a sleep so profound that I had to poke him vigorously with a stick before he awoke, looked up, and rushed forth with a little puff of wrath, and a little puff of smell."

Feces and urine were ordinarily deposited in one particular place by each of the captive weasels that I have observed. Hamilton (1933:294) records that a large male M. f. noveboracensis, in a week, averaged 10 evacuations every twenty-four hours, that urination immediately precedes defecation, and describes the feces as black or brown, long and narrow and often spiral-shaped owing "to the matted fur of some rodent that had been eaten." Quick (1944:77) writes, concerning four winter dens in Michigan, that "The latrines of weasels were in the entries of used dens and scats could be collected there by the handful." Polderboer, Kuhn and Hendrickson (1941:116) in the spring of 1939 at Ames, Iowa, gathered scats "from latrines found at the entrances of burrows and from latrine chambers found within burrows." Scats were found by them in the linings of some nests.

Courage of a high order might be credited to the long-tailed weasel because individuals have attacked animals much larger than the weasels. Actually, however, in few if any of these instances was the motive for attack known. That a hawk was attacked is suggested by Soper's (1919:45) account of Mustela frenata noveboracensis wherein he repeats a story told to him of a hawk observed in unsteady flight, and obviously in distress, which when it plummeted to earth was with a weasel which escaped from the observer. Charles Tatham, Jr., of Cambridge, Massachusetts, according to Seton (1929 (2): 630, 631) observed one that attacked his dog.

Persons and long-tailed weasels have figured in some rather strange encounters. For example, Oehler (1944:198) recounts that in the autumn of 1940 at Cincinnati, Ohio, an animal, mistakenly thought to be a chipmunk, was seen to dash into a hollow log whereupon pounding on the log brought out the weasel which bit and clung to the hand of one man whose companion was bitten when he attempted to free the man that was bitten first.

Seton (1929 (2): 631) writes that on the night of September 5, 1897, on Roosevelt's old ranch, near Medora, North Dakota, a man turned over his saddle (which was lying on the ground) to dislodge what was thought to be a pack-rat. The animal was a long-tailed weasel which attacked him. It ran up his legs a number of times aiming at his throat before being killed by a dog.

Criddle and Criddle (1925:146) wrote: "August 20, 1919.—A longicauda in the Insectary ran at me this morning apparently with a view to intimidating. It uttered a shrill cry while making the attack, but retreated after advancing within two feet." The same authors (op. cit.: 147) further write that a "Long-tailed Weasel was caught in a trap set for gophers, and, on being released by Miss M. Criddle, at once turned upon its liberator and bit savagely at her boot. It then moved a short distance away to a tub of water, where it drank thirstily, merely glancing at the observer from time to time while doing so, and then ran off out of sight.

"Mr. T. Criddle records a similar experience. After liberating a large weasel from a trap, it immediately rushed at him and persisted in its attack with such ferocity that it was three times picked up and thrown, on each occasion to a greater distance, before it finally abandoned its offensive.

"We have no record of a weasel making an unprovoked attack upon anyone."

Wight (1932: 164) in Michigan, detected a weasel attacking a hen. The weasel fled at Wight's approach but returned and attacked him several times. Finally the weasel went around Wight to reach the hen. In Wight's words "There was no evidence of infuriation, but rather a well directed offense at the one object, regardless of its size, which stood between the weasel and an opportunity to satisfy its desire to kill, which was probably based upon the uncontrollable urge of hunger pangs."

Weasels of each of the three North American species have been successfully kept in captivity. A type of cage satisfactory for keeping the animals in the laboratory is described by Bissonnette and Bailey (1940:761-763). Some of the captives used their teeth to break glass water-containers and to gnaw slivers of wood from the cages. Ingested slivers of wood and bits of broken glass caused the deaths of some of the captives. Weasels kept by me all were of the species Mustela frenata. They thrived on a meat diet but I was always careful to give them, every few days, if not each day, some small rodents entire, thinking that the bits of bone and fur ingested might, in some way unknown to me, keep the digestive tract in better condition than would flesh devoid of hair and bone.

Three young weasels approximately the size of mice, in the Okefinokee Swamp of Georgia, were obtained by a hunter who, according to Harper (1927:303), raised them by feeding "milk for a few days, and then fresh meat." Litters of young born in captivity have been successfully raised by the mothers (Hamilton, 1933) and success in getting the animals to breed in captivity and to rear their young is recorded by Wright (1948A). He has found, however, that the majority of his captive adult males show no interest in mating when placed with females in heat. He, therefore, uses only selected males and when a female in heat is to be bred, he places one of his responsive males with her one day, another of his responsive males with her the second day and thus alternates a couple of males for three or four days. Even so, slightly fewer than half of the females which were thus bred produced young.

A weasel in the white winter coat was used by Audubon and Bachman (1856:177, Quarto edit.) to drive rabbits out of their burrows in the same fashion that ferrets commonly are used. Although these naturalists refer to their animal as an ermine it probably was Mustela frenata noveboracensis, the long-tailed weasel. The animal's teeth (probably canines) were blunted and a long cord tied on its neck. With the aid of this weasel 12 rabbits were caught in one morning and more than 50 in four weeks.

Enemies

Little is recorded concerning enemies of weasels and it may be that other vertebrates are not an important factor in removing the annual increase. Errington (1935:195-198), in Iowa, found four, putrid weasels about dens of red foxes, Vulpes fulvus. No remains of weasels were found in the feces of the foxes and it appears that the foxes do not eat the weasels. The label on an adult female specimen of M. f. spadix from Boone County, Iowa, bears the date May 10, 1938, and the annotation, by T. G. Scott, "fox-killed." Bailey (1931:328) recounts that "Weller saw a coyote carrying one in its mouth" at an elevation of 11,500 feet in the Pecos Mountains of New Mexico. The type specimen, a young female, of M. f. peninsulae from Hudsons, Florida, according to Rhoads (1894:155) ". . . was caught in the woods by a cat." Barber and Cockerell (1898:189) mention one that was killed by a dog in Mesilla Park, New Mexico. Moore (1945:258) records the death of a weasel in Florida. Circumstantial evidence indicated that it was killed by the bite of a water moccasin. In the Biological Surveys Collection of mammals in the United States National Museum, the label with the skull of an adult male weasel, No. 160663, from Banning, California, carries the information that the skull was taken from the stomach of a Crotalus (rattlesnake).

In reporting on a study of owl predation in Delaware County, Pennsylvania, Pearson and Pearson (1947:143) mention that "weasels are found throughout the county but . . . were never eaten by the owls."

The Uinta spermophile at some places and times probably is a prey sought by the long-tailed weasel but Warren (1924:265) records Citellus armatus repeatedly chasing weasels in August, at Camp Roosevelt, Yellowstone National Park, and how the ground squirrels at one time ignored the weasel even when it came within a few inches of a squirrel.

Warren (1932:71), on August 2, 1931, at Grand Mesa, Colorado, obtained a large male weasel with two porcupine quills in it; one was near the mouth and another "in the skull." Osgood (1935:156) writes that near Rutland, Vermont, a male weasel "taken in April, was heavily parasitized and had several short porcupine quills embedded in its neck, head, and shoulders." The remainder of Osgood's account implies that the weasel may have turned to porcupine because the normal food for weasels was scarce at the time. Porcupine quills, then, are a hazard for weasels although it is unlikely that the porcupine is ever to be classed as an enemy of the weasel.

An accident of another sort, which must at the very least have been annoying to the weasel that suffered it, was recorded by Soper (1921:37). The animal had a stick lodged crosswise between the fourth upper premolar teeth.

The recorded actions of several kinds of animals which are too small to be dangerous to the weasel suggest that they recognize that the weasel is a danger to them. Borell and Ellis (1934:21) mention that a weasel in Nevada caused a great disturbance among the chipmunks. Long (1938:250) heard pikas give evidence of terror by a peculiar cry when a weasel was in a rock slide occupied by the pikas. Seton (1929 (2):629) writes "On June 14, 1915, as I prowled around the south side of the lake on my homeland at Greenwich, Conn., my attention was called to a pair of song sparrows and a male towhee that were noisily mobbing a Weasel, twittering around and darting at him, as though they knew full well his evil ways. The weasel paid little heed, but soon dived from sight in a stone wall."

No account has been found of an American weasel or ermine rolling, tumbling and frolicking in a manner that aroused the curiosity of birds to a degree which permitted the weasel to come within leaping distance of the birds. Accounts of such behavior are on record for the English stoat (ermine).

Food and Hunting

Weasels are active both in the daytime and at night. Whether the time of activity varies with the season, with the locality, with the sex or with other conditions, I do not know. Adult, live, free-living, actively moving weasels that I recall having seen all were observed in the daytime: two were in Alameda County, California, two were in White Pine County, Nevada, one was in Scotts Bluff County, Nebraska, and one was in Laramie County, Wyoming. I recall ten adults, from the same three states, and one from Washington State, that got into my traps; two of these certainly got in the traps in the night; one certainly got in the trap in the daytime; the other eight were found in traps which may have caught the weasels either in the night or in the daytime. Soper (1946:136) in speaking of M. f. longicauda north of the International Boundary in Canada remarks that it has the "habit to some extent of hunting at all times of day." Criddle and Criddle (1925:144) in writing of Mustela frenata longicauda in Manitoba record that "The shrill cry of a rabbit [Lepus americanus] in the dark is nearly always due to the weasel's attack. Indeed, we have often watched the latter at work during the twilight hours. First would come the almost noiseless run of the small rabbit with its characteristic dodging and this would be followed by the appearance of the agile foe which, at times, would leap high over obstacles and at others move swiftly beneath them. Then there would follow intermittent cries of the rabbit as the weasel secured a temporary hold of its quarry, for be it noted that this hunter apparently bites anywhere to begin with and it is probable that the blood made to flow acts as an aid to tracking as well as weakening the prey. Several similar close encounters might occur before the rabbit would be finally overcome, but weasels are very persistent when they once get into contact with their victims and it is therefore very seldom that the latter escape. In killing, they either penetrate the brain with their teeth, or dislodge the vertebrae behind the head." These and more than two score other observations which record the time when weasels were seen make it clear that some were active at night and that some were active in the daytime.

As to the routes traveled while the weasels are hunting, Quick (1944:77) says of four individuals that he studied in Washtenaw County, Michigan: "The weasels appeared to prefer hunting certain coverts with noticeable regularity, but rarely cruised the same area on two consecutive nights."

The killing technique of fifteen captive Mustela frenata noveboracensis was studied by Glover (1943A). For the weasels he released 19 mice, 3 brown rats, 6 cottontails and 4 ring-necked pheasants. Most of the mice were killed by a bite on the back of the head, with the body and legs of the weasel hugging the back of the victim. "The weasel shoved the prey in close to the stomach with the hind legs, and the kill was made in a reclining semi-curled-up position." On each of the rats (Rattus) an initial grip was secured at the base of the ear. When the rat rested, a new hold was taken by the weasel. Finally the weasel secured a hold at the base of the skull and near the ear, and a light crushing sound followed. Four of the six cottontails were killed by bites on top of the head and ear; two cottontails succumbed from neck wounds. In three instances, neither of two weasels could be induced to make a determined attack on the cottontails or to kill them. At times the cottontails proved to be able opponents for weasels by striking out with their front feet and by kicking with their strong hind legs. In killing the pheasants the teeth of the upper jaw of the weasel pierced the top of the braincase and the teeth of the lower jaw entered the region of the auditory process. The forelegs hugged the neck of the pheasant, the body of the weasel was extended in a riding position on the back of the bird and no amount of kicking or rolling dislodged the weasel.

Polderboer, Kuhn and Hendrickson (1941) describe a cottontail cached by a weasel as having the muscles of the neck severed from the region behind the right mastoid process and noted "that hemorrhage in the region of the right jugular vein had occurred."

Concerning the methods of killing mammals smaller than cottontails, the accounts by Nichols and Nichols (1935:297-299) and that by Svihla (1931) corroborate Glover's (1943A) account, as do also the accounts of Miller (1931B:164) and Moore (1945:257). The latter says that his captive male, from Gainesville, Florida, customarily bit its rodent prey at the base of the skull and used the feet to manipulate the live prey. Miller (loc. cit.) emphasized that his male weasel (M. f. longicauda) grasped where it could, used its snakelike body to coil over the prey and shifted the grip of its teeth to the nape of the neck or back of the skull. The captives that I have had [one from Salt Lake City, Utah; three from Contra Costa County, California; and the same individual reported upon by Miller (1931:150)] customarily employed the techniques of killing small rodents that were described by Glover and Miller (loc. cit.).

Allen (1938:225-229) experimented with the ability of four different males of M. f. noveboracensis from Michigan to kill adult cottontails. The method used was to place the weasel in a cage of quarter-inch hardware cloth approximately three feet long, two feet wide, and two feet high. The bottom of the box was covered with several inches of straw. One cottontail was offered to each weasel. In two instances the weasel attacked and bit the cottontail, was struck by the hind feet of the cottontail, retired from the attack and died a few hours later as a result of the blows of the cottontail's hind feet. In the other two instances the weasel rendered the cottontail helpless by severing the neck muscles from the skull. Subsequently an incision made by the weasel, in each of the two instances gave access to blood on which the weasel fed until it was full, in one instance by licking "blood as a cat laps milk." One rabbit was subdued in 10 minutes and the other in 15 minutes. Allen (op. cit.) points out that cottontails form a considerable portion of the weasel's food and thinks that they are killed in burrows more easily than they were in the cage.

In writing of the three species of weasels, including Mustela frenata, found at Treesbank and vicinity, Manitoba, Norman Criddle and Stuart Criddle (1925:143, 144), in my opinion, correctly explain the killing of more prey than weasels need. "The fact that weasels frequently kill many more animals than they require for immediate use has been universally interpreted as a lust for killing—a supposition which we believe to be quite erroneous. It is true that weasels often kill more than they need, but the surplus is not necessarily wasted because the animals always store it for future use, in much the same way as do badgers, minks or skunks, and with the same object in view as squirrels have in gathering nuts. We have observed many such stores, but as far as our observations go, the habit of killing in excess occurs much more prominently in the late summer and autumn months than in the spring. Indeed, we have no records of excessive spring slaughter and this indicates that the supposedly blood-thirsty habit of weasels is no more a lust for killing than is the woodsman's foresight in providing his larder with meat for the winter months. It should be noted in this connection that members of the weasel family, when undisturbed, do not leave their victims scattered about, but carefully store them away, and in many instances the bodies are buried with earth or taken under ground to preserve them. We suspect that this instinct for preserving food for future use accounts for most of the excessive killing by carnivorous animals instead of this latter indicating an aimless desire for slaughter which would unnecessarily deplete the food supply of the future. This instinct, however, does not seem to be as definite as that of some rodents, and there is no doubt that much of the stored meat decays before it can be utilized."

Criddle and Criddle (1925:146) note that a weasel in the vicinity of Treesbank was carrying a rat [Rattus] and that "Two small punctures in the throat were the only evidence of the manner in which its death had been brought about."

Considerable information has been recorded concerning the food of Mustela frenata and a little information is on record as to kinds of foods not taken that could have been taken. For example, Ingles (1939:253, 254) on May 14, 1938, near Shasta City, California, noted that nestlings of russet-backed thrushes were ignored by an adult weasel and four young weasels which were feeding instead on meadow mice and a mole. Howard (1935:322, 323) records that a weasel in Michigan which carried bits of meat from beef bones on a porch ignored a red squirrel which drew on the same food supply but which retreated to the end of the porch when the weasel appeared. Quick (1944) records that in the winter of 1940 on a 640 acre area in Washtenaw County, Michigan, four resident weasels did not kill any of the 10 rabbits or several pheasants but subsisted on smaller animals. Glover (1943A) thought that M. frenata kills only a few adult cottontails in the wild. To judge from these observations, M. frenata chooses small mammals as prey in greater measure than it does birds or larger mammals.

Records of prey taken, attacked or pursued by Mustela frenata include the following:

Broad-footed mole (Scapanus latimanus).—One was fed on by an adult M. frenata and four young, on May 14, 1939, "near Shasta City," California (Ingles, 1939:253, 254).

Dusky shrew (Sorex cinereus).—A female weasel, at Majestic, Long Island, N. Y., was shot when carrying a Sorex cinereus that had a small hole in the top of its head (Nichols and Nichols, 1935:297-299).

Big short-tailed shrew (Blarina brevicauda).—One was taken from the stomach of a weasel (Hamilton, 1928:249).

Townsend ground squirrel (Citellus townsendii).—Alcorn saw a weasel five miles west of Fallon, Nevada, carrying a squirrel (Hall, 1946:192).

Richardson ground squirrel (Citellus richardsonii).—The attempted capture of one of these squirrels in Saskatchewan is recorded by Seton (1929 (2):625).

Belding ground squirrel (Citellus beldingi).—Grinnell, Dixon and Linsdale (1937:233) recount that at Tuolumne Meadows, California, a weasel killed a ground squirrel of this species.

Thirteen-lined ground squirrel (Citellus tridecemlineatus).—Errington (1936:406, 407) found a den in Palo Alto County, Iowa, on June 22, 1934, where he collected 32 fecal pellets. Sixteen samples contained thirteen-lined ground squirrels, 9 contained rabbits, 9 contained mice (7 Microtus, 1 Peromyscus and 1 unidentified); red-winged blackbirds and unidentified fringillids were represented as also were ground beetles, grasshoppers and other insects. One red-winged blackbird lay near the entrance of the den.

Franklin ground squirrel (Citellus franklinii).—Sowls (1948:126) records that at Delta, Manitoba, a weasel was observed killing one of these squirrels and that "the weasel had taken the squirrel from its hibernating burrow as evidenced by tracks in the snow." On July 19, 1917, in the vicinity of Treesbank, Manitoba, T. Criddle saw a weasel attacking one of these ground squirrels which was in mortal terror and squeaking continuously. Eventually the squirrel was thrown on its back "and would have been speedily killed but for an interruption" (Criddle and Criddle, 1925:146).

Golden-mantled ground squirrel (Citellus lateralis).—On August 15, 1941, along the Kaweah River in Sequoia National Park, Boyer (1943:99, 100) saw a weasel chasing a Citellus lateralis; three or four times the weasel grasped the back of the neck of the squirrel which each time threw off the weasel until the two, weasel after the squirrel, plunged into the river. The squirrel, bleeding at the base of the skull, was rescued and entered a hole; the weasel got out of the water and under a rotting log. Follett (1937:365) at 2 p.m. in Plumas County, California, saw a weasel have hold of the lower jaw of a golden-mantled ground squirrel near its throat. Alcorn watched a weasel chase a golden-mantled ground squirrel in Nevada (Hall, 1946:192) and Grinnell and Dixon (1919:681) record that on August 4, 1911, near Monache Meadows in eastern Tulare County, California, a weasel pursued, captured and killed a golden-mantled ground squirrel.

Eastern chipmunk (Tamias striatus).—Pearce (1937:483) in central New York State, on July 29, 1931, saw a chipmunk scamper up a tree pursued by a weasel.

Chipmunk (subgenus Neotamias).—Stanford (1931:363) on November 11, 1931, at Fish Lake, Utah, saw a weasel pursuing a chipmunk. On August 5, 1910, "near Independence Lake," Nevada County, California, Louise Kellogg recorded that a weasel seized and ran off with a chipmunk (Grinnell, Dixon and Linsdale, 1937:233). Allen (1938:228) observed that a chipmunk (whether Tamias striatus or T. minimus not specified) was killed in 30 seconds whereas 10 to 15 minutes were required by the caged, male Mustela frenata noveboracensis to kill a cottontail.

Red squirrel (Tamiasciurus).—Seton (1929 (2):625) records the capture of one in Pennsylvania, and Grinnell, Dixon and Linsdale (1937:232), at Cisco, California, saw one closely pursued by a weasel.

Flying squirrel (Glaucomys).—Burroughs (1900:77, 78) records remains of one of these squirrels along with the remains of other animals in a food cache of a Mustela but his account does not make clear whether Mustela frenata or Mustela erminea was the species of weasel involved.

Northern pocket gopher (Thomomys).—In "July, 1939, near Stillwater [Nevada], Alcorn pursued . . . [a] weasel and caused it to drop . . . a pocket gopher [Thomomys bottae] which was about two-thirds grown" (Hall, 1946:192). Grinnell, Dixon and Linsdale (1937:233) write that "at least twice, weasels in the [Yosemite] Valley were seen carrying pocket gophers." Relative to Thomomys talpoides in the vicinity of Treesbank, Manitoba, Criddle and Criddle (1925:146) record that on September 11, 1918, an individual of Mustela frenata longicauda took seven pocket gophers dead. . . . It seized the rodents by the middle of their back and held them high while carrying them away. They were stored in a gopher burrow some two hundred yards distant. On February 17, 1921, "Came across the marks of a weasel carting some object over the snow. An investigation revealed a recently-killed pocket gopher with its captor still in possession." Criddle (1930:279), at Aweme, Manitoba, "frequently observed this weasel [M. f. longicauda] . . . carrying a pocket gopher to its larder, and twice it has been encountered in mid winter with freshly killed gophers in its possession." The evidence already presented that weasels levy heavily on pocket gophers is strengthened by the many references in the literature to weasels having been caught in traps set for pocket gophers in the burrows of those rodents and by the many statements, not quoted here, that living quarters of weasels are in burrows made originally by pocket gophers. For example, the present writer, in an account of the Mammals of Nevada (Hall, 1946:191, 192), has said of the long-tailed weasel, Mustela frenata nevadensis, that "All the three dens that were excavated . . . were originally burrows of pocket gophers. . . . Although we have found weasels in many situations in Nevada, . . . they most often were obtained from the burrows of pocket gophers." Excluding the weasels taken by Alcorn, more specimens of the remaining lot were caught in traps set in the burrows of pocket gophers than by all other means combined. All of the 22 weasels taken by Alcorn [within a radius of 10 miles of Fallon] were obtained in gopher traps.

Mexican pocket gopher (Cratogeomys).—At Chalchicomula, 8000 feet, Puebla, Nelson (1918:470 and letter dated March 9, 1928) saw a weasel fastened to a pocket gopher. Nelson obtained the pocket gopher and found that its neck muscles were torn loose from the skull.

Grasshopper mouse (Onychomys).—Barber and Cockerell (1898:189) found remains of this mouse in the stomach of a weasel at Mesilla Park, New Mexico.

White-footed mice (Peromyscus).—Green (1936) saw a weasel in Gratiot County, Michigan, in May, carrying a Peromyscus. Quick (1944:76), in winter, in Michigan, found one dead, probably killed by a weasel. From Washtenaw County, Michigan, Quick (1944:77) examined 294 scats of free-living weasels and found Peromyscus in 189 scats, Microtus in 83, small birds in 20, red squirrel in 3, and hair of weasels in small quantities (probably from the animals which deposited the scats) in 36. He concludes (op. cit., 78) that the winter food was 65 to 70 per cent Peromyscus, 23 to 33 per cent Microtus, and 2 to 7 per cent small birds.

Wood rats (Neotoma).—A female long-tailed weasel weighing 250 grams was taken one mile north of Kent, Texas, while eating a Neotoma albigula (Davis and Robertson, 1944:263). A wood rat house under observation by Vestal (1937:364) in Contra Costa County, California, was invaded by one weasel which ate two adult wood rats (Neotoma fuscipes) and one young. In the same area he saw a weasel in a wood rat nest some months later (Vestal, 1938:5). Three miles east of Reno, Nevada, on May 13, 1936, W. B. Richardson watched a long-tailed weasel carrying a half-grown round-tailed wood rat (Neotoma lepida) across a rock slide (Hall, 1946, 192). Harper (1927:303) records three wood rats [Neotoma floridana] and two cotton rats [Sigmodon hispidus] found dead in the den of a female weasel and her three young in the Okefinokee Swamp of Georgia. Another female and three young approximately half grown were found in the swamp in a hollow pine log. Contents of the den as described to Harper were nearly a peck of wood rats, whole and in pieces; remains of several kinds of birds including robins and quail, and a piece of joint snake (Ophisaurus ventralis).

Meadow mice (Microtus).—Polderboer, Kuhn and Hendrickson (1941), in 1939, at Ames, Iowa, identified "A total of 118 items . . . in 97 winter scats and 48 in the 38 spring scats." Their combined data are as follows:

Frequency Percentage

Meadow mouse 71 42.85

Harvest mouse 36 21.75

Deer mouse 17 10.23

Mearns cottontail 14 8.42

Short-tailed shrew 9 5.42

House mouse 3 1.86

Tree sparrow 2 1.02

Grasshopper 1 .60

Shaw pocket gopher 1 .60

Least weasel 9 5.40

Unidentified material 3 1.85

Polderboer, Kuhn and Hendrickson divide their data into two categories, winter and spring. Items recorded in winter but not in spring are house mouse, tree sparrow, and grasshopper. Items recorded only in spring were pocket gopher and least weasel. The samples of cottontail and least weasel all were from the scats of one large male weasel. Of a total of 14 pheasants, 24 quail and 35 cottontails on the 160 acres involved in the study only two cottontails appear to have been killed by the weasels—really by one weasel of four which lived on the area.

Food items taken from the nests (3) and adjacent caches of food in the dens, were as follows: meadow mouse, 30; short-tailed shrew, 4; pocket gopher, 2; deer mouse, 2; least weasel, 1; tree sparrow, 1. The authors remark that the abundance of several prey species does not cause the weasels to ignore the shrews which are said to be distasteful to carnivores.

Two horned larks, apparently killed by weasels, were found on the 160 acre area studied; the horned larks were not in caches of food, nor were remains of horned larks found in scats.

Dearborn (1932:34, 37) for Michigan, on the basis of contents of (37?) intestinal tracts and "feces collected partly in winter and partly in summer" found that, by frequency of occurrence, mammals comprised 83 per cent of the food, birds 10 per cent and insects 7 per cent. Frequency indices for the genera of mammals in percentages of food items of all kinds were as follows: Microtus, 31 per cent; Peromyscus, 24 per cent; Sylvilagus, 14 per cent; Sorex, 7 per cent; Blarina, 5 per cent; Scalopus, 2 per cent.

Criddle and Criddle (1925:146), for the vicinity of Treesbank, Manitoba, record that on October 3, 1913, a weasel was seen to take a field mouse down a hole. They add (op. cit.:147) that "Once while ploughing, we observed a Long-tailed Weasel carrying a field mouse. . . ." Ingles (1939:253, 254), in June, 1938, near Mt. Shasta City, California, found an adult and four young weasels which fed on several Microtus montanus montanus. Green (1936) in May, in Gratiot County, Michigan, in the vicinity of a nest in which there were four young weasels, found "several" dead Microtus. Hamilton (1933:330) records that in New York State a male weasel, on April 5, 1932, at Ithaca, had eaten a Microtus and that in May, 1927, a female weasel was seen carrying a Microtus in its mouth.

Hamilton's (1933:333) study of the contents of the digestive tracts of bodies of weasels obtained from fur trappers and fur buyers enabled him to publish the following "Frequency Indices of Mammal Genera in Fall and Winter Food of 163 Mustela noveboracensis": Microtus, 33.6 per cent; Sylvilagus, 17.3; mammals undetermined to genus but principally mice, 17.1; Peromyscus, 11.3; Rattus, 9.1; Blarina, 5.9; Sciurus, 2.7; Tamias, 1.0; Condylura, 0.8; Ondatra, 0.8.

Grinnell, Dixon and Linsdale (1937:233, 234) quote W. Fry concerning a weasel which reared six young at Giant Forest, California, in 1919, as follows: "This parent weasel, after the birth of her young, remained at the premises for a period of thirty-seven days; during which time, from actual count, the following numbers of mammal species fell victim to her: mice [genera not specified] 78; gophers 27; moles 2; chipmunks 34; wood rats 3; ground squirrels 4. This is a total of 148 animals for the . . . thirty-seven days . . . not a bird was captured during the period."

Rats (Rattus).—Criddle and Criddle (1925:146), on the farm at Treesbank, Manitoba, record a long-tailed weasel, on July 2, 1918, running away from the farm buildings carrying a rat; July 11, 1919, "Two longicaudas . . . have been seen running off with rats on several occasions."; July 11, 1920, "There are two large weasels about the buildings[;]. . . . Each has been noted with rats and this afternoon one of them was seen running into the woods carrying a rat, followed by two excited swallows." The authors (op. cit.:147) add "In the fall of 1924, Mr. A. Cooper, a prominent poultryman of Treesbank, observed a large weasel carrying a freshly killed rat which it stored below ground and then returned towards the poultry-house, causing no little apprehension to the owner. Within a short time, however, the weasel reappeared with another rat which it hid as before. In this way several rodents were accounted for during the afternoon, and Mr. Cooper assures us that the weasel 'kept up the good work for some days'." Hamilton (1933:330) in New York State in May, 1927, saw a male weasel in possession of a rat.

Big jumping mouse (Zapus major).—In the Warner Mountains of California, on Parker Creek, H. C. Bryant frightened a weasel that dropped a freshly killed jumping mouse (Grinnell, Dixon and Linsdale, 1937:232).

Snowshoe rabbit (Lepus americanus).—Adolph Murie (1935:321-322) writes that: "Four miles north of Funkley, Minnesota, early on the morning of November 13, 1921, . . . watched from the top of a 30-foot spruce a weasel. .. hunting a varying hare. . . . The ground was covered with six inches of fresh snow . . . both animals . . . [had] their [white] winter pelage.

"My attention was first attracted to the hare as it came hopping steadily but unhurriedly from the north. Directly in front of me, about 75 feet from the tree I had climbed, the hare crisscrossed back and forth at various angles over an open area about 20 feet in diameter. After producing a maze of tracks, the hare 'froze' near one edge of the pattern. In a few minutes the weasel appeared, all his faculties focused on the warm trail. Expertly he followed its convolutions, passing at times within a few feet of the watching hare. Not until the weasel had followed every turn of the trail to within three feet of its termination did the hare skip off. It came out to the road almost directly below me, turned at right angles northward and was soon out of sight. At the road the weasel lost the trail, . . . and then ran parallel with it, once more in hot pursuit.

"Ten minutes later the hare emerged from the north as before, came on directly to the tracked-up area, and continuing its stratagem, leisurely hopped about to leave its zigzag trail. Then it sat down quietly to wait. . . . The weasel['s] . . . nose led him through the network with little trouble. He was almost upon the hare before it jumped off and followed the same path [as] . . . before. . . .

"The hare had to show his big heels [a third time] . . . as the weasel approached him. This time the weasel failed to follow. . . . After examining a few brush heaps he vanished into the woods behind me."

Seton (1929 (4):723, 724) writes that in December of 1886 in the sandhills northeast of Carberry, Manitoba, he saw a weasel chasing a snowshoe rabbit which took refuge near his feet under the sleigh and so escaped the weasel. Thurber (1940:356) mentions a month-old varying hare that was rescued from a weasel and of approximately the same size as the weasel.

Criddle and Criddle (1925:146) for the vicinity of Treesbank, Manitoba, record "August 21, 1921.—Heard cries of a small rabbit at dusk to-night, which investigation showed was being attacked by a large weasel. The rabbit was later carried to the weasel's store chamber below ground." They record further (op. cit., 146, 147): "November 8, 1924.—Shot a bush rabbit and left it lying. Two hours later [it] . . . was found to have been dragged beneath a brush pile and partly eaten. Innumerable weasel tracks left no doubt as to the identity of the thief." In describing a weasel that wintered in a nest in a threshing machine, the same authors (op. cit.:143) say that no bird remains were found in the pile of approximately three pounds of droppings adjacent to the nest. In a store chamber some 140 yards away from the nest, two bush rabbits (Lepus americanus) had been dragged to the entrance and numerous smaller rodents were taken below ground. The rabbits were buried beneath the snow and eaten as necessity arose. Narrow selectivity on the part of the weasel in choosing food is almost always shown in instances where the food of weasels has been studied. For example, the weasel which lived in the threshing machine ate rodents and rabbits and not poultry although the weasel had ready access to the poultry building. The weasel which lived in the bag of feathers in the basement of Stuart Criddle's house ignored grouse, approximately 20 in number, in favor of other non-avian food.

Cottontail (Sylvilagus).—Polderboer, Kuhn and Hendrickson (1941) mention that one of 4 weasels which they studied on a 160 acre area at Ames, Iowa, in 1939, had a cache of food in a pocket gopher burrow 10 rods distant from the weasel's den. The cache contained only two cottontails, one partly eaten. Leopold (1937) records seeing a Mustela (probably a long-tailed weasel but possibly an ermine) kill a third-grown cottontail by biting it at the base of the skull. Leopold describes the blood sucking or licking, suggesting that he shared the popular misconception that weasels suck blood. The supposition that weasels suck blood has been refuted by many observers, for example by Svihla (1931). My own observation of captives makes me think that weasels do not suck blood. Seton (1929 (2):626) quotes B. H. Warren as seeing a weasel dragging a freshly killed, still warm, rabbit that contained nine embryos almost ready for birth. A young rabbit was seen being carried by a weasel in Hidalgo County, Texas, in March, 1935 (Mulaik, 1938:104). An instance of a cottontail being chased in June in South Carolina is recorded by Hamilton (1933:330). Addy (1939:372, 373), in Virginia, on August 14, 1939, shot a large weasel which was pursuing a Sylvilagus that was only a foot and a half ahead of the weasel. The rabbit stopped when a shot was fired and permitted itself to be stroked and petted. Tracking showed that the weasel had chased the rabbit for a half mile. On November 20, 1942, at Lake James, Indiana, a weasel was seen by Grosjean (1942:443) attacking a "young rabbit" in the throat of which the weasel had made five large holes from which there was no obvious bleeding. Seton (1929 (4):798) recounts that in 1910 at Base Lake, Michigan, F. C. Hicks saw a cottontail with a weasel hanging to its legs rush to the cottage. When only four feet from Hicks the weasel loosed its hold and the cottontail escaped under the cottage. Burroughs (1939:253) on May 14, 1939, in Saginaw County, Michigan, records that a young cottontail weighing between 200 and 250 grams was carried from the nest and killed. Burroughs was attracted first by the "hissing scream" of the weasel, strode toward the sound, flushed an adult cottontail, and discovered the empty nest from which the weasel had taken the young cottontail.

Brush rabbit (Sylvilagus bachmani).—Vestal (1937:364) in Contra Costa County, California, found a brush rabbit that appeared to have been killed by a weasel.

Reptiles.—Grinnell, Dixon and Linsdale (1937:234) recount that in July, 1889, in Wilson Canyon, near Pasadena, California, a weasel killed a red racer by severing the backbone of the snake. In April, 1935, in Hidalgo County, Texas, a half grown bull snake (Pituophis sayi sayi) was regurgitated by a young weasel. Russell (1930:504, 505) has recorded finding in California a male weasel and a king snake (Lampropeltis getulus boylii) three feet five inches long in mortal combat. The weasel killed the snake but the weasel, incapacitated by the conflict, was easily picked up by hand and was also saved as a specimen.

Wild birds.—In the spring of 1940, in Washtenaw County, Michigan, one bobwhite, of 10 bobwhite living on a 640 acre area, was killed by one of four weasels that lived on the area. No other quail was killed there. The one unfortunate bird was killed in the mouth of an abandoned den where the quail roosted (Quick, 1944:76). A male weasel, subspecies M. f. effera, was seen by Booth (1946:439) attempting to enter the nesting hole of a pair of flickers, Colaptes. One song sparrow (Melospiza melodia), and one slate-colored junco (Junco hyemalis) were recorded by Quick (1944:76) as killed by weasel in Michigan.

Chicken (genus Gallus).—Quick (1944:78) writes that in one year (1938-1939) weasels were reported to have killed 1.03 per cent of all chickens in one township of Washtenaw County, Michigan, and that of the total damage to all kinds of poultry, 59 per cent was done by weasels. Weasels entered 19 per cent of the chicken coops on the study area. Farmers killed 68 per cent of the weasels seen in barn yards. Spring and summer were the seasons in which most of the weasels were observed in barn yards. Internal evidence in Quick's (op. cit.) account leads me to suspect that some losses of poultry were charged to weasels when Rattus was actually responsible.

Criddle and Criddle (1925:146), quote a neighbor in the vicinity of Treesbank, Manitoba, as recording that on October 29, 1917, "A weasel last night made its way into our fowl-house, the door being inadvertently left open. The weasel killed eleven fowl, some of which were dragged into the yard. All the largest fowls were selected, the pullets remaining untouched though they were in the majority. Next night the weasel dug a hole beneath the building and killed a hen and two cocks, returning for another during the day, making a total of fourteen in all." Criddle and Criddle (1925:146) remark that the weasel proved to be a large one, probably an old male. The same authors (op. cit.:147) record that at their farm at Treesbank, Manitoba, on January 31, 1925, "A Long-tailed Weasel killed three hens last night, and rather severely bit a cock about the neck. This, or another weasel, had been around the farm-yard for sometime (The specimen was a large male). . . .

"In the fall of 1924, Mr. A. Cooper, a prominent poultryman of Treesbank, observed a large weasel carrying a freshly killed rat which it stored below ground and then returned towards the poultry-house, causing no little apprehension to the owner. Within a short time, however, the weasel reappeared with another rat which it hid as before. In this way several rodents were accounted for during the afternoon, and Mr. Cooper assures us that the weasel 'kept up the good work for same days'.

"Being a farmer of many years' standing, Mr. Cooper has naturally lost some poultry through the agency of weasels, but while he remarks that 'there are good as well as bad actors among weasels', he has the practical good sense to recognize the value of an efficient ratter even though it be a weasel.

"Our sister, Maida Criddle, writes under date of March 4, 1925:

"'There is another weasel (longicauda) in the fowl-house, a well-behaved one this time. It came and took a piece of meat out of my hand quite nicely, which it carried down a hole and then came and sniffed all over my mitt to see if there was any more. I thought it had been killed when I visited the farm buildings next day as there was a strong smell of musk on the cat and in the fowl house, but the weasel was there as cheeky as ever. It got hold of my skirt twice and tried to pull me down its hole. I think it wanted the cloth for a bed, as it was taking straw and other material down the burrow. The poultry were very frightened at first, but they are getting used to the weasel's presence now'."

In commenting on the economic role of the long-tailed weasel in Manitoba, Criddle and Criddle (1925:145) write as follows: "Supply and demand are prominent factors in governing our weasels' food habits. The two smaller species, as we have already pointed out are so dependent upon mice for a living that they increase or diminish with the fluctuation of these creatures. The Long-tailed Weasel, however, is not so easily checked by the temporary disappearance of any particular kind of game. If mice are scarce it devotes greater attention to gophers or bush rabbits and if these are not in sufficient numbers to satisfy its appetite, the animal raids a poultry house as a last resource. In nine years out of ten, this weasel will find sufficient food about the fields and woods, but on the tenth it may be obliged to temporarily turn to domestic animals. It is at such times that the weasel is seen and its deeds recorded. A thousand mice may have been killed in the meantime, but the destruction of half a dozen hens is alone used as evidence of the weasel's economic standing.

"In the last twenty years we have permitted weasels to frequent the farm buildings at will and the poultry house has been no exception. In that time rats and mice suffered severely from the weasels, while the total number of poultry taken were six. Many times that number, however, have been killed by rats.

"When we review our experiences of the past, we are astonished to find what few poultry have been killed by weasels. Our own losses in forty-two years have not exceeded fifteen birds and even these were usually eatable. There have been reports of losses from time to time from neighbors, but on looking into details we find that there are very few farmers who have experienced more than three separate occasions of weasel depredation and the total loss per farmer in the last thirty years does not, we are sure, exceed ten birds. This is surely a remarkably small payment to weasels in general for the great good done by them in killing rodents.

"We wish to point out, too, that only the exceptional weasel becomes a poultry killer. In most cases apparently it is a fully-grown male that does the killing. There are exceptions, of course, but when we see a large weasel actively engaged in rodent hunting within a few feet of a brood of newly hatched chickens and not even looking at them, we must at least pause to ask if this animal is the enemy that we were taught to believe it to be."

A suggestion that weasels sometimes obtain the prey killed by hawks is offered by Criddle and Criddle (1925:147) who write: "Hawks are not always the aggressors, as is shown by an incident reported by Mr. H. L. Seamans, of Lethbridge, Alberta. Mr. Seamans noted a large buzzard suddenly fly straight upwards from a fence post, and then alight upon another one some distance away. A little while afterward this bird once more arose in the same manner as before, and presently repeated the performance again. An investigation then followed and revealed that a Long-tailed Weasel was following the hawk from post to post.

"We should hardly expect a weasel to attempt to capture a bird of the above type. On the other hand, it is possible that these animals might be able to startle a hawk sufficiently to cause it to drop its prey, which would thus provide food for the weasel."

The following frequency index is compiled from the foregoing data on prey of Mustela frenata.

• Moles (family Talpidae), 5
• Shrews (family Soricidae), 26
• Pigmy weasel (Mustela rixosa), 1
• Ground squirrels (genus Citellus), 23
• Chipmunks (genus Tamias), 38
• Tree squirrel (possibly all Tamiasciurus), 8
• Flying squirrel (genus Glaucomys), 1
• Pocket gophers (family Geomyidae), 34
• Mice (order Rodentia), 96
• Harvest mice (genus Reithrodontomys), 36
• Grasshopper mouse (genus Onychomys), 1
• Deer mice (genus Peromyscus), 235
• Cotton rat (genus Sigmodon), 2

• Wood rats (genus Neotoma), 14
• Meadow mice (genus Microtus), 248
• Muskrat (genus Ondatra), 1
• Old World rats (genus Rattus), 19
• House mouse (genus Mus), 1
• Jumping mouse (genus Zapus), 5
• Varying hare (Lepus americanus), 5
• Rabbits (genus Sylvilagus), 48
• small birds, 32
• chickens, 17
• lizard, 1
• snakes, 4
• insects, 3

More significant than the above compilation, of course, are the results of careful studies of the food of the long-tailed weasel in restricted areas. Examples of such studies are those of Polderboer, Kuhn and Hendrickson (1941) and Hamilton (1933:333).

According to Hamilton's (1933:332) observations on captive weasels, "There seems to be little relative difference in the amount they eat, regardless of their activities.

"In general, more food is taken in summer than in winter. Usually about a third their weight every 24 hours is eaten, but a growing young weasel will consume much more. A young male noveboracensis, weighing 145 grams, consumed an entire chipmunk, fur and bones, weighing 85 grams, in 24 hours. A day later it ate all of a partly grown rat, 105 grams, in the same length of time."

Moore (1945:253) records that a captive male that he obtained at Gainesville, Florida, consumed, on the average, between 63 and 70 grams of flesh and blood per day. The weasel itself weighed approximately 320 grams.

Sanderson (1949:413), concerning seven young weasels from Manitoba, that he raised in captivity, writes: "From the fifth to the seventh week of age, they consumed approximately 22 per cent of their body weight per day; from the eighth to the tenth week (just before reaching mature size) they consumed approximately 24 per cent; but after reaching maturity they consumed only 18 per cent. When given all the food they would take in one day, they ate as much as 40 per cent of their body weight."

Criddle and Criddle (1925:143, 146) say that weasels drinking at a bird trough "held their mouths very close to the water and as far as we could see, lapped the liquid up with rapid movements of the tongue. As a rule, after drinking, they would merely spring to the ground and vanish amid a bunch of scolding birds, but occasionally we have seen an animal slowly drag itself through the water and follow this performance by some rapid gambols, or a quick run, a method of drying which most of us have practiced in our youth." According to Hamilton's (1933:332) observations on captives, "Weasels are great drinkers, and while they take but little at a time, about 25 c.c. is drunk by a large animal during a day. . . ."

Reproduction

Philip L. Wright's several papers (1942A, 1942B, 1947, 1948A, and 1948B) reporting on his detailed studies of Mustela frenata (subspecies oribasus and longicauda) in captivity have yielded a large share of the precise information that we have concerning breeding and reproduction in this species. He has found that a single litter, of up to 9 young is born in the spring, usually in April. At three months of age the females "are full grown." The young males remain sexually immature during the first summer but the young females, as well as the females which are more than a year old, come into heat in the midsummer and are bred by the adult males. After a long period of quiescence lasting for several months, the embryos resulting from these matings become active in early spring and develop to full term in less than 27 days after they become implanted. The adult males are sexually active from April into August, when the testes are at maximal size and are conspicuous in the scrotum. A gradual regression takes place starting in August and extending into September. By October the testes may be fully regressed and the molt to white may start in this month. The white winter weasel, of either sex, is sexually inactive. The testes of the sexually active male in early spring and late summer are seven to eight times the size of the fully regressed testes. Females which had borne and suckled young were first found to be in oestrus 65 to 104 days after birth of the young. Lactation lasts for approximately 5 weeks. In 18 litters the length of the gestation period varied from 220 to 337 days with an average of 279 days. The female in heat has the vulva much swollen and she will remain in this condition for several weeks if not bred. Wright (1948A) describes the actual mating as beginning with a scuffle after which the male grabs the female by the scruff of the neck with his teeth and holds her until she becomes subdued when he clasps her lower abdomen with his front feet and arches his back over her posterior regions. The two animals remain locked in this position usually for two hours and sometimes for longer than three hours. If the animals are left together, copulation may take place again on the same day or upon succeeding days.

Hamilton (1933:316-321) writes of a freshly born M. f. noveboracensis that it ". . . was pink and much wrinkled. The wetness . . . did not entirely obscure a few sparse, rather long, white hairs . . . over its back and head. It had the pronounced and extraordinarily long neck of the adult." At one day of age the average weight of six individuals in the litter was 3.1 grams, which is 3 per cent of the weight of the adult female and 1-1/2 per cent of the weight of an adult male. At two weeks of age "The silky white hair . . . obscures the general flesh color of the skin, evident a week earlier. The hair on the back of the head and neck, also over the shoulders, is slightly longer than that of the back . . ." but there is no crest or mane or pompadour at this or any other age such as characterizes the juvenal ermine. When 21 days old one young male "hurried from the nest chamber and commenced to eat some meat." At three and a half weeks "They are all eating small pieces of meat. . . . The canine teeth have made their appearance in both the upper and lower jaw, but just a hint of the incisors show. Some of the cheek teeth are through, as the meat appears to be thoroughly masticated by the little ones." On the 36th and 37th days the eyes opened. Sanderson (1949:415) found that a litter of seven young of Mustela frenata longicauda, from Manitoba, raised in captivity, "reached the peak of their growth" at approximately ten weeks of age.

Several nests have been found. In Manitoba, Sanderson (1949:412) excavated a burrow at the mouth of which he had trapped the adult female and in which he found eight young approximately five weeks old. The "burrow was about three inches in diameter, with two chambers at a depth of twelve inches. One of these was empty, the other contained the young. The two surface-openings were but two feet apart and the entire burrow was no more than three feet long. . . . The meager nest material consisted entirely of finely chopped grass. There was no mouse hair present, no accumulation of fecal material, and no storehouse containing food."

Charles O. Handley has written me that on January 25, 1929, on the Sinkola Plantation, Thomas County, Georgia, he investigated the living quarters of a family of five weasels, four of which had been shot five days before by a hunter. According to the hunter each of the four which had been killed was approximately two-thirds the size of one which escaped into a hole in the ground. Handley found that the weasels had been using as headquarters a burrow in the trunk of an old uprooted oak as well as a nearby gopher burrow. The burrow in the oak was approximately ten feet long and had been excavated in the rotten wood. In a distance of fifty feet along the gopher tunnel there were several used openings with pathways leading away from each. On February 6, Handley, with the help of a friend, trapped a large male weasel near this place.

Criddle and Criddle (1925:143) describe a female which, one winter, slept in a bag of feathers in a basement of a house occupied by one of the authors; another weasel in winter made its headquarters in a threshing machine. The nest of the latter "was somewhat roughly constructed and consisted of a convenient bunch of straw and chaff under the cylinder."

Harper (1927:303) in the Okefinokee Swamp of Georgia dislodged a weasel from the house of a wood rat and was told of a den found in the swamp "in the trunk of a hollow cypress tree" from which a mother weasel and three young "about the size of mice" were obtained. "The bed contained, I suppose, a bushel or more of rabbit hair, rat hair, and squirrel hair. It looked like it must have been used as a den for several years, although there was no stink that I could detect except the musk from the old Weasel." Another female and three young approximately half grown were found in a hollow pine log.

Between January 6 and April 12, 1940, on 640 acres of land, in Washtenaw County, Michigan, four weasels were studied and each weasel used only one den in this period (Quick, 1944:78). Criddle (1930:279) remarks that M. f. longicauda at Aweme, Manitoba, often makes its temporary headquarters in the burrows of pocket gophers (Thomomys). A female and three young weasels were found by Shaw (1921:167) using a nest of a mountain beaver in the burrow of that animal. Green (1936), in May, in Gratiot County, Michigan, saw a weasel enter a hole under a decayed log and investigated finding four young weasels in a nest mostly of Microtus fur.

In the early part (winter and spring) of 1939, at Ames, Iowa, Polderboer, Kuhn and Hendrickson (1941) studied four weasels living in four separate dens on 160 acres typical of Iowa farmland and excavated three of the dens. One den was in a weed patch in an old mole run. The nest chamber, approximately nine inches in diameter and six inches below the surface of the ground "was filled with grasses packed in a layer-like formation. In the center of this mass was a nest hollow lined with patches of mouse and shrew fur. Beneath this layer of fur and at the sides of the nest were skins, various bones, and skulls of partially eaten mice and shrews . . . scats [were in the nest]. . . . At intervals, layers of clean grass had been laid over the filth of the former bed, thus giving the nest a stratified appearance." A second den, of a large male, was in a field of sweet clover two feet high in the former burrow of a Franklin's ground squirrel. The nest cell, seven inches in diameter and nine inches below the surface of the ground, "was lined with grasses mixed with much rabbit and mouse fur. Some scats, and bones and fur of mice and shrews were matted together in layers at the bottom of the nest." When this den was abandoned the male weasel occupied, for a month, another burrow, 20 rods distant, of a Franklin ground squirrel, in the field of sweet clover. The nest cell measured 11 by nine inches and was 11 inches below the surface of the ground. "Two nest layers were present. The first, composed chiefly of coarse straw and grass, had apparently been occupied at some time by a spotted skunk. . . . On top of the skunk nest was the weasel nest composed of fine grasses, mouse fur, and skeletal remains of mice."

Relation of the Sexes to each other and to the young

Quick (1944:75) writes that on March 28, in Michigan, he found the tracks of a male and those of a smaller animal, supposedly a female, meeting. The two "then led along the fence for about 18 chains and both entered the den of the male. . . . Only the tracks of the smaller weasel left the den on the same date. Observation on April 12 showed that the large male still occupied the den." I am at a loss to explain this behavior since breeding would not be expected to occur in late March and since I suppose that the male and female do not live together except in the breeding season. Consequently, I wonder if the sign was wrongly read.

Hamilton (1933:328), however, writes that M. f. noveboracensis is to "be found in pairs when caring for the young. During mid-May, 1927, I several times saw a male of this species carrying food to a den of young ones." Green (1936), in May in Gratiot County, Michigan, remarks that while he was uncovering and examining a nest of four young weasels, two adults ran about excitedly and one removed a young weasel. In instances where several nearly full-grown young have been obtained from one den it has been my experience (Hall, 1946:191) that the only adult trapped there was the female; no adult male was found or in the one instance when found he was living alone in a den 200 yards away from the den of the female and her young. Data are too few to warrant a definite conclusion about the extent to which the male aids in rearing the young, but I have wondered if he might not do so when the young were less than half grown and then live alone when they were more than half grown.

Mustela frenata noveboracensis (Emmons)

Long-tailed Weasel

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

Putorius Noveboracensis Emmons, Quadrupeds of Massachusetts, p. 45, 1840.

Mustela fusca DeKay, Zool. of New York, Pt. 1, Mammalia, p. 34, 1842.

Putorius fuscus Audubon and Bachman, Journ. Acad. Nat. Sci. Philadelphia, 8 (Pt. 2):288, 1842; Audubon and Bachman, Vivip. quadrupeds of N. Amer., 3:234, pl. 148, 1853 (pl. 1848).

Putorius noveboracensis, DeKay, Zool. of New York, Pt. 1, Mammalia, p. 34, 1842; Baird, Mamm. N. Amer., p. 166, 1858; Merriam, N. Amer. Fauna, 11:16, pl. 4, figs. 1, 1a, 2, 2a, pl. 5, figs. 3, 3a, text figs. 4-6, 30, June 30, 1896; Bangs, Proc. Biol. Soc. Washington, 10:13, pl. 1, figs. 2, 2a, pl. 2, figs. 2, 2a, and pl. 3, figs. 3, 3a, February 25, 1896; Cory, Mamm. Illinois and Wisconsin, p. 366, plates, 1912.

Putorius erminea, Audubon and Bachman, Vivip. quadrupeds of N. Amer., 2:56, pl. 59, 1851.

Putorius agilis Audubon and Bachman, Vivip. quadrupeds of N. Amer., 3:184, pl. 140, 1853.

Putorius richardsonii, Baird, Mamm. N. Amer., p. 164, 1858 (part).

Putorius (Gale) erminea, Coues, Fur-bearing animals, p. 109, 1877 (part).

Putorius noveboracensis notius Bangs, Proc. New England Zool. Club, 1:53, June 9, 1899. Type from Weaverville, Buncombe County, North Carolina.

Mustela noveboracensis noveboracensis, Miller, U. S. Nat. Mus. Bull., 79:97, December 12, 1912; Soper, Journ. Mamm., 4:251, November 1, 1923.

Mustela cicognanii, Henninger, Journ. Mamm., 2:239, November 29, 1921; Seton, Lives of game animals, 2:584, 1929 (part, Ohio); Hamilton, Amer. Midland Nat., 14:290, July, 1933 (part, Ohio); Lyon, Amer. Midland Nat., 17:109, January, 1936 (part, Ohio).

Mustela noveboracensis, Jackson, Journ. Mamm., 3:15, February 8, 1922.

Mustela frenata noveboracensis, Hall, Carnegie Instit. Washington Publ., 473:104, November 20, 1936; Hall, Amer. Midland Nat., 18:304, March, 1937.

Type.—Williamstown, Berkshire County, Massachusetts. Type specimen not known to be in existence.

Range.—Altitudinally, sea level to highest parts of mountains of eastern United States; Canadian Life-zone of Ontario and Quebec southward through eastern United States in Canadian, Transition and Upper Austral life-zones to and including upper edge of Lower Austral Life-zone in the Carolinas and northern parts of Georgia, Alabama, and Mississippi; westward from the Atlantic Coast to St. Croix and Mississippi rivers. See figure on page .

Characters for ready recognition.—Differs: From M. f. olivacea, in males, by width of tympanic bulla which is less than rather than more than 8.5 mm., and in adult females by total length which is less than rather than more than 345 mm. and by mastoid breadth which is less than rather than more than distance between articular faces of exoccipital condyle and glenoid fossa; from M. f. occisor by a number of average differences including smaller size, relatively shorter tail and relatively narrower skull (see measurements); from M. f. spadix by least width of color of underparts amounting to less than 41 per cent of greatest width of color of upper parts, absence of color of underparts on ankles and feet, adults with hind foot less than 50 in males and 40 in females, orbitonasal length less than 15.5 in males and 13.5 in females, length of tooth-rows less than 18.0 in males and 15.7 in females, mastoid breadth less than 25.5 in males and 22.0 in females; from M. f. primulina in males by interorbital breadth averaging more than 24 per cent of basilar length, orbitonasal length averaging more than 34 per cent of basilar length or 64 per cent of mastoid breadth, tympanic bullae less inflated anteromedially, than posteromedially, and in females by orbitonasal length amounting to more than two-thirds of mastoid breadth, by zygomatic breadth averaging less than 21, and by anterolateral margin of tympanic bullae not projecting below squamosal; from M. f. arthuri in males, by zygomatic breadth more than distance between anterior palatine foramen and anterior margin of tympanic bulla and by convex dorsal outline of skull in longitudinal axis.

Description.—Size.—Male and Female:

Locality Number of specimens averaged Total length Length of tail Per cent of body- length Length of hind foot

Massachusetts 8 ad. ♂ 415 (390-432) 146 (127-159) 54% 46.0 (41.0-48.0)

4 ad. ♀ 311 (298-321) 104 (95-114) 50% 33.9 (31.5-37.0)

Liberty Hill, Conn. 10 ad. ♂ 411 (379-438) 141 (124-155) 52% 47.1 (43.0-51.5)

6 ad. and sad. ♀ 318 (303-338) 105 (80-123) 49% 33.0 (31.7-36.0)

Beaver Dam, Wisc. 10 ad. ♂ 407 (372-431) 130 (113-143) 47% 46.0 (42.0-50.0)

4 ad. ♀ 326 (303-338) 99 (86-108) 43% 35.6 (34.6-38.0)

Washtenaw Co., Mich. 10 ad. ♂ 371 (350-405) 130 (115-140) 54% 45.0 (40.0-50.0)

10 ad. ♀ 306 (290-335) 97 (90-120) 46% 34.0 (30.0-40.0)

The length of the hind foot averages more than the basal length in males whereas the reverse is true in females. The tail, relative to the length of the body, is longer in males than in females. The average differences in external measurements of the two sexes in Massachusetts, are: total length, 104; length of tail, 42; length of hind foot, 12.1. In Michigan, where the males are smaller, corresponding differences are only, 65, 33, and 11. Weight of 19 adult males from New York (Hamilton, 1933:294), 225 (196-267) grams and in 13 adult females, 102 (72-126) grams. Weights of 2 adults from Michigan are: ♂ 258; ♀ 101 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 as shown in figure .

Color.—Upper parts, in summer, Vandyke Brown or darker than tone 4 of Burnt Umber of Oberthür and Dauthenay, pl. 304. Sometimes approaching tone 2 of Warm Sepia of Oberthür and Dauthenay, pl. 305. Underparts, in summer, ranging from white through Napthalene Yellow (Peterboro, N. Y.), Pale Orange Yellow (eastern Mass.), near Primuline Yellow (unusual specimen from Leelanau Co., Mich.) to near (c) Deep Chrome (no. 19053, U. S. Nat. Mus., Roan Mts., N. C.) In winter, all white except tip of tail, or upper parts near (12" 1) Rood's Brown and tone 2 of Raw Umber of Oberthür and Dauthenay, pl. 301, with underparts white or sometimes tinged with yellowish. 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 to foot and sometimes over upper sides of toes and on medial sides of hind limbs only to knees. Least width of color of underparts averaging, in a series of twenty-two males, mostly in full winter pelage, from Liberty Hill, Connecticut, 21 (11-40) per cent of greatest width of color of underparts. In eleven females from the same place, corresponding percentages are 20 (14-29). Black tip of tail in same series of males, most of which are in full winter pelage, 70 (60-75) mm. long; thus longer than hind foot and averaging 50 per cent of length of tail-vertebrae.

Skull and teeth.—Male (based on ten adults from Massachusetts): See measurements and plates -; weight, 3.6 (3.3-4.4) grams; basilar length, 44.6 (43.3-46.0); zygomatic breadth less than distance between condylar foramen and M1 or than between anterior palatine foramen and anterior margin of tympanic bulla; mastoid breadth less than postpalatal length; postorbital breadth more or less than length of upper premolars and greater than width of basioccipital measured from medial margin of one foramen lacerum posterior to its opposite; interorbital breadth more or less (usually more) than distance between foramen opticum and anterior margin of tympanic bulla; breadth of rostrum less than length of tympanic bulla; least width of palate less than length of P4; anterior margin of tympanic bulla as far posterior to foramen ovale as width of 3 to 6 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 or shorter than rostrum; anterior margin of masseteric fossa behind or directly below posterior fourth of m1.

Female (based on five adults from Mass.): See measurements and plates -; weight, 1.7 (1.2-2.1) grams; basilar length, 36.5 (35.2-38.1); zygomatic breadth less than distance between condylar foramen and M1 or than between anterior palatine foramen and anterior margin of tympanic bulla; postorbital breadth more or less than length of upper premolars and more than width of basioccipital measured from medial margin of one foramen lacerum posterior to its opposite; least width of palate more or less (usually less) than greatest length of P4; tympanic bulla as far posterior to foramen ovale as width of 4 to 5-1/2 upper incisors; height of tympanic bulla less than distance from its anterior margin to foramen ovale; length of tympanic bulla more than length of lower molar and premolar tooth-row and longer than rostrum.

The skull of the female averages 53 per cent lighter than that of the average male.

Comparisons of the skull with those of M. f. olivacea, M. f. spadix, M. f. primulina, and M. f. arthuri, are made in the accounts of those subspecies. As compared with that of M. f. occisor the skull of adult male noveboracensis, is of smaller average size with relatively (to basilar length of Hensel) lesser mastoid and zygomatic breadths. In addition to the zygomatic arches of noveboracensis being less widely bowed outward they seem to be more rounded posteriorly. Comparisons of subadult females indicate that these differences exist in the females as well as in the adult males.

Remarks.—The earliest of the post-Linnaean references to this weasel mostly were under the specific name erminea in the belief that the American animal was the same as the larger of the two common species of weasel in the Old World. The name noveboracensis, now in use for this subspecies, was applied in 1840 and since that time the males usually have borne that name; the females, because they are smaller, were more frequently confused with some other species. Audubon and Bachman in 1853 even proposed the name agilis for the female in the mistaken belief that it was a species distinct from the male. After 1896, when Bangs correctly classified the weasels of the eastern United States, the males have been correctly identified and the females, except by a few authors, likewise have been correctly named. Because many early American naturalists did their first collecting of mammals in the geographic range of noveboracensis, the person who examines labels of specimens of this subspecies can find data written in the hand of Spencer Fullerton Baird, Theodore Roosevelt, and other naturalists famous for their work as scientists or accomplishments otherwise. The material is more nearly adequate than is that of many other subspecies and the number of specimens is exceeded—and only slightly—by that of the subspecies nevadensis, which like noveboracensis has a relatively large geographic range.

Intergradation with Mustela frenata spadix is indicated by subadult males from western Wisconsin, namely, one from Gordon, three from Colfax and one from Meridean. Linear measurements of the teeth of these specimens are exactly intermediate between those of spadix from Elk River, Minnesota, to the west, and noveboracensis from, say, Beaver Dam, Wisconsin, to the east. The specimens from western Wisconsin show approach to spadix also in that the length of the tooth-rows and breadth of the rostrum are slightly greater than in noveboracensis from farther east, say, Beaver Dam, Wisconsin.

Indeed, animals from as far east as Beaver Dam itself might be thought of as showing some approach to spadix. Although, along the eastern seaboard, the upper lips, with rare exceptions, are the same color as the underparts, farther west, in Michigan and Wisconsin, the lips more often than not are white. Animals from Beaver Dam have slightly shorter black tips on the tails, broader extent of the light color of the underparts and females average slightly larger than typical noveboracensis, say, those from Massachusetts. Each of these differences reflects characters found better developed in the spadix-longicauda stock to the west.

Toward the southern part of its range where noveboracensis meets M. f. olivacea there is a marked increase in yellowness of the underparts. This coloration of the underparts, since it is not so well marked in the northern part of the range of noveboracensis, might be regarded as showing intergradation with olivacea and primulina, each of which has far more intensely colored underparts than does noveboracensis. Excepting this increase of yellow on the underparts, however, there are few if any characters of noveboracensis which undergo marked change as approach to the range of olivacea is made. Indeed, the characters of noveboracensis remain constant to within a relatively short distance of the geographic range of olivacea.

Notwithstanding the state of affairs described above, intergradation seems to take place. Three specimens referred to noveboracensis but which at the same time are regarded as intergrades with olivacea are as follows: No. 28.300, Charleston Museum, from five miles east of York, South Carolina, is an adult female with a badly crushed skull. In external measurements the specimen agrees with noveboracensis. The underparts, as regards color and width, are intermediate. The general proportions of the skull and tympanic bullae agree with those of noveboracensis but the skull is larger than in any female of true noveboracensis and approaches that of olivacea. The same can be said of a young female, no. 80, Ohio State Museum, from Roswell, Georgia.

Another female, no. 171559, U. S. Nat. Mus., from Lookout Mountain, 1500 ft., Fort Payne, Alabama, is barely subadult. The external measurements are nearer those of olivacea. The color and narrowness of the underparts are typical of noveboracensis. The proportions and especially size of the skull show approach to olivacea, though they are nearer to noveboracensis when all features are taken into account. In the northern part of its range individuals of noveboracensis attain larger size than farther south. This tendency reaches its extreme, in males at least, in M. f. occisor of Maine. Specimens of noveboracensis from the Adirondacks of New York average larger (see cranial measurements on page ) than those from farther south, and thus approach occisor in size as well as in geographic position. Also, occasional individuals which strongly show characters of occisor are found even farther south than the Adirondacks of New York. This is true of no. 96518, U. S. Nat. Mus., ♂ ad., from Lunenburg; Massachusetts. The animal has a large skull of relatively great width much as in occisor, although its external measurements, relative length of tail and long, terminal, black brush place it with noveboracensis rather than with occisor. Of a pair of specimens from Ossipee, New Hampshire, the male, no. 77108, U. S. Nat. Mus., has a long (175 mm.) tail, and short (60 mm.) black pencil as in occisor, although otherwise it is referable to noveboracensis. Still another specimen, a subadult male, no. 4193, Mus. Comp. Zoöl., from Upton, Maine, has a longer (51 mm.) hind foot than noveboracensis although it otherwise agrees with that subspecies. As remarked by Bangs (1899:55), other than fully adult specimens from the range of occisor are "troublesome," and would not be selected as distinct from noveboracensis if placed in a series of that subspecies, say, from New York State. In view of the facts that several specimens from intermediate localities combine the characters of noveboracensis and occisor, that noveboracensis in the northern part of its range averages larger than it does farther south and thus approaches occisor in size, and that occasional large specimens resembling occisor in several, but not all, features sometimes crop up in the northern part of the range of noveboracensis, it appears that noveboracensis and occisor intergrade. Therefore they are treated as two subspecies of the single species, Mustela frenata.

Intergradation with M. f. primulina has been commented on in the discussion of that subspecies. Female, no. 159980, U. S. Nat. Mus., from Golconda, Illinois, has many characters of primulina but two young males from there agree better with noveboracensis.

Examination of 283 adult and subadult skulls for malformation of the frontal sinuses revealed only ten that were not obviously malformed. Two were from New York, one from Massachusetts, one from Pennsylvania, and six from the 52 specimens from Michigan and Wisconsin. In addition, skulls of many young and even juveniles were malformed.

Specimens examined.—Total number, 555, arranged alphabetically by states and provinces and, unless otherwise noted, from north to south by counties in each state. Except as otherwise noted specimens are in the United States National Museum.

Alabama. DeKalb County: Fort Payne, 1.

Connecticut. Litchfield County: Riverton, 1; Gaylordsville, 1. Hartford County: East Hartford, 4 (3); Glastonbury, 2; South Glastonbury, 4. Windham County: Plainfield, 2 (1). Fairfield County: Greenwich, 2. New London County: Liberty Hill, 35 (33, 2).

District of Columbia. Washington, 3; near Washington, 1; Eastern Branch, 1; Congress Heights, 1; Benning, 1; no definite locality, 1.

Georgia. Towns County: Young Harris, 1. Cherokee County: Canton, 1. Cobb County: Roswell, 1.

Indiana. St. Joseph County: Notre Dame, 2. Porter County: Hebron, 1. Miami County: Denver, 5 (4, 1). Wells County: Bluffton, 1. Howard County: Russiaville, 1. Jay County: Salamonia, 1. Boone County, 1. Knox County: Bicknell, 3.

Illinois. Lake County: Camp Logan, 3; Fort Sheridan, 1. Cook County: W Northfield, 2; Flossmoor, 1; no locality more definite than county, 1. Du Page County: Bloomingdale Spg., 1. Carroll County: Savanna, 1. McLean County: Normal, 1. Champaign County: Harwood Township, 1. Pike County?: Milton Spring, 1. Pope County: Golconda, 3.

Kentucky. Woodford County: Midway, 1. Hancock County: Hawesville, 1.

Maine. Oxford County: Upton, 1; Bethel, 1.

Maryland. Howard County: Long Corner, 1; Hanover, 1. Montgomery County: Gaithersburg, 1; Garret Park, 1; Chevy Chase, 1; Bethesda, 1. Prince Georges County: Laurel, 18; Plummer Island, 3; Oxon Hill, 1. Talbot County: Easton, 1. Dorchester County: Cambridge, 5.

Massachusetts. Middlesex County: Wilmington, 6; Burlington, 6; Lexington, 1; Wayland, 2. Berkshire County: New Marlboro, 1. Worcester County: Lunenburg, 2; Lancaster, 1; Princeton, 2. Norfolk County: So. Weymouth, 1. Plymouth County: Wareham, 5.

Michigan. Marquette County: Michigamme, 1. Charlevoix County: Thumb Lake, 1; 1/2 mi. N Thumb Lake, 1. Leelanau County: Leland, 3; Duck Lake, 2 mi. S Leland, 1; Lost Pond, 8-1/2 mi. S Leland, 1. Osceola County: Le Roy, 2. Huron County: Rush Lake, 1. Saginaw County: East Saginaw, 1. Oakland County: Royal Oak, 4; South Lyon, 1. Livingston County: Portage Lake, 1. Washtenaw County: Portage Lake, 6; Waterloo, 2; Lima, 1; Ann Arbor, 11; 3 mi. E Ann Arbor, 1; 2 mi. SE Ann Arbor, 1; 2 mi. S Ann Arbor, 1; 3 mi. S Ann Arbor, 1; Dixboro, 1; Pittsfield, 3 (2); Saline, 1; near Saline, 2; 1 mi. S Saline, 2; York, 2; Manchester, 2. Lenawee County: Morenci, 1. Cass County: Marcellus Township, 1. Berrien County: Harbert, 1; Warren Wood Preserve, 1; Warren Woods, 1.

New Hampshire. Grafton County: Franconia, 1. Carroll County: South Chatham, 4 (3); Ossipee, 2; Intervale, 1. Merrimack County: Webster, 2.

New Jersey. Morris County: Morristown, 1. Essex County: West Orange, 1. Mercer County: Princeton, 1. Ocean County: Point Pleasant, 1. Camden County: Haddonfield, 1. Cumberland County: Millville, 2.

New York. St. Lawrence County: Ogdensburg, 1. Clinton County: Rouses Point, 1. County?: Adirondacks, 12. Essex County: Elizabethtown, 1; Schroon Lake, 1; no locality more definite than county, 1. Lewis? County: Locust Grove, 4; Lyons Falls, 1. Warren County: Lake George, 6; Caldwell, 1. Hamilton County: Beaver Brook, 1/2 mi. above mouth Indian Lake, 1. Oswego County: Scriba, 2; Palermo, 1. Monroe County: Penfield, 3. Madison County: Peterboro, 6 (2). Schoharie County: Schoharie, 1. Rensselaer County: East Shodack, 1. Tompkins County: Taughannock Falls, 2; Ithaca, 4 (3); Glenside, Ithaca, 1; 6 mi. Creek, Ithaca, 1. Green County: Lanesville, 1. Orange County: Poplopen's Pond, 1; Highland Falls, 1. Putnam County, 1. Westchester County: Sing Sing, 4; Armonk, 1; Hastings, 3 (2, 1). Nassau County: Flushing Meadows, 1; Flushing, 1; near Flushing, 1; Oyster Bay, 2. Long Island: Cold Spring Harbor, 1; Bridgehampton, 1. County in question: Severance, 3; Lake Grove (Long Island?), 1.

North Carolina (east to west by counties). Wake County: Raleigh, 4 (1, 1, 2). Mitchell County: Magnetic City, foot of Roan Mountain, 6; Roan Mt., 1; Roan Mt., 6000 ft., 3. Buncombe County?: Valley of Black Mts., 1. Madison County, 2.

Ohio. Trumbull County: Warren, 1. Seneca County: Tiffin, 1. Summit County: Ira, 2. Crawford County: Galion, 1. Ashland County: Loudonville, 1. Auglaize County: New Bremen, 3. Franklin County: 3 mi. N Columbus, 1; Minerva Park, Columbus, 5. Fairfield County: Sec. 32, Pleasant Twp., 1; Lancaster, 1. Clinton County: Reesville, 1; 1/2 mi. S and 1/2 mi. W Wilmington, 2. Pike County: Waverly, 1.

Ontario. Sudbury District: Metagama, 2. Carleton County: Ottawa, 2. Muskoka County: Lake of Bays, 1; Bracebridge, 1. Haliburton County: Gooderham, 1. Simcoe County: Orillia, 4 (2, 2.) Prince Edward County: Bloomfield, 1. York County: Toronto, 1. Waterloo County: Branchton, 3; Preston, 2; no locality save county, 1. Welland County: Ridgway, 1. Elgin County: St. Thomas, 1. Essex County: Kingsville, 1; Point Pelee, 1.

Pennsylvania (east to west by counties). Crawford County: Pymatuning Swamp, 3-1/2 mi. W Linesville, 1; Meadville, 2. Beaver County: Beaver, 1; Raccoon Creek, 1. Butler County: Mars, 1; Leasuresville, 4. Allegheny County: Allegheny, 1. Warren County: Bear Lake, 2. Westmoreland County: Kingston, 1; Laughlinstown, 2. Somerset County: Confluence, 1; Tub Mill Run, 2 mi. N Springs, 1. Jefferson County: Siegel, 1. Clearfield? County: Penfield, 1. Cambria County: Cresson, 1. Fulton County: Well's Tannery, 1. Clinton County: near Round Island, 2. Cumberland County: Carlisle, 1. Snyder County: 5 mi. S Selinsgrove, 1. Northumberland County: Pottsgrove, 1. Union County: Mifflinburg, 1. Sullivan County: Lopez, 7 (4, 3). Chester County: Westtown, 1; Valley Forge, 1; W Bradford Township, 1; no locality more definite than county, 3. Philadelphia County: Holmsburg, 2. Bucks County: 1. Pike County: Milford, 1.

Rhode Island. Providence County: Chepachet, 1. Washington County: Lake Warden, 2.

Quebec. Megantic County: Black Lake 1. County in question: Meach Lake, 1.

South Carolina. York County: 5 mi. E York, 1. Laurens County: Laurens, 1.

Tennessee. Campbell County: Highcliff, 1. Carter? County: Roan Mts., 1. Hamilton County: Walden Ridge, near Soddy, 3.

Vermont. Windsor County: Hartland, 1.

Virginia. Shenandoah County: Toms Brook, 1. Arlington County: Arlington, 1; Ballston, 1; Alexandria, 1. Fairfax County: Falls Church, 3; Mt. Vernon, 2; no locality more definite than county, 1. Prince William County: Occoquan, 1. Essex County: Montague, 1. Prince George County, 1. Norfolk County: Wallaceton, 1. Grayson County: Mt Rogers, 3. County in question: Dismal Swamp, 1; Massanutten Mt., 1.

West Virginia. Hardy County, 1. Pendleton County: radius of 2 mi. Smoke Hole, 1. Greenbriar County: White Sulphur, 2.

Wisconsin. Douglas County: Gordon, 1. Vilas County: Mamie Lake, 4. Dunn County: Colfax, 4; Meridean, 1. Door County: state game farm, 17; no locality more definite than county, 1. Dodge County: Rolling Prairie, 1; Beaver Dam, 52. Dane County: Wingra Lake, 1. Waukesha County: Pewaukee, 2. Racine County: Waterford Township, 2. Rock County: Milton, 1; Bowers Lake, 1. Walworth County: Lane's Mill, 8 mi. N Elkhorn, 7 (1, 6); Delavan, 7.

Mustela frenata occisor (Bangs)

Long-tailed Weasel

Plates , , , , and

Putorius occisor Bangs, Proc. New England Zool. Club, 1:54, June 9, 1899.

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

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

Type.—Male, adult, skull and skin; no. 9102, coll. of E. A. and O. Bangs in Mus. Comp. Zoöl.; Bucksport, Hancock County, Maine; January 15, 1899; obtained by Alvah G. Door but measured and sexed by O. Bangs.

The skin is well made and in good condition. It is in full, white winter-dress with black-tipped tail. The skull has the posterior half of the left zygomatic arch broken away; otherwise the skull is unbroken and complete. Left I3 and right P3 are missing. The teeth otherwise all are present and entire.

Range.—Maine; possibly north locally to south side of St. Lawrence River in Quebec and possibly occurring in western New Brunswick. Zonal range Canadian and probably Transition. See figure on page .

Characters for ready recognition.—Differs from M. f. noveboracensis by a number of average differences including larger size, relatively longer tail and relatively wider skull (see page , and measurements on pages , ).

Description.—Size.—Male: Five adults yield average and extreme measurements as follows: Total length, 443 (430-465); length of tail, 163 (154-175); length of hind foot, 50 (47-54). Tail averages 58 per cent as long as head and body. Length of hind foot averages more than basal length.

Female: Measurements of two subadult female topotypes are as follows: Total length, 346, 318; length of tail, 116, 110; length of hind foot, 39, 35.5.

Tail amounts to 50 per cent and 54 per cent of body-length respectively. Length of hind foot more or less than (about equal to) basal length.

The average differences in external measurements of the two sexes are: Total length, 111; length of tail, 50; length of hind foot, 12.5.

Externals.—As described in Mustela frenata noveboracensis.

Color.—As described in Mustela frenata noveboracensis except that black tail-tip in series of 10 males in full winter pelage 60 (45-80) mm. long; thus averaging 39 per cent of length of tail vertebrae.

Skull and teeth.—Male (based on 3 adults): See measurements and plates -. As described in Mustela frenata noveboracensis except that: Weight, 4.2 (4.1-4.3) grams; basilar length, 45.7 (44.9-46.9); zygomatic breadth more or less than (about equal to) distance between anterior palatine foramen and anterior margin of tympanic bulla; least width of palate rarely less than length of P4; anterior margin of masseteric fossa behind or directly below posterior half of m2.

Female (based on 2 subadults): See measurements and plates -. As described in Mustela frenata noveboracensis except that: Weight, 2.0 (1.9-2.1) grams; basilar length, 37.3, 38.2.

Comparison of the skull with that of M. f. noveboracensis is made in the account of that subspecies.

Remarks.—Excepting a specimen in the Academy of Natural Sciences of Philadelphia, obtained in 1893, and two in the Boston Society of Natural History, obtained in 1925, I have seen no material of this subspecies in addition to that examined by Bangs at the time he prepared the original description in 1899.

Anderson (1945:56, 57) records a specimen, Canadian National Museum Catalogue Number 18426, from Kamouraska County, Quebec, as of this subspecies and thinks that occisor occurs north of Maine "locally to south side of lower St. Lawrence River in Quebec; probably also in western New Brunswick."

So far as the available material of occisor permits one to judge, it is distinguished from noveboracensis by a combination of characters no one of which invariably can be relied upon as diagnostic. Employing adult males, average differences indicate that M. f. occisor is larger in each of the external and cranial measurements; tail relatively longer; black tip of tail relatively shorter; mastoid and zygomatic breadth relatively greater and zygomatic arches more nearly square posteriorly.

Considering the large number of specimens of noveboracensis which are available in comparison with the few of occisor it is not surprising that some noveboracensis should be found which exceed in size those of occisor. This is the case as regards the basilar length of a very old male, no. 96518, U. S. Nat. Mus., from Lunenburg, Massachusetts. Also, the skull is actually broader than any of those of occisor. However, this specimen is much older than any occisor examined. In a female, no. 4260, Mus. Comp. Zoöl., from Liberty Hill, Connecticut, the skull is longer (but narrower) than in either of the two available females of occisor.

The average differences pointed out above which characterize this extreme northern population of noveboracensis-like weasel in comparison with true noveboracensis without much question are geographic variations. Whether or not these variations are of a degree sufficient to warrant nomenclatural recognition is debatable. With equally scanty material from other regions I have not named variations seemingly as great as those shown by occisor. The paucity of material of occisor is a handicap in making a decision in this instance.

Each of the adult and subadult specimens, except the one from Perry, shows malformation resulting from the infestation of the frontal sinuses with parasites.

Specimens examined.—Total number, 18, listed by counties from west to east and unless otherwise indicated in the Museum of Comparative Zoölogy.

Maine. Oxford County: South Andover, 2 (Boston Soc. Nat. Hist.); Umbagog Lake, 1. Franklin County: Seven Pd. Township, 1. Piscataquis County: Grenville, [= Greenville?], 1. Hancock County: Bucksport, 10. Washington County: 3rd Mopang Lake, 1 (Acad. Nat. Sci. Phila.); Perry, 1 (Boston Soc. Nat. Hist.). County in question: Moosehead Lake, 1.

Mustela frenata primulina Jackson

Long-tailed Weasel

Plates , , , , and

Mustela primulina Jackson, Proc. Biol. Soc. Washington, 26:123, May 21, 1913.

Putorius noveboracensis, Baird, Mamm. N. Amer., p. 166, 1858 (part).

Mustela longicauda longicauda, Dice, Journ. Mamm., 4:108, May 9, 1923.

Mustela longicauda primulina, Linsdale, Journ. Mamm., 9:141, May 9, 1928.

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

Mustela frenata, Leopold and Hall, Journ. Mamm., 26:143, July 19, 1945.

Type.—Male?, young, skull and skin; no. 168006, U. S. Nat. Mus., Biol. Surv. Coll.; 5 miles northeast of Avilla, Jasper County, Missouri; May 11, 1905; obtained by Hartley H. T. Jackson; original no. 7869X.

The skin lacks the distal part of the tail—the part which bears the black tip. Otherwise the skin is complete and well preserved. The teeth of the permanent dentition all are present and entire. The lower jaws are complete and unbroken. The skull is broken transversely through the interorbital region, transversely through the braincase and longitudinally through the basioccipital. Both zygomatic arches are gone. The type is judged to be a male rather than a female as stated by the original describer, Jackson (1913:123), whose measurements of hind foot, interorbital constriction, maxillary tooth-row, and mandibular tooth-row agree with those of males and are larger than those of any female seen of this subspecies.

Range.—Upper and Lower Austral life-zones west of the Mississippi River in Missouri and Arkansas, the southeastern half of Iowa, eastern half of Kansas and Oklahoma, northern Louisiana and northeastern Texas. Southern and southwestern limits of range undetermined. See figure on page .

Characters for ready recognition.—Differs from M. f. noveboracensis in males by interorbital breadth averaging less than 24 per cent of basilar length, orbitonasal length averaging less than 34 per cent of basilar length or 64 per cent of mastoid breadth, tympanic bullae as much inflated anteromedially as posteromedially, and in females by orbitonasal length amounting to less than two-thirds of mastoid breadth, by zygomatic breadth averaging more than 21 mm., and by anterolateral margin of tympanic bulla projecting below squamosal; from M. f. spadix by least width of color of under parts amounting to less than 40 per cent of greatest width of color of upper parts, by absence of color of underparts on hind leg below knee, and by smaller size (hind foot less than 50 in males and 40 in females; orbitonasal length less than 15.5 in males and 13.5 in females; length of tooth-rows less than 18 in males and 15.7 in females; mastoid breadth less than 25.5 in males and 22 in females); from M. f. longicauda by Brussels Brown rather than near (h) Clay Color of upper parts, least width of underparts less than 40 per cent of greatest width of color of upper parts, absence of color of underparts on hind leg below knee, zygomatic breadth less than 28.8 in males and 24.1 in females; from M. f. neomexicana by Brussels Brown rather than Buckthorn Brown color of upper parts, in absence of white frontal spot and broad white bands on sides of head, in anterolaterally rounded, rather than "square," tympanic bullae and in zygomatic breadth of less than 30 in males and 24 in females; from M. f. frenata and M. f. texensis by absence of white facial markings and smaller size (basilar length of adult males less than 47, tail-length less than 155 in males, and hind foot less than 40 in females); from M. f. arthuri by less evenly spreading zygomatic arches (see pls. 16, 17 and 18), greater inflation of the tympanic bullae anteromedially and more nearly straight (less convex) dorsal outline of skull.

Description.—Size.—Male: Eighteen adults and subadults from Douglas County, Kansas, yield average and extreme measurements as follows: Total length, 397 (371-440); length of tail, 133 (120-147); length of hind foot, 43 (40-47). Tail averages 50 per cent as long as head and body. Length of hind foot averages less than basal length. Corresponding measurements, originally taken in inches and fractions thereof, of 9 adults and subadults from Boone County, Arkansas, are as follows: 413 (384-438); 138 (127-155); 41 (38-44).

Female: Six adults and subadults from Douglas County, Kansas, yield average and extreme measurements as follows: Total length, 339 (317-355); length of tail, 107 (95-115); length of hind foot, 35 (34-37). Tail averages 46 per cent of length of head and body. Length of hind foot less than basal length. Corresponding measurements, originally taken in inches and fractions thereof, of 5 adults and subadults from Boone County, Arkansas, are as follows: 355 (332-397); 113 (101-127); 33 (29-38).

The average differences in external measurements of the two sexes, in Douglas County, Kansas, are: Total length, 58; length of tail, 26; length of hind foot, 8. An adult male from Boone Co., Iowa, weighed 293 grams.

Externals.—Longest facial vibrissae black or dark brown (often both colors in same specimen) and extending beyond ear; carpal vibrissae colored either like underparts or upper parts and extending to apical pad of fifth digit; hairiness of foot-soles as shown in figure .

Color.—Upper parts, in summer, Brussels Brown to near (14 n) Brussels Brown, or tones 2 to 4 of Raw Umber of Oberthür and Dauthenay, pl. 301. Chin and rarely upper lips white. Remainder of underparts Picric Yellow to Primuline Yellow. In winter, color essentially the same except for smoke-gray effect in upper parts and more whitish in underparts. Tip of tail at all times black. Upper parts of uniform color except for occasional darkening of nose and mid-dorsal region. Color of underparts extends distally on posterior sides of forelegs onto antipalmar faces of toes, on medial sides of hind legs only to a point between knee and ankle. Least width of color of underparts averaging, in a series of 21 males from Lawrence, Kansas, 23 (9-35) per cent of greatest width of color of upper parts. Black tip of tail in same series, most of which are in full winter pelage, 52 (40-70) mm. long; thus longer than hind foot and averaging 39 per cent of length of tail-vertebrae.

Skull and teeth.—Male (based on ten adults from Douglas County, Kansas): See measurements and plates -; 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 -; 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; 3 mi. SE Bergman, 1; 3 mi. S Bergman, 1; 3 mi. SW Bergman, 1; 4 mi. SE Bergman, 2; 5 mi. SE Bergman, 1; 4-1/2 mi. SE Bergman, 3; 5 mi. SE Bergman, 1; 5 mi. S Bergman, 2; 5 mi. SW Bergman, 2. Washington County: Fayetteville, 1. Crawford County: 10 mi. S Winslow, 1. Sebastian County: Fort Smith, 1.

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

Kansas. Riley County: near Winkler, 1. Pottawatomie County: Onaga, 1. 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). Woodson County: 1-1/2 mi. S Neosho Falls, 1. Greenwood County: 8 mi. SW Toronto, 2. Pratt County: Pratt, 1.

Louisiana. Quachita Parish: Swartz, 1.

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

Oklahoma. Payne County: 8 mi. NW Stillwater, 1. Cleveland County: Norman, 1.

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

Mustela frenata arthuri Hall

Long-tailed Weasel

Plates , and

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 -) 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 on page .

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 .

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 -. 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; 5 mi. W Eagle Lake, 1; 3 mi. S Garwood, 1.

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

Mississippi. Harrison County: Saucier, 1.

Mustela frenata olivacea Howell

Long-tailed Weasel

Plates , , , , and

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 -), 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 on page .

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 .

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 -; 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 -; 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. Colbert County: Leighton, 1. Autauga County: Autaugaville, 1. Dale County: Midland City, 1. Mobile County: Mobile River, 12 mi. NE Mobile, 1.

Florida. Alachua County: Gainesville, 4. 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; 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. County in question: Billy's Island, Okefinokee Swamp, 1; Okefinokee Swamp, 1.

Mississippi. Clark County: Souinlonie Creek, 1.

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

Mustela frenata peninsulae (Rhoads)

Long-tailed Weasel

Plates , and

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 on page .

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 .

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 -. 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. Pinellas County: Tarpon Springs, 1. Hernando County, 1. Polk County: Auburndale, 1; no locality more definite than county, 1. Orange County: Apopka, 3. Volusia County: Enterprise, 1. Seminole County: Osceola, 1.

Mustela frenata spadix (Bangs)

Long-tailed Weasel

Plates , , , , and

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 on page .

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 .

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

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

Nebraska. Cuming County: Beemer, 5.

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

South Dakota. Roberts County, 1. Marshall County: Fort Sisseton, 1. Douglas County: Armour, 1. Clay County, 1.

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 on page .

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 .

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 -; 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 -; 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; Battle River, south of Camrose, 1; Daysland, 1; Moose Pass, 1; Blindman River, 2 (1, 1); Red Deer, 3 (2, 1); Bearberry Creek near Sundre, 1; Canad. Nat. Park, N.W. Territory, 1; Red Deer River, Didsbury, 1; Canmore, 1; Calgary, 11 (6, 2, 1); Red Deer River, 3; Little Sandhill Creek, Red Deer River, 1; Waterton Lake Park, 2; Sweetgrass Hills, 1; Alberta, 1.

Colorado. Yuma County: Wray 4 (1, 3).

Kansas. Rawlins County: 7 mi. N, 3 mi. W Beardsley, 1; 6 mi. S and 2 mi. E Atwood, 1; 15 mi. SE Atwood, 1. Thomas County: near Brewster, 2; no locality more definite than county, 2. Trego County, 2 (1). Cloud County: Glasco, 1.

Manitoba. Portage la Prairie, 3; Carberry, 2 (1, 1); Carman, 1; Max Lake, Turtle Mt., 1.

Montana. Glacier County: St. Marys Lake, 1; Blackfoot, 1: Blackfoot Agency, 1. Blaine County: 6 mi. east Chinook, 1. Pondera County: 1/2 mi. SE Conrad, 1. Toole County: Shelby Junction, 1. Hill County: Havre, 1. Fergus County: Moccasin Mts., 5 mi. NW Hilger, 1; 7 mi. NE Hilger, 1. Rosebud County: 3/4 mi. N Ingomar, 1. County in question, Milk River, 2.

Nebraska. Dawes County: Chadron, 2. Cherry County: Kennedy, 1; no locality more definite than county, 1. Brown County: Long Pine, 1. Antelope County: Neligh, 1. Adams County: Hastings, 2. Clay County: Inland, 4.

North Dakota (arranged by counties from west to east). Divide County: Crosby, 1. Mountrail County: Lostwood, 1. Little Missouri River, 1. Golden Valley County: Sentinel Butte, 1. Billings County: Medora, 1. McLean County: 3 mi. W Elbowoods, 1. Oliver County: Ft. Clark, 2. Morton County: Mandan, 1. Sioux County: 3 mi. N Cannonball, 1. Logan County: 6 mi. SW Napoleon, 1. Rolette County: Turtle Mts., 1; Fish Lake, 1. Benson County: Ft. Totten, 3; Sully Hill Nat. Park, 1. Ramsey County: Devils Lake, 2. Stump Lake or Turtle Mts., 2. Nelson County: Stump Lake, 1. Grand County: Larimore, 1. Walsh County: Grafton, 11 (4, 3, 2). Stutsman County: Jamestown, 1. Barnes County: Valley City, 1.

Saskatchewan. Wingard, 5; Osier, 2; Simpson, 1; Touchwood Hills, 4; South arm Last Mountain Lake, 1; Rush Lake (Assiniboia, N.W.T.), 2.

South Dakota. Pennington County: Rapid City, 1.

Wyoming. Goshen County: Fort Laramie, 2.

Mustela frenata oribasus (Bangs)

Long-tailed Weasel

Plates , , , , , and

Putorius (Arctogale) longicauda oribasus Bangs, Proc. New England Zoöl. Club, 1:81, December 27, 1899.

Putorius longicauda, Coues, Fur-bearing animals, p. 136, 1877 (part).

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

Mustela longicauda oribasa, Hall, Univ. California Publ. Zoöl., 40:368, November 5, 1934.

Mustela frenata oribasa, Hall, Carnegie Instit. Washington Publ. 437:105, November 20, 1936.

Type.—Female, adult, skull and skin; no. 9058, collection of E. A. and O. Bangs, but now in collection of Mus. Comp. Zoöl.; source of Kettle River, 7500 feet [the summit between middle fork of Kettle River and Cherry Creek at Pinnacles—oral information from the collector, Feb. 12, 1936], British Columbia; September 10, 1898; obtained by Allan Brooks; original no. 1368.

The skull (plate ) is complete and unbroken. The teeth all are present and entire except right I which has the anterior half broken away. The skin is complete, fairly well made, and in summer pelage.

Range.—Canadian and Hudsonian life-zones from near 56°N in the Rocky Mountains of British Columbia and Alberta and Ootsa Lake along the Fraser and Chilcotin rivers south to Alta Lake, in the Caribou and Monashee mountains, probably in the Selkirks and Rockies, and through the Rockies of Montana into extreme northern Wyoming. See figure on page .

Characters for ready recognition.—Differs from M. f. longicauda by near (14 n) Brussels Brown rather than near (h) Clay Color of upper parts and in males by relatively shallower 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 59 per cent of the mastoid breadth; from M. f. nevadensis by greater average size, see measurements.

Description.—Size.—Male: Two adults from Florence, Montana, measure as follows: Total length, 440, 440; length of tail, 165, 161; length of hind foot, 47, 49. Corresponding measurements of an adult male from Quesnel, British Columbia, are: 443; 168; 55. Tail amounts to 60, 58, and 61 per cent as long as head and body. Length of hind foot averages more than basal length.

Female: The type specimen, the only typical adult or subadult specimen of this sex of which external measurements are available, measures: Total length, 392, length of tail, 150, length of hind foot, 46. Tail is 63 per cent as long as head and body. Length of hind foot amounts to more than basal length.

The differences in external measurements, between the one female and the average of the three males are: Total length, 49; length of tail, 15; length of hind foot, 4.

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

Color.—Upper parts, in summer, near (14 n) Brussels Brown, more blackish and less reddish than tone 4 of Burnt Umber of Oberthür and Dauthenay, pl. 304; in type near tone 4, pl. 301 of Oberthür and Dauthenay. Underparts, in summer, Buff Yellow or near (20 c) Amber Yellow. In winter, all white except tip of tail which is at all times black. Upper parts of uniform color except for occasional slight darkening of top of head and along mid-dorsal line of back. Color of underparts extends distally on posterior sides of forelegs over feet, on medial sides of hind limbs over antiplantar faces of toes and over proximal two-thirds of ventral side of tail. Least width of color of underparts amounting to 43 per cent of greatest width of color of upper parts, 75 per cent in male from 4 miles northeast of Quesnel, British Columbia, and 52 (33-66) in four males from Montana. Black tip of tail in four males from Montana averaging 50 (44-60) mm. long. Thus averaging approximately as long as hind foot and 33 per cent of length of tail-vertebrae.

Color not different than in many specimens of M. f. nevadensis. Color comparison with M. f. longicauda has been made in the account of that subspecies.

Skull and teeth.—Male (based on 5 adults and 2 subadults from British Columbia and 4 adults from Montana): See measurements and plates -. As described in Mustela frenata longicauda except that: Weight, 5.0 (3.8-6.0) grams; basilar length, 46.7 (43.6-48.8); postorbital breadth in one of nine instances less than width of basioccipital measured from medial margin of one foramen lacerum posterior to its opposite; interorbital breadth more or less than distance between foramen opticum and anterior margin of tympanic bulla; breadth of rostrum less than length of tympanic bulla; anterior margin of tympanic bulla as far posterior to foramen ovale as width of 2-1/2 to 5 upper incisors; length of tympanic bulla not less than length of lower molar and premolar tooth-row and shorter than rostrum.

Female (based on the type, specimen): See measurements and plates -, 40. As described in Mustela frenata longicauda except that: Weight, 3.5 grams; basilar length, 41.6 mm.; zygomatic breadth more than distance between anterior palatine foramen and anterior margin of tympanic bulla; postorbital breadth more than width of basioccipital measured from medial margin of one foramen lacerum posterior to its opposite; least width of palate more than outside length of P; tympanic bulla as far posterior to foramen ovale as width of 4-1/2 upper incisors; height of tympanic bulla less than distance from anterior margin of tympanic bulla to foramen ovale; length of tympanic bulla less than length of rostrum. If more than one skull were available of the female of oribasus it is believed that the description would agree with that of longicauda in nearly all features.

The skull of the female is 30 per cent lighter than that of the average male.

Comparison with longicauda reveals that, on the average, skulls of males are larger, relative to the basilar length broader across the mastoids, shallower through the braincase as measured at the anterior end of the basioccipital exclusive of the sagittal crest, with longer rostrum. Compared with nevadensis, the skull averages larger in all measurements taken, and has a relatively broader rostrum, relatively greater mastoid breadth and a braincase which is shallower relative to the basilar length. By weight, the skull of nevadensis is a fourth lighter, and in linear measurements 5 to 18 per cent smaller.

Remarks.—Some of the specimens from Montana, which here are referred to oribasus, more than half a century ago were listed by Coues (1877:138) under the name longicauda. It was not until 1899 that this race was given a name by Bangs, who at that time (1899B:81) accurately made out the distinctive color features. Distinctive cranial characters cannot be described with assurance even now because there still are too few specimens.

The type specimen, at one time examined by the present writer, has on the stuffed skin no well-developed mammae, scrotal pouch, or other visible sexual part. Probably the collector's sex mark for female is correct.

As judged by the two skulls of subadult males from the Barkerville region, individuals of this race attain larger size than do those of longicauda. On the basis of larger size than either longicauda or nevadensis, the specimens from the Rocky Mountains of Montana and two from northern Wyoming are referred to this race. The short, wide, flat, tympanic bullae, relatively great mastoidal breadth, and some other features of the specimen from Donovan, Montana, point toward oribasus, whereas nearly as many more cranial features, in this instance mainly differences in size, are indicative of nevadensis to which race the specimen might almost equally well be referred. Another male from Darby, in the Bitterroot Valley of Montana, has a slightly longer hind foot than those from Florence, but a female from Hamilton, agrees more nearly with nevadensis. The average of all the specimens from the Bitterroot Valley is a little nearer oribasus. Four skulls from Buffalo, Wyoming, here referred to nevadensis show approach to oribasus in size of skull. The specimens from Big Snowy Mountains, and the Highwood Mountains of Montana are too young clearly to show size of the adult skull, but are distinctly darker colored than longicauda of the plains country proper. Of two subadult females from Tacy, Montana, the color of the one in summer pelage is distinctly nearer that of oribasus and nevadensis than it is to that of longicauda to which some approach in color might be expected. The reduced size of both of the specimens is further suggestive of nevadensis and it may be that adult specimens from these more eastern mountainous areas in Montana will show that nevadensis is the name proper to apply to animals of this region.

Intergradation with nevadensis is suggested by specimens collected from along the upper reaches of Okanagan Lake, British Columbia, by Major Allan Brooks and Mr. J. A. Munro and by a series of skulls from Ione, Pend Orielle County, Washington, lent me by Mr. Walter Dalquest. At each place, the average of all specimens is nearest to that of nevadensis.

Specimens from near Waterton Lake show several steps in the transition from the light-colored longicauda type of coloration to the darker coloration characterizing oribasus. One taken here, at a time when the body of water referred to seems to have been known as Chief Mountain Lake, is barely dark enough to be placed with oribasus. Two other specimens from across the Canadian Border labeled as "Waterton Lake Park" are slightly lighter colored above, and on this account are placed with longicauda.

The two adult males from Lillooet, British Columbia, are referable to oribasus although neither is quite typical. One has a saturated coloration suggestive of that of altifrontalis and the skull is shorter and broader than in other specimens of oribasus. The female from Lillooet, skin alone, no. 916, Prov. Mus., B. C. is small for oribasus. The female, no. 1539, collection of Kenneth Racey, from Alta Lake, in brown winter pelage, in almost every measurement falls nearly midway between altifrontalis and oribasus but slightly nearer the latter. The skull from Chezacut and 3 animals from Wistaria, British Columbia, probably are females and show a greater average size than specimens from farther to the southeast. For example, the basilar length of the skull, 44.8 (44.3 to 45.1), exceeds that of the type specimen. The animals from Wistaria on Ootsa Lake furnish the northwesternmost station of occurrence of which I have record for this subspecies.

The northernmost records of occurrence, at "Clearwater River, Peace River, B. C," and at Little Prairie, are furnished by a white skin without skull, no. 257450, U. S. Nat. Mus., purchased on August 2, 1932, at the place mentioned by W. H. Sheldon and Richard Borden, and a skull with white winter skin, no. 3585, Provincial Museum, British Columbia, respectively. The characters distinguishing longicauda and oribasus are not shown by white winter skins; the skull shows some features of longicauda, and the reference of these specimens to oribasus rather than longicauda is tentative.

Only the skull from Little Prairie shows evidence of infestation of the frontal sinuses by parasites. In the Barkerville area of British Columbia, Mr. and Mrs. Thomas T. McCabe obtained only 2 skulls of this subspecies from a total of 238 weasel skulls gathered by local trappers. The others were Mustela erminea.

Specimens examined.—Total number, 46, listed by localities from north to south and unless otherwise indicated, in the United States National Museum.

British Columbia. West of Hudson Hope, 1; Clearwater River, tributary to Peace River, 1; Little Prairie, a few miles south of Peace River and about 40 miles west of the main highway between Dawson Creek and Fort St. John, 1; Wistaria, 3; Four Mile Creek, 4 mi. NE Quesnel, 1; Isaacs Lake, 3200 ft., 1; Barkerville region, 1; Clear River, 4800 ft., 1; Chezacut, 1; Lillooet 3 (2, 1); Alta Lake, 1; source of Kettle River, 7500 ft., 1; E side Beaverfoot Range, 4000 to 4500 ft. between Fraser Creek and 6 mi. SE of Fraser Creek, 1; Cranbrook, 1; head of Cross River, 10 mi. below Assiniboine Pass, 1; camp east of "Kootanie," 1; camp east of Kootanie River, 1.

Alberta. Thoral Creek, 7000 ft., 50 mi. NE Jasper, 1.

Montana. Glacier? County: Chief Mt. Lake (= Waterton Lake), 1. Flathead County: Columbia Falls, 1. Chouteau? County: Highwood Mts., 1. Fergus? County: Big Snowy Mts., 1. Wheatland County: Harlowton, 1. Ravalli County: Florence, 2; Hamilton, 1; Darby, 1; Carlos [= Charlos] Heights, 2; Tin Cup District, 2; no locality more definite than county, 2. Beaverhead County: Donovan, 1. Madison County: Sheridan, 1. Gallatin County: Ranch 7-11, Eldridge, 1. Stillwater County: Tacy, 2. County in question: Gallatin Valley, 1; Yellowstone Park, 1.

Wyoming. Glen Creek, Mammoth Hot Springs, 1. Park County: Four Bears, 1.

Mustela frenata alleni (Merriam)

Long-tailed Weasel

Plates , , , , and

Putorius alleni Merriam, N. Amer. Fauna, 11:24, June 30, 1896.

Mustela alleni, Miller, U. S. Nat. Mus. Bull., 79:99, December 31, 1912.

Mustela frenata alleni, Hall, Carnegie Instit. Washington Publ. 473:106, November 20, 1936.

Type.—Male, adult, skull and skin; no. 186451, U. S. Nat. Mus. (formerly 4485/5120, collection of Dr. C. Hart Merriam); Custer, South Dakota; obtained by Vernon Bailey; original no. 90.

The skull is complete and unbroken. The upper incisors are missing. All the other teeth are present although the premolars, and especially the canines, are much worn, possibly as the result of the animal's efforts to free itself from a trap. The skin is fairly well made, in a good state of preservation, and entire.

Range.—Canadian, Transition and Upper Sonoran life-zones of the Black Hills of South Dakota and adjacent semi-bad-land territory of Wyoming and Nebraska southward to Mitchell, Scottsbluff County. See figure on page .

Characters for ready recognition.—Differs from M. f. longicauda in smaller size, adult males having a total length of less than 400, hind foot less than 45, basilar length less than 43.5, and in adult females total length less than 375, and basilar length less than 40; from M. f. nevadensis in near Clay Color rather than near (14 n to l) Brussels Brown of upper parts in summer.

Description.—Size.—Male: External measurements of the type specimen are: Total length, 372; length of tail, 137; length of hind foot, 44. Tail is 58 per cent as long as head and body. Length of hind foot more than basal length.

Female: No external measurements for typical adults are available. No. M1 #41 from Mitchell, Scottsbluff Co., Nebraska, an adult female which is an intergrade with the larger M. f. longicauda, measures as follows: Total length, 367; length of tail, 120; length of hind foot, 41.

Externals.—Longest facial vibrissae dark brown or white 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 .

Color.—Winter pelage unknown; probably white except, of course, tip of tail. Summer pelage as described in Mustela frenata longicauda except that: Least width of color of underparts averaging, in 3 males from Black Hills, 54 (38-62) per cent of greatest width of color of upper parts. Black tip of tail averaging 43 (40-45) mm. long. Thus, averaging approximately same length as hind foot and in type specimen amounting to 33 per cent of length of tail-vertebrae.

Skull and teeth.—Male (based on the type and no. 7440 Amer. Mus. Nat. Hist., from Hill City, S. Dak.): See measurements and plates -. As described in Mustela frenata longicauda except that: Weight, 3.1 (3.0-3.2) grams; basilar length, 41.0 (40.9-41.0); mastoid breadth not less than postpalatal length; breadth of rostrum more than length of P4; anterior margin of tympanic bulla as far posterior to foramen ovale as width of 4 to 5 upper incisors; height of tympanic bulla more or less than distance from its anterior margin to foramen ovale.

Female (based on no. 7441, American Mus. Nat. Hist., from Black Hills, S. Dak.): See measurements and plates -. As described in Mustela frenata longicauda except that: Weight, 2.0 grams; basilar length 37.6. The skull of the female is 35 per cent lighter than the average for the two males.

Comparison with M. f. longicauda and M. f. nevadensis reveals that the tympanic bullae average more nearly flat and that the skull is smaller.

Remarks.—Animals of this subspecies were described and named by Merriam in 1896 as a distinct species on the basis of two or possibly three specimens from the Black Hills of South Dakota and the name seems never to have been applied to specimens from other regions. Vernon Bailey obtained only the one specimen, the type, on his trip in 1888, but two more were obtained for the American Museum of Natural History by Walter Granger in 1894.

Mustela frenata alleni combines the light coloration of M. f. longicauda with the small size of M. f. nevadensis. Indeed, the size may average less than that of nevadensis. M. f. alleni seems to reach its extreme of small size in the Black Hills of South Dakota. Specimens from Mitchell, Scottsbluff County, Nebraska, here referred to alleni are of larger size and in this respect are intermediate between the subspecies alleni and longicauda. Of the two specimens available from Chadron, Nebraska, and here referred to as longicauda, the female, M1 #6, is almost exactly intermediate in size between alleni and longicauda, whereas the male, M1 #11, is as large as the average-sized longicauda.

None of the nine skulls (5 adults) shows malformation resulting from the infestation of the frontal sinuses with parasites.

Specimens examined.—Total number, 10, as follows.

Wyoming. Crook County: Sundance, 1.

South Dakota. Pennington County: Hill City, 1; 20 mi. N Elk Mt, 1. County in question: Black Hills, 1. Custer County: Custer, 2 (1, 1).

Nebraska. Scottsbluff County: Mitchell, 4.

Mustela frenata arizonensis (Mearns)

Long-tailed Weasel

Plates , , , , and

Putorius arizonensis Mearns, Bull. Amer. Mus. Nat. Hist., 3:234, June, 1891; Merriam, N. Amer. Fauna, 11:22, fig. 12, June 30, 1896.

Mustela arizonensis, Miller, U. S. Nat. Mus. Bull., 79:99, December 31, 1912.

Mustela frenata arizonensis, Hall, Carnegie Instit. Washington Publ. 473:106, November 20, 1936.

Type.—Female, adult, skull and skin; no. 2490/1886, Amer. Mus. Nat. Hist.; San Francisco Forest [then (1886?), Yavapai County], Arizona; June 20, 1886; obtained by Edgar A. Mearns.

The skull (plates -) is complete and unbroken save for a small puncture in the right squamosal. The incisors above and below and M and P on each side are missing. Four canines are preserved separately. Otherwise the teeth are in place. The skin has been taken down from a mount. Some hair has been lost from in front of the ears. Seven mammae are evident and show the animal to have been nursing young. The slightly faded color was mentioned by Mearns in the original description. He says (1891:234): "The memorandum of the colors was made before skinning, the specimen having been subsequently preserved in a solution of alum and salt, which extracted much of the coloring matter."

Range.—Transition to Hudsonian life-zones of Arizona and extreme western New Mexico, along the Colorado River, and south of the Little Colorado River, from San Francisco Mountain region along Mogollon Plateau to extreme western New Mexico. See figure 29 on page .

Characters for ready recognition.—Differs from M. f. neomexicana by near (14 n) Brussels Brown rather than Buckthorn Brown color of upper parts, in absence rather than presence of white frontal spot continuous with color of underparts, in basilar length of less than 44 in males and 39.3 in females; from M. f. nevadensis in that total length averages less than 375 in males and 330 in females, basilar length averaging less than 41 in males and less than 36.7 in females.

Description.—Size.—Male: No. 24679/32071, from Springerville, and no. 248993 from the Kaibab Plateau, measure respectively, as follows: Total length, 363, 367; length of tail, 140, 143; length of hind foot, 41.5, 41.0. Tail is 63, and 64 per cent as long as head and body. These males, the only specimens of that sex of which external measurements are available, probably are grading toward nevadensis and therefore are nontypical.

Female: Three specimens, one young from Little Spring, a subadult from Deadmans Flat and the type specimen, measure respectively as follows: Total length, 323, 296, 302; length of tail, 110, 101, 109; length of hind foot, 38, 33, 36. These average, 307, 107, 36. Tail averages 53 per cent as long as head and body.

Differences in external measurements of the two sexes are: Total length, 56; length of tail, 39; hind foot, 5.5.

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) about as shown in figure .

Color.—Winter pelage unknown. Summer pelage with upper parts near (14 n) Brussels Brown or tone 2 of Raw Umber of Oberthür and Dauthenay, Pl. 301, darker on top of head from nose to line connecting posterior margins of ears. Tip of tail always black. Chin and upper lips white. Remainder of underparts Buff Yellow to Straw Yellow and rarely Ochraceous Buff. Color of underparts extends distally on posterior sides of forelegs over toes onto antipalmar faces of feet and wrists, on medial sides of hind legs to ankles and over antiplantar faces of toes, medial third of tarsus, and over proximal fifth to fourth of ventral side of tail. Least width of color of underparts averaging, in 8 specimens, 44 (29-54) per cent of greatest width of color of upper parts. Black tip of tail, in four females averaging 35 (33-38) mm. long. Thus, averaging shorter than hind foot and 32 per cent of length of tail-vertebrae. Three of the eight specimens before me (no. 242671 from 25 mi. SE Flagstaff, not available at time of this accounting) have the dark spot near the angle of the mouth faintly indicated, whereas the other five lack the spots. The color is as in M. f. nevadensis.

Skull and teeth.—Male (based on 55211, 65231, and 248993; see page ): See measurements and plates -; weight 2.7 and 3.1 grams; basilar length, 40.4; 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 more or less than distance between foramen opticum and anterior margin of tympanic bulla; breadth of rostrum less than length of tympanic bulla; least width of palate more or less than medial length of P4; anterior margin of tympanic bulla as far posterior to foramen ovale as width of 3-1/2 (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-premolar tooth-row and longer or shorter than rostrum; anterior margin of masseteric fossa below talonid of m1.

Female (based on the type specimen): See measurements and plates -; weight, 1.6 grams; basilar length, 35.5; zygomatic breadth less than distance between condylar foramen and M1 and more than distance between anterior palatine foramen and anterior margin of tympanic bulla (nearly equal in each instance); postorbital breadth less than length of upper premolars and greater (7.1-8.4) than width of basioccipital measured from medial margin of one foramen lacerum posterior to its opposite; least width of palate equal to inside length of P4; tympanic bulla as far posterior to foramen ovale as width of 3 (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-premolar tooth-row and greater than length of rostrum.

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

Compared with the skull of M. f. nevadensis, that of arizonensis is smaller, less heavily ridged and has more inflated tympanic bullae and a relatively greater mastoid breadth. Comparison with the skull of M. f. neomexicana is made in the account of that subspecies.

Remarks.—In 1891 Mearns (234-235) named this weasel as a full species on the basis of two individuals taken by him in 1886 and 1887. Since that time only a few additional specimens have been preserved. Only four are adults. Although this material does not permit of a definition of the subspecies as precise as could be wished, still, it clearly shows that the animals from the plateau region of Arizona are recognizably different from those farther north in the Sierra Nevada of California and those of the Rocky Mountains and Great Basin region northward to the Canadian border. These more northern animals have gone by Mearns' name, arizonensis, since the date of its proposal until 1939 when the name nevadensis was proposed.

The smaller size, especially of the skull, and the greater inflation of the tympanic bullae are the outstanding characters which distinguish arizonensis from the similarly marked nevadensis. The bullae are relatively much inflated throughout but especially so on the posteromedial parts.

Although the three adult males and two subadult females available of this subspecies are smaller in most parts measured than any of the scores of nevadensis of similar age that have been measured, overlap in size probably will be found as additional specimens of arizonensis become available. A young female, no. 18513, coll. D. R. Dickey, from Little Spring, does have certain cranial measurements as large as are found in the minimum-sized nevadensis from farther north.

Intergradation with the two subspecies whose geographic ranges adjoin that of arizonensis is indicated by specimens at hand. One of these is the adult male from 25 miles southeast of Flagstaff, which shows decided approach to neomexicana, in color and in possessing white facial markings less well developed than in neomexicana. Even better developed white facial markings, with intervening blackish coloration, are displayed by no. 148271, U. S. Nat. Mus., from 8500 feet altitude on Willow Creek, New Mexico. This subadult female shows approach to neomexicana also in larger size of the skull and entire animal. The great inflation of the posterior part of each of its bullae and the dark color of the upper parts are characters of arizonensis. The color of the underparts stops at the ankles leaving the hind feet dark colored, in which respect the specimen is unlike either neomexicana or arizonensis. If additional specimens showing the same characters as this one be found at other nearby localities they probably should be given recognition as a separate subspecies. For the present it seems best to regard the specimen merely as an intergrade. Although it might, with almost equal propriety, be referred to either neomexicana or arizonensis, the specimen is here placed with the latter. The subadult male from Springerville, Arizona, is of larger size than the topotypical male of arizonensis and in this respect shows slight approach to nevadensis. The narrower mastoidal breadth and slightly less inflated tympanic bullae of the male from the Kaibab Plateau may reflect merely individual variation or may represent intergradation in these features with nevadensis.

The statement made by Merriam (1896:22) that, "The type specimen . . . is an immature female and is of unusually small size. A male obtained by him [Mearns] near the same place is of the normal size, as is another male in the Department collection from Springerville, Ariz., collected by E. W. Nelson," needs correction. The female is not immature. The specimen obtained by Mearns near the same place probably refers to Amer. Mus. No. 2489, from Quaking Asp Settlement, which lacks both the skull and external measurements. As stuffed it is of small size for a male. The male from Springerville, as shown by the external and cranial measurements, is not of normal (i. e. average) size, but is smaller than the average for the other populations of similarly colored weasels referred to by Merriam (op. cit.) as arizonensis but here described under the name nevadensis.

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

Specimens examined.—Total number, 17, arranged alphabetically by states and from north to south by counties in each state. Unless otherwise indicated specimens are in the collection of the United States National Museum.

Arizona. Coconino County: VT Park, Kaibab Plateau, 1; Deadman Flat, 6400 ft., 1; Little Spring, 1; Government Prairie, near Parks, 1; Coconino? County: San Francisco Forest (Yavapai Co., in 1886), 1; 25 mi. SE Flagstaff, 1; Quaking Asp Settlement, 1. Apache County: Springerville, 1; North Fork White River, White Mts., 8200 ft., 4; head San Francisco River, Judd Ranch, Alpine, 1; 2 mi. SE Big Lake Knoll, 8700 ft., 24 mi. S Springerville, 1. Greenlee County: S end Blue Range, 9000 ft., Prieto Plateau, 1; Beaver Creek, 7000 ft., 1.

New Mexico. Grant County: Mogollon Mts., Willow Creek, 8500 ft., 1.

Mustela frenata nevadensis Hall

Long-tailed Weasel

Plates , , , , , and

Mustela frenata nevadensis Hall, Carnegie Instit. Washington Publ. 473:91, November 20, 1936.

Putorius longicauda, Coues, Fur-bearing animals, p. 136, 1877 (part); Merriam, N. Amer. Fauna, 5:83, July 30, 1891.

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

Putorius arizonensis, Merriam, N. Amer. Fauna, 11:22, figs. 13, 14, June 30, 1896 (part); Stephens, Mammals of California, p. 247, 1906.

Mustela arizonensis, Grinnell and Swarth, Univ. California Publ. Zoöl., 10:376, October 31, 1913; Whitlow and Hall, Univ. California Publ. Zoöl., 40:247, September 30, 1933.

Mustela arizonensis arizonensis, Grinnell, Univ. California Publ. Zoöl., 40:102, September 26, 1933.

Mustela frenata, Boyer, Journ. Mamm., 24:99, February 20, 1943.

Type.—Female, adult, skull and skin; no. 41053, Mus. Vert. Zoöl.; three miles east Baker, White Pine County, Nevada; May 30, 1929; obtained by E. R. Hall and W. C. Russell; original no. 2674, E. R. H.

The skull (plates -) is complete and unbroken. The teeth all are present and entire. The skin is fairly well made. Eight mammae are evident and show the animal to have been nursing young.

Range.—Altitudinally, 700 feet at Wenatchee, Washington, to the highest parts of the mountains of the western United States; Upper Sonoran Life-zone to Arctic Alpine Life-zone; southern British Columbia in the Cascades and territory west to Monashee Mountains, and Nelson, southward in the Cascades of northern Washington, over western Washington, Idaho, Utah, and Nevada to northeastern Arizona and northern New Mexico; westward from the eastern base of the Rocky Mountains in Colorado to the western base of the Sierra Nevada and Cascades of California and to the Cascades of southern Oregon. See figures and on pages 221 and 314.

Characters for ready recognition.—Differs from M. f. oribasus by smaller average size, see measurements; from M. f. longicauda by near (14 n to l) Brussels Brown rather than near (h) Clay Color of the upper parts, and in males by a shallower 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 59 per cent of the mastoid breadth; from M. f. alleni by near (14 n to l) Brussels Brown rather than near (h) Clay Color of upper parts in summer; from M. f. neomexicana by near (14 n to l) Brussels Brown rather than Buckthorn Brown color of upper parts, in absence of white frontal spot continuous with color of underparts, in basilar length of less than 46 in males and 40 in females; from M. f. arizonensis by total length averaging more than 375 in males and 330 in females, basilar length averaging more than 41 in males and 36.7 in females; from M. f. inyoensis by absence of white facial markings; from M. f. pulchra by absence of light facial markings, near (14 n to l) Brussels Brown rather than near (16 j) Buckthorn Brown color of upper parts, and lesser size, hind foot less than 40 in females and basilar length averaging less than 46.0 in males; from M. f. xanthogenys by absence of light facial markings and near (14 n to l) Brussels Brown rather than Buckthorn Brown color of upper parts; from M. f. munda by absence of white facial markings, presence of color of underparts on ventral face of proximal third of tail, and hind foot of less than 50 in males; from M. f. saturata by presence of light color of underparts on tail and ankle and in lesser average breadth across mastoid processes of skull (see measurements); from M. f. oregonensis by absence of nasofrontal white patch, presence of light color of underparts on ventral face of tail, and shorter skull, which, relative to its length in males, is deeper through the braincase; from M. f. washingtoni by presence of light color of underparts on ventral face of tail, by skull which in male relative to basilar length is shorter in the preorbital region and wider across the zygomata and mastoid processes, and in female has longer preorbital region and larger bullae (see measurements); from M. f. altifrontalis by lighter colored upper parts which are tones 1 to 3 of Raw Umber, pl. 301, rather than tone 4 of Brownish Drab, pl. 302, of Oberthür and Dauthenay, by Buff-Yellow to Straw Yellow rather than near (14´ a to 16´ c) Ochraceous-Buff color of underparts, by least width of color of underparts amounting to more than 37 per cent of greatest width of color of upper parts, by presence of color of underparts on ventral side of tail and on hind leg over ankle, and by lesser depth of skull through frontal region; from M. f. effera by larger size, males averaging 12-1/2 per cent larger in external measurements, 8 per cent larger in linear measurements of skull, and 22 per cent heavier in weight of skull, total length averaging 400 rather than 360, basilar length averaging 43.6 rather than 40.5.

Description.—Size.—Male: Twenty-one adults from the southern half of the Sierra Nevada of California yield average and extreme measurements as follows: Total length, 400 (356-428); length of tail, 150 (125-178); length of hind foot, 46.1 (42-50). Tail averages 60 per cent as long as head and body. Length of hind foot averaging more than basal length. Corresponding measurements of twelve adults from extreme southern and southwestern Colorado are as follows: 407 (355-431); 150 (133-170); 46.0 (42-49).

Female: Ten adults from the Sierra Nevada of California yield average and extreme measurements as follows: Total length, 349 (336-362); length of tail, 127 (120-133); length of hind foot, 36.3 (32-39). Tail averages 57 per cent as long as head and body. Length of hind foot less than basal length. Corresponding measurements of ten adults from the Rocky Mountains of central Colorado are as follows: 347 (325-375); 123 (111-141); 40 (32-43).

The average differences in external measurements of the two sexes, in the Sierras of California are: Total length, 51; length of tail, 23; length of hind foot, 9.8. Weight of 7 adult males from California is 267 (226-345) grams. Two adult females from there weigh 148 and 115 grams and 3 from White Pine County, Nevada, 134, 122 and 124, giving an average of 129 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) about as shown in figure .

Color.—Upper parts, in summer, near (14 n to l) Brussels Brown or tones 1 to 3 of Raw Umber of Oberthür and Dauthenay, pl. 301, darker on top of head from nose to line connecting posterior margins of ears. Chin and upper lips white. Remainder of underparts Buff-Yellow to Straw Yellow and sometimes Ochraceous-Buff especially in young, and in some adults from southern Colorado. In winter, all white, except tip of tail, or upper parts near (j) Snuff Brown or lighter than Brussels Brown with a smoked effect, and underparts white. Tip of tail at all times black. Color of underparts extends distally on posterior sides of forelegs over toes onto antipalmar faces of feet and wrists, on medial sides of hind legs to ankles, over antiplantar faces of toes, medial third of tarsus and usually over proximal tenth to three-fourths of ventral side of tail. Least width of color of underparts averaging, in a series of twenty males from the southern half of the Sierra Nevada of California, 59 (37-76) per cent of greatest width of color of upper parts. In seven males from southern Colorado corresponding percentages are 55 (37-71). Black tip of tail in series from Sierra Nevada averaging 50 (40-60) mm. long; thus longer than hind foot and averaging 33-1/3 per cent of length of tail-vertebrae.

Skull and teeth.—Male (based on 25 adults, from Sierra Nevada of California): See measurements and plates -; weight, 3.7 (2.9-4.9) grams; basilar length, 43.6 (40.6-46.1); zygomatic breadth more than distance between condylar foramen and M1 (save in four instances) and more than distance between anterior palatine foramen and anterior margin of tympanic bulla (save in two specimens); mastoid breadth more (80 per cent of specimens) or less (20 per cent) than postpalatal length; 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; interorbital breadth more or less 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 medial length of P4 (except in two specimens); anterior margin of tympanic bulla as far posterior to foramen ovale as width of 3 to 5 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 not carried farther forward than point directly below hypoconid of m1.

Female (based on ten adults from Sierra Nevada of California): See measurements and plates -; weight, 2.2 (1.8-2.4) grams; basilar length, 38.2 (36.7-39.5); zygomatic breadth more (except in one specimen) than distance between condylar foramen and M1 and more (save in two specimens) than distance between anterior palatine foramen and anterior margin of tympanic bulla; postorbital breadth less than length of upper premolars and less than (except in one specimen) width of basioccipital measured from medial margin of one foramen lacerum posterior to its opposite; least width of palate more or less than either outside or inside length of P4 but generally less than inside length; tympanic bulla as far posterior to foramen ovale as width of 3 to 5-1/2 upper incisors; height of tympanic bulla more or 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 more or less than length of rostrum.

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

Compared with the skull of M. f. longicauda, that of both sexes averages smaller in every measurement taken. Males of nevadensis, on the average, relative to the basilar length, are narrower in the interorbital region and across the zygomata but have the orbitonasal length greater. Stated in another way, the rostrum of longicauda appears to be shorter and broader and the zygomata are more expanded. Females of nevadensis, on the average, relative to the basilar length are narrower across the mastoid processes and zygomata and have the braincase deeper at the anterior margin of the basioccipital. Also in nevadensis the mastoid processes do not project so far laterally beyond the braincase, the lambdoidal crest and postorbital processes are less well developed and except in the interparietal region, the temporal ridges hardly meet and they form a sagittal furrow rather than a low sagittal crest which characterizes adult females of longicauda. Each of these differences separating the females of longicauda from those of nevadensis are of the same nature, although not necessarily of the same degree, as those which appear in longicauda with increasing age. The differences mentioned above are readily appreciable when series of specimens are compared. However, none of the differences is of great degree, and most parts of the skulls of the two subspecies are of similar relative proportions. Even so, there is but little overlap in actual size. Comparisons with the skulls of M. f. oribasus, alleni, neomexicana, arizonensis, inyoensis, pulchra, xanthogenys, munda, saturata, oregonensis, washingtoni, altifrontalis, and effera are made in the accounts of those subspecies.

Remarks.—The populations to which the name nevadensis at present is assigned have gone by the name arizonensis since Mearns proposed this name in 1891. Before that time Coues (1877:141) had included individuals of this race under the name Putorius longicauda.

Among the populations here assigned to M. f. nevadensis, there is some geographic variation but it is of lesser degree than in most other species of mammals which range over the same region. Comparison of 20 adult males from the Rocky Mountains of Colorado with 25 adult males from a place as far distant as the Sierra Nevada of California shows that the two populations closely resemble each other. The specimens from Colorado average a trifle wider across the zygomata, have a longer body and therefore relatively shorter tail, and, except in southern Colorado, a slightly longer hind foot. Comparison of ten adult females from each of the two areas reveals that those from Colorado have a markedly longer hind foot, and a tail somewhat shorter relative to the length of the body. The mentioned differences are the only ones found among the great number of points investigated, except that as remarked by Merriam (1896:23) the Sierran animal has the yellow of the underparts reaching farther up under the chin, the underside of the tail on the average is more suffused with yellowish and the white on the upper lip is more extensive. As regards the last mentioned feature, my check of 34 skins from Colorado reveals that the white extends all the way around the upper lip in every specimen but one, whereas in 69 specimens from the Sierra Nevada the white extends all the way around the upper lip in only 39. However, as further remarked by Merriam (loc. cit.), not only this but the other color features are inconstant in addition to being slight. When the occurrence of the dark spots near the angles of the mouth are tabulated, it is found that in 33 Colorado-taken specimens they are absent in 19, faintly indicated in 13, and well developed in 1. In 62 California-taken specimens they are absent in 37, faintly indicated in 20, and well developed in 5.

In northwestern Colorado, southern Wyoming, and possibly through the Bear River Divide into southeastern Idaho, long-tailed weasels here referred to nevadensis approach longicauda in large size and occasionally in other features, more closely than do specimens of nevadensis from most other places in its range. This tendency is thought to be significant for much of the area in question lies in or below the Transition Life-zone, the same life zones in which farther to the eastward true longicauda occurs.

One specimen that illustrates this approach to longicauda is an adult male, no. 2334, collection of E. R. Warren, from 6160 feet, Lay, Routt [now Moffat] County, Colorado. In large size and, relative to the basilar length, shorter rostrum and shorter tympanic bullae, it agrees with longicauda but the darker color and, relative to the basilar length, narrowness of the rostrum, interorbital region, zygomatic expanse and the shallowness through the region of the postorbital processes place it with nevadensis. Of two other specimens from Steamboat Springs, Routt County, a young male, no. 4010, in the collection of E. R. Warren, has a hind foot (50 mm.) as long as in longicauda; and the other, no. 138195, U. S. Nat. Mus., an adult male, agrees well enough in size and proportions with nevadensis but has the coloration typical of longicauda.

From Wyoming, one subadult female, no. 177553, U. S. Nat. Mus., from Garrett, is intermediate in size and coloration but is nearer to nevadensis in these particulars, as it is in all other points considered except size of the molar teeth which are as large as in longicauda and larger than in any female nevadensis from Colorado or California. Another female, an adult, no. 179304, U. S. Nat. Mus., from Lonetree, Wyoming, agrees with longicauda in size of skull. Indeed, ten of seventeen cranial measurements exceed the maximum for Colorado-taken nevadensis. Where differences exist in relative proportions of the skull as expressed in percentages of the basilar length, the specimen approaches nevadensis in 5 instances and longicauda in only 3. The color is intermediate but much nearer that of nevadensis with which the animal agrees also in external measurements. Ten subadults (5 of each sex) from within 12 miles of Laramie (not Fort Laramie) show greater resemblance to nevadensis but definitely approach longicauda. Average external measurements are: ♂, 408, 155, 44; ♀, 361, 134, 40. The two other specimens examined from this general locality, a young female, no. 2711, Mus. Vert. Zoöl., from Fort Bridger, and a subadult female, no. 188377, U. S. Nat. Mus., from Bridger Pass, show no departures from nevadensis of similar age.

The specimens from scattered localities in the Transition Life-zone of northwestern Colorado and southern Wyoming are larger than nevadensis is elsewhere, and also in certain other features resemble longicauda of the plains to the eastward. Everything considered, the animals in question are much more like nevadensis than longicauda. Study of more specimens, especially from Wyoming, might provide grounds for recognizing as a different subspecies the animals in this large area comprising parts of Colorado and Wyoming from which so few specimens now are available. Possibly the name Putorius culbertsoni Coues would apply. Decision on that point will require adequate material from the type locality, Fort Laramie. See discussion of this name under M. f. longicauda.

In southeastern Idaho males are larger than they are at most other places within the range of nevadensis. An average of 7 adults and subadults from Pegram, Montpelier, Springfield, and the vicinity of Pocatello, reveals, when compared with the average of nevadensis from Colorado and that of longicauda from the Great Plains, that this population from southeastern Idaho is nearest to longicauda in linear measurements of the orbitonasal length, mastoid breadth, length of tympanic bullae, and as expressed in percentage of the basilar length, length of tooth-row, breadth of rostrum, and zygomatic breadth. In all other points of size, relative proportions and color, the animals approach nearer to, or actually agree with, nevadensis.

The specimens commented upon clearly show intergradation between nevadensis and longicauda. Similarly, the specimens from Scottsbluff County, Nebraska, here referred to M. f. alleni, by their larger size suggest intergradation of that subspecies with the larger nevadensis-longicauda stock although the approach is more toward longicauda than nevadensis. Between oribasus and nevadensis, however, there is no lack of material showing intergradation. As set forth in the account of oribasus, specimens from Montana are truly intermediate structurally as well as geographically.

Intergradation with washingtoni is shown by specimens from the northern part of the Cascade Range in Chelan and Okanogan counties, Washington. The adult male, U. S. Nat. Mus., no. 235183, from Bald Mountain, is referable to washingtoni on the basis of cranial characters but all the other adult and subadult specimens examined from Chelan and Okanogan counties are nearer nevadensis on the basis of cranial characters. Indeed, some show no approach to washingtoni in cranial characters. As might be expected on geographic grounds, the specimen from Easton, U. S. Nat. Mus., male subadult, no. 116870, shows approach to washingtoni. This is true of the coloration of the hind limbs, small size of the tympanic bullae, and relatively greater length of the preorbital part of the skull. However, the greater width of the light color of the underparts and relatively great breadth across the mastoid processes and zygomatic arches are points of agreement with nevadensis. Similarly, a series of 7 specimens from the Entait River, 20 miles above its mouth, in tone of color is nearer to washingtoni, as is one of the two skulls of adult males in length of the preorbital region. However, in greater breadth of the skull otherwise, and in the relatively great width of the light color of the underparts, the animals are nearer to nevadensis, to which they are here referred. Some of these characters mentioned above in which departure is shown from typical nevadensis are characters that show approach to altifrontalis. This is especially true of the more intense coloration and restriction of the color of the underparts.

Complete intergradation with effera is shown by specimens from southern Oregon. The change from small effera to the larger nevadensis here is gradual; consequently in northeastern California and southern Oregon the size increases gradually to the northward. Specimens showing complete intergradation with oregonensis and saturata are wanting. However, one specimen from Crescent Lake suggests oregonensis in having near (18) apricot yellow underparts such as occur frequently in oregonensis. Also some specimens from northern California approach saturata in having the color of the underparts reduced in the extent to which it reaches out on the under side of the tail. This fact and the consideration that the two races are less different from one another than are other kinds which definitely are known to intergrade leave no doubt but that material from the intervening localities would show complete intergradation.

Intergradation between nevadensis and munda is indicated by specimens from South Yolla Bolly Mountain, Trinity County, which are commented on at greater length in the account of M. f. munda. M. f. inyoensis is so closely related to nevadensis as to leave no doubt that specimens from suitable localities will show actual intergradation. That intergradation occurs directly with the bridled weasel of the interior valleys of California, M. f. xanthogenys, is shown by specimens from along the west-facing flank of the southern part of the Sierra Nevada. Probably intergradation occurs all along the Sierra Nevada on the western slope but specimens are lacking to show this. Weasels are known to occur in the foothill territory and the lesser attention given to this region by mammal collectors than to the higher parts of the mountains may explain the lack of preserved specimens. Individual specimens, here referred to nevadensis, but, showing varying degrees of approach to xanthogenys are as follows: A female from Hume; a male and a female from 8000 feet elevation, Monache Meadows; a male from 9800 feet elevation on the east fork of the Kaweah River; and 7 specimens, probably one family, from one-half mile south of Mineral King, 7850 feet. Of the specimens from 7850 feet, the adult male has no light facial markings and the head is only slightly darker than the back. The adult female has much restricted, light facial markings and the intervening areas are darker than in the male. The five juveniles trapped in the same burrow as the female, each has more extensive light facial markings than the adult female although the area of this varies from only slightly more than in the female to as much as in typical specimens of xanthogenys. Also, the dark color of the head in these five specimens averages darker than in nevadensis and more as in weasels to the southwestward especially latirostra. One of the five juveniles is lighter colored over all of the upper parts than nevadensis and is suggestive of xanthogenys in this respect. Finally, the adult male has on the underparts small spots of ochraceous orange suggestive of latirostra and some individuals of pulchra. No. 30655/42628, U. S. Nat. Mus., taken on Mount Whitney, also shows white facial markings and some other features of the valley-inhabiting xanthogenys. A suggestion of intergradation with arizonensis is furnished by specimens, referred to that race, from Springerville and the Kaibab Plateau. No specimens happen to be available from the region in which intergradation would be expected between nevadensis and neomexicana. Since neomexicana and arizonensis intergrade it is probable that nevadensis also will be found to intergrade with neomexicana. In summary, nevadensis is judged to intergrade with each of the subspecies of Mustela frenata whose range adjoins that of nevadensis.

This subspecies is remarkably free from injury to the frontal sinuses such as result from the presence of parasites. In 98 adults from Oregon, California, Nevada, and Colorado, no malformation was noted. Only 1 of the 26 specimens from Washington was malformed and it was an intergrade with washingtoni. The single adult from New Mexico was diseased, as were 3 of the 6 from British Columbia, 1 of the 20 from Idaho, and 1 of the 7 from Utah.

Specimens examined.—Total number, 568, arranged alphabetically by provinces and states and from north to south by counties in each state. Unless otherwise indicated specimens are in the collection of the United States National Museum.

Arizona. Apache County: 15 mi. E Luka Chu Kai Navajo School, 8000 ft., 2.

British Columbia. Monte Cr., 20 mi. E Kamloops, 1; Sicamous, 2; Okanagan, 18 (7, 6, 1, 1); Monashee Pass, 1; Swan Lake, near Okanagan Landing, 1; Okanagan Landing, 11 (2, 3, 3, 3); Vernon, 1; Hope-Princeton Summit, 5600 ft., 1; Hope, 1; Similkameen, 1; Osoyoos-Bridesville Summit, 1; Anarchist Mt., Osoyoos, 1; Myer's Creek, 1; Rossland, Mt. Glory, 7000 ft., 1; Cascade, 1; Nelson, 1.

California. Siskiyou County: Hornbrook, 1; Tule Lake Refuge, 5; Upper Mud Creek, 6700 ft., Mt. Shasta, 3; Mt. Shasta, 1. Modoc County: Goose Lake, 1; Joseph Creek, 1; 5280 ft., Parker Creek, near Alturas, 1; Warner Mts., near Alturas, 1; 5 mi. NW Eagle Peak, 7000 ft., 2; Shields Creek, 5000 ft., 1; Jess Valley, 1. Shasta County: Cassel, 1. Lassen County: 3 mi. W Eagle Lake, 5800 ft., 1; 4 mi. S Eagle Lake, 6000 ft., 2; Mill Creek, 5000 ft., S base Mt. Lassen, 1; 6 mi. SW Calneva, 1. Tehama County: Dale's, 600 ft., on Paines Creek, 1. Plumas County: Kelly's, 2 mi. S Willow Lake, 5200 ft., 3; Quincy, 4; Beckwith, Sierra Valley, 1. Butte County: Jonesville, 1. Sierra County: Little Truckee River, 6500 ft., 3 mi. N Independence Lake, 2. Nevada County: Independence Lake, 1. Placer County: Donner, 3; 2 mi. W Soda Springs Station, 6500 ft., 1; Blue Canyon, 5000 ft., 2 (1); 4 mi. S Tahoe City, 1. Eldorado County: 5 mi. S Tallac, 6300 ft., 1; Gilmore Lake, Mt. Tallac, 2; Mt. Tallac, 1; Phillips, 1. Alpine County: 8000 ft., Hope Valley, 1; 8000 ft., Silver Creek, 1. Tuolumne County: Strawberry, 5200 ft., 1; 9300 ft., Ten Lakes, Yosemite Park, 1; Tuolumne Meadows, 8600 ft., Yosemite Park, 1; Tuolumne Meadows (Soda Springs), 1; Tuolumne Meadows, 8500 ft., Yosemite Park, 1; Sequoia, 1. Mariposa County: Chinquapin, 6256 ft., 2; Merced Grove Big Trees, 5400 ft., 1; Wawona, 1; no locality more definite than county, 1. Madera County: Bass Lake, 1. Mono County: Tioga Crest, near Tioga Pass, 4; Warren Creek, 1; Tioga Lake, 1; Ellery Lake, 9600 ft., 1; Mono Lake P. O., Mono Lake, 1; Walker Lake, 8000 ft., 2; Pine City, 1; Mammoth, 13 (12, 1); 10300 ft., near Big Prospector Meadow, White Mts., 2. Inyo County: Little Onion Valley, 7500 ft., 1; N Fork Bishop Cr., 10500 ft., 1; S fork Bishop Cr., Andrews Camp, 8000 ft., 1; South Lake, S Fk. Bishop Cr., 9750 ft., 1; Lamarck Cr., 9900 ft., 15 mi. SW Bishop, 1. Fresno County: Hume, 1. Tulare County: Mt. Whitney, 2; Whitney Meadow, 9800 ft., 1; Monache Meadow, 8000 ft., 3; E fork Kaweah River, 9800 ft., 1; 1/2 mi. S Mineral King, 7850 ft., 7; Quaking Aspen Meadow, 7500 ft., 1.

Colorado. Moffat County: Lay, 1. Routt County: Steamboat Springs, 2 (1); no locality more definite than county, 1. Jackson County: Higho, North Park, 8400 ft., 1; Buffalo or Illinois Creek, "near Rand," 6. Washington County: 6 mi. NE Hillrose, 1. Larimer County: Estes Park, 2 (1, 1); Pinewood, 1; Loveland, 2 (1); no locality more definite than county, 1. Rio Blanco County: Compass Creek, 9000 ft., 1; White River, 6200 ft., 1; Piceance Creek, 6200 ft., 1; Dry Fork, 6200-6600 ft., 4; Meeker, 1; Marvine, 1. Grand County: Crembling [= Kremmling?], 1; Middle Park, 1. Boulder County: Foot Mt. Meeker, 8700 ft., 1; Silver Lake Mine, 1; Boulder, 1; Dixie Lake, 2 (1, 1); Caribou, 1; no locality more definite than county, 1. Clear Creek County?: Grays Peak, 1. Jefferson County: 7000 ft., Mt. Parks, 1; 6 mi. W Denver, 1. Adams County: Barr, 1; near East Lake, 2. Denver County: Denver, 2 (1, 1). Arapahoe County: Littleton, 1. Summit County: Breckenridge, 1. Eagle County: Eagle, 9500 ft., 1. Park County: Jefferson, 4 (1); 12800 ft., Mt. Bross, 1. Mesa County: Tunnel, 1. Montrose County: near Crawford, Clear Fork of Smiths Fork, 1; Coventry, 3 (1); Naturita, 1; Paradox, 1; West Paradox Valley, 1. Pitkin County: Placita, 2. Gunnison County: Marble, 1, Crested Butte, 2; Deckers Ranch, Crested Butte, 2; Sapinero, 7245 ft., 1. Chaffee County: Buena Vista, 1; Hancock, 1; Salida, 5. Teller County: Glencore, Pikes Peak, 1. El Paso County: Monument, 1; Seven Lakes, 1; Lake Moraine, 10250 ft., 1; Colorado Springs, 6000 ft., 1; 5 mi. E Sand Creek, Colorado Springs, 1; no locality more definite than county, 1. Saguache County: Villa Grove, 1; Pierce Place, Cochetopa Nat. Forest, 1; Houselog Creek, Cochetopa Nat. Forest, 1; P. Tevebaugh's Ranch, near Cochetopa Pass, 1; P. Tevebaugh's Ranch, 9 mi. S Cochetopa Pass, 1. Rio Grande County: between Monte Vista and Del Norte, 1. Archuleta County: Upper Navajo River, 2; Navajo River, 5 (4, 1); Chromo, 2. Conejos County: Osier, 3. Montezuma County: Ure Peak, 1. County in question: Del Norte Peak, 1; no locality more definite than state, 4.

Idaho. Latah County: Cedar Mt., 4000 ft., 12 mi. NE Moscow, 1; Moscow and 1/2 mi. W, 2. Idaho County: Lochsa River (= Locksaw Fork), 1; between Selway Riv., and S Fork Clearwater Riv., 8; Selway Divide, 8; Pilot Creek, 2; Newsome Cr., 1. Lemhi County: Salmon River Mts., (now Lemhi Mts.), 8000 ft., 5; Leadore, 3. Adams County: summit Smith Mt., 7500 ft., 1. Washington County: Midvale, 2. Custer County: Pahsimeroi Mts., 1; Double Springs, 16 mi. NE Dickey, 1; Mackay?, 1; Stanley Lake, 1. Payette County: 2 mi. S Payette, 1. Fremont County: 17 mi. E, 4 mi. N Ashton, 6275 ft., 2. Teton County: 3 mi. S Victor, 1. Jefferson County: 20 mi. W Camas, 1. Blaine County: Sawtooth City, 1; Ketchum, 5 (3, 2). Canyon County: Nampa 3. Clark County: Dry Creek, Targhee Nat. Forest, 1; Birch Creek, 2. County in question: North fork of Teton River, 1. Bingham County: Shelley, 1; Alridge, 2; Springfield, 1. Lincoln County: Shoshone, 1. Minidoka County: 1/4 mi. E Heyburn Bridge, 1. Power County: 4 mi. NW American Falls, 1. Bannock County: 3 mi. N Schutt's Mine, Ross Creek, 1; 3 mi. N Pocatello, 1; near (within 10 miles of) Pocatello, 1; 3 mi. S Pocatello, 1; 1 mi. E Portneuf, 1; 2 mi. up Mink Creek, 2 (1, 1); Inkom, 2; Swan Lake, 1. Owyhee County: 5 mi. SE Riddle, 1; Three Creek, 2. Cassia County: Elba, 1. Bear Lake County: Geneva, 6171 ft., 1; Montpelier, 1; Paris, 6000 ft., 1; Pegram, 2.

Nevada. Humboldt County: Alder Creek, 7000 ft., Pine Forest Mts., 1; head of Big Creek, 8000 ft., Pine Forest Mts., 1; Cottonwood Range, 1; Calico Mt., Little Owyhee R., 1; Mahogany, Little Owyhee R., 2; Sulphur, 1. Pershing County: Lovelocks, 1. Elko County: Mountain City, 3; Three Lakes, Ruby Mts., 1. Washoe County: Pyramid Lake, 1; 3 mi. E Reno, 1; Incline Creek, 7100 ft., 1; 2-1/2 mi. S Incline, 6250 ft., 1; E side Marlette Lake, 8000 ft., 1; Marlette Lake, 8000 ft., 1. Ormsby County: 1/2 mi. S Marlette Lake, 8150 ft., 1. Churchill County: 4 mi. W Fallon, 1; 3 mi. W Fallon, 1; 2 mi. W Fallon, 1; Fallon, 3970 ft., 1. 5 mi. S Fallon, 4000 ft., 1; 8 mi. S and 3 mi. E Fallon, 1. Douglas County: Mt. Siegel, 1. Mineral County: Lapon Cañon, 8900 ft., Mt. Grant, 1. Nye County: Arc Dome, 1; 10700 ft., 1/2 mi. SW Jefferson Peak, Toquima Range, 1. White Pine County: 3 mi. E Baker, 1; Baker Creek, 6600 ft., 4; Baker Creek, 8400 to 8450 ft., 4; Gleason Creek, 7500 ft., 1. Esmeralda County: Arlemont, 4850 ft., Fish Lake Valley, 1. Lincoln County: 3 mi. S Crystal Spring, 3900 ft., Pahranagat Valley, 1.

New Mexico. Taos County: 2 mi. N Twining, 10500 ft., 1; Taos, 2. Santa Fe County: 11600 ft., Pecos Baldy, 1. San Miguel County: 8000 ft., above Willis, Pecos River, Forest Reserve, 2; Ribera, 1.

Oregon (by counties from west to east). Jackson County: Rustler Peak, Crater Nat. Forest, 1; Siskiyou (probably south of), 2. Klamath County: 20 mi. W Crescent, 1; Anna Creek, Mt. Mazama, 2; S Boundary Crater Lake Nat. Park, 1; Fort Klamath, 15; Upper Klamath Lake, 2; Klamath Falls, 1. Lake County: Dog Lake Ranger Station, 30 mi. SW Lakeview, 1. Harney County: Camp Harney, 2; Burns, 2 (1); 20 mi. S Burns, 1; Narrows, 1; Voltage, 1; Shirk P. O., 2; Keiger Gorge, Steen Mts., 4. Malheur County: Riverside, 1; 2 mi. NW Riverside, 2; Barren Valley, Cord, 1; Cedar Mts., 2; Cow Creek Lake, 1; Jordan Valley, 1. County in question: Sageview, 1.

Utah. Cache County: Logan, 1. Rich County: 8000 ft., near Laketown, 1. Boxelder County: Willard, 1. Salt Lake County: Salt Lake City, 1; Barclay, 6500 ft., Wasatch Mts., 1; Mill Creek, 1. Utah County: Provo Bench, 2; Aspen Grove, Mt. Timpanogos, 1; Payson, 1. Juab County: between Santaquin and Starr, 1. Uinta County: Dry Fork Canyon, 20 mi. NW Vernal, 1. Carbon County: Sunnyside, 1; Range Creek, 1. Millard County: Deseret, 1. Sevier? County: Fish Lake Plateau, 1. Grand County: Warner Ranger Station, La Sal Mts., 1. Beaver County: Britts Meadows, 11000 ft., Beaver Range, 1; Britts Meadows, Beaver Range, 1; Puffer Lake, 1. Garfield County: Boulder, 2. Washington County: Pine Valley, 1; St. George, 1. San Juan County: Geyser Pass, La Sal Mts., 2. County in question: Salt Lake, 2; Wasatch Mts., 1; La Sal Mts., 11000 ft., 1.

Washington. Okanogan County: Bald Mt., 6800 ft., 1; Bauerman Ridge, 6800 ft., Tungsten Mine, 1; Hart Pass, Methow River Trail, 1; Conconully, 2 (1, 1); 5 mi. NW Loomis, 1; Molson, 3800 ft., 1; Tunk Mt., 3500 ft., 1. Whatcom County: Barron, 5000 ft., 1. Stevens County: Colville, 1; Orin, 11. Pend Oreille County: Ione, 6. Chelan County: Chelan Mts., 1; Lake Chelan, 1; Manson, 1; Entiat River, 1680 ft., 20 mi. from mouth, 7; Dryden, 2; Wenatchee, 1. Kittitas County: Easton, 2 (1); Ellensburg, 1; 4 mi. E Ellensburg, 1. Grant County: Neppel, 1. Lincoln County: Sprague, 1. Spokane County: Spokane, 1; Cheney 2. Whitman County: Pullman, 11 (6, 1, 1); 6 mi. S Pullman, 1. Garfield County: Snake River, 1. Yakima County: Yakima, 1; 1 mi. W Moxee, 1.

Wyoming. NW Wyoming, 1. Yellowstone National Park: Lamar River, 1; Yellowstone Lake, 1. Park County: Greybull River, 1. Teton County: Crystal Creek, 2; Jackson, 1; Whetstone Creek, 2. Johnson County: Buffalo, 4 (2). Fremont County: Continental Divide, 20 mi. NW Dubois, 1. Sublette County: Bronx, 1. Carbon County: Medicine Bow Mts., 1; 15 mi. SE Parco, 1. Albany County: Garrett, 1; 12 mi. W Laramie, 1; 7 mi. W Laramie, 2; 5 mi. W Laramie, 4; "near" Laramie, 1; 3 mi. SW Laramie, 1; 12 mi. S Laramie, 1. Uinta County: Fort Bridger, 6800 ft., 1; Lonetree, 1; Bridger Pass, 1. County in question: Laramie River, 2. No locality more definite than state, 1.

Mustela frenata effera Hall

Long-tailed Weasel

Plates , and

Mustela frenata effera Hall, Carnegie Instit. Washington Publ. 473:93, November 20, 1936.

Mustela arizonensis, Dice, Journ. Mamm., 1:12, November 28, 1919.