(4.) It is assumed that the old hair is not shed, but suddenly assumes new vitality and proceeds to "increase in length," and to take on an entirely changed molecular structure, for he says elsewhere (1. c., p. 233) that "to this [increase in the length of the autumnal hairs] must be added that the blanching shaft, in the majority of cases, has also augmented in thickness....the increase being consequent upon a more than usual number of series of cells entering into its composition." And again (1. c., p. 234): "It would seem that the rapid development of new hairs involves the autumnal outer fur in the same process, leading to an increased length and thickness in the shaft of the hair by the superposition of layers of the same colourless cells entering into the structure of the new growth-perhaps combined also with an arrested production of pigmentary matter." This is further elaborated in such detail of explanation, based on microscopical examinations, that it would seem to rest on a solid basis, but I must confess that to me the case is simply incomprehensible. My material certainly demonstrates that the summer coat is shed, and not transformed, either in structure or color. On the other hand, it does not need even the aid of a magnifying glass to show that the winter coat of long white overhair is of a finer and softer texture, with a much smaller (instead of larger) diameter of shaft than the summer coat, which is coarser, harsher to the touch, and of a different structure, the particolored (not the wholly black to the same degree) hairs of the summer coat being subapically thickened, the thickened portion including the fawn-colored band, and extending slightly above and below it. Hence, the only explanation that occurs to me is that the thickened hairs were a part of the old summer coat, which it was assumed were in process of change, or about to change to white, simply from the fact that the brown hairs were disappearing, and that he compared these old hairs with the new hairs of the winter coat. Again, in specimens well advanced in change, on casually parting the pelage of the back, the new white hairs are so abundant and conspicuous that all of the hairs seem white below the surface, but on removing some of the hairs still remaining of the summer coat they are found to be colored to the base and still unchanged.

In a January specimen (No. 3277, Rutland, Vt., Jan. 8) a small proportion of the hairs over the dorsal region are pure white for

rather more than their apical half, and then pass into a broad band of very pale horn-color, which usually fades out basally, or may persist faintly to the base of the hairs. In texture and size they do not differ from the pure white hairs with which they are intermixed. A very few of these basally faintly horn-tinted hairs can be found in another specimen taken at the same time and place, and in one out of three early March (March 2) specimens from the same locality. Also a specimen from Locust Grove, Lewis Co., N. Y., taken March 21, 1884, and kindly loaned me by Dr. Merriam, is a fine example of this phase of coloration. In this specimen many hairs still remain wholly black, others are only black, or more or less blackish, basally. These hairs are evidently, however, in each case, a part of the true winter coat, and not a remnant of the summer coat, any more than are the fulvous hairs on the fore legs and hind feet of certain midwinter specimens, or the basally fulvous hairs on the nose, or the fulvous hairs on the ears, of these same specimens. They simply grew particolored instead of white, and have not either the texture or the form of the particolored hairs of the summer coat. Yet, although so exceptional, they might, considered alone, seem to give some support to the 'blanching' theory of the autumnal change of color.

So much attention would not be given, in this connection, to Mr. Welch's paper, in view of the overwhelming evidence of its. erroneous character, were it not that it has recently been made prominent by Mr. E. B. Poulton as the chief basis for his theory respecting Variable Protective Resemblance in Vertetrates,'' in which he either quotes at length or summarizes from Mr. Welch most of the passages above quoted, and proceeds to theorize from this insecure basis.

In conclusion it may be well to correct a time-honored error respecting the geographical distribution of Lepus americanus, since it figures prominently in the matter of seasonal change of color. Thus Mr. Poulton (1. c., p. 97), in speaking of this species, says: "In Hudson's Bay Territory it changes early and carries. the winter coat till June, while no change of colour takes place

Cf. The Colours of Animals,' Chap. VII. Intern. Sci. Ser., Vol. LXVII. New York, Appleton & Co., 1890.

in the winter in the southern parts of the United States," basing the statement probably on Welch (1. c., p. 235), who in turn quotes it directly from Sir John Richardson. Richardson seems to have derived the statement from Pennant, as he says (Faun. Bor.-Am., I, p. 218): "The white color is less perfect in more southern districts, and to the southward of New England, according to Pennant, the brown dress endures all the year." Pennant, in his Arctic Zoology,' recognized only two species of North American Hares-the Varying Hare and the American Hare. The first is the Arctic Hare of the present day; the other was primarily the Varying Hare here under consideration, but included also all other species of North American Hares then known. Hence when he says, "From New England southward they retain their brown color the whole year," he is evidently speaking of the 'Cotton-tails' and Swamp Hares of the South, which, as every one knows, never turn white. It is needless to tell intelligent mammalogists that the southern limit of distribution of the Varying Hare is the southern half of the Alleghanian Fauna-in other words, it is not found at ordinary levels south of Massachusetts. Furthermore, there is little evidence to show that it does not practically become white in winter to the very southern limit of its range, although less perfectly so than further north, the change in color sometimes remaining more or less incomplete and superficial, so far as can be determined at this writing.

1 Not italicized in the original.

241.6

2001894 8086

Article V. OSTEOLOGY OF PATRIOFELIS, A MIDDLE

EOCENE CREODONT.

By J. L. WORTMAN, M.D.

PLATE I.

HISTORY AND SYNONOMY.

The genus Patriofelis was originally established by Dr. Leidy, in the 'Proceedings' of the Philadelphia Academy, March, 1870, p. 10, upon the fragmentary portions of the rami of both lower jaws, which were obtained by Dr. Hayden in the Bridger Baisin, Wyoming, the year previous. In August, 1872, Prof. Marsh described (Amer. Jour. Science, Vol. IV, p. 10) from the same locality, some remains of a "gigantic Carnivore which he referred to a new genus and species under the name of Limnofelis ferox. According to Prof. Marsh's statement, his specimen consists of portions of the skull, fragments of the lower jaw, some vertebræ, and other less important parts of the skeleton. In the same paper he describes a second species under the name of Limnofelis latidens, from a last upper premolar, which was obtained in the same horizon. Prof. Cope, in the 'American Naturalist' of 1880 (Vol. XIV, p. 745), proposed a third genus from teeth and limb bones, which were collected by the writer in the Wind River Baisin in the summer of 1879. To these remains Prof. Cope gave the name Protopsalis tigrinus. Prof. Scott has described in the 'Journal of the Philadelphia Academy,' 1886 (Vol. IX, p. 174), some remains of a large Creodont from the Bridger formation, which he referred to Prof. Cope's genus Protopsalis. A new species of this genus was proposed by the writer (Bull. Amer. Mus. Nat. Hist., Vol. IV, p. 98, 1892), under the name of P. leidyanus, from a specimen in the Princeton Collection.

The material collected by the American Museum Expedition into the Bridger Baisin now enables me not only to give an unusually full description of the osteology of the species of [May, 1894.]

[129]

9

Patriofelis, but to establish the synonomy of the three generic names that have been proposed.

A careful comparison of Marsh's description of Limnofelis ferox with Prof. Cope's figures and description of Protopsalis tigrinus leaves little doubt of the generic identity of the two specimens. Both of them, moreover, agree so perfectly with our specimens, that I do not hesitate to refer them to the same genus. Marsh's description is characteristically brief and imperfect, but enough is stated to indicate that the last lower molar of Limnofelis is the same as that of Protopsalis, which is in turn like that of Patriofelis. Marsh's specific name ferox is therefore adopted, since it has priority. A comparison of our specimens with Leidy's type of Patriofelis ulta reveals a difference only in size, our largest specimen being at least one-third larger in every way. I am therefore convinced that Limnofelis and Protopsalis do not present any characters, so far as known, which will enable one to separate them generically from the genus originally proposed by Leidy, namely, Patriofelis.

The specimens upon which this paper is based were found in the Bridger Baisin at widely separated localities. The most complete is from the Henry's Fork region; several other specimens of both the species were found at Twin Buttes, and their remains are likely to occur wherever the exposures contain fossils. The larger species, P. ferox, is one of the largest Creodonts known, and equaled in size a full-grown black bear. The head was disproportionately large and massive, almost equaling in this respect an adult lion. The smaller species, P. ulta, was about one-third smaller. In both, there was a long and powerful tail, broad plantigrade feet, which, together with other characters presently to be considered, lead to the conclusion that they were aquatic in habit.

The subject is considered under the following heads: History and Synonomy; Osteology; Comparison with other Creodonts; Comparison with the Seals; Probable Habits; Classification and Species.