Page images
PDF
EPUB

a mean of all known observations, it seems proper to choose those which give the slowest rates of thermal increment, for these are the ones which have been least interfered with by local volcanic action, and which most nearly correspond to the deeper and more stable gradients. The substitution of these slowest rates increases the time scale.

Clarence King has given a different method, depending on the fusion temperature of basalt as determined by Barus, and assumptions as to the average composition of the earth's outer layers, which are supposed to be basaltic. But the eruptive rocks are just those which have been, as it were, sweated out from the crust, leaving the more refractory materials behind. Grant this, and the average melting point of the residual rocks must be raised, increasing the time-estimate by this method. Taking the initial temperature of solidification as 1741° C., Clarence King finds for a thermal gradient of 1° F. in 50 feet, 20×10 years, and for 1° F. in 75 feet, 46×10 years, on the hypothesis of a solid earth. Neither estimate takes any account of the greater conductivity of the deeper layers. If this were done, it might be necessary to extend the computed durations several times.t

Finally, if the gradual modification of living forms by natural selection is a fact, the longest durations named are none too long for the numerous successive faunas and floras made known to us by paleontology.

We need an earth-duration of several hundred million years to meet the requirements of evolutionary theory and of thermotics. The sun, by its contraction, will give only 20,000,000. Hence the dilemma.

The hypothesis of the development of solar energy by atomic dissolution greatly extends the duration of the sun, in all probability doing away with this dilemma.

The companion hypothesis of an atomic genesis by luminous impulses, in addition, fulfills the philosophic necessity for a cosmic cycle which shall restore the ancient order, and answers the question: What becomes of the light from these myriads. of stars?

Washington, D. C.

*This Journal (3), vol. xlv, p. 1, 1893.

+ See Professor John Perry, "On the Age of the Earth," Nature, vol. li, p. 582,

1895.

ART. XVII.

Studies of Eocene Mammalia in the Marsh Collection, Peabody Museum; by J. L. WORTMAN. (With

Plate VI.)

[Continued from p. 128.]

Limnocyon Marsh.

Limnocyon Marsh, this Journal, August, 1872, p. 6, Separata;
Thinocyon Marsh, this Journal, August, 1872, p. 12, Separata ;
Oxy@nodon Wortman, Bull. Amer. Mus. Nat. Hist., 1899, p. 145;
Telmatocyon Marsh, this Journal, May, 1899, p. 397.

A group of small or medium-sized Creodonts ranging in time from the beginning of the Bridger to the close of the Uinta epoch, and, as far as known, having the following principal characters: Dental formula I. C. Pm. M.; first upper and lower premolars two-rooted (except in Limnocyon dysodus); last superior molar transverse and little reduced; two subequal inferior molars with internal cusps and moderate-sized basin-shaped heels; fibula articulating with calcaneum; astragalus moderately grooved; femur with small third trochanter; deltoid crest of humerus reduced; distal end of humerus broad, with prominent supinator ridge and an entepicondylar foramen; metapodials of fore feet short and phalanges elongated; carpus unknown.

The genus Limnocyon was described by Professor Marsh in Angust, 1872, from a series of superior teeth which were not in place in the maxillary. A second species was proposed in the same paper upon a specimen consisting of both mandibular rami, one of which contains the last premolar and first molar, together with all the alveoli. In the same paper Professor Marsh proposed a second genus, Thinocyon, upon an entire left mandibular ramus, containing a few of the teeth in good condition and the alveoli and roots of all the others. In June, 1899, I proposed the genus Oryanodon upon a well-preserved half of a skull, in which both the upper and lower teeth are present. In May, 1899, Professor Marsh, upon my advice, placed Limnocyon verus as synonymous with Sinopa, and proposed for the second species L. riparius, the generic name Telmatocyon. The reason for this advice was as follows: The type of the genus consists of the dissociated upper teeth, in which the superior molars are almost, if not quite, indistinguishable, in structure at least, from those of certain species of Sinopa, and as the number in this latter group is three and no two-molared type was at that time known, it was quite naturally supposed that the type specimen of L. verus was a Sinopa. AM. JOUR. SCI.-FOURTH SERIES, VOL. XIII, No. 75.-MARCH, 1902.

The type of L. riparius, having but two subequal lower molars, was otherwise unknown, and was regarded as a distinct genus. The relationship with Oxyanodon was entirely overlooked. The unstudied part of the collection affords much additional material, and it is now quite evident that the number of superior molars in L. verus is two instead of three. The association of upper and lower teeth in this material renders it clear, moreover, that the type of L. riparius is the lower jaw of L. verus, and on this account I do not hesitate to unite them. I also arrange Thinocyon and Oxyanodon in this genus, but, as already remarked, it may be found, with more complete information, that they represent distinctive generic modifications.

Limnocyon verus Marsh.

Limnocyon verus Marsh, this Journal, 1872, p. 6, Separata.

Originally established upon a superior series of teeth of the right side, figure 71, with the first premolar only, in place.

[blocks in formation]

FIGURE 71.-Series of right superior teeth of Limnocyon verus Marsh ; crown view; three halves natural size. (Type.)

The two middle incisors are present, but they do not present any characters of unusual importance; their roots are much compressed from side to side, like certain members of the Mustelidæ, and their crowns are obtusely pointed and rather narrow. The canine is represented by fragments only, but these are sufficient to indicate that it had the usual size and form of this tooth among the Carnivora; its surface is traversed by characteristic longitudinal grooves and ridges, much as in certain species of reptiles. There are apparently no traces of these grooves and ridges upon the canines of the other species of the genus (except very faint ones in L. velor), and it may be taken to be diagnostic.

The first premolar is two-rooted, with a conical, more or less recurved, principal cusp, and a rather extended heel bearing a small cusp; it is placed immediately behind the canine without the intervention of a diastema. The second premolar is only partially preserved in the type, but is present in many other specimens in the collection. The third premolar is missing in

the type, but in other specimens is similar to the first but larger. There is no internal cusp as in Oxyæna. The crown of the fourth is made up of the usual elements found in the typical sectorial premolar, although certain parts are little developed. There are two main external cusps, together with a small but distinct anterior basal cusp; to these is added a relatively large internal lobe, which is placed opposite or a little anterior to the large principal external cusp. The posterior external cusp is proportionally small, and the posterior

[merged small][graphic][subsumed][subsumed][subsumed][subsumed][subsumed][subsumed]

FIGURE 72.-Left mandibular ramus of Limnocyon verus Marsh (type of Limnocyon riparius Marsh); side view; three halves natural size.

border of the main cusp is little produced, so that the shear of this tooth cannot be said to be very perfectly developed. The first molar is the largest tooth of the series; its crown is composed of two principal external cusps, well separated, a strong blade-like postero-external spur, between which and the postero-external cusp is a deep vertical fissure, a basal antero-external ledge, and a large lunate internal cusp. The large postero-external spur, together with the postero-external cusp, furnish the principal shear, and constitute the chief sectorial organ of the superior series of teeth. The last molar is relatively large, three-rooted, and transverse; its crown is composed of one main external cusp, external to which, and separated by a slit-like fissure, is a sharp ridge, which has more or less of a sectorial function. The usual internal lunate cusp, together with distinct anterior and posterior intermediates, are present, but the postero-external cusp is ves- and fourth premolar of the tigial or wanting. preceding figure; inside

73

pᏎ

FIGURE 73.-First molar

The type of Limnocyon riparius, fig- view; three halves natural ures 72, 73, which I take to be the same

size.

as L. verus, consists of the greater part of both mandibular rami and a portion of a first superior molar. The specimen

does not exhibit very clearly the number of incisors, but in others which agree with it very closely, the number is three. The first premolar is two-rooted and is placed close behind the canine, without diastema. All the other premolars are likewise implanted by two roots, and are much crowded in the jaw. The fourth has a distinct posterior, but no anterior basal cusp. The crowns of all the premolars are rather thick from side to side, and the cusps are inclined to be obtuse and rounded. The two subequal molars, of which the first is preserved in the specimen, have the following characters: A moderately elevated trigon, with a large external and smaller internal and anterior cusps; a moderately well-developed shear, and bladelike modification of the external and anterior cusps; and a medium-sized basin-shaped heel. These two teeth are nearly equal in size, the last being a little the larger. The jaw is shallow vertically and thick from side to side. The symphysis is enlarged and extends to beneath the middle of the third premolar. There are two mental foramina, the larger of which issues beneath the anterior border of the second premolar. The coronoid is rather large and the angle is not inflected.

74

75

In one specimen a considerable part of the skull, Plate VI, together with certain parts of the skeleton, are preserved, from which the following characters may be stated: The face is rather short; the orbital cavity is relatively small; there is a distinct postorbital process and a sagittal crest. of moderate proportions; there is apparently no anterior glenoid process, but a distinct postglenoid foramen; the mastoid is well exposed upon the postero-lateral wall of the skull, but the position of the stylomastoid foramen cannot be ascertained; the trochlear surface of the astragalus, figure 74, is moderately grooved and there FIGURE 74.-Astragalus and is a distinct astragalar foramen, of calcaneum of Limnocyon verus about the same size and position as Marsh; front view; three seen in that of the otter; the fibula is

[graphic]

halves natural size.

FIGURE 75.-Fibula of Lim- little reduced and has a considerable nocyon verus Marsh; end contact with the calcaneum, figure 75, view; showing articular sur- as in the otter; the calcaneum has a face for calcaneum; three moderately short tuber, a broad sustentaculum, and a prominent calcaneal tubercle; there were apparently five toes in the hind foot, and the phalanges were elongated, being but little shorter than those of Lutra.

halves natural size.

« PreviousContinue »