P. M. Ryzhkov,
Instructor at the Siberian Technological Institute.

Elephas primigenius Blum. and Bison priscus H. v. Meyer from the Area of the Kemerovo Mine on the Tom River

Introduction

The present article is the result of work on a small collection of bones found by me in the summer of 1923 at the Kemerovo coal mine on the Tom River. The conditions under which the bones were discovered were as follows. Workers of the American Industrial Colony “Kuzbass,” while driving the Vladimirov inclined shaft (see fig. 18), at the sixth meter from the surface, in a layer of sandy-clayey deposits, came upon by chance an accumulation of bones of various kinds. The mine administration reported this find to the Tomsk Technological Institute, whose administration sent me to Kemerovo to carry out excavations and recover the bones. As a result of the excavations I extracted the following bones: two lower molars m₃, belonging to one individual of Elephas primigenius Blum. (mammoth), and 34 different bones, likewise belonging to a single animal—Bison priscus H. v. M. (fossil bison). The mine administration donated the recovered bones to the Tomsk Technological Institute, where, after their description, they are kept in the museum of the paleontological cabinet.

It must be said that finds of remains of Elephas primigenius and Bison priscus in the area of Tomsk Province¹), in particular along the Tom River²), had been noted in the literature previously as well. Thus, Prof. M. E. Yanishevsky indicates that separate bones of these animals were found along the Tom River in the villages of Saltymakovo, Zmeinka, Krapivina, and Bedareva. At the same time he explains that most of the bones were acquired from various persons who had found them near their own places of residence. The find of bones at the Kemerovo mine represents a more fortunate case, since all the bones were found in primary deposits, are distinguished by comparatively considerable quantitative completeness, and belong to a single individual. Hence the geological and paleontological value of the remains found is beyond doubt, and therefore I consider it necessary to dwell in greater detail on the characterization of the bones found, both of Elephas primigenius Bl. and especially of Bison priscus H. v. Meyer.

All the bones described below were discovered in a layer of yellow clayey sand, with faintly noticeable diagonal bedding, at a depth of 6 meters (see table V, fig. 19). The thickness of this layer does not exceed one meter. The bone-bearing horizon is overlain by a layer of loess up to five meters thick, and is underlain by yellow plastic clay up to 6.5 meters thick. Beneath the clay layer lies well-sorted river sand, passing deeper into pure gravel. All these more or less loose deposits rest upon the heads of dense sandstones, clayey and carbonaceous shales with seams of hard coal.

¹) Yanishevsky, M. E. Brief report on excursions in 1907 and 1908. Tomsk. 1909; 1–8.
²) Yanishevsky, M. E. Preliminary report on a student excursion along the Tom River in 1909. Izv. Tomsk. T. I. Vol. XXIV. No. 4. 1911; 18.

DESCRIPTION OF THE FOSSIL REMAINS.

Elephas primigenius Blum.

(Table 1, fig. 1).

Two lower molars \(m_3\), belonging to Elephas primigenius, were found on the same horizon and in the same deposits as the bones of Bison priscus H. v. M. The teeth are distinguished by a dirty-gray color. Their size, form, and the character of wear indicate that both teeth belong to a single adult individual. The general preservation of the teeth is quite satisfactory, although it must be noted that the roots, with the exception of the anterior one, which is roughly hook-shaped, have been lost. At the posterior end there is visible, in the form of an indentation, the mark of retarded development of the tooth, which is due to the pressure of the following molar \(m_6\), which was pressing it out. Each tooth consists of eighteen plates, of which fourteen fall within the chewing surface. Each plate-ridge on the chewing surface occupies 0.8 cm, of which 0.4 cm is dentine and 0.2 cm each layer of enamel. The greatest width of the chewing surface is 8.6 cm, with a length of 17.5 cm. With the total length of the tooth, in the plane perpendicular to the plates, being 20 cm = 7.8 dm, each loima contains about \(2 \tfrac{1}{2}\) plates—this, as Samuel Calvin¹ indicates, is a quite sufficient criterion for assigning these teeth to the species Elephas primigenius Blum.

Bison priscus H. v. Meyer²).

Fossil bison.

(Table I, figs. 2 and 3; Tables II, III and IV; figs. 4–17)

In the post-Pliocene epoch, bison of various species inhabited Europe, Asia, and North America in great numbers, as is evidenced by their numerous remains found in many places in Siberia, Europe, and North America. These remains are often encountered together with the mammoth, rhinoceros, and ancient bison (Bos primigenius Boj.). At the present time, thanks only to special protection, the bison survived³ before the war in a small tract of old virgin forest known as the “Białowieża Forest” (Grodno Province), and in the mountains of the northwestern Caucasus⁴); and the closely related species—the American bison (Bison americanus)—is kept in small numbers in the Central Yellowstone Park of North America under the protection of the government of the States. In addition, a small colony of it, artificially bred and carefully maintained, is found in the Miedzyrzecz grove in Silesia, and individual specimens of bison are kept in some zoological gardens of Europe⁵). But despite the abundance of these animals in the post-Pliocene epoch, all finds—at least in Siberia—are reduced to separate scattered bones of the skeleton, collected in most cases by entirely accidental persons, often not even indicating the nature of the strata in which particular remains were discovered. Thus, at least, matters stand with the fossil remains of animals—

¹ Samuel Calvin. Aftonion mammalian fauna. Bulletin of the Geological Society of America. Vol. 20. 1909; 351–352.
² H. v. Meyer. Ueber fossile Reste v. Ochsen. 1835.
³ Prof. Ruzskii M. D. “The bison as an expiring representative of our fauna.” Scientific Notes of the Kazan Veterinary Institute, vol. XV, 1898; 11, 45.
⁴ The Caucasian bison has already separated from the Białowieża bison and was described by K. Satunin as a distinct species (Bison caucasicus).
⁵ Among other places, in our country in the government reserve “Askania-Nova” (in the south of European Russia).

...kept in many museums of Siberia, as is attested by brief journal notes which, in most cases, are limited to a simple enumeration of the bones of fossil vertebrates. It is true that one must note the substantial works of Chersky¹˒² and Pavlova³˒⁴, in which, among other things, the remains of Bison priscus H. v. M., collected in the far north of Siberia, as well as in the environs of the city of Irkutsk and in Transbaikalia, are described in considerable detail. In addition, mention must be made of the single complete skeleton of a fossil bison, found near the city of Tyumen, which is preserved in assembled form at the museum of the former Tyumen Realschule.⁵ This skeleton was described in considerable detail by M. V. Pavlova⁶ in 1906. With regard to the remains of fossil bison found at individual localities of Tomsk Province, as I indicated above, Prof. Yagnishevsky reports on them.

The find in the area of the Kemerovo mine, in terms of the number of bones belonging to one and the same individual, is more or less exceptional. Here we have the following bones of Bison priscus H. v. Meyer: the core of the right horn, both rami of the lower jaw, 10 vertebrae, the sacrum, right humerus, right radius et ulna and left radius et ulna, right femur, right tibia, right and left metacarpus, all phalanges of the right forefoot, right and left astragalus, right and left calcaneum, right metatarsus, and phalanges of the right hind foot: IVʹ and IV₃. The majority of these bones are in quite satisfactory preservation, which makes it fully possible to describe them.

The right horn core and part of the frontal bone, bounded by a suture vertically dividing the frontal part of the skull in half, reveal (pl. I, fig. 2) a forehead that is typically convex for bison and almost horizontally directed to the side, and a horn. The completely smooth neck, thickening, passes into a coarse ribbing that extends to the end of the horn. At the same time, on the anterior side and at the end this ribbing is weaker than on the posterior, upper, and lower sides of the horn. The exact direction of the horn is as follows: from the base it rises noticeably upward, then from the middle—somewhat forward and again upward along a weakly spiral line. The section of the horn core at the base is almost round, while from the middle to the end it is sharply elliptical, with a large anteroposterior axis. Measurements of the described bones give the following values for the most characteristic elements.

¹ Izv. East-Siberian Department of the Russian Geographical Society. 1886. Vol. XVI.
² Chersky, I. D. Description of the collection of post-Tertiary mammals collected by the Novo-Sib. exped. in 1885–86. 1891; 109–213.
³ Zapiski Ak. Nauk. Vol. XXI, No. 1. 1906.
⁴ Trudy Troitskosavsko-Kyakhtinsk. otd. Priam. Russ.-Geogr. O-va. Vol. XII, issue I. 1910; 21–60.
⁵ Pavlova, M. V. Description of fossil remains of mammals of the Troitskosavsko-Kyakhta Museum. 1910; 23.
⁶ M. Pavlow. Etudes sur l’histoire paléontologique des ongulés. Sélénodontes posttertiaires de la Russie. 1906; 57–67.

The lower jaw (plate I, fig. 3) is represented by both rami, broken off at the ends. On the dental part of each fragment only the true molars and traces of deciduous teeth are preserved, in the form of alveoli for the roots of the teeth. The teeth, judging from those that remain, have a masticatory surface sloping toward the outer edge. On the outer side of each molar tooth there is an accessory basal column, and behind the last tooth, moreover, a strongly developed heel. The enamel layer, covering the tooth externally, also forms loop-like structures within it, elongated along the tooth row and having an irregular shape.

Vertebral column. Of the vertebral column, my collection contains the following vertebrae of Bison priscus H. v. M.: epistropheus, the IV, V and VI cervical vertebrae, 5 thoracic vertebrae, one lumbar vertebra, and the sacrum. Of these, the cervical vertebrae, one thoracic vertebra, and the sacrum may be considered well preserved; the others permit only partial measurements to be made and only probable conclusions to be drawn.

The difficulty of determining the position of one or another vertebra in the vertebral column is pointed out by many investigators of fossil vertebrates, in particular by Chersky¹), who described the remains of extinct representatives of the family Bovidae from the far north of Siberia. True, on this question he gives some

¹) Chersky, I. D. Description of the collection of post-Tertiary mammals collected by the New Siberian Expedition of 1885–86. 1891; 133–134, 176.

the indications of Rütimeyer¹ and, in greater detail, Chersky,² which make it possible, with a certain accuracy, to indicate the positions only of the cervical vertebrae. Having before me vertebrae from different parts of the vertebral column, known to belong to one individual, I can give here certain general statements characterizing the separate parts of the spinal column of representatives of the family Bovidae. The medullary canal, on leaving the skull, rapidly narrows within the first vertebra, assuming in section a rounded form. Further on, its form through all the cervical vertebrae and almost to the end of the row of thoracic vertebrae does not change; only before the lumbar vertebrae does its section slowly pass from a rounded form first into a roughly triangular one, and then into an elliptical one, with the major axis placed horizontally. In the sacrum the canal rapidly narrows and in the caudal region passes into a narrow rounded-triangular tube. Neural canals may be observed on the atlas, epistropheus, thoracic vertebrae, and sacrum. On the IV, V, and VI cervical vertebrae available to me, and on one lumbar vertebra, there are no neural canals; but on these same cervicals, parallel to the body of the vertebra, there pass vascular canals of fairly considerable size and oval form, which are absent on the thoracic, lumbar, and sacral vertebrae. The size and form of the spinous processes vary greatly throughout the whole length of the vertebral column. The spinous process of the epistropheus is fairly broad, but its height does not exceed the length of the body of the vertebra. Further, up to the seventh thoracic, the length of the processes increases smoothly; but, beginning with the first thoracic, by leaps up to the fourth it becomes disproportionately large, sometimes exceeding by 9–10 times the length of the body of the vertebra. Then, beginning with the fourth thoracic, the length of the processes decreases, and on the last sacral again does not exceed the length of the body of the vertebra. The body of each vertebra anteriorly and posteriorly has articular surfaces, the anterior ones being convex and the posterior ones concave. Beginning with the third cervical, the convexity, from an elongate shield-like form, gradually passes into a rounded one, which it attains fully in the last lumbar vertebrae. Further toward the tail, the form of the articular surfaces again becomes elongate-rounded, but now with the major axis directed horizontally.

After these general remarks I pass to a detailed characterization of each vertebra separately.

Epistropheus (pl. II, fig. 4). In its external form the second cervical vertebra of Bison priscus differs sharply from the other vertebrae in the skeleton. In front of the body of the vertebra, more than halfway around the canalis vertebralis, there projects the so-called odontoid process, which, when articulated with the atlas, enters the expanded part of the medullary canal. Below the odontoid process, a semicircular articular surface is situated. Along the lower part of the body of the vertebra runs a crest; on both sides of the crest there are comparatively large depressions, the outer margins of which posteriorly pass into poorly developed lateral processes. On the posterior part of the body of the vertebra there is a deeply concave articular surface for articulation with the third cervical vertebra. From the medullary canal, where this surface has no rims, its margins gradually rise and downward pass into a triangularly elongated form. Above the canalis vertebralis, along the entire length of the body of the vertebra, there projects a spinous process, which from the anterior margin toward the posterior gradually thickens and lengthens, and on the posterior edge, almost perpendicular to the axis of the medullary canal, ends in a sharp crest. At the base of the spinous process, to the sides of the posterior entrance into the medullary canal, the posterior articular processes depart at an angle, articulating with the anterior articular processes of the third cervical vertebra. The medullary canal, of rounded section, from the posterior entrance to the anterior noticeably

expands, and not far from the exit on the sides has, on each side, one nerve canal extending outward, of comparatively large size and round in shape. Below is given a table of measurements of the individual elements of the epistropheus.

IV, V, and VI cervical vertebrae (pl. II, figs. 5—7). Here, first of all, the strong development of the upper and lower transverse processes is striking, the upper of these being directed to the sides, and the lower downward. In external appearance, in general outline, all three vertebrae are similar to one another, and only details make it possible to distinguish them from each other. Thus, the upper transverse processes in the IV vertebra are comparatively longer than in V and VI, and are directed horizontally to the sides from the body of the vertebra; in V they are somewhat shorter, and their direction is noticeably upward; in VI they are still shorter, and their ends are directed distinctly upward. The lower lateral processes in the IV and V vertebrae, compared with VI, are thinner and directed more downward and to the sides. In front and behind, above the spinal canal, are the upper and lower articular processes; on the inner side they pass directly into the spinous process. Projecting backward, the upper articular processes at the base of the spinous process form a triangular notch, which in the IV vertebra is expressed rather considerably, in V less so, and in VI is scarcely noticeable at all. The length of the vertebral body from IV to VI, though slowly, decreases. The crest observed on the lower part of the body of the epistropheus is also present on IV and V,

on VI it is completely absent. The appearance of the anterior articular surface is distinctly convex, in the form of a shield, bounded above by the lower platform of the canalis vertebralis. The posterior articular surface has the appearance of an oval depression with its long axis directed vertically. The canalis vertebralis is of rounded form; its lower part is expressed as a platform. The neural canals, which are observed in the epistropheus and in the thoracic vertebrae, are absent here; instead, on all the above-mentioned vertebrae, along the body, on both sides, there pass vascular canals of rather considerable size and oval section, their position differing in the different vertebrae. In the IV vertebra the openings of the vascular canals are situated, if viewed from the anterior side, considerably above the upper lateral processes; in V, half of the openings lie opposite the same processes; in VI they cut through the processes exactly at the middle of their bases. Concerning the spinous processes, owing to their loss on all three vertebrae, it is impossible to say anything. The dimensions of the individual elements are shown in the table below, common to all three vertebrae.

Of thoracic vertebrae (Table III, fig. 9) there are only five in the Kemerovo collection; of these, only one vertebra has a completely preserved spinous process, while in the others the processes have been lost from their very bases. The complete disarticulation and incompleteness of the vertebrae do not permit even an approximate indication of their positions in the vertebral column. The absence of parapophyses, the diapophyses weakly developed in comparison with the cervical vertebrae, the presence of articular surfaces on the side of the upper part of the bodies for the capitulum of the rib, and similar facets, of a somewhat different appearance, on the diapophyses for the tuberculum of the rib, further, the excessively long spinous processes directed backward at an angle, and finally the weak development of the articular processes—these are the principal characters that generally characterize the thoracic vertebrae.

With more detailed study one may point out a further series of features inherent only in the thoracic vertebrae. Thus, the bodies of the vertebrae, in comparison with the cervical and lumbar ones, are noticeably shorter, and the ridge observed along the lower part of the bodies of the IV and V cervical vertebrae, and disappearing in the VI, here reappears and continues along all the thoracic and lumbar vertebrae as far as the sacrum, where it again disappears. On the sides of the body, in its upper part, in front and behind, are articular surfaces for the capitulum (head) of the rib, the anterior ones being somewhat convex, and the posterior ones concave. The articular surfaces for articulation with the neighboring vertebrae are not so convex and concave as can be observed on the cervical vertebrae described above. The shield-like elongated form of the articular surfaces on the cervical vertebrae here changes into a more rounded one. The neural canal in form and size differs little from the canal of the last cervical vertebrae, although toward the end of the series of thoracic vertebrae it noticeably assumes the form of an ellipse with its major axis directed horizontally. Along the lower part of the canal there passes a small ridge, which was not observed on the cervical vertebrae. Not far from the posterior outlet of the canalis vertebralis, on both sides, small-diameter nerve canals depart, subsequently dividing into two branches, one going upward, the other downward. The transverse processes, in comparison with the diapophyses of the last cervical vertebrae, are weakly developed and have a strictly horizontal direction with a slight upward bend of the ends. The facets for the tuberculum of the rib are located on the lower side of the transverse processes, nearer the ends; their form varies greatly depending on the position of one or another vertebra in the skeleton. The anterior and posterior articular processes, which, as we saw on the last cervical vertebrae, were in the form of large projections, are here barely noticeable. The form and position of the articular facets in different vertebrae are variable: at times they are elongated, roughly elliptical and situated close beside one another; at times they are almost rounded and separated from one another by a broad but shallow groove. The spinous processes, unfortunately, have not been preserved, with the exception of one, which in its length exceeds the vertebral body ninefold. The section of the process at the base is triangular; approximately from one-third of its length it changes into a lenticular form. In front, along its entire length, there passes a sharp ridge, while on the posterior side at the base there is a groove, which soon likewise changes into a ridge that continues to the end of the spinous process. Measurements of individual vertebrae are given in the following table.

Of the lumbar vertebrae (Table II, fig. 8), the Kemerovo collection contains only one, and that one is not in entirely good preservation.

The features characterizing the lumbar vertebrae are: the absence of parapophyses and of rib facets; the sharply elliptical form of the section of the neural canal; and a spinous process that is comparatively small. It may further be noted that the body of a lumbar vertebra is considerably longer than in the thoracic vertebrae, and on the lower side also has a longitudinal ridge. The articular surfaces are not so convex and concave as in the thoracic and cervical vertebrae; moreover, their form is more flattened from top to bottom. The diapophyses are moderately developed. The articular processes, both anterior and posterior, have a rather original form, namely: the posterior articular process has facet-like projections in the form of an S-shaped stamp; the anterior one, corresponding to the facet of the posterior articular process, articulating with it, embraces the articular facet on three sides.

All the indicated elements are clearly expressed in our specimen. The neural canal in transverse section has the form of an ellipse; along its lower side, lengthwise of the vertebra, there passes a sharp ridge. Neither vascular nor nerve canals are observed here. The spinous process, unfortunately, is broken off at the root. The dimensions of the individual elements of the vertebra in our collection are given in the following table:

Sacrum. Bison priscus (pl. III, fig. 10), in a quite well-preserved state, consists of six vertebrae, whose centra, upper arches, and spinous processes are tightly fused with one another. On the lower, slightly concave side, where in the thoracic and lumbar vertebrae a crest was observed, there passes a small groove. The centra of the individual vertebrae, elliptical in cross-section, become noticeably shorter toward the tail. The diapophyses, together with the anterior and posterior articular processes, having fused, form the appearance of wings; the length of the wings decreases from the first to the fourth vertebra, and then, toward the end of the sacrum, gradually increases again. The spinous processes from the first vertebra to the third gradually become longer, broader, and thicker, and farther toward the tail again become shorter, narrower, and thinner. The neural canal along the entire length of the sacrum has an elliptical cross-section, with the major axis directed horizontally. From the posterior boundary of the fusion of the vertebrae, a nerve canal departs from the canalis vertebralis, which immediately divides into two branches, one going upward, the other downward; the lower exit foramen is considerably larger than the upper. Dimensions of the sacrum:

Bones of the extremities. Humerus—the humeral bone (pl. III, fig. 11) in the skeleton of Bison priscus, among the bones of the extremities, is one of the massive ones; like all tubular bones, it consists of a diaphysis, or shaft, and of epiphyses. The more thickened proximal epiphysis, by means of which the humerus articulates with the scapula, consists of an articular head and tubercles: tuberculum major and tuberculum minor. Somewhat lower, in the same upper part of the bone anteriorly, there is a shallow but rather broad depression—fossa bicipitis, serving for the attachment of the tendon—musculus biceps. Here also, approximately from the middle of the diaphysis toward the outer side, a crest projects rather sharply—crista deltoidea. Narrowed downward, the humerus at the end, again widening, terminates in a block-like articulation (trochlea) projecting sharply forward. Along the inner half of the block run two grooves and between them a ridge, while the outer half is completely smooth. The diameter of the inner edge of the block considerably exceeds the diameter of the outer. Behind the block and above it there adjoins a rather deep fossa anconea for the reception of the oleocranon of the ulna.

Rütimeyer¹ and Bojanus² point out certain features in the structure of the humeral bone of Bison priscus that distinguish it from the humerus of other

closely related representatives of the family Bovidae. All these features amount to the following: 1) the diaphysis of the humerus of Bison priscus is more slender than in Bos primigenius and Bos taurus and is supplied with a narrower groove for the tendon of the biceps brachii, as well as with a wider, but less deep, fossa above the posterior surface of the trochlea (fossa supratrochlearis posterior); 2) the trochlea is shorter, and its narrowing from the inner end to the outer proceeds more rapidly than in Bos primigenius and especially in Bos taurus; and 3) the ridge separating the outer part of the trochlea from the inner is less sharp, but broader, and the angle formed by the axis of the trochlea with the axis of the entire bone is straighter than in the aurochs. But Cherskii¹, in studying the remains of fossil bison, does not consider these characters entirely stable, which is also confirmed by my observations in comparing the humerus of Bison priscus from Kemerovo with a humeral bone of Bos taurus kept at the museum of the paleontological cabinet of the Siberian Technological Institute. Measurements of individual elements of the humerus show the following:

Radius et ulna (Table III, fig. 12) in representatives of the family Bovidae are always fused with one another; in Bison priscus one can always observe, at the proximal end in the last quarter of the fusion, a free space of an inch, and sometimes more, separating the radius from the ulna.

Radius, being a tubular bone more developed than the ulna, has, almost throughout its entire length, an approximately elliptical cross-section; moreover, it thickens noticeably toward both the proximal and the distal ends. The proximal end has one depression situated nearer the inner margin of the bone, another of smaller dimensions nearer the outer margin, and between them a groove, which, taken together, corresponds to the block of the humerus. At the distal end the relief of the articular surface is more complex than at the proximal. Here, beginning from the inner side, the bones of the first row of the carpus articulate: the scaphoid (scaphoideum), the lunate (lunare), and the pyramidal (piramidale); moreover, the facet for the pyramidal bone is situated partly on the radius and partly on the ulna.

The ulna, from the distal end to the articular area of the proximal, is closely fused with the radius; this fusion is observed more intensively, almost without traces of sponginess, on the lower part, opposite the facet for the pyramidal bone. Higher up, the trace of sponginess is clearly visible in the form of small grooves running along the fusion. In the last quarter of contact, as I indicated above, the ulna over the course of one or two inches departs from the radius, but then again grows tightly into it. Still higher, before the articular surface, one can see not fusion, but only articulation with the radius, expressed in the form of a zigzag-like fissure. The ulna, in comparison with the radius, is a weakly developed bone; its thickness, changing little in the first quarter of its length, higher up, gradually widening, ends in a broad process—the oleocranon. The oleocranon, having a deviation toward the inner side, in its anterior part, above the upper boundary of fusion with the radius, has an arcuately concave facet for articulation with the humerus. Measurements of the separate parts of the radius et ulna give the following values:

Femur (Pl. IV, fig. 13) in the bison skeleton is one of the long tubular bones; its length closely approaches that of the tibia. The cross-section of the femoral diaphysis in its middle part is nearly rounded in form. Only on the posterior side, at the boundary with the external side, a small ridge projects weakly; toward the distal end, somewhat below the middle of the bone, it divides into two branches, one of which goes to the external side, while the other emerges toward the middle of the external articular condyle. The proximal epiphysis consists of the head, the neck, and the trochanter major. Between the neck and the greater trochanter, on the posterior side, there is a deep oval depression, below which, toward the inner part of the diaphysis, lies a cone-shaped projection—the lesser trochanter (trochanter minor). The head of the femur, serving for articulation with the pelvis, has an ovoidly convex surface, sharply deviating to the inner side from the longitudinal axis of the femur. Almost at the summit of the articular head one can observe a small, round depression. The greater trochanter projects as a sharp crest from the external side of the proximal epiphysis, rising above the head. At the distal end there are two articular condyles, departing from the posterior side of the femur: entocondylus and ectocondylus, which are separated from one another by a deep fossa intercondyloidea. On the anterior side of the distal epiphysis there projects the patellar facet, expressed by two rounded crests and a groove between them. The inner crest projects considerably more than the outer one and, in its upward continuation, passes directly into the crest of the femoral diaphysis. At the boundary between the condylus and the crests of the patellar facet there are two rather large depressions, the external of them being larger than the internal. A similar depression may be observed closer to the external side of the lower part of the ectocondylus. The measurements of the individual elements of the femur are as follows:

Tibia Bison priscus (Pl. IV, fig. 14) is only slightly shorter than the femur in length; it belongs to the tubular bones and consists of a shaft and epiphyses. At the distal epiphysis and above it the section of the diaphysis is elliptical; but, beginning from the boundary of the start of the second third of the shaft, it becomes sharply triangular, the posterior part of the diaphysis being flat, while the anterior projects in the form of a large crest (crista tibiae). The upper epiphysis has two articular surfaces, with two small crests projecting between them, which, when articulated with the femur, enter the fossa intercondyloidea. The inner articular surface is noticeably wider than the outer. The distal, somewhat thickened end has an articular surface in the form of two large boat-shaped depressions and two small facets situated on the outer edge of the epiphysis.

Of the small bones of the limbs I have: metacarpus right and left, all phalanges of the right foreleg (Pl. IV, Fig. 16), astragalus right and left, calcaneum right and left (Pl. IV, Fig. 15), metatarsus right and phalanges of the right hind leg—IV₁ and IV₃ (Pl. IV, Fig. 17). Unfortunately, all these bones have undergone considerable weathering, and therefore it is not possible to dwell in detail on their full characterization. Below I give tables of measurements of individual elements of the above-mentioned bones.

Astragalus—talus:

Calcaneum—calcaneal bone (pl. IV, fig. 17):

Phalanges of the fore and hind limbs:

Conclusion. Besides the purely paleontological value of the collection described, which is comparatively rare in its completeness, it also has general geological significance, since it makes it possible to determine rather precisely the age of that loose sandy-clayey series in which our collection was found.

Prof. M. E. Yanishevsky¹ divides the Cenozoic deposits of Tomsk Province into two parts: the weakly cemented sandstones lying at its base, and the sands and clays, up to the pebble layer, that overlie them and are developed in the area of the city of Tomsk. On the basis of the plant remains studied by him he assigns the former to the Miocene; the pebble beds and the overlying loose sandy-clayey formations he places in the post-Pliocene, on the basis of finds in them of freshwater and terrestrial mollusks, as well as remains of bones of Mammalia. However, these considerations still do not give an exhaustive answer to the question of the age of the loose sandy-clayey series of Tomsk Province, since M. E. Yanishevsky gives no precise indications either of the localities of the post-Pliocene fossils or of the stratigraphic position of the beds that yielded these remains.

The fossil bones of Elephas primigenius and Bison priscus described above make it possible to confirm and refine the opinion regarding the post-Pliocene age of the loose series of Tomsk Province, since all of them were found in one definite bed, underlying the loess-like loams and constituting the upper horizon of one lacustrine formation, as is shown in fig. 19.

¹ Yanishevsky, M. E. On the Miocene flora occurring in the environs of the city of Tomsk. Transactions of the Geological Committee, New Series, issue 131, 1915.
Idem. Preliminary report on a student excursion along the Tom River in 1909. Proceedings of the Tomsk Technological Institute, vol. XXIV, no. 4, 1911.

It is interesting to compare with this the fact of my finding remains of Elephas primigenius in the vicinity of the city of Tomsk, along the B.-Kirgizka River, likewise directly beneath loess-like loams, in a layer of weakly diagonally bedded clayey sands which, as at Kemerovo, constitute the upper member of the Tomsk lacustrine formation.

Thus, the lacustrine regime in Tomsk Province ended in the age of Elephas primigenius and Bison priscus; in that same age there began the regime of a continental-steppe climate, which produced a thickness of loess-like loams, likewise in places containing remains of Elephas primigenius Bl.¹).

Explanation of the Plates.

Plate I.

Elephas primigenius Blum.
Fig. 1. Lower molar tooth, m₃, from the side; ⅓ natural size.

Bison priscus H. v. Meyer.
“ 2. Part of the skull; ¼ natural size.
“ 3. Lower jaw (part); ½ natural size.

Plate II.

Bison priscus H. v. Meyer.
“ 4. Epistropheus from the side; ⅓ natural size.
“ 5. Sixth cervical vertebra from behind; ⅓ natural size.
“ 6. Fifth “ “ “ ⅓ “ “
“ 7. Fourth “ “ “ ⅓ “ “
“ 8. Lumbar vertebra from the front; ⅓ “ “

Plate III.

Bison priscus H. v. Meyer.
“ 9. Thoracic vertebra; ¼ natural size.
“ 10. Sacrum from below; ⅓ natural size.
“ 11. Humerus, right; ⅕ natural size.
“ 12. Radius et ulna, right; ⅕ natural size.

Plate IV.

Bison priscus H. v. Meyer.
“ 13. Femur, right; ⅕ natural size.
“ 14. Tibia, right; ⅕ natural size.
“ 15. Calcaneus (right leg); ⅕ natural size.
“ 16. Os metacarpi et phalangi digitorum (right foreleg); ⅕ natural size.
“ 17. Os metatarsi et phalangi digitorum (right hind leg); ⅕ natural size.

Plate V.

“ 18. Plan of the Kemerovo mine.
“ 19. Geological section at the place where the bones were found.

¹) Kashchenko, N. On the finding of mammoth remains near the city of Tomsk. 1896.

Plate I.

Fig. 1

\( \frac{1}{3} \)

Fig. 2

\( \frac{1}{4} \)

Fig. 3

\( \frac{1}{2} \)

Plate II.

Fig. 4

Fig. 5

Fig. 6

Fig. 7

Fig. 8

Plate III

Fig. 9

Fig. 10

Fig. 11

Fig. 12

Plate IV

Fig. 13

\( \frac{1}{5} \)

Fig. 14

\( \frac{1}{5} \)

Fig. 15

\( \frac{1}{3} \)

Fig. 16

\( \frac{1}{5} \)

Fig. 17

\( \frac{1}{5} \)

Plate V

Fig. 18

Plan of the Kemerovo mine
Scale: 0.01–200 sazhen.

Legend:

• Brick and stone buildings and structures.
• Boundary of the ore allotment.
• Place where the bones were found.

Fig. 19

Geological section

Coal.

Post-Pliocene.

• Forest.
• Dark loess with another layer.
• Layer of clay.
• River sand.
• Pebbles.
• Talus.

Carboniferous deposits:

• Clayey shales.
• Coal.
• Sandstones, clayey shales.

1. Cherskii, I. D. Description of the collection of post-Tertiary mammals, collection of the Novo-Sib. Exped. 1885–86. 1891; 178–183. ↩↩↩↩

2. Nova Acta Acad. Nat. Cur. XIII. 2. 1827. ↩↩