of the proximal phalanges. It is perfectly whole, or "solid", as seen in the figure. In frout, there is a slight, though evident, vertical line of impression along the middle, indicating its composition from lateral halves. On the sole of the hoof, there is a broad, angular elevation of horny substance, apex forward, and sides runding backward and outward to the lateral borders of the hoof, the whole structure being curiously like the frog of the horse's hoof. In fact, it is a frog, though broad, flattened, and somewhat horseshoe-shaped, instead of being narrow, deep, and acute, as in the actual frog of the horse. This arcuate thickening of the corneous substance occupies about the middle third of the whole plantar surface of the hoof.

Viewing the apparent establishment of this pseudo-perissodactyle structure in an artiodactyle, the question arises whether we have not, under our eyes, an example of a way in which a solidungulate may be evolved from a pluridigitate stock-though of course the one case is by enlargement of a single median digit and reduction of lateral digits, while in the present instance a bone in the axis of the limb is produced by failure of fission between lateral paired digits. Nothing is more certain than that the present solid-hoofed horse has come by direct descent, with modification, from its several-toed ancestors of the Tertiary. In the present case, we seem to have the initial steps of an actual transformation which may in time result in modifications to which ordinal value may attach. It may be suggested that this modification is one of progressive adaptation of the animals to their freely-ranging state on the prairies of the country, just as the series of modifications which the primitive horse's foot has undergone in adaptation to the making of the most serviceable hoof for running on hard ground at the expense of any other function.

ART. XIV.-PROFESSOR OWEN ON THE PYTHONOMORPHA.

BY E. D. COPE.

The British Museum has recently obtained the Van Breda collection of fossils, which includes a valuable series of Mosasauroid remains from Maestricht, the locality which furnished to Cuvier the typical specimen of the Mosasaurus giganteus. Professor Owen has improved the opportunity to study this material with that already in possession of the museum, some of which was derived from North American sources.

In pursuing this subject, Professor Owen has done me the honor to study my contributions to it, a summary of which appears in the second volume of the Final Report of the United States Geological Survey of the Territories under Dr. F. V. Hayden. He follows my determinations and conclusions, and criticises them in the light of his long experience. As a portion of this criticism is adverse to what he supposes my conclu. sions to be, I propose on the present occasion to give such a brief review of Professor Owen's paper* as my other immediate occupations will permit. I premise that this cannot now include a complete review of the subject, nor the exposition of several parts of it which have not yet received the attention of Professor Owen or of any one else.

Professor Owen's references to my work may be included under three beads, viz:-First, as to matters of fact or observation; second, as to determination of homologies of parts; third, as to the estimation of affinities as derived from the preceding branches of the subject. I now consider

I. QUESTIONS OF FACT.

The many observations as to the structure of the order of Pythonomorpha recorded by me in the volume already referred to are confirmed by Professor Owen with a single exception. He correctly describes the vertebræ of the genus Mosasaurus as without the zygantrum and zygo sphene articulation, and proceeds to say (p. 709), in reference to my ascription of this structure to the genus Clidastes, that the structure of Mosasaurus "is repeated in plates xviii, xix, xx, xxi, xxiii, xxiv, xxvi, xxvii, xxix, XXX, xxxiv, and xxxv of Professor Cope's great work; in every figure the zygosphene and zygantrum are absent." And again,—“In the plates xviii and xxiii given to the vertebræ of the species [Clidastes] stenops and planifrons, the parts and processes are as usual not indicated." All this is a remarkable oversight on the part of Professor Owen. He *Quarterly Journal of Geological Society, London, 1877, p. 682.

will find the zygosphene distinctly represented on figs. 5 a and 5 b, pl. xviii; figs. 3 b, 3 d, 6 b, 6 c, pl. xix; fig. 15 d, pl. xxi; figs. 3 c and 3 d, pl. xxiii.; fig. 4, pl. xxiv; and the zygantrum in nearly as many figures. He will also find them well represented in the figures of vertebræ of Clidastes on plates v and xii of the Extinct Batrachia and Reptilia of North America. In order to substantiate his position, he copies from my work a figure of a vertebra of Clidastes stenops from which the zygosphene has been accidentally broken away.

Professor Owen places me in the attitude of committing error in questions of fact in regard to the limb-bones and their arches in the Lacertilia and Ophidia. My statement is," As there are many Lacertilia without limbs, and some serpents with them, their presence in this order is irrelevant in this connection, especially as the arches supporting them are most like those of tortoises and Plesiosaurs." Professor Owen then proceeds to state that there are only twenty-three geuera of Lacertilia with reduced limbs, and "extremely few" where they may be considered to be rudiments. Professor Owen can hardly have had in mind the developments of herpetology during the last five or ten years in making this assertion; for the genera of lizards now known in which the limbs are rudimental may safely be said to be numerous, and those without even rudiments are not a few. Professor Owen appears to have overlooked the entire suborder of the Amphisbania, which are all limbless with the exception of one genus. He then criticises my reference to serpents with limbs, and observes:-"In certain Ophidia dissection has revealed a small styliform bone on each side the cloaca; in a few it is tipped with horn in the shape of a claw. . . Whether these appendages to the generative parts be homologous with the claspers' of sharks or with the ventral fins, and, if the latter, with the hind limbs of lizards, is yet an open question." Reference to the numerous genera and species of serpents which possess rudimental hind limbs, as well as to the two suborders which possess a pelvis, is here entirely omitted, and the demonstration of the homology of the anal claws above mentioned with true hind limbs appears to be unknown to Professor Owen. Besides the Boida, Pythonidæ, and Xenopeltidæ known to Professor Owen as possessing these rudimental limbs, there are the Lichanurida, Tortricidæ, and Stenostomida; while the Typhlopida and Stenostomida possess a pelvis-the latter family with ilium, ischium, and pubis, as ascertained by Peters. This pelvis is more complete than that of various Lacertilian genera of the Diploglossa group, or of the suborder of the Amphisbænia, which consists, according to Stannius, of a rudimental ilium only. My statements on this point are borne out by the facts. My assertion as to the resemblance of the scapular and pelvic arches to those of tortoises and Plesiosaurs is true in view of the fact that the former has no inferior connection with a sternum, so far as known, an element absent in the orders named and the Ophidia, but present in the lizards, although not universally so.

Another question of fact is raised in regard to the possibility of the lateral horizontal flexure of the mandibular ramus in the various genera of Pythonomorpha. My critic states,-"In Python the outer plate of the dentary is deeply notched behind by a long angular depression which receives a process of similar shape of the angulo-surangular element. In Mosasurus as in Monitor, the outer plate of the dentary terminates in a subvertical line; this is curved in Iguana, less so in Monitor, still less in Mosasaurus, where it seems to have suggested to Professor Cope the idea of a movable articulation with the hinder part of the ramus: but the relative overlapping position of the mandibular elements, causing the angular break of the line" [of the posterior border of the dentary] "on the outer side of the ramus, and in a great degree of the inner surface of the ramus, must have as effectually opposed such flexion in Mosasaurus, as is the case with Lacertians and a fortiori with Ophidians."

I have not had the opportunity of studying a perfect mandibular ramus of a species of the genus Mosasaurus; but I have numerous mandibles of Platecarpus, Liodon, and Clidastes. In all of these, the mobility is indicated by the character of the adjacent extremities of the segments of the lower jaw, as well as by the form of the proximal end of the os quadratum, by which that jaw is mediately articulated with the skull. There is no "overlapping of the mandibular elements causing the angular break" in these genera, either in the horizontal or vertical lines, although the inferior portion of it, where the ball-and-socket articulation is found, forms a slight angle with the remaining portion of the hinge. The anterior extremities of the surangular and coronoid are contracted to an obtuse edge, which fits into a groove or rabbet of the dentary and splenial elements, so as to form a movable joint, the two segments of the ramus being held together by a lamina of bone which in life was doubtless perfectly flexible. This flexure is rendered necessary when the jaw is opened widely by the form of the proximal end of the os quadratum. This extremity forms a sliding joint with the inferior face of the opisthotic; and as it is bent or curved in form, its movement necessarily causes a 1otation of the quadraté round its vertical or long axis. This rotation of course throws the proximal part of the mandibular ramus outward; and to permit this movement, the joint near the middle of the latter is clearly adapted. The degree of flexure is dependent on the degree of rotation, and that in turu on the curvature of the prox imal end of the quadrate. This curvature depends on the development of the "proximal internal angle", which is very large in Clidastes and Liodon, and smaller in Mosasaurus. It is possible that the power of flexure was small in the latter genus, and that Professor Owen's conclusions in the matter may be due to imperfect material.

Under the head of matters of fact may be mentioned a few points in the history of the discovery of the structure of the Pythonomorpha. I have claimed in my work that the discovery of the hind limbs and much