470 THE STRUCTURE AND RELATIONSHIPS Ot ANCODUS. 
desceuds less below the inferior border of the horizontal ramus. The coronoid pro- 
cess is less elevated and recurved than in the American species ol /Incodus, more 
than in the European, and the condyle is much less extended transversely and 
especially on the outer side. The symphyseal region is much more steeply inclined 
and less procumbent. In all of these particulars the later oreodont genus, Meryco- 
cluerus, presents a decidedly closer approximation to A^icodus than does Oreodon 
itself, though never attaining such an extreme elongation of the face. 
II. THE VERTEBRAL COLUMN. 
The atlas differs from that of Oreodon in being proportionately longer and less 
extended transversely. The anterior cotyles are separated dorsally by a wide and 
deep emargination of the neural arch ; the latter is quite strongly convex from side 
to side, and the neural spine is represented by a prominent rugose tubercle. A lyrate 
area, formed by ridges, encloses the spine and descends abruptly at the sides, which 
are perforated by foramina for the first pair of spinal nerves. The transverse pro- 
cesses are not greatly extended laterally, nor do they reach so far back of the sur- 
faces for the axis, as is the case in Anoplotheriim, though their shape is more like 
what we find in the latter genus than that in Oreodon. So far as I can determine, 
the transverse process is not perforated by the vertebrarterial canal. The facets for 
the axis are large and but little oblique in position, presenting more backward than 
inward. 
If Kowalevsky has correctly referred to Ancodiis the axis from Puy (Ronzon), 
which he has hgured (No. 3, PL XXXIV, fig. 7), then the American species of the 
genus differ very radically from the European with regard to the character of this 
bone. There can, however, be but little doubt that Kowalevsky’s specimen has 
been erroneously identified, and that it belongs to some very difterent genus, proba- 
bly a perissodactyl {Ronzotheriumf). In A. brachyrhynchus the axis has a lom>- 
broad, and much depressed centrum, which is but feebly keeled upon the ventral 
side and has no hypapophysial tubercle. The anterior cotyles for the atlas are very 
broad, extending out laterally much beyond the rest of the centrum, but, at the same 
t.,„e.ll,eyl,ave„o great vertical height and do not reach far enough upward to 
enclose more t „n a small part of the neural canal. The articular surfaces of these 
cotyles .re s^dle-shaped, slightly concave transversely and convex dorso-ventrnlly. 
ca ie, he wlTT"’ of the shape which might almost be 
pXlriZ lZlZf h’ '■ '-’f Agriock«:rus, Ancodus, 
betlen the two n 1 .“.“ j "" but iulermediate 
ve„t::,:r":ctrfxt:t^iirrrfyZttr 
and the process gradually controls a„terlor(y uli’ti t endnifa hi®'‘; ^ T 
have elsewhere shown the great probabilitv of il • n j ^ ^ 
charachr h, 
