SCHWARZ — EUROPEAN FOSSIL HORSES 
151 
The evidence afforded by fossil bones and teeth found in Europe, 
Asia, and Africa appears to admit only two general types. In the 
species named stenonis^ atlanticus, persicus, etc., the anterior pillar 
(protocone) of the premolars and molars is short (generally only one 
third of the total length of the tooth) with the anterior lobe scarcely 
developed and the interior margin mostly convex, straight or only 
slightly bent in at the middle. The enamel folds are moderately 
developed, the paracone and metacone are thick and their lateral sur- 
face convex, and the parastyle and mesostyle narrow and, if at all, 
only doubled in the anterior premolars. The eye sockets are some- 
what projecting, and the angle of the lower jaw is deep. In the limbs 
the lower portions (radius-metacarpal and tibia-metatarsal) are com- 
paratively long and narrow, the upper portions (humerus, femur) 
shorter than in the next species.^*’ 
In the other type to which such names as caballus, plicidens, robus- 
tus, siissenbornensis, mosbachensis, przewalskii, etc., have been applied, 
the anterior pillar of the cheek teeth is generally longer (usually more 
than one third of the total length of the tooth) with the posterior lobe 
always and the anterior one in many cases of considerable size, the 
interior margin being always bent in or even sharply divided at the 
middle. The enamel folds are more pronounced, the paracone and 
metacone distinctly lunate and their outer border concave, and the 
parastyle and mesostyle more imposing and doubled in most of the 
cheek teeth; the eye sockets are not projecting and the angle of the 
lower jaw is much less deep. The metatarsals and metacarpals are 
as a rule comparatively shorter and broader in specimens from older 
strata but about as slender as in the former species in more recent ones. 
The upper portions of the limbs (humerus, femur) are, however, some- 
what longer compared to the lower portions (tibia, metatarsus, etc.). 
It has hitherto been supposed that Equus sivalensis belongs to the 
first or stenonis type. But a renewed examination of Mr. Lydekker’s 
beautiful plates^^ makes me believe that it is a member of the caballus 
series. The shape of the protocone, paracone and metacone and of 
the para- and mesostyle are decidedly of this type. This is still more 
so in Equus namadicus which is probably the direct descendant of 
E. sivalensis and apparently similar in the Ceylon tooth described by 
Lydekker, Guide to the specimens of the Horse family, p. 26 (1907). 
Mem. Geol. Surv. India (Palaentologia Indica), (10) II, pi. XIV, XV (1882). 
