22 PROCEEDINGS OF THE GEOLOGICAL SOCIETY. [June 16, 
Had all the molars above described belonged to the permanent 
series, they would have differed from those of the Paleeothere not 
only in their minor number—six instead of seven—but also in the 
simpler form of the last molar, as well as by the tubercle at the sum- 
mit of the posterior ridge in the three last molars. A small orifice, 
however, in the alveolar border (fig. 3, 3), behind the last molar 
in place, led me to suspect the existence of another molar in a con- 
cealed alveolus, and on removing a portion of the outer wall of 
the jaw, the germ of such a molar was exposed, consisting of the 
shell of the crown (fig. 4, 3). This seemed to make the number 
of grinders the same as in the Paleotherium, but the hidden grinder 
resembled the crown of the last one im place, and had not the third 
lobe which characterizes the last true molar of the Paleotherium ; 
whilst the second grinder (D 2) still opposed its small size and uni- 
lobate structure to an accordance with the generic dental character of 
the Paleotherium. Observing, next, that the fourth grinder (D 4), 
in the jaw from Hordle, had been much more worn by mastica- 
tion than the succeeding tooth (M 1)—now proved to be the first of 
the three true molars—I was led to suspect that D 4 might be a de- 
ciduous tooth, knowing that the permanent one when in this position 
has its crown less worn im all known Herbdivora than the first true 
molar behind it ; and that it comes into place either at the same time 
as the last true molar or a little later. The degree of attrition of 
the third tooth in place (D 3) indicated that it also might be one 
of the deciduous series; and the extent to which the fangs of all 
the four anterior grinders were exposed, together with some minute 
foramina on the inner border of their sockets, determimed me to search 
for traces of successors beneath them. ‘The search was successful, 
and the germs of the permanent premolars, not indeed equal in num- 
ber, but almost co-extensive, with the four anterior grmders which 
they would have displaced and succeeded, were discovered (fig. 4, 
P 2, 3,4). These germs consist of the entire crown widely excavated 
at the base. The first (2b. P 2) is chiefly composed of one semi- 
conical lobe with a basal ridge interrupted at the outside of the base 
of the cone, but developed behind into a wing-like appendage, o!: the 
apex of the crown points below the interspace of the first and second 
deciduous molars, both of which it would have displaced. The second 
premolar (74. P 3) lies beneath the third milk-molar; it resembles 
the first premolar in shape, but is of larger size, and has the hinder 
wing-like process (0’) more developed : this process presents a flat and 
nearly vertical surface outwards, and is convex at the inner and back 
part of its base. The third premolar (6. P 4) consists of two 
semicylindrical lobes like the true molars and like the last milk- 
molar which it would have succeeded: like them, also, it has no 
basal ridge on the inner side of the crown, and that on the outer side 
is interrupted and represented by the oblique anterior and posterior 
ridges, the latter of which terminates in the characteristic tubercle ¢. 
In all existing hoofed quadrupeds (Proboscidians excepted), the last 
milk-molar presents and foreshadows, so to speak, the peculiarity in the 
form of the last true molar, when this tooth differs from the others; 
