P. N. Bose—History of the Extinct Carnivora. 275 
instance, Glisorex ferrugineus, is striking; at the same time such a 
view would not be open to the objection just raised, for in the 
Insectivora the molars very generally decrease in size and differentia- 
tion from the first to the third. -dmphicyon has seven molars on each 
side like Cynodon; the lower carnassial is well specialized, and the 
back molars assume the normal “ tuberculous ” form. 
Exceptional as the molars of Palceonictis, Cynodon, and Uintacyon 
are, those of Proviverra, Hyenodon, Pierodon, Ambloctonus, Stypo- 
lophus, and Oxyena, are doubly exceptional. In these eight genera, 
to the peculiarity of a succession of carnassials, is added that of 
increase in differentiation, and very generally also in size, from the 
homologue of the normal flesh-cutting tooth, to that of normally the 
most ‘tuberculous,’ and generally the smallest molar. In Pro- 
viverra, the three molars increase rapidly in size from the first to the 
third; each is composed of three pointed cusps and a horizontal 
“heel.” And it is interesting to note that the first molar, in which 
the primitive antero-external and the accessory cusp in front are the 
least differentiated, has the “heel” best developed. In the last 
molar both these cusps are developed in the form of a blade, some- 
what as in Hyenodon, and the ‘heel,’ at the same time, becomes 
most insignificant. The second molar beautifully presents an inter- 
mediate condition. The Stypolophus of Cope differs but slightly 
from Proviverra, and I propose to incorporate the former with the 
latter. Oxyena differs from Proviverra only in having two molars 
instead of three. In Ambloctonus the first two molars are subequal, 
and the last molar is a little smaller. But the latter is the most 
trenchant and best differentiated, the blade being very much like 
that of Hyenodon. If in each of the three molars of Proviverra the 
posterior and inner denticule (the representative of the original 
antero-internal cusp) be entirely suppressed, the anterior accessory 
cusp better developed, so as tou form with the primitive antero- 
external cusp a blade-like tooth, and the “heel” correspondingly 
reduced, we would have asa result the molars of Pterodon. The 
step from Pterodon to Hyenodon is very easy. If in the former the 
teeth became more trenchant, the anterior accessory cusp still more 
differentiated, and, as a necessary consequence, the “heel” all but 
non-existent, we would have the teeth of Hyenodon. This progres- 
sive modification from Proviverra to Hyenodon is not simply a matter 
of speculation. In America remains of Proviverra have been met 
with in beds which are homotaxial with the Woolwich and Reading 
beds of England and the “ Lignites du Soissonais” of France. In 
Europe Proviverra has been found associated with Plerodon in 
Middle Eocene deposits of the same age as the Bracklesham Series 
of the Hampshire basin. Hyenodon appears at the commencement 
of the Upper Hocene (or Oligocene) Period, being first met with 
in the Hordwell beds (Barton Clays). Both Proviverra and Pterodon 
die out at the commencement of the Miocene Period. Hyenodon held 
out much longer, living into the Middle Miocene. 
Our knowledge of the teeth of Putriofelis and Sinopa is very 
imperfect. We are not much more fortunate with the dentition of 
