293 
the teeth of all the mammals of the Lance Creek stage, except those of 
the Allotheria, are triangular, showing that the possessors were either in- 
sectivorous or flesh-eating in their habits. 
On the other hand, there are several genera of Puerco mammals that 
possess a well developed hypocone and internal cingulum. In some cases, 
where the hypocone had no great development, the hinder internal part of 
the tooth had swollen so as to reduce much the gap between the successive 
teeth and produce a broad triturating surface. In Polymastodon, which 
must have been a vegetarian, an extensive triturating surface was secured 
in another way. It presents a great advance over the teeth of any of the 
Lance Creek Allotheria. If it is considered how slowly changes in tooth 
structure had advanced during the Mesozoic era we must conclude either 
that a considerable interval had elapsed between the Lance Creek epoch 
and that of the Puerco or that the animals of the latter were not de- 
scendants of the former. 
There are important differences between the mammals of the Lance 
Creek beds and those of the Puerco as regards the size attained. Most of 
the former are of insignificant proportions, resembling in this respect those 
of the Jurassic; while many of those of the Puerco are large. Further- 
more, there was in the mammals of the Puerco a far greater variety of 
form, structure, and systematic relationships than among those of the 
Lance Creek mammals. Of the latter, there have been described about 
twenty-five genera and about forty-five species, most of them by Marsh. 
Osborn has regarded himself as justified in reducing these to about ten 
genera, these representing a very few families. From the Puerco Matthew 
(Bull. 361, U. S. Geol. Sury., 1909, p. 91) recognizes twenty-nine species, 
belonging to eighteen genera and nine families. To what extent this in- 
creased diversification of the mammalian life of the Puerco is due to im- 
migration we can not now tell; but it does not seem to be necessary to 
assume that it was due to invasion of mammals from some other region. 
Tor, in view of the interval between the two formations that is indicated 
by the plants and reptiles, it is possible that the Puerco mammals are the 
direct descendants of those of the Lance Creek epoch. 
In case there was no serious interruption in deposition between the 
Lance Creek beds and the Puerco and Fort Union, one might expect to 
find close relationships between the reptiles of the two levels. Crocodiles 
are not abundant in either and, so far as known, no species passes from 
