their leaves. Many of the Angiosperms are evergreen, so
that it cannot be said that the one change entailed the other. In fact,
a careful study of the leaves preserved in the rocks seems to show
the deciduous Angiosperms gaining on the evergreens at the end of the
Cretaceous. The most natural, it not the only, interpretation of this is
that the temperature is falling. Deciduous trees shed their leaves so as
to check their transpiration when a season comes on in which they cannot
absorb the normal amount of moisture. This may occur either at the
on-coming of a hot, dry season or of a cold season (in which the roots
absorb less). Everything suggests that the deciduous tree evolved to
meet an increase of cold, not of heat.

Another suggestion is that animals and plants were not "climatically
differentiated" until the Cretaceous period; that is to say, that they
were adapted to all climates before that time, and then began to be
sensitive to differences of climate, and live in different latitudes.
But how and why they should suddenly become differentiated in this way
is so mysterious that one prefers to think that, as the animal remains
also suggest, there were no appreciable zones of climate until the
Cretaceous. The magnolia, for instance, flourished in Greenland in the
early Tertiary, and has to live very far south of it to-day. It is much
simpler to assume that Greenland changed--as a vast amount of evidence
indicates--than that the magnolia changed.

Finally, to explain the disappearance of the Mesozoic reptiles without a
fall in temperature, it is suggested that they were exterminated by
the advancing mammals. It is assumed that the spreading world of the
Angiospermous plants somewhere met the spread of the advancing mammals,
and opened out a rich new granary to them. This led to so powerful
a development of the mammals that they succeeded in overthrowing the
reptiles.

There are several serious difficulties in the way of this theory. The
first and most decisive is that the great reptiles have practically
disappeared before the mammals come on the scene. Only in one series of
beds (Puerco) in America, representing an early period of the Tertiary
Era, do we find any association of their remains; and even there it
is not clear that they were contemporary. Over the earth generally the
geological record shows the great reptiles dying from some invisible
scourge long before any mammal capable of doing them any harm appe