uld be generally in one direction, both might be varied at particular
seasons. So far as concerns the distribution and thickness of the
glacial deposits, there is not much to choose between either hypothesis;
but on that of land-ice it is extremely difficult to explain the
intercalation of perfectly stratified sands and gravels and of
boulder-clay, as well as the not infrequent signs of bedding in the
latter."

Now with regard to the land-ice theory, several serious difficulties
present themselves in connection with the origin of the European fauna.
In the first place, as the climate renders Northern Siberia almost
uninhabitable for mammals at the present day, how much more severe must
it have been during the time of the maximum glaciation in Europe. As the
then existing fauna was not driven into Europe, where could it possibly
have survived? Secondly, how can we reconcile the contemporaneous
existence of a great inland sea (the Aralo-Caspian) containing survivals
of mild Sarmatic times with an immense glacier almost touching it on its
northern shores? How did one of the most characteristic species of that
sea, _Dreyssensia polymorpha_, come to make its appearance in the lower
boulder-clay of Prussia and then disappear in the upper? And finally,
how are we to explain the sudden appearance of a Siberian fauna after
the deposition of the lower boulder-clay, except by the removal of a
barrier which had prevented their egress from Siberia?

If we assume that the continental boulder-clay of Russia has been formed
in the manner so ably explained by Murchison, de Verneuil, and von
Keyserling, viz., by a sea with floating icebergs, the temperature of
Siberia might have been higher than at present, and have supported a
fauna in more northern latitudes.

The contemporaneousness of the deposits of this sea with those of the
Aralo-Caspian is also rendered more intelligible. If we suppose,
moreover, the connection between the Aralo-Caspian and the White Sea
(Fig. 12, p. 156) to have existed at this time, we possess an
explanation of the method of migration of the Arctic marine species into
the Southern and of the Caspian species (_Dreyssensia_) into the
Northern Sea.

An inter-glacial phase is believed to have supervened after the
deposition of the lower boulder-clay, and it is during this period that
the Siberian species first appeared in Central Europe. If we assume then
that the retreat of the Northern Sea (Fig. 13, p. 170) o