so during the Cretaceous epoch, and it is assumed that the great
Australia-Antarctic-Patagonian land was severed first to the south of
Tasmania and then at the South Georgian bridge. No connection, and
this is important, is indicated between Antarctica and either Africa or
Madagascar.

So far we have followed what may be called the vicissitudes of the great
Permo-Carboniferous Gondwana land in its fullest imaginary extent,
an enormous equatorial and south temperate belt from South America to
Africa, South India and Australia, which seems to have provided the
foundation of the present Southern continents, two of which temporarily
joined Antarctica, of which however we know nothing except that it
exists now.

Let us next consider the Arctic and periarctic lands. Unfortunately very
little is known about the region within the arctic circle. If it was all
land, or more likely great changing archipelagoes, faunistic exchange
between North America, Europe and Siberia would present no difficulties,
but there is one connection which engages much attention, namely a land
where now lies the North temperate and Northern part of the Atlantic
ocean. How far south did it ever extend and what is the latest date of
a direct practicable communication, say from North Western Europe
to Greenland? Connections, perhaps often interrupted, e.g. between
Greenland and Labrador, at another time between Greenland and
Scandinavia, seem to have existed at least since the Permo-Carboniferous
epoch. If they existed also in late Cretaceous and in Tertiary times,
they would of course easily explain exchanges which we know to have
repeatedly taken place between America and Europe, but they are not
proved thereby, since most of these exchanges can almost as easily
have occurred across the polar regions, and others still more easily by
repeated junction of Siberia with Alaska.

Let us now describe a hypothetical case based on the supposition of
connecting bridges. Not to work in a circle, we select an important
group which has not served as a basis for the reconstruction of bridges;
and it must be a group which we feel justified in assuming to be old
enough to have availed itself of ancient land-connections.

The occurrence of one species of Peripatus in the whole of Australia,
Tasmania and New Zealand (the latter being joined to Australia by way of
New Britain in Cretaceous times but not later) puts the genus back
into this epoch, no unsatisfactory as