from those of Dr. Croll, it is not
from any want of appreciation of his facts and arguments, since for many
years I have upheld and enforced his views to the best of my ability. But a
careful re-examination of the whole question has now convinced me that an
error has been made in estimating the comparative effect of geographical
and astronomical causes on changes of climate, and that, while the latter
have undoubtedly played an important part in bringing about the glacial
epoch, it is to the former that the mild climates of the Arctic regions are
almost entirely due. If I have now succeeded in approaching to a true
solution of this difficult problem, I owe it mainly to the study of Dr.
Croll's writings, since my theory is entirely based on the facts and
principles so clearly set forth in his admirable papers on "Ocean Currents
in relation to the Distribution of Heat over the Globe." The main features
of this theory as distinct from that of Dr. Croll I will now endeavour to
summarise.

Looking at the subject broadly, we see that the climatic {94} condition of
the northern hemisphere is the result of the peculiar distribution of land
and water upon the globe; and the general permanence of the position of the
continental and oceanic areas--which we have shown to be proved by so many
distinct lines of evidence--is also implied by the general stability of
climate throughout long geological periods. The land surface of our earth
appears to have always consisted of three great masses in the north
temperate zone, narrowing southward, and terminating in three comparatively
narrow extremities represented by Southern America, South Africa, and
Australia. Towards the north these masses have approached each other, and
have sometimes become united; leaving beyond them a considerable area of
open polar sea. Towards the south they have never been much further
prolonged than at present, but far beyond their extremities an extensive
mass of land has occupied the south polar area.

This arrangement is such as would cause the northern hemisphere to be
always (as it is now) warmer than the southern, and this would lead to the
preponderance of northward winds and ocean currents, and would bring about
the concentration of the latter in three great streams carrying warmth to
the north-polar regions. These streams would, as Dr. Croll has so well
shown, be greatly increased in power by the glaciation of the south polar
land; and whenever any con