80 Glacirrs (nuf (ihwial Hadiants ClaypoU. 
contltieiifc over more or less of tlie Liinl lying within the Arctic 
circle. Indeed the climate o£ this area will warrant us in be- 
lieving that the ice was continuous over large districts around 
the pole. But unless the configuration of the land and water 
was very different from that which now prevails we can scarcely 
in accordance with physical laws admit a solid mass of ice even 
in this extreme latitude. For glaciers do n(}t form at sea and 
ice-bergs cannot be born where glaciers are not. ■ Floe-ice and 
sheet-ice of even considerable thickness may form and float but 
no known conditions can produce a massive continental ice-sheet 
over a sea-area. And so far as we can judge from our present 
knowledge the region of North America toward the pole con- 
sists of an archipelago whose islands are not of great hight, 
Avhile to the extreme north there is apparently a polar sea ex- 
tending perhaps round the globe. Such an area would afford a 
not very good gathering-ground for snow and ice and conse- 
(juently not a very good birth-place for an extensive glacier. 
Should it eventually prove to be the case that the polar area 
is occupied by a deep and open sea nothing less than the severest 
evidence —proof beyond all controversy — could bring us to the 
belief in a polar glacier of enormous thickness. No case can be 
♦quoted from the existing geography of the earth where an open 
ocean is or has been a glacial radiant. For the production of a 
glacier in such a position ice must form on the surface and 
gradually thicken downward until the sea is frozen solid to the 
very bottom. Then the accumulation of snow could begin and 
the formation of an ice-sheet might become possible. 
But the greater warmth of the sea-bottom would constantly 
dissolve off the roots of the ice-floe and the scanty snow-fall of 
those high latitudes would scarcely be able to keep pace with 
the continual melting below and the powerful action of a con- 
stant sunshine of six months' duration. 
While therefore not denying the possibility that an ice- radi- 
ant existed at the very pole we submit that there is no evidence 
sufficient to su|)port it but a very high probability against it. 
That huge ice-floes and heavy sheet-ice were formed there dur- 
ing the ice-age, as now, we fully admit. That these ice-floes 
may have been both constant and continuous so as to be unable 
to flow away through the narrow intricate channels of the polar 
archipelago we also freely allow. Thatcui-rents may have borne 
