abundant supply of vapour to create there by precipitation and the liberation of latent heat a degree of rarefaction sufficient to cause a general movement of the air polarward for the distance of 40° of latitude all round. That there is an immense volume of comparatively warm water going into the Arctic Ocean is abundantly shown by observation, and the rising up there of this water to the surface would afford heat and vapour enough for a vast degree of rarefaction.
459. The middle ice.—The records of arctic explorations, together with the whalemen's accounts of "middle ice" in Baffin's Bay and Davis' Straits, go to confirm this view, which is further elaborated in the next chapter (§ 475). The facts there stated, and this "middle ice," go to show that every winter a drift takes place which brings out of the Frozen Ocean a tongue of ice a thousand miles or more in length: it is the compact and cold "middle ice." In our fresh- water streams it is the middle ice that first breaks up; that which is out of the way of the current remains longest. Not so in this bay and strait; there the littoral ice first gives way, leaving an open channel on either side in spring and early summer, while the "middle ice" remains firm and impassable. The explanation is simple enough: the middle ice was formed in the severe cold of more northern latitudes, from which it has drifted down, while that on the sides was formed in the less severe climates of the bay and straits. This winter tongue of ice, which we know by actual observation is in motion from December till May, must, during that time, be detached from the main mass of ice in the Arctic Ocean, consequently there must be water between the ice that is in motion and the ice that is at rest. Not only so. In early summer the whalemen will run up to the north in the open water at the side of the "middle ice" in Davis' Strait and Baffin's Bay, even as far sometimes as Cape Alexander in 78°, to look for a crossing-place. Here, though so far north, they will find the "middle ice" gone, or so broken up that they can cross over to the west side. They trace it up thus far, because at the south, and in spite of a higher thermometer, they find the "middle ice" compact and firm, so much so as to be impassable. In this fact we recognize another circumstance favouring the theory of an open sea at the north, and giving plausibility to the conjecture that this "middle ice" drifts out from the southern edge of the open sea as fast as it is formed during the winter. According to this
