This ice must be removed by the operation of some natural cause; and observation having shown that this actually takes place, Rendu occupies himself with methods to discover how Nature has performed the task; and he comes to this very rational conclusion: That the glacier and the river are in effect the same; that between them there is a resemblance so complete that it is impossible to find in the latter a circumstance which does not exist in the former; and as the river drains the _waters_ which fall upon the hill-sides to the ocean, so the glacier drains the _ice_ which forms from the snows on the mountain-sides down to the same level: And he closes his argument with declaring the Law:-- [Sidenote: THE LAW OF CIRCULATION.] "The conserving will of the Creator has employed for the permanence of His work the great Law of _Circulation_, which, strictly examined, is found to reproduce itself in all parts of Nature." And, in illustration of this law, we see that the waters circulate from the ocean to the air by evaporation, from the air again to the earth in the form of dews and rains and snows, and from the earth back again to the ocean through the great rivers which have gathered up the little streams from every hill-side and valley. Now this law of Circulation is, in the icy regions of the Alps, of the lofty Himalayas, of the Andes, of the mountains of Norway and of Greenland, the same as in the lower and warmer regions of the earth, where the rivers drain the surface-water to the sea. A glacier is in effect but a flowing stream of frozen water; and the _river systems_ of the Temperate and Equatorial Zones become the _glacier systems_ of the Arctic and Antarctic. We have now seen that a part of the snow which falls upon the mountains is converted into ice, and this ice, strange though it seems, is movable. By what exact principle of movement has not yet been decided to the mutual satisfaction of the learned, but it is nevertheless true. Rendu truly remarks:-- [Sidenote: MOVEMENT OF THE GLACIERS.] "There is a multitude of facts which would seem to necessitate the belief that the substance of glaciers enjoys a kind of ductility, which permits it to mould itself to the locality which it occupies, to grow thin, to swell and to narrow itself like a soft paste." And this, true of the Alpine passes, is true also of the Greenland valleys. A great frozen flood is pouring down the east and west slopes of the Greenland continent; and, as in the Alps, what is gained in height by one year's freezing is lost by the downward flow of the mobile mass.