In order fully to discuss this question, I must begin by referring to some of the newest theoretical opinions entertained on the glacial question. When treating of this subject in the "Principles of Geology," chapter 15, and in the "Manual (or Elements) of Geology," chapter 11, I have stated that the whole mass of the ice in a glacier is in constant motion, and that the blocks of stone detached from boundary precipices, and the mud and sand swept down by avalanches of snow, or by rain from the surrounding heights, are lodged upon the surface and slowly borne along in lengthened mounds, called in Switzerland moraines. These accumulations of rocky fragments and detrital matter are left at the termination of the glacier, where it melts in a confused heap called the "terminal moraine," which is unstratified, because all the blocks, large and small, as well as the sand and the finest mud, are carried to equal distances and quietly deposited in a confused mass without being subjected to the sorting power of running water, which would convey the finer materials farther than the coarser ones, and would produce, as the strength of the current varied from time to time in the same place, a stratified arrangement. In those regions where glaciers reach the sea, and where large masses of ice break off and float away, moraines, such as I have just alluded to, may be transported to indefinite distances, and may be deposited on the bottom of the sea wherever the ice happens to melt. If the liquefaction take place when the berg has run aground and is stationary, and if there be no current, the heap of angular and rounded stones, mixed with sand and mud, may fall to the bottom in an unstratified form called "till" in Scotland, and which has been shown in the last chapter to abound in the Norfolk cliffs; but should the action of a current intervene at certain points or at certain seasons, then the materials will be sorted as they fall, and arranged in layers according to their relative weight and size. Hence there will be passages from till to stratified clay, gravel, and sand. Some of the blocks of stone with which the surfaces of glaciers are loaded, falling occasionally through fissures in the ice, get fixed and frozen into the bottom of the moving mass, and are pushed along under it. In this position, being subjected to great pressure, they scoop out long rectilinear furrows or grooves parallel to each other on the subjacent solid rock. Smaller scratches and striae are made on the polished surface by crystals or projecting edges of the hardest minerals, just as a diamond cuts glass.