Most of the earlier observers considered the structure to be due to the stratification of the mountain-snows--a view which has received later development at the hands of Mr. John Ball; and the practical difficulty of distinguishing the undoubted effects of _stratification_ from the phenomena presented by _structure_, entitles this view to the fullest consideration. The blue veins of glaciers are, however, not always, nor even generally, such as we should expect to result from stratification. The latter would furnish us with distinct planes extending parallel to each other for considerable distances through the glacier; but this, though sometimes the case, is by no means the general character of the structure. We observe blue streaks, from a few inches to several feet in length, upon the walls of the same crevasse, and varying from the fraction of an inch to several inches in thickness. In some cases the streaks are definitely bounded, giving rise to an appearance resembling the section of a lens, and hence called the "lenticular structure" by Mr. Huxley and myself; but more usually they fade away in pale washy streaks through the general mass of the whitish ice. In Fig. 39 I have given a representation of the structure as it is very commonly exhibited on the walls of crevasses. Its aspect is not that which we should expect from the consolidation of successive beds of mountain snow. [Illustration: Fig. 39. Veined Structure of the walls of crevasses.] Further, at the bases of ice-cascades the structural laminae are usually _vertical_: below the cascade of the Talefre, of the Noire, of the Strahleck branch of the Lower Grindelwald Glacier, of the Rhone, and other ice-falls, this is the case; and it seems extremely difficult to conceive that a mass horizontally stratified at the summit of the fall, should, in its descent, contrive to turn its strata perfectly on end. Again, we often find a very feebly-developed structure at the central portions of a glacier, while the lateral portions are very decidedly laminated. This is the case where the inclination of the glacier is nearly uniform throughout; and where no medial moraines occur to complicate the phenomenon. But if the veins mark the bedding, there seems to be no sufficient reason for their appearance at the lateral portions of the glacier, and their absence from the centre. [Sidenote: ILLUSTRATIVE EXPERIMENTS.]