40. Great additional force is given to this argument, in many cases, by the nature of the materials of which the stratified rocks are composed. The beds of breccia and pudding-stone, for instance, are often in planes almost vertical, and, at the same time, contain gravel-stones, and other fragments of rock, of such a size and weight, that they could not remain in their present position an instant, if the cement which unites them were to become soft; and therefore they certainly had not that position at the time when this cement was actually soft. This remark has been made by mineralogists who were not led to it by any system. The judicious and indefatigable observer of the Alps, describing the pudding-stone of Valorsine, near the sources of the Arve, tells us, that he was astonished to find it in beds almost vertical, a situation in which it could not possibly have been formed. "That particles," he adds, "of extreme tenuity, suspended in a fluid, might become agglutinated, and form vertical beds, is a thing that may be conceived; but that pieces of stone, of several pounds weight, should have rested on the side of a perpendicular wall, till they were enveloped in a stony cement, and united into one mass, is a supposition impossible and absurd. It should be considered, therefore, as a thing demonstrated, that this pudding-stone was formed in a horizontal position, or one nearly such, and elevated after its induration. We know not," he continues, "the force by which this elevation has been effected; but it is an important step among the prodigious number of vertical beds that are to be met with in the Alps, to have found some that must certainly have been formed in a horizontal situation."[14] [Footnote 14: Voyages aux Alpes, tom. ii § 690.] 41. Nothing can be more sound and conclusive than this reasoning; and had the ingenious author pursued it more systematically, it must have led him to a theory of mountains very little different from that which we are now endeavouring to explain. If some of the vertical strata are proved to have been formed horizontally, there can be no reason for not extending the same conclusion to them all, even if we had not the support of the argument from the parallelism of the layers, which has been already stated.