"Let us consider," he says, "those more considerable changes to which glaciers are subject, or rather, the immense extent which they had in the prehistoric period. This former immense extension, greater than any that tradition has preserved, is proved, in the case of nearly every valley in the Alps, by facts which are both many and well established. The study of these facts is even easy if the student is looking out for them, and if he will seize the least indication of their presence; and, if it were a long time before they were observed and connected with glacial action, it is because the evidences are often isolated and occur at places more or less removed from the glacier which originated them. If it be true that it is the prerogative of the scientific observer to group in the field of his mental vision those facts which appear to be without connection to the vulgar herd, it is, above all, in such a case as this that he is called upon to do so. I have often compared these feeble effects, produced by the glacial action of former ages, with the appearance of the markings upon a lithographic stone, prepared for the purpose of preservation, and upon which one cannot see the lines of the draughtsman's work unless it is known beforehand where and how to search for them. "The fact of the former existence of glaciers which have now disappeared is proved by the survival of the various phenomena which always accompany them, and which continue to exist even after the ice has melted. These phenomena are as follows: "1. Moraines.--The disposition and composition of moraines enable them to be always recognized, even when they are no longer adjacent to a glacier nor immediately surround its lower extremities. I may remark that lateral and terminal moraines alone enable us to recognize with certainty the limits of glacial extension, because they can be easily distinguished from the dikes and irregularly distributed stones carried down by the Alpine torrents, The lateral moraines deposited upon the sides of valleys are rarely affected by the larger torrents, but they are, however, often cut by the small streams which fall down the side of a mountain, and which, by interfering with their continuity, make them so much more difficult to recognize.