The ground over which we rode in a SW. direction, parallel to the base of the escarpment, was very irregular in form, rising in places into mounds sloping inwards towards the cliffs as well as outwards towards the plain; and, although in great part covered with vegetation, it appeared pretty certain that the whole was composed of irregular masses of sandstone intermixed to some extent with fragments of the rocks forming the barrier beside us. To those who did not admit the probability of the boulders before seen being deposited by glacial action, the phenomena here presented offered strong confirmation. A glacier descending from a main valley necessarily flows down the slope towards the plain, and could not turn aside at right angles to its previous course, and to the line of maximum inclination, unless there had been a barrier of solid rock stopping the way, of which there was here not the slightest indication. Whether or not materials that are borne down a steep incline by sub-aërial denudation form a talus with a diminishing slope resting against the face of the escarpment, or form mounds at a greater or less distance from the base, is a question depending upon the momentum with which they descend; and this again depends on three elements—the weight of the blocks, the steepness, and the length of the slope. If the greater portion of the materials consist of large blocks launched down a steep and long incline, these will travel to a considerable distance from the base of the cliff, and gradually form a barrier that will stop the course of other similar masses, until these accumulate into considerable mounds, as may be seen in many instances of berg-falls in the Alps. Whatever be the origin of these accumulations in this part of the Great Atlas, it would appear that the conditions that gave rise to them have now ceased, as we saw no instance of any large block that appeared to have been recently borne to its present position. As it was important to reach our night quarters by daylight, we collected few plants during the afternoon ride; a fine _Asperula_, with numerous flowers varying from white to pink, seemingly not different from the Spanish _A. hirsuta_, was a great ornament here, and in several other places on the skirts of the Atlas.