§ 20. Of the aqueous curvatures of this crest, we shall have more to say presently; meantime let us especially observe how the providential laws of beauty, acting with reversed data, arrive at similar results in the aiguilles and crests. In the aiguilles, which are of such hard rock that the fall of snow and trickling of streams do not affect them, the inner structure is so disposed as to bring out the curvatures by the mere fracture. In the crests and lower hills, which are of softer rock, and largely influenced by external violence, the inner structure is straight, and the necessary curvatures are produced by perspective, by external modulation, and by the balancing of adverse influences of cleavage. But, as the accuracy of an artist's eye is usually shown by his perceiving the inner anatomy which regulates growth and form, and as in the aiguilles, while we watch them, we are continually discovering new curves, so in the crests, while we watch them, we are continually discovering new straightnesses; and nothing more distinguishes good mountain-drawing, or mountain-seeing, from careless and inefficient mountain-drawing, than the observance of the marvellous parallelisms which exist among the beds of the crests.
§ 21. It indeed happens, not unfrequently, that in hills composed of somewhat soft rock, the aqueous contours will so prevail over the straight cleavage as to leave nothing manifest at the first glance but sweeping lines like those of waves. Fig. 43, p. 196, is the crest of a mountain on the north of the valley of Chamouni, known, from the rapid decay and fall of its crags, as the Aiguille _Pourri_; and at first there indeed seems little distinction between its contours and those of the summit of a sea wave. Yet I think also, if it _were_ a wave, we should immediately suppose the tide was running towards the right hand; and if we examined the reason for this supposition, we should perceive that along the ridge the steepest falls of crag were always on the right-hand side; indicating a tendency in them to break rather in the direction of the line _a b_ than any other. If we go half-way down the Montanvert, and examine the left side of the crest somewhat more closely, we shall find this tendency still more definitely visible, as in Fig. 62.