Sec. 46. If the reader will now examine Turner's work at the point _x_ in the reference figure, and again on the stones in the foreground, comparing it finally with the fragment of the rocks which happened fortunately to come into my foreground in Plate +20+, rising towards the left, and of which I have etched the structure with some care, though at the time I had quite forgotten Saussure's notice of the peculiar M-shaped zigzags of the gneiss at the spot, I believe he will have enough evidence before him, taken all in all, to convince him of Turner's inevitable perception, and of the entire supremacy of his mountain drawing over all that had previously existed. And if he is able to refer, even to the engravings (though I desire always that what I state should be _tested_ by the drawings only) of any others of his elaborate hill-subjects, and will examine their details with careful reference to the laws explained in this chapter, he will find that the Turnerian promontories and banks are always simply _right_, and that in all respects; that their gradated curvatures, and nodding cliffs, and redundant sequence of folded glen and feathery glade, are, in all their seemingly fanciful beauty, literally the most downright plain speaking that has as yet been uttered about hills; and differ from all antecedent work, not in being ideal, but in being, so to speak, pictorial _casts_ of the ground. Such a drawing as that of the Yorkshire Richmond, looking down the river, in the England Series, is even better than a model of the ground, because it gives the aerial perspective, and is better than a photograph of the ground, because it exaggerates no shadows, while it unites the veracities both of model and photograph. Sec. 47. Nor let it be thought that it was an easy or creditable thing to treat mountain ground with this faithfulness in the days when Turner executed those drawings. In the Encyclopaedia Britannica (Edinburgh, 1797), under article "Drawing," the following are the directions given for the production of a landscape:-- "If he is to draw a landscape from nature, let him take his station on a rising ground, where he will have a large horizon, and mark his tablet into three divisions, downwards from top to the bottom; and divide in his own mind the landscape he is to take into three divisions also. Then let him turn his face directly opposite to the midst of the horizon, keeping his body fixed, and draw what is directly before his eyes upon the middle division of the tablet: then _turn his head, but not his body_,[96] to the left hand and delineate what he views there, joining it properly to what he had done before; and, lastly, do the same by what is to be seen upon his right hand, laying down everything exactly, both with respect to distance and proportion. One example is given in plate clxviii.