These general considerations are further strengthened when we compare the finished plate (I am speaking only of the etching in the published states, and not of the rare ‘first state’ of the etching in the late Mr. J. E. Taylor’s collection) with the preliminary study made for it, now in the National Gallery (No. 879). The size and proportions of the plate differ from those of the drawing, so the etching could not have been traced from it, while the etching nowhere follows the drawing with the accuracy one would expect if it were merely an engraver’s copy. There is no authority in this drawing for the shapes given to the crests of the distant mountains in the centre, nor for the shapes of the upper portions of the nearer mountains on the left. The lower parts of the design are also modified from the forms in the drawing in exactly the way Turner habitually recast all the drawings he etched. It is also hard to suppose that Dawe or anybody but Turner could have recast the vague shapes of the disappearing ridge of the glacier on the left in the masterly way this has been done in the etching, or that anybody but Turner could have invented, on the strength of the loose indications in the drawing, the masterly lines that give definition to the stretch of valley against which the ice of the glacier is relieved. The shapes of the [Illustration: _PLATE XLII_ VIEW OF A RIVER, FROM A TERRACE SOMETIMES CALLED “MACON” SEPIA. ABOUT 1818] [Illustration: _PLATE XLIII_ CROWHURST, SUSSEX SEPIA. ABOUT 1818] first two upright pines near the centre have also been recast by the mind and hand of the master, not copied by another hand from the indications given in the drawing. Alterations have also been introduced in the character of the stems and their branches which a professional engraver would not dare to make in an ostensible copy; the same remark applies also to the tops of the pines on the right. For these reasons, I think, those critics are mistaken who deny that the workmanship of the whole plate is Turner’s. And the mistake has arisen to a large extent, I feel inclined to add, through attaching too much importance to _a priori_ notions of technical mastery. As the late Mr. Arthur Strong very justly said, we are inclined to start with an idea that masters are always masterly and classical, and we often end by finding that they have left nothing behind them quite worthy of our preconceived ideas of what they ought to have done.