The method of our study must be determined by the general character of our subject-matter. Our main business is with fragmentary records, hurried memoranda, half-formed thoughts, and tentative designs. We must not and cannot treat these dependent and embryonic fragments as independent entities; we cannot pick and choose amongst them, or love or dislike them entirely for their own sakes, as we can with complete works of art which contain within themselves the grounds of their own justification or insufficiency. To grasp the significance of our sketches and studies we must study the goal towards which they are striving. We must not be content to admire even the most beautiful of these sketches entirely for its own sake, but must study them for the sake of their connection with the works which they were instrumental in producing. These considerations have also weighed with me in the selection of the numerous illustrations with which the publishers have generously enriched this volume. On the whole I have chosen the illustrations rather for the light they throw on Turner’s conception of art and methods of work than for their own individual attractiveness; but the glamour of execution is so invariably present in all that came from Turner’s hand, that few of these drawings will be found which do not possess a very powerful aesthetic appeal of their own. In dealing with Turner’s work from the point of view I have indicated, we are forced to touch upon problems which the prudent art critic is apt to avoid. In studying the relation between the preliminary sketches and studies and the finished works into which they were developed, we find ourselves plunged into the midst of some of the most baffling difficulties of psychology and aesthetic. In attempting even to describe the relation between the more rudimentary and the more fully articulated processes of artistic expression, we are forced, whether we like it or not, to face the problems of the relation between form and content, between treatment and subject, between portrayal and portrayed; and we cannot go far without finding ourselves obliged to reconsider the common-sense ideas of Truth, Nature, and Art. We cannot avoid such problems if we would. If I face them, therefore, instead of emulating the discretion of my elders, it is, I am sure, from no ingrained love of abstractions, but rather from an overpowering interest in all the concrete forms of pictorial art.