ten of, some other phase of the reader's understanding must be appealed to than that of visual imagination.
If the elementary content of subjective reality be unpicturable, much less can its abstract aspects be rendered in spatial forms and relations. Thus scientific and logical analysis, explanation and philosophical reflection, and the whole literature of appreciation are debarred by the nature of their subject-matter from direct appeal through illustration. Works on logic and metaphysics—proverbially hard thinking—are rendered more unrelievedly so by the pages of close-packed type which follow one another in unbroken succession from beginning to end of the book. Yet it is just in such fields as this that illustration is most needed. Discussion of concrete things is readily apprehended by the ordinary mind, for it finds little difficulty in representing its substance in imagination; but to follow a process of abstract thought for any continuous period necessitates a more sustained act of attention and a mind disciplined by reflection. To make them generally comprehensible such abstract works imperatively demand illustration; and since a pictorial commentary is out of the question exposition must proceed by an exemplification of the concept or law to which the term illustration in its wider sense is of course commonly extended.
Pure science labors under the same difficulty as logic and metaphysics. The work of reflective thought, in all its forms alike, is to extract from a multitude of vivid but confusing facts some common type or law to which they conform, and the elaboration of norms and principles in this field is found by the average mind hardly less repellant than in philosophy itself. The process is laborious; it is insecure; it is unsubstantial. Works on pure science are not read as are those on experimental science and natural history, because their abstractness makes them more difficult; and the writer, especially if he appeal to a technical class of readers, feels that his concern is primarily in making clear the theory he is developing, together with its evidence—not in illustrating each aspect of it by concrete examples.
The limits of illustration and the function it logically performs thus involve a certain contradiction. Its availability and its desirability stand in inverse relations. Where it is least needed—that is, when the concrete things of the sensible world are dealt with, it is entirely feasible to introduce it; but where it is most needed—in facilitating apprehension of abstract conceptions, its use is practically impossible. There is therefore to be expected a surfeit of pictures in the one case and a dearth, if not a complete absence, of them in the other.
Within the field where pictorial illustration offers itself as an adjunct to literary treatment the further question arises of the relation which text and picture logically bear to one another, and the specific service which illustration renders to the reader. It may be assumed that the function is thus specific, and that text and picture theoretically form
