It has been shown that to attempt to make an illusion of nature is neither within the power of monochrome nor has been the chief aim of draughtsmen, but that the art of drawing consists in giving a plain statement of one's ideas, be they slight or studied, of the solid forms of nature. But the question may still be asked: Why is it that a rigorously accurate and finished drawing by a student or artist with _no_ such ideas or conception is not good drawing, containing as it must do all that can be seen in the original, missing only its complete illusion? Why, in a word, is not a photograph a work of art? The common explanation of the above important question is that the artist "selects and eliminates from the forms of nature." But surely this is the principle of the caricaturist and virtuoso? A beautiful drawing, however slight, is but the precipitate of the whole in the artist's mind. And a highly finished drawing by a master does not show even any apparent selection or elimination. The adoption of the principle of selection to differentiate art from mechanical reproduction is fundamentally vicious, and could be shown to be wholly inapplicable to the so-called formative arts. Nor could the theory of "selection" be used as a principle of teaching, for if to the first question the pupil would make, "What am I to select?" it were answered, "Only the important things," then the next question, "What are the important things?" could be answered only by saying, "That alone the real artist knows, but cannot teach." Certainly there are important things that can be taught the student in the initial stage of "laying-in" a figure, but _when_ to begin selecting or eliminating no teacher could tell him, simply because he must be aware that a true draughtsman can afford to eliminate nothing when the truth of the whole is at stake. The artist's conception and its expression may be slight or elaborate, but in neither case can selection or elimination take place, for a true conception must be founded upon the character of the whole, which is determined by the entire complex of all the parts.