image, which may be missed if it remains in its natural purity, (as, I suppose, few in looking at the Cephalus and Procris of Turner, note the sympathy of those faint rays that are just drawing back and dying between the trunks of the far-off forest, with the ebbing life of the nymph; unless, indeed, they happen to recollect the same sympathy marked by Shelley in the Alastor;) but the imagination is not shown in any such modifications; however, in some cases they may be valuable (in the Cephalus they would be utterly destructive,) and I note them merely in consequence of their peculiar use in religious art, presently to be examined. Sec. 19. Exaggeration. Its laws and limits. First, in scale of representation. The last mode we have here to note in which the imagination regardant may be expressed in art is exaggeration, of which, as it is the vice of all bad artists, may be constantly resorted to without any warrant of imagination, it is necessary to note strictly the admissible limits. In the first place, a colossal statue is necessarily no more an exaggeration of what it represents than a miniature is a diminution, it need not be a representation of a giant, but a representation, on a large scale, of a man; only it is to be observed, that as any plane intersecting the cone of rays between us and the object, must receive an image smaller than the object; a small image is rationally and completely expressive of a larger one; but not a large of a small one. Hence I think that all statues above the Elgin standard, or that of Michael Angelo's Night and Morning, are, in a measure, taken by the eye for representations of giants, and I think them always disagreeable. The amount of exaggeration admitted by Michael Angelo is valuable because it separates the emblematic from the human form, and gives greater freedom to the grand lines of the frame; for notice of his scientific system of increase of size I may refer the reader to Sir Charles Bell's remarks on the statues of the Medici chapel; but there is one circumstance which Sir Charles has not noticed, and in the interpretation of which, therefore, it is likely I may be myself wrong; that the extremities are singularly small in proportion to the limbs, by which means there is an expression given of strength and activity greater than in the ordinary human type, which appears to me to be an allowance for that alteration in proportion necessitated by increase of size, of which we took note in Chap. VI. of the first section, Sec. 10, note; not but that Michael Angelo always makes the extremities comparatively small, but smallest, comparatively, in his largest works; so I think, from the size of the head, it may be conjectured respecting the Theseus of the Elgins. Such adaptations are not necessary when the exaggerated image is spectral; for as the laws of matter in that case can have no operation, we may expand the form as far as we choose, only let careful distinction be