Another instance occurs to me, where equal liberty may be taken, in regard to the management of light. Though the general practice is, to make a large mass about the middle of the picture surrounded by shadow, the reverse may be practised, and the spirit of the rule may still be preserved. Examples of this principle reversed may be found very frequently in the works of the Venetian School. In the great composition of Paul Veronese, THE MARRIAGE AT CANA, the figures are, for the most part, in half shadow; the great light is in the sky; and, indeed, the general effect of this picture, which is so striking, is no more than what we often see in landscapes, in small pictures of fairs and country feasts; but those principles of light and shadow, being transferred to a large scale, to a space containing near a hundred figures as large as life, and conducted to all appearance with as much facility, and with an attention as steadily fixed upon _the whole together_, as if it were a small picture immediately under the eye, the work justly excites our admiration; the difficulty being increased as the extent is enlarged. The various modes of composition are infinite; sometimes it shall consist of one large group in the middle of the picture, and the smaller groups on each side; or a plain space in the middle, and the groups of figures ranked round this vacuity. Whether this principal broad light be in the middle space of ground, as in THE SCHOOL OF ATHENS; or in the sky, as in THE MARRIAGE AT CANA, in THE ANDROMEDA, and in most of the pictures of Paul Veronese; or whether the light be on the groups; whatever mode of composition is adopted, every variety and license is allowable: this only is indisputably necessary, that to prevent the eye from being distracted and confused by a multiplicity of objects of equal magnitude, those objects, whether they consist of lights, shadows, or figures, must be disposed in large masses and groups properly varied and contrasted; that to a certain quantity of action a proportioned space of plain ground is required; that light is to be supported by sufficient shadow; and we may add, that a certain quantity of cold colours is necessary to give value and lustre to the warm colours: what those proportions are cannot be so well learnt by precept as by observation on pictures, and in this knowledge bad pictures will instruct as well as good. Our inquiry why pictures have a bad effect may be as advantageous as the inquiry why they have a good effect; each will corroborate the principles that are suggested by the other.