VII The application of this idea does not end with the simple proportions of measurement between the limbs and parts of the figure; it is also concerned with what is called the modelling, and the treatment of surfaces such as the draperies, the hair, the fleshy portions and those beneath which the bony structure comes to prominence; in painting it may be applied to the chiaroscuro and colour. Reynolds' remarks on the Venetians in his Eighth Discourse well illustrate this fact. He says: It ought, in my opinion, to be indispensably observed that the masses of light in a picture be always of a warm mellow colour, yellow, red, or a yellowish-white; and that the blue, the grey, or the green colours be kept _almost_ entirely out of these masses, and be used only to support and set off these warm colours; and, for this purpose, a small _proportion_ of cold colours will be sufficient. If this conduct be reversed, let the light be cold, and the surrounding colours warm, as we often see in the works of the Roman and Florentine painters; and it will be out of the power of art, even in the hands of Rubens or Titian, to make a picture splendid or harmonious.[86] Here we see a great colourist attempting to establish a canon for colour. Had he lived at an earlier period, before expression had become generally a subject of criticism, he would have described his discovery in less guarded and elastic language, such as is now applied to scientific laws. And then he might have been as excusably misunderstood as Leonardo and Dürer have been; as it is, the misunderstanding dealt out to him is quite without excuse.