EASTLAKE’S HISTORY OF OIL-PAINTING 275

tone of their colouring in the flesh; produced by the mode in which they often prepared their works, viz. by a green under-painting. The appearance was neutralized by the red sandarac varnish, and pictures executed in the manner described must have looked better before it was removed.”-Ib., p. 252.

Farther on, this remark is thus followed out:-

“The paleness or freshness of the tempera may have been sometimes calculated for this brown glazing (for such it was in effect), and when this was the case, the picture was, strictly speaking, unfinished without its varnish. It is, therefore, quite conceivable that a painter, averse to mere mechanical operations, would, in his final process, still have an eye to the harmony of his work, and, seeing that the tint of his varnish was more or less adapted to display the hues over which it was spread, would vary that tint, so as to heighten the effect of the picture. The practice of tinging varnishes was not even new, as the example given by Cardanus proves. The next step to this would be to treat the tempera picture still more as a preparation, and to calculate still further on the varnish, by modifying and adapting its colour to a greater extent. A work so completed must have nearly approached the appearance of an oil picture. This was perhaps the moment when the new method opened itself to the mind of Hubert Van Eyck. ... The next change necessarily consisted in using opaque as well as transparent colours; the former being applied over the light, the latter over the darker, portions of the picture; while the work in tempera was now reduced to a light chiaroscuro preparation. ... It was now that the hue of the original varnish became an objection; for, as a medium, it required to be itself colourless.”-Ib., pp. 271-273.

21. Our author has perhaps somewhat embarrassed this part of the argument, by giving too much importance to the conjectural adaptation of the tints of the tempera picture to the brown varnish, and too little to the bold transition from transparent to opaque colour on the lights. Up to this time, we must remember, the entire drawing of the flesh had been in tempera; the varnish, however richly tinted, however delicately adjusted to the tints beneath, was still broadly applied over the whole surface, the design being seen through the transparent glaze. But the mixture of opaque colour at once implies that portions of the design itself were executed with the varnish for a vehicle, and therefore that the varnish had been entirely changed both in colour and consistence. If, as above stated, the improvement in the varnish had been made only after it had been mixed with opaque colour, it does not appear why the idea

[Version 0.04: March 2008]