ht behind a colored and transparent medium, and aimed at
giving brilliancy to their tints by allowing the white ground to shine
through them. If those painters and their followers erred, it was in
sometimes too literally carrying out this principle. _Their lights are
always transparent_ (mere white excepted) and their shadows sometimes
want depth. This is in accordance with the effect of glass-staining, in
which transparency may cease with darkness, but never with light. The
superior method of Rubens consisted in preserving transparency chiefly
in his darks, and in contrasting their lucid depth with solid lights
(p. 408).... Among the technical improvements on the older process may
be especially mentioned the preservation of transparency in the darker
masses, the lights being loaded as required. The system of exhibiting
the bright ground through the shadows still involved an adherence to the
original method of defining the composition at first; and the solid
painting of the lights opened the door to that freedom of execution
which the works of the early masters wanted." (p. 490.)

* * *

131. We think we cannot have erred in concluding from these scattered
passages that Mr. Eastlake supposes the brilliancy of the high lights of
the earlier schools to be attributable to the under-power of the white
ground. This we admit, so far as that ground gave value to the
transparent flesh-colored or brown preparation above it; but we doubt
the transparency of the highest lights, and the power of any white
ground to add brilliancy to opaque colors. We have ourselves never seen
an instance of a _painted brilliant_ light that was not loaded to the
exclusion of the ground. Secondary lights indeed are often perfectly
transparent, a warm hatching over the under-white; the highest light
itself may be so--but then it is the white ground itself subdued by
transparent _darker_ color, not supporting a light color. In the Van
Eyck in the National Gallery all the brilliant lights are loaded; mere
white, Mr. Eastlake himself admits, was always so; and we believe that
the flesh-color and carnations are painted with color as _opaque_ as the
white head-dress, but fail of brilliancy from not being _loaded enough_;
the white ground beneath being utterly unable to add to the power of
such tints, while its effect on more subdued tones depended in great
measure on its receiving a transparent coat of warm color first. This
_may_ have