PREFACE
that I should so have failed perhaps of a pretext for my present insistence.
Since I do insist, at all events, I find this ghostly interest perhaps even more reasserted for me by the questions begotten within the very covers of the book, those that wander and idle there as in some sweet old overtangled walled garden, a safe paradise of self-criticism. Here it is that if there be air for it to breathe at all, the critical question swarms, and here it is, in particular, that one of the happy hours of the painter's long day may strike. I speak of the painter in general and of his relation to the old picture, the work of his hand, that has been lost to sight and that, when found again, is put back on the easel for measure of what time and the weather may, in the interval, have done to it. Has it too fatally faded, has it blackened or "sunk," or otherwise abdicated, or has it only, blest thought, strengthened, for its allotted duration, and taken up, in its degree, poor dear brave thing, some shade of the all appreciable, yet all indescribable grace that we know as pictorial "tone"? The anxious artist has to wipe it over, in the first place, to see; he has to "clean it up," say, or to varnish it anew, or at the least to place it in a light, for any right judgement of its aspect or its worth. But the very uncertainties themselves yield a thrill, and if subject and treatment, working together, have had their felicity, the artist, the prime creator, may find a strange charm in this stage of the connexion. It helps him to live back into a forgotten state, into convictions, credulities too early spent perhaps, it breathes upon the dead reasons of things, buried as they are in the texture of the work, and makes them revive, so that the actual appearances and the old motives fall together once more, and a lesson and a moral and a consecrating final light are somehow disengaged.
All this, I mean of course, if the case will wonderfully take any such pressure, if the work does n't break down under even such mild overhauling. The author knows well enough how easily that may happen—which he in fact frequently enough sees it do. The old reasons then are too
xii
