Page:Popular Science Monthly Volume 29.djvu/104

94 the work, and who began by cleaning the pictures carefully and putting them all on new canvases. In this state the new canvas showed through all the holes like the skin of a pauper through his shirt, and every one of these little islands of new canvas had to be colored up to the tint of the surrounding paint, or rather to be colored like the paint which had disappeared, the nature of it having to be guessed from what remained round about it. When there is no detail, as often happens in draperies and backgrounds, this is not extremely difficult, though it requires a well-trained eye to color; but when detail has to be invented exactly in the style of the picture, that is a different matter, which taxes the skill of the restorer to the utmost. However, there can be no question that when a picture is so injured as to present hiatuses, whether by holes in the canvas or by mere removal of the paint, it is an absolute necessity to have them filled as well as we can. Painting is not in the same position as literature in this respect. There are numerous unfinished lines in the "Æneid," and after the death of Virgil we are told that Augustus appointed a literary commission, empowered by him to remove those parts which were glaringly unfinished and defective (as Virgil himself had died before his own intended revisal of the poem); but we are also told that Augustus strictly forbade the revisers to add anything whatever of their own. We all feel that no hand but that of the author should add anything to a poem; we all prefer certain fragments of Coleridge and Shelley to any finishing that would involve additions by a reviser. In a minor degree we object to restoration in sculpture, though here we tolerate it to some extent. When a nose is broken from a bust, it is generally restored, and so is a finger on a hand; but prudent conservators of museums do not often attempt the restoration of an arm that has entirely disappeared. These distinctions, as well as our greater desire for the restoration of paintings, are all perfectly logical. A hiatus does not make a poem intolerable. The numerous small gaps in the "Æneid" have but a very slight effect in diminishing the reader's satisfaction, the reason being that they occur one at a time, and each little gap is forgotten in the interest of the next perfect opening of two pages; but in a picture all the gaps are seen at the same time, and distract our attention from the general beauty of the work. A Greek bust, however lovely, is a torment to us without its nose, and though the restored nose may not be so good as the lost original, it allows us to admire the beauties of the brow and chin in peace. If we shrink from the restoration of an arm, it is because we do not know enough about the arm that has been lost to replace it satisfactorily, but the lost arm is not spoiled; it is simply absent, and though there are loss and mutilation,