9 o THE POPULAR SCIENCE MONTHLY. 



white-lead. It has been remarked that a certain kind of decorative 

 work used in the middle ages consisted of paint applied on tin-foil and 

 protected by glass. Here was a double protection against damp, the 

 glass before and the tin-foil behind, the glass answering to the varnish 

 on a picture, but with more complete efficacy. 



Glass is now largely used in the National Gallery for the protection 

 of oil -pictures, but, unfortunately, the common objection that it does 

 not allow the spectator to see the picture easily is but too well founded. 

 What we really see is too often the reflection of a group of visitors to 

 the gallery, almost as in a looking-glass. This happens especially 

 when the picture is a dark one, and many of the finest old pictures are 

 dark. We are sometimes told that it is an affair of focusing the eye, 

 and that if we look as we ought to do at the picture itself, and not at 

 the reflections, we shall not see the reflections. What really happens 

 is this : If we look at the reflections of the visitors we see them won- 

 derfully well, down to the most minute inventions of the feminine cos- 

 tume, and if we look at the picture we see it in a confused way inter- 

 mingled with the reflections. This being so, it follows that private 

 owners are not much encouraged to put their pictures under glass. It- 

 may be objected that water-colors are habitually protected in this way, 

 and that no one complains. True, but in the first place, with regard 

 to water-colors we have no choice, as any fly could spoil an unprotected 

 water-color in a minute ; in the second place, a drawing in water- 

 color is usually of small dimensions, so that it is more easily seen ; 

 and, lastly, water-colors are generally paler than oil-pictures, so that 

 they do not make such perfect mirrors. A dark old oil-picture w T ith a 

 sheet of plate-glass before it is, in certain lights, almost as good a 

 mirror as if the glass were lined w r ith quicksilver. We can hardly, 

 then, include glass among the means to be recommended for protect- 

 ing oil-pictures from damp, and must trust rather to the dryness of the 

 atmosphere in which the pictures are kept ; and yet it is necessary to 

 avoid excessive heating, which in certain cases produces or favors 

 cracking and destroys by detaching paint from the priming of the 

 canvas. 



Canvas may not seem a very durable material, and yet, on the 

 whole, it is preferable to wooden panels, for it may truly be said of 

 wood, as it was said of the arch in architecture, that it is never at rest. 

 It is always either swelling or contracting, and if a composite panel is 

 not quite scientifically constructed, it is sure to tear itself and show 

 fissures. Panels are therefore usually employed for small works only, 

 and for these copper would be better still, though it has been used 

 rarely. If a panel is well painted on the back, it will absorb damp 

 less readily, and this precaution is very easily taken. 



The art of removing a painting from an old to a new canvas is now 

 so well understood that the operation, which many years ago seemed 

 formidable, is now performed every day without attracting attention. 



