OILS, ETC., USED IN THE PAINTING OF PICTURES. 465 



The next question to be considered is the occasional and capri- 

 cious cracking of j^ictures. The explanation of this has, I confess, 

 so far completely baffled me, but there are certain matters of interest 

 in connection Avith it worth mentioning. 



The great difficulty I have found in investigating this matter is 

 due to the fact that I have been unable, under any condition, to 

 produce cracking. The first experiment I tried was more than 

 twelve years ago, when I painted out two, three, and four coats, 

 following each other quickly as soon as the last coat was sufficiently 

 dry on the surface, of flake white (1) ground stiff with oil, (2) 

 diluted Avitli more oil, (3) diluted with copal varnish and (4) diluted 

 with petroleum. 



The same set of experiments were repeated with lead sulphate 

 and zinc oxide paint, and with pure zinc oxide, thus making in all 

 48 different j^anels representing the different conditions. They are 

 all perfect and show no signs of disintegration to-day. I have also 

 tried i^ainting on ordinary jDrimecl canvas with yellow ocher, and 

 then as soon as it was sufficiently dry, laying over it strips of umber, 

 a quick-drying pigment — result, no cracks. 



I have also tried the following combinations: Undermost oil 

 paint -{- mastic or pure mastic, second coat oil paint, third coat oil 

 paint -)- mastic, umber -|- mastic, then yellow ocher, then mastic 

 varnish, oil paint -\- olive oil, umber in oil on top, mastic on top, 

 shellac varnish on top, and other similar combinations. All were 

 hurried, no proper time for drying was allowed, and after twelve 

 months they were free from cracks. 



In no instance did any crack, with the exception of the shellac 

 varnish, on the paint mixed with olive oil. I noticed, however, one 

 curious result, the yellow ocher on the top of mastic varnish cracked 

 while still wet, owing to surface tension effects, but changed no more 

 after it was dry. 



These experiments, then, were all negative in their results, and cer- 

 tainly eliminate many of the causes to which cracking is supposed 

 to be due. 



I had the good fortune to be presented with two pictures which 

 had cracked badly within a few months of painting. In the first 

 picture mastic had been used as a medium, and the cracking was 

 confined to the parts where thin liquid painting had been done and 

 mastic probably freely used, as the surface here was hard and brittle. 

 The canvas was of very poor quality, hardly closer in mesh than 

 coarse muslin. A section through a crack when placed under the 

 microscope showed the crack to be merely through the upper paint- 

 ing, and to be a broad crack with straight edges perpendicular to 



SM 1906 30 



