1890.] the old Masters in Painting. 51 



much has, I think, been proved by Eastlake, and we have next to 

 ask what varnish was used, or would any varnish do. On this 

 point it is apparently impossible to get any definite historical 

 evidence on account of the looseness with which terms were used. 

 Nevertheless Eastlake does come to a conclusion, and states his 

 belief that an oil varnish ground with the colour is sufficient 

 protection whether made from amber, copal, sandarac, or Venice 

 turpentine, and discards the tradition that Van Eyck made use of 

 amber varnish alone. He points out that while receipts exist in 

 the MSS. for preparing amber varnish, they also exist for the 

 preparation of other varnishes, and there is nothing to show that 

 amber was exclusively used. In fact it is improbable that amber 

 varnish alone was used, as it is difficult to prepare any which is 

 not very dark in colour. So difficult in fact is it to prepare that 

 it can hardly be said to be an article of commerce, though occa- 

 sionally varnishes claiming to be amber are sold to artists. This 

 theory of Eastlake's is easily put to the test, and he quotes an 

 experiment with gamboge, by Sir Joshua Reynolds, which seems 

 to confirm it. Unfortunately, however, Prof. Church has tested 

 this point, and finds that Eastlake's theory is not tenable. He 

 finds that gamboge fades as quickly when mixed with copal varnish 

 (the best of resins now in use) as it does with oil alone. We are 

 left then with this question still to solve, "What varnish did 

 Van Eyck use to mix with his colours ? " 



I have made some experiments which I think point to the 

 correct solution of this problem. My object has been to get some 

 rapid means of deciding whether a given vehicle was or was not 

 permeable to moisture, and I finally hit on the following device. 

 If we ignite sulphate of copper it loses all its water of combination, 

 leaving a white powder which is very hygroscopic. If this powder 

 is exposed to the air for a short time it turns green, owing to the 

 absorption of water. I ignited therefore some sulphate of copper, 

 and using it as a pigment ground it with boiled oil, and painted 

 it out on three pieces of glass. One of these I placed in a desicc- 

 ator, one in a warm dry room, and one in a room with the window 

 open. In 12 hours the sulphate of copper in the damp room had 

 turned completely green, that in the dry room slightly green, and 

 that in the desiccator remained white. I then exposed all three in 

 the damp room, and they were soon equally green. This showed 

 me that I had here a delicate test of the permeability of such 

 mediums to moisture, though this first experiment was hardly 

 fair to the boiled oil, as it had not been allowed to harden before 

 exposure to moisture. For no pigment used in practice is so 

 fugitive as to be affected by exposure to moisture during the time 

 the oil is drying, and therefore it was obviously necessary to try 

 an experiment after the vehicles had dried completely in the 



VOL. VII. PT. II. 5 



