THE BOOK OF THE LANTERN. 161 



a good even sky-blue tint, he has only to try the same pro- 

 cedure with his yellows and reds to produce all kinds of 

 brilliant, ethereal, striped-petticoat effects. But, on trying 

 these colours, he will soon find out his mistake, and will 

 also find that he must add varnish to them before he can 

 work with them at all. Moreover, they seem to be espe- 

 cially prone to attract any little unconsidered trifles in the 

 way of dust which may be seeking rest. 



I was so impressed with these difficulties with regard to 

 sunset skies, when first I began glass painting, that [ sought 

 for another means altogether for gaining what I wished. 

 I was attracted by the brilliant hues of the aniline, or coal- 

 tar colours, and at once endeavoured to enlist them into my 

 service. As others may be tempted to work in the same 

 groove, I may at once state why, after patient trial, I dis- 

 carded them. Most of these colours can be readily dissolved 

 in alcohol, and, therefore, it is not difficult to make 

 coloured varnishes with them. But when I tried to paint 

 my gelatine picture with the splendid tints, I found it next 

 to impossible to confine them within the boundaries of any 

 outlines whatever. They would flow over the edges on 

 their own account, do what I might. The fact is that these 

 aniline colours have a kind of greedy affinity for gelatine, 

 and there seem to be no means of controlling their advance 

 when once they come into contact with it. By flooding 

 an entire picture with a yellow or red varnish, I was able 

 to gain (sometimes) some wonderful effects. But the action 

 of the dye upon the gelatine was of too uncertain a nature 

 to tempt me to adopt that method of working as a perma- 

 nent resource. Lastly, aniline colours are fugitive. 



M 



