176 G. Tait, Esq., on producing White or Neutral Lights 



veniently produced by having the glass made at first a little 

 too dark, by making it a little thicker than necessary, and 

 afterwards reducing it sufficiently by grinding, and, if neces- 

 sary, by polishing. 



If " coloured" glass of the proper depth of blue cannot be 

 obtained, coloured glass of a lighter blue, or colourless glass, 

 may have the necessary colour applied to it with transparent 

 paint. I have found the blue commonly called French blue 

 or factitious ultramarine, to ans\ver the purpose very w^ell. 

 It is very transparent, but sometimes has a slight tinge of 

 red. Genuine ultramarine, which is prepared from the pre- 

 cious stone lapis lazuli, and which is very expensive, at least 

 the best that I could procure in Edinburgh as such, however 

 much superior it is for ordinary painting, is much less trans- 

 parent, and has a dusky slaty appearance by transmitted 

 light. Cobalt is still more objectionable in both respects. 

 Prussian blue is transparent, but produces a considerable 

 tinge of green, which it is necessary to neutralize with red, 

 the consequence of which is a considerable shade of grey. 



The French blue has this disadvantage, that till the oil and 

 the varnish used in applying it are completely dry throughout 

 (which in the case of the oil may be very tedious), there is a 

 very wide refraction and dispersion of a large portion of the 

 rays producing red ; which is easily observed by holding the 

 glass close to a flame. The consequence of this is, that when 

 a limited beam of light is thrown on a perpendicular object, 

 there is an offensive green tinge in the direct line of the light, 

 caused by the blue and the yellow not being neutralized by 

 that portion of the red which has been refracted. This con- 

 sequence does not take place where the light is used in a 

 diffused manner, for example, a flame surrounded by blue 

 glass, because the whole of the component parts of the trans- 

 mitted light reach every object, either directly from the flame 

 through the intervening part of the glass, or by refraction 

 from other parts of it. The refraction of red takes place also 

 with Prussian blue having the green tinge neutralized by 

 cochineal lake. It does not take place with genuine ultra- 

 marine or with cobalt; but their want of transparency and 



