by means of ordinary Artificial Liyht, 179 



" coloured" or painted, may be adjusted by comparing their 

 efifects with its effect, by artificial light. 



If the paper for which a glass has been adjusted have a 

 slight tinge of colour, paper of the same tinge ought always 

 to be used along with that glass, when used with a similar 

 light. 



The excess of orange is greater as the flame is less bright ; 

 for example, greater in a swallow-tail flame of gas from par- 

 l"ot or cannel coal, than in the flame of a well- constructed 

 argand lamp with such gas or fine oil ; greater in the flame 

 of a wax or composition candle ; and still greater in a duller 

 flame, such as that from coarse tallow or common oil ; as may 

 be easily seen by comparing the colours of contiguous shadows 

 from two lights on a white surface ; so that a glass adapted 

 for a bright flame is not accurately suitable for one less bright. 

 At the same time, the excess of orange which would thus be 

 transmitted if the flames do not diff^er materially in brightness, 

 is comparatively so small, that, for many purposes, it might 

 be disregarded ; but it would be undesirable to use a glass 

 adapted for a less bright flame, as it would absorb more than 

 necessary of the orange produced by the flame used, and 

 would produce an unpleasant blue tinge. 



A very large portion of the artificial light is absorbed by 

 the blue of the medium. By repeated observations, which 

 did not diff'er materially, made by comparing (by means of 

 shadows in the usual manner) the light of a bright swallow- 

 tail flame of the gas from parrot or cannel coal, made by the 

 Edinburgh Gas-Light Company, transmitted through glass 

 painted blue, so as to be tolerably transparent, with the same 

 light transmitted through glass of the same quality, brought 

 to apparently the same degree of transparency without colour, 

 I found that the former, the light transmitted through the 

 blue medium, was equal to about one-fifth part only of the 

 latter. And as the former, being white light, consists of 

 equal parts of orange and blue, and as the portion absorbed 

 consists entirely of the excess of orange, it appears (if that 

 portion be about four-fifths), that the ratio of orange to blue 

 in such a flame is about nine times that in white light. The 

 portion now stated as being absorbed, is given merely as an 



