by means of ordinary Artificial Light. 175 



mary colours occur, yellow and red, yellow and blue, or red 

 and blue, they form the secondary colours, orange, green, or 

 purple ; and when they occur in the ratio in which they are 

 produced by white light, they form perfect or normal second- 

 aries. The colours of those objects which are called coloured, 

 is caused by their having the quality, from their peculiar 

 structure, or otherwise, of reflecting or transmitting different 

 parts of the rays forming the compound light to which they 

 are exposed, the remainder being absorbed and extinguished. 

 When light is composed of rays which produce colours in a 

 different ratio from that in which they neutralize each other, 

 the colour of the excess is produced on a white surface, and 

 the appearance of coloured objects is affected according to the 

 excess or the deficiency of the rays producing particular 

 colours. For example, an intensely blue object exposed to 

 a perfect orange light appears black, as there are no blue 

 rays which it may reflect or transmit, and it absorbs and ex- 

 tinguishes the red and the yellow. If some l*ays producing 

 blue be combined with the orange light, this effect is modi- 

 fied according to their quantity. 



In ordinary artificial light, the rays producing red and 

 yellow appear to occur nearly in the same ratio in which they 

 occur in white light. And as there is an excess of them over 

 the rays producing blue, all that is necessary in order to pro- 

 duce white light, is to interpose a transparent medium of 

 such a depth of blue as to absorb that excess, allowing such 

 a portion of them to pass as is necessary to neutralize the 

 rays producing blue, all of which have been allowed to pass. 



It appears that that excess may be best absorbed by inter- 

 posing coloured glass, of a proper depth of blue, according to 

 the ratio of orange in the flame. It ought to be "coloured" 

 glass, that is, coloured in the melting-pot. Glass " stained" 

 blue, that is, partially coloured by applying the colouring 

 matter to the siirface, and fixing it by a heat sufficient to 

 soften the glass, but not to melt it, is not transparent ; in 

 which respect blue differs from other stained colours in 

 general. It is very difficult to get ** coloured" glass made 

 of any precise tint. Perhaps the tint required could be con- 



