346 Analysis of Scientific Books and Memoirs. 



light. Be this as it may, the remark that aqueous vapour present in a 

 flame increases the quantity of yellow light, which, we believe, no one 

 before Dr B. had made, is curious and important, and furnished him with 

 what he was in search of — a monochromatic flame. Diluted alcohol is the 

 pabulum he proposes to employ, and his paper contains a description and 

 drawing of a convenient lamp, for maintaining and managing its combus- 

 tion with perfect facility. 



Mr Herschel appears to have entered on a somewhat similar inquiry, 

 having felt the want of some standard homogeneous light in the course of 

 optical researches of another nature. He remarks, that the flame of an or- 

 dinary spirit lamp consists of two portions, a yellow cone enclosed in a blue 

 envelope, but projecting above it, so that the upper part is purely yellow, 

 the lower a mixture of yellow and other faint rays. If, however, it be 

 viewed through a combination of a pale green with a pale orange glass, it 

 appears purely yellow, and if enclosed in a lanthorn of such glass, becomes 

 a monochromatic lamp. 



In the course of their investigations, both authors were led to examine 

 the action of differently coloured media on the spectrum. The results 

 at which they arrive agree in many points. We shall state the principal 

 of them. 



1. All coloured media absorb some rays of the spectrum in preference 

 to others, and the quantity absorbed depends on the thickness of the 

 medium. 



2. The quantity of any coloured ray,* transmitted by a homogeneous 

 medium, decreases in geometrical progression as the thickness increases in 

 arithmetical progression. 



3. Every medium has its own peculiar scale of action on the series of 

 differently refrangible rays, or its own peculiar ratio of the geometrical pro- 

 gression above mentioned for each degree of refrangibility. 



4. In consequence, as the thickness of a medium varies, the tint changes, 

 and this truth, which at first sight appears paradoxical, and never fails to 

 surprise when experimentally shown, is general. That ray in the spec- 

 trum which is least energetically absorbed will, of course, penetrate 

 through the greatest thickness, and its ultimate tint will be a homogeneous 

 one of this particular refrangibility. 



5r» The energy with which coloured media attack the different rays is not 

 only not the same, in all parts of the spectrum, but, moreover, follows no 

 regular law of progression in proceeding from one end of the spectrum to 

 the other. Dr Brewster has given coloured drawings (to which the engrav- 

 er or colourer cannot have done justice) of spectra, seen through various 

 media. Mr Herschel represents the intensity of a ray transmitted through 

 a medium of given thickness by the ordinate of a curve, taking that of the 

 intromitted ray for one, and the length of the spectrum for the abscissa. 

 This curve appears in numerous cases to have several maxima and minima, 

 in consequence of which, the media it represents have really two or more 

 distinct colours, and undergo not merely a change of shade by an increase 



11 



