a Dispersion-Photometer. 



119 



be assured of there being no reflection from the inside of the 

 box. Beside the instrument, and containing the standard 



|3B 



candle, is a box blackened inside, in one end of which is a 

 screen similar to D. If D is the distance from the light to 

 the lens and d is the distance from the principal focus of the 

 lens to the screen; if 8 is the focal length of the lens, then, 

 roughly, a bundle of rays of unit solid angle gets to have, after 

 refraction, an angle 



D 2 



as D is always great in comparison with 8. If L is the total 

 light, then the unit angle of incident rays and the unit angle 

 of refracted rays have amounts of light 



L , L S 2 . 



so that a screen at the distance d from the focus of the lens 

 has illumination of the intensity 



_£_ ^ A 



4tt ' D 2 " d 1% 



If, now, another screen has the same illumination from a candle 

 whose total light is unity, at the distance Dj this illumination 

 is 



1 _L 



47rJD'j ~~ 4-7T 



and hence 



T D 2 d 2 



A double-concave lens, of focal length 1 centimetre, being 

 employed, and d being capable of variation from 40 to 10 cen- 

 timetres, and D being as much as five times D 1? we can mea- 

 sure in a very small space a light which is from 40,000 to 100 

 times the standard candle. 



We propose to make careful experiments on the absorption 



D 2 



