213 On the Decomposition of Light. 



sorption, appear to give some preference to the latter 

 method. 



Slill I must repeat, that the observations I have here 

 made are only to show the possibiUty of the thing; the 

 question can be decided only by the direct examination of 

 the rays of light on the spectrum in its simplest state, and 

 this remains for me to give. 



As few have the means of procuring this very simple 

 spectrum, and there is some difficulty in applying them, I 

 shall enter into this subject somewhat at larg;e. This I con- 

 ceive to be the more necessary, as few appear to have re- 

 peated experiments of this kind since Newton, at least with 

 due precision. Treatises on optics indeed do not mention 

 this repetition formally, many philosophers having attempted 

 it without success, and others having persuaded themselves 

 a little too hasty, that thev had completely succeeded ; as 

 Abbe Nollet, for instance, whose name has been quoted as 

 an authority. 



I should not myself have had the means I long wished, 

 but from the politeness and enlightened assistance of Mr. 

 Tremery. Fortunately his study was provided with every 

 thing necessary ; but I sh;.ll first briefly describe the nature 

 of the experuucnt, and tiie conditions indispensably re- 

 quisite. 



The business was to repeat the experiment, in which 

 Newton obtained a well defined solar spectrum, the breadth 

 of which, by concentrating the pencil of light, was reduced 

 to -Jf^, or y^ of its length ; and which consequently exhi- 

 bited the homogeneal rays incomparably more distinct from 

 each other than in the common spectrum. Opt. Book I. 

 Part. 1. Exp. 11. 



I have already hinted above, that the success depends, 

 1st, in opciating on a pencil of light that is very small be- 

 fore it readies llie prism; Sdly, in producing by the prism 

 a considerable dispersion of the coloured rays ; and 3dly. 

 in receiving their dissected ihiage on a plane very distant 

 from the point of th(y angle of dispersion. 



But these three conditions are not of themselves suffi- 

 cient. It iS almost impracticable to attain the desired object 



by 



