Composition of White Light. 97 



it is well known that the dispersion of the red light in glass 

 is less than the violet, and, therefore, the red ray after three 

 reflexions will be less elevated than is here represented. 



The red ray after six reflexions, will emerge below the 

 direct ray, and parallel to it ; so as to form with it, and any 

 other reflexions, multiples of six, an elongated image in the 

 spectrum ; but as the angle of dispersion in red is less than 

 in violet light, the reflected rays approach nearer to the 

 direct ray, and the elongation of the red light is less con- 

 siderable ; the images being in a greater degree superposed. 



We thus account for the formation of blue and yellow 

 light, by the intersection of rays nearly parallel, though of 

 different refrangibility ; we give a probable reason for the 

 presence of chemical rays at one end of the spectrum, and 

 heating rays at the other ; both being produced by reflected 

 images, invisible from their low intensity, when, if our 

 preliminary observations be correct, the light is more readily 

 absorbed ; we also account for the increased elongation of 

 the spectrum in strong light ; and for various appearances 

 in it, when subjected to the action of absorbing media. 



Such a concurrence of circumstances can scarcely be 

 attributed to mere chance ; and, connected with the con- 

 siderations previously advanced, although it may not be 

 sufficient to prove the truth of our hypothesis, at least 

 gives it a claim to farther inquiry. It must be observed, 

 however, that although, upon further investigation, this 

 hypothesis may fail to account for those appearances in the 

 spectrum, which may be supposed to support the idea of 

 innumerable degrees of refrangibility, it will not by any 

 means invalidate the proofs that are advanced in this paper 

 of the erroneous character of this doctrine, which are per- 

 fectly independent of it; the only difference this failure 

 would make, would be, that we should have to direct our 

 inquires to other probable causes ; and several readily offer 

 themselves to our consideration. 



When the spectrum is formed by suffering the unob- 

 structed light of the sun to fall on a prism, and its coloured 

 images fall perpendicularly on a screen, there will be 

 observed a disposition in the red part of the spectrum, par- 

 ticularly in the lower part of it, to assume a circular form ; 

 and upon looking at the spectrum through coloured glasses, 



VOL. III. JI 



