MEMOIR V.] INTERFERENCE OF RADIATIONS. 95 



upon the violet, so that the colors should half overlap 

 each other, I found that at the point of concourse the 

 plate instantly solarized, and assumed a splendid green 

 metallic color. 



I have now explained the acceptation in which I re- 

 ceive the negative ray as a synonym (in this instance of 

 iodide of silver) for the yellow ray, and alluded to the 

 mechanism which seems to be the cause of protecting ac- 

 tion generally. Perhaps on a review of his own experi- 

 ments M. Becquerel may find reason to believe that there 

 are in reality antagonizing actions in different parts of 

 the spectrum ; actions not limited to the daguerreotype, 

 but occurring in all kinds of cases. They have been met 

 with by every one who has examined the spectrum with 

 sensitive papers, and, in a different series of phenomena, 

 M. Becquerel has himself furnished a conclusive illustra- 

 tion. He shows that when sulphide of lime and other 

 phosphorescent bodies in a shining state are exposed to 

 the spectrum, the more refrangible rays increase the glow, 

 but the less extinguish it. 



It is proper to observe that some of the phenomena 

 recorded in this communication which seem to be in op- 

 position to the principle set forth are not so in reality. 

 All reasonings founded on the decomposition of light by 

 the prism, and the action of the prismatic spectrum on 

 changeable surfaces, are liable to error. The only meth- 

 od free from these difficulties is to employ the diffraction 

 spectrum formed by a ruled surface or a grating, a meth- 

 od which was proposed eight years ago by Herschel 

 with a view of getting rid of the disturbing agencies 

 arising from the ideal coloration of glass, and which I 

 first carried into effect in 1844 with so much success 

 that the resulting daguerreotype impressions contained 

 Fraunhofer's lines, even with microscopic minuteness. 

 With this spectrum we avoid a far more serious diffi- 



