92 INTERFERENCE OF RADIATIONS. [MEMOIR V. 



digo and violet spaces, which are in a highly solarized 

 condition. 



But a totally different result is obtained when the 

 daylight is not allowed to fall on the plate, either before 

 or during its exposure to the spectrum. Under these 

 circumstances the rays which would otherwise protect 

 now act on the plate and slowly whiten it. A daguerre- 

 otype spectrum formed in darkness and without pre- 

 vious exposure to the light exhibits a white stain over 

 all the less refrangible regions, and bears a marked con- 

 trast to one formed under the simultaneous action of a 

 weak daylight. For brevity I will call the former the 

 spectrum of darkness, and the latter the spectrum of day- 

 light. The following are some additional observations : 



In the spectrum of darkness there is in the white stain 

 a point of maximum action. This corresponds with the 

 maximum of protection in the spectrum of daylight. 



The white stain of the spectrum of darkness is appar- 

 ently narrower than the protected space in the spectrum 

 of daylight. 



Rays of luminous or of non-luminous heat projected 

 on the dai'kness or daylight spectra during their forma- 

 tion appear to exert no kind of special influence on the 

 result. 



The white fringe which borders the solarized portion 

 is not due to anything analogous to conduction. These 

 chemical changes, unlike thermal changes, cannot be con- 

 ducted. 



By interposing between the prism and the daguerreo- 

 type plate a small convex lens of short focus, so as to 

 intercept in succession each of the colored rays, I threw 

 all over the plate, while the spectrum was in the act of 

 being impressed upon it, red, orange, yellow, and other 

 lights in succession ; the object being to ascertain how 

 far the impressed spectrum would change when these 



