90 INTERFERENCE OF RADIATIONS. [MEMOIR V. 



mercurialized it would have whitened uniformly all 

 over, on being made to receive the spectrum the less 

 refrangible rays actually carried it back to the unaffect- 

 ed condition, reversing what had been already done. 

 While the more refrangible rays were forcing it on to 

 the solarized state, these were returning it into the con- 

 dition of shadow : they therefore not only protect, but 

 seem even to exert a negative or antagonistic action. 



Sir J. Herschel has critically examined one of these 

 specimens, and has suggested an explanation of their 

 appearance on the Newtonian principle of the tints ex- 

 hibited by thin films (Phil Mag., Feb., 1843). But I 

 found that it was immaterial whether the exposure to 

 the spectrum was for thirty seconds or one hour the 

 result was the same. The final action had been pro- 

 duced, the less refrangible rays had carried their region 

 to the unaffected state, while the more refrangible had 

 solarized theirs. Now if the phenomenon were due, as 

 M. Becquerel supposes, to an unequal action of the same 

 kind in the different rays, it is obvious that the final 

 result ought to depend on the time of exposure; the 

 red ray, aided by the daylight, should carry its portion 

 through the various shades of white, and solarize it at 

 last. But this in the longest exposure never takes place; 

 that part of the plate remains as though a ray of light 

 had never fallen upon it. 



Such are the facts I observed, and they seem to have 

 been reproduced by MM. Foucault and Fizeau; but 

 there are also others of a much more singular nature. 

 In these Virginia specimens the same protecting action 

 reappears beyond the violet. 



The only impressions in which I have ever seen this 

 protecting action beyond the violet are those made in 

 Virginia in 1842; they were made in the month of July. 

 Struck with this peculiarity, on my return to New York 



