212 EFFECT OF THE SUX'S ATMOSPHERE. SECT. XXIV. 



rays, and an oxydating action in the less refrangible part of the 

 spectrum, is manifest from the blackening of one and the 

 bleaching effect of the other ; but the peculiar action of the red 

 rays in the experiments mentioned shows that some other prin- 

 ciple exists different from contrariety of action. These opposite 

 qualities are balanced or neutralized in the region of the mean 

 yellow ray. But, although this is the general character of the 

 photographic spectrum, under certain circumstances even the red 

 rays have a deoxydating power, while the blue and violet exert 

 a contrary influence ; but these are rare exceptions. 



The photographic action of the two portions of the solar spec- 

 trum being so different, Sir John Herschel tried the effect of 

 their united action by superposing the less refrangible part of 

 the spectrum over the more refrangible portion by means of two 

 prisms ; and he thus discovered that two rays of different refran- 

 gibility, and therefore of different lengths of undulation, acting 

 simultaneously, produce an effect which neither, acting separately, 

 can do. 



Some circumstances that occurred during the analysis of the 

 chemical spectrum seem to indicate an absorptive action in the 

 sun's atmosphere. The spectral image impressed on paper pre- 

 pared with nitrate of silver and Rochelle salt commenced at, or 

 very little below, the mean yellow ray, of a delicate lead colour ; 

 and when the action was arrested, such was the character of the 

 whole photographic spectrum. But, when the light of the solar 

 spectrum was allowed to continue its action, there was observed 

 to come on suddenly a new and much more intense impression of 

 darkness, confined in length to the blue and violet rays ; and, 

 what is most remarkable, confined also in breadth to the middle 

 of the sun's image, so far at least as to leave a border of the 

 lead-coloured spectrum traceable, not only round the clear and 

 w r ell -defined convexity of the dark interior spectrum at the less 

 refrangible end, but also laterally along both its edges ; and this 

 border was the more easily traced, and less liable to be mistaken, 

 from its striking contrast of colour with the interior spectrum, 

 the former being lead gray, the latter an extremely rich deep 

 velvety brown. The less refrangible end of this interior brown 

 spectrum presented a sharply terminated and regularly ellip- 

 tical contour, the more refrangible a less decided one. " It may 

 seem too hazardous," Sir John continues, " to look lbi\ the cause 



