1885.] on Sunlight and the Earth's Atmosphere, 271 



to an army divided into numerous brigades, eacli wearing a distinct 

 uniform, one red, one green, one blue, so that all the colours are 

 represented each by its own body. If, to re2)resent the light absorbed 

 as it progresses, we suj^posed that the army advances under a fire 

 which thins its numbers, we should have to consider that (to give the 

 case of nature) this destructive fire was directed chiefly against those 

 divisions which were dressed in blue, or allied colours, so that the 

 army was thinned out unequally, many men in blue being killed otf 

 for one in red, and that by the time it has advanced a certain distance 

 under fire the proportion of the men in each brigade has been altered, 

 the red being comparatively unhurt. Almost all absorption is thus 

 selective in its action, and often in an astonishing degree, killing off, 

 so to speak, certain rays in jDreference to others, as though by an 

 intelligent choice, and destroying most, not only of certain divisions 

 (to continue our illustration), but even picking out certain files in 

 each company. Every ray, then, has its own individuality, and on 

 this I cannot too strongly insist ; for just as two men retain their 

 personalities under the same red uniform, and one may fall and the 

 other survive, though they touch shoulders in the ranks, so in the 

 spectrum certain parts will be blotted out by absorption, while others 

 next to them may escape. 



To illustrate this selective absorption, I put a piece of didymium 

 glass in the path of the ray. It will, of course, absorb some of the 

 light, but instead of dimming the whole spectrum, we might almost 

 say it has arbitrarily chosen to select one narrow part for action, in 

 this particular case choosing a narrow file near the orange, and letting 

 all the rest go unharmed. In this arbitrary way our atmosphere 

 operates, but in a far more complex manner, taking out a narrow file 

 here and another there, in hundreds of j^laces, all through the spectrum, 

 but on the whole much the most in the blue, the Frauenhofer lines 

 being merely part of the evidence of this wonderful quasi-intelligent 

 action which bears the name of selective absorption. 



Before we leave this spectrum, let us recall one most important 

 matter. We know that here beyond the red is solar energy in the 

 form of heat which we cannot see, but not on that account any less 

 important. More than half the whole power of the sun is here 

 invisible, and if we are to study completely the action of our atmo- 

 sphere, we shall have to pay great attention to this part, and find out 

 some way of determining the loss in it, which will be difficult, for the 

 ultra-red end is not only invisible, but compressed, the red end being 

 shut up like the closed pages of a book, as you may notice by com- 

 paring the narrowness of the red with the width of the blue. 



Now refraction by a prism is not the only way of forming a 

 spectrum. Nature furnishes us colour not only from the rainbow, 

 but from non-transparent substances like mother-of-pearl, where the 

 iridescent hues are due to microscopically fine lines. Art has lately 

 surpassed nature in these wonderful "gratings," consisting of pieces 

 of polished metal, in which we see at first nothing to account for the 



