SUNLIGHT AND THE EARTH'S ATMOSPHERE. 645 



tion clearer, observe that this blueness is not a pure spectral blue. It 

 has in it red, yellow, blue, and all the colors which make up white, but 

 blue in superabundance ; so that, though the white is, so to say, latent 

 there, the dominant effect is blue. The glass colored veil does not put 

 anything in, but acts I repeat like a sieve straining out the blue, and 

 letting through to us the white light which was there in the bluishness, 

 and so may not our air do so too ? 



I think we already begin to see that it is at any rate conceivable 

 that we may have been hitherto under a delusion about the true color 

 of the sun, though of course this is not proving that we have been so, 

 and it will at any rate, I hope, be evident that here is a question 

 raised which ought to be settled, for the blueness of the sun, if proved, 

 evidently affects our present knowledge in many ways, and will modi- 

 fy our present views in optics, in meteorology, and in numerous other 

 things. In optics, because we should find that white light is not the 

 sum of the sun's radiations, but only of those dregs of them which 

 have filtered down to us ; in meteorology, because it is suggested that 

 the temperature of the globe and the condition of man on it depend 

 in part on a curious selective action of our air, which picks out parts 

 of the solar heat (for instance, that connected with its blue light), and 

 holds them back, letting other selected portions come to us, and so 

 altering the conditions on which this heat by which we live depends ; 

 in other ways, innumerable, because, as we know, the sun's heat and 

 light are facts of such central importance that they affect almost every 

 part of scientific knowledge. 



It may be asked, What suggested the idea that the sun may be blue 

 rather than any other color ? 



My own attention was first directed this way many years ago when 

 measuring the heat and light from different parts of the sun's disk. 

 It is known that the sun has an atmosphere of its own which tempers 

 its heat, and, by cutting off certain radiations and not others, produces 

 the spectral lines we are all familiar with. These lines we customarily 

 study in connection with the absorbing vapors of sodium, iron, and so 

 forth, which produce them ; but my own attention was particularly 

 given to the regions of absorption, or to the color it caused, and I 

 found that the sun's body must be deeply bluish, and that it would 

 shed blue light except for this apparently colorless solar atmosphere, 

 which really plays the part of a reddish veil, letting a little of the 

 blue appear on the center of the sun's disk where it is thinnest, and 

 staining the edge red, so that to delicate tests the center of the sun is 

 a pale aqua-marine, and its edge a garnet. The effect I found to be 

 so important, that if this all but invisible solar atmosphere were dimin- 

 ished by but a third part, the temperature of the British Islands would 

 rise above that of the torrid zone, and this directed my attention to 

 the great practical importance of studying the action of our own ter- 

 restrial atmosphere on the sun, and the antecedent probability that our 



