Solai* and Luna?' Spectrum. 507 



It is not easy to give an adequate idea of the 

 difficulties of observation which lead to this apparently para- 

 doxical result, particularly as physicists are so far from having 

 yet investigated this region that even the barriers which have 

 closed it to research are themselves of an unfamiliar kind. I 

 can perhaps best illustrate it by analogy. Every spectro- 

 scopist knows how very hard it is to view the lines below A; 

 and that even A itself, though very large, is not an easy object 

 to see without special precautions. This arises not so much 

 from the fact that the very deep red light here, like that of 

 dull glowing iron, feebly affects the eye, but, in a still greater 

 degree, because yellow and orange light exists in relatively 

 enormous quantity in the neighbouring parts of the visible 

 spectrum, and that irregularly diffused and reflected portions 

 of this light reappear where they do not belong and over- 

 power the radiation legitimately there. Still we can put a 

 coloured glass before the slit and cut off the intruding light 

 in a great measure, and we can see the extraneous light which 

 comes in, and allow in some degree for its effects ; but here, 

 in the actual case of the unseen heat in the far more remote 

 spectral region we are about to describe, all radiations, both the 

 feeble ones we would study, and the intruders on them which 

 we would avoid, are alike invisible, and we are, of course, 

 unable in any case to use glass, since this is opaque to all the 

 rays now in question. If any one familiar with the visible 

 spectrum will imagine himself as trying to discriminate with 

 his eyes shut between these different components of the appa- 

 rent radiation just below Fraunhofer's A, and endeavouring 

 while blindfold to say how much of it legitimately belongs 

 there and how much does not, he will haA'e a better concep- 

 tion of the difficulties peculiar to our actual field of research, 

 though still an inadequate one, since the total heat radiation 

 here is at best less than the hundredth part of that in the 

 vicinity of the A line, which we have used in illustration. 



For the clearer understanding of this I must, in anticipa- 

 tion of what follows, remark that while in the solar spectrum 

 the maximum heat, as we all know, appears not -very far from 

 the red, so that the heat corresponding in a general sense to 

 the short waves is great, and to still longer ones small, in the 

 lunar invisible spectrum the reverse is the case ; for here, 

 speaking generally, the solar reflected heat found in the upper 

 part of the lunar spectrum is less than the heat apparently 

 radiated from the moon's own soil, which is of great wave- 

 length, and which we have found in the extreme region of the 

 spectrum we are now studying. In other words, the typical 

 solar-spectrum heat is greatest in the relatively short wave- 



