508 Prof. S. P. Langley on the Invisible 



lengths ; the typical lunar-spectrum heat is greatest in the 

 long wave-lengths. The explanation of the curious fact that 

 this particular quality of heat may he more easily recognized 

 where it exists in a less degree as in the lunar spectrum than 

 where it is found in a relatively great degree as in the solar, 

 will be still clearer if we consent (in continuance of the 

 illustration) to further compare this lunar invisible radiation 

 of great wave-length to the deep -red light from a piece of 

 scarcely luminous hot iron. This peculiarly deep red is seen 

 with little difficulty in the iron in a dark room, but never in 

 daylight ; yet it is of a quality which we know from theory 

 must exist in far greater degree in the daylight itself ; nor 

 do we, even when we would isolate it in a certain part of the 

 solar spectrum, see it there, because it is now obscured by the 

 inevitable diffusion or reflexion of part of the neighbouring 

 brilliant light which the prism ought to keep wholly away, 

 but (owing to inevitable instrumental defects) does not. The 

 dull glowing iron carries no white light along with it, and 

 therefore its feeble peculiarly deep red is easier seen than the 

 far stronger corresponding red in the solar spectrum. 



By the aid of this analogy in the case of light, and passing 

 now to the actual case of wholly invisible radiation, I hope it 

 may be clear how the feebler heat in the lunar extreme infra- 

 red spectrum was at first recognized more easily than the 

 stronger corresponding heat in that of the sun. 



It may be asked why (if we cannot cut off the diffused heat 

 in the solar infra-red spectrum by the use of an absorbing 

 glass) we cannot put a prism in front of the slit after the plan 

 of Helmholtz. This is practically impossible here (owing to 

 instrumental conditions which we need not now explain), 

 unless we find some way of keeping the axis of the spectro- 

 bolometer either motionless or always parallel to itself, in 

 spite of the varying direction of the rays from such a prism, 

 and of automatically limiting the kind of radiation to be ob- 

 served in any part of the spectrum, to that legitimately 

 belonging there. The following arrangement was, after 

 various trials, adopted with success. Its immediate purpose 

 is to overcome the difficulty which we have just explained at 

 such length — that is, to sift out the extraneous heat which 

 remains after the ordinary action of the prism ; but it can of 

 course be used for light also. 



Description of Sifting Train (see fig. 1). 



Let N8 be a massive beam, resting on two piers, and 

 immovably fixed in the meridian. Let AE be a second 

 beam, movable on a turn-table, placed centrally beneath N S. 



