466 Intelligence and Miscellaneous Articles. 



to the extreme end of the visible part, shows, among other things, 

 big absorption-bands. The two more-refrangible ones have already 

 been split up into groups of bands and very fine lines ; and it is 

 probable that a more complete investigation would lead to a similar 

 result for the two others. In my former researches I indicated 

 these bands, starting from A, by the letters A', A", A'", and A IV . 



The bands A', A", A'" appear to have been observed by most 

 physicists who have examined the infra-red region of the solar spec- 

 trum by thermoscopic methods, notably by J. Herschel, Fizeau and 

 Foucault, Desains, and Lamansky. The same bands were observed 

 by my father in 1866 by applying methods relating to phospho- 

 rescence ; but the values (varying considerably among themselves) 

 which different physicists have obtained for the wave-lengths of 

 these rays do not allow of the identification, with any certainty, of 

 the results of their researches. In 1847 Fizeau gave the number 

 0*001445 millim. for the wave-length of a band which appears to 

 be the band A'". On the other hand, my father obtained for this 

 same band, by the phosphorographic method, numbers varying 

 between 0*00i400 and 0*001200 millim., and gave the number 

 0*001220 millim. for its most-refrangible border. It was therefore 

 necessary to determine these wave-lengths directly by using a 

 grating. 



In 1879 Oapt. Abney prepared a very beautiful map of the infra- 

 red portion of the normal spectrum ; it was obtained by photo- 

 graphy, and extended to about the wave-length 0*000980 millim. 

 In my previous researches I was enabled, by the application of 

 phosphorescence, to measure, in the diffraction-spectrum of a gra- 

 ting, to about 0*001000 millim. ; but the feeble brilliancy of the 

 spectra did not permit of my going further, and I adopted the 

 number 0*001220 millim. for the most-refrangible border of the 

 band A'" which can be easily observed in the spectrum formed by a 

 prism. The numbers above 0*001000 millim., published in the paper 

 cited above, were obtained by means of an interpolation based upon 

 this assumption. 



In a memoir published in 1883 Mr. Langley deduces the wave- 

 lengths of the bands A', A", A'", A IV from their position in the 

 spectrum formed by a prism by an ingenious method of graduation 

 of his spectrobolometer ; and the numbers obtained by him, starting 

 from 0*001000 millim., differed considerably from those given by 

 our hypothetical interpolation. I then proposed to make these 

 determinations again, by throwing the diffraction -spectra obtained 

 by a grating on phosphorescent substances considerably more sen- 

 sible for infra-red rays than those which were at my disposal during 

 my former researches. These substances have allowed of my mea- 

 suring, with a sufficiently near approximation, the wave-lengths 

 of the bands in question, as well as several of the finer lines ; the 

 numbers given by these experiments are near enough to those ob- 

 tained by Mr. Langley with his bolometer, but the delicacy of the 

 method of observation used by me is much greater. 



The solar rays, concentrated upon a narrow slit, in the focus of a 



