408 Prof. S. P. Langley on Invisible Heat-spectra. 



lines could be seen in its spectrum which the rock-salt prism 

 would not also show. The probable error of one setting of 

 the micrometer was less than \" . From the agreement of 

 the different measurements made in this way, it is believed 

 that the deviations throughout the visible spectrum are correct 

 to within h" . Those in the ultra-violet and infra-red cannot 

 of course pretend to this degree of accuracy. The positions 

 of p 9 t (invisible) given by two independent series with the 

 bolometer differed by 30 /x ; those of <j> (invisible) by ]/, those 

 of (f> (invisible) by 30", and those of O (invisible) agreed 

 exactly. 



We have thus obtained incidentally the data for constructing 

 a table of refractive indices of rock-salt throughout the entire 

 range of the solar spectrum, with an accuracy which we 

 believe to be greater than has heretofore been attained, and 

 which we deem of sufficient interest to give in full below, for 

 the convenience of others having occasion to work with this 

 material, and for testing theories of dispersion. 



Refk acting Angle of Prism — 59° 57 / 54/ 



Line. 



Wave-length. 



Deviation. 



A 



Refractive Index. 



M 



0-3727 



43 50 57 



1-21 



1-57486 



L 



0-3820 



43 35 27 



1-20 



1-57207 



H a 



03933 



43 19 32 



119 



1-56920 



H a 



0-3968 



43 14 44 



119 



1-56833 



G 



0-4303 



42 36 7 



1-16 



1-56133 



F 



0-4861 



41 51 47 



1-13 



1-55323 



b, 



0-5167 



41 33 43 



112 



1-54991 



b\ 



0-5183 



41 32 52 



1-12 



1-54975 



!>! 



5889 



41 2 41 



1-10 



1-54418 



Do 



0-5895 



41 2 29 



110 



1-54414 







0-6562 



40 42 56 



1-09 



1-54051 



B 



0-6867 



40 35 49 



109 



1-53919 



A 



0-7601 



40 22 25 



1-08 



1-53670 



p s r ... 



0-94 



40 1 26 



107 



1-5328 



6 



113 



39 49 11 



106 



1-5305 



* 



1-38 



39 39 56 



105 



1-5287 



Q 



1-82 



39 29 21 



1-05 



1-5268 



Temp. =24° C. Barometer, 733'1 mm. 



The wave-lengths of the M and L lines are from Cornu, 



those of the lines between H and A inclusive from Angstrom, 

 and those of the infra-red bands from the Allegheny observa- 

 tions. The column headed A was prepared at the suggestion 

 of Mr. Keeler, and is for the purpose of facilitating the 

 reduction of observations made with a different prism-angle 

 from that for which the table is computed and for which our 



