APPENDIX VL 



INDICES OF REFRACTION OF ROCK SAIvT. 

 (Fig. 132.) 



It will be noticed that in the present work, using a rock-salt prism, 

 the maximum of the absorption band of water occurs at 2.95 fi, while 

 Paschen (loc. cit.) found it at from 2.95 /^ to 3.0 /x, and Aschkinass 

 (loc. cit.) found it at 3.0 /x, both using a fluorite prism. Furthermore, 

 it will be noticed that in the present work the conspicuous alcohol bands 

 occur at 2.95 ju and 3.43 fx, while Ransohoff (loc. cit.) found them at 

 3.0 IX and 3.43 fi. 



That this discrepancy can not be due entirely to errors in the observa- 

 tions has been shown in the discussion of the " sources of errors." 

 From the nature of the dispersion curve, as well as the magnitude of the 

 dispersion of fluorite, which is about twice as great as that of rock salt 

 for the present work, and, furthermore, from the fact that the rock-salt 

 dispersion curve passes through a double curvature near this point, it 

 would appear that this discrepancy is due to errors in the indices of 

 refraction of rock salt in the region of 2.5 /*. 



Now, it so happens that we have no observational data for this region. 

 The most recent and most reliable measurements of Rubens (loc. cit.) 

 are for wave-lengths 1.71 /x, 2.35^11, and 3.34 /a (fig. 132). LangleyV 

 values are all much larger, which can be interpreted, for the present, 

 as a shifting of his zero of wave-lengths by about o.i /* toward the infra- 

 red. The COo absorption band occurs at about 4.28/1,, while that of 

 CO occurs at 4.58 fi. Unless it can be shown that the great atmospheric 

 absorption band found by Langley at 4.4 fi is the composite of CO and 

 CO2, it would appear that his greater value, 4.4 /*, instead of 4.28 fi, is 

 due to the shift noted above. 



As indicated elsewhere, the great bench-mark in the infra-red is the 

 emission band of CO2, at 4.40 fx, using a Bunsen burner. The accuracy 

 of the location of this band has also been noticed; and it was shown 

 that all observers agree in the location of this band for fluorite, includ- 

 ing myself, using rock salt. Langley (loc. cit., p. 215) finds this band 

 at 4.6/1, using a Welsbach mantle heated by the flame of a Kitson lamp, 



^Langley : Annals of Astrophys. Obs., vol. i, p. 261. 

 132 



