Dr. Ketteler on the Dispersion of Light in Gases. 341 



pressures p x being observed as the pressure was increasing, and 



p 2 as it was falling. The quotients - (neglecting the two first) 



are perfectly constant, with the exception of a slight decrease, 

 which is sufficiently explained by conditions of temperature of 

 the various parts of the apparatus. 



As a mean value for the indices of refraction, I obtained, from 

 two only slightly differing experiments with common air, the fol- 

 lowing number applicable to the temperature 0° C. and the nor- 

 mal barometric pressure : — 



w Na = 1-00029470, 



a result which is quite accordant with those previously given by 

 Delambre, Arago, and Biot. 

 Hence we get also 



n u =1-000293669, 



w Na = 1-000294704, 



w T i =1-000295669. 



By multiplying these numbers into the above-given wave- 

 lengths, we obtain, lastly, for the wave-lengths in space free from 

 air and destitute of dispersive power, the values 



X Li =67081-2, \ Na =58897-3, \ T1 =53466-8. 



For what relates to the indices of the other gases, I must 

 refer to my complete paper, published separately under the title 

 " Observations on the Chromatic Dispersion of Gases" (Beobach- 

 tungen uber die Farbenzerstreuung der Gase) : Bonn, 1865. 



Supplement. 



To the foregoing communication, which was printed in the 

 Monthly Notices (Monatsberichte) of the Royal Academy of 

 Berlin for the year 1864, I thought myself justified in adding a 

 formula which expresses the index of refraction as a very simple 

 function of the wave-length and density, and whose two con- 

 stants retain approximately the same value for all states of ag- 

 gregation of one and the same substance. The equation in ques- 

 tion was 



where / denotes -, and A(= , J and /3 2 are constants, and /3 = \ 

 may be regarded as a wave-length. 



