290 Miss Marian Baxter on the Dispersion 



previously suggested *. The results obtained with the lines 

 of the principal series of the Alkali metals and also the 

 work o£ Ladenburg and Loria on dispersion in luminous 

 hydrogen \ appear to confirm this theory. If the atoms 

 are differentiated in this way, then the relative number of 

 atoms required for the absorption corresponding to higher 

 members of the series of lines would be likely to decrease 

 with increasing temperature, and the ratio a x \a 2 would 

 increase. From values of the constants a A a 2 for sodium 

 and potassium calculated in the paper first mentioned, there 

 appeared to be some evidence of this increase in the value 

 of ai/a 2 with rising temperature ; but the measurements 

 of the deviated spectrum, which were made from photo- 

 graphs obtained by Wood's method of crossed prisms, would 

 not admit of great accuracy and the temperature range was 

 not wide. 



In the following year Roschdestwensky, in a paper on the 

 anomalous dispersion of sodium vapour J, pointed out the 

 importance of the question of the constancy of ai/a 2 under 

 alteration of temperature .and density of the vapour. His 

 measurements of a l and a 2 were made from photographs of 

 interference fringes produced by the method of Puccianti, 

 using a modified form of Jamin interferometer, a cylinder 

 of sodium vapour being placed in the path of one of the 

 interfering beams. The apparatus used and the method 

 of measurement employed were such as to admit of con- 

 siderably greater accuracy than was possible in Bevan's 

 measurements. But as regards the temperature effect he 

 arrived at no definite conclusion. From four photographs, 

 taken with increasing vapour density, an increase of aja 2 

 was obtained, but the highest value was uncertain and the 

 other three showed no definite increase, being approximately 

 the same. Taking all four values, the maximum increase 

 was not more than 10 per cent, when the vapour density 

 increased 50-fold. These results appear to show that for 

 sodium the value of a x ja 2 changes extremely little — if 

 at all. 



The advantages of the interferometer method, however, 

 seemed to justify another attempt with potassium vapour. 

 The dispersion was observed at the first two pairs of lines of 

 the principal series— 7699 and 7665; 4047 and 4044. The 

 positions of these lines enabled the measurements of the 

 fringe shifts to be taken over almost the whole range of the 



* Proc. Roy. Soc. A, vol. lxxxiv. (1910). 



t Verh. der deutschen phys. Ges. x. p. 858 (1908). 



% Ann. der Physik, Vierte Folge, Band 39 (1912). 



