296 Miss Marian Baxter on the Dispersion 



were reduced to two or three photographs at moderate 

 temperatures. Those which were taken with Dr. Bevan in 

 the first stage o£ the work gave results which made it impos- 

 sible to come to any conclusion with regard to the change in 

 a l a 2 . With increasing temperature and vapour density both 

 increasing and decreasing values of a 1 ja 2 were obtained, and 

 generally the values were so close together that the change 

 in either case was doubtful. One series of three photographs 

 gave values which were more widely separated, and this 

 showed a decrease with increasing temperature instead of 

 the looked for increase. This series (I. 1, 2, 3.) will be 

 referred to later. 



When I was able at a later date to make another trial of 

 the method, I obtained the series of photographs numbered 

 II. 1, 2, 3, 4, 5, 6, 7, 8. II. 3 and II. 5 are printed as ex- 

 amples of these (PI. XI.) . This appears to be a reliable series. 

 Photographs with sufficiently good definition for measure- 

 ment were obtained over a much wider temperature range 

 than previously (about 308° 0.-401° C), the fringes being- 

 narrower and clearer than before. The curves for all these 

 photographs are good. There is a gap in the last two 

 curves of the series for the reason already mentioned, but 

 from the shape of the curve on either side of this gap the 

 correct position of the upper part of the curve can be fairly 

 accurately judged. In photograph 7 only a few fringes are 

 indistinct, in photograph 8 the absorption region extends 

 further and the position of the curve for higher wave- 

 lengths is consequently more doubtful. It is, however, 

 evident that the calculated values a Y and a 2 may be relied 

 upon since in both cases the curves drawn give good 

 agreement between the observed and calculated values 

 of n. 



The values of the ratio a x ja 2 as seen from the tables 

 which follow 7 show a definite decrease with rising tem- 

 perature. 



Series II. 1, 2, 3, 4, 5, 6, 7, 8. 



The temperature increases from 1-8. The following table 

 gives the approximate temperature for each photograph and 

 also the readings of points where the fringes, taken in order, 

 cross a horizontal line on the negative. This line was 

 obtained by stretching a fine silk fibre across the slit of the 



