Page:Proceedings of the Royal Society of London Vol 69.djvu/119

Rh required, it would be sufe to make all the necessary observations and readings.

Our experience has been that if by any chance the prism-contents were altering in temperature the edges of the D-line appeared some- what blurred, but if the temperature was constant the line was per- fectly sharp and well defined. Though this method of working is accompanied by some personal discomfort, yet it is believed to be susceptible of greater accuracy than when only the temperature of the prism is varied.

As a preliminary experiment determinations were made of /* for water, purified by the methods adopted by us ; the values found were pjw == 1-33376 and /V s = 1 '33328; the latter does not differ widely from the mean value /i D 20 = 1 '33334 obtained by former observers, whose results have been recently collated by Rudolphi.*

The method of conducting a determination was as follows :

(i) The prism was washed out four times with portions of the pre- pared sample of acid, then finally charged, allowed to stand for a short time and the thermometer inserted ;

(ii) Spectrometer telescope adjusted for viewing the collimator slit directly, the point of intersection of the cross lines being made to coincide with the left-hand edge of the image of the slit, and a reading taken ;

(iii) Charged prism placed upon the table and the position of minimum deviation for the D-line found, the temperature being noted ;

(iv) Prism removed and (ii) repeated ;

(v) Repetition of (iii) after a short interval. After several hours a second set of experiments was made.

The following are the details of one particular set of determinations :

Date, 12.4.1900. Concentration of acid = 71 '57 per cent. First set, temp. 15 C< 3 (corr.). Hour, 2.50 P.M.

Micrometer No. I. Micrometer No. II.

Direct readings ......... 180 0' G" 360 0' 8"

Deviation 150 36 4 330 35 56

Difference = Min. Dev. 29 24 2 29 24 12

Mean... 29 24' 7".
 * 'Zeits. Physikal. Cbem.,' vol. 37 (1901), p. 433.