IN QUARTZ OF LIGHT IN THE VISIBLE REGION OF THE SPECTRUM. 
287 
8. The readings were taken in accordance with the principle utilised by Von Lang 
(‘ Wien. Ber.,’ 1876, vol. 74, 209-214) and by Soret and Sarasin (‘ Geneva Archives,’ 
1882, p. 54), according to which the measurements are most trustworthy when made 
by contrasting dextro- and Isevo-quartz, rather than by reading either crystal separately 
against a zero taken in the ordinary way. This principle was regarded as specially 
important in the present investigation, because the glare of satellites, &c., was entirely 
without effect when reading a zero at which all the lines were extinguished together, 
whereas it became almost the dominant factor when reading long columns of quartz. 
For this reason it was felt that better results would follow from reading a Imvo-column 
of 226‘3670 mm. against the dextro-column of 181'4382 mm., than by attempting to 
read the whole of the available 500‘8 mm. of Isevo-quartz, either against a column of 
quartz of less than one-half this length, or against a zero. Valuable check-readings 
could be obtained by reading the two columns of quartz against one another with both 
a positive and a negative half-shadow angle. The readings of the dextro- and Isevo- 
quartz, read separately against a zero, are, however, of value in considering the relative 
accuracy of the different experiments, since their concordance could not be seen until 
all the measurements had been taken and the results reduced to unit lengths 
7. Tabulated Ohservatio7is. Form of the Dispei-sion Curve. 
Table I. shows for a series of 24 lines :— 
A. The wave-lengths in Angstrom units ; 
B. The actual rotations produced by the dextro- and Isevo-columns of quartz, read 
against a zero, with both positive and negative half-shadow angles; 
C. The ratio of the readings for each wave-length relatively to those for mercury 
green in the case of— 
{a) dextro-quartz, average of + and — half'-shadow angles; 
(6) Isevo-quartz, „ „ „ „ ,, ; 
(c) dextro against Isevo, half-shadow angle -f- ; 
(^) 5) 55 5 5 5 5 5 5 5 
D. The mean values of these dispersion-ratios and the average deviation (ranging 
from 1 to 13 parts per million) of the four ratios from the mean ; 
E. The rotation in degrees per millimetre taking for mercury green the value 
25'5371 discussed and adopted in § 5 ; 
F. The rotations in degrees per millimetre calculated for each of the 24 wave¬ 
lengths from the formula 
11-6064 , 13-42 4-3685 
^ — —ii-—T -:: — - o — ’ 
X'-Xy X' 
where 
= 0-010627, Xs^ = 78-22, X == wave-length in microns. 
G. The differences between the calculated and observed rotations in degrees per 
millimetre. 
