294 
DE. T. M. LOWEY: NATUEAL AND MAGNETIC EOTATOEY DISPEESION 
with what would be expected in a substance obeying the same dispersion-law but 
possessing a rather lower dispersive power. The figures are of importance as showing 
that the defects reduced the dispersive power of the specimens, as well as the absolute 
values of the rotations. No such steady difference could be detected between the 
values for the specimens of dextro- and Imvo-quartz finally selected. These gave the 
numbers 
Dextro. 
Lsevo. 
Li . . 
. 0-647508 
0-647539 
(- 0-000031) 
Hg . . 
. 1-627001 
1-626987 
(+ 0-000014), 
corresponding with errors of reading of about 0°‘16 and 0°'07 respectively; the 
differences do not exceed the possible range of experimental error and (taking the 
laevo-cylinders as the standard) are also in the opposite direction to those recorded 
above for specimens that were known to be faulty. 
9. Tests of Wiedemann’s Law. 
The law of the proportionality of natural and magnetic rotatory powers over a range of 
wave-lengths was discovered by G. Wiedemann in 1851 (‘ Pogg. Ann.’, vol. 82, 215) 
from experiments on turpentine, using a series of five Fraunhofer lines. It was 
tested after an interval of over half a century by Disoh (‘ Ann. d. Physik,’ 1903 [IV.], 
vol. 12, 1153), whose results are given in a memoir which is of further interest as 
being the earliest in which the use of the Arons mercury lamp in polarimetric work is 
described. Disch found a steady difference between the two dispersions as is shown 
by the following ratios :— 
Wave-length. 
. 6563 
5893 
5780 
5461 
4916 
4359 
4050 
Ethyl valerate 
f natural . 
. 1-000 
1-258 
_ 
1-500 
1-924 
2-573 
[ magnetic 
. 1-000 
1-167 
— 
1-380 
1-746 
2-267 
— 
Quartz. • • 
f natural . 
. 1-000 
1-254 
1-307 
1-475 
1-846 
2-400 
2-826 
L magnetic 
. rooo 
1-251 
1-301 
1-464 
1-821 
2-316 
2-672 
It is remarkable that Disch should have concluded from these figures that “ both 
the tables and the curves show that Wiedemann’s Law holds to a very close 
approximation : the quotient njm is as good as constant, the ratios for the optical 
and magnetic rotations are nearly equal to one another and accordingly the two curves 
fall almost together.” The deviations, which amount to 12 per cent, in one case and 
5 per cent, in the other, were attributed by Disch to a lack of optical homogeneity in 
the ester and to a lesser extent in the quartz. 
As this matter was one of considerable importance, especially as regards the theory 
of magnetic rotation, experiments were made to test the law in the case of quartz as 
well as of a series of organic liquids whose optical and magnetic rotatory dispersions 
