of Edinburgh, Session 1879-80. 
473 
6. On some Applications of Rotatory Polarization. 
By Professor Tait. 
Since last meeting of the Society I have found that Broch (in 
Dove’s Repertorium) employed the combination of prism, Nicols, 
and quartz plate, for the purpose of measuring the rotatory power of 
quartz for different wave-lengths. I do not find, however, that he 
suggests its use for the determination of wave-lengths according to 
one definite standard. Nor does he seem to have used great thick- 
nesses of quartz, which is essential to accuracy in the application I 
have proposed. In the Annales de Chimie , 1846, there is a transla- 
tion of a part of Broch’s paper, with the remark that the process, 
which he called a new one, was due to Pizeau and Foucault. Their 
paper, however, refers to quartz cut parallel to the axis ; but it in- 
troduces the question, very important so far as my object is con- 
cerned, of the interference of two polarized rays after one has been 
retarded more than the other by very many wave-lengths. It is 
quite possible that this consideration, which I had for the time 
forgotten, may be found fatal to my method when very great 
thicknesses of quartz are employed on the bright lines given by 
glowing gases, for the purpose of estimating the velocity of the indi- 
vidual particles. It will not affect the method as applied to the 
spectra of auroras, comets, &c. 
Prof. Niven {Phil. Mag. 1878) speaks of rotation of plane of 
polarization as the most delicate test of change of wave-length. It is 
so in theory, but in practice it cannot be compared to a train of prisms. 
I have within the last fortnight operated with pieces of quartz 
from 4 to 8 inches in length, and have found that the sharp- 
ness of extinction of the red line of lithium is greater than that 
of the green line of thallium. The breadth of the latter must of 
course come more into play. The range of uncertainty for the orange 
sodium line is very much greater still. This was to be expected 
from its being double. With thick plates of quartz it cannot be 
extinguished at all. In fact, in order to extinguish it, a plate would 
be required which would make the difference of rotation for the two 
constituents one or more semi-circumferences. The least thickness 
of quartz for such a purpose would be somewhere about 13 feet, and 
about 500 successive oscillations of the luminous particles would 
