598 
MR. R„ T. GLAZEBROOK ON THE REFRACTION OF PLANE 
biquartz, and when everything was adjusted the line of separation was seen clearly 
and distinctly. 
Let us consider the ordinary image. Owing to the dispersion of the planes of 
polarization produced by the biquartz,, the light of different colours in the incident 
wave is polarized in different planes, and a position can be found for the polarizer such 
that the plane of polarization of light of a certain colour on emerging from one-half 
of the biquartz is so related to the angle of incidence on the spar prism that the light 
of that colour is absent from the refracted beam. A dark band will be seen across 
od e-half the spectrum in the place of this colour. Light of this colour emerging from 
the other half of the biquartz is in general not polarized in the same plane, and 
therefore in general* though a dark band will be formed in both halves of the 
spectrum, it will occupy different positions in the two. By turning the polarizer 
these bands appear to move in opposite directions across the field, and for one position 
of the polarizer the one can be brought vertically below the other. This position can 
be determined with great accuracy, 
When this is the case the wave length of the light destroyed is clearly such that 
it has been rotated through 90° in opposite directions by the two halves of the 
biquartz. It is light then of a definite wave length, and we are thus able to place 
our prism with great accuracy in a position such that no light of one certain definite 
wave length is present in the ordinary wave. If, then, we are able to observe the 
angle of incidence and the deviation of the light of the same wave length in the 
extraordinary spectrum we shall have enough data to determine the azimuth of the 
plane of polarization of the incident light according to the electro-magnetic theory. 
It is easy enough to observe the angle of incidence. To find the deviation of the 
corresponding wave in the extraordinary spectrum, rotate the polarizer through about 
90°; the dark bands will move out of the ordinary into the extraordinairy spectrum, 
and the polarizer can be adjusted till they are brought to coincidence in it. When 
this is the case we know that it is light of the same wave length as before (viz.: that 
whose plane of polarization is turned through 90° by the biquartz), which is absent, 
and we have in the extraordinary spectrum a well-marked dark band, whose centre 
can easily be determined, and to which the needle point or cross wires of the observing 
telescope can be set with all the accuracy required. If we observe then the deviation 
of this dark band, we obtain the deviation in the extraordinary spectrum of the light 
which in the first part of our observation was wanting from the ordinary spectrum. 
To escape the difficulty of having continually, when making observations for the 
determination of the position of the plane of polarization, to turn the polarizer through 
about 90° in order to get the deviation of the light in the extraordinary spectrum when 
the ordinary was quenched, or vice versa, I divided the operation into two parts. 
In the first I set the spar prism at a known angle of incidence and turned the 
polarizer until the dark bands coincided in the ordinary spectrum and then observed 
the deviation. I then turned the polarizer until the bands coincided in the extra- 
