336 
been polished particularly well. Neither is it now necessary, that 
these faces are perfectly plane and exactly parallel to each other; 
it suffices when the glass-plates are rather exactly parallel to each 
other. We had only to take care that these plates are not exposed 
to a too high pressure, so that the closeness (with caoutchoue rings 
between the glass plates and plane metal rims) was not perfectly 
tight. This however was of no importance. 
5. When after having been fixed in this way the crystal was 
placed between crossed Nicols the light reappears again and when 
we fix the eye on the crystal we see irregular light spots over its 
extension (about 1 cm’). These are due to inner structure deviations 
and tensions and might retain us from further investigation, when 
not the differences in phase in this ‘accidental’ double refraction 
proved to be rather small. When anywhere they reached the value 
of half a wavelength we should see dark spots between parallel 
Nicols. There is no question of this; a rotation of one of the Nicols 
over a few degrees from their crossed situation sufficed to obtain a 
uniform distribution of the light. 
The irregular differences in phase being so small, it might be 
expected that the effect of a regular anisotropy, the same over the 
whole cross-section of the light beam could be observed when it was 
superposed on the irregular differences. In fact this was proved to 
be the case when a thin plate of mica was adjusted before the 
crystal and rotated in its own plane. 
6. Now the experiment was made on the following way. A small 
round aperture into which the rays of a glow-lamp with opaque 
bulb are folling is placed in the focus of a collimator lens. After 
passing this lens the rays fall in a telescope that is focussed for 
parallel rays. 
We then see a sharp image of the lighted aperture, which image 
is extincted by two Nicols placed between the collimator lens and 
the telescope. 
When now the crystal is placed between the Nicols and when 
these are rotated, we see in positions differing 90° minima of the 
intensity of the light. These minima are not always equally pro- 
nounced, but they are always easy to recognize. 
That in the mentioned positions the light is not perfectly extin- 
guished, must of course be ascribed to the accidental double refrac- 
tion, which now however causes a uniform illumination. As we 
bave namely focussed not at the crystal but at the lighted aperture, 
