depending on the Polarimtibn of Light by Rejteximi. 14T 
angle less than 56°, the polarising angle for glass, the reflected 
ray CD will not be polarised, arid c*onsequently, an image of 
that candle, not very bright, will be seen by reflection from D. 
When an observer, therefore, looks into the two mirrors B, D 5 
he will see no image of the candle in B, and a distinct image 
in D. While he is looking at these reflectors, let another, per^. 
son breath gently upon the glass reflectors A, C, and image 
of the candle will instantly appear in where there was none 
before^ while the image of the candle that was seen in 1) will he 
extinguished. The effect, therefore, of the same film of vapour 
has been to extinguish the image of the candle in D, and to 
create a new image in C. When the aqueous film has evapo- 
rated, which takes place very rapidly, the new image will be ex- 
tinguished, and the extinguished one wilf reappear. 
The cause of thi^ paradoxical effect may be easily under- 
stood. When a film of aqueous vapour is laid upon the reflec- 
tor A, the surface of the glass is, as it were, converted into- 
water, and the ray RA is no longer polarised, because it is re- 
flected from an aqueous surface, whose polarising angle is about 
53*. Hence, as the ray AB is not polarised, a portion of it 
will be reflected from B, and will cause the candle from which 
it proceeds to be visible. 
In the reflector C, where the ray CB is not polarised, in con- 
sequence of its being incident at an angle less than the polaris- 
ing angle for glass, the substitution of a film of aqueous vapour 
will polarise the ray CD comp etely, as the ray RC is incident 
upon the surface of it at an angle of 53°, wliich is the polarising 
angle for water. Hence, since the reflector D is placed in the 
position which will not reflect a polarised ray, no part of the 
ray CD wrill now be reflected from D, and, consequently, the 
image of the candle, which was formerly seen, will be extin- 
guished by throwing a film of vapour upon the reflector C. 
This experiment, independent of its paradoxical character, 
affords a very simple and striking illustration of the relation be- 
tween the refractive power of bodies, and the angle at which 
they polarise the incident light. 
K 2 
