TRANSLATIONS AND SELECTED ARTICLES. 43 
reflected light. Newton says, “water reflects the violet, blue, and 
green rays, but readily transmits the red.” Count Xavier de Maistre 
considers the colour of water to be blue in reflected, and yellowish 
orange in transmitted light. Arago, that it is blue in reflected, and 
green in transmitted light. The view that the blue of water only 
occurs in reflected light is common to all three statements. 
In the experiments which Bunsen made to ascertain the colour of 
distilled water, transmitted light was alone concerned, and yet he 
found the colour blue. In order to look through still longer columns 
of water I used the following apparatus:—A box, the bottom and 
and sides of which are made of plates of gutta percha, is closed at both 
ends by parallel plates of very white thin glass. Directly inside these, 
two similar glass plates are fixed, which are covered with a silver re- 
flecting surface, by Liebig’s method. A narrow slit is scratched in the 
covering at dandd’. If apencil of direct sunlight is projected upon slit 
d, this will be reflected several times backward and forward between the 
two mirrors; if the box is filled with a liquid, the light is compelled 
to traverse this liquid repeatedly, and it is easy to increase or diminish 
the length of the layer to be traversed, by altering the angle of inci- 
dence. This experiment may be made either objectively or subject- 
ively. If the pencil is allowed to fall into the slit d, so that after a - 
certain even number of reflexions it falls directly upon the slit d’, it . 
can be caught upon a screen after its emergence. The number of 
reflexions may be altered by gradually rotating the box. But if the 
observer uses the illuminated slit d as a self-luminous object, and looks 
through d' into the box, he sees, close to one another, a series of nar- 
row subjective pictures of the slit; they are gradually smaller and 
nearer each other, and correspond to the different numbers of reflex- 
ions. In making some experiments, I had, at first, so placed the 
mirrors that the uncovered glass surfaces were opposite each other. 
The light must then, at each reflexion, pass twice through the glass 
plates themselves. If the box contained no liquid, then the image 
appeared almost white after six to eight reflexions ; but still, on com- 
paring the subjective images with one another, it could be seen that 
each following one had a somewhat yellowish tint. I supposed that 
this coloration was to be ascribed to the tolerably thick layer of glass 
which the light had to traverse, and therefore turned the mirrors, 
which were once more polished on the silvered side. Yet even in this 
case each following image showed a yellower colour, though in a less 
