In the experiment just described, the eye would seem to have some
share in the colours of the rainbow, since these bubbles in the
glass do not display the colours except through the medium of the
eye. But, if you place the glass full of water on the window sill,
in such a position as that the outer side is exposed to the sun's
rays, you will see the same colours produced in the spot of light
thrown through the glass and upon the floor, in a dark place, below
the window; and as the eye is not here concerned in it, we may
evidently, and with certainty pronounce that the eye has no share in
producing them.