blue and yellow pigments. No natural colour is _pure_. A blue liquid,
or a blue powder, permits not only the blue to pass through it, but a
portion of the adjacent green. A yellow powder is transparent not only
to the yellow light, but also in part to the adjacent green. Now, when
blue and yellow are mixed together, the blue cuts off the yellow, the
orange, and the red; the yellow, on the other hand, cuts off the
violet, the indigo, and the blue. Green is the only colour to which
both are transparent, and the consequence is that, when white light
falls upon a mixture of yellow and blue powders, the green alone is
sent back to the eye. You have already seen that the fine blue
ammonia-sulphate of copper transmits a large portion of green, while
cutting off all the less refrangible light. A yellow solution of
picric acid also allows the green to pass, but quenches all the more
refrangible light. What must occur when we send a beam through both
liquids? The experimental answer to this question is now before you:
the green band of the spectrum alone remains upon the screen.

The impurity of natural colours is strikingly illustrated by an
observation recently communicated to me by Mr. Woodbury. On looking
through a blue glass at green leaves in sunshine, he saw the
superficially reflected light blue. The light, on the contrary, which
came from the body of the leaves was crimson. On examination, I found
that the glass employed in this observation transmitted both ends of
the spectrum, the red as well as the blue, and that it quenched the
middle. This furnished an easy explanation of the effect. In the
delicate spring foliage the blue of the solar light is for the most
part absorbed, and a light, mainly yellowish green, but containing a
considerable quantity of red, escapes from the leaf to the eye. On
looking at such foliage through the violet glass, the green and the
yellow are stopped, and the red alone reaches the eye. Thus regarded,