167.
This glass may be adapted for experiments in various ways: it may either be made in a very slight degree non-transparent, in which case the light seen through various layers placed one upon the other may be deepened from the lightest yellow to the deepest red, or, if made originally more opaque, it may be employed in thinner or thicker laminæ. The experiments may be successfully made in both ways: in order, however, to see the bright blue colour, the glass should neither be too opaque nor too thick. For, as it is quite natural that darkness must act weakly through the semi-transparent medium, so this medium, if too thick, soon approaches whiteness.
168.
Panes of glass throw a yellow light on objects through those parts where they happen to be semi-opaque, and these same parts appear blue if we look at a dark