we annex to the number of each paragraph the number of the former corresponding one. But we set out by observing generally that the reader must consult the plates, that the scientific investigator must be familiar with the apparatus in order that the twin-phenomena in one mode or the other may be placed before them.
XXI.
REFRACTION WITHOUT THE APPEARANCE OF COLOUR.
306 (195, 196).
That refraction may exhibit its effects without producing an appearance of colour, is not to be demonstrated so perfectly in objective as in subjective experiments. We have, it is true, unlimited spaces which we can look at through the prism, and thus convince ourselves that no colour appears where there is no boundary; but we have no unlimited source of light which we can cause to act through the