The yellow and red squares, on the other hand, are comparatively reduced by the heterogeneous edges in this position of the figures, and their colours are, to a certain extent, vitiated. The blue edge in both is almost invisible. The violet border appears as a beautiful peach-blossom hue on the red, as a very pale colour of the same kind on the yellow; both the lower edges are green; dull on the red, vivid on the yellow; the violet border is but faintly perceptible under the red, but is more apparent under the yellow.
Every inquirer should make it a point to be thoroughly acquainted with all the appearances here adduced, and not consider it irksome to follow out a single phenomenon through so many modifying circumstances. These experiments, it is true, may be multiplied to infinity by differently coloured figures, upon and between differently coloured grounds. Under all such circumstances, however, it will be evident to every attentive observer that coloured squares only appear relatively altered, or elongated, or reduced by the prism, because an addition of homogeneous or heterogeneous edges produces an illusion. The inquirer will now be enabled to do away with this illusion if he has the patience to go through the experiments one after the other, always comparing the effects together, and satisfying himself of their correspondence.
Experiments with coloured objects might have been contrived in various ways: why they have been exhibited precisely in the above mode, and with so much minuteness, will be seen hereafter. The phenomena, although formerly not unknown, were much misunderstood; and it was necessary to investigate them thoroughly to render some portions of our intended historical view clearer.
In conclusion, we will mention a contrivance by means of which our scientific readers may be enabled to see these appearances distinctly at one view, and even in their greatest splendour. Cut in a piece of pasteboard five perfectly similar square openings of about an inch, next each other, exactly in a horizontal line: behind these openings place five coloured glasses in the natural order, orange, yellow, green, blue, violet. Let the series thus adjusted be fastened in an opening of the camera obscura, so that the bright sky may be seen through the squares, or that the sun may shine on them; they will thus appear very powerfully coloured. Let the spectator now examine them through the prism, and observe the appearances, already familiar by the foregoing experiments, with coloured objects, namely, the partly assisting, partly neutralising effects of the edges and borders, and the consequent apparent elongation or reduction of the coloured squares with reference to the horizontal line. The results witnessed by the observer in this case, entirely correspond with those in the cases before analysed; we do not, therefore, go through them again in detail, especially as we shall find frequent occasions hereafter to return to the subject.—[Note P.]
[1] [Plate 3], fig. 1. The author always recommends making the experiments on an increased scale, in order to see the prismatic effects distinctly.