9. Let the two halves of a sheet of unglazed paper, about twenty inches square, be coloured clear grey, by a mixture of chalk and black; fix them, in any way, upon a piece of unbleached linen, placed across a window, at the distance of twelve inches asunder. Take two halves of another piece of similar paper B Bʹ, but of a darker grey, and coloured with the same substances. Fix A next to B and place Bʹ twelve inches from B. (See Fig. 1.) [Illustration: PLATE I.] Upon looking at the four half-sheets for a few seconds, it will be seen that A contiguous to B will be lighter than Aʹ, while B will seem darker than Bʹ. 10. It is easy to demonstrate that the modification is not equally intense over the whole of the surfaces A Aʹ and B Bʹ, but that it becomes gradually feebler from the line of contact. This may be proved by placing a card, so cut, that O and P may each present three grey stripes, as shown in Fig. 2, Plate 1. The stripes 1 1 are more modified than the stripes 2 2, and these are more so than the stripes 3 3. However, in order that this modification may be effected, it is not absolutely necessary that O and P should touch; for if the stripes 1 1 be covered, the stripes 2 2, 3 3 will be modified. 11. The following experiment, which is simply the result of the two preceding (9 and 10), is well suited to demonstrate the extent of contrast of tone. Upon a sheet of cardboard divided into ten stripes, each about a quarter of an inch broad, lay a uniform tint of Indian ink. As soon as it is dry, lay a second tint on all the stripes except the first. As soon as the second is dry, lay a third on all the stripes except the first and second, and so on of all the rest, so as to have ten flat tints, gradually increasing in depth from the first to the last. (See Fig. 3.) If ten strips of paper of the same grey, but each of a different tone, be laid upon a cardboard, in the preceding gradation, it will serve the same purpose. Upon looking at the cardboard, it will be seen that the strips, instead of presenting flat tints, will each appear of a tone diminishing in intensity from the edge _a a_ to the edge _b b_. In the stripe 1, the contrast is produced simply by the contiguity of the edge _b b_ with the edge _a a_ of the stripe 2; in the stripe 10, it is simply by the contiguity of the edge _a a_ with the edge _b b_ of the stripe 9. But in each of the intermediate stripes 2, 3, 4, 5, 6, 7, 8, 9, the contrast is produced by a double cause; partly by the contiguity of the edge _a a_ with the edge _b b_ of the stripe which precedes it, partly by the contiguity of the edge _b b_ with the edge _a a_ of the darker tint which follows it. The first cause tends to raise the tone of the half of the intermediate stripe; while the second cause tends to lower the tone of the other half of the same stripe.