[263.]

The extreme red edge will identify itself with the vermilion colour of the square, which will thus appear a little elongated in this direction; while the yellow border immediately underneath it only gives the red surface a more brilliant appearance, and is not distinguished without attentive observation.

[264.]

On the other hand the red edge and yellow border are heterogeneous with the blue square; a dull red appears at the edge, and a dull green mingles with the figure, and thus the blue square seems, at a hasty glance, to be comparatively diminished on this side.

[265.]

At the lower outline of the two squares a blue edge and a violet border will appear, and will produce the contrary effect; for the blue edge, which is heterogeneous with the warm red surface, will vitiate it and produce a neutral colour, so that the red on this side appears comparatively reduced and driven upwards, and the violet border on the black is scarcely perceptible.

[266.]

On the other hand, the blue apparent edge will identify itself with the blue square, and not only not reduce, but extend it. The blue edge and even the violet border next it have the apparent effect of increasing the surface, and elongating it in that direction.

[267.]

The effect of homogeneous and heterogeneous edges, as I have now minutely described it, is so powerful and singular that the two squares at the first glance seem pushed out of their relative horizontal position and moved in opposite directions, the red upwards, the blue downwards. But no one who is accustomed to observe experiments in a certain succession, and respectively to connect and trace them, will suffer himself to be deceived by such an unreal effect.