But still more unfortunately, the fact supposed to have been observed, on which these ingenious speculations were based, has not only been called in question, but has been altogether negatived. More exact observations have shown that the supposed darkening of the floor of Plato is a mere optical illusion. When the sun has lately risen at that part of the moon, the ringed wall surrounding this great plain throws long shadows across the level surface. These shadows are absolutely black, like all the shadows on the moon. By contrast, therefore, the unshadowed part of the floor appears lighter than it really is; but the mountain ring which surrounds this dark grey plain is of light tint. So soon as the sun has passed high above the horizon of this region, the ring appears very brilliant compared with the dark plain which it surrounds; thus the plain appears by comparison even darker than it really is. As the long lunar afternoon advances, however, black shadows are again thrown athwart the floor, which therefore again appears by contrast lighter than it really is. All the apparent changes are such as might have been anticipated by anyone who considered how readily the eye is misled by effects of contrast. To base any argument in favour of a regular change in the floor of Plato on evidence such as this, would be as unwise as it would be to deduce inferences as to changes in the heat of water from experiments in which the heat was determined by the sensations experienced when the hands were successively immersed, one hand having previously been in water as hot as could be borne, the other in water as cold as could be borne. We know how readily these sensations would deceive us (if we trusted them) into the belief that the water had warmed notably during the short interval of time which had elapsed between the two immersions; for we know that if both hands were immersed at the same moment in lukewarm water, the water would appear cold to one hand and warm to the other.