But after these are all accounted for, there is still a real difference.

Two stars of the class known as double stars, that is, so little separated that considerable optical power is necessary to divide them, show these different tints very nicely in the same field of the telescope.
"Then there comes in the chance that the colors are complementary; that the eye, fatigued by a brilliant red in the principal star, gives to the companion the color which would make up white light.

This happens sometimes; but beyond this the reare innumerable cases of finely contrasted colors which are not complementary, but which show a real difference of light in the stars; resulting, perhaps, from distance,--for some colors travel farther than others, and all colors differ in their order of march,--perhaps from chemical differences.
"Single blue or green stars are never seen; they are always given as the smaller companion of a pair.
"Out of several hundred observed by Mr.Bishop, forty-five have small companions of a bluish, or greenish, or purplish color.

Almost all of these are stars of the eighth to tenth magnitude; only once are both seen blue, and only in one case is the large one blue.

In almost every case the large star is yellow.