PS YCHOLOGICAL LITERA TURK. 313 •curve symmetrically somewhat beyond the part which can be laid down from actual measurements; but that can safely be done, because the area taken in by this means is only a small fraction of the whole area, and has therefore only a small effect upon the result. The entire number of discernible brightnesses is in this way found to be about 660. These results differ very little for the eyes of different normal individuals. For Brodhun, who is green-blind, the

more »

... rences in sensitiveness to change of brightness were within the limit of probable error ; in color-tone his number of distinct sensations is about 140. This number does not differ much from that of the normal eye, for the reason that, although his spectrum does not •change beyond X 550, he has a keener sense for change of color-tone in the blue-green region than has the normal individual. In this connection it may be interesting to point out a most extraordinary statement which occurs in a book which is otherwise an admirable example of good scientific method-Havelock Ellis' Men and Women. It is there stated that Newton was able to distinguish seven different colors in the spectrum, but that most people since his time have only been able to see six. To how many of the 165 colors which are actually discernible by the ordinary person, it may be desirable to give a separate name for popular use is a question the answer to which may change from time to time ; but not to be able to distinguish between this question and the question of the number of colors which can be separated in sensation, is to have a mind which is abnormally incapable of drawing distinctions. The question of the number of differences of saturation which are just perceptible for different colors has been treated by Aubert, Woinow and J. J. MUller ; but, of course, the subject lacks all interest when the investigation is not made with spectral lights. To determine by a corresponding method with the one here considered, the total number of distinguishable sensations caused by light of all kinds would require in effect the integration by mechanical means of a solid body in space of four dimensions-what it is not beyond the powers of the mathematician to accomplish.