576
L. T. Troland
[J.O.S.A. & R.S.I, VI
to heterogeneous, and wave-length of the former, are adjusted until a match is obtained (66). The measured color is then designated in terms of (1) luminosity, (2) dominant wave-length, and (3) “per cent. hue.” It is apparently not essential, although advisable, in this method that the heterogeneous radiation
should have an invariable constitution, e.g., that of “normal gray light,” so long as it always produces an achromatic color with the given observer. Results obtained by this method are subject to serious errors with deviations, of common occurrence, of the observers’ visual systems from the average normal. It is