Using apparatus designed by Priest and Gibson, the writer has determined sensibility to wave-length difference for his right eye: (1) For stimuli of unit purity (spectral light) from 450 to 645 mµ (2) For stimuli consisting of artificial noon sunlight plus homogeneous light of some selected wave-length, the wave-lengths being 455, 470, 481.5, 493, 530, 580, 635 mµ, and the purities ranging from unity to a few percent. (3) For stimuli consisting of homogeneous light of wave-length 455 plus some one of heterogeneous stimuli specified as follows: (a) equal energy, (b) color temperature 2570°K, (c) color temperature over 24,000°K. (4) For a stimulus consisting of homogeneous light of wave-length 530 mµ plus heterogeneous light of color temperature over 24,000 °K. The results of (1) are concordant with previous determinations by others. The least perceptible difference (LPD) in wavelength for (2) shows, for wave-lengths 635, 580, 530 and 493mµat first a slow and then, near zero purity, a rapid increase as purity is decreased. For wave-length 455 mµ with decreasing purity, the LPD at first increases and then decreases to a pronounced minimum at about fifteen percent purity, increasing rapidly beyond this. Wavelengths 470 and 481.5 mµ show a similar but less marked effect. The experiments noted under (3) and (4) were undertaken in order to study the effect of the spectral distribution of the heterogeneous stimulus in modifying the peculiar results just described for wave-lengths less than 482 mµ. It is shown that the form of the curve showing LPD as a function of purity depends upon the spectral distribution of the heterogeneous stimulus. Two hypotheses are proposed to account for the peculiar results for short wave-lengths.
E. P. T. TYNDALL, "Chromaticity Sensibility to Wave-Length Difference as a Function of Purity," J. Opt. Soc. Am. 23, 15-23 (1933)