A colorimeter was used in which the adjustable primaries can be extinguished separately, and in which the comparison stimulus is synthesized by an equivalent optical system, thus assuring a physical as well as a visual match. After a few matches were obtained in the customary manner, two of the adjustable primaries were left undisturbed and the third was merely flashed intermittently, five seconds off, one-half second on. After observation of one or two cycles, a great difference is apparent between the two halves of the field during the half-second when all three primaries are on. When the red primary is intermittent, the lower semicircle appears bluish-green for five seconds, and much too reddish to match the upper half during the brief flash. The observers then readjusted the primaries until the two halves of the field appeared to match during the flash. Each primary was made intermittent, in turn, and also each pair of primaries, giving measurements of the effect for six different directions around each comparison color. Achromatic and moderately saturated red, green, and purple comparison colors were investigated. The intermittent stimulus which appears to match the steady stimulus lies very nearly on the straight line in the chromaticity diagram connecting the latter with the adapting stimulus.
DAVID L. MACADAM, "Measurement of the Influence of Local Adaptation on Color Matching," J. Opt. Soc. Am. 39, 454-459 (1949)