Abstract
In the present expt. 2 equivalent shadows, one encircled by a line and the other unencircled, were used, and measurements of brightness difference were made by means of rotating discs inserted within the shadows. It was found that the characteristic difference in brightness was present when neither shadow was perceived as shadow; that this difference persisted even after full explanation of the exptl. conditions to the subjects; that the difference persisted even when the subjects were instructed to disregard the phenomenal difference in brightness and equate for the physical properties of the situation. It is argued that the impression of anomalous illumination is not a necessary condition of some color-constancy phenomena; that some of these phenomena are more properly referable to conditions of field-organization; and that the concept of constancy is now of dubious value in the study of the determinants of color.