IT has been shown that the reactions of the xanthophores (yellow pigment cells) in Fundulus heteroclitus are independent of the responses of the melanophoresblack pigment cells (Connolly, 1925; Fries, 1931). For example, both the xanthophores and the melanlophores are contracted in a fish which has been kept for some time in an illuminated white vessel, and both are expanded in one kept in an illuminated black vessel. In an animal adapted to a blue background, however, the melanophores are mainly expanded and the xanthophores contracted, while on a yellow background, the xanthophores are expanded and the melanophores are contracted. Thus this inclependent behavior of the yellow and black pigment cells to yellow and blue backgrounds led Fries to conclude that the two types of chromatophores were controlled by separate sets of neurons. In the following year, Parker (1932) formulated the neurohumoral hypothesis for the control of chromatophores on the basis of the responses of denervated melanophores. Such responses were first clearly described by Fries (1931), who declared that they must be governed by humoral substances. Mills (1932), who studied in detail the same responses, concluded also that humoral factors were responsible for the activity of enervated melanophores and stated further that these humoral substances were secretions from intact nerve terminals surrounding 1 These experiments were performed at the Marine Biological Laboratory, Woods Hole, Massachusetts, during the summers of 1934 and 1935. For the privilege of working at the laboratory in 1935, I am especially indebted to its director, Dr. M. H. Jacobs.