178 BULLETIN OF THE BUREAU OF FISHERIES. 



when mechanically stimulated or when some object, as, e. g., another fish comes within 

 the range of vision. In general, the difference in the shade of the various regions of the 

 skin becomes, under such conditions, much more pronounced. Such changes are usuallj' 

 referred to as psychic. They may possibly be protective by way of warning off enemies. 



(2) Many authors have observed that during the breeding season the colors in 

 various fishes, especially the males, become much brighter and the entire surface 

 becomes more conspicuous. These changes are generally supposed to be functional 

 in sexual selection. Hess (191 3), however, has recently offered serious objections to 

 this view. He maintains that in some species which spawn at a depth of 60 meters 

 the males become brilliantly colored during the breeding season, and that red and 

 yellow predominate, although at that depth these colors do not exist, for all of the 

 longer waves of light have been absorbed. 



(3) It is held by many that fishes simulate the background, not only in shade but 

 also in color. Stark (1830), De Vescovi (1886), Van Rynberk (1906), Frisch (1912), 

 Secerov (1913), and others have come to this conclusion regarding a considerable number 

 of genera (Blennius, Gobius, Labrus, Crenilabrtis, Solea, Rhomboidichthys, Nemachilus, 

 Phoxinus, etc.). Van Rynberk (1906, p. 549) says that particularly in Rhomboidichthys 

 the skin assumes a color strikingly similar to that of the background, "exquisite anpas- 

 sung an den Farbenton des Bodens," and Frisch is equally positive in asserting that 

 there is color adaptation in Phoxinus. 



These conclusions, however, have not gone unchallenged. Schondorff, on the 

 basis of results obtained in experiments on trout of the same species studied by a num- 

 ber of the investigators referred to above, concludes that there are no adaptive color 

 changes in fishes. He writes (1903): "Wenn friiher einige Autoren wie z. B. Stark 

 (54) die Behauptung aufstellten, die Farbe der Fische richte sich nach deni Grunde des 

 Gefassess, in dem sie gehalten wiirden, so beruht dies auf einem Irrthum." Hess is even 

 more positive in his denunciation of the idea of adaptive coloration in fishes, although 

 his criticism is directed primarily toward the work of Frisch. He repeated and ex- 

 tended much of Frisch's work and found no evidence indicating production of color in 

 the skin in harmony with that of the environment. He says (1913, p. 439): "Frisch's 

 Angaben iiber die Farbenanpassung dcr Pfrille sind samtlich unrichtig. Die Farbe des 

 Grundes hat keinen Einfluss auf die Farbung der Pfrille." 



Not only has it been maintained that fishes simulate the background in brightness 

 or shade and in color, but it has also been asserted, particularly by Sumner, that the 

 pattern in the skin changes, so as to continually harmonize with that of the background. 

 Sumner found in experiments on some of the flatfishes, especially Rhomboidichthys, that 

 the size of the figures in the skin changes to correspond to a most remarkable degree 

 with those in the bottom on which they lie. This is illustrated in his excellent photo- 

 graphs (191 1, p. 481-505). A careful study of these photographs shows clearly that 

 if the light and dark areas in the background are small the figures in the skin are also 

 small, and if the areas are large the figures are correspondingly large. But the form 

 of the figures in the skin does not appear to depend upon that of the areas in the back- 

 ground. It is essentially the same in fishes on a bottom containing alternate black 

 and white squares as it is on one containing alternate black and white stripes, or black 

 spots on a white field, or white spots on a black field, or an irregular pattern as is found 

 in nature on gravel bottoms. Sumner says (p. 468) : " Squares, crossbands, circles, 



