236 BULLETIN OF THE BUREAU OF FISHERIES. 



Fig. 34. Individual (C) on black and white background (2 mm. squares) from August 25, 8 a. m., 

 to August 27, 10.18 a. m. Note that the patterns produced by these two ver)' different backgrounds 

 are essentially the same. 



Plate XXVIII. 



Fig. 35. A small section of the background shown in figure 31, enlarged. X 9. 



Fig. 36. Enlargement of a small section of figure 31 taken slightly above the ventro-anterior ocellus. 



X9- 



Fig. 37. Enlargement of a section taken from the same relative place in figure 32. X 9. 



Fig. 38. A small section of the background shown in figure 32, enlarged. X 9. Note that the 

 details of the pattern in the skin (fig. 36, 37) assumed on these two different backgrounds (iig. 33, 38) 

 are strikingly similar. Practically all of the dark and the light areas found in one are also found in the 

 other, and the forms of these are, respectively, very nearly the same, although in the two backgrounds 

 the dark and the light areas differ greatly, both in form and in spacial arrangement. The same is true 

 with reference to individual (C), figures 33, 34, 39, and 40. On neither background is there any evi- 

 dence of an actual reproduction in the skin of the configuration in the background. 



Fig. 39. Enlargement of a small section of figure 33 located in the same relative position as that 

 reproduced in figures 36 and 37. X 9. 



Fig. 40. Enlargement of the same relative section of figure 34. X 9. Note that the patterns shown 

 in figures 39 and 40 are essentially the same, although the backgrounds on which they were produced 

 differ greatly. Note also that the patterns shown in figures 36 and 39 differ considerably in detail, 

 although produced by the same background but in different individuals. While the large features in 

 the pattern of the skin of all individualsof this species are essentially the same, there is sufficient specific 

 individuality in the smaller features to make it possible to recognize every individual by a thorough 

 study of the pattern found on a square millimeter or less of any part of the pigmented surface. 



Plates XXIX, XXX. — The photographs on these plates show that simulation of the back- 

 groutid is dependent upon stimuli received through the eyes a7id that the stimuli from 

 each eye are distributed over the entire pigmented surface. 



Plate XXIX. 



Fig. 41. Individual (B) (same specimen shown in many other figures), on black and white back- 

 ground (2 mm. squares) from July 29, 11 a. m., to July 30, 3 p. m. 



Fig. 42. Individual (B) on black and white background (circles i sq. cm.) from July 30, 3 p. m., 

 to July 31, 11.40 a. m. 



Fig. 43. Individual (B), one eye continuously on a coarse and the other on a fine-grained back- 

 ground, on August 29, 8 to 11.20 a. m. No further appreciable change occurred in the pattern after 

 8.30 a. m. Thus it is clear that the fish was fully adapted long before the photograph was taken. This 

 rapid adaptation was no doubt due to the fact that this animal had been used in the study of patterns 

 produced by different backgrounds almost continuously for a month or more. Note that the pattern is 

 the same over the entire surface and that it is intermediate in texture between those shown in figtu-es 

 41 and 42. It appears to be a sort of superimposition of these two patterns. 



Plate XXX. 



Fig. 44. P. albiguttus (G), 22 cm. long, head continuously on white and tail on black, on August 

 21, 2.30 to 4.55 p. m. This specimen was taken from a black aquarium and was fully adapted to black 

 in the beginning of the experiment. 



Fig. 45. Individual (G), one eye on white the other on black continuously, on August 22, 10 to 11.55 

 a. m. The fish was nearly maximum black at 10 a. m. 



Fig. 46. Individual (G), continuously with head on white, body on black, on August 23, 6 to 9.20 

 a. m. 



Fig. 47. Individual (G), head on black, body on white, on August 23, q.20 a. m. to 12 m. 



