_ COLOR CHANGES AND ADAPTATION IN FISHES. 235 



pronounced immediately after stimulation. They become less conspicuous rapidly but it usually re- 

 quires several miniites for them to disappear completely. They usually also appear when the animal 

 is feeding. 



Plates XXV, XXVI. — Adaptation on artificial black and white backgrounds, showing that 

 the patterns on them is not actually reproduced in the skin. 



Plate XXV. 



Fig. 23. Individual (B) on black and white background (5 by 10 mm. rectangular), long axis of the 

 fish parallel with short axis of rectangles, continuously August 21, from 10 a. va. to 2.30 p. m., and 

 August 22 from 11 to 11.30 a. m. No change in pattern could be detected from that assumed when the 

 fish was free on this background. 



Fig. 24. Individual (B) on same background, long a,xis of fish continuously parallel with long axis 

 of rectangles, August 22 from 3.10 to 5 p. m., and August 23 from 6 to 9.15 a. m. 



Fig. 25. Individual (B) on the same background continuously rotated on a disk under the crystal- 

 lizing dish August 21 from 2.30 to 4.45 p. m. Note that the pattern is essentially the same as that shown 

 in figures 23 and 24, although the apparent forms of the figures were continuously changing from long 

 narrow to short and broad rectangles and vice versa. 



Fig. 26. Individual (B) on black and white background (i sq. cm. circles, area of white equal to 

 that of black) from August 7, 11.30 a. m., to August 8, 10.15 ^- ™- 



Plate XXVI. 



Fig. 27. Individual (B) on black and white background (i sq. cm. circles, area of black equal to 

 that of white) from July 30, 3 p. m., to July 31, 11.40 a. m. 



Fig. 28. Individual (B) on black and white background (stripes i cm. wide) from August 4, 5.15 

 p. m., to August 5, 9.45 a. m. 



Fig. 29. Individual (B) on black and white background (stripes 5 mm, wide) from August 5, 

 4 p.m., to August 6, 4 p. m. Note that the pattern produced by stripes is essentially the same as that 

 produced by white circles on a black field (fig. 26) or black circles on a white field (fig. 27) or square or 

 elongated rectangles, foreshortened in either direction or constantly rotated so as to continuously change 

 the perspective. There is no evidence whatever indicating an actual reproduction in the skin of the 

 figures in the background. 



Fig. 30. P. albiguttus (A), 15 cm. long, on black and white background (5 mm. square) from July 

 23, 2 p. m., to July 25, 10 a. m. This specimen was tested on larger and on smaller squares and also on 

 stripes. The pattern produced in the skin was essentially the same as in individual (B), shown in the 

 preceding figures. 



Plates XXVII, XXVIII. — Photographs showing more in detail the relation between the 

 pattern in the i:kin and thai in the backgroutid. 



Plate XXVII. 



Fig. 31. P. albiguttus (}^), 14 cm. long, in large, shallow granite pan (variegated, dark blue and white) 

 from July 31 to August 14. Note the small white spots shown in irregular rows across the fish. 



Fig. 32. Individual (F) on black and white background (dots 2 mm. in diameter) from August 23 

 to 27, 10 a. m. The white spots in the irregular rows have become more conspicuous, but the pattern 

 is essentially the same, although the spacial interrelationship of the light and dark areas in the back- 

 ground is very different. This is very clearly shown in the enlarged photographs reproduced in figures 

 35. 36. 38. and 39. 



Fig. 33. Individual (C) same specimen shown in figure 19 in the granite pan represented in 

 figure 31 from August i, 9 a. m., to August 14, after having been on various backgrounds during the 

 preceding week. All the individuals in the granite pan after being fully adjusted were so inconspicuous 

 that it required considerable attention to see them. (See autochrome fig. 66.) 



