224 BULLETIN OF THE BUREAU OF FISHERIES. 



The only results which do not support this statement were obtained with Ancylopsetta (B). 

 This individual, as the table shows, turned frequently and persistently toward a given 

 shade for a considerable period of time and then suddenly stopped and turned as per- 

 sistently in the opposite direction. The results obtained with all of the other individuals 

 are, however, so consistent that not much weight can be placed on those obtained with 

 this one. The tabulated results also show that, especially in Ancylopsetta, there is a 

 tendency to turn toward the ventral surface. Thus it is evident that these creatures 

 select, to some degree, that background on which they are least conspicuous. Whether 

 or not this selection has reference to color and pattern as well as to shade remains to be 

 ascertained. 



COLORED BACKGROUND. 



After this section of the work was completed opportunity presented itself to extend 

 this investigation so as to include color. This study was carried on in the same way as 

 that described above, bearing on the selection of backgrounds on the basis of difference 

 in shade. The details regarding the experiments and the results obtained are presented 

 under color-vision in the section on vision. By referring to table iv in that section, it 

 will be seen that the individuals adapted to blue and to green showed a strong tendency 

 to go toward the background with which they harmonized in color, that those adapted 

 to yellow showed very little of this tendency, and that those adapted to red showed none. 

 Thus it would seem that there is some evidence indicating that color is a factor in the 

 selection of backgrounds. The whole subject, however, is in need of further investiga- 

 tion before hard and fast conclusions can be reached. 



