COLORATION OF FISHES LONGLEY. 485 



Some patterns which are not changeable seem to be correlated 

 with definite habits. In the case of others which are changeable there 

 is conclusive evidence that they are displayed under specific condi- 

 tions. One may almost dare state it as a law, that when any species 

 has alternative patterns of longitudinal stripes (or self-color) and 

 transverse bands, the former is shown when the fish that displays 

 it is in motion, while the latter tends strongly to appear whenever 

 it comes to rest. 1 



To be able to say so much is decidedly encouraging. The color 

 patterns of fishes are more than variegated pigment patches com- 

 pounded at random; they possess biological significance. Whoever 

 reads the riddle of their organization, and comprehends the order 

 of their changes, will discover the essential principles of a natural 

 system of camouflage. In the changeable colors, and even more in 

 the changeable patterns of fishes, he should find, too, a delicate physi- 

 ological indicator, through changes in which variation in psychic 

 states should be observed to advantage. 



1 Since this was written it has been observed that like many fishes the squid (Sepia 

 sp.) when at rest in the water is transversely banded, but replaces these bands by 

 longitudinal stripes when it begins to move. 



136050°— 20 32 



