552 W. H. LONGLEY 



How largely independent of activity, Estates of mind,' and 

 internal stimuli the color phases are, may best be illustrated by 

 reference to specific instances. 



On July 28, 1915, I broke open a large, long-spined sea urchin 

 ( Centre chinus) over clear sand in a small bight surrounded by 

 large coral heads, which were at no place within 8 or 10 feet of the 

 bait. A small hogfish about 10 inches long came and circled 

 about many times, for the broken test was lying spines upper- 

 most, so that it could get nothing, although smaller fishes could 

 go under and feed from the inside. Finally it found a detached 

 spine which it picked up and masticated base foremost. It 

 remained consistently in the gray phase throughout the whole 

 period during which it was under observation (fig. 6). 



I next moved to a dense gorgonian patch 3 or 4 feet in diam- 

 eter, and placed food beneath the brown branches of the clustered 

 colonies. Within a few minutes a larger hogfish, 18 or 20 inches 

 in length, went in under them and displayed an almost uniform, 

 brown color, or, as it moved about, replaced this by a mottled 

 pattern of rich reddish brown and gray (fig. 7). When it was 

 driven out over the open bottom, it turned gray, and swam 

 off in that phase mth its mouth full of food. It went under 

 gorgonians nearby and turned brown for the second time and 

 finally swam away low down over mixed bottom in its mottled 

 livery. Coming back after a little, it approached the original 

 station, ate and swam about in brown phases. Then a broken 

 Centrechinus was placed about 10 feet away on bare sandy 

 bottom ; the fish was driven out to it with a long-handled dipnet, 

 and assumed its gray phase in the open. The food was moved 

 back and forth a number of times, and color and pattern were 

 changed regularly with each significant change of position, but 

 neither variation in activity, nor any incipient alarm engendered 

 by the repeated seizure and removal of its food seemed to have 

 any effect in modifying the fish's appearance. The facts re- 

 garding the color changes of the other species mentioned in 

 table 1 are perfectly comparable with these and seem to point 

 to only one rational conclusion. 



Whether or not a familiar object minus its color might by 

 association induce a color change appropriate to its normal but 



