550 W. H. LONGLEY 



no evidence that in nature any other factor than the color and 

 shade of the surroundings exercises notable control over those 

 changes in coloration which occur by day, modifications in 

 pattern excepted. 



Fishes in tanks, it is true, often show color phases which baffle 

 interpretation, but when they are unconfined this is not commonly 

 so. This suggests that in the former case the behavior is not 

 entirely normal, and two sets of observations lend some support 

 to this conclusion. In the first place, the whole behavior of 

 some species seems to be modified by restraint, to which they 

 do not become accustomed in weeks of confinement. It is 

 also known that in some animals processes underlying growth, 

 secretion, etc., are directly affected by comparatively slight 

 disturbances in their environment. The rate of growth of rats, 

 for example, is distinctly modified, if unfamiliar operations are 

 carried on by strangers in or about the creatures' quarters. That 

 is to say, modified color reactions may be anticipated wdth 

 reason in fishes in confinement, since they sometimes show its 

 effects plainly in other ways, and since, in addition, in other 

 animals processes apparently no more complex than those con- 

 cerned in adaptive color changes are inhibited by no more ein- 

 dent stimuli, and without more marked departure from normal 

 behavior in other respects. 



But wholly apart from complications which might originate 

 in disturbance of organic function through confinement, there 

 is a possible source of error in experimentation upon fishes 

 under such conditions. This lies in the fact that the color- 

 complex in which most captive fishes find themselves, only 

 remotely resembles that in which they normally occur, and to 

 which they are able to adjust themselves accurately. But 

 when to match a given background color is wholly impossible, 

 and there is bound to be maladjustment in any case, much greater 

 variation is to be expected in one fish at different times, and in 

 different ones at the same time, than if a stereotyped response 

 well within the creatures' capacity were demanded. Finally, 

 exhibition tanks such as those in the New York aquarium 

 unquestionably fall under the head of mixed en\dronments, in 



