ASSOCIATION AND COLOR DISCRIMINATION 493 



color, mudminnows, and even more especially the sticklebacks 

 were confused by the reversal of the food and paper in relation 

 to the colors. The reversed combination seemed to be more 

 difficult to master than the original one. 



GENERAL CONCLUSIONS FROM THE EXPERIMENTS DESCRIBED 



Compared with land vertebrates, fishes live in a constant 

 medium where little premium is put upon sensory specialization. 

 None of the sense organs have a marked degree of perfection. 

 The behavior of fishes is stereotyped. Their associations are 

 simple, few in number, and are not readily modified, though 

 they are often fairly permanent when once formed. ' Learning' 

 seems to consist for the most part in the gradual elimination of 

 useless movements and the establishing of useful ones. 



There is little evidence of the formation of new types of move- 

 ments. A considerable period of training in concentrating the 

 attention upon the object in view seems to be necessary before 

 associations can be formed. 



That instincts may be inhibited is demonstrated in the color 

 experiments where the fishes gradually overcame the impulse 

 to leap out after paper which resembles their food in light of 

 a certain color when they had thus been several times unsuc- 

 cessful in obtaining food. Stickleback no. 57, after being re- 

 peatedly offered gray paper alternately with pieces of liver,, 

 refused to spring out after the paper. 



Fishes do not seem to be capable of anything which might 

 properly be called a concept, nor to exhibit memory in the sense 

 of having ideas about absent objects. They do react more 

 readily to present objects with which they have had experience 

 in the past, particularly if this experience has been several times 

 repeated. There is nothing to indicate that they are able to 

 recall an image of their past experience. After some repetition 

 of a reaction they form associations and habits. 



Imitation in fishes is well described by Hobhouse's definition 

 of feensory-motor imitation: ''the perception of what is done 

 discharges a .motor impulse to do the same thing quite apart 

 from any purpose to be served by doing it." Sticklebacks which 



THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 27, NO. 4 



