A NOTE ON THE PSYCHOLOGY OF FISHES. 
EDWARD THORNDIKE. 
Numerous facts witness in a vague way to the ability of 
fishes to profit by experience and fit their behavior to situa- 
tions unprovided for by their innate nervous equipment. All 
the phenomena shown by fishes as a result of taming are, of 
course, of this sort. But such facts have not been exact 
enough to make clear the mental or nervous processes involved 
in such behavior, or simple enough to be available as demon- 
strations of such processes. It seemed desirable to obtain evi- 
dence which should demonstrate both the fact and the process 
of learning or intelligent activity in the case of fishes and 
demonstrate them so readily that any student could possess the 
evidence first-hand. 
Through the kindness of the officials of the United States 
Fish Commission at Woods Holl, especially of the director, 
Dr. Bumpus, I was able to test the efficiency of some simple 
experiments directed toward this end. The common Fun- 
dulus was chosen as a convenient subject, and also because 
of the neurological interest attaching to the formation of 
intelligent habits by a vertebrate whose fore-brain lacks a 
cortex, 
The fishes studied were kept in an aquarium (about 4 feet 
long by 2 feet wide, with a water depth of about g inches) rep- 
resented by Fig. 1. The space at one end, as represented by 
the lines in the figure, was shaded from the sun by a cover, 
and all food was dropped in at this end. Along each side of 
the aquarium were fastened simple pairs of cleats, allowing the 
experimenter to put across it partitions of wood, glass, or wire 
screening. One of these in position is shown in the figure by 
the dotted line. These partitions were made each with an 
opening, as shown in Fig. 2. If now we cause the fish to leave 
his shady corner and swim up to the sunny end by putting a 
973 
