INTELLIGENCE IN LOWER VERTEBRATES 221 
one of the latter was placed in the compartment of the larger 
fish. It was treated at first much as before, but after a little 
“the perch became more demonstrative toward it, but re- 
strained themselves from striking, so long as it quietly 
avoided them.” When the minnow made a quick movement 
the perch were more apt to dart after it and on one occa- 
sion it had to.be rescued. After a time the pursuit of the 
minnows became less eager and although they were followed 
the instinct to dart at them was generally inhibited. 
The formation of an association in relation to the glass 
partition was shown by the fact that when the partition was 
removed the fish behaved much as if it still remained. ‘The 
male swam out to the place, stopped, made little bumps 
forward as if expectng to strike the usual obstruction, and 
was plainly at a loss. He then turned and swam down as 
if following the glass.” Often the fish would swim to the 
mark where the plate had been and then turn back. 
These experiments showed that the natural instinct of the 
perch to charge after the minnow was inhibited, although not 
completely so, and that the fish came to associate the appear- 
ance of a certain region of the aquarium with the experience 
of being bumped. Pieron found that a cyprinoid fish, 
Carassius auratus, would snap at live worms placed in a glass 
tube, but the number of attacks in a given time gradually 
grew less on the following days. The glass tube was placed 
in the water for twenty-seven minutes each time. In the 
first experiment the fish made 117 attacks, on the three 
successive days the attacks were 58, 38, and 25 in number. 
Several days later the number of attacks had diminished 
still more. The tube was then left in the aquarium and the 
fish soon came to ignore it entirely, and even refused to eat 
the worms when they were placed free in the water. When 
the number of attacks on the tube diminished they could be 
