PHYSIOLOGICAL EFFECTS 123 



and would continue to jump even when the wire 

 loop held only tasteless filter paper in place of the 

 usual worm. 



The group behaved thus: just as a fish would 

 be moving up preparatory to jumping for the food, 

 another fish, or more than one would approach, all 

 making the same preliminary motions which if 

 they had been alone would have led to a jump for 

 food. When they came near together, however, 

 they turned their attention to each other rather 

 than to the food and pushed or gave other indi- 

 cations of a combat and so did not jump. Jump- 

 ing under these conditions rarely occurred unless 

 the fish chanced to be relatively alone. It is 

 apparent that even with fishes, the effects of 

 numbers present on the rate of learning depends 

 in part on the problem set and there is good 

 reason for thinking that it also depends on the 

 sort of fishes that is being tested. 



There is good evidence also that groups of these 

 fishes eat more than would the same individuals if 

 they were isolated. These effects are apparently 

 the result of a certain more or less nebulous sort 

 of group solidarity which reaches a much higher 

 state of development in the schooling species. 

 Incidentally it is worth recording that with the 

 black catfishes that form close aggregations when 

 they are young, the grouped fishes consume 



