INTELLIGENCE OF EARTHWORMS 



339 



experimenter it seemed extremely improbable that the worm 

 had acquired a perfect direction-habit as a result of only 20 

 trials in the T. For the present it must suffice to state that no 

 satisfactory explanation for the correct series was revealed by 

 check experiments. The worm apparently profited by ex- 

 perience with surprising quickness. 



Between October 12 and November 6, as may be noted in 

 the accompanying summary of experiments with No. 2, the 

 worm was given 200 trials in series of 10 each. The results 

 were variable. Thus we find that the perfect record for October 

 17 was followed by a record of contact with the sandpaper in 

 three trials and with the salt in one trial on October 18, and 

 that, in turn, by a perfect record on October 19. 



Summary of Experiments with Worm No. 2 



The training on the basis of 10 trials per day (with omission 

 of experimentation occasionally) failed to establish a definitely 

 predictable mode of response (perfectly dependable habit) in 

 No. 2, so the number of trials per series was increased to 20, 

 beginning on November 7. Thus the experimentation was 

 continued until November 17, when No. 2 was given its four- 

 hundredth trial. The results varied extremely from day to 

 day. There appeared to be "good" and "bad" days. The 

 data of tables 2 and 3 are offered in support of this statement. 



