AIoRSE, Reactions of the Larva of Tenehrio Molitor. 725 



counted. In order to be certain that the experiment was conducted 

 in as normal and natural a manner as possible, a long tin box, a meter 

 long and two decimeters square, blackened within and provided with 

 a window at either end, one of which being used by the observer 

 from a hood, was employed and the tube discarded. After covering 

 the ocelli, of one side (the right), and placing the animal within 

 the box with its long axis in the long axis of the box, facing the light, 

 the number of times it turned towards the right or the left was ob- 

 served and the results are recorded below : 



1 Right 14 Eight 



2 Left 15 Right 



3 Left 16 Left 



4 Left 17 Left 



5 Left 18 Right 



6 Right 19 Right 



7 Left 20 Left 



8..... Left 21 Right 



9 Left 22 Right 



10 Left 23 Left 



11 Left 24 Left 



12 Left 25 Left 



13 Right 



About one-third of the reactions, it will be seen, are dextral while 

 two-thirds are sinistral. It is evident from this that unilateral stim- 

 ulation has no orienting effect or else such an effect is compensated. 

 If there were such an effect, there would scarcely be such a large 

 proportion of reactions not amenable to the rule. Moreover, judg- 

 ing from the circling 'experiment, Avhere the circling is towards the 

 covered side, the data here given are exactly the reverse of what we 

 should expect. Similar indefinit^ness of reaction is to be seen in 

 the behavior in general. These random movements will now be 

 considered. 



Random movements. As the larva crawls away from the light, 

 the head is seen to sway from side to side rhythmically, as if per- 

 forming testing movements. The amplitude of the sway is such 



