LOCOMOTIONS OF SURFACE-FEEDING CATERPILLARS. 139 



the arms of the maze were arranged in a variety of ways. In 

 the latter experiments with them and in the experiments with 

 all others the maze was arranged as in Fig. 2 ; but, the bands G 

 and H were used only on special occasions. In some of the 

 experiments the plain maze was used ; in others one or two odorous 

 bands of filter paper were attached to certain portions of the 

 maze. In experimenting one or more caterpillars were placed 

 on the ground at the foot of the post, or else upon the top of the 

 post, or on some intermediate portion of the maze. 



When placed on the ground at the foot of the plain maze, the 

 tent-caterpillar almost invariably letisimulated. In the first 

 few experiments with the tent-caterpillars, on recovering from the 

 feint, the larva moved toward and up the post. If its initial 

 movements were away from the post, after a few steps it turned 

 about and moved towards the maze. This striking behavior 

 recalled some experiments by which H. B. Weiss 1 discovered 

 that certain water insects when turned loose several yards from 

 their home, would immediately go directly to the pond. It 

 looked as though these caterpillars were being mechanically led 

 to the post. To test the matter, a series of experiments, extend- 

 ing over several days, were performed. In these experiments 

 precautions were taken to see that the heads of the caterpillars 

 pointed in a variety of directions. In some cases the majority 

 of the larvae moved towards the post, in others the majority 

 moved away from the post, in yet others about as many went 

 away from the post as toward it. Evidently the behavior noticed 

 in the first experiments was merely a coincidence, the uniformity 

 of which may have been due to some factor which escaped ob- 

 servation. 



Once upon the maze, each caterpillar displayed marked indi- 

 viduality. It might continue all the way to the top of the post 

 without mounting any of the arms; it might, on its way up, ex- 

 plore a few or all of the arms; or it might turn about before reach- 

 ing the top and return to the ground. At intervals it was sure 

 to pause and reach upward and outward with the front end of its 

 body and move it, in a wabbly manner, from side to side. [For 



1 Notes on the Positive Hydrotropism of Gerrls marginatus and Dineutes assimilis. 

 Canad. Entom., Vol. XLVL, pp. 269-271. 



