112 PAUL S. WELCH. 



The larvae under observation showed a definite response to 

 currents of low rapidity. When placed in the open water of a 

 stream having a current of 10 feet per minute they almost in- 

 variably swam down stream. This accounts for the prevalence, 

 at the mouth of Bessey Creek, of infested leaves showing attack 

 only at the junction of the petiole with the leaf an indication 

 of the work of the older larvae. 



Floating. Instances were observed where larvae were trans- 

 ported from place to place by floating down stream on detached 

 leaves and other floating objects. Occasionally a swimming 

 larva became passive and, since it remained on the surface, it was 

 carried along with the current. 



Craiding. Change of position on the leaf is accomplished by 

 crawling and, in case lily leaves are contiguous, means is thus 

 provided for the passage from one leaf to another. When in 

 search of a new leaf the larva usually explores the periphery of 

 the leaf, taking to the water and swimming only when it is not 

 possible to reach another leaf from the supporting one. The 

 explorations at the edge of the leaf consist in projecting one- 

 third to one-half of the body over the edge and swinging it about 

 as if seeking another support. If nothing is within reach, the 

 body is withdrawn and the same performance repeated at another 

 place. If, however, an object is within reach the larva immedi- 

 ately crawls upon it. 



In cases where larvae were deserting the old leaves it was not 

 always possible to account for the departure. The two most 

 common incentives apparently were (i) the presence of other 

 larvae on the same leaf, since only one larva can occupy the burrow 

 in the petiole, and (2) the deterioration of the lily leaf and petiole 

 which results in an unfitting of the plant for the larva. 



ENEMIES. 



These larvae were eagerly snapped up by sunfish whenever 

 there was opportunity. They were comparatively safe in the 

 burrows but whenever they left the lily leaves and swam in open 

 water the mortality was often high. Specimens removed from 

 the petioles and thrown out into open water were soon discovered 

 and captured. The undulatory swimming motion appeared to be 



