HABITS OF THE LARVA OF BELLURA MELANOPYGA. 1 03 



In all cases the initial point of attack was at the top of the flower. 

 Both opened and unopened flowers were affected. Floating 

 rootstalks in the aquarium were always attacked and, if left 

 for any length of time, were ultimately excavated through and 

 through. This utilization of the substance of the rootstalk was 

 not always caused by famine. A number of specimens were 

 kept under conditions where rootstalk substance alone could be 

 secured as food and apparently thrived on it. It thus appears 

 that such materials may serve as food for these larvae. 



It sometimes happened that a fall in the level of the water 

 caused portions of the long petioles to be bent above the surface 

 of the water, or at least to lie on the surface. Whenever larvae 

 discovered these emergent petioles they attacked them, making 

 an entrance through the side of the petiole. 



Excrement. In the field the best mark of recognition of the 

 work of these larvae is the heap of excrement which accumulates 

 around the margin of the burrow on the upper side of the leaf. 

 As would be expected the quantity depends upon the length of 

 occupancy and the activity of the larva. In August it was a 

 common thing to find hundred of leaves with conspicuous heaps 

 of excreta around the hole on the upper surface. 



The excrement is always deposited outside of the burrow re- 

 gardless of the relation of its- opening to the level of the water. 

 Since the position of the larva is such that the posterior end is 

 towards the leaf, the excrement is always deposited by thrusting 

 the terminal somites out through the entrance of the burrow. 

 That the level of the water has nothing to do with determining 

 the place of deposition of excrement is shown by the following 

 observations: (i) In floating, partially submerged rootstalks 

 which had been tunneled by the larva? it often happened that 

 much of the upper part of the burrow contained no water but 

 invariably all of the waste matter was deposited on the outside 

 around the entrance. (2) Instances were observed where the 

 entrance to the burrow was submerged for about an inch, due to a 

 raising of the level of the water. The waste matter was always 

 deposited at the entrance of the burrow and under water. 



Microscopical examination of the excrement at all stages of the 

 petiole period showed a composition similar to that described for 



