THE TOILERS OF THE NIGHT. 51 



Some of these last journeys were merely short flights aroimd 

 her domicile, and not for the purpose of seeking prej. 



We now left her but came back at half past two in the after- 

 noon. She was working, and she kept up her goings and com- 

 ings until four o'clock when she suspended operations for the 

 <iaj. On the next morning we were called away and know 

 nothing of what she did, but on the following day, July thirty- 

 first, we resumed our observations. She worked hard all the 

 morning, but in the afternoon her trips were few and were made 

 int long intervals. On the morning of August first she worked 

 from eight to nine, when she departed and never returned. We 

 watched for her, at intervals, all through that day and the 

 next, when we were forced to conclude that our faithful little 

 worker had fallen a victim to some bird or beast. We did not 

 disturb the nest until the fifth, when we cut the stalk and ex- 

 amined it. 



We found that the tunnel was thirty-nine centimeters in 

 length. This was a long distance for her to excavate, and, all 

 things considered, her progress had been rapid. We have 

 opened a number of stems that had been stored by this species 

 and all the excavations were from thirty to forty centimeters 

 in length, the width of the gallery being about three and one- 

 half millimeters, while on each side there was from one to one 

 find one-half millimeters of pith that had not been cut away. 

 Of course these points varied with the diameter of the stem and 

 also with the size of the worker. 



We found that our little stirpicola had stored one cell, had 

 laid an egg, and had built a partition of pith across the stem as 

 a floor to the second cell, before her untimely taking off. Had 

 she lived, ten or twelve cells would have been stored, one above 

 the other. The completed cell contained a larva and parts of 

 eighteen flies of different sizes, four species being represented, 

 Opthirsia punctipennis Wlk,, Anthomyia sp., CalUphora vom- 

 itaria, and another that we could not identify. The flies had all 

 been attacked by the larva, the abdomens of some and the tho- 

 races of others having been eaten. The larv^a continued to eat 



