EXPERIMENTS IN REARING. 79 



and so far as could be perceiv^ed they rested quietly in the 

 interior of the boxes upon the net. (As a matter of fact, 

 the rushes, etc., were gathered in a basket at this time and 

 placed in the tub on my return that evening.) The col- 

 lecting boxes and fly traps were also placed in the bath 

 room that evening, so that the insects could liberate them- 

 selves the next morning. The floor below was darkened 

 by closing the shutters, in order that none of the insects 

 should be induced by the light to pass down the stairway. 

 The next day was Sunday, August 1 8th, a bright, pleasant 

 day, and soon after dawn I stationed myself at convenient 

 points of observation, remaining as motionless as possible 

 consistent with an occasional change of position. The re- 

 sult was a disappointment, so far as pairing, ovipositing, or 

 even the destruction of the flies were concerned. Tlie flies 

 began to circle around the rooms as soon as they were 

 fairly light, and as morning advanced flew towards the 

 windows seeking an outlet, and alternated between the 

 windows and the objects in the room, crawling over the 

 walls and ceilings, lighting on the plants, and apparently 

 searching for food. The dragon flies were indisposed to 

 activity until the sun began to heat and light the rooms, 

 about 10 A. M., when they, too, flew toward the windows 

 and skylights, and though frequently flitting around the 

 ceilings always returned to the former, and when appar- 

 ently fatigued with their exertions, or convinced that their 

 eflbrts to escape were futile, would remain resting on the 

 sills, unless disturbed by the buzzing flies, which crawled 

 fearlessly over them. There were no attempts at ovipos- 

 iting or pairing, and the improvised pool with its rush 

 lined banks seemed to offer no attraction, probably be- 

 cause it was lacking in three particulars, viz., a muddy 



