78 DRAGON FLIES VS. MOSQUITOES. 



The next experiment was undei-taken for the purpose of 

 testing the willingness of the dragon fly to mate or to 

 oviposit while in partial confinement. 



A portion of the upper story of my house on Prospect 

 Heights, in Brooklyn, consisting of bath room with large 

 skylight, hall with skylight, and adjoining hall room with 

 southern exposure, was utilized as a temporary nursery for 

 dragon flies during August and again in September of this 

 year, my family being absent. Every article of furniture 

 was removed, the carpet covered with newspapers, the 

 bath tub filled nearly to the brim with water and plenti- 

 fully fringed with rushes and aquatic grasses, the roots of 

 which were inserted in several pans of previously well 

 washed sand placed on the bottom of the tub ; some boxes 

 and water jars containing large weeds transplanted from 

 an adjoining field were distributed at intervals along the 

 hall and hall room, and a wire fly trap was set in the 

 kitchen to provide a supply of food for such dragon flies 

 as might be taken. 



In the afternoon of that day (August 17th) some brack- 

 ish pools a quarter 'of a mile from the beach at Coney 

 Island were visited, and between thirty and forty speci- 

 mens of Libellula pulchella and Diplax rubicundula were 

 captured with a large net (eighteen inches diameter), 

 care being taken to secure some pairs in copulo. These 

 were immediately inserted, according to size of insect, in 

 one of two confectionery boxes lined with mill net, with 

 sliding covers, each having a hole one and one-half inches 

 in diameter, through which the dragon flies could be 

 thrust without removing the cover. The darkness of the 

 box as well as the lateness of the hour prevented the in- 

 sects from attempting to fly and thus injuring themselves, 



