174 THE CANADIAN ENTOMOLOGIST 



ing their erratic course in the open or sometimes resting on low- 

 branches. 



On August 18 I witnessed the largest flight of ^^shna I have 

 ever seen. They were first noticed about 5 p.m., flying a few feet 

 from the ground over the grassy spaces and roadway just behind 

 the cottages along the lake front. They were hawking after other 

 flying insects such as midges (Chironomidae), which were very 

 abundant, and each dragonfly appeared to restrict its movements 

 to a more or less definite area. In certain spots they were so 

 abundant that there seemed to be about one for every square 

 yard of ground surface. The day was hot but the sun somewhat 

 obscured by the smoke of distant forest fires. The insects flew 

 continuously, never being seen to rest. A number were captured, 

 all proving to be constricta, males predominating. As the sun set 

 they rose higher and began to disperse, and at 7.10 p.m., though 

 still numerous, their numbers had greatly diminished and they 

 were flying at about 30 or 40 feet from the ground. At 7.1.5 p.m. 

 they had almost entirely disappeared. 



On the same day earlier in the afternoon I visited Wilson's 

 Creek to ascertain whether the swarms of A. canadensis, which 

 had left the woods, had migrated there for breeding purposes, 

 and also whether A. constricta or its exuviae were present. I 

 found the former species in considerable numbers but nothing was 

 seeVi of constricta. The numbers of canadensis, however, were 

 quite insignificant as compared with those seen about the woods 

 earlier in the season. 



From now on A. canadensis gradually diminished in numbers, 

 while constricta continued to be plentiful throughout August and 

 in early September. A small number were seen at Wilson's Creek 

 on Sept. 3, but they were flying high and apparently none were en- 

 gaged in oviposition. Their numbers now began to dwindle at 

 De Grassi Point, although there were still a few about the place 

 when I left on Sept. 23. As they evidently did not go to Wilson's 

 Creek to oviposit, or at most in small numbers, I went on Sept. 

 13 to the wide marshes at the mouth of the Holland River, about 

 33^ miles from the Point. It was too cool for dragonflies to be 

 flying in large numbers, but I saw a considerable number of ^'Eshnas 

 on the wing and found several others resting in the marsh grass. 



