NO. 1692. DRAG0NFLIE8 OF THE MISSISSIPPI VALLEY— WILSON. 557 



chella Drury were observed going to roost for the night in the tall 

 ironweed along a dried-up overflow bottom. When roosting they 

 flatten back against the vertical stem of the weed instead of holding 

 their bodies horizontal as is done when they alight in the daytime, 

 possibly as a protection against rain. JThis was not the right kind of 

 a shore for Gomphus^ and hence only the single pair was seen. 



XXIII. QUINCY, ILLINOIS, AUGUST 9. 



Only a half-hour could be spent here, and in that time the following 

 species were either seen or secured : Lihellula pulchella Drury, 

 JEshna juncea verticalis (Hagen), Pachydiplax longipennis (Bur- 

 meister), Gomphus amnicola Walsh, Argia tibialis (Rambur), and 

 A. moesta putrida (Hagen). 



XXIV. HANNIBAL, MISSOURI, AUGUST 10. 



Two hours in the afternoon and the same period the next forenoon 

 were spent here in collecting, but with limited results. There were 

 hundreds of Lihellida pidchella Drury flying across the river and over 

 the inland fields, but the only other species found were Gomphus 

 externus Selys, G. amnicola Walsh, Pachydiplax longipennis (Bur- 

 meister). Argia moesta putrida (Hagen), A. tibialis (Rambur), and 

 A. apicalis (Say), and of these there was only a single specimen of 

 each of the first three. 



XXV. THE ILLINOIS RIVER, AUGUST 12. 



In passing up the river from Grafton to Hardin two distinct 

 colonies of Erythemis simplicicoUis (Say) were found. The first 

 was 10 miles above Grafton, where the east bank of the river was 

 covered with hundreds of this species, including both sexes, while 

 many were flying across the river. 



The other colony was 4 miles farther up the river, at the head of 

 an island. Here the island seemed to be the headquarters from which 

 the dragonflies flew out in every direction. 



Just below Coon Creek an abundance of Libellida pulchella Drury 

 was observed, and they could be seen flying over the inland fields. 

 Repeated observations seemed to indicate that in general the flight 

 was from the shady to the sunny side of the river, from east to west 

 in the forenoon and from west to east in the afternoon. 



XXVI. COON CREEK, ILLINOIS, AUGUST 12. 



This was between the colony of Libellula and one of Erythemis^ 

 and there were found here, naturally, these two species, though in 

 limited numbers, and beside them Pachydiplax longipennis (Bur- 

 meister), fully as numerous as either of the preceding, Celithemis 

 eponina (Drury), Tetragoneuria cynosura (Say), and Gomphus 

 plagiatus Selys. 



