DRAGONFUES AND DAMSELFUES IN PONDFISH CULTURE. 



imago was found anywhere in the vicinity until the last of July, and then only a few. 

 On the other hand, the imagos of both Gomphus externus and G. -vastus were common 

 about the ponds, but neither nymphs nor nymph skins were secured from the ponds 

 themselves. 



The nymphs of Anax junius and Mschna constricta were abundant in all the ponds, 

 but comparatively few imagos could be found at any one time. This migratory scat- 

 tering profoundly affects the numbers of imagos found about the ponds. 



3. The Weather. — "All dragonflies are most active in hot, quiet, sunshiny 

 weather" (Needham and Hart, 1901, p. 1 1), and consequently they appear more abundant 

 at such times than on a dull, lowering day, when many remain in the shelter of the grass 

 and weeds. Hence if we are to estimate the number of dragonflies correctly we must 

 take into account the kind of weather during which the observation is made. 



4. Concealment. — Large numbers of the smaller damselflies, especially such species 

 as Ischnura verticalis, frequent the dense grass and are sure to escape observation. 

 Hence the actual number of damselflies in any locality is almost certain to be underesti- 

 mated, unless the grass and weeds are closely examined. A vigorous sweeping of the 

 net over them always reveals far larger numbers than were first seen and often yields 

 species that would otherwise escape detection. 



Careful estimates, made by repeatedly walking around the ponds and counting such 

 specimens as could be seen, making due allowance for repetitions and keeping in mind 

 the considerations just presented, indicate that there is for each of the ponds a fairly 

 constant average of 100 to 150 dragonflies, while the numbers of damselflies vary with 

 their periods of emergence. Just after emergence there will be 300 to 500 for each pond, 

 but between times the number may fall to 50 or even less. These numbers hold through 

 the latter part of June, July, and the first of August, but become greatly reduced by the 

 •last of August. 



RELATIVE ABUNDANCE OF DIFFERENT SPECIES. 



Turning now to the relative abundance of the various species we can secure an 

 exact numerical basis for our estimate. While it would be manifestly impossible to 

 count the imagos of the different species and obtain any result worthy of record, good 

 results can be obtained by gathering and counting the nymph skins. Such collections 

 were made at intervals of two weeks during the summer of 191 6 with the results shown 

 in the following table. The successive counts were made along the north shores of 

 ponds 4, 3, 2, and i, respectively, the length of shore covered by each count being about 

 the same. Over 2,000 nymph skins were obtained in the four counts, which indicates 

 that the numbers given above under actual abundance are not too large. 



