DRAGONFUES AND DAMSELFUES IN PONDFISH CUI.TURE. 21 7 



there happens to be a small brook at the bottom of the gully fully developed imagos of 

 Lihellula pulchella and Plathemis will be found, as is usual in such places. These hordes 

 of tenerals are simply resting and feeding in comparative security until they become 

 ready for the active duties of propagation. The time thus occupied is apparently about 

 five or six days but is difficult to determine, because new tenerals are arriving and 

 matured individuals are leaving all the time, and it is practically impossible to tell just 

 how long any of them remain. 



It is probably a similar retirement from the vicinity of the water during the interval 

 between emergence and sexual activity that accounts for the disappearance of some 

 of the Gomphus species. (See p. 188.) 



The food obtained in these gullies and out on the prairie is of necessity somewhat 

 different from that captured around the ponds, but still consists largely of flies, gnats, 

 and mosquitoes, with an occasional lepidopter. 



Periodic Eating. — Since most of their prey is captured while on the wing, when 

 it is bright and sunny the imagos are eating much of the time, while in dull and cloudy 

 weather they eat very little, if at all. 



Microscopic examination of the digestive tract shows that the newly emerged 

 tenerals do not eat anything for a day or two, until about the time their color pattern 

 is fuUy formed. Probably the tenerals of L. ltu:tuosa just mentioned take no food before 

 reaching the gullies and the prairie. Other tenerals that remain near the ponds of 

 course obtain their first food there. 



Again, such an examination shows that, while the imago is voracious and often 

 feeds all the time, we can still distinguish two periods of maximum eating, fairly well 

 marked. Whatever is eaten during any day is all digested long before the next morn- 

 ing, so that imagos captured before leaving their roost in the morning will have nothing 

 in their gizzards and very little in the intestine. Accordingly the first period of maximum 

 eating comes in the forenoon, shortly after the imago leaves its roost, as soon as the 

 insects which constitute its food begin to swarm. This is followed by a lull, or at least 

 a diminution in the amount eaten, which lasts until well into the afternoon, and dur- 

 ing this period they are occupied with mating and egg laying. The eating then increases 

 again, and the second period of maximum feeding, which is more intensive than the 

 first, comes toward sunset. 



Of course, it will be understood that there is no intention of implying that imagos 

 eat two meals a day, or anything of the sort. There are simply more of them feeding in 

 the morning and late in the afternoon and more of them depositing their eggs through 

 the middle of the day. Moreover, the gizzards of those captured at 9 or 10 a. m. and 

 toward sunset are more apt to be well filled. They are thorough believers, however, 

 in eating between meals, and are not restrained in their desires by any irksome rules of 

 hygiene. 



Food Found in tun Alimentary Canal op 218 Imagos of Pantala FlavescEns in Hawaii. 

 Diptera: 



Midges, Chironomus hawaiiensi^ i 



Night mosquitoes, Culex fatigans r 



Day mosquitoes, Stegomyia scutellaris i 



Undetermined mosquitoes ^ 



Pomace flies, Drosophilidae ^ 



Undetermined flies 140 



