DRAGONF^UES AND DAMSElvFUES IN PONDFlSH CULTURE. 



207 



The nymph eats voraciously as long as it remains light colored; while the color 

 pattern is disappearing the amount eaten also diminishes, and after it has turned brown 

 the nymph's gizzard contains little if any food. All the nymphs whose gizzards were 

 empty were dark-colored, but the intestine usually contained indigestible debris from food 

 previously eaten. One very dark nymph of L.luctuosa contained only a single tiny Cypris 

 incquivala, another had but one Ceratopogon. larva, while a third yielded two Ischnura 

 nymphs with nothing in the intestine. Similarly two dark-colored nymphs of Anax had 

 each eaten but one Ischnura nymph, and the intestines were empty. A dark nymph of 

 E. simplicicollis contained only half a dozen short alga-1 filaments v/ith nothing in the 

 intestine. This condition is particularly true of nymphs as they approach their final 

 transformation, when they apparently fast for quite a long period before crawling out 

 of the water. Hence we conclude that an empty digestive tube is one of the essential 

 prerequisites for the great change which then takes place. 



Warren (1915, p. 8) has given the lengths of the various instars, meaning by that the 

 periods between molts, during the entire nymphal life of four specimens of Pantala fla- 

 vescens. These periods are about the same length for the first 9 or 10 molts and then 

 increase greatly the last two molts. The final period, corresponding to the pupal stage in 

 insects which have a complete metamorphosis, lasted for a month, while the first nine 

 molts were only five or six days each. He did not record the periods of fasting, but it is 

 probable that these nymphs ate very little during the last week before transformation. 



Balfour-Brown (1909, p. 270) similarly found that the periods between molts tend 

 to lengthen as the nymph grows larger. In some of his damselflies there were nearly two 

 months in the last period ; in such cases the fast preceding transformation would also be 

 lengthened. 



Besides these fasts which accompany the various molts and the one which precedes 

 the final emergence, the nymph is able to go without food for long intervals when necessary 

 and apparently suffer no injury. It is doubtful, however, if a nymph could transform 

 when the fast preceding emergence was considerably lengthened. It is quite possible 

 that some of the failures to fully emerge from the nymph skin (p. 222) may be caused by 

 an insufficiency of food. 



Food of Hawaiian Nymphs. — Warren examined the contents of the alimentary 

 canal of 253 Anax and Pantala nymphs captured in the vicinity of Honolulu, Hawaii, and 

 it is interesting to compare his results with those recorded for Fairport. (See table, 

 p. 201.) 



Food Found in AIvIme;ntary Canal of 253 Odonate Nymphs from Honolulu, Hawaii, 

 Examined by A. Warrkn. 



Ants and bees: 



Pheidole megacephala (Myrmicidae) 2 



Ants, undetermined species 11 



Dragonflies: Pantala flavescens, nymphs 6 



Crustacea: 



Cypris, sp 108 



Shrimps, sp. 3 



Protozoa: Euglena, sp 30 



Worms: Nereis, sp i 



Amphibians: Tadpoles 8 



Fish: Top minnow (?) i 



Mollusca: Spiral shells 14 



Beetles: Dytiscidae 16 



Flies: 



Chironomid larvae 16S 



Chironomid adults 4 



Mosquito larvae and pupae 12 



Mosqtiito adult i 



Dolichopodid fly i 



Adult fly, undetermined i 



Bugs: 



Merragata hebroides (Nsogeidae) i 



Microvelia vagans (Veliidse) 2 



