MISCELLANEOUS NOTES. 237 



an exceptional one, as the level sun shows up clearly both the hunter and its 

 prey but the flight of the prey is so distinct one can easily recognise the 

 insect. In most cases it is practically impossible to determine by observation 

 what dragonflies are getting. 



Another method has suggested itself; in the field one sees dragonflies sitting 

 on a convenient plant or support and darting off every now and then on the 

 chase. Below such a point, to which the same dragonflies come back constantly, 

 one finds their excreta; we examined some of these, crushing them gently in 

 water on a slide and examining them with a quarter and a sixth inch objective. 

 In those we saw, the excreta were a solid mass of chitinous remains of insects' 

 legs, antennas, the facetted coverings of eyes, odd pieces and so on. These 

 are not determinable to species, but they are more or less referable to 

 groups ; the characteristic legs of Orthoptera, the hairy legs of Aculeates, the 

 scales of Lepidoptera, the elytra of beetles, the spinous tissues of flies, the 

 scales of Mosquitoes, the scutella or hemelytra of bugs, the legs of Homoptera 

 are more or less distinguishable, even after exposure to the digestive fluids of 

 a dragonfly. In some we saw the remains were exclusively dipterous, so far 

 as anything definite could be traced, but there was no sign of anything belong- 

 ing to a mosquito. 



In a single instance recently we saw a very small pale-blue Agrionid deliber- 

 ately eating a Caddis-fly (Trichoptera), the common ^Ethaloptera sexpunctata 

 Kol., nearly as big as the little slender dragonfly. 



The subject is interesting as we see dragonflies constantly in great abund- 

 ance; we believe them all to be predaceous; we believe that nearly always they 

 feed on only small insects they can dispose of on the wing, and such small in- 

 sects are probably in many cases mosquitoes and other small insects which fly 

 in the still air and are easily caught. The mosquito in these days is so import- 

 ant, and so many investigators are on his track that even incomplete observa- 

 tions may give a clue and clearly the dragonfly question is worth attention. 

 The only other method of investigation that occurs to me is to catch and kill 

 dragonflies and investigate their stomach-contents. The direct observation of 

 dragonflies, the investigation of their stomachs and excreta should give data 

 on which to decide what it is dragonflies do eat. 



In Poulton's "Predaceous Insects and their Prey" (Trans. Ent. Soc, London 

 1906, p 399 et seq) are given records of the food of dragonflies; 26 cases only 

 are cited showing how very little is known and recorded; the prey includes a 

 fly, a dragonfly, a wasp, a tabanid fly, a Malacoderm beetle, a wasp, another 

 dragonfly, a Coccinellid beetle, a Lycaanid butterfly, a May-fly, a Crambid 

 moth, a Syrphid fly, a Galerucid beetle, a Skipper (Hesperiid) butterfly, a 

 Danaid butterfly, and a Termite. The records are probably very one-sided as 

 the prey to be identified must be large enough to be clearly seen, and 

 dragonflies more usually prey on small insects. The subject is one that may 

 be of importance and may perhaps commend itself to some observer in India 

 gifted with patience and leisure. 



