Drajoiijlies and their Pfcy. 503 



fulling into two different families, but in this c.ise Dr. Oiir- 

 penter, who made the observation', was paying special 

 attention to tlie bionomics of the dreaded carrier of sleeping 

 sickness. The same entomologist obhiiiied two specimens of 

 /c/i«Ms/e?'oa; preying upon tlie same African honey-bee; but 

 he also found that Dragonfly feeding upon other Jlymenn- 

 ptera as well as Rhynchota. Again, the two specimens of 

 the blood-sucking fly Ucematopota longa, sent home from 

 Nyasaland by Mr. Neave, were being fed upon by difPereut 

 species of Ortheirum. 



So far as our enquiry has proceeded, it would appear that 

 Odonata, whether considered by species or as a group, are 

 omnivorous feeders among other flying insects *, and I am 

 not aware that any evidence is forthcoming to show that 

 apterous or larval insects contribute in any way to a 

 Dra<i:onfly's diet. Even Danaine and Aciasine butterflies, 

 which are known to be distasteful to many insect-feeding 

 vertebrates, are not rejected by the voracious Dragonfly. 

 Danaida chrysippus may be taken as a typical example of a 

 butterfly specially protected against the attacks of vertebrate 

 enemies ; but it does not enjoy tlie same immunity from the 

 assaults of Odonata. In addition to Dr. Simpson's rrcord 

 (No. 25), ]\Ir. Guy Marshall observed in Natal a very large 

 red Dragonfly (now recognized by him as Anax sper dus, 

 Hagen) devouring an imago of the same s|>ecips of butterfly 

 (Trans. Ent. Soc. London, 1902, p. 329). We have also 

 seen that honey-bees are sometimes hunted down for food 

 (see cases Nos. 14, 15, and 21). Indeed, the 'Field' for 

 21st Marcii, 1908 (p. 486), mentions the complaint of a bee- 

 keeper in Australia against " the dragon-fly, which is a 

 greater pest than any of the birds, and sucks bees dry by the 

 dozen in one summer da}'." Moreover, toll is taken ot" 

 certain other prcdaceous insects, such as Pom[)ilid Hymeno- 

 ptera and Reduviid bugs ; but, although Asilid flies not 

 infrequently attack Dragonflies, I cannot ascertain that 

 Odonata ever attack Asilidre. 



There is necessarily some correspondence between the 

 size of the captor and the size of tiie prey. All the largr- 

 bodied Hymenopfera which we have hud under review have 

 fallen victims to Dragonflies of the family 7Escluii(he, which 

 includes the largest members of the order, while the small 



* Mr. E. B. AVilliamson says that *' Mr. F. S. Webster has observeil 

 LiheHula auripennis feeding on fresh crocodile fl'ish " (Indiana Geol. 

 Reports, xxiv. p. 235, 1809). 



