125 



ness), and if the position of its victim was favourable, it pounced 

 upon it immediately. Then the frog-hopper hopped; in some 

 cases the fly lost it ; in some cases the fly ^reappeared instantly 

 from the place to which the frog-hopper hopped. In one case I 

 saw the frog-hopper land with the fly still on its back ; I caught 

 both, without waiting, as perhaps I should have done. If the po- 

 sition was unfavourable, or stems got in the way (e. g. Ccraiuiiiii 

 Robci'fiaiinni), the fly would circle round with its head towards 

 the victim, like a male Dolichopns, seeking a point fromvvhich 

 to pounce. On several occasions it failed to get a hold. Once 

 a flv pounced on a frog-hopper which did not hop\ the fly imme- 

 diatelv left it. Another came up and looked at it, but went 

 awav without touching it. Was the frog-happer already enter- 

 taininig an egg, or was it a male, or for some reason unsuitable? 

 Clialanis is common in my garden, but I have not been able to 

 see anything of its oviposition. 



"These observations are crude and inconclusive on several 

 points; but I send them on the 'chance that others who can 

 aflford the time will complete them. The flies are still common, 

 and the frog-hoppers now less scarce." 



Verrall remarks that he considers the Pipunculidae to be the 

 most exquisite fliers that exist in Diptera. However true this 

 may be of the European, I do not think it can be said of the 

 Australian or Hawaiian species, which cannot compare in power 

 and beauty of flight with many Syrphidae or other flies; indeed 

 to cite but one instance they are vastly inferior to the Austra- 

 lian species of Baccha in this respect. It is certain however, 

 that on capture they often exhibit to the full their power of 

 hovering within the meshes of the net, where more sturdy spe- 

 cies, which wdiile free can remain poised and motionless in the 

 air or dart forward with speed that baiTles the vision, would 

 under like circumstances beat wildly against the sides in their 

 endeavour to escape. 



The Pipunculidae are known to affect various families of 

 Homoptera, the spittle-insects or Cercopidae, as well as the Jas- 

 sidae and Fulgoridae, and perhaps the Cicadidae, being subject 

 to their attacks. It is not known whether in any case the same 

 species of parasite will attack leaf-hoppers of more than one 

 family, e. g. Fulgoridae and Jassidae. In this connection, how- 

 ever, one may mention the fact that the resemblance between 

 such species as P. hclliio and P. Koebtdci is extreme, both 

 structurally and superficially, although the former is parasitic 



