126 



on Siphaiita and the latter on a Jassid somewhat alhed to Pcn- 

 thiinia. The fact that the two flies are so extremely similar 

 while the hosts are so utterly unlike leads me to suspect that 

 species of Pipunculus may be found that attack species of either 

 family of leaf-hoppers indiscriminately. However this may be, 

 it is certain that some of these flies will attack very different 

 species of leaf-hoppers within the limits of a family. For exam- 

 ple that most common Australian species P. beneftciens was bred 

 from nymphs of three or four very different Jassids, and P. 

 cniciator from two widely different genera, of the same family. 

 It may also be noted that nymphs and adult leat-hoppers are 

 both subject to attack. 



In Australia all the species of Pipunculus, that we bred, were 

 from Jassidae, excepting two, one of which {P. helluo) is a para- 

 site, as above mentioned, on Siphanta and allied genera, while 

 the other {P. xanthocnemis) attacks a species of Libiirnia. The 

 Hawaiian species here described are probably all attached to 

 Delphacine Fulgorids. Three of the five most certainly are, 

 since one of the latter has been bred by Mr. O. H. Swezey, and 

 the other two are found in company with the same leaf-hoppers, 

 in places where Jassids are quite absent. 



There is, so far as I know, no means of telling with absolute 

 certainly whether a living leaf-hopper contains a larva of Pipun- 

 culus or not. In some species of a green colour, when the para- 

 sitic larva becomes large, the green of the hopper is not infre- 

 (juently changed to a sickly yellow, or even dark discoloration 

 may be seen, but in most cases a slight sickly appearance, evi- 

 denced by indescribably minute differences in its superficial ap- 

 pearance, or by its more lethargic movements, alone inform 

 one, or at least lead one to suspect, that the leaf-hopper is para- 

 sitized. Practically it is not difficult to detect parasitized indi- 

 viduals, when one is familiar with the species of leaf-hopper that 

 is infested. If a green Jassid containing a large Pipunculus 

 larva be placed in alcohol, so that the colouring matter is large- 

 ly removed, and greater transparency is attained, it is some- 

 times possible to see the parasitic larva quite plainly through 

 the cuticle of its host. The head of the larva appears to be al- 

 ways turned towards the head of the leaf-hopper, and generally 

 the parasite fills up by far the greater part of the abdomen of 

 its host. Such a specimen, a species of Hecalus containing a 

 larva of P. cruciator, is figured on PI. VII, fig. i. 



W'hen the larva of the Pipunculus quits its host, it usually 



