437 



a most interesting and significant discovery, showing that these 

 gall-making psyllids have been present here and living npon 

 Metrosideros for a considerable period of time. 



Probably the original immigrant species was one inhabit- 

 ing leaf galls of Metrosideros and, as new species have evolved 

 from this, some have retained the gall-making habit on the 

 same plant, others have taken to living free on the leaf sur- 

 face in the nymphal stages, while still others have gone off to 

 other plants, making leaf galls or living free. 



From this ancestral species, a THoza,- have been derived 

 thirteen species in three genera. Five of these belong to the 

 cosmopolitan genus Trioza,- five to an endemic genus, Hevaheva, 

 and the remaining three to a more or less cosmopolitan genus, 

 Kiiwayama, members of which have arisen independently it 

 would seem in various countries. These three species here 

 appear to be not related directly to other species of the genus 

 in other lands but rather to species of Trioza here. 



Hevaheva is the most specialized of the genera and indi- 

 cates a long evolution. The five species are closely inter- 

 related, but H. giffardi show^s considerably more specialization 

 than the others. The genus is more closely related to Trioza 

 iolani than to the other species of the Islands, but this rela- 

 tionship shows a gap not now bridged over. The relationships 

 of the species of each genus are discussed later. 



The other two genera are both Polynesian and apparently 

 have come in at a much later date than the ancestral forms 

 of the other assemblage. Megatrioza is abundantly represented 

 in the Malay and Philippine archipelagoes and probably else- 

 where in the tropics of the Old World. The one species here 

 probably has not arrived recently and probably is endemic, for 

 it occurs only in the higher mountains on the native palm 

 (Pritchardla). It does not occur on lowland introduced palms 

 and hence it seems that its arrival here must have long ante- 

 dated the modern epoch. Less is known of the other genus, 

 Cerotvioza, both here and elsewhere. It is possible that our 



