t'RESIDENX's ADDRESS. 2l 



the stigma closed. The lever of the connective was nearly an 

 inch long. 



In "The Fauna Hawaiiensis"(10), Perkins gives a long account 

 of the pollination of endemic flowers by native birds belonging to 

 the Families Drepanidid(e and Meliphayidc. The former contains 

 thirty-five species in seventeen genera, and the latter five species 

 in two very distinct genera. The birds of the first family vary 

 from entirely honey-eating to entirely insect-eating, and the 

 Meliphagid(H appear to be entirely honey-eating. All the Dre- 

 paiiididoi have the tubular tongue, which shows descent from a 

 common ancestor; and the author considers that that ancestor 

 was a honey-eater, but that, as insects became more common, the 

 characters of some of the birds gradually altered. He states 

 that nectar is absolutely necessary to the existence of five of the 

 genera, and that these can be kept alive on nectai- ami sugar- 

 cane juice. The nectar-feeding birds ai-e characterised by a 

 slender beak, as well as the tubular tongue. "All, or practically 

 all, the plants visited by these liirds for food had bell-shaped or 

 tubular blossoms, in which nectar was more or less hard to reach. 

 Of these tubular-rtowered plants, there are several predominant 

 genera, some of which are themselves restricted to the islands, 

 and belong to various families, comprising hosts of peculiar 

 species. Most striking of all are the arborescent Lobeliacete, 

 not closely related to forms found in other countries. The mul- 

 tiplicity of these plants, and their isolation from foreign forms 

 bears a striking i^esemblance to that of the Drepanid birds them- 

 selves, indicating likewise an extremely ancient occupation of 

 the island." This seems to me to show that the flowers (Jjobe- 

 liacea) anfl the birds developed in dejjendence upon each other, 

 and the author holds the same view, for he says: — "How easily 



the extraordinary lengthening of the bill may have taken 



place, side by side with the increasing length of the tubular 

 flowers, is apparent from tht; fact that, even now, in some of the 

 bii'ds, tlierc is iiuHvidual \ai"iation in this respect. ... A series 

 of observatiims made on one of the most superb of the Lobelias 

 showed that it could only be fertilised by these highly specialised 



