27 



number are iindoiibtedly due to extremely varied conditions in these islands, pro- 

 duced by the enormous rang:e of altitude, and eonse((ucnt variable climatic condi- 

 tions, which proved favorable and remained so, to the development of distinct 

 forms from an early ancestor. Such widely ranging conditions were not available 

 in the islands of the eastern Pacitic as Tahiti and Rarotonga, which probably ac- 

 counts for the vei-.v small number of Lobdioidcac : — in each island one or two 

 species at most. That the genera ScJcrothica and AprfaJiia were derived from 

 Australian species there is probably no doubt. In .judging the length of time 

 that the Lohclioidcac have existed in Hawaii it is only necessary to look, next 

 to their wonderful development, at the fauna to which they are host. Some of 

 the birds which are dependent on the Lobdinidi m are quite as remarkable as the 

 plants themselves, and like them are not known in any other part of the world. 



The early species of the Hawaiian baccate genera of Lobelioidcae or their 

 ancestors must have belonged to a t>-pe that was able to subsist in the less humid 

 forests or even preferred the more open localities on the higher mountains of this 

 group. The first immigrants probabl.v have settled at high elevations whence they 

 descended into the humid forests below, where they found conditions favorable 

 to the remarkable development which they have now attained. Such old rem- 

 nants of high mountain forms we find in Clermontia nalcakalensis, which grows 

 on the upper slopes of I\It. Haleakala at an elevation of 7,000 feet whence one 

 would never expect a baccate Lobelioid. The fact is that Clermontia Haleaka- 

 hiisis is the most di.stinet and curious of all species belonging to that genus, and 

 represents proI)ably the oldest type of CIrnnonfia. 



Another high mountain foi'm is D(liss<a iiiidiilata whicli the writer found on 

 the slopes of Jlauna Loa at an elevation of about 5,()fl0-li.fl()0 feet in the dreary 

 forests of Mijoporuni saiidicicrii.'<r. Acacia Koa and Suphara chrusophyUa. Clcr- 

 mnntia Halcakalensis which has the aspect of a Dracaena- grows in company of 

 a similar vegetation, as Dclissra uiidulala. the ])lants growing nearest to it were 

 Rubus, Stenogyne rugosa, Santalum, Halcakalac. Sopliora, etc. It is to be re- 

 gretted that the uplands of these high mountains were given over to cattle graz- 

 ing. Many more such types as Clermontia Ualeakalciisiti occurred no doubt at 

 the higher levels, but have now become extinct ; these could have given us a clue 

 as to the origin or evolution of the Hawaiian genera. CI. Haleakalcnsi^i owed its 

 survival to a few deep lava gulches on the slopes of the crater Puunianiau 

 between which it grew, and which were more or less inaccessible to roaming 

 cattle. Thousands of people have ascended Mt. Haleakala, among them a number 

 of naturalists, biit it was left, to the writer to discover this rare and curious 

 species of which only three trees are, or were then in existence. How many of 

 these curious plants have disappeared from the uplands from where they 

 descended to the more liumid forest belts below we can not even guess, for the 

 present number of survivors of such ty'pes is too small. It is quite possible that 

 the baccate Hawaiian Lobelioideae have originated from two ancestral immi- 

 grants, of which that of the genus Cyanca antedated that of Rollandia. The 

 very- fact that Cyanca possesses so many species, almost as many as the other 

 baccate genera together and the remarkable types of the section palmaeformes, 

 as Cyanca Icpfostcgia, (' arliorca. and C. Giffardii, point to the great anti(piitv 

 of that genus, which undoiditedly antedates the other baccate genera. The 

 peculiar distribution of Rollandia with several species on Oahu, and only one on 



