Systematic Botany.] SUBANTARCTIC ISLANDS OF NEW ZEALAND. 467 



In the foregoing pages a large number of facts dealing with the composition and 

 affinities of the flora of the islands to the south of New Zealand have been arranged 

 and tabulated. I think that the study of this material proves that there are three 

 main elements in the flora. These we may particularise as under : — 



First, an endemic element. This clearly consists of two parts, the first and 

 most important comprising plants like Stilbocarpa, Pleurophyllum, the lonopsis 

 section of Celmisia, &c., which are probably the remnant of an ancient flora 

 which ages ago occupied the islands, and possibly other lands in the vicinity ; the 

 second consisting of species closely allied to others at present living in New Zea- 

 land. Whether the presence of the latter is to be accounted for by an immigra- 

 tion of old date from New Zealand over an intervening ocean, or whether it is 

 due to an ancient land connection between the two countries, is not at all easy to 

 decide. 



Second, an element which includes by far the greater portion of the flora, and 

 which is composed of species at present living in New Zealand. These have all the 

 appearance of recent immigrants, and, in my opinion, have arrived from New Zealand 

 since the islands were separated from one another and had assumed much of their 

 present size and configuration. This element shows affinities with Australia, and, in 

 the case of certain ferns, with Polynesia also. But, as has been previously pointed 

 out, the Australian relationship, so far as living species are concerned, is a conse- 

 quence of the connection of the flora of the islands with that of New Zealand, and 

 not an independent factor. 



Third, a Fuegian element. This also consists of two parts, one much older 

 than the other, and resting on the presence of species of such genera as Colobanthus, 

 Abrotanella, Phyllachne, Rostkovia, &c., the introduction of which cannot be of much 

 later date than that of the Stilbocar'pa-PleurophyUum element of the flora. The 

 second portion, including plants living at the present time in both countries, such 

 as Ranunculus biternatus, Azorella Selago, &c., is of much more recent origin. I have 

 already stated my belief as to the manner in which these^'species have found their 

 way into the southern islands. 



The occurrence of several species (at least twenty-five) with an exceptionally 

 wide distribution, although highly interesting from the point of view of plant- 

 dispersal, is not of much use in explaining the origin of the flora, and I do not 

 propose to take their presence into consideration in this inquiry. Nor do I intend 

 to bring in any conclusions based upon the cryptogamic portion of the flora, apart 

 from the ferns and lycopods. Their spores are so easily spread by the agency of 

 the wind that no difficulty exists in accounting for their presence in any country, 

 however remote. 



The conclusion to be derived from the above facts is that the flora of the islands 

 to the south of New Zealand is in its main characters and alHances nothing more 

 than a branch of the New Zealand flora, to some extent changed and modified by 

 long ages of isolation. Its history, so far as we can read it, shows clear evidence 

 of successive immigrations from New Zealand — some tolerably ancient, others of 

 recent date. These immigrations overshadow in numbers and importance those 

 of a comparatively small number of Fuegian or South American plants. The 

 ancestors of some of these may have been derived from the Antarctic Continent 

 at the time when it possessed a copious vegetation; the remainder have mostly 



