DERIVATION OF THE FLORA AND FAUNA 257 



ing to Arcto-tertiary). An Antarcto-tertiary taxon is not necessarily of Antarctic 

 ancestry; strictly spoken it certainly should be, but there are numerous cases 

 where the point of origin was either the South American or the Australasian- 

 New Zealand centre, and Antarctica only served as a trans-continental route of 

 migration. 



Trisetum chromostachymn is Chilean, but the genus is both N. and S. temper- 

 ate, perhaps originally Arcto-tertiary? Danthonia collhta and Koelerm micrathera 

 also are Chilean, but the distribution pattern of the genera indicates that Antarc- 

 tica eventually was involved in their history. 



Megalackne, temporarily referred to Bromus but once more stated to be an 

 independent genus, is a relict type with unknown history, eventually a remnant 

 of a pre- Andean flora (comp. above p. 217). 



The genus Chusqiiea belongs to the neotropical element in the flora of Chile, 

 and the same is true of the species of Cyperus. Scirpiis nodosus and cernuus are 

 circumpolar seaside plants and most likely thalassochorous; if Antarctica had a 

 share in their earlier history is impossible to say, but it is not improbable. Eleo- 

 charis is another large world-wide genus, perhaps too wide-spread to allow us to 

 locate its place of origin. 



Oreobolus. The Antarcto-tertiary character of Oreobolus can hardly be dis- 

 puted; it is often referred to as a classical example of an Antarctic genus. Of the 

 6 species recognized by KUKENTIIAL, i (with 2 varieties) is found in S. and E. 

 Australia, Tasmania and New Zealand, 2 in N.E. New Guinea, one of them 

 extending to Borneo, i in N. Sumatra and Malacca, i in Hawaii, and O. obtusangu 

 lus in Chile from the Cordilleras of Valdivia to Fuegia, the Falkland Is. and Juan 

 Fernandez. [A seventh species, 0. pfeifferianus Barros, was identified by KUKEN- 

 THAL [16^. vill) with pjunilio var. pectinaius\ 



Cladiiim (16^. Xll) is wide-ranging but it is not cosmopolitan in spite of the 

 large areas occupied by C. 7nariscus L. and its varieties, among v^\\\q\\ j amaiceiise 

 is circumpolar and distributed also south of the equator. The main distribution 

 of subg. Machaerina (inch Vincentia), where C. scirpoideum of Juan Fernandez 

 belongs, is palaeotropical with 11 species (Madagascar, Mascarene Is., Australia, 

 Lord Howe I., New Guinea, Indonesia, Oceania); 5 are neotropical (W. Indies, 

 Brazil). Its austral character is clear enough. The closest relative of scirpoideum 

 is not an American species but C. angustifolium (Gaud.) Benth. et Hook. fil. (New 

 Guinea, Tahiti, Hawaii). Subg. Baumea, with 29 scattered from Australia, where 

 there are 18 species, over the Indomalayan region west to Ceylon, north to Hong- 

 kong and Japan, east to Melanesia and Hawaii, has 15 in Australia-Tasmania- 

 New Zealand and i in the region of Madagascar and Mascarene Is. Together the 

 two subgenera cover the 3 sectors, the African, the Australian-Malaysian and the 

 American. In the centre of this vast area lies Antarctica or, as it were, Gond- 

 wana Land. The history of Cladium may well lead back to the Mesozoic, and 

 it seems natural to refer the genus to the Antarcto-tertiary element. The same 

 applies to Uncinia, one of the generally recognized Antarctic genera, represented 

 in Juan Fernandez by 5 species, 2 of them endemic. Within the South American 

 sector are 13 species distributed along the Andes with a concentration toward the 



17 "557857 The Nat. Hist, of Juan Fernandez and Easter Isl. Vol. I 



