THE VKGKTATION OF riASTl^K ISLAND 



493 



tion witli dominant Sporoboliis <ioni^a/iis} Similar types hava been described 

 from other warm coiintres. Under the heading » Subtropical and tropical grass- 

 lands and meadows» WARMING and GRAE15NER (Lehrbuch der oekologischen 

 Pflanzengeographie, p. 587) mention Brazil, the West Indies fSporoboliis, Kyl- 

 lifiiia), Hawaii etc., and a comparison may also be made with the Impirata 

 arundinacca- Ass. in the Malayan region. It is stated that these grasslands are 

 due mainly to human influence and the disappearance of forests. Similar com- 

 munities occupy much ground in Juan Fernandez where they are, at least in 

 part, natural. In the system of Brockmann-Jerosch and RiJBEL (Die Ein- 

 teilung der Pflanzengesellschaften, Leipzig 1912, with additions and corrections, 

 Ber. d. deutsch. bot. Ges. 33, 191 5) we have to look for this type under »prata» 

 or possibly under »deserta». Among the former only two groups can be dis- 

 cussed, >Kluriprata» and »sempervirentiprata». The grassland in Easter Island 

 has very little to do with the former, with their pronounced winter s rest, and 

 the latter includes such different things as the northern mesophilous meadows, the 

 »Alpenmatten», »SchneetaIchen» etc. We know practically nothing about the 

 ecology of the vegetation in Easter Island, but it seems impossible to bring it 

 to the steppe types, neither to the prairies nor to the »siccideserta». There 

 is no period of drought or cold causing periods of rest, there are no bulbose 

 species except two species of Ophioglossuni and very few wild annuals, if any. 

 There is no marked periodicity in the floral phenomena. Nothing is known 

 regarding the water economy. Of the common grasses, Sporobolus presents 

 xeromorphic features and so do the rare DantJionia paschalis and Stipa hoj-ri- 

 dula, both considered to be endemic. 



The type of grassland which offers itself for comparison is the oceanic 

 meadow developed along the coast of northern Norway, on the north Atlantic 

 islands and in the south temperate and subantarctic regions (tussock grassland), 

 living under corresponding temperate oceanic conditions. RUBEL (Handworter- 

 buch der Naturwiss. IV, p. 887) brings these types to the »sempervirentiprata». 



The Tristan da Cunha group (see the Challenger Report, Botany, vol. i, 

 1885) and St. Paul's and New Amsterdam Islands (see Wiss. Ergebn. d. deutsch. 

 Tiefsee Exp. »Valdivia», vol. 2, 1905) offer beautiful examples of a similar kind of 

 vegetation. In Tristan, the Sparthia anindinacea- Kss. with groups of the dwarf 

 tree PJiylica iiitida forms one of the main features of the islands, and in St. 



' In my report on tlie Phanerogams ot Easter Island, this volume, p. 65, this grass was 

 named S. indiciis (L ) R.Br, on the authority of Dr. Pilger. Very often .S". elongaius R.Br, is re- 

 ferred to indictcs, also in manv recent works. But Dr. Hitchcock regards them as distinct. 

 He calls the species occurring in Hawaii S. elongatus, and according to him it is not very 

 closely related to .S". inditits. I have seen Hitchcock's material, and there cannot be the 

 slightest doubt that the Hawaiian and Easter Island plants are absolutely identical. I have also 

 examined many more specimens from various parts of the globe, also from tropical America, 

 where S. iiidiciis is supposed to be indigenous, and some of these look exactly like S. elon- 

 gatus in all details. If thev represent the true itidiciis, we have good reason to unite the two 

 species. It remains to be stated what Lin'NE's Agrostis indica was, and as long as 1 do not 

 know this, I prefer to use the name S. elongatus for the Easter species. It is widely distri- 

 buted in southern Asia, the Malay region, Australia and New Zealand. Setchell, Tahitian 

 Spermatophytes (Univ. of Calif, publ. in bot., vol. 12, p. 147) lists S. indicus, which is said to 

 be of recent introduction. Schumann and Lauterbach, Flora der deutschen Schutzgebiete in 

 der Siidsee, p. 182, mention S. elongatus only. 



