494 CARl. SKOTTSBERG 



J'aul and New Amsterdam Poa Novarae and Scirpus nodosus, the latter a 

 member of a somewhat similar community in Juan Fernandez, are dominant in 

 addition to Spartiiia. SCHIMPER writes that St. Paul is »bedeckt von Gras- 

 steppe». There are no trees, while in New Amsterdam Phylica nitida is found, 

 and ScHIMPER speaks of the Phylica- Spartina community as a »savanna». It 

 may or may not be convenient to include this community in the savanna type, 

 and I believe that it is better to avoid the use of the term »steppe» for the 

 Spartina- Ass. SCHENCK showed that its leading species are xeromorphic in 

 structure, but the ecological factors and the water economy must be very dif- 

 ferent from those found under true steppe conditions. SCHlMPERs notes and 

 the very fine photographic plates are quite suggestive, and the physiognomic 

 likeness between these islands and Easter must be extraordinary. It can hardly 

 be altogether accidental. In New Zealand, grassland occupies wide areas, 

 especially E. of the mountain ranges in the South Island. This tussock-steppe, 

 as it is termed by Cockayne (Vegetation of New Zealand, Veget. der Erde 

 XIV, 192 1 ), enjoys a smaller rainfall and is said to be decidedly xerophytic. 

 Among the members of Cockayne's PoaFcstuca tussock association is Di- 

 chelachne ci'inita and perhaps also the very similar D. sciurea: both seem to 

 be common grasses in Easter Island. 



I believe that we are entitled to regard the grassland in Easter as a 

 tropical facies of oceanic grassland. I have already told that a similar 

 type is found in Hawaii, and I suspect that communities of the same kind exist 

 in Polynesia. In the Atlantic, Sporobohis indicus is reported »indigenous on 

 the hills» in St. Helena (it is most likely introduced), and there is some sort 

 of grassland also in Ascension. 



Along the north coast back of Hanga Ho Orno and Anakena the ground 

 is very stony with outcrops of rock where vegetation is restricted to the crevices. 

 Sporobohis is not so common here as Paspalmii scrobiadatnin var. orbiailare, 

 but on the slopes of the hills the former takes the lead. On the small tufa 

 cones is a more or less pure cover of Eragrostis clongata, but no close in- 

 spection of this association was made. Between Rano Aroi and Rano Kao 

 Sporobohis seems to give way to Axoiwpns pascJialis and to the introduced 

 Sorglmm halepcnsc, while the slopes of Rano Kao are covered with Sporobohis. 

 The low country along the south side of the island is very stony in places: Spo- 

 robohis is dominant and Sorgluan not uncommon. Round Rano Raraku the 

 Sporobohis- Kss. is well developed. Accumulations of boulders so commonly 

 met with on the low land are accompanied by the bright green bunches of the 

 fern Microhpia strigosa. Kyllinga brcvifoha is nearly always present in the 

 grassland and increases in number as we advance uphill, where the moisture 

 is greater. 



On the land leased by the farm of Mataveri, the grass is burnt in order 

 to improve the grazing. It is a pity that no botanical investigation was made 

 previous to the introduction of sheep and cattle. Animals disturb the balance 

 between different species of grasses, and I was told that they much prefer 



