20 



Anisotome intermedia var. oblongi folia, A. flabellata (Epacri- 

 daceae) Dracophyllum Pearsoni; (Gentianaceae) Gentiana 

 Gibbsii; (Scrophulariaceae) Veronica, Laingii, Ourisia modesta; 

 (Compositae) Olearia divaricata, Celmisia glabrescens, Raoulia 

 Goyeni, Abrotonella muscosa. 



An interesting point is that, notwithstanding the Stewart 

 flora being separated from the South Island by 15 miles of sea at 

 the nearest point, it has only 6 per cent, of its florula locally en- 

 demic, much less than might be expected when the florulas of the 

 North-western and North-eastern Botanical Districts are con- 

 sidered, the former with 4 per cent., and the latter with 4.4 per 

 cent, of locally endemic species. Again, the Chatham Islands, 

 situated 500 miles from the New Zealand mainland, have only 13 

 per cent, of locally-endemic species. 



(e) Effect of altitude upon plant distribution. 



It has already been noted that New Zealand has a high- 

 mountain flora very distinct from that of the lowlands in that 

 out of its 950 species no less than 500 are entirely mountain- 

 dwellers, while 100 others descend to the lowlands only under 

 exceptional circumstances. These high-mountain species are not 

 arranged haphazard, but each has its special altitudinal range; 

 while, as in mountains everywhere, the flora as a whole is ar- 

 ranged in altitudinal belts. Thus, beginning at sea-level, there 

 is the lowland belt which extends to an altitude of about 

 1000 feet throughout most of the South Island, but in the 

 North Island it will be from 500 to 1000 feet higher. 

 Above the lowland belt comes the montane belt, and it 

 extends upwards for 2000 feet. Then the next thousand feet is 

 the lower subalpine belt, above which for one thousand feet is 

 the upper subalpine belt, and above this, up to the limit of vege- 

 tation, is the alpine belt. These figures are altogether approxi- 

 mate, for latitude, aspect and climate all affect altitudinal distri- 

 bution, so that there is no uniformity even on any one mountain. 

 This is evidenced by the sheep-farmer with his "summer" and 

 "winter" country, and in this he recognises the fundamental fact 

 regarding vertical distribution, namely the average length of 

 time that the snow lies on the ground during the cold months 

 of the year. Thus the alpine belt is that area where the winter 

 snow lies for some six months; on the subalpine belt it may lie 

 for two months or so in its upper part and a week or two in its 

 lower part ; on the montane belt the snow will lie usually only a 

 few days ; while on the lowland belt there may be snow only occa- 

 sionally, or perhaps not at all. These facts regarding the effect 

 of snow stand out clearly from an examination of the vegetation 

 of gullies and hollows where the snow lies for a long time far 

 into the summer it may be. A few examples may be of interest. 



On Mount Egmont, in the subalpine belt, there are two 

 well-marked plant-formations in close proximity; the one tall 

 tussock-grassland of Danthonia Raoulii, var. rubra, the tussocks 

 densely crowded, but with here and there open spaces occupied 

 by a turf of prostrate herbs, semi-woody plants and shrubs ; the 



