Re 
| 
# 
‚iS.& pretty sure sign, as SpEiGHT has pointed out to me, that the ice-period of that group wasnot 
= Pe t9 increased el. a en a a 
Repeopling the new ground during the Retreat of the Glaciers. 177 
Calamagrostis pilosa, Poa novae-zealandiae, P. Cockayniana, Schoenus panci- 
| forus (especially where water lies), Karina autumnale, Carmichaelia grandi- 
fora, Coriaria ruscifolia, Metrosideros lucida, Gunnera albocarpa, Gaultheria 
rupestris, Dracophyllum longifolium, Veronica subalpina, V. Lyallii, Coprosma 
rugosa, Celmisia bellidioides, Olearia avicenniaefolia (almost the first-comer), 
O. zlieifolia, O. arborescens, O. Colensoi. ‘The transforming of rock, thus oc- 
cupied, into a closed association is a very slow process, so that there are 
many extensive bare patches of considerable age. Where moraine, even if 
‘quite thin, is deposited upon the rock, a closed association is quickly produced. 
Thus Harper Rock, a roche moutonnee still partly embedded in the terminal 
face of the glacier, is quite bare except on the 'summit where, on a patch of 
moraine, there is an embryonic scrub of Arundo conspicua, Carmichaelia 
grandıflora and some other plants. 
n older moraine-covered rock, at some distance back from the rock now 
being invaded, is a broad belt of tall scrub consisting of the shrubs already 
mentioned (subalpine-scrub species) together with rain-forest species, especially, 
— Asplenium bulbiferum, Blechnum lanceolatum, Polystichum vestitum, Histiop- 
teris incisa, Carpodetus serratus, Weinmannia racemosa, Melicytus ramıflorus, 
Fuchsia excorticata, Coprosma lucida and Coprosma Foetidissima. Within, is 
more or less Metrosideros lucida, i. e. the association is potential Southern-rata 
forest. Such an association forms the next belt which. extends upwards per- 
hapseto the scrub-line and 'marks a comparatively recent advance of the ice. 
| According to BELL, “probably not more than 150 years ago”'), the glacier 
extended 820 m northwards down the valley, depositing on its retreat exten- 
sive terminal moraines. On these, and the old river-bed, can be seen vege- 
tation at different stages of formation, the climax, so far, being a scrub about 
3.6 m high (Plate XXXIII, Fig. 45). On older moraine still, there is Metro- 
sideros lucida forest but the climax-association of the valley is taxad-forest 
‚with Daerydium cupressinum, Podocarpus ferrugineus, P. Hallii and the ordinary 
trees, shrubs, tree-ferns?) and ferns of the Western district rain-forest. 
At the greatest extension of the glaciers in the Pleistocene period, both 
On the extreme E. and W. of the Southern Alps, there would be peaks and 
slopes, not necessarily of great altitude, free from ice and still harbouring the 
Pleistocene alpine flora many of the species of which we may conclude were 
identical with, or closely resembled, those of today. But as such havens of 
refuge would be of limited extent, the struggle for existence would greatly 
reduce the number of species. Some certainly, if the glacier-extension were 
due'to elevation of the land and not to increased cold®), would migrate on to 
the Canterbury Plain &c., where, under intensified steppe conditions, they 
en RR 
1) A Geographical Report on the Franz Josef Glacier, Wellington, N. zZ, ia 
2) Hemitelia Smithii also occurs in the forest above the glacier. 
3) The presence of a tree-fern, HYemitelia Smithii, in one valley of the Lord Auckland Islands, 
‚-Oekayne, The Vegetation of New Zealand, 
