of appropriate age. But the pole stands along ridge crests in the Mill Creek catchment are between 100 and 140 years old. They are on exposed sites marked with forest dimples and occasional traces of old beech stumps and are, therefore, believed to occupy old windthrows. The seedlings and saplings that are usually present at the junction of beech forest with shrubland or grassland are often not backed by any substantial number of poles, indicating that though there is a tendency to slow marginal advance it is periodically nullified by fires sweeping along the forest edge. However, in a few places there are well-established marginal pole stands which do not appear to be on land recently in beech. One at 1,500ft above Woodside occupies what 50 years ago is stated to have been a snow-tussock spur, and still contains moribund plants of flax. The Silver Peak forest has, in places, a pole stand 2–3 chains in width at its margin.

In the South Waikouaiti R. there are areas of pole beech containing large numbers of dead and moribund broadleaf (Griselinia littoralis) and kaikawaka indicating a recent advance of beech into kaikawaka forest, and beech seedlings can be found in gaps in podocarp forest where the two adjoin. In the Silver Peak beech forest also there are moribund kaikawaka trees, but they are surrounded by old beech. It appears here that beech has been dominant continuously, but the forest has passed through a phase favourable to a limited invasion by the conifer. We can conclude that, outside the main coastal block of podocarp forest, silver beech is and has for over a century at least been capable of vigorous regeneration and of slow marginal extension usually into snowgrass, but in some areas into coniferous forest.

Conifers and accompanying Dicotyledonous Trees.

Table II indicates the regenerative capacity of the dominant species of podocarp-dicotyledonous forest by listing the average number of germinated seeds, seedlings, saplings, poles and living trees occurring per ¼-acre (0.1 hectare). The series of sample plots on which these figures are based were not standardized in shape or area but were chosen to convey as fair a picture as possible of the general situation. Their total area was about six acres.

These figures suggest that both the mountain and the true totara are maintaining, perhaps even increasing their density, and that matai is vigorous in the establishment of seedlings but poorly represented in the young tree classes. But over the wetter part of the lowland forest the principal podocarps have been rimu (Dacrydium cupressinum) and miro (Podocarpus ferrugineus). They now exist as widely spaced emergents above a continuous canopy of broadleaf and smaller dicotyledonous trees. Partial milling accounts for much of this thinning—at Woodside, for instance, where the original stands are said to have been dense—but it is questionable whether it accounts for all of it. The present inadequate regeneration rate may be of long standing.

A feature of all kaikawaka stands is that there are nearly half as many still identifiable dead trees as live ones. As this small tree is not particularly durable it is clear that the stands have recently been thinning fast. Seedlings are common, but an impression that the species is migrating to a lower altitude is given by the fact that sapling and pole classes are absent except in the belt transitional with wet lowland forest. Here there is vigorous regeneration of broadleaf and some of miro.

Though old logs of pink pine are widespread in the snow-tussock country living trees occur only east of Leith Saddle—i.e., in the Mt. Cargill area. Here it appears to be stable, as does the mountain toatoa (Phyllocladus alpinus) whose main occurence is in this region also.

Broadleaf is the most widely ranging of the larger dicotyledonous trees, and is nowhere failing to establish seedlings though outside the 35in isohyet these are not plentiful. However, saplings and poles are not always present. Nowhere have seedlings or saplings of the pokaka been observed. This suggests that it is becoming extinct locally and that the rarity of kamahi (Weinmannia racemosa) and complete