16

C—ll

fell continuously in such numbers as to make a sound like the patter of raindrops on the roof of my tent. Nothofagut Menzie.sii is slightly less xtrophytic than N. cliffortioides, while N. fusca, with its thin leaves, is distinctly hygrophytic. Thus, according to increasing altitude the forest zones become more xerophytic, the strongly hygrophytic mixed taxad forest at 2,000ft. altitude, with its tree-ferns, woody lianes, and epiphytic asteliads, offering a great contrast to the mountainbeech forest 1,500 ft. above, with its undergrowth of shrubs, many identical with those ..I the wind-swept subalpine scrub. Leaving the zone of A', fusca out of the question for the present, the under-shrubs of thti upper fo'rest have frequently coriaceous leaves, and in addition certain of them are of the markedly rerophvtio divaricating type already described in the case of the very typical Suttonia divaricate. Ke.rophily is likewise shown in the filmy fern Hymenophyllwm multifidum having its leaves so frequently curled up, a condition I have already proved by experiment to be brought about directly by excessive transpiration. The elegant fern Hypolepis millefolium is truly herbaceous, the brightgreen fronds dying to the ground in winter; and Polystichum vestitum and Blechnum penna manna have each Coriaceous leaves. Also, it must be pointed out that where the same species appear as both forest and subalpine scrub or shrub-steppe plants, so different are they at times thai they might be considered by one unversed in the flora as different species. How much shade- ,. litiom can alter form I have shown elsewhere with regard to the 'South Island form of Pittosporum rigidum (5). This plant is, out in the open, of the most dense and divaricating habit conceivable, but in the forest its leaves are pinnatitid and not entire, thin and not coriaceous, and its habit is twiggy and open. Still more interesting is the behaviour of Aristotetia fruticosa. This, in its juvenile form, exhibits a most remarkable leaf-variation, but in the open usually it finallv becomes a divaricating sh»ub with small leaves and Leanest branch-apices which are semispiny. But almost invariably in the forests of the volcanic plateau the plant in question is found with comparatively broad leav< b, and remains of a hygrophytic rather than a xerophytic form. So, too, with Corokia Cotoneaster, a very rare plant for this locality. ' In contradistinction to the xeiop'hvtic form is the hygrophytic, which is exhibited by several of the comm..nest plants. Thus there is nothing xerophytic about Coprotma fcetidissima and C. ten ii!folia. In fart, the competition of the forettt depend* far more upon the history of the vegetation i.e.. "linn tin plants which by chance mine to settle down on the neie ground—than n/>on ami special adaptations lime may hair posteued. Probably, the main requisite was the physiological one of their frost enduring' limit. Any xerophily would, however, stand them in good stead on fust settling down on the new and excessively porous ground, while when coming finally into moist forest-conditions they assumed at once, according to their plasticity, more or less hygrophytic The light-relation regulates the density of the undergrowth. This is well shown by the powerful crop of saplings wherever a few of the old trees have died. Where more light still can pene trate, the undergrowth ai once becomes abnormally thick, and certain plants enter in which are usually uncommon in the formation, especially Hie liane Eubtu auttralU (bush-lawyer) and the aggressive herb Acama Sanguisorba (piripiri). Increase of moisture also changes the forest - charaoter, a sha.lv hank or the bottom of a gully having a richer vegetation, and there certain species may appear which are absent elsewhere. The soil factor of course is of great importance regarding the undergrowth. Within the forest it consists of a surface layer 2 in. or more deep of fallen beech-leaves, partially or altogether decayed, mixed with rotted moss-cushions, decaying twigs, and so on. Beneath this is a sandy soil with little or no humus, in which the various elements pumice, scoria, and lava are mixed together, and beneath this again at a varying distance a certain amount of sandy clay and stones. Su.h a soil becomes excessivelj dry in summer. Even digging in a gully will usually fail to teach water. It is the capacity of the top layer of decayed ami decaying vegetable matter to hold water on which the plants must in considerable measure rely, and this moisture, which is renewed even by the slightest shower, encourages the growth of low moss cushions or mats, these playing a notable part M soil -makers and water-conservers. One especially, dying and decaying as it glows, builds up broad low cushions of a yellowish-green colour, only the peripheral portion being alive for a trifling depth. Through such moss stoloniferous plants penetrate; young seedlings find there the moisture they require, and finally, acquiring sufficient vigour, can penetrate into the sandy ground, sending down long roots and growing eventually into thickets. Regarding the (lowers of the forest-plants, more than 70 per cent, have insignificant flowers, mostly of a dull colour, while about fifi per cent, have unisexual flowers. Alseuosima macrophylla, Myrtm peduncvlata, Nothofagus cliff ortioides, Lagenophora petiola, and Styphclia arerosa are the only members which can lav claim to any degree of showiness, and of these only the Alseuotmia and beech are highly coloured, the remainder being white and small. Tn many cases wind will play the chief part in cross-fertilisation, but in some instances — Astelia montana, <•.;/., —flies will lie the fertilising agent. Heterophylly, so common a phenomenon amongst Now- Zealand plants, has been already discussed in the case of certain divaricatingly branched shrubs. Nothopanax simplex is another important example. This small tree is especially distinguished by its having two quite distinct juvenile forms. The more common of the two has a compound deeply cut leaf, which, to be sure, varies a good deal in form, depth of cutting, A-c. The second form is a tern ate or simple leaf with merely toothed leaflets. At one time I felt assured that the two forms were related, and that each would appear at a definite stage in the life-history of the plant. But this does not altogether appear to be the case, the two fornis growing side by side and thus under the same environment.*

* Pertain specimens which I collected, however, have simple-toothed, ternate-toothed, and ternate-pinnatified leaves on the same plant, these fatter in some cases showing transitions towards a toothed margin. k The toothed and cut leaves therefore arc fur from l>ein<j fixed and constant structures.