C—ll

26

where little <>r no vegetation of any kind is present, a small spring of water issues from the ground, trickling over and keeping the substratum constantly moist, Here is a rich collection ef nlaiits (Photo. No. '23). Quite at a distance this oasis can be seen, and the idea thai strikes the observer a< first is that it is a deposit of sulphur. The yellow colour, however, arises from a moss of that hue, while great masses of the mountain water willow-herb (Epilobiuni macropus) are present in January, a really beautiful sight, covered with the large white flowers, pink in the bud. Blooming at the same time are quantities of the large-flowered Craspedia (C. uniflora var. robusta), the heads cream-coloured ami quite showy in the mass. Here, too, i> Montia fontana, Carex pyrenaica, Poa nova-tealawlias, Agrostis Muelleri, and a small Epilobium of the pedunculate type —mostly species extremely rare or wanting elsewhere in the district. There is a turf of Coprotma repent. Present also are Galamagrottit setifolia, Orenholus pectinatut, Heliehrysum bellidioides, Ligusticwm aromatieum, and Gnaphalium paludosuvi, another turf-builder, silvery in colour. Hut more remarkable still is where from beneath the great scoria desert at the base of Xgauruhoe there issues forth on to the scoria plain forming the bed of a branch of the Waihohonu a number of streams. These, keeping the scoria damp in their vicinity, actually favour the growth of bog-plants, and led masses of Drosera tpathulata, as noted above, are dotted here and there all over the black scoria. Where stones are scattered over the deserts the vegetation gets a better ohanoe, as these not only provide shelter, but keep the ground moist beneath. Veronica Hookeriana is a plant especially of such a position. Heliehrysum bellidioidet forms considerable sheets on the ground under like conditions. Finally, under the lee of the larger stones are combinations of the steppe-shrubs. The desert is in part a primary and in part a secondary formation. The scoria slopes, with or without rocks, are a geological feature at a certain stage in the evolution of the earth-surface. Hut where landslips, " wash-outs, , ' m , broad river-beds have destroyed the plant formations a secondary desert will arise— i.e., we have a reversion. Also, in some places there is a gradual movement of large areas of scoria and sand which bury the shrubby steppe or the more fully occupied desert (Photo. Xo. 'i I), and here again ,1 rejuvenation takes place. Hut it is, in all these instances, no new formation thai occupies the ground, but merely a reversion. New formation* alone occur where introduced animals and plants can gain a foothold, the former destroying or damaging certain species, thus upsetting the balanct of nature, and the /after bringing in new competitors in the struggle for existence. So far as the deserts of the volcanic plateau are concerned, there are no changes of this kind of any moment. (•"5. ) The Onetapu Desert. The Onetupu desert, being in part situated near the coach-road, is the best-known and at the same time of a somewhat different character to anj of the other barren tracts. Situated on the highest part of the eastern plateau, the surface of the ground is exposed to the full fviry of the wind. The melting snows and rain also cut furrows into the ground, undermining the surface, ami exposing the bare ground to the violence of the gales. In consequence, a large area has become in course of time a sandy waste. That this has not been so always is shown by the curious mounds here and there, covered with grass or shrub steppe, which stand like islands in the sea of sand (see also Colenso, 8, p. 40). All transitions are seen, from such to low mounds where dead stems and roots alone persist. At its lower portion many acres of the desert are quite without plant-life — stretches of sand alone exist, or. where this has blown away, there is consolidated pumice or soft rock. Here lie scattered the bleached remains of former shrubs. But in other places a few plants have gained a foothold. Fire) comes Danthonia temianwularis var. setifolia, and the sand backing up against it raises low mounds. Single plants may stand far isolated from one another, and. becoming more numerous, a colony is formed, other plants entering in— e.ij., the silvery mats of Raoulia austrails, the vivid-green semi-cushions of I'inielni lavigata, and the stiff, erect." pale-green, flattened shoots of Carmiehaelia Enytii var. orbieulatq. As Ruapehu is approached the plant covering becomes denser ami denser, the tussocks closer and closer, until finally an open grass-steppe results. In some places the surface is covered with numerous small pieces of lava, and in others the surface is all pumice, many pieces, large and small, lying on the ground. On this latter, Gentiana bellidifolia is very abundant. This desert is by no means only a feature of to-day. It changes its position from time to time, but at the time of Coknso's visit it evidently had exactly the same appearance as at present. (4.) Rock-vegetation. The vegetation of tlie rocks is usually very scanty, and consists chiefly of those desert plants which can insert their roots into the cracks, such as Oavltheria rupestris, Danthonia semiannularis var. setifolia, Ligusticum aromatieum, Helichrysum bellidioides, Veronica spathvlata, all of which were noted growing on a hard lava cliff on Huapelm at about 7,000 ft. altitude. Among the earliest plants of rocks arc mosses, especially species of Andniia. and lichens, both of which form soil on which seds of ipermaphytes germinate. These latter plants arc usually rather denizens of the chinks, hollows, and depressions than true chasniopliYtes. The lava-Hows may lie either- solid rock, or consist of great blocks, the one piled upon the other. Certain Hows at a high altitude on Ngauruhoe are quite without plants, not even a lichen being present. Others, such as the one in the central crater of Tongariro, are occupied by very few plants indeed, and these are chiefly on the terminal face, Veronica spathulata and Danthonia nt mi annularis var. setifolia being tho principal occupants. The older lava in the Oturere Crater has a richer plant covering, and on it are most of the ordinary desert plants and shrubs growing amongst the blocks of lava. Perhaps the most interesting lava-flow is that from Tβ Mari through the forest, which consists of immense blocks piled on one another. The vegetation there depends entirely upon the shelter afforded, and whore this is strongest is a tin.' subalpine scrub made up of the usual con-