are to the south; but we got a few trumpeter and sea perch of excellent quality; and also several small species of fish, which I have not seen elsewhere on the coast. On the 17th August, there having been several days of fine weather with S. E. wind, I made another attempt to examine the Cleddau River, taking with me three men, a tent, and provisions for some days. The woods were very dry and pleasant, and the stream so much lower than during the previous weeks, that we were able to skirt it in many places where I had previously to wade across it.

Following up the middle of the three branches which join to form this river, by evening we had made about 8 miles—the latter part of the journey being very rough work, on account of the great size of the boulders which block the channel, and over which we had to scramble at the risk of slipping into the torrent; this did happen to two of the party, but fortunately with no worse result than a thorough drenching in the icy water.

The fall of the river is very great; and the bed of the stream is everywhere composed of glacier detritus, sometimes rudely stratified, and filling the valley to the height of 1500 feet above the sea level, the immediate river valley being excavated between this accumulation and the steep smooth wall of rock against which it rests.

Next day we followed up one of the branches to its source. The upper part of its valley is cut to the depth of 540 feet, through a true moraine, consisting of earthy clay, containing regular blocks of rock of all sizes up to 30 and even 40 feet in diameter.

The stream ends quite abruptly against a glacialised surface of rock, which slopes to a height of 3000 feet, at an angle of from 30° to 40°. The snow, which falls from the mountains, is unable to lie on this polished surface, and sliding down, wedges in at the back of the moraine, forming a miniature glacier, though without the true ice structure, at an elevation of only 1000 feet above the sea level. The depth of the grove, which has been cut by this snow bank between the rock and the moraine, is not less than 400 feet.

By a slightly dangerous climb, we got up the glacialised surface of the rock and on to the top of the great moraine, which is heaped up against it. The frequent landslips which take place from the face of the moraine cliff does great havoc among the trees that grow on top, leaving their roots bare, so that they die, and are easily thrown over. The forest is very open, and some of the trees are of good size. The principal trees which I observed at an altitude of 1800 feet were the black birch, the iron wood, or batta, the remu, totara, cedar (a second species of Podocarpus), broad leaf, New Zealand holly (Eurybia dentata), moka, and several others.

At this altitude, on westerly exposures there are few lichens, or mosses, as the woods are well aired and the soil dry. We were now in the third great longitudinal valley, which runs north and south, crossing the main valley, which is continuous with that of the Sound. As these valleys conform to the trend of the strata, they probably indicate lines of softer rock, along which the erosion was more easily effected by the descending glaciers. In these valleys, the moraine matter is heaped principally on the eastern side, being opposite to that upon which the greatest accumulation of ice must always have taken place.

The study of these deposits which fill these valleys possesses great practical interest from their intimate relation to the gold drifts on the eastern side of the mountains.

Although the mountains rise so precipitously from the valleys, there are not so steep towards their summits, where there is generally a large area, presenting slopes on which snow could rest under circumstances favourable for its accumulation, and form the source of glaciers which would descend into the lower valleys. Pembroke Peak (6823 feet) is covered with perpetual snow, which on its south-east face extends as low as 4000 feet, with a slope of 20° to 30°, and there terminates in a cliff of true glacial ice, judging by its intense blue tint compared with that of the surrounding snow; and did it not overhang a precipice, this ice would doubtless descend as a glacier to a very low altitude. Now the average height of the mountain ridges is nearly 6000 feet and with the present conditions of climate, an elevation of the land equal to 2000 feet would, according to the best estimate I can form, raise about six-tenths of the area of this mountain district to that altitude, which is certainly considerably above the snow line in the strict sense, or from where the snow never disappears during the Summer, unless by evaporation after assuming the glacier form by regelation.

It is a mistake to estimate the size of glaciers generated from a mountain range merely by its altitude, as it is truly the area which in the district is elevated above the snow line that determines their extent. If this be the case, the area must always be diminishing rapidly, from the eroding action of the descending ice, and therefore the extent of the glaciers must also diminish. Judging from the structure of the Sounds on the west side of the mountains, and that of the Lake district on the east side, I am inclined to think that the opposite sides of this mountain range have undergone repeated and alternate oscillations to the extent of at least 1000 feet in either direction from a nominal point; and that the Western district being at present near to the period of greatest depression, the re-elevation of the land to the other extreme would be almost sufficient to extend the glaciers to their