107

α-i

built track, owing to the somewhat shattered nature of the rocks and the frequent heavy winter avalanches. I came over a saddle slightly west of FitzGerald's Pass, aud consider that a more practicable line might be found on the spur west of his route from the Hooker Glacier to the small ice-field at the top of the ridge, after which the travellers could be taken to the left for, say, half an hour to FitzGerald's Pass, from which the descent is very easy. In any case, it will be a matter of considerable expense, necessitating, I believe, much solid masonry, and a telephone-wire between the termini. Though the small ice-field is a drawback, it does not form an insurpassable obstacle, as I know passes in Switzerland with longer stretches of ice over which horses are taken. But it will probably be the source of delay and expense keeping a way open after snow even in the summer. For the present requirements of the country it appears to me that if a good horse- or even foot-track is made up the Copland River on the western bank to Welcome Flats, and thence on the eastern bank to the grass-line near the Marchant Glacier, or even a little beyond, and if the ascent on the Hermitage side is made easier, that tourists could, with little exertion beyond an hour or two's walking, reach Mr. Scott's house in one day comfortably. The expense of a horse-track on the Copland side would not be great, but it is in descending the Hermitage side that the heavier cost would be incurred ; and, even though a horse-track is not attempted, a telephone-wire would be required to communicate with the West Coast to have horses brought up the river, and vice versa for the Hooker Track. We cut good blazed tracks when in the valley above Welcome Flats, the entrances to which are marked with crossed sticks whenever the river-bed is too rough to travel on, and, given fine weather, a good walker should be able to reach Mr. Scott's house in eighteen hours. If the weather is fine and the river low the best line to take below Welcome Flats is to ford the river on the flats, and keep down the west bank, where the travelling is far easier, provided a party keeps back from the river on the terrace. Above the flats there is no difficulty, as our blazed tracks should be easily picked up for three or four years to come by any one with ordinary powers of observation. The map will have the route marked, which I recommend foot-travellers to follow until some further steps are taken to open up a track. For studying the action of ancient glaciers on rocks and hills the Karangaroa and its branches give as good.opportunities as any district I know, and there is a fine field for a scientific man, practically untouched, since I am only able to form theories and opinions, which are necessarily of little value from a scientific point of view. From the inflow of the Copland River to Cassell's Flat the valley is, roughly speaking, a mile broad, filled with glacial deposit, which descends in gentle slopes from the lower part of the hills to the centre of the valley. Through this moranic drift the river has cut a channel, leaving terraces on each side from 20ft. to nearly 100 ft. high, while the course is full of large erratic boulders, in some places completely blocking the valley. The top of the terraces is generally flat for some chains back, cut through here and there by deep channels, and in places covered with large boulders. At the end of the spur opposite the forks the glacial drift is piled up to the height of some 400 ft., while the spur itself above and behind the drift shows in places smooth ice-worn rock. The timber is chiefly kamahi, until the lower hill-slopes are reached, when it changes to totara, cedars, and rata, the latter prevailing, but up Cuff's Creek I saw some good patches of pine forest. With regard to the travelling to Cassell's Flat, a horse can with little trouble be taken to within 20 chains of Cuff's Creek, after which to Haskell's Creek is good open going for foot-travelling, rough in places, but nothing out of the common. We put a blazed track from there to Poison Camp and another from Poison Camp to the flat, along the top of the terrace, to use when carrying heavy loads, as the boulders in the river for some 40 chains are bad ; this latter track is, however, really unnecessary, and was only cut as the men who took Douglas's place were unused to rough boulder work. At Cassell's Flat the valley still retains its width, but is free of all the drift and debris which has been piled up below, and the hills descend nearly sheer oh to the level ground, showing magnificent bare faces of rock at an angle of about 70 degrees up to the height of 3,000 ft., on which the scrub has little hold. Here three branches mcct —namely, Regina Creek, draining the lower end of the Sierra Range with the ice-field off V 11.5 ; the Twain River, which drains the immense ice-field of the Douglas Glacier off Mount Sefton ; and the two other glaciers above mentioned —namely, the Horace Walker and FitzGerald; and lastly, the Karangaroa branch from the saddle into the Landsborough Valley, draining no ice-field beyond a small one on Mount Fettes. Of these branches the Twain River is by far the largest, and in reality is the main branch of the river. Coming as it does from the dividing-range, and bringing down a large volume of water, in the summer it is rarely fordable even at the forks, and above that it is unfordable for its whole course. Each branch descends through narrow gorges blocked with immense boulders, which form fine cataracts, of some 400 ft. descent in about a quarter of a mile. The gorge of the Twain I have already mentioned as being especially striking. The north side, on which I was twice brought to a stop, reminded me of the north side of the Francis Joseph Valley, in so much as there are bluffs running from the top of the range to the bottom, like those at Points D and Arch Creek, on the Waiho, facing up the valley, and forming blocks to the flow of ice. I should not like to state that ice had not come through the Twain Gorge in ancient days, because the round hill between it and Regina Creek is evidently formed by ice-action, and the north bank of the gorge is distinctly ice-worn. But I believe there was comparatively little iceflow through the present gorge, which seems more like a cleft in the hill by some convulsion of nature than an ice- or water-cut channel. The main flow of the old glacier was, I think, over the low saddle behind the rounded hill, and into the Regina Creek basin, joining forces with another but smaller glacier from V 11.5, and then flowing through the present Regina Creek Gorge into the great glacier which came off the dividing-range over the Karangaroa Saddle. The Regina Creek cataract is as fine as the other two, but it has not such a good gorge, the mouth of the valley being U-shaped, and full of immense erratics, either in the.course of the stream