Go to National Library of New Zealand Te Puna Mātauranga o Aotearoa
Volume 38, 1905
This text is also available in PDF
(511 KB) Opens in new window
– 560 –

Art. LXIII. —Geological Notes on the Country North-west of Lake Wakatipu.

[Read before the Otago Institute, 14th November, 1905.]

The country to the west of Lake Wakatipu has been but seldom visited. It has, however, been twice reported on in a general way by geological observers.

Professor Park in 1887 made a flying reconnaisance through the district, and gave a general description of the geological formations he met with in the Reports of the Geological Survey. Professor Ulrich in 1890 published in the “Quarterly Journal of the Geological Society” a general account of the rocks of the district, compiled from examination of specimens collected by several explorers.

An expedition was made in January, 1905, by myself, with a party of students from the Otago University School of Mines, and Mr. A. E. Flower, of Christ's College. As the district is still without any tracks, and is traversed by mountain-ranges of an extremely precipitous nature, while the valleys support a dense growth of forest, it was recognised that difficulties greater than those that usually impede a geologist would have to be encountered. Fifteen days' provisions were taken, as well as an ordinary camping-outfit, so each man had to carry a load of 45 lb. at the start. The object was to cross over to the west coast via the Rock Burn, Hidden Falls, Olivine, and Pyke Streams, and to ascend the Red Hill Mountain of olivine formation. The weather experienced entirely prevented this result, and the expedition was unable to proceed much further than the head of the Olivine Stream. Geological and physiographical observations were made on the way, in the hope that some additional information might be gleaned as to the origin and nature of the surface features and structure of this out-of-the-way part of the country.

– 561 –

The valley of the Rock Burn proved to be in all its features one of the typical glacial valleys of the west of Otago. Bounded by precipitous mountains rising to 6,000 ft. or 7,000 ft., on the sides it exhibited well the typical U shape of such valleys. At no point is this valley entered by a tributary at its own level. Its waters are fed by tributaries that enter its valley over waterfalls. They are true hanging valleys. The floor of the valley is comparatively flat—about a quarter of a mile wide; it is covered with dense forest except in its upper portions, and everywhere it is strewn with immense angular blocks of rock, some morainic in their origin, others have fallen from the frowning cliffs above. Covered as they are with dense growth of ferns and forest, they offer formidable obstacles to the traveller, and the difficulties taken together often limit the rate of progress to half a mile per hour, even when strenuous exertions are made. From time to time steps in the floor of the valley were met with; over these the stream rushes with impetuous fury, but has cut but a narrow gap in the rock of which the floor is formed. These steps are result of a sudden increase in the erosive power of the glacier which filed out the valley. The increase is usually due to the addition of a further quantity of ice from a small tributary, which made a material difference to the weight of ice, and hence the power of erosion of the glacier beneath the junction of the tributary. The valley is about fifteen miles long, and in this distance its bed descends from an altitude of 3,520 ft. to 1,150 ft. where it enters the Dart. The valley terminates in a cirque, with precipitous walls on the north and east, but on the west the slope is more gentle.

On the east side the snout of a glazier terminates at an elevation of 1,200 ft. above the valle-floor. The glacier supplies the greater portion of the water of the Rock-burn Stream. The east side of the valley is usually 65° in slope. It is often a flat, bare surface coincident with the foliation plane of the schist.

The west side of the cirque is comparatively low, rising to only 4,490 ft. The west, therefore, forms a low pass over the Humboldt Range. The pass is only a quarter of a mile wide, and then descends precipitously to the valley of the Hidden Falls Stream, whose bed is here 1,000 ft. lower than that of the Rock Burn, the barometer giving a reading of 2,560 ft. above sea-level.

There is no bush in the Rock Burn Valley at a greater height than 3,000 ft., but on the east side of the Hidden Falls Valley it rises to 4,000 ft. The valley of the Hidden Falls is less U-shaped than that of the Rock Burn. This is mainly due to the huge slips and scree slopes that have broken away from the steep faces and litter the sides of the valley.

The Hidden Falls Stream obtains most of its water from

– 562 –

glaciers on the Barrier Range. The snouts of the glaciers are about 1,000 ft. above the valley-floor, and, as in the case of the Rock Burn, the waters from their melting ice tumble down precipices into the valley below. This ice fills hanging valleys from which ice was under other conditions supplied to the main valleys.

From the head of the Hidden Falls Stream there is a low, flat pass with mountain walls on either hand to the head of the Olivine Stream. On the west side a mountain wall rises without a break, but on the east deep valleys cut far into the hills, and each appears to terminate beneath the snout of a glacier, though we were unable to spend the time necessary to prove this in every case. There can be no doubt that all the flat of this deep pass was once covered by ice, for in many points of vantage on the mountain-side perched blocks, evidently of a different nature from that around them, are to be seen. A large mass of moraine extends across the top of the Olivine Stream. This dammed up the stream, and is the main cause of the formation of the wide flat area of the pass beyond it.

The Olivine Stream is in all essentials a duplicate of the Rock Burn. Its main features are certainly due to glacial erosion.

From the Olivine there is a low saddle leading to the valley of Lake Alabaster. The slope on the Olivine side is extremely steep, and that stretching towards Lake Alabaster also appears precipitous. On the saddle itself the effects of ice-action are clear and pronounced. Rounded and worn rock-surfaces are everywhere in evidence, though the actual grooves and polished surfaces are here, as elsewhere, somewhat indistinct.

The profound valley occupied by Lake Alabaster is entered by the Pyke River from the north-east. This river is, at a distance of twenty-five miles above Lake Alabaster, separated from the waters of Big Bay by a flat area of morainic matter no more than 200 ft. above sea-level. This morainic matter fills a wide valley between the Skipper's and McKenzie Ranges. The Pyke River, instead of following this obvious straight course to the sea, turns sharply to the south-west towards Lake Alabaster, joins the Hollyford River, and after a course of thirty-five miles enters the ocean near Martin's Bay.

Evidently some special explanation of this eccentricity is required. The explanation is to be found in a consideration of the movements of the ice of the great Hollyford glacier. This prehistoric glacier received the ice from twenty-five miles of mountain-ranges on either hand, and, judging from the depth and width of its valley, must have attained enormous dimensions. In the lower part of its course the ice-stream underwent “diffluence,” and a portion of it passed over a low saddle then occupying the site of Lake Alabaster, and entered the

– 563 –

sea at Big Bay. The ice that followed this course was of great quantity, and soon eroded the saddle away, and formed the deep basin of Lake Alabaster on its way. As in this low country the ice would be gradually melting away, a “reverse” slope would be formed, and if the terminal face were near the sea in Martin's Bay morainic matter would be deposited there. Finally, when the climate became more genial and the ice melted, the streams that took the place of the glaciers were blocked from Big Bay by the moraine, and flowed down the reverse slope to Lakes Alabaster and Mackerrow, whose waters were then continuous. Since then the detritus carried down by the Hollyford has advanced for and separated Lakes Mackerrow and Alabaster from one another. The course of the upper part of the Pyke, of the Olivine, and Barrier Streams clearly points to Big Bay as the original outlet for their water or their ice.

The complete filling of the Hollyford Valley with ice also appears to offer an explanation of the formation of the remarkable flat pass between the Olivine and Hidden Falls Streams. This would decrease the grade down the Hidden Falls Valley, for the surfaces of tributary and main ice-streams are always on the same level, consequently the flow of ice was partly dammed back, and some of it flowed over a saddle into the Olivine and eroded it to its present low level. Still further on some of this ice appears to have passed over gaps from the Olivine to the Alabaster Valley, where the level of the ice-surface, owing to the small amount of gathering-ground in that locality, and its proximity to the glacier-snout, was less.

One important fact in regard to the physiography of this region is the very different levels of streams on opposite sides of mountain-ridges. For instance, the Rock Burn floor is 1,000 ft. higher than that of the Hidden Falls Stream, on the other side of the pass. The Olivine floor is 3,500 ft. above Lake Alabaster, though only separated from it by a narrow rock ridge. Instances might be multiplied indefinitely. They seem to point to the probability that here some cause has completely upset the usual drainage conditions of areas of high land. Since this disturbing agent has ceased to act the normal relations of drainage valleys and systems have not been established, because sufficient time has not yet elapsed.

That ice is capable of just this disturbance that is here found I believe to have been fully established by geologists. To my mind no writer has more fully stated the peculiarities of glacial valleys than Professor A. Penck, of Vienna, and Professor W. M. Davis, of Harvard. More especially has this been done lately in the “Journal of Geology,” where Professor Penck, writing on the “Glacial Features on the Surface of the

– 564 –

Alps,” has described facts that are point for point repeated time and again in the glacial regions in the south-west of Otago. It is inconceivable to me that earth-movements, sudden or gradual, could account for the many peculiar physiographical features which have been alluded to in the previous pages.

The general geological structure of the district is indicated in the small sketch-map on p. 566. It is here advisable only to state that the district is one of mica-schist and phyllites, with a dip and strike that varies little throughout the area. Their age is still doubtful, for while Captain Hutton has lately classed them as of Arcæan formation, Sir James Hector has classed them as Silurian, and other divergent statements are not wanting. Through this schist there is intruded an igneous mass which outcrops at the Olivine Saddle, as afterwards described.

The following rock-types have been distinguished, and their petrographical characters are described:—

Diorite.—Bryneira Saddle, between Olivine Creek and Lake Alabaster. Hand-specimen medium-grained. Hornblende and feldspar can be distinguished. Section: All the feldspar is completely saussukitised, and the hornblende is uralitic. The exact original nature of this rock can only be guessed at.

Gabbro.—Cow Saddle. A coarse-grained rock showing decomposed feldspar and large cleavage-surfaces of diallage. In section the diallage is fairly fresh, but the feldspar is completely changed into saussurite.

Pyroxenite.—Cow Saddle. Cleavage surfaces of diallage numerous and conspicuous. The mineral appears to constitute nine-tenths of the rock. In section the spaces between the diallage-grains are small, and are entirely filled with serpentine, mingled with which is some pyrite.

Lherzolite.—Cow Saddle. A pale-yellow rock showing many cleavage surfaces of a pyroxene imbedded in the olivine. Sections show large irregular grains of olivine nearly fresh. Pyroxene, both monoclinic and orthorhombic, is present. The former is a very pale-green, and is referred with certainty to diopside. The orthorhombic pyroxene is quite colourless, and is certainly enstatite. In addition there is a fair quantity of a yellow mineral with a high index of refraction. It is far more transparent than the chromite of the dunite of Nelson and of Milford Sound, and is probably a chrome picotite. The mineral has not yet been separated for analysis. This description of the rock will be found to agree satisfactorily with the description of rocks from the Red Hill by Professor Ulrich.* In this description,

[Footnote] * Quart. Journ. Geol. Soc., vol. xlvi, 1890, p. 627.

– 565 –

however, the green mineral that was evident in hand-specimens was not represented in any of the sections, and Professor Ulrich provisionally referred it to enstatite, which was present in quantity. Ulrich referred the rock to the group of saxonites of Wadsworth—that is, the hartzbergites of Rosenbusch and other authors. The demonstration of the fact that the green mineral is diopside, which is easily done in the sections now before me, justifies the classification of this rock with the Iherzolites. A section of this rock was sent some months ago to Professor Rosenbusch of Heidelberg. The section, however, was in some respects exceptional, for but little olivine was present; and Professor Rosenbusch stated that the rock was a websterite, a connecting-link between the hartzbergites and the pyroxenites. A partial cataclastic structure was also remarked upon. Seeing that in the greater part of the rock olivine is very abundant, I feel justified in classing the rock with the Iherzolites in spite of the statement of the eminent authority. I would wish, however, to put on record my deep sense of gratitude to Professor Rosenbusch for the kindness and assistance that he has so readily extended to me.

Dunite.—A large mass of the rock is of a dark-slate colour, weathered on the surface to a dull-brown. In hand-specimens it appeared perfectly dense; sections, however, showed at once that the rock is really a dunite in process of serpentinisation. A little chromite and pyroxene are present, but the greater portion is a mass of olivine-grains traversed in all directions by small veins of serpentine, forming a very complete mesh structure.

The arrangement of these rocks is shown in the plan on the following page.

The weather effectually prevented an exact demarcation of the boundaries of the different rock-types, and the approximate boundary of the Iherzolite was made out observing the colour of the rocks where they emerged above the snow in the distance.

It will be noticed that the igneous rocks are bounded by schists on the east and by the Te Anau breccia on the west. The latter, however, may quite possibly be the effusive type and associated fragmentary rocks of the diorite magma adjacent to it. From the boundary of the diorite there is a regular increase of basicity until the lherzolite is reached. At the south end of the area the lherzolite is certainly in direct contact with the schist on the east side. So far as could be seen the division-lines between the different rocks correspond in direction and inclination with those of the schist. If this should finally prove to be the case, further work will be necessary to decide whether the igneous mass was a laccolite that has since been subject to the same

– 566 –

folding movements as the surrounding rocks, or whether the mass is an intrusion formed subsequent to the main rock-movements of the region. The apparent abrupt change on the last from lherzolite to schist, and the gradual change of rock-type to the west, seem to support the former view. At any rate it is apparent that here there is abundant material for studying the differentiation of a magma from which rocks varying from diorite to lherzolite have crystallized. However, the exceeding remoteness of the locality, the roughness of the country, and the great difficulty of obtaining supplies render it unlikely that the district will be revisited for some time, hence it was considered advisable to place on record even these vague and ill-determined facts.

The probable transition from the peridotite type to gabbro was already supposed by Professor Ulrich, from information supplied by Butement, who had gathered it from observations made nearer the coast-line.

– 567 –

The expedition to this region was made with the object of making observations on the nickel-iron alloy awaruite; but so far as that object was concerned it was without result. No awaruite was found, and no nickel could be found in the rock when chemical tests were made. It appears, therefore, that awaruite does not occur in the periodotite rocks of the south end of this extensive magnesian region.

The relationship of this area to the magnesian rocks of Milford Sound, described by me last year, is not a very close one, though of course the two olivine rocks belong to the same group and occur in the same petrological province. The dunite of Milford Sound contains much darker chrome-ore, and is almost destitute of diopside. The hartzbergite of Milford Sound contains a chrome-bearing magnetite, but no picotite or chromite or diopside. In the present peridotite the picotite is palecoloured, diopside is abundant, and enstatite less frequent.