Go to National Library of New Zealand Te Puna Mātauranga o Aotearoa
Volume 42, 1909
This text is also available in PDF
(5 MB) Opens in new window
– 589 –

Art. LXI.—The Great Ice Age of New Zealand.

[Read before the Otago Institute, 22nd September, 1909.]

New Zealand geologists and physiographers have always agreed that there was a period of intense refrigeration in the South Island in very late Tertiary times, and all have agreed that this refrigeration was caused by an elevation of the land. There has, however, been some disagreement as to the extent and date of the glaciation caused by this refrigeration.

With respect to the extent of the glaciation, Mr. J. T. Thomson and Sir Julius von Haast maintained that the whole of the South Island was covered with a continuous ice-sheet; while Sir James Hector, Captain F. W. Hutton, Mr. A. McKay, and Mr. W. T. L. Travers held the view that the main Alpine divide was occupied by an immense icefield from which gigantic glaciers descended to the sea on the west coast, and on the east coast reached the Canterbury and Southland plains. The glacier which occupied the Taieri Basin from near Dunedin to the south of Milton was admitted by Hutton to have crept over the coastal hills to the present coast-line. In the late “seventies” Haast modified his first view as to the extent of the glaciation. Up till the year 1907, but without having made any special review of the evidence, I gave a general support to the contention of Hector and Hutton.

The Glacial period of New Zealand was believed by Hector, Haast, and McKay to have taken place in the Pleistocene—that is, contemporaneously with the Ice Age of the Northern Hemisphere. Hutton, on the other hand, placed it at a somewhat earlier date—that is, in the Pliocene; but in my review* of the “Marine Tertiaries” of Otago and Canterbury I showed that the Greta and Awatere beds of the South Island were older Pliocene, and not Miocene, thus removing Hutton's difficulty as to the Post-pliocene glaciation.

During the progress of my geological survey of the Wakatipu region in the summer of –8, among new evidences of glaciation I discovered that the mountains on the north-east side of the lake were covered with a stony boulder drift up to the height of 4,000ft., mainly composed of erratics transported from the Livingstone Mountains, and that the mountains were ice-shorn and cut into shelves up to a height of 6,500ft. above the sea. As a result of my observations, which are fully recorded in my bulletin on the Wakatipu district, I estimated that during the glacial period the ice formed a continuous sheet that stretched over the greater part of the South Island, in the Wakatipu Basin attaining a thickness of over 7,400ft. In support of my contention that the ice-sheet reached the sea on the east coast, I referred to the glacial boulder-clays near Dunedin, the glacial drift overlying the coastal hills between the Taieri Plain and the sea, and the loess at Oamaru and Timaru.

[Footnote] * J. Park, Trans. N.Z. Inst., vol. xxxviii, 1904, pp. 536 and 547.

[Footnote] † J. Park, “Geology of Wakatipu Area,” Bulletin No. 7 (New Series), N.Z. Geological Survey, 1909.

– 590 –

I ascribed the cause of the glacial period to an elevation of the land, which I estimated at not less than 3,000ft. Reviewing all the evidence, I expressed the opinion that during the period of maximum elevation it was not improbable that the polar ice-sheet reached as far north as New Zealand.

In my presidential address* to the Otago Institute in May, 1909, I reviewed the evidence submitted in my official report, and reaffirmed my belief that the glacial period of New Zealand was Pleistocene, and in all respects parallel with that of Northern Hemisphere. And, arguing from the evidences of glaciation in the Auckland, Campbell, and other islands lying south of New Zealand, in Tasmania, and in South America, I hazarded the opinion that all the land-areas in the Southern Hemisphere up to corresponding latitudes were glaciated at the same time as New Zealand. I further suggested that during this great southern Ice Age there was probably an invasion of the Antarctic ice-sheet, in the same way as in the Glacial epoch of the Northern Hemisphere there was an advance of the Arctic ice-sheet.

A close scrutiny of the topographical features of those portions of Southland, Otago, and Canterbury lying on the east side of the axial divide discloses evidence on all hands that the present configuration of the surface has been moulded by ice erosion.

The slopes and summits of the Hector and Eyre Mountains, overlooking the South Arm of Lake Wakatipu, are ice-shorn up to a height of 5,500ft., the smooth, flowing outlines of the former in particular standing up in conspicuous and striking contrast to the sharply serrated and rugged summits of the adjoining Remarkables, that stood above the surface of the Pleistocene ice-plateau.

The Wakatipu region was the centre of movement of the ice-sheet which flowed southward across Southland and eastward across Otago, the track of the ice being plainly marked in a thousand places.

The smooth, flowing contours, dome-shaped summits, truncated spurs and ridges, and U-shaped valleys that are so well developed in the Wakatipu district can be traced from the Hector Mountains, opposite Kingston, across the Hokonui Hills to the Southland Plain, from which rise the Bluff Hills, themselves merely gigantic ice-shorn roches moutonnées. From the Upper Nevis, Nokomai, and Mataura the same smooth, rounded outlines can be followed eastward to Gore, and thence northward to Otago Peninsula, without a break except where recent fluviatile erosion has modified the glaciated surfaces. For example, near Athol the Mataura River and its branches have eroded deep V-shaped channels in the older glaciated slopes; and so sudden is the change from the smooth, rounded contours to the characteristic tent-shaped ridges and V-shaped valleys that the contrast but serves to emphasize the startling difference that exists in landforms modified by ice and by running water—a difference so fundamental and conspicuous that it needs not the trained eye of the topographer to discern it.

Sir Archibald Geikie, in his charming book on the “Scenery of Scotland,” when discussing the origin of the smooth, rounded outlines of the Scottish mountains, asks the question, “What agency other than ice is capable of producing these forms?” The ice-worn outlines have been traced down to the present coast-line between Southland and Dunedin.

[Footnote] * James Park, “The Physiography and Glaciation of Otago,” Otago Daily Times, 12th May, 1909.

– 591 –

And it may be as well at this point to inquire what other evidence we have that the ice-sheet reached thus far eastward.

At Munro's Gully, Blue Spur, and Weatherstone's, in the lower Clutha, there is a thick deposit of fluvio-glacial drift resting on a mica-schist floor. The drift is mainly composed of rounded boulders of greywacke and angular masses of mica-schist. The greywackes are erratic, and must have been transported from a distance. The mica-schist is local. Hutton thought that the parent source of the greywacke was the Tapanui Range, lying some twenty miles to the westward; and in this surmise I think he was probably right.

The well-known Taieri Moraine forms the front of the coastal range lying between the Taieri Plain and the sea, extending from Mount Misery, south of Milton, to Allanton, a distance of twenty-five miles. It attains its greatest development between Waihola and Otokaia, where it forms hills varying from 250ft. to 700ft. high.

It is easily divisible into two unequal parts—namely, an upper and a lower series. The upper division is mainly composed of coarse angular fragments and blocks of semi-metamorphic mica-schist, mingled with gritty silt and a certain proportion of semi-rounded schistose gravel. The whole mass is partially consolidated into a breccia-conglomerate. The pervading colour is light red.

The thickness of this division is some 450ft. In the upper portion of it there occur many masses of schist ranging from 6ft. to 12ft. in diameter. These are perhaps more abundant behind Waihola and Henley than elsewhere.

The whole of the material shows a certain rude stratification that can be easily distinguished along the bank of the Taieri River and in the high slips near the crown of the hills behind Henley Inn.

The lower division comprises the main portion of this glacial drift. It is beautifully exposed in the road-cuttings that follow the north bank of the Taieri River from the bridge at Henley to the upper end of the gorge. It consists of a great succession of red clays, gritty and sandy clays, and beds of well-waterworn gravel that commonly contain a large proportion of small angular fragments of mica-schist. In a few places there are beds of angular masses of rubbly schist in which many blocks are over a foot in diameter.

The whole of the material is well stratified, and diagonal bedding is not uncommon. The dip is towards the north-north-west for a distance of two miles, at angles varying from 10° to 33°. The mean angle is not less than 16°, giving an apparent thickness of over 3,000ft. At Cotter's Gully, and also opposite Crab Island, there is a distinct steeping of the inclination of the beds, which, taken in conjunction with the long lines of escarpment on the south side of the river, may be taken to indicate faulting. If the suggested faulting has taken place to such an extent as to cause the repetition of the beds, I estimate that the thickness must in any case exceed 1,000ft. But I wish to make it clear that it was the extraordinary apparent thickness of the beds that first suggested repetition rather than any physical appearance of faulting, of which there is no satisfactory evidence. I wish further to say that Sir James Hector,* in his diagrammatic representation of this section, shows a continuous dip from the upper end of the gorge to the summit of the coarser upper member of the series facing the Taieri Plain.

[Footnote] * J. Hector, Rep. Geol. Exp., –91, p. lx.

– 592 –

Hector* and McKay found the Taieri glacial beds so well stratified that they placed them in the brown-coal series of Lower Tertiary age. The better opportunity for examination now given along the new road following the north bank of the Taieri, described above, shows that there is no evidence to justify this correlation.

South of Waihola and north of Allanton the glacial material dwindles down to a thin sheet that towards its outer limits in the south is represented by only a few scattered blocks of basalt. Passing eastward towards the sea, the eastern limit of the syncline rests against a highly denuded surface of the mica-schist. To what distance the glacial drift formerly reached in this direction it is impossible to say. All that now remains of it is a thin sheet of gravel and clay that appears here and there on the hill-tops fronting the sea.

[Note.—Hutton, Hector, and McKay have recorded the Taieri Moraine on the hills overlooking the Tokomairiro Plain. Last July, Mr. A. Gordon Macdonald, B.E., and I traced it to the north bank of the Tokomairiro River; and since then Mr. Macdonald has pointed out that the front of Mount Misery up to a height of 1,000ft. above the sea is covered with a thick deposit of glacial morainic drift. I have lately verified this further extension of the Taieri Moraine.]

Briefly stated, we have in the Taieri Moraine a glacial deposit varying from 0 to 1,500ft. or more in thickness. The upper layers, 450ft. or 460ft. thick, are composed of coarse, angular, and gravelly drift rudely stratified; the lower and major portion, over 1,000ft. thick, consists of well-stratified silts, sands gravelly drift. The deposit can be traced over a distance of nearly twenty-two miles in length, and over a width varying from one to three miles.

The Taieri glacial deposit is, according to Dr. Marshall, the terminal moraine of a glacier that flowed across Central Otago, entering the Taieri Basin from the west. The beautifully glaciated slopes of the mica-schist uplands lying immediately west of the Taieri Plain, and the domed summit of Maungatua, bear eloquent witness of the path of this ancient ice-sheet; but I cannot agree that the moraine is terminal. At any rate, it bears no resemblance to such typical ancient terminal moraines as those at Clyde, Lower Kawarau Gorge near Cromwell, Kingston, Von, Waitaki, Lakes Guyon, Tennyson, Rotoroa, Rotoiti, and Brunner. Nor can I find in the literature of glaciation a reference to any terminal moraine consisting of such an assemblage of stratified deposits.

The terminal moraines of ice-sheets consist of tumbled masses of rock dumped over the face of the ice, and seldom show the sorting of fluviatile action. The terminal moraines of valley-glaciers, on the other hand, often show rude sorting of the coarser matter, effected by the river issuing from the bottom of the glacier, the muds and finer material being generally carried away in the flowing stream. But from first to last the tumbled, angular, unsorted material predominates.

In the Taieri Moraine the finer material forms 80 or 90 per cent. of the whole deposit. It is well stratified, and the stratification can be traced over a length of twenty miles. Subglacial and fluvio-glacial agencies alone seem competent to account for the formation of this pile of bedded glacial drift,

[Footnote] * J. Hector, Rep. Geol. Exp., –91, p. lx.

[Footnote] † A. McKay, l.c., p. 45.

Picture icon

Glaciation of the North Island.—Park.

Picture icon

Glaciation of the North Island.—Park.

Picture icon

Glaciation of the North Island.—Park.

Picture icon

Glaciation of the North Island.—Park.

Picture icon

Glaciation of the North Island.—Park.

Picture icon

Glaciation of the North Island.—Park.

Picture icon

Glaciation of the North Island.—Park.

– 593 –

I have described the Taieri glacial deposit as a ground-moraine, as opposed to the view that it is a terminal moraine. Dr. Marshall* challenges this conclusion, asserting that a boulder-clay or ground-moraine is relatively thin because formed between the ice and the bed-rock, whereas the Taieri Moraine is, he says, at least 700 ft. thick. In making this statement he would seem to have fallen into an error. The thickness of boulder-clays, as recorded by eminent geologists, proves conclusively that such deposits are not relatively thin, but, on the contrary, relatively thick.

The till of Scotland, according to Sir Archibald Geikie, varies from 0 to 160 ft. thick; and that of North America, according to Professor Salisbury, from 0 to 500 ft. At Back Valley, in South Australia, the Committee on Glaciation of the Australasian Association for the Advancement of Science reports a Permo-Carboniferous glacial drift or till 1,500 ft. thick. In Germany§ the Pleistocene glacial drift varies from 0 to 670 ft. thick.

In Greenland, Geikie tells us, “Nansen found enormous accumulations of ground-moraine on the edge of the inland ice at Austmannatjern, where there were no nunataks, and not a vestige of surface moraine was-visible.”

The Cambrian glacial beds of South Australia are described by Mr. W. Howchin as being 1,500 ft. thick in the gorge of Appila, including, to use his own words, “860 ft. of characteristic till with boulders.” Thus we find that from the earliest geological times up to the Pleistocene a glacial till is not relatively thin, but may be many hundred feet thick.

Professor Marshall also states that a boulder-clay is more or less continuous in its disposition over the whole area covered by the ice-sheet. This statement also is at variance with known facts. In Scotland, which was completely covered with the northern ice-sheet in the Glacial period, the till is not only notoriously irregular, but it covers a relatively small portion of the country. In writing of the unequal distribution of the glacial drift, or till, in North America,** Professors Chamberlin and Salisbury state that “The thickness of the drift ranges from zero to 500 ft., and the variations are often great within short distances. One hill may be composed of drift, while the next has no more than an interrupted mantle of drift. The drift may be thick on hills and thin in valleys, but more commonly the reverse is the case.” Irregularity is, without doubt, one of the distinctive features of a glacial till.

The Taieri Moraine reaches its northern limit south of Saddle Hill, but small areas of glacial till occur, as first pointed out by Mr. J. T. Thomson,†† in the Kaikorai Valley, more particularly at Fernhill Coal-mine, Abbotsford, and Burnside. The till consists for the most part of confused deposits of angular and semi-angular boulders of igneous rocks and clays, generally unstratified; but in some places the material is irregularly bedded, as may be seen in the steep escarpment overlooking Fernhill Coalmine.

The moraine at Abbotsford is a typical example of a glacial till, It begins at Abbott's Saddle, overlooking the Taieri Plain, and extends from

[Footnote] * P. Marshall, Otago Daily Times, 13th May, 1909.

[Footnote] † Chamberlin and Salisbury, “Geology,” vol. iii, p. 346.

[Footnote] ‡ Proc. A.A.A.S., vol. vii, p. 126; 1898.

[Footnote] § A. Geikie, “Earth Sculpture,” 1902, p. 191.

[Footnote] ∥ A. Geikie, l.c., p. 176.

[Footnote] ¶ Walter Howchin, Trans. A.A.A.S., vol. xi, p. 266.

[Footnote] ** Chamberlin and Salisbury, “Geology,” vol. iii, p. 346.

[Footnote] †† J. T. Thomson, Trans. N.Z. Inst., vol. vi, p. 313.

– 594 –

the railway-line northward for half a mile or more, occupying the lower portion of the Kaikorai Basin. It was known to and figured by Hutton* as far back as 1876. Its south-east slope is knobby and hummocky, the hummocks in many cases enclosing undrained hollows between them, while its upper surface presents smooth, flowing contours. The material consists of peaty clays containing moa-bones, generally much decomposed, stiff yellowish-brown clays containing an irregular bed of bouldery gravel, followed by clays and a confused mass of boulders.

At present, however, I am not so much concerned with the particular manner in which the Taieri and Kaikorai glacial deposits were laid down or formed as with what the deposits stand for. They mean that a gigantic glacier or ice-sheet descended from the distant main divide in western Otago to the present coast-line, occupying the Taieri and Tokomairiro basins from Dunedin to the Clutha, and presenting a continuous ice-face over forty miles long. The débris it has left in its track proves that it rode over the coastal range on its way to the sea. This ice-sheet crept across the uplands of Central Otago, which present no dominating or leading valley leading to the Taieri Basin.

The present canñon of the Taieri River, in Central Otago, and gorges of the Waipori and Clutha, as well as the lower gorges of the Tokomairiro and Taieri, that cut through the coastal range, are admittedly features of comparatively recent date, excavated since the recession of the ice.

The glacial deposits in the Waikouaiti Valley, in the lower valleys of the Waitati and Canterbury rivers, at the south end of Lakes Rotoroa and Rotoiti, and at Boulder Lake, all attest the eastward advance of the land-ice, and the widespread magnitude of the glaciation. Such intense refrigeration as this implies could not but affect the North Island; and of this we have ample evidence provided by the great glacial till in the Hautapu Valley, on the Wellington-Auckland Main Trunk line.

In the vicinity of Dunedin we find that Saddle Hill, Maungatua, and the uplands facing the Taieri Basin, Flagstaff, Swampy Hill, and the hills on each side of Otago Harbour, everywhere present smooth, flowing contours, in many places impressed with fluviatile features of later date. Flowing contours, domed hills, and truncated spurs are the work of no known agency but ice. It is admitted that an ancient glacier or ice-sheet of enormous size occupied the Taieri Basin. How, then, can we, if we view the position without prejudice or bias, deny that ice was the active agency in forming these flowing features when we know that a gigantic glacier existed so near at hand? Whatever the cause of the Ice Age, we cannot escape the conclusion that when the Taieri glacier extended to the sea, the climatic conditions on the east coast of Otago must have been antarctic; and all polar discovery has shown that under such intense refrigeration every valley would be filled with glacier-ice, and the whole region a waste of ice and snow.

In my Bulletin on the Wakatipu region I referred to the Kaikorai boulder-clays as glacial, and not, as assumed by Dr. Marshall, to the clays that lie along the summit of the ridge on which Maori Hill, Roslyn, and Mornington stand. These clays are obviously residual, and as such have always been described by me.

[Footnote] * F. W. Hutton, “Geology of Otago,” 1876, pl. vii.

[Footnote] †J. Park, Bulletin No. 7, N.Z. Geol. Surv., p. 43; 1909.

– 595 –

The Rev. James Christie* has stated that Saddle Hill owes its distinctive form to ice erosion; and after a careful examination of its topography I am entirely at one with him. The seaward slopes of Saddle Hill exhibit beautiful, smooth, undulating contours, and I am satisfied that the two main knobs of the hill, as well as the small knob lying to the east of these, are roches moutonnées in a good state of preservation.

Waihola Lake, and the Taieri Basin, of which it is a part, are good examples of U-shaped valleys, as also is the Kaikorai Valley, which heads into a fine glacial cirque at the source of the Leith Stream. The Tokomairiro is also a glacial basin. It is a continuation of the Taieri Basin, and is separated from the Clutha glacial valley by some low ice-shorn hummocky hills.

At Wingatui, following the foot of the mica-schist hills, there is a deposit of brick-clay some 50 ft. thick, consisting of thin horizontal alternating layers of rock-flour and gritty clay. The material is not residual, but aqueous, and is almost identical with the fine bedded glacial clays at Lake Hayes and Kawarau Falls, near Lake Wakatipu. The existence of the glacial moraines at Allanton, a few miles to the south, and at Ferntown, a few miles to the north, and the existence of an ancient glacier in the Taieri Basin, are strongly suggestive of the glacial origin of these clays.

The slopes and crests of Flagstaff, Swampy Hill, Pine Hill, and of Otago Peninsula, when viewed from almost any standpoint, exhibit smooth, sweeping contours, domed crests, and truncated spurs, in many places dissected by the later streams draining the slopes. Moreover, the higher slopes of Swampy Hill, facing the Upper Leith; the slopes of the Otago Peninsula behind Anderson's Bay; and the higher slopes of Sidey's Hill, opposite Abbotsford, are excavated into tiers of more or less parallel shelves—a feature that I have shown is often associated with ice erosion in the Wakatipu region.

No one has ever claimed that the Otago Harbour is a submerged valley of fluviatile erosion. Such a hypothesis would at once be faced with insuperable difficulties, not the least being a satisfactory explanation of the sudden cessation, or ending, of the valley at Dunedin without leaving a trace of its former existence to the westward of the city. It seems not improbable that the site of North Dunedin, embayed as it is on three sides with encircling hills, was a glacial basin from which one stream of ice passed seaward through the gap between Lawyer's Head and St. Clair, while another stream flowed down Otago Harbour.

As a matter of fact Otago Harbour consists of two distinct basins—namely, the Dunedin Basin and the Port Chalmers Basin—which are separated by a chain of islands that extend from side to side with only narrow, shallow channels between them.

The Leith is obviously a young stream that has eroded its present narrow rift-like gorge down to the floor of the Dunedin cirque since the glacial period.

Mr. Christie has expressed the opinion that Otago Harbour was the work of ice erosion. I am so far in agreement with him in believing that the harbour was occupied by a glacier which deepened it and moulded the adjacent hills, imparting to them the flowing contours and beautiful catenary curves that are so conspicuous on each side.

[Footnote] * J. Christie, Otago Daily Times, May, 1909.

– 596 –

The glacier that occupied the North-east Valley formed the fine amphitheatre-like cirque between Mount Cargill and Signal Hill, and, flowing southward, joined the Dunedin glacier, truncating Logan's Point spur into a broad platform as it turned around the lower slopes of Signal Hill

Dr. Marshall supplemented his remarks with respect to the supposed inability of ice-thrust to bend or break the least resistant schist with some observations apparently meant to convey the meaning that the ancient ice in the Wakatipu region was incompetent to cause the bending of the schists. In my Bulletin on that area, speaking of the outcrop-bending the weak mica-schist of the Arrow and Shotover has suffered, I stated (p. 43), “The bending of the rocks is the work of the ice-sheet that covered the mountains in the glacial period. It often obscures the true dip.” Further on I mentioned that on account of this bending it was impossible to get trustworthy observations for dip and strike over considerable areas.

Referring to the ability of ice to crumple and disrupt rocks, Geikie* says, “While the general surface of the land has been abraded by the ice-sheets, more yielding portions of the rocks have been broken off, bent back, or corrugated by the pressure of the advancing ice.” Further on he continues, “The laminæ of shales or slates are observed to be pushed over or crumpled in the direction of ice-movement.” And, again, speaking of the bending of fissile strata, he says, “Similar effects, with even proofs of contortion, may be observed under boulder-clay, or in other situations where the rocks have been bent over and crushed by a mass of ice.”

In many places the freshness of the glaciated slopes and hummocky moraines is as conspicuous on the coast-line as it is at Wakatipu, which is the more remarkable when we observe that contiguous areas of hard rock have been deeply dissected by streams. This differential erosion is a common feature of glaciated regions, and has been a subject of much discussion and speculation in Europe and America. Mr. G. W. Lamplugh, F.R.S., of the British Geological Survey, in a recent paper on the “Surface Features of Glaciated Areas,” when dealing with this matter, makes the following interesting statement: “In the detailed examination of districts overspread by glacial deposits, I have been constantly impressed by the great discrepancy between the effects of erosion on contiguous tracts. On the one hand, drumlins of boulder-clay, esker-ridges and kamemounds of sand and gravel, and moraines of incoherent rubble, though all composed of material yielding readily to erosion, have retained with very slight impairment their original shape and position in spite of the steepness of their slopes, so that one might imagine that the forces of denudation had been almost in abeyance in the district since the close of the glacial times. On the other hand, a neighbouring stream-channel will often show in the clearest manner that during the same period the erosive agencies have been particularly energetic, so that the stream may not only have cut its trench through a thickness of drift, but may also have notched deeply into the underlying solid rock. My attention was particularly directed to these divergent conditions in Ireland, where they are everywhere prevalent. But I have since recognised in varying degree the same anomalies in every glaciated region that I have studied.”

[Footnote] * A. Geikie, “Geology,” vol. ii, p. 1309.

[Footnote] † A. Geikie, l.c., vol. i, p. 669.

[Footnote] ‡ G. W. Lamplugh, Geographical Journal, July, 1909, pp. –56

– 597 –

Lamplugh thinks that the preservation of the glacial forms was due to the protection afforded by permanent snowfields. He argues that during and for some time after the recession of the ice the refrigeration would still be sufficiently intense to cause the formation of fields of permanent or nearly permanent snow that would protect the ground on which it lay from subaerial waste or denudation, fluviatile action being confined to the defined watercourses. The contention seems not unreasonable.

When considering the genesis of the present topography of the Otago Peninsula and neighbourhood we must not forget the influence the rock formations would be likely to exercise in the development of characteristic surface-forms. The rocks consist of a pile of basic and semi-basic lavas—emissive and hypabyssal—alternating with beds of tuff and breccia.

These rocks rise to a height of 2,000 ft. or 3,000 ft., and from their character and origin one would naturally expect them to form broken ridges, diversified with frowning precipice and steep escarpment, as do the piles of andesitic rocks in the Hauraki Peninsula. Instead of these, we have everywhere the smooth outline and flowing contour commonly presented by intensely glaciated surfaces, diversified by stream - dissected features.

From Dunedin to Timaru the dominant feature of the topography is the smooth, flowing contours, dome-shaped hills, and truncated crests; while everywhere the valleys are U-shaped. The amount of recent-stream erosion is surprisingly insignificant, and where it had taken place is easily distinguishable from the glaciated surfaces.

Glacial till is well developed between Puketeraki and Merton, and there, as near Seacliff, it forms conspicuous hummocks and hollows. The moraine at the former is doubtless a continuation of that spoken of by Mr. Christie* as occurring at Cherry Farm.

The land-forms in the lower end of Waikouaiti Valley, and of the encircling hills and ranges, everywhere bear unmistakable evidence of glacial erosion. On the north side of the valley stand the basalt-covered hills, Hawkswood, Bald Hill, and Mount Watkins, all of them presenting beautiful ice-worn domes. The high spurs descending from Mount Watkins are finely terraced in front and sharply truncated on the erest.

From Moeraki to Oamaru, and thence northward to Timaru, the surface of the land everywhere bears the impress of ice erosion, and no one can view the landscape, with its beautiful smooth contours, domed and truncated crests, without being struck with the small part fluviatile erosion has played in modifying the ice-worn surfaces of the Glacial period.

The floor of the Shag Valley from Bushey Park to the river, as first shown by Captain Hutton, is occupied by glacial clays.

The Waitaki is a U-shaped glacial valley. The great glacier which occupied this valley descended from the main divide to the present coastline. It filled the Mackenzie Plain, occupying the present site of lakes Tekapo and Pukaki. It received great accessions from the Hopkins, Hobson, and Ahuriri. It scooped out the basin of Lake Ohau, and ploughed out the tiers of glacial shelves seen on the slopes of the adjacent ranges

The Waitaki glacier received its chief a accessions from the Tekapo, Pukaki, and other basins that were filled with gigantic glaciers the surface of which, judging from the ice erosion and truncation of the mountains around the Mackenzie Plains, of the ranges lying east and west of Lake

[Footnote] * J. Christie, Otago Daily Times, May, 1909.

– 598 –

Pukaki and bounding the Jollie River, stood at a height of 6,000 ft. above the sea.

The Omarama Basin was the gathering-ground and centre of movement of the ice-sheet in this region. From Omarama a stream passed southward through the Lindis Pass, joining the Clutha glacier a short distance above Bendigo; but the main Waitaki glacier turned eastward, flowing down the present valley to the sea.

The ice surmounted and flowed around the south side of Kurow Hill, at the mouth of the Hakataramea River, its course being clearly marked by ice-shorn surfaces and morainic drift for some six miles below the junction of that river, as first noted by Haast. The present gorge of the Waitaki River from Kurow to Awakino is of post-glacial date, the old glacial valley lying behind Kurow Hill.

On emerging from the mountains the Waitaki glacier, which was 118 miles long, deployed to the north and south, to the southward passing over and deeply eroding the triangular area of older marine Tertiaries lying between Oamaru and the Kakanui Range; to the northward flowing over the foothills lying between the Hakataramea and Waimate.

The hills along the seaward front of the Waitaki glacier are covered with a thick deposit of yellowish-brown clay or loess. In a few places near Oamaru, and on the old Holmes Estate, the loess contains intercalated beds of gravel; and along the present sea-face, from Oamaru past Cape Wanbrow to the mouth of the Awamoa Stream, a bed of beach-gravel containing many recent marine shells occurs at its base.

The Tertiaries on the north side of the Waitaki are truncated into plateau-like hills and ridges, while the spurs on the side of the range behind Waimate are in places excavated into glacial shelves, and everywhere exhibit beautiful, smooth, flowing contours. Behind Waimate the glacial shelves extend along the flank of the range for thousands of yards, about half-way up to the summit.

The Rangitata, Ashburton, Rakaia, and Waimakariri valleys are, as reportd by Haast and Speight, U - shaped valleys, the present deep channels in which the rivers in places run having been excavated along the floor of the glacial valleys since the close of the glacial period. Where these rivers emerge from the mountains the spurs and slopes are, as noted by Haast, Hutton, and Speight, ice-shorn and truncated, and in places terraced to a height of 3,500 ft. or 4,000 ft. above sea-level. The ice must have stood 1,000 ft. or more above that level to give it effective eroding-power at the 3,500 ft. contour. But even if we assume that the surface of the ice stood no higher than 3,500 ft., it is obvious, on gravational grounds, that the Waitaki, Rangitata, Rakaia, Ashburton, and Waimakariri glaciers, when they deployed from their mountain gorges, must have united into one great ice-sheet that presented a continuous face towards the sea along the present site of the Canterbury Plains.

Haast estimated, from the evidence of glacial terraces and roches moutonnées, that the Waitaki glacier was about 5,000 ft. thick in its middle portion.

The evidences of ancient glaciation on the east side of the main divide are conspicuous at Lake Coleridge, in the Rakaia Valley; at Lake Summer, in the Hurunui Valley; at Lake Guyon, in the Upper Waiau; at Lake Tennyson, in the Upper Clarence; at Lake Rotoroa, in Nelson; and at Boulder Lake, in Collingwood. At these places, it should be noted, there are no existing glaciers. The mammillated Gouland Downs; the ice-

– 599 –

shorn Whakamarama Range; the domed crests of Lead Hill and Mount Olympus; the beautiful roches moutonnées at Boulder Lake—itself a perfect rock-basin; and the glacial drift scattered over the Aorere Valley, afford conclusive proof that the Pleistocene glaciers of the South Island reached as far north as Cook Strait in latitude 41° S.

But Pleistocene glaciation was not confined to the South Island. There is abundant evidence that a large portion of the Province of Wellington suffered intense glaciation in that period. In the month of July of this year I examined a portion of the Rangitikei Valley and country lying around Mount Ruapehu. When I examined this region in –87 it was covered, except in the Inland Patea and Murimotu plains, with a dense, almost impenetrable, forest, with few tracks or means of access. The construction of the Main Trunk Line Railway, the spread of settlement, and the clearing of the forest now render the examination of the topography and geology comparatively easy.

From the slopes of Ruapehu and Kaimanawa mountains the country is occupied by soft marine blue clays of Pliocene age, containing a thin band or two of rubbly shelly limestone. The surface of these rocks has everywhere the distinctive features of a land-surface subjected to ancient glaciation. The smooth, flowing outlines, the hummocky and mammillated valley-bottoms, the truncated crests, the terraced slopes, are always present.

The Rangitikei is a U - shaped glacial valley, as also is that of its tributary the Hautapu. The south and west slopes of Ruapehu are drained by the Wanganui and Wangaehu rivers; but it is obvious that a stream of ice no less than 4,000 ft. or 5,000 ft. thick flowed from the Karioi Basin across the divide into the Hautapu Valley, joining the great glacier that filled the Rangitikei Valley, a few miles south of Taihape.

The course of the Hautapu glacier is marked not only by the characteristic erosion of moving ice, but by the sheet of volcanic debris spread along its course for a distance of nearly thirty miles. The debris is mainly composed of erratic blocks of various andesites transported from the slopes of Ruapehu. It varies from 0 to 60 ft. thick, as seen in the railway cuttings; and surmounts the hills and ridges bounding the old glacial valley, forming a more or less continuous sheet, everywhere resting on a deeply eroded surface of the underlying marine clays or papa. Near Ruapehu—that is, in the Karioi Basin—the till is mainly composed of clays, blocks of andesite being absent or seldom numerous; but after crossing the divide into the Hautapu the blocks or boulders become more and more abundant, until in the lower twenty miles of the valley they compose the bulk of the till. Below Taihape they become less and less numerous, and are seldom seen as far down as the junction of the Hautapu and Rangitikei. The blocks are always angular, and vary from small pieces up to masses 6 ft. and 8 ft. in diameter. The largest blocks are more numerous at the southern limit of the deposit than elsewhere.

In many places the Hautapu has cut its way through the glacial drift into the underlying papa, and in the progress of this erosion has formed beds of resorted andesitic gravels that are well exposed on the banks of the main stream near Turanga-a-rere and other places. These gravels also contain well-worn greywacke pebbles brought down by the branches of the Hautapu that reach northward to the Kaimanawas. The difference between the tumbled rubbly glacial sheets of angular andesite blocks lying over hill and dale, and the recent gravels derived from it by erosion, is so

– 600 –

conspicuous that it seems hardly possible for the one to be mistaken for the other.

The area at present known to be covered with this glacial till is some two or three hundred square miles, but it is not improbable that the clearing of the forest will show that it stretches far to the westward of the Hautapu.

The Rangitikei glacier descending from the Kaimanawa Range carried only greywacke blocks, and on account of its greater mass it forced the Hautapu glacier to the westward. The effect of this deflection is well seen in the fluvio-glacial drift that forms the great coastal plain extending from the Manawatu to the Wanganui Valley. In the Manawatu area the drift is entirely composed of greywacke material; in the Rangitikei area, of greywacke and andesite, the former predominating; while in the Wanganui area the boulders are mainly andesitic.

In the North Island the evidence of ancient glaciation has been described by me as far north as 39° 15′ S. latitude, and I am strongly of the belief that further investigation will extend the northerly limit considerably beyond this.

When discussing my paper on the glaciation of Ruapehu region and Rangitikei Valley,* Dr. Marshall dissented from my views both as to the evidence of glacial erosion and the glacial origin of the sheet of andesitic detrital material covering the Hautapu Valley. It is, of course, open for any one to dissent from my interpretation of the genesis of the surface forms—the acceptance of evidence as to glacial erosion is proportional to the observer's ability to interpret aright the signs of extinct glaciers; but in respect to the sheet of andesitic material the case is altogether different. When I discovered this deposit in 1886, so numerous, large, and angular were the piles of andesitic blocks of which it was composed, as then exposed in slips and where cut through by streams, in the dense forest then covering the Hautapu Valley, that I described and mapped it as “volcanic agglomerate and tuff,” derived either from some local volcanic centre or from Ruapehu. I am confident that any one who now views the piles of angular volcanic material exposed in the railway cuttings along the Main Trunk line between Mataroa and Taihape will pardon the error I unwittingly fell into in the year 1886.

Dr. Marshall also objected to the glacial origin of the till because I had made no mention of the occurence in it of perched blocks. This was merely an omission on my part. As a matter of fact, many large erratic blocks of andesite occur perched on the slopes and summit of the papa hills near Kaikoura Stream, a short distance south-west of Utiku—that is, at the extreme south limit of the deposit, over forty miles from Ruapehu.

He further objected that no striated boulders had been found in the drift. So far no striated stones have been seen, and I think it is improbable that such will ever be found, as the underlying papa or blue clay is so soft that it could not, in my opinion, offer sufficient resistance to cause the striation of the harder andesite blocks-dragged along by the ice.

But striated blocks are not always present in glacial deposits, their occurrence depending on the character of the bed-rock and the resistant

[Footnote] * J. Park, “On the Glaciation of Ruapehu and Rangitikei Valley, Wellington,” read before Otago Institute, 10th August, 1909.

[Footnote] † J. Park, “On the Geology of the Western Portion of Wellington Province and Part of Taranaki,” Rep. Geol. Exp., –87, p. 69.

– 601 –

character of the transported blocks. The Taieri Moraine, for example contains no striated boulders, and their absence has not been advanced by any one as an argument that the deposit is non-glacial.

Dr. Marshall* agrees with me that the material in the Hautapu glacial drift has been derived from Ruapehu, but says it is a river-gravel. I am confident that no one who has examined this deposit could possibly mistake it for a river-gravel; and I can only assume that Dr. Marshall has never seen the angular, rubbly, high-level andesitic glacial till discovered by me, and so well exposed in the railway cuttings, or that he has mistaken the resorted stream-gravels on the banks of the Hautapu and its tributaries for the glacial deposit itself. I have elsewhere stated that these stream-gravels are merely a local rewash of the great andesitic glacial deposit formed at the places where the streams have cut through it.

As Dr. Marshall still maintains that the glacial drift in the Hautapu Valley is a river-gravel, it becomes necessary to state the criteria of a river-gravel as defined by recognised authorities. They are as follows:—


The material is water-worn, and becomes more and more rounded as it travels away from its source.


The material is coarsest near the source, and gets finer and finer with increasing distance from the source.


The material is sometimes sorted into irregular beds that often exhibit current-bedding.

Judged by these criteria, we find that the Hautapu glacial deposit is everything that a river-gravel should not be.

In the first place, the material is not water-worn, but angular; and, in the second place, it is not coarsest near its parent source, but at its extreme south limit. It contains but little water-worn material, but, where typically developed in the Hautapu Valley, consists of a rock-rubble of angular andesite blocks, set in a matrix of gritty sands and rubbly clays.

The first to state that Otago was covered with a continuous ice-sheet was Mr. J. T. Thomson, for many years Chief Surveyor of Otago, and afterwards Surveyor-General of New Zealand. Mr. Thomson was a keen student of the development of the surface-forms and topographical features, and brought to New Zealand a knowledge of the evidences of ancient glaciation as seen in Scotland, and of the work of valley glaciation in the Himalayas in India. By profession, by natural bent and experience, Mr. Thomson was perhaps better qualified to express an opinion on the glaciation of Otago than any other physiographer since his time.

In a letter to the Otago Daily Times of the 13th May, 1909, criticizing my presidential address to the Otago Institute, Dr. Marshall states that Mr. Thomson “went hot-headed for the covering of the whole southern portion of the Island with an ice-sheet.” I hold no brief for Mr. Thomson, but, as one who knew him and the thoroughness of his work, I have no hesitation in saying that Dr. Marshall's statement does less than justice to the temperament and mental attitude of Mr. Thomson, who was an accomplished geographer and distinguished mathematician. Mr. Thomson was no hot-head, nor were his views as to the ancient glaciation of Otago hot-headed or precipitate, for he tells us that his conclusions as to the glaciation of the South Island were arrived at as the result of observations

[Footnote] * P. Marshall, Evening Star, Dunedin, 12th August, 1909.

[Footnote] † J. T. Thomson, Trans. N.Z. Inst., vol. vi, p. 312; 1876.

– 602 –

extending over seventeen years. In a paper read before the Otago Institute* in 1873 he made special mention of the furrowing of the hills and ranges so remarkably general in Otago.

Mr. Thomson's view as to the magnitude of the ancient glaciation of Otago did not receive the support of Hutton or Hector; but the only active opposition, except as to the striated boulders at Kaikorai Valley, came from Mr. W.T.L. Travers, the well-known advocate, of Wellington. Mr. Travers disagreed with Thomson's views, but did not visit Otago or even challenge the facts adduced by the latter as to the evidence of ancient glaciation. He did not attempt to explain the origin of the smooth, flowing contours, the furrowed and truncated hills and ridges, or the morainic deposits. He merely contented himself with the general statement that in the South Island there was during the glacial period a great extension of the valley glaciers, but no ice-sheet.

Mr. Travers, in discussing the cause of the greater extension of the glaciers, thought it was due to an elevation of the land amounting, in his opinion, to 5,000 ft. or 6,000 ft. In dealing with the glaciation of Otago, I only asked for an uplift of 3,000 ft. The 5,000 ft. or 6,000 ft. uplift conceded by Travers will do all and more than all Thomson asked for. In the first place, it would lower the mean temperature of the South Island below freezing-point throughout the whole year; and, in the second place, it would unite New Zealnd with the Antarctic continent by a nearly continuous land connection. With such a low temperature and increase of land-area it is obvious that not only the South Island, but in all probability the greater portion of the North Island, would be in the grip of King Frost and dominated by ice.

Dr. Marshall objects to the silts or clays of Oamaru and Timaru being given a glacial origin. He states that we have the authority of Haast, Heim, Hardcastle, and others that they consist of wind-blown dust or loess. But to say the silts are loess does not disprove their glacial origin. The origin of the loess of Northern China, Russia, and North America has, like that of New Zealand loess, been a subject of keen discussion. Some have maintained that it was purely wind-blown, others that it was of subaqueous origin; but all have recognised that wherever it was found it occurred in the neighbourhood of glaciated regions. Sir Archibald Geikie and other geologists believe that it is “a flood-loam of glacial times.”

Writing of the loess deposits of Europe, Geikie says the vast accumulations of loess in southern and south-east Russia doubtless owe their origin chiefly to the flood-waters escaping from the margin of old land-ice.

Chamberlin and Salisbury say that “The best-known portions of the loess in America and Europe are associated with glacial formations.”

The contention that the clays of Oamaru and Timaru are loess is, after all, a proof of their glacial origin, which adds further confirmation of my contention as to the extension of the Pleistocene glaciers towards the present coast-line.

The thickness of the ice-sheet at the lower end of Lake Wakatipu was not less than 7,000 ft.—was probably much greater. But, taking it at 5,000 ft., it would surmount everything between that place and the east

[Footnote] * J. T. Thomson, Trans. N.Z. Inst., vol. vi, p. 312.

[Footnote] † A. Geikie, “Earth Sculpture,” p. 192; 1902.

[Footnote] ‡ Chamberlin and Salisbury, “Geology,” vol. iii, p. 405.

– 603 –

coast; and the smooth contours, domed crests, and U-shaped valleys in Southland and eastern Otago would indicate that this was actually the case.

The maximum thickness of the ancient ice-sheet in North America has been estimated at 7,000 ft.; and the southern limit of the ice in Illinois was not less than 1,500 or 1,600 miles south of the Labradorian* centre of movement, reaching down to 37° 30′ N. latitude.

The distance from the Livingstone centre of movement west of Wakatipu to the present east coast is about a hundred and ten miles, and from Kingston about eighty-five miles.

If the North American ice-sheet, 7,000 ft. thick, was able to travel 1,500 or 1,600 miles from its centre of movement, does it seem impossible for the Wakatipu sheet, 6,000 ft. thick at Kingston, to travel a hundred miles? The answer is that it not only reached the present coast-line, but in all probability extended far beyond it as the result of gravitation stress.

Nansen found that the surface of the Greenland ice-sheet rose with a steep gradient at either margin, and flattened as the summit was reached. In the central parts of the icefield he found that the gradient varied from 26 ft. to 27 ft. per mile. For larger sheets Chamberlin and Salisbury believe that the slope was probably less.

Discussing the thickness of the ice-sheet that covered northern Europe, Geikie says that the ice may have been 6,000 ft. or 7,000 ft. thick in Scandinavia, and 4,000 ft. or 5,000 ft. as far south as Scotland, where ice grooves have been found at a height of 3,000 ft. above the sea.

Now, we know that the southern limit of the ice-sheet in Great Britain was the Thames Valley, a distance of over three hundred miles from the Scottish Highlands—the centre of movement. This distance has an important bearing on the gradient of the ice-sheet, for, if we take the thickness at 5,000 ft., it is obvious that the gradient of the British sheet must have been only 1 in 17.

If we take 500 ft. as the height of the abruptly sloping ice at the terminal face of the Otago ice-sheet, we have a vertical height of 5,500 ft. remaining; which, with a gradient of 50 ft. per mile, would give a flow of 110 miles, more than the distance from Wakatipu to the present coastline. The Labradorian ice-sheet that extended southward to the mouth of the Missouri could not have had a gradient exceeding 5 ft. per mile, except we are to believe that the thickness of the ice-sheet greatly exceeded 8,000 ft., which Salisbury says was not the case. With a gradient of 20 ft. per mile, the thickness would amount to the exaggerated thickness of 32,000 ft. The gradient of 50 ft. per mile I have assumed is probably far in excess of the actual mean gradient of the Otago ice-sheet.

It is noteworthy that Captain Hutton, in his map of the ancient glaciated area§ of New Zealand, shows nearly half of the South Island covered with a continuous ice-sheet.

My discoveries in the Wakatipu region in 1907–8 prove that the glaciation of the South Island attained a magnitude hitherto unknown to Hutton and other New Zealand geologists. My own view of the

[Footnote] * Chamberlin and Salisbury, “Geology,” vol. iii, p. 357.

[Footnote] † Chamberlin and Salisbury, “Geology,” vol. i, p. 358.

[Footnote] ‡ A. Geikie, “Geology,” vol.

[Footnote] § F. W. Hutton, pamphlet, “Report of the Research Committee appointed to collect Evidence as to Glacial Action in Australasia in Tertiary or Post-Tertiary Times,” p. 12.

– 604 –

New Zealand glacial period up to the year 1907 was the same as that of Hutton; but the fresh evidence discovered at Wakatipu and Wanaka, a re-examination of the glaciation of the coastal land between the Clutha and Canterbury, and the glacial deposits of the North Island, have led me to the belief that the glaciation of New Zealand was on a scale too gigantic to be accounted the work of valley-glaciers. Hutton has shown that nearly half of the South Island was glaciated with an ice-sheet. I have ventured, in the light of fresh evidence unknown to Hutton, to extend the limits of the glaciation over the greater portion of the South Island.

I am in agreement with Hutton and Hector that the glaciation was caused by a general elevation of the land, the uplift amounting to 3,000 ft. or more. This elevation would extend the limits of New Zealand many hundreds of miles to the eastward, and six or seven hundred miles to the southward, the southward limit reaching within seven hundred miles or less of the Antarctic continent. If the New Zealand uplift affected the south polar region, then glacial New Zealand would be separated from the Antarctic land by a relatively narrow stretch of sea.

The question that at once arises is, could New Zealand suffer such intense glaciation without lands in the same latitudes in other parts of the Southern Hemisphere being equally glaciated? The evidence on this point is clear and emphatic, leaving to my mind no further room for doubt that the whole of the higher latitudes of the Southern Hemisphere were glaciated contemporaneously with the Pleistocene glaciation of the Northern Hemisphere.

The land-areas that extend southward to the glaciated latitudes of New Zealand are South America and Tasmania. It will be interesting to see what is said by recognised authorities as to the glaciation of these lands.

The glaciation of Tasmania in the Pleistocene has now been placed beyond all doubt. Evidence of glacial action in that State was first discovered by Mr. Charles Gould,* in the Cuvier Valley, and reported by him in 1860; but, though challenged and denied for some years, the high-level glaciation was at last established by Dunn and Moore. Mount Tyndall, on the authority of Moore, is polished and striated at an altitude of 3,850 ft. Mr. A. Montgomery, sometime Government Geologist of Tasmania, has described important glacial phenomena in the vicinity of Mount Pelion and Lake Eyre. Morainic deposits and perfect roches moutonnées, together with erratics, are described by him as traceable from 2,000 ft. to 2,792 ft. above the sea.

The West Coast Range, Eldon Range, and Mount Ida were covered with glaciers which flowed westward into the valley of the King River, the Macintosh River, and the Henty. The lowest moraines formed by these glaciers, on the authority of Professor Gregory, occur at a height of about 400 ft. above the sea. This is confirmed by Professor David, who states that in the Pleistocene time the glacial ice came to within a few hundred feet of sea-level, if not down to sea-level itself. Important contributions to the glaciation of Tasmania have also been made by Johnston, Twelve-trees, Kitson, Waller, and others. An excellent summary of glacial action in that State and in Australia is given by Mr. Johnston in a paper on the “Glacial Epoch of Australasia.”

[Footnote] * C. Gould, “A Report of the Exploration of the Western Country,” Parl Paper, Tasmania, No. 6, 1860.

[Footnote] † R. M. Johnston, Papers and Proceedings, Royal Society of Tasmania, 1893.

– 605 –

Evidences of ancient glaciation have been also discovered in Victoria, New South Wales, and South Australia.

In 1872, Professor Agassiz, the celebrated Swiss naturalist and glacialogist, as the result of his explorations, announced that the southern portion of the South American continent, as far north as 37° S. latitude—that is, for a distance of nearly 1,400 miles north of Cape Horn—was covered with a continuous ice-sheet extending from the Atlantic to the Pacific Ocean.* He stated that the movement of the ice was to the north, and independent of the present slopes of the land.

But, long prior to this, Darwin had called the attention of geologists to the thick masses of boulder-clay and other evidences of glaciation at Tierra del Fuego; while, in 1870, Professor Dana, a distinguished American geologist, in his well-known work on “Geology,” had stated that glacial drift in South America is met with as far north towards the Equator as 41° S. latitude.

Professor Agassiz was the father and exponent of the theory of ancient glaciation of northern Europe and America. It was he who in 1840 startled the geological world with the announcement that northern Europe had been another Greenland, lying under a continuous sheet of land-ice—a conclusion arrived at as the result of a close study of the European Alps.

The clamorous opposition to his views regarding the invasion of the polar ice-sheet, mostly advanced by zoologists, was soon silenced; and in the early “forties,” as a result of a tour through Scotland, accompanied by the veteran Dr. Buckland, he further announced that the northern kingdom everywhere contained clear evidence of ancient glaciation.

Subsequently, Darwin, Buckland, Sir Charles Lyell, Sir A. C. Ramsay, Sir Archibald Geikie, and Professor Geikie verified the conclusions of Agassiz, and made fresh discoveries in England and North Wales, which, in the late “forties,” led to complete agreement among British geologists as to the Pleistocene glaciation of Great Britain and Ireland by a sheet of land-ice which was an extension of the north polar ice.

Agassiz carried his observations to America, and in the early “fifties” reported the discovery of evidence showing that a continuous ice-sheet had extended down the Missouri Valley as far south as 41° N. latitude—a conclusion now accepted by all American geologists. Later discoveries have shown that the southern limit of the Labradorian ice-sheet reached as far as 37° 30′ N. latitude.

The opinion of Agassiz, announced in 1872, that the Southern Hemisphere had passed through a glacial period, or great Ice Age, in all respects parallel with that in the Northern Hemisphere, is quoted approvingly by Sir Archibald Geikie in his monumental text-book, by Professors Chamberlin and Salisbury in their comprehensive manuals, by Professor Prestwich in his well-known text-book; and is challenged by none. On the contrary, Prestwich in his glacial map of the Antarctic region shows the southern portion of South America as glaciated with a continuous ice-sheet up to 37° S. latitude, contemporaneously with the Glacial period of the Northern Hemisphere. On the same map he shows that the greater portion of the South Island of New Zealand was glaciated at the same time.

[Footnote] * Am. Jour. Sc., vol. iv, p. 135; 1872.

[Footnote] † J. D. Dana, “Geology,” 2nd ed., p. 540; 1870.

[Footnote] ‡ Chamberlin and Salisbury, “Geology,” vol. i, p. 330.

– 606 –

Speaking of the evidences of ancient glaciation in South America, Professor Prestwich* says, “Those facts are nevertheless decisive as to the prevalence in late geological times of a glacial epoch which there is reason to suppose was synchronous in the two hemispheres.”

If the Southern Hemisphere suffered a period of glaciation synchronous with that of the Northern Hemisphere, how could New Zealand escape ? South America shows clear evidence of ice invasion as far as 37° S. latitude: why not look for similar evidences in corresponding latitudes in New Zealand ? My discoveries of gigantic and hitherto unrecorded evidences of glaciation in the South Island and in the Province of Wellington seem to show that New Zealand did participate in the general glaciation of the Southern Hemisphere as far north as 39° S. latitude.

Speaking of the criteria of glaciation, Professors Chamberlin and Salisbury say, “It is not strange that the glacial theory was resisted half a century.” It is perhaps not strange to find resistance still alive in this remote Dominion, where the modern thought of the northern world must of necessity permeate slowly, and take time to find a place in our literature.

Having shown that the South Island and a portion of the North Island were glaciated during the Pleistocene, I have ventured to suggest that the glaciation was caused by a general uplift of the land to the extent of 3,000 ft. or more—an uplift giving New Zealand continental dimensions, and bringing it close to the Antarctic land. The natural corollary of this hypothesis was my suggestion that the glaciation of New Zealand was caused, or, at any rate, accompanied, by an advance of the Antarctic ice-sheet in the same manner as the Glacial epoch of northern Europe was caused by an invasion of the Arctic ice-sheet, which was followed by a great extension of the Alpine glaciers.

The distance from New Zealand to Wilkes Land is 1,320 miles, or about the same distance as Melbourne. The Campbell, Auckland, Macquarie, and other subantarctic islands stand on the same marine platform as New Zeland, and, according to Sir Joseph Hooker§ and Captain Hutton, were at one time a continuation of that area. As we have already seen, New Zealand geologists are agreed that the Glacial period was caused by elevation of the land, the majority—myself included—agreeing that an uplift of 3,000 ft. or 4,000 ft. was required to give the necessary refrigeration. It is obvious that an uplift of that amount would link up all the southern and outlying islands with the mainland of New Zealand, thereby reducing the sea gap between glacial New Zealand and the Antarctic land to less than seven hundred miles, besides correspondingly reducing the depth of the sea in the intervening gap.

If the New Zealand elevation was accomplished by uplift in the Antarctic to a corresponding extent, then Ross Sea, which is relatively shallow, would be dry land, thereby still further reducing the stretch of water intervening between the extended New Zealand area and the Antarctic continent.

[Footnote] * Prestwich, “Geology,” vol. ii, p. 465.

[Footnote] † Chamberlin and Salisbury, “Geology,” vol. i, p. 337.

[Footnote] ‡ J. Park, “Glaciation of Wakatipu Region,” Bulletin No. 7, N.Z. Geol. Surv., 1909; and presidential address to Otago Institute, 12th May, 1909, “On the Block Mountains and Great Ice Age of New Zealand,” Otago Daily Times, 12th May, 1909.

[Footnote] § Rose, “Voyages,” vol. i, p. 160; 1847.

[Footnote] ∥ Hutton, Trans. N.Z. Inst., vol. xxxii. p. 182.

[Footnote] ¶ Dr. K. Fricker, “The Antarctic Region,” p. 208; 1900.

– 607 –

Dr. Marshall* states that in the “most highly glaciated tracts the ice does not extend over deep-sea areas.” This statement is entirely contrary to well-attested facts. In the Antarctic region the Great Ice Barrier extends over the sea for many hundreds of miles.

Ross found a depth of over 2,460 ft. on the edge of the barrier, and, commenting on this, he says, “So great a depth of water seemed to remove the supposition that has been suggested of this great mass of ice being formed upon a ledge of rock, and to show that its outer edge, at any rate, could not be resting on the ground.”

Further on he mentions the almost magical power of the sea in breaking up land-ice or extensive floes from 20 ft. to 30 ft. thick; but this extraordinary barrier of ice of probably more than a thousand feet in thickness crushes the undulations of the waves and disregards their violence. It is, he says, “a mighty and wonderful object, far beyond anything we could have thought or conceived.”

Captain Scott, of the “Discovery” Expedition§ found soundings ranging up to 482 fathoms, or 2,892 ft., on the edge of the ice barrier, concerning which he says, “What was the thickness of the ice-sheet to the south, or what lay beneath it, was obviously impossible for us to determine; but on collecting all the indirect evidences which bear on these points, I came to the conclusion—which I still hold—that the greater part of it is still afloat; and, strange as it is to imagine that the sea should run beneath such a solid sheet for so many hundreds of miles, I have yet to learn any reasonable argument against such an idea.”

The ice-sheet stretched far beyond the most southerly point reached by Scott. Its extent is perhaps not less than five or six or even seven hundred miles.

Speaking of the former extent of the ice-sheet, Scott says, “On all sides of us and everywhere were signs of the vastly greater extent of the ancient ice-sheet.”

In the Glacial epoch of the Northern Hemisphere the ice-sheet extended across the North Sea for many hundreds of miles, until it reached Scotland.

Referring to the former extension of the ice-sheet in Ireland, Geikie says, “There can be no doubt from this evidence that even in the south of Ireland the ice-sheet continued to be so massive that it went out to sea as a great wall of ice.”

The Antarctic Ice Barrier still extends over the sea as a solid sheet “for many hundreds of miles,” even though the Glacial period there is now past. Have we any reason to believe that it did not extend seven hundred miles farther north—that is, to the limit of the extended New Zealand area—during the period of maximum refrigeration ?

Or, even assuming the Glacial period to have been due to some other cause than elevation of the land, do we know of any reasonable argument why the ice-barrier in the Glacial period could not extend over the sea for the 1,300 miles that separates New Zealand and the Antarctic land ? On the contrary, it seems only reasonable to suppose that the intense refrigeration which caused the gigantic glaciation of New Zealand would assist the ice barrier to invade these latitudes.

[Footnote] * P. Marshall, Otago Daily Times, 13th May, 1909.

[Footnote] † Ross, “Voyages,” vol. i, p. 222.

[Footnote] ‡ L.c., p. 228.

[Footnote] § Scott, “Voyage of the ‘Discovery,’” vol. ii, p. 309.

[Footnote] ∥ Scott, l.c., p. 314.

[Footnote] ¶ A Geikie, “Geology.” vol. ii, p. 1329.

– 608 –

The extent of a marine ice-sheet would seem to be mainly dependent on the degree of refrigeration. It is difficult to see where the depth of the sea comes in when once the ice is afloat, except it be supposed that the ice can only advance where it rests on the sea-floor. But Ross, Scott, and others have shown the fallacy of this contention, proving by soundings that the Antarctic ice barrier in Ross Sea is afloat. as stated above, for many hundreds of miles.

Dr. Marshall* further objects that no boulders foreign to New Zealand have been found in any glacial deposit. In making this statement Dr. Marshall must surely have failed to recognise the fact that the movement of the ice in the New Zealand area would everywhere be away from the centre of movement, at the main divide, towards the sea. I never suggested that the Antarctic ice advanced to New Zealand and crept over its highlands and mountains. Polar ice cannot disregard the laws of gravitation. Ice obeys all the laws of motion of a viscous fluid, and hence flows away from the gathering-ground to a lower level like a slowly moving river. Obviously, as the New Zealand land-ice alone flowed over Otago, it would be futile to look for Antarctic boulders in the local moraines, as the pressure of the land-ice would certainly keep the Antarctic sea-ice away from our shores.

The local origin of morainic drift, or till, is emphasized by all writers of text-books on geology. Thus Sir Archibald Geikie, in his text-book, says, “The great majority of stones in boulder-clay are of local origin, not always from the immediate adjacent rocks, but from points within a distance of a few miles.” Further on he says, “No Scandinavian blocks have been met with in Scotland, for the Scottish ice was massive enough to move out into the basin of the North Sea until it met the northern ice-sheet streaming down from Scandinavia, which was kept from reaching the more northerly parts of England.”

Again, he says, on the authority of Professor Salisbury, that the general local character of the glacial drift is as marked in Canada and the United States as in Europe.

In other words, blocks can only be carried from their gathering-ground, or centre of movement, within the limits of flow of the ice-stream, and in all cases the pressure of land-ice will be superior to that of ice supported on water.

Boulder-clays or till, and all infraglacial drifts, must from their mode of formation be composed of rocks of local origin. On the other hand, the rocky material borne on the back of the glacier or interbedded in an ice-sheet may be carried a great distance from the parent rock.

British geologists recognise two distinct epochs of glaciation in the late Tertiary, separated by an interglacial period of milder climatic conditions—namely, an older epoch during which Scotland and the greater part of England were covered with an ice-sheet, and a later epoch of less intense refrigeration than the first, as shown by the more local character of the glaciation. The last was distinguished by many minor phases of intensity.

As far back as 1876 Captain Hutton§ also distinguished two epochs of glaciation in the late Tertiary in New Zealand. To the older, or period of

[Footnote] * P. Marshall, Otago Daily Times, 13th May, 1909.

[Footnote] † Sir Archibald Geikie, “Text-book of Geology,” vol. ii, p. 1310.

[Footnote] ‡ L.c., p. 1311.

[Footnote] § F. W. Hutton, “Geology of Otago,” 1876, pp. –67.

– 609 –

maximum refrigeration, he ascribed the Taieri Moraine and Blue Spur deposits, and to the younger the terminal moraines scattered throughout Otago, as well as the Kaikoria and Shag Valley deposits. Hutton's view has been supported by Mr. Hardcastle,* of Timaru.

I am entirely at one with these writers in distinguishing two epochs of late Tertiary glaciation in New Zealand. To the older period I ascribe the Blue Spur, Taieri, and Kaikorai glacial deposits in Otago, and the great andesitic boulder-till in the Rangitikei Valley in Wellington; and to the later epoch the valley moraines, glacial dams, and fluvio-glacial drifts in the South Island. The first was the period of maximum refrigeration, causing widespread glaciation; the second, the epoch of gigantic valley glaciers, which showed many minor advances and retreats before the last and final recession to the Alpine strongholds.

The parallelism between the Glacial epochs of Great Britain and New Zealand is too remarkable to be a mere coincidence, and would tend to show that the refrigeration of the Pleistocene was due to some secular cause, and not to local elevation, as New Zealand geologists have commonly believed.

In Europe a succession of great ice-sheets, radiating from north Norway as a centre, crept over the lowlands of northern Europe, and, crossing the basin of the North Sea, was met by the ice radiating from the mountains of Great Britain. The ice-sheets from the Scandinavian glacial centre also crept northward, joining the ice of the Arctic region.

In North America four centres of glacial radiation have been recognised—namely, the Greenlandian, Labradorian, Keewatin, and Cordilleran—the last three being situated in the continental area somewhere between the parallels 52° and 55° N. latitude. The Labradorian centre lay about 1,800 miles east of the Keewatin centre, and the Cordilleran about 1,000 miles west of it. The ice-streams from these centres, though so far apart, united as they spread outward, covering altogether an area of some 4,000,000 square miles. The Cordilleran ice-sheet crept southward to 47° N. latitude; and the Labradorian to 37° 30′ N. latitude, or 1,600 miles from the place of dispersion. From these glacial centres the ice-sheets also spread northward, joining the advancing ice of the Arctic region.

The evidence seems to leave little reasonable room for doubt that the Alpine region of the South Island was a centre of dispersion from which ice-sheets in like manner radiated outward towards the sea on all sides. The European and North American ice-sheets deployed northward, joining the Arctic ice. There seem to be no valid grounds for supposing that the New Zealand ice-sheets did not spread southward and join the advancing Antarctic ice.

Criteria of Glaciation.

Having thus reviewed the evidences of ancient glaciation presented in New Zealand, more particularly as seen in Otago and on the east coast of the South Island, and in the Provincial District of Wellington, we are better able to determine the value of the criteria of an ice-sheet mentioned by Dr. Marshall. He places these criteria under five headings, which I propose to consider seriatim:—


A continuous line of morainic deposits at the extreme limit of the ice-sheet.

[Footnote] * Hardcastle, Trans. N.Z. Inst, vol. xxiii, p. 311.

[Footnote] † Chamberlin and Salisbury, “Geology,” vol. iii, p. 331.

– 610 –

This is contradicted in England, where, as Geikie tells us, no terminal moraine marks the southern limit of the ice-sheet that crept over Britain. But in New Zealand we have the Taieri coastal moraine, for size without a parallel in any other part of New Zealand, or even in Great Britain; therefore the balance of evidence as to an ancient ice-sheet lies with New Zealand as against England, which, although possessing no terminal moraine, is admitted to have been covered with an ice-sheet.


The occurrence of till or boulder-clay over almost the whole area covered by the ice-sheet.

When tested against British experience we find that this criterion, like the first, also fails. In Scotland, although ice-covered in the Pleistocene from end to end, the till is mainly confined to the maritime benches, coastal plains, and lowland valleys. Altogether it does not cover more than 25 per cent. of the total area. England and Wales were covered with a continuous ice-sheet as far south as the valley of the Thames, and yet we find that the glacial drift is most irregularly distributed, altogether covering less than 12 per cent. of the glaciated area. In New Zealand we have the great Taieri Moraine, the Kaikorai boulder-clay, and the loess covering the coastal hills at Oamaru and Timaru, the Lyttelton Hills, and Motanau Downs; loess being, as we have seen, a deposit of purely glacial origin.

The Taieri Moraine is admitted to have been formed by a glacier that descended from the highlands between the Clutha and Taieri rivers. The glacier that filled the Taieri Basin and formed the great coastal moraine must have presented a front from thirty to forty miles long at the least. And, since this gigantic glacier surmounted hills over 1,000 ft. high near Milton, it must also have flowed southward into the lower Clutha Valley—a conclusion strengthened by the smooth, flowing contours and chain of shallow lagoons on the Clutha side. To those not familiar with the geography of this place, it should be mentioned that the rise between Milton and the Clutha Valley is only some 150 ft., the area between the Toko-mairiro Plain and the lower Clutha being occupied by low, round, ice-shorn ridges varying from 200 ft. to 300 ft. above sea-level.

My object is to show that this Taieri ice-sheet must have been of gigantic proportions. Now, if the climatic conditions in the arid region of Central Otago allowed the accumulation of this great ice-sheet, what, I ask, must have been the climatic conditions on the higher, less arid mountainous region of western Otago ? Would not that area, as already shown by Hutton, be also covered with a continuous ice-sheet ? Dr. Marshall seems to have anticipated this difficulty, for he tells us—and this is only a hypothesis suggested by Geikie and Chamberlin—that the supposed non-glaciation of eastern Europe, Siberia, and Alaska was due to the aridness of the climate—that is, to the absence of precipitation. And from this he argues that one valley may contain a glacier and the next to it no glacier. Unfortunately for the application of this argument to New Zealand, the site he has selected for the gathering-ground of the Taieri ice-sheet is situated in the most arid region of Central Otago, where, as in Siberia, the snowfall should have been too small to maintain such a mass of ice as that which radiated from the highlands between the Clutha and Taieri watersheds, filled the Taieri and Tokomairiro basins, and crept over the coastal hills to the present coast-line.

I contend that the great ice-sheet that descended from the highlands of Central Otago was a part of the greater ice-sheet that occupied the Wakatipu and Wanaka areas.

– 611 –

If Dr. Marshall's second criterion were made to read that “a portion of the glaciated area is covered with till or boulder-clay,” it would be nearer the truth, and be applicable equally to Great Britain and New Zealand.


The occurrence of rounded hills over the glaciated area.

This is admitted by geologists to be a common feature of glaciated areas. Rounded hills are conspicuous along the east coast of Otago from the Kakanui River south to Waikouaiti, and from near Dunedin south to the Clutha. Of examples north of Dunedin we have Hawkswood Hill, Bald Hill, and Mount Watkins on the north side of Waikouaiti Valley, There are many equally fine domed hills in the Palmerston district, but I have selected these, as they are well seen from any point on the railway near Puketeraki. South of Dunedin we have Saddle Hill, Scrogg's Hill, and Stony Hill; Waihola Hill and Tuna Hill on the west side of Lake Waihola—the first opposite Waihola Railway-station, the second at the south-west end of the lake; and farther south the beautiful rounded domes and whalebacked mounds near Crichton Railway-station, six miles south of Milton.

Maungatua is a good example of a domed hill when looked at from the Taieri Basin; but, as viewed in profile from any point south of Milton, it is found to present the characteristic features of the well-known “slope and crag” which is so distinctive of highly glaciated areas. Viewed from the Taieri Plain it appears to be a high mountain with steep slopes, but when seen in profile it appears as a fine dome sloping gently westward to the Barewood Plateau, on the edge of which it stands. On the east it descends suddenly into the Taieri Plain. The gentle slope is the stossseite, the steep slope the lee-seite.


The occurrence of grooved or ice-worn surfaces of rock throughout the district covered by ice.

Ice-worn surfaces, as I have already shown, abound on every hand, in some areas as smooth and fresh as if the ice had only yesterday retreated, but in most areas dissected and eroded by the younger present day streams.

Of the undisturbed glaciated surface we find beautiful examples around Hampden, and immediately south of the Tokomairiro Plain, where between Crichton and Lovell's Flat the smooth, flowing outlines are still finely preserved on the long, gentle slope rising towards Central Otago.

Of the glaciated ice-worn surfaces that are more or less impressed with later stream erosion I will here mention only the splendidly ice-worn slopes on the west side of the Taieri and Tokomairiro plains, extending as far south as Lovell's Flat and Balclutha, a distance of altogether some forty miles. These smooth, gently rising slopes were eroded by the great ice-sheet that descended from the Barewood Plateau and the Clutha glacial valley, and can be seen to great advantage from the Main South Railway, more particularly on the west side of Lake Waihola and on the south-west side of the Tokomairiro Plain after passing Milton.

Recent strems have cut many narrow gutters and shallow gullies in the glaciated slopes; but the two generations of erosion can be easily traced by even the merest novice in physiography from the circumstance that the flat glaciated surfaces are generally covered with yellowish-coloured tussock, while the gullies worn out by later stream-action are commonly sharply outlined by the darker-coloured bracken that grows in them.

– 612 –

The evidences of glacial erosion are perhaps seen to better advantage on the west side of the Tokomairiro Plain than elsewhere in Otago. The panorama of ice-worn slope, rounded dome, mammillated ridge and hummock, as seen from the summit or slopes of Mount Misery, near Milton, is unsurpassed by anything in the Wakatipu district, or by anything I know of in Scotland or England.

Briefly stated, Dr. Marshall's fourth criterion postulates that the area covered by ice is grooved and ice-worn; and no one, I think, will be found to dispute it. Further on he admits that the Taieri Moraine was formed by an ice-sheet or glacier that descended from the highlands between the watershed of the Taieri and Clutha rivers.

The pertinent question is, does he deny the competency of this particular ice-sheet to erode the surface over which it flowed ? Did it present an exception to the general law ? I have, I think, sufficiently proved that the surface over which it flowed is, as we should naturally expect, deeply glaciated, and the evidence is there for every one to see.


Boulders of rock foreign to the locality would be found in great numbers in positions where they could not have been deposited by streams that formerly existed, or on prehistoric sea-benches.

I have already quoted references from Sir Archibald Geikie and Professor Salisbury to show that the rocks in boulder-clays are of local origin. A large proportion of the boulders in the Taieri Moraine consists of rocks carried from the west side of the basin. The absence of foreign rocks in this moraine has not been held by Hector, McKay, or Hutton to be proof of its non-glacial origin. No rocks foreign to New Zealand have yet been found in the coastal moraines, and I think it improbable that such will ever be found. Geikie has shown that the pressure of the land-ice descending from the Scottish Highlands thrust the Scandinavian ice-sheet, flowing over the North Sea, away from the shores of Scotland; and it is certain that the Otago land-ice would in the same manner thrust the invading polar ice away from our shores.

Dr. Marshall states that diligent search for eight years has failed to show him on the coastal hills of Otago any of the evidences of glaciation that I have shown to be so abundant in the Wakatipu region. Perhaps, after all, this is not surprising when we remember that such veteran geologists as Sir Charles Lyell, Sir Roderick Murchison, Sir Andrew C. Ramsay, and Dr. Buckland failed to observe the widespread and conspicuous evidences of glacial action in Scotland until these were first brought under their observation by Agassiz.

The glaciation of northern Europe was at first denied by zoologists. It was claimed that an advance of the polar ice would have caused an overlapping of Arctic, Sub-arctic, and Temperate faunas. That some such overlapping did take place appears to have been established. Be that as it may, the difficulties of the zoologists have long since been removed.

The glaciation of the South Island and of a large portion of the North Island will doubtless present difficulties to the biologist. What these difficulties are I cannot suggest. All that I can do is to record the evidences of an ancient glaciation that occurred on a scale so great as to suggest that the South Island was a centre of movement from which the ice-sheet descended to the sea, in the southward direction probably meeting the advancing polar ice.