Go to National Library of New Zealand Te Puna Mātauranga o Aotearoa
Volume 6, 1873
This text is also available in PDF
(1 MB) Opens in new window
– 297 –

Art. L.—On the Extinct Glaciers of the Middle Island of New Zealand.

[Read before the Wellington Philosophical Society, 13th October, 1873.]

There are few points of geological interest more strikingly brought under the notice of the traveller in the great mountain range of the Middle Island of New Zealand, than the evidences of the former extension of a glacier system, of which the numerous glaciers of the first order still occupying the valleys radiating from Mount Cook are, without doubt, a continuing remnant. Indeed, it is scarcely too much to say, that every great valley stretching into the main range, from one end of the island to the other, affords unmistakable proofs of having, at some time, been occupied by ice; and it is my purpose in the present paper, after making some general observation as to the bearing of this fact upon other geological questions affecting both islands, to describe, in some detail, the particular evidences of glacier action which are to be seen in the valleys of the Buller and the Dillon, two of the largest rivers in the Province of Nelson.

Now it must be evident that the disappearance of the enormous glaciers which, as will be seen in the sequel, formerly filled the upper parts of these two valleys—as well as of those which occupied the valleys of the Hurunui, the Waimakariri, and the Rakaia, in the Canterbury Province—must be attributed either to a singular change in climate, or to a great diminution in altitude above sea level of the mountain chain in question.

Those who are curious upon the first point, as a possible cause, will find

– 298 –

abundant suggestions for discussing it in the last edition of Sir Charles Lyell's “Principles of Geology”; but although the circumstances there indicated, as being calculated to affect the climate of the North Pacific, may have been instrumental, in some measure, in determining the height of the snow line in the latitude of New Zealand, I am nevertheless inclined, for the purposes of the present description, altogether to discard them from consideration, and to look to depression alone in order to account for the disappearance of the ice masses in question.

In this connection it must be observed that, except the summits of Mount Franklin (which is certainly not under 10,000 feet in height) and of a few of the higher peaks by which it is immediately surrounded, no part of the Middle Island range to the northward of the Mount Cook system at present reaches a greater altitude than 8,500 feet above sea level. In the Mount Cook system, however, it rises abruptly, attaining its greatest elevation (13,600 feet) * in Mount Cook itself; whilst the lower mountains in its immediate vicinity vary from 11,000 to 12,000 feet in height. It is, moreover, worthy of note—having regard to the continued existence of glaciers of the first order in this part of the Middle Island range—that its present altitude is very much the same as that of the greater portion of the Pennine Alps, a chain comprising the highest ground and the most colossal mountains in Europe, and which has always been distinguished by the number and extent of its glaciers.

We are unfortunately without special data for determining the actual position of the snow line in New Zealand, but many circumstances concur in inducing me to adopt, for the Middle Island mountains at all events, the same height above sea level as that which has been fixed by observation for the Swiss Alps, namely, about 9,000 feet. But it has also been ascertained that, in those portions of the latter mountains in which glaciers of the first order occur, the average depth of perpetual snow, taken over the whole surface above the snow line, is not less than 300 feet, and we may therefore fairly conclude—looking to the fact that some of the glaciers of the Mount Cook system may compare in extent with some of the largest of those which now occupy the valleys radiating from Mont Blanc—that the snow fall and the average depth of perpetual snow upon and around Mount Cook are much about the same in extent as in the case of the Swiss Alps. I need scarcely say, however, that these assumptions (as in the case of all others where no exact data exist) may contain elements of error, but not, as I think, to such an extent as materially to affect the general conclusions which I propose to deduce from them,

[Footnote] * The altitude of Mount Cook, as trigonometrically determined by Mr. T. R. Hacket, is 12,364 feet. I am not aware if this observation has been verified or disproved. See Geological Survey Report, 1869, p. 12.—[Ed.]

– 299 –

especially when taken in connection with the ascertained facts to which I shall call attention in the sequel. Assuming, then, that under existing climatal conditions an average elevation of not less than 13,000 to 14,000 feet would be necessary, in those parts of the Middle Island range which do not now exceed 9,000 feet in height, for the formation and existence of such glaciers as undoubtedly once occupied the valleys of the Hurunui, the Waimakariri, and the Rakaia—the highest summits in the vicinity of which do not now exceed the latter altitude—we must either accept a change in climate of a very remarkable character, but of which we have no evidence whatever, or attribute the disappearance of such glaciers to a diminution of not less than 4,000 to 5,000 feet in the general height of the range in question, as compared with its altitude when the glaciers referred to attained their greatest extension.

I may add that I am the more inclined to adopt the latter hypothesis, not only because the evidences in support of it are precisely the same as those which have led to similar conclusions respecting the former extension of the Swiss glaciers, but also because it is more in accordance with the principles which govern sound geological enquiry. One circumstance, moreover, is very noticeable in connection with the extinct glaciers to the north of Mount Cook, namely, that the extent of each appears to have borne a distinct relation to the altitude of the mountains in which it arose; for we find, not only with those which still occupy the Mount Cook valleys, but also with those which formerly occupied the valleys radiating from the Spencer Mountains, that the lateral moraines occur at far greater heights, and the terminal moraines extend to far greater distances, and are much more extensive in their dimensions, than those which were deposited by the glaciers which occupied any part of the range intervening between these two great mountain masses.

Assuming then that—at the time when the valleys above referred to were occupied by glaciers of the first order—the Middle Island range, generally, stood at an additional elevation of not less than 4,000 to 5,000 feet above sea level, not only must the present islands of New Zealand proper have been connected, but an immense area of dry land must have existed in all directions around them, probably extending, to the eastward, far beyond the chain of islands which curves round them on that side, from Raoul Island in the north, by the Chatham Group, to the Antipodes Islands in the south, all of which still bear a vegetation nearly identical with that of the parent land. To what extent the depression which led to the disappearance of the glaciers in question may have exceeded the maximum above referred to, I am not prepared to say, and, although both Captain Hutton and Dr. Haast have mentioned facts which lead to the belief that the eastern side of the Middle Island has risen since the last great depression, the extent to which this is indicated in their statements is too trifling to settle the question.

– 300 –

But the existence, in the localities referred to, of the ordinary evidences of glacier action, such as huge lateral and terminal moraines, of roches moutonnées, blocs perchés, etc., is not the only or even the chief circumstance of interest brought under our observation in connection with the former extension of the glaciers. On looking at a map of the Middle Island we cannot fail to observe a chain of lakes extending in an almost direct line from north to south, occurring chiefly, however, on the eastern side of the great range, and comprising Lakes Howick and Arthur, to the north of the Spencer Mountains; Lakes Tennyson and Guyon, on the eastern side of the same group; Lakes Sumner and Taylor, lying between the Provinces of Nelson and Canterbury; Lakes Coleridge, Lyndon, Heron, Acland, Tripp, and others, in the latter Province; and the more extensive Lakes Wanaka, Wakatipu, Hawea, and others, to the south of the Waitaki River. Now, it has never been doubted that all these lakes owe their existence as such, more or less, to the action of glaciers; those which occur to the north of the Waitaki, at all events, all lying in valleys above the lines of huge terminal moraines which have been deposited across them, and which have formed dams in many instances several miles in length and several hundred feet in depth.

It is, moreover, a matter of extreme interest that many of the larger valleys which, during the former extension of the glaciers, were occupied by ice, and are now filled with ordinary alluvial deposits, must for a long period after the disappearance of the ice have been filled with water to the height at which the glacier streams had then cut through the terminal moraines. In this condition they resembled, in every respect, the great majority of the existing valley lakes to the northward of the Waitaki River. An admirable example indicating the former existence of such a lake, in which the water has been replaced by alluvium, is to be seen in the upper part of the valley of the Dillon. In this case the moraine which stretches across the valley has an average width of about a mile, and extends down it for upwards of three miles, the fall from the point at which the river has cut through it on the upper side to that at which it discharges itself on the lower side being fully 180 feet, whilst the average slope of the valley for several miles above the moraine is less than 20 feet to the mile, but increases to at least 35 feet below it, indicating the great depth to which the moraine deposit extends below the present general level of the valley. The moraine itself rises, at its greatest height, about 100 feet above the level of the upper valley, and exhibits, in the angle which it forms with the mountains on the eastern side, and at a height of about 30 feet, a former lake margin, as fresh and clean, and as free from vegetation and all other marks (except the recent tracks of cattle and sheep), as if the lake had been emptied only a week ago. I have designated margins of this kind, which are usually composed of sub-angular shingle, as wave-margins,

– 301 –

indicating, as they do in all the existing lakes, the direction of the most prevalent winds, and usually running across the valleys in which the lakes are situated.

The moraine above referred to is about 24 miles below the main source of the river, and the lake which succeeded the ice could not have been less than 14 miles in length in the principal valley, with a branch at least five miles long in the tributary valley of the Ada. Of course it is impossible to determine the actual depth of the moraine deposit at its upper face, but even assuming it not to exceed 150 feet below the lake margin, we have an area of 19 miles in length and a mile in width, with an average depth of 60 feet, which has been filled with river alluvium (independently of the immense quantity of matter which must at the same time have been carried below the moraine) within the period which the river has occupied in cutting down the comparatively loose material of this dam, to a depth of 35 feet only.

It may, in view of such a fact, appear remarkable that the beds of the lakes on the northern side of the Spencer Mountains—which present on that side precisely similar conditions—should not also have been filled up; but I attribute the rapid accumulation of alluvium in the case of the Dillon Valley to the facts that the mountains bounding it are much steeper, are composed of more easily disintegrated rock, are in their upper parts very bare of vegetation, and therefore exposed to the alternate action of frost and heat, and moreover present in many places, for thousands of feet in height and for miles in length, little else than continuous slopes of broken stone; whilst those on the opposite side of the range are in a great measure densely wooded, and are chiefly composed of hard, crystalline rocks. I ought, however, to state that in assuming the moraine of the Dillon glacier to have a depth of only 150 feet below the lake margin above referred to, and in further assuming that the bed of the valley rises gradually from that depth to 0′, I am doing so without any ascertained facts.

I am not aware whether any measurements have been made in order to ascertain the depth of any of the lakes between the Spencer and Mount Cook ranges, with reference to the fall of the rivers flowing from them below the lines of their moraine dams; but the depth of Lakes Arthur and Howick, on the northern side of the Spencer Mountains, is very great as compared with the apparent depth of the bed of the valley of the Buller; whilst that of some of the larger lakes in the Otago Province, and notably of the Wakatipu, exceeds 1,100 feet, their beds, indeed, extending below the present level of the sea.

From a consideration of these facts, and of others which I have not thought it necessary to mention in so general a sketch, I think we are justified in concluding that these extinct glaciers originated during an upheaval of the

– 302 –

land, which extended to an altitude exceeding that of the present mountain chain by at least 4,000 to 5,000 feet, and that they attained their maximum extension coincidently with the maximum of elevation. They also lend strength to the assumption (founded on other independent grounds) that during the period of maximum elevation the land of which the present New Zealand islands proper are the chief remnant had a quasi-continental extension, chiefly to the eastward, and embraced, at least, the chain of islands above referred to. And they further justify us in attributing the disappearance of the glaciers in question to a depression of the land continued up to a comparatively recent period, a circumstance which, I think, is chiefly indicated by the following facts, namely:—That the larger number of the valley lakes which still exist above the terminal moraines of glaciers of the first order in those portions of the range in question which intervene between Mount Cook and the Spencer Mountains (as for example, Lake Sumner on the line of the Hurunui River), have been only partially filled up with alluvium, although the rivers which feed them are all shingle-bearing torrents; whilst, on the other hand, the lake which succeeded the great glacier of the Dillon has actually been filled with alluvium within the period which its outlet has occupied in cutting down the moraine dam to a depth of 35 feet only, for a distance, having regard to the general fall of the valley, of little more than half a mile.

With regard to the probable time at which the upheaval referred to took place I can offer but few observations. So far as I can understand from the reports of the Geological Survey, we have no evidence of any upper marine pliocene beds in the Middle Island, whilst the upper portions of the miocene series are found at elevations varying from 1,200 to 1,800 feet above sea level. I therefore assume that the elevation in question commenced at the close of the miocene period, and that the more recent pliocene deposits within the Novo-Zealandian Province would only have been found on the outside boundaries of the quasi-continental area which existed when the elevation referred to attained its maximum, and were submerged during the subsequent subsidence of the land. The total elevation most probably approached 5,000 to 6,000 feet, the subsequent depression exceeding that by at least from 1,200 to 1,500 feet. At what time the depression ceased it is difficult to say, but it probably continued in pleistocene times, when a re-elevation of the land again took place.

The period which I have ventured to assign for this great oscillation of level would, no doubt, be considered immense if it could possibly be counted in years, but is certainly not too great for the effects produced. That the connection between the main islands and the Chatham Group must have been severed at a comparatively remote period is indicated by the fact that much

– 303 –

greater differences exist between their respective flora and fauna than exist between those of England and the continent of Europe, the connection between which was only severed in pleistocene times. It will at once be seen by those who have had an opportunity of perusing Dr. Haast's elaborate report on the Canterbury plains (presented to the Provincial Government of Canterbury in September, 1864), that the views contained in the foregoing brief sketch are altogether at variance with those which he there propounded in reference to what he has termed “the pleistocene glaciation of New Zealand.”

Whilst giving reasons for his belief that the southern island of New Zealand has never been higher than it is at present, he nevertheless asserts that it was subjected, in earlier pleistocene times, to a general glaciation analogous to that of Greenland. His words are: “It is not necessary to give a picture of the desolate aspect of the country in those pleistocene times; but when reading the descriptions of Dr. Kane, of Greenland, and of other arctic and antarctic explorers, it brought visibly before my mind that this island during that'era would have presented a very similar appearance.” He, however, adduces no evidence whatsoever in support of this statement, nor does he attempt to account for the suggested glaciation otherwise than by a loose assertion “that the climate had changed by some physical causes, and assumed an antarctic character.” For my own part I have never seen—at least in those portions of the South Island mountains which I have personally visited —the slightest evidence which could support such a statement, or which would have led me to the belief that, even during the greatest elevation of the land of which any indication remains, it presented features of glaciation differing (except in such degree as would naturally follow in this latitude) from those which it now presents where glaciers of the first order still exist.

In this connection the following extracts from the Duke of Argyll's address (in February of this year), as President of the Geological Society of London, have a distinct application to the existing physical features of the Middle Island mountains. His Grace says:—“If I may judge from a paper lately contributed by Professor Ramsay to ‘Macmillan's Magazine,’ upon the valley of the Po, and from the recent discussion on Mr. J. F. Campbell's very interesting paper on the glaciation of Iceland, it seems to be admitted by Professor Ramsay that no larger amount of work can be assigned to the glaciers of the glacial epoch than that of greatly deepening the valleys which existed before. If this be admitted, then the question of the effects of glacial denudation in determining the existing configuration of the surface of the earth becomes a comparatively narrow question. The existence of a glacial epoch, at least over a large part of the Northern Hemisphere, which, in its coming, its duration, and its passing away, has been the latest in the great agencies of change, is perhaps one of the most firmly established doctrines of

– 304 –

geological science; and if it be admitted, on the one hand, that when the period began it found the existing systems of hill and valley in the main determined, it must also be admitted, on the other hand, that it cannot have left them exactly as it found them. The intensity given to denuding agents by frost, or rather by the alternations between frost and thaw, is well known to be enormous; and it is impossible that a glacial period should have come on, should have endured for a long period of time, and should have gradually given way to a more genial climate, without having left upon the pre-existing surface powerful and lasting effects. But the conclusion that the glacial epoch deepened within certain limits pre-existing valleys, degraded to a like extent pre-existing hills, filled up estuaries with moraine matter, or with sand and gravel, or covered a great extent of country with boulder-clay, all this is very different from the conclusion that our existing systems of hill and valley, and even of sea and coast, have been all cut out of the solid by some great ice sheet of enormous thickness, which was quite independent of local glaciers, and which did not derive either the cause or the direction of its motion from the mountains which we now see.” Further on, after referring to the present glacial conditions of Greenland and of the great antarctic continent, his Grace says:—“From observations such as these we may be assured, I think, of the truth of the theoretical conclusion that lofty mountain chains, with all their characteristic variety of surface, must, in all ages and in parts of the globe, have preceded the development of glacial conditions, and that in these chains the unequal elevations and depressions, which are the work of subterranean force, have ever been the guiding and controlling cause of glacial action.”

Moreover, such a glaciation as Dr. Haast suggests in the report above alluded to must necessarily have obliterated all but the scantiest fragment of the fauna and flora of the country, leaving, indeed, at most but a few alpine forms struggling for existence amidst the inhospitable conditions by which they were surrounded; whilst, on the other hand, the study of the existing forms of life, and of those which have certainly become extinct within pleistocene times, has led all who have engaged in it to a conclusion entirely at variance with any such assumption. I propose, however, to deal with this question more fully in a future paper.

I will now proceed to describe, in some detail, the glacier phenomena presented to us in the upper parts of the valleys of the Buller and the Dillon, which I have selected as well marked types of those which are exhibited in other parts of the great mountain range of the Middle Island.

These two rivers, as well as many other of the larger rivers in the northern part of the island, have their sources in the great mountain system named by me the “Spencer Mountains,” which occupy the centre of the tract of country comprising the Provinces of Nelson and Marlborough. The highest point of

– 305 –

this range is Mount Franklin—not over-estimated at 10,000 feet—whilst around it are several minor peaks averaging from 8,000 to 9,000 feet in height. The Buller River, which rises on the north side of the range, has its sources in Mount Travers, and empties itself, after a course of about 15 miles, into Lake Arthur, which, with the valley above it, lies between a spur of Mount Robert on the west, and the base of the St. Arnaud range on the east; the general trend of the valley and lake being due north. After leaving the lake the river flows to the westward until it reaches the sea, its waters being increased on its course by numerous large rivers, several of which have their sources on the western side of the Spencer Mountains. A line of road from Nelson, leading originally only to the Wairau Valley, now branches, in the heart of a great forest familiarly known as the “Big Bush,” to the gold fields of the Inangahua and Lyell, passing through the valley of the Buller and the rugged gorge of the Devil's Grip. Almost immediately after leaving the point of junction the Buller road enters the line of an ancient moraine, along the flank of which it runs for several miles, until it opens a small valley excavated in the moraine itself, in a direction nearly parallel to the northern shore of Lake Arthur. Crossing this valley and ascending the moraine on the southern side of it the lake opens out, forming a noble sheet of water, which is seen stretching for miles into the great mountain range. The waters of the lake are of a rich blue colour, and throw up in bold relief the massive spurs of Mount Robert and of the St. Arnaud range, whose lower slopes, covered to the very edge with dense forest, dip abruptly into them; whilst far to the northward, closing in the view, rise the rugged snow-flecked peaks of Mount Travers. It is impossible to imagine a scene of greater beauty, unless we can look forward to the time when the shores of the lake will be studded with villas, giving to it that appearance of life and animation which are alone wanting to complete its loveliness.

Along the front of the lake, parallel to the small valley above referred to, lies a considerable portion of the huge moraine, rising from 100 to 150 feet above the level of the water, the outlet of which has made its way through it at its south-west corner, cutting it down to the depth of from 100 to 120 feet. On the opposite side of the valley, and about a mile and a half from the water, is a range of hills rising from 300 to 400 feet higher than the upper surface of the moraine, and which are composed chiefly of boulder beds, gravels, and sands, in no degree cemented, very little inclined in stratification, and in many places exhibiting perpendicular sections several hundred feet high, particularly in places where the foot of the hills has been washed by a river. The materials are all water-worn, and exhibit the common appearance of gravels and sands which have been deposited in quiet water basins. Captain Hutton, who accompanied me during a late visit to this locality, treats these beds as of

– 306 –

miocene age, but, as yet, there are no data from which this can be satisfactorily ascertained. They certainly overlie sandstones and shales of miocene age.

When the great moraine in question reached the flank of these hills it was deflected to the east and west, stretching in the former direction for three or four miles, and even crossing the watershed into the Wairau Valley, and in the latter for several miles down the valley of the Buller. This moraine is of stupendous dimensions, and was evidently deposited by a glacier which occupied the site of the lake and of the valley above it, and the surface of which, judging from the height of the lateral moraines, must have stood at least 1,000 feet above the present level of the water. The lake itself is several hundred feet in depth in its deepest part, the slope increasing from each extremity but most rapidly from the lower end. It is difficult to account for the great depth of this lake as compared with the general slope of the valley of the Buller, unless we assume that before the elevation of the land its bed was filled with the same materials as compose the hills in front of it, and that these were gradually ploughed out or otherwise removed by the glacier. There can be no doubt, indeed, that a glacier will easily remove loose materials from a pre-existing depression to a depth considerably below the level of their surface on the lower side of the terminal moraine, or, in other words, will scoop out such materials to a depth greatly exceeding the general slope of the valley, but they cannot be removed unless forced more or less up a slope, and brought within the influence of the stream which issues from the foot of the glacier.

If, therefore, the site of Lake Arthur and of the valley above it, as well as of that part of the Buller which is now occupied by the moraine, was filled before the formation of the glacier with the same gravels and sands as compose the hills on its northern side, or with any other loose materials, I see no difficulty in believing that the portion of those materials which lay in the lake depression below the level of the general slope of the valley has been removed by the glacier, leaving the lake basin to be refilled by the alluvium which has, since the disappearance of the ice, been and is still being carried into it by the main river and by the innumerable streams which furrow the ranges on each side of it.

I have already alluded to certain facts in relation to the action of the glacier which formerly occupied the valley of the Dillon, but there are some circumstances of a special character in connection with it, which render it necessary that I should give more details of the physical features of the district, in order that my subsequent remarks may be understood.

The Dillon has its principal sources in the Pyramid Mountain, a huge peak to the north of Mount Franklin, and for the first ten or twelve miles of its course is fed by innumerable small torrents which drain the rugged slopes

– 307 –

of Mount Franklin and of the lower mountains between it and the Pyramid. About four miles below the former mountain it is joined by the waters of the Ada, a large stream, the sources of which are in Mount Una. The main valley lies nearly due north and south, and that of the Ada, which enters it about three miles above the great moraine already referred to, about northwest and south-east. The glacier which descended from the Pyramid and Mount Franklin being joined by an immense branch from Mount Una, passed down the main valley to a point below that of the Henry, damming back the waters flowing through the latter, so as to reverse the drainage and cause them to flow over a low col into the river Boyle, which joins the Dillon many miles below the great moraine. In the range of mountains which forms the eastern side of the main valley, and about five miles above the moraine, is a low col or saddle leading-into the valley of the Stanley, which, after flowing for several miles on the eastern side of the dividing range, falls into the Dillon below the moraine. Before the glacier had filled the main valley the drainage of the mountains to the eastward of this col was carried into the Stanley River, but, as the col must have been considerably lower than the upper surface of the glacier, a branch of the latter passed over it, filling the valley below it (which lies about north-west and south-east) as far as the valley of the Stanley, in front of which it deposited an enormous load of moraine matter. This col is about three quarters of a mile across, and, no doubt, before it had been invaded by the ice, presented the ordinary features of a mountain saddle, namely, a smooth, rounded summit, with steep pitches into the valleys on either side. But the ice in its passage across it planed it down on the lower side to an even, uniform, and gentle slope for a distance of about half a mile, from which point it plunged abruptly into the valley below, not only scooping out all the material which had previously lain in its bed, but also, in all probability, deepening it somewhat, as a basin. Passing on then to the main valley of the Stanley, it deposited at its edge, and completely across the lateral valley, a huge mass of moraine matter, which extends back into the latter for about three-quarters of a mile.

The space between the col and the inner line of the moraine is now occupied by Lake Guyon, a sheet of water about a mile and a quarter in length and half a mile broad, and (by actual measurement) 60 feet deep in its deepest part, gradually shoaling, however, towards both ends. As the upper surface of the moraine is fully 100 feet higher than the highest part of the col, the drainage has been reversed, the waters of the lake flowing into the Dillon along a channel worn through the col. The scooping effect of the glacier is very evident in this case, for even the surface of the lake lies at least 60 feet below the general slope of the valley of the Stanley, which, in many places close to the great moraine, flows through and over solid rock.

– 308 –

Lake Guyon is being gradually filled with alluvium, carried into it by the streams which flow from the mountains on each side, and, indeed, it is already nearly divided into two separate sheets of water by a peninsula many acres in extent, which has been formed on its northern side by two of the largest of these streams. The amount of solid matter carried into it during heavy rains is enormous, and the circumstance that it has not already been filled is one of extreme interest, when considered in connection with the rapid accumulation of alluvium in the valley of the Dillon.

On the lower slopes of the mountain, immediately above the col, are innumerable roches moutonnées, many of them of great extent, beautifully rounded in general outline, but nowhere presenting the least striation. This, however, may be attributable partly to the fact that they are composed of the same sandstones as the great mass of the blocks deposited on the moraine, and partly to the circumstance that they weather with great rapidity under the alternate action of frost and heat, the disintegrated material being removed by rain.

The general conclusions which I have arrived at from a consideration of the foregoing, and other grounds, are :—

1.

That the land of which the Middle Island is composed began to rise at the close of the mioeene period, and attained its greatest elevation during pliocene times.

2.

That it rose to an elevation of not less than from 4,000 to 5,000 feet greater than its present height above sea level.

3.

That during the period of maximum elevation the land had a quasi-continental extension, chiefly to the eastward.

4.

That during this period the land assumed its present physical aspect, and that as it rose glaciers were formed in the great mountain valleys.

5.

That all the later marine pliocene deposits then formed were formed on the outskirts of the continental area.

6.

That the glaciers, of which the evidences are to be found in those parts of the Middle Island mountains to the north and south of Mount Cook, owe their disappearance to a subsequent depression of the land, which continued during pleistocene times, during which a fresh upheaval subsequently took place.

7.

That during this depression all the later marine pliocene beds of the Middle Island tract were again submerged.

8.

That the Middle Island presents no evidence whatsoever of any such pleistocene glaciation as is mentioned by Dr. Haast.

I may add that, in addition to its picturesque beauty, the tract of country

– 309 –

referred to in the foregoing paper, and especially the part surrounding the Spencer Mountains, presents features which, properly investigated, are calculated to assist materially in solving many moot points in connection with the action of glaciers.