Art. LVI.—On the Date of the Last Great Glacier Period in New Zealand.

Transactions and Proceedings of the Royal Society of New Zealand, Volume 5, 1872, Unnumbered Page

Art. LVI.—On the Date of the Last Great Glacier Period in New Zealand. By Capt. F. W. Hutton, F.G.S. [Read before the Wellington Philosophical Society, 18th September, 1872.] It is an acknowledged fact that the glaciers of the South Island of New Zealand have been at some former time of much larger dimensions than they are at present, and to the period of their greatest extension Dr. Haast has applied the term “Glacier Period.” This term is a very convenient one, and I shall here adopt it with the understanding, however, that it has no relation to, and implies no contemporaneity with, the “Glacier Epoch” of the Northern Hemisphere; for although Dr. Hector (“Geo. Mag.,” 1870, p. 70; N.Z. Exhibition Jurors' Reports, p. 263; Trans. N.Z. Inst., II., p. 372; “Quar. Jour. Geo. Soc.,” 1865, p. 128; and Anniversary Address to the Well. Phil. Soc., 1872), and Dr. Haast (Formation of the Canterbury Plains, pp. 7, 14, etc.) refer our last glacier period to pleistocene times, that is to about the same time as the glacial period of Europe, I think I shall be able to show that it is in reality far older, or at any rate that the subject is one that admits of discussion. No New Zealand geologist advocates now a cold or glacial period to account for the former extension of our glaciers, for, as Dr. Hector has pointed out in his anniversary address for this year, there are no signs of any till or marine formed boulder-drift to be seen, and our pleistocene and newer-pliocene fossils show that no very great reduction of temperature has occurred in these latitudes since those times.* It can by no means be inferred from this that there has been no “Glacial Epoch” in the Southern Hemisphere, for no glacial (as distinguished from glacier) phenomena are found in Europe south of 50°, and it is probable that if no land now existed north of that parallel of latitude the occurrence of a glacial epoch would never have been suspected. In the pleistocene deposits of Wanganui we find Triton spengleri, Cassis pyrum, and another extinct species of the same genus, Ancillaria australis, Turbo granosus, Imperator imperialis, Rotella zealandica, and Labio zealandicus, as well as Mesodesma chemnitzii and M. cuneata, none of which probably would have been able to survive a reduction of temperature sufficient to cause so great an extension of our glaciers as we know to have taken place. In the newer-pliocene beds which form the lower series at Wanganui (Geo. Reports, 1872, p. 182) we also find a Typhis, the same extinct species of

Cassis that is found in the upper series, Ancillaria australis, two species of Cladopoda one of which is still living, Imperator imperialis and Rotella zealandica, showing that here also we cannot call to our aid any great diminution of temperature. We have no marine deposits in New Zealand of older pliocene date, for, as I shall subsequently show, the land then stood at a much higher level than it does at present, consequently we have no proofs here, either one way or the other, of a change of climate, but as the elevation of the land would, if high enough, be able by itself to account for all the phenomena, there is no necessity for calling to our aid any other cause. During miocene times our climate was warmer than at present, as is proved by fossils of the genera Conus, Mitra, Marginella, Crassatella, Limopsis, Perma, and the large species of Cucullœa and Cardium which then inhabited our seas.* This is the usual palæontological argument, but I believe that, when applied to extinct species, it may lead to very erroneous deductions, and that when opposed to physical arguments it is of no weight at all. We must therefore necessarily infer that the greater extension of the glaciers was caused by greater elevation of the land†, † An elevation of from 2,000 to 3,000 feet would be sufficient to account for all the phenomena, while an elevation of 550 feet would connect the two islands. and their subsequent reduction in size was caused by subsidence, and so far Dr. Hector, Dr. Haast, and myself agree. Dr. Hector, however (“Jour. Roy. Geograph. Soc.,” 1864, p. 103, and “Geo. Mag.,” 1870, p. 70), and Dr. Haast (“Cant. Plains,” p. 14, and “Quar. Jour. Geo. Soc.,” XXI., p. 131) appear to attach considerable importance to the erosion of the vallies by the glaciers reducing the area of land above the snow line. This appears to me to be an unnecessary and exaggerated view of the rapidity of glacier erosion, ‡ On this subject see an excellent paper by the Rev. T. Bonney in the “Quar. Jour. Geol. Soc.,” 1871, p. 312. and the fact that eocene or miocene marine rocks are found far up many of these vallies, such as at the Rakaia, Canterbury, and some twelve miles above Queenstown in Lake Wakatipu, proves beyond dispute that they had attained to nearly their present size in eocene times. Dr. Hector also assumes (Trans. N.Z. Inst., II., p. 373) that the chief erosion, by which the vallies are eaten back by the glaciers, takes place at the abrupt fall known as the “ice cascades.” But the friction, and therefore the power of erosion, of any solid body like ice must vary as the cosine of the angle of inclination, and consequently the greater the slope the less the erosion. The maximum of erosion must necessarily be at the upper angle of the ice cascade, where the ice bends downward by its own weight, and consequently the effect would be the gradual reduction of an abrupt fall to one of gentle inclination, and this is fully borne out by the fact

that the majority of glaciers have no ice cascade at all, and nearly all partake more of the character of “ice-rapids” than “ice-falls.” As this point, however, is of minor importance it is not necessary to pursue it any further. Dr. Hector and Dr. Haast base their opinion that our last glacier period was in pleistocene times on the supposition that since then great subsidence has taken place. I will, therefore, in the first place give the reasons that have led me to an opposite conclusion, viz., that during the whole of the pleistocene period elevation has been more or less continuous over the greater part of New Zealand, and I will then give the evidence in favour of the glacier period having been in older-pliocene times. It is so universally acknowledged among geologists that river terraces prove elevation, that it is quite unnecessary for me to go over again such well trodden ground.* Consult, inter alia, Lyell's “Elements of Geology,” 6th ed., pp. 118 and 120, and his “Student's Elements,” p. 79; Dana's “Manual of Geology;” “Geological Magazine,” 1871, pp. 75, 190, 239, 333, 430, 526, 574; ditto, 1872, p. 48; Ramsay's “Physical Geology of Great Britain,” p. 109; Jukes and Geikie's “Manual of Geology,” p. 402. There may be some cases where, in a mountainous country, rock-bound lakes have been filled up before the outflowing river had cut down through the rocky barrier to its normal level, and where consequently terraces might be afterwards formed in the old lacustrine deposits as the level of the river was lowered, or where a stream of lava from a volcano has blocked up the course of a river, and thus led to a similar result; but these are exceptional cases which can always be recognized by the terraces being formed of stratified silt or fine sand, but never of shingle, and it cannot possibly apply to rivers running through plains or broad vallies. Now throughout the South Island of New Zealand on both sides of the Alps, and in the central portions of the North Island, all the rivers appear to be deeply terraced. I know, from personal observation, that this is the case with the rivers of Southland, with the Aorere in Golden Bay, and with the Waipa and Waikato in the province of Auckland. Dr. Haast describes the same thing with the rivers flowing into the Canterbury Plains. (Report on Cant. Plains, Christchurch, 1864.) Mr. Buchanan describes terraces in the vallies of the Awatere and Eden rivers (Geo. Reports, 1866–67, p. 41), Dr. Hector describes those of the west coast of the South Island (Progress Report, 1866–67, p. 29) including the Buller (l.c., p. 32), and also of the Mohaka (Geo. Reports, 1870–71, p. 160) in the province of Napier, and Dr. Hochstetter (“New Zealand,” p. 467) and Mr. Travers (“Quar. Jour. Geo. Soc.,” 1866, p. 259) describe those of rivers in the province of Nelson. But there are many other proofs of recent elevation besides that afforded by the river terraces. The Southland plains show unmistakable marine action. Towards its mouth the Jacobs River runs through extensive deposits

of limestone and calcareous sandstone of upper-eocene age; these rocks are all planed down to a uniform level and covered by a thin layer of silt and gravel, which is at the same altitude as the thick gravel beds that form the other parts of the plain; the same thing on a smaller scale occurs in the Waimea plains, in Southland, where the eocene limestone is also covered with gravel, and cannot be distinguished in outline from the terraces; we know of no agency but marine denudation that could effect this. The seaward plains show their origin still more distinctly by their uniform level all round the southern face of the Hokanuis, and from the terraces being sometimes arranged more or less parallel to the present coast. It is no objection to the marine theory that beds of lignite are found under these deposits, on the contrary it is much in favour of it, as it is well known that vegetable remains are very sparingly distributed in river alluviums, for they are scattered widely by the currents, while, on the other hand, we know that most coal seams are covered by marine beds. Indeed the occurrence of vegetable remains on a large scale below alluvial plains is a certain proof that those plains were formed either by lakes or by the sea, and not by rivers. These vegetable deposits accumulated during the depression which preceded the elevation. Of the Canterbury plains I speak with much diffidence as I have not visited them, and because Dr. Haast, after a careful examination, has come to the conclusion that they have been formed by the rivers during a long course of depression (Report on the Canterbury Plains, 1864); but judging from the sections that he gives in his report I cannot understand how they can have an entirely fluviatile origin, for in a line parallel to the coast they are nearly level from the Waimakariri to the Rangitata, the highest portion being about the Waimakariri. Now each of the rivers must have poured out an amount of detritus proportional to its size, and therefore the plains should be higher about the larger rivers than about the smaller ones; but the fact is that the country about the Hinds and Ashburton, two small rivers, is higher than that about the Rangitata, a large river, and nearly as high as that about the Rakaia, the largest river on the plains. If we suppose that the larger rivers after raising their own beds wandered about the plains helping the smaller ones, I can then see no reason why all the smaller rivers should have afterwards left the common channel, and each pursued its own way direct to the sea. Neither does Dr. Haast explain how it is that the gravel formation of the plains wraps round the spurs of the hills at the same level that it has at the river gorges, nor how it is that the plains of the Rakaia and Waimakariri are nearly at the same level on each side of the Malvern hills, while the beds of the rivers are at very different levels, nor why the tertiary rocks between the junction of the Kowai and the Gorge hill, and at the gorge of the Rakaia, are levelled on the top. Dr. Haast not only believes in a general

subsidence, but also that this subsidence has been greater on the west than on the east coast; consequently, according to his theory, the velocity of the rivers must have been considerably reduced, and he has not informed us how it is that they have been enabled with their reduced velocities to cut through and remove the alluvium which they could carry no further, but deposited when their velocity was greater. All these things, as well as the occurrence of vegetable deposits below the gravel, are readily explained by supposing the plains to be a marine formation since elevated, but are, I think, quite inexplicable on the river formation theory alone. I might also fairly ask, if rivers form such large level plains in New Zealand why do they not form the same in other countries? Why are there no broad level gravel deposits like the Canterbury plains round the foot of the Himalaya, Alps,* The gravel deposits of Switzerland are of quite a different character, and are the grundmoränen, or moraines profondes of glaciers., etc.? My answer would be because none of these places have been lately submerged below the sea. That the greater part of the shingle of the Canterbury plains has been brought down by the rivers from the mountains I do not dispute, and I also acknowledge that, as the plains were elevated, the rivers must have often changed their courses and wandered over a large part of the plains near the then shore line, all that I contend for is that the materials brought down by the rivers have been rearranged by the sea, and the shape of the stones would therefore depend upon the length of time that they had been subjected to wave action, and on the amount of sand in which they are imbedded. The silt deposit upon which a large part of the town of Lyttelton is built is also evidently a recent marine deposit, but I do not know to what height it extends above the sea. Mr. W. T. L. Travers has pointed out to me that the land side of the hills forming Banks Peninsula shows no trace of marine erosion, and this is the most formidable objection to the elevation theory that I have as yet met with. It would be very easy to say that as Banks Peninsula is volcanic it may have been thrown up or elevated more rapidly than the plains, and at a later date, but there is no proof of this, and until that can be given I could not accept it as an escape from the difficulty, but we must remember that the land side would not have been exposed to a heavy surf, and that the rapid decomposition of the volcanic rocks might soon obliterate all traces of a sea cliff. Mr. C. Forbes states (“Q. J. Geo. Soc.,” 1855, p. 526) that “there is abundant evidence to prove that at a very recent period the Peninsula was an island.” The absence of fossils in the Canterbury plains is easily explained, indeed we could hardly expect any to occur, for those shells that were not completely pounded to pieces on the shingle beach would be rapidly dissolved out, on

emergence above the sea-level, by the ready percolation of rain-water containing carbonic acid through such porous strata. Dr. Haast also describes at Timaru (Report on Timaru district, p. 4) silt, underlaid in places by fine clay or gravel, covering basalt, and sloping up from the sea to a height of 686 feet, and containing recent marine shells near the sea (Cant. Plains, p. 8). Mr. Hacket gives evidence (Geo. Reports, 1868–69, pp. 10 and 11) of a rise of the land near the Okarita and Waikukupa Rivers, on the west coast of the South Island. It is evident that the deposits he here describes are not ordinary morainic accumulations, but it remains yet to be proved whether these beds were deposited in a lake or in the sea, or whether they are of morainic origin at all. Dr. Hector, in describing the gold fields of the west coast of the South Island (Progress Report, 1866–67, p. 29), says that the gold drifts have been “carried out from the mountains by the rivers, and deposited upon a gradually changing coast line. They thus have a general distribution parallel to what was the western shore of the island at the epoch of their deposit; and by tracing the successive lines of elevation, and allowing for the consequent changes which have occurred in the direction of the drainage channels of the country, we are enabled to form an opinion as to the extent and position of the auriferous leads.” Further on he speaks of the first group of auriferous alluviums as being the “earliest formed and most elevated of these drifts,” but he does not give the height to which it attains; but the third group he calls “beach terraces which extend to an altitude of 220 feet, and mark several changes in the level of the shore line within a comparatively recent geological period,” so that we must infer that the first group attains a greater height than 220 feet. At Taranaki, Dr. Hector also describes (Progress Report, 1866–7, p. 3), “pleistocene deposits consisting of stratified gravels and sand-rock, with beds of lignite,” reaching an altitude of 150 feet above the sea, which he says “must be regarded as in some way connected with an ancient coast line, and from the circumstance that at the base of this formation in many places, and underneath the lignite seams, there is a layer of rolled broken shells of existing species, we may infer that these gravels have been deposited in lagoons parallel with the coast line during a gradual elevation of the land, and that they have been overtaken, as it were, by the encroachment of the sea, and exposed in the sea cliffs after they are 80 to 100 feet above the present level of the tide.” Mr. R. Pharazyn, in a paper read to this Society (Trans. N.Z. Inst., II., p. 158), gives evidence of recent elevation near Wanganui; and Dr. Haast (Report on the Cant. Plains, p. 8) at Timaru. In the Nelson province Mr. W. T. L. Travers also (“Quar. Jour. Geo. Soc.,”

1866, p. 256) describes extensive deposits of post-pliocene gravels and sands, the materials of which “are all water-worn; and exhibit the common appearance of river or beach shingle.” These deposits attain an altitude of more than 2,000 feet above the sea; they “are in no way cemented, very little inclined in stratification, and in many places exhibit perpendicular sections several hundred feet high;” they can, therefore, hardly be due to river action.* Since writing the above I have examined these beds and find that they are of præglacier date, and consequently have nothing to do with the present argument.—F.W.H. The sandy beds at Wanganui contain about 10 per cent. of extinct shells, and must therefore be referred to the earlier part of the pleistocene period, that is, the glacial epoch of Europe. I do not know to what height above the sea they attain, but Dr. Hector states (“Cat. Col. Museum,” 1870, p. 172) that they are 100 feet thick at Wanganui. Raised beaches of pleistocene, or of almost recent, age are found at Motanau, in Canterbury, and on the north-west side of Cape Kidnappers, in Hawke Bay, but I do not know their altitude. On the north-west side of Hicks Bay, near the East Cape, there is a very distinctly marked line of inland cliffs; and the same thing is seen in Cook Strait, near Wellington. Besides all this evidence, Dr. Hector admits that recent raised beaches, from 15 to 25 feet above the sea level, are found in places all round the coast. The objections which Dr. Hector raised, in his anniversary address to our Society for this year, to a recent elevation of New Zealand, or rather the reasons he adduced in favour of recent subsidence, are:— 1. That vallies on the west coast of Otago, in the northern parts of New Zealand, and in other places, are depressed far beneath the level to which they could have been eroded, and that they are prolonged beneath the water level, forming deep water inlets and harbours. I quite agree with Dr. Hector that these vallies must have been eroded when the land stood far higher than it does now; but they afford no proof of a recent depression, for, as I have already stated, we have unmistakable evidence in the eocene and miocene tertiary rocks found far up some of them that they were formed in the earliest tertiary times, and they no more prove recent subsidence in New Zealand than similar fiords do in Norway, Tierra del Fuego, and the west coast of Scotland, all of which are known to have risen during pleistocene times. 2. That “the low shelving and sandy parts of the coast have a heaped up shore line, that appears as if encroaching on the alluvial deposits.” I do not know to what particular parts of the coast Dr. Hector alludes, but this appears to me to be no evidence, either one way or the other, for if the conditions were favourable for the formation of sand-dunes, they would appear to encroach on the recent marine deposits, as they rose above high water-mark, in the same way that they would be formed over alluvial deposits when sinking.

3. That with the exception of a raised beach, nowhere raised more than 20 feet above the sea level, “there is a total want of any inland cliffs, lines of sand-dunes, and ridges, and other familiar evidences of an emerged coast line.” I do not think that inland cliffs can by any means be called “familiar evidences of an emerged coast line,” because everywhere they are the exception and not the rule. Nothing is so soon obliterated as an inland cliff, and few things are rarer to find, except close to the sea; but I have already quoted Dr. Hector himself as mentioning beach terraces on the west coast of the South Island which attain a height of 220 feet above the sea; and I have also mentioned inland cliffs in Cook Strait, and near East Cape, both of which are certainly more than 100 feet above the sea level. Mr. Traill also describes (Trans. N.Z. Inst., II., p. 169) an old sea beach near Oamaru, “elevated considerably above the present one.” Mr. P. Thompson describes (Trans. N.Z. Inst., III., p. 263) two series of sand-dunes at Wickliffe Bay, in Otago, one much older than the other, and covered with grass; and in his “Catalogue of the Colonial Museum,” Dr. Hector classes the Upper Wanganui series as a raised beach. Mr. Buchanan also gives a section (Geo. Reports, 1866–7, p. 36) showing pleistocene gravels containing marine fossils at a considerable elevation above the Clarence River, in Marlborough. 4. That “the low country is invariably formed of marine strata of higher antiquity than the period of the extension of the glaciers.” As Dr. Hector supposes the extension of the glaciers to have taken place in pleistocene times it is of course difficult to find any strata younger than this; but even on his view the pleistocene beds at Wanganui and Taranaki must be either younger than, or of the same age as, his glacier period; and if I am right in referring the glacier period to older-pliocene times, we have the newer-pliocene beds of Shakspeare Cliff and Patea, younger than the glacier period. 5. That the Canterbury plains have been “overwhelmed by shingle deposits brought from a higher level by the rivers,” and have an old drift-wood bed below them at 80 to 90 feet below the level of the sea. This point I have already discussed, and shown, I think, that as yet it is far from certain that these plains have been formed altogether by the rivers. The drift-wood bed simply proves oscillation of level, and it was probably formed during the subsidence that took place in newer-pliocene times, and is perhaps older than the lower Wanganui beds of Shakspeare Cliff. 6. That there is no pumice drift at high altitudes in land-locked harbours like Wellington, although it is found at low levels. On this I would remark, that the seaward slopes round all these harbours are very steep, and that pumice is very light, and easily washed away. It also very rapidly decays when kept wet, much of that which comes down the Waikato being half rotten already. These causes are, I think, sufficient to account for the disappearance of pumice in the course of time.

7. As a last reason, Dr. Hector cites my paper on the Lower Waikato deposits, read before the Auckland Institute (Trans. N.Z. Inst., III., p. 244), in which I state that there is no evidence of the sea having ever been in the Lower Waikato valley. This I certainly think shows that the Lower Waikato district has not risen more than 50 feet during the pleistocene period, but it does not affect other parts of the Island. Indeed, there can be no doubt that the elevation has been very unequal in different districts. The central portion of the North Island appears to have risen most, and next to that the central portion of the South Island, while the whole of the northern portion of the province of Auckland does not seem to have risen more than 20 or 30 feet, but we are almost without data at present to estimate these differences correctly. I do not think, therefore, that the reasons brought forward by Dr. Hector by any means prove that subsidence has been going on during the pleistocene period, on the contrary I believe that nearly the whole of the evidence is in favour of elevation. At Shakspeare Cliff, Wanganui, and at Patea, in the province of Wellington, we find marine strata containing fossils of which about 24 per cent. are extinct. These beds must be referred to the newer-pliocene period, and this, therefore, cannot have been the time of elevation and extension of the glaciers. The next set of beds, however, below these contain about 59 per cent. of extinct species, thus proving that a long interval of time must have elapsed between their deposition and the newer-pliocene period, which is quite unrepresented in New Zealand by marine strata. I refer these lower beds, which are found at the Awatere, the Port hills at Nelson, the White Cliffs of Taranaki, Awamoa, etc. (see Geo. Reports, 1872, p. 183), to the upper-miocene period; and it is therefore the older-pliocene period that is unrepresented. But not only is there a great difference between the fossils of these two formations, but there is also a great difference in their stratigraphical position, and in the amount of sub-aerial denudation that they have respectively undergone. The older formations always show a broken outline, deeply eroded into hills and vallies, and in some places the beds are tilted at high angles; while where the newer-pliocene beds form the surface level plains cut by narrow ravines only are found (Pharazyn, Trans. N.Z. Inst., II., p. 158). These facts are, I think, sufficient to prove that the older-pliocene period was a period of upheaval, and it is therefore to this time that I refer the last great extension of our glaciers. If Dr. Hector's views are correct as to the glacier period having been in pleistocene times, we shall have to find some reason for the newer-pliocene deposits not being more denuded than they are; for, according to this theory, they once stood at a much higher level than they have at present.

But independently of all these reasons we find, I think, in the marks of former glaciation themselves strong evidence of a very ancient date. In Otago, about Lake Wakatipu, which is the only glacier region in New Zealand that I have had an opportunity of examining, the evidence of the former extension of the glaciers rests almost entirely on the presence of moraines and roches moutonnées, which are the most permanent marks that a retreating glacier leaves behind it. All the more perishable ones, such as blocs perchés and striæ, are almost entirely absent. I searched in vain on some beautifully rounded surfaces of rock near Queenstown for striæ, but all had been obliterated by decomposition, and the only striæ that Mr. J. McKerrow and myself could find were on a few loose boulders at the head of the lake. The absence of striæ, blocs perchés, and other well-known glacier marks, forms, a remarkable contrast to what obtains in the Alps, North Wales, and the south-west of Ireland, all of which districts I have personally examined, and this alone would make me refer our glacier period to a time long antecedent to the glacial period of Europe. Striæ may be more common in other parts of New Zealand where the rocks are harder than they are in the South, but the absence of any descriptions of them in the reports of Dr. Hector and Dr. Haast, beyond general statements that such exist, makes me think that they must be far from common. I am therefore of opinion that the last great extension of our glaciers was in older-pliocene times, when the land stood far higher than it does now; that the newer-pliocene was a period of subsidence, followed by elevation in the pleistocene period, and that that elevation is probably still going on. Note.—November, 1872. The fact that several species of birds and insects are different on the two islands of New Zealand would be considered by nearly all naturalists as a good proof that these islands have been separated longer than Great Britain from Europe, that is to say, previous to the pleistocene period;* Godwin-Austen, “Quar. Jour. Geo. Soc.,” VI., p. 94. but an elevation of 500 feet would obliterate Cook Strait and join the two islands together, consequently New Zealand cannot have stood at an elevation 500 or 600 feet higher than at present since the pliocene period. We are, therefore, driven to adopt one of two suppositions, viz., either that the former extension of the glaciers was caused by an intense cold or glacial period, or by elevation of the land in præ-pleistocene times.—F. W. H.

The formation of Lake Wakatipu has been ascribed by Dr. Hector (“Mining in New Zealand,” Trans. N.Z. Inst., II., p. 374) to unequal subsidence, while Mr. J. McKerrow (Trans. N.Z. Inst., III., p. 256* In this paper Mr. McKerrow points out, I believe for the first time, the very important fact that the constrained flexure of a solid body like ice, when passing from one angle of inclination to another, would greatly increase the friction at this particular point.) ascribes it to the erosive action of a former glacier, in accordance with the theory put forward by my esteemed friend and instructor Professor Ramsay, which theory he has, in my opinion, reduced to a demonstration such as is rarely seen in geological inquiries. I need therefore scarcely say that I agree with Mr. McKerrow as to the glacier origin of Lake Wakatipu, but in order to establish my opinion I will examine what would be the results that an acceptance of the unequal subsidence theory would lead us to. Lake Wakatipu is a rock-bound basin, lying in a nearly north and south direction, with a sharp east and west deflection in the middle; consequently, if it has been formed by unequal subsidence, this subsidence must have been most in a northerly direction, for if it had been in a westerly direction the end of the lake would have been at Queenstown. Of course, a southerly elevation would produce the same effect as a northerly depression. Now the deepest part of Lake Wakatipu is 1,400 feet, off Collins Bay, about 16 miles north of Kingston (McKerrow, l.c., p. 254), consequently, to change a horizontal surface to one having this slope, Kingston must have been elevated 1,400 feet more than Collins Bay, and this would give an elevation to the coast of Southland of 8,400 feet, and to Stewart Island an elevation of about 11,000 feet; or, if the lake was caused by depression, Collins Bay must have been depressed 1,400 feet more than Kingston, which would give a depression to Jackson Bay on the west coast, due north of Kingston, of 9,800 feet. These are the very least movements that can account for the phenomena, for if Kingston was not the axis of the movements, or if the movements had taken place in any other direction than that of the length of the lake, they would have to be immensely increased in order to bring about the same result. But most of the sounds, the origin of which Dr. Hector also ascribes to the same movements, are south of Lake Wakatipu, and are therefore situated in that part which would have undergone the least depression, so that they are just where they ought not to be if this unequal depression has occurred over the whole country. Lake Monowai also requires that the depression should have been towards the south instead of the north. If also movements on such an extensive scale had taken place since the pleistocene period, the river system of Otago would be certain

to show evidence of it by some, at least, of the rivers being deflected to the north; but the very contrary is the case, for the Jacobs River, Oreti, and Mataura have all been deflected towards the south. The evidence, indeed, of the river system goes to show that the central part of the South Island has been more elevated than the southern part, a movement which must have tended to empty Lake Wakatipu. The next supposition that we can make is that this unequal elevation and depression was not universal but local, the country north of Lake Wakatipu having alone been depressed. If, however, this local depression occurred between the head of the lake and the west coast it would have emptied Lakes Wanaka and Hawea which lie north of it; and if the depression was north of Lake Wanaka it would have emptied, in the same way, Lakes Pukaki and Tekapo, which are further north again, for all these lakes lie in a more or less north and south direction, with the south end dammed up. There appears, however (Hochstetter's “New Zealand,” p. 484), to be one place, off Cliffy Head, from which many of these lakes radiate, so that if the depression had taken place there it might perhaps have formed them all. But if we assume this, we again encounter those difficulties that I at first pointed out; for by this theory the central part of the Island must have been depressed at least 15,000 feet more than the north and south, and the deep sounds instead of being found on the south-west of Otago and the north of Marlborough, should occur in Canterbury, and the rivers should be deflected to the north in Otago, and to the south in Nelson and Marlborough; for the fact of all the rivers on the Canterbury plains having cut deep gorges through the alluvial deposits, shows clearly that their volocities have not been reduced by a greater sinking of the west than of the east coast. The Nelson lakes, moreover, would require some different arrangement again to account for them. We may still make a third supposition as to the formation of Lake Wakatipu by supposing it to be owing to a small local subsidence in that area alone, but this is disproved by the regularity of the strike of the rocks from one end of the lake to the other, and the dip of the beds is so slight that any movement by which the upper or central portion of the lake had been depressed could not possibly escape detection, and I had this constantly in my mind when examining the district last summer. In order, therefore, to explain the formation of these lakes by unequal subsidence, and at the same time to account for other phenomena observed round the coast, we should have to imagine such a complicated system of local depressions and upheavals that they would more resemble the contortions produced by lateral pressure than any movements that we know, or have any right to assume, are going on at the surface of the earth. In the section given by Dr. Hector (Trans. N.Z. Inst., II., p. 372) we see

that there are two other rocky barriers below the one at Kingston, and that each of these barriers is capped by morainic accumulations. Now, by Professor Ramsay's theory of the glaciers having excavated the lake basins, this is just what we might expect; but on the subsidence theory we have to account for these three moraines having all been deposited exactly at the places where the former glacier having got over the flatter ground was just commencing to descend a steeper slope. These are the reasons which lead me to think that Lake Wakatipu was scooped out by a glacier, and that it cannot possibly have been caused by unequal depression; indeed, if Professor Ramsay's theory was in want of further evidence to prove it, I know of no place equal to the province of Otago for obtaining that evidence, for the rocks there have been so little disturbed that recent elevations and depressions could be generally proved or disproved.