Go to National Library of New Zealand Te Puna Mātauranga o Aotearoa
Volume 23, 1890
This text is also available in PDF
(1 MB) Opens in new window
– 311 –

Art. XXX.—On the Drift in South Canterbury.

[Read before the Philosophical Institute of Canterbury, 3rd July, 1890.]

The Transactions of the earlier seventies show that fifteen to twenty years ago the glacial or great glacier period, of New Zealand was a favourite topic of discussion, with members of the Institute. Since then the subject appears to have been laid aside, though it was by no means threshed out, differences of opinion remaining as to both the cause and the extent of the glaciation. In the course of the controversy the following remark was made by Mr. A. D. Dobson: “I believe that a careful examination of the drift-formation of New Zealand generally will throw a great deal of light upon many points at present unsettled, and that until the age of the drift is satisfactorily settled we may speculate in vain upon the origin of the glacial period.”* Having long been familiar with what I take to be portions of the formation Mr. Dobson referred to in South Canterbury, it occurred to me on meeting with Mr. Dobson's remark recently that a few notes on the drift of this district would be acceptable to members of the Institute. I propose to go over somewhat wider ground than the drift-formation only, sufficiently to trace, to the best of my ability, its relations to the preceding and succeeding formations, and., viewing the drift as the record of a particular condition of climate, my notes will have special reference to the changes of climate registered by the deposits dealt with.

In the plains and downs of South Canterbury east of the ranges, omitting recent and incomplete deposits, we have, in descending order,—

1. The great shingle-fans of the plains, including smaller fans laid down by the smaller rivers, such as the Hare, Waihi, and Opihi.

2. An extensive but irregular deposit of loess, upon areas elevated above the reach of the fan-building rivers

3. Sheets and streams of doleritic lavas at two points, Timaru and Geraldine.

4. The drift, a vast mass of stream-laid gravel and sand, with interstratified clays and silts, the shingle highly oxidized and decomposed, the whole formation greatly denuded, and more or less disturbed by subterranean action.

5. A series of fresh-water beds of totally different kind, these being almost wholly of fine, material. These beds also show local, and almost certainly independent, disturbance.

[Footnote] * Trans. N.Z. Inst., vol. vii., 1874, p. 444.

– 312 –

6. Fine-bedded, loose, rusty, unfossiliferous sands, which appear to be rearrangements by fresh-water of portions of (7).

7. Fossiliferous sands of the Pareora (marine) formation.

Working upward again, at the upheaval of the Pareora sea-bed as a starting-point, we commence with a temperate, or rather warm, climate, as shown (according to the authorities) by the character of the Pareora fossils. This evidence of the shells is fully corroborated by that of the physical character of the marine beds close to steep mountains. The inorganic material is uniformly fine; there is no evidence of frost-work in gravels, conglomerates, or breccias.

The emergence of a sandy sea-bed would certainly be followed by the rearrangement of some portions of it by rains and streams, and to such a process, I think, must be referred certain loose, rusty, fine-bedded, unfossiliferous sands, seen in terrace sections in the valley of the Tengawai between Pleasant Point and the Cave,; under the gouged western, face of Mount Horrible, and elsewhere. These fine-bedded sands are of no great thickness, under 20ft. probably, as a rule, and they rest upon sands of the same mineral character, but fossiliferous, and massive instead of fine-bedded.

Deposited conformably upon the rearranged sands in the Tengawai sections is a series of fresh-water beds, showing 150ft. or 200ft. in thickness in one exposure—Howell's Bluff —and they may have been much thicker. These beds consist of alternations of silts, clays, white sands and grits, and dirt-beds. Some of the clays are full of beautiful plant-impressions, and these should tell an instructive story to the botanist. The physical characters of the beds, however, sufficiently attest the warmth of the climate during their deposition. The white grits perhaps supply the clearest and simplest evidence. They aggregate, in the Howell's Bluff section, 30ft. or 40ft., in three principal layers. The coarser grits are quite loose, and well sorted and stratified. The fragments range in size up to that of small peas. Most of them are well waterworn, but a considerable percentage are quite angular, only a little worn. These are easily recognized as vein-quartz, and they suggest that the whole of the quartz—grits and sands—was derived from the highly metamorphosed, intricately-veined, slaty or schistose rocks of The Brothers Mountain, four or five miles away. Save for some rolled pellets of clay containing carbonized vegetable matter, the grit beds are of pure quartz, and the total absence of slate pebbles shows clearly that in the denudation of the mountain, by which the quartz was set free, frost played no part. The whole series appears to have been of slow growth, and there are marks of several pauses in its deposition, in the dirt-beds and drought-veined surfaces of

– 313 –

clay-beds. Beyond the Cave, on the opposite side of The Brothers Mountain, is a seam of good lignite, about 2ft. thick, associated with a dirt-bed of greater thickness, and white sands and clays. These rest upon a denuded surface of blue fossiliferous (Pareora) sand-rock, and the two formations have been steeply tilted together. The fresh-water and carboniferous beds must belong to the same period as those lower down the river, and I feel sure some of the lignites of the Upper Kakahu are of the same age. Silts, clays, and sands resembling those of the Tengawai (but with less whitish material), with a dirt-bed, and a seam or pocket of lignite near the base, are exposed in the south bank of the Hare River, near Geraldine. This section contains two or three thin beds of quartz-gravel, with some admixture of sandstone pebbles. The quartz is evidently vein-quartz; much or most of it is angular, and, as the section is only two or three miles from hills of slaty rocks that yield readily to frost, these beds, equally with those of the Tengawai, show that during the time of their deposition a practically if not perfectly frostless climate prevailed in this region. The base of the Hare beds is not exposed, but in sinking through the denuded crown of the arch into which they have been thrown, some years ago, in search of lignite, I soon reached loose red sands, resembling those I have since seen beneath the Tengawai fresh-water beds. I have no doubt there are other exposures of these beds, but the two described are the only ones I have examined.

The Tengawai beds were disturbed and denuded before any later beds of different material were deposited upon them. No such gap of unconformability exists in the Hare section, where the fine-grained series is regularly overlain by red gravels of the succeeding formation. The junction has been brought into view by denudation following a disturbance which elevated both together into an independent anticlinal fold.

This brings us to the red gravels, which I take to be the “drift,” or “dispersed gravels,” of writers in the Transactions and Geological Reports. As these terms, however, seem to imply a partial or complete ignorance of the mode of origin and deposit of this formation, neither of which does there seem any necessity for admitting so far as the red gravels of South Canterbury are concerned, I prefer to avoid them, and to use the descriptive, term “red gravels.” They form “red cliffs” everywhere.

In South Canterbury the red gravels, usually with a covering of loess, form a large part of the country east of the ranges, generally in the shape of downs, whose easy contours and good loess soil make them first-rate agricultural country. They

– 314 –

have been greatly denuded, cut into by streams, and their structure is exposed in many fragmentary natural sections, and on a smaller scale in numerous road-cuttings. I have examined them in the Geraldine, Waitohi, and Cave—Albury downs, and in the Timaru plateau. The same formation stretches from Albury to Fairlie Creek, and from Timaru to Waimate. Everywhere near the hills the gravels present the same broad features—beds of stream-laid subangular shingle, and sand, separated by clays and silts. In places the formation is hundreds of feet in thickness; in Geraldine Hill, I believe, nearly, if not over, 1,000ft., and in the Cave—Albury downs it must be over 500ft. Generally the shingle is coarser and the clays less numerous towards the top. Wherever I have examined them the gravels are unquestionably stream-laid, and generally, I believe, are of quite local origin. The Geraldine downs, however, contain pebbles of a volcanic rock, which must have come from somewhere northward, and across existing drainage-lines. (These are pebbles of a dense, tough, micro-crystalline rock, black in colour, with minute white specks; the stone weathers to a pale blue, of chalky softness.) I have read that there is volcanic rock at the head of the Orari. There is none in the hills opposite the Geraldine downs. At Timaru, the point furthest from the hills where the red gravels appear, only the upper portion is exposed to observation, the formation sloping just down to sea-level. Here the shingle is largely replaced by finer material. About two miles inland is a gravel-pit, disclosing about 30ft. of the uppermost beds, two-thirds or so of which are of silts and clays, the rest of small shingle, in two layers. The silts are fine-bedded, and at more than one level contain vestiges of having been vegetated. A boring put down at the freezing-works—commencing at the bottom of a bricked well sunk 43ft. into the gravels, the nature of which was not recorded—passed through 40ft. of shingle, 39ft. of clay, and 25ft. of sand, each in several layers. The thickest bed was of fine yellow clay, 12ft., the base of which for about 1ft. was stained dark-brown as if by humus, and contained the rust-granules common in wet soils. The clays near the hills vary from a few inches to 10ft. or 12ft. in thickness, and they vary much in character, some being fine and white, others blue, most of them yellow, and of all degrees of coarseness up to silt. In some places beds of clay show abrupt changes of thickness, as if from contemporary denudation. Some of the beds show drought-veins in their upper surfaces, the mark of their having been for a considerable period persistent land - surfaces under a dry climate. The silts have the character of flood-deposits, and some of the clays probably had a similar origin. Others require further examination to decide whether they are not

– 315 –

really loess of the same kind which afterwards overspread so much of the formation, i.e., wind-blown dust. One of the clay-beds related to the red gravels appears to be of this nature. It underlies the dolerite at Mount Horrible, and directly overlies rearranged marine sands, at a spot which must always have been above the reach of rivers. The only doubt as to this being a loess-bed is founded on the possibility of its being a slope-deposit, or rain-wash, from some eminence which has since been removed.

The shingle is everywhere highly oxidized and much decomposed. It is somewhat harder as well as coarser near the hills; but generally, and the smaller shingle everywhere, is too rotten to be of any use for road-making. At Timaru, pebbles whose smooth surface shows that they were once hard enough to take a fine polish can now be cut with a knife like chalk.

The great quantity of the red gravels implies a correspondingly great denudation of the neighbouring mountains, from which the material was derived, plus a liberal allowance for finer material carried further away. The greater part of the formation being of shingle is proof that frost was the agent of denudation. Large deposits of river-shingle can only be produced in one way—by frost and thaw breaking up the surface of mountains steep enough to enable gravity to deliver the spoil to streams sufficiently powerful to transport it so long as their channels are confined and the fall great, but not powerful enough to transport it all when the fall is reduced and the channels become indefinite. The form of deposit—except in deltas, and these are not deltas—is necessarily the fan, modified, it may be, by the work of other streams or by obstructive eminences. From my observations I conclude that in the red gravels of South Canterbury we have the decomposed, disturbed, and denuded remains of a coalescing system or systems of fans, laid down by small streams issuing from the neighbouring low ranges, as the larger fans of the plains were laid down by the larger rivers at a later date.

Mr. Sandham Gillingham, who owns an estate on the Albury-Fairlie Creek downs, and who has paid much attention to their formation, is also of opinion that those downs were piled up by the streams which now intersect them. It appears, therefore, that at some time long subsequent to the upheaval of the Pareora sea-bed, a warm climate having prevailed during the interval, a profound change of climate set in, and King Frost began a long and vigorous reign, though perhaps not continuous, whose history is writ large in these red gravels.

The question now to be asked is: What degree of cold was concerned in the production of these gravels? Are they the

– 316 –

product and record of a glacial period in the Southern Hemisphere?

The gravels themselves, so far as I am acquainted with them, contain no evidence of even the proximity of glaciers. It seems, however, difficult to believe that an area in which was produced such a massive record of frost-work could have been far removed from a region where glaciers and ice-fields existed at as low a level. In default of direct evidence on the point, we could hardly doubt that the frosts which effected such great denudation were much more severe than those experienced in the same localities to-day. We may safely assume that the eastern ranges, having been exposed to the action of the sea as promontories or islands for a prolonged period under a warm climate, had become extremely rugged and precipitous, and that, consequently, when the frosts did begin to work upon them, they would do so under great advantages.

Still, such frosts as we now have, though working under such favourable conditions, could hardly produce chips for shingle-making in such quantities as to overload the streams and compel them to lay out fans. There is, however, some direct evidence in the gravels that, at a certain stage at all events, the frosts were much more severe than they now are. In two or three of the beds of shingle, separated by clays, in the Geraldine downs, there are numerous lumps of white silty clay, so large, so irregular in shape, and so distributed among the shingle, that their sizes, forms, and pell-mell positions can only be explained by the supposition that they are pieces of the bed of a lake or pool, picked up and transported by being frozen to and buoyed up by thick ice, and were deposited where they are by stranding or the loss of their buoys. There are also numerous smaller rolled pieces, some of silt so incoherent—some even of coarse sand—that they could not have withstood rolling along a shingle river-bed unless they were cemented by ice. The larger pieces—some of them 6ft. long by 2ft. thick—could not have been floated, ice-buoyed, through the rough channels of the stream within the mountains; they must have been picked up below the gorges. In them, therefore, we have evidence of winters in the lowlands so severe that they may well have been related to the existence of great ice-fields and glaciers in the alpine regions and at lower levels in the south of the island.

As moraines and boulder-clays are perhaps the most trustworthy and most easily-recognized marks of glacier work, it is desirable that search should be made for such deposits of the red-gravel age, as indices of the extent of the glaciation during that age. Such indices might be hopefully looked for in Otago; and it appears to me that one has been found there.

– 317 –

In the Progress Report of the Director of the Geological Survey for 1887, p. xxxix., there are described as existing on the eastern side of the Taieri Plain “coarse, angular breccia beds, apparently of glacier origin, the age of which has never yet been clearly determined.” These beds cover a good deal of ground, and range up to 800ft. in thickness. “At the back of Henley,” says the report, “the deposit is composed of exceedingly angular material, often containing blocks of great size, 5ft. to 10ft. and 12ft. in diameter, and is but loosely compacted together, the irregular spaces between the confusedly-heaped angular blocks being often filled with finer material.” This is a description of a glacier moraine, and of nothing else. These beds, it is stated, have been disturbed and invaded by volcanic rock; so also have the red gravels here. In Haast's “Report on the Formation of the Canter-bury Plains” (1864)—Dr. Haast having in his mind but one glaciation, and clear distinctions could not be expected—there are one or two glacial deposits mentioned in such wise as to suggest that they probably belong to the red-gravel age, and if so are indices of the severity of the climate of that age. Writing of the earliest work and greatest extension of the glaciers of Canterbury, he says (p. 5), “The glaciers began to fill all the existing valleys to the plains, and even advanced beyond, spreading in the plains in a fan-like shape. Of this occurrence, however, we have very little proofs, if it be not that the older glacier deposits in the bed of the Rangitata, several miles below the gorge, and some others rising above the plains between the Malvern Hills and the Waimakariri, belong to that period.” As the Rangitata deposits are in “the bed of the river,” they probably, as we shall see further on, are of red-gravel age. The other deposits alluded to are no doubt Little Racecourse Hill and Trigpole Hill, described on page 46 as morainic; and on page 47 Haast clearly refers them to some period antecedent to his Pleistocene glaciation.

If these morainic deposits do belong to the red-gravel age, they seem to show that that period was more severe than the later cold age. This is not in agreement with my observations in the South Canterbury lowlands so far, for I have found no signs of local glaciation within the red-gravel formation, while there are proofs that the same area was glaciated during the second cold age. It is, however, quite possible that the contradiction is only apparent. A less severe climate prevailing for a long time may produce longer glaciers, where glaciers are produced at all, than one which, whilst of sufficient severity to glaciate lower lands, persists but a short time.

I have not had time or opportunity to look through other

– 318 –

reports for mention of morainic beds that may be of red-gravel age, but I have somewhere met with mention of “large angular blocks” in the drift of North Canterbury. The glacier relics of one age are apt to be destroyed or obscured by glaciers of a later and severer age—and there has been a severer age; but as there appears to have been a considerable amount of seismic disturbance in the interval between the two cold periods, moraines or other glacier beds of the earlier or red-gravel period may here and there have been lifted above, or otherwise shifted out of the way of destruction

The total duration of the first cold age must have been very great, to permit of the accumulation of so massive a record of the work of frost by streams having only small drainage-basins, and the duration must have been all the greater, as the accumulation was certainly not continuous. Some of the clays were evidently for long periods persistent land-sur-faces; in the Fairlie Creek downs there are two lignite-beds in the formation, and, I am informed, in the Pareora downs remnants of a forest or bush in petrified wood. I have not yet given much attention in the field to the evidences of variations of climate within the period, but, from what I have seen, I think it likely that a good deal of evidence may be obtained on this point.

If we compare the red gravels of South Canterbury with the later fans in the same localities, it is clear that the older formation, as to bulk, is much the more important of the two. This being so, we are compelled to ask, What has become of the related larger masses which must have been laid down by the larger rivers north of the Orari? The reply must be that they are buried under the later fans which form the surface of the plains; and I find in Haast's report, above quoted, some facts recorded which tend to prove, not only that the Canterbury Plains belong to two periods, but that the greater portion of the whole was laid down during the earlier period. In describing sections presented by vertical faces of the north bank of the Rangitata, the report says (page 39): “The shingle of the Pleistocene fans shows an equal change from the natural bluish tint to one of a rusty dark-yellowish colour, which coats the whole deposit. I have found in several rivers that the change from one colour to the other is often rapid, the dark-yellow shingle underneath being divided from the bluish at the summit by a very distinct line, sometimes with a small bed of loam between them, over which, at numerous spots, small watercourses are dripping down.” Under the head of “Ash-burton,” page 40, Haast writes of terraces on the north bank, in the broad valley of the river, and says: “In some slips in these terraces a distinct line of demarcation is visible between the shingle and gravel of different colours, blue shingle in its

– 319 –

natural colour lying above, with ferruginous beds of the same material beneath.” On page 63 is an Appendix, a special report on the coast cliffs at the mouth of the Ashburton, which accords with my recollection of them, and shows clearly, to my mind, that these cliffs are wholly built up of the red gravels of the first cold age, the blue or grey gravels of the second cold age not having reached the present coast-line there. I do not know the coast-country between the Ashburton and Rakaia, but I venture to assert that investigation will show that the western margins of the swamps between the Ashburton and Rangitata, and between the Rakaia and Waimakariri, are, approximately, the eastern margins in those localities of the shingle-fans laid down during the last cold age.

The “distinct line of demarcation,” visible between the upper and lower shingle in the Ashburton terraces, the “very distinct line” dividing the bluish from the yellow shingle in several rivers, the small bed of loam between them over which water drips (an old soil), these are marks we should expect to find at the junction of the surface of plains of the red-gravel age with later deposits upon them. Had the pioneer geologist of Canterbury recognized the meaning and importance of these marks, and of the difference of colour and condition of the gravels above and beneath them, he would have supplied quite another context to his italicized sentence (on page 16): “From that moment the formation of the Canterbury Plains began.” Indeed, he must barely have missed seeing the necessity for doing so, for (on page 46) he writes, under the head of “River Waimakariri,” “On that part of the plain between the junction of the Kowai and the Gorge Hill, there is.… evidence.… that before the Pleistocene fan was formed older formations of a similar nature existed here.” The Canterbury Plains, I conclude, not only belong to two periods, but the grey gravels of the later fans are only superficial, which brings them into proportion with the small and shallow fans of the smaller rivers of South Canterbury.

To explain the existence of the red gravels, to account for the severity of climate necessary for their production, we must suppose either that New Zealand was elevated some thousands of feet or that the cold was due to a general refrigeration of the hemisphere—a glacial epoch. I do not intend to do more on this head than to make one remark and offer one suggestion. In view of the proofs of glacier action since discovered in the mountains of southern Australia and southern Africa, together with the previously known proofs of mighty frostwork in Patagonia, the elevation theory could scarcely be so readily adopted or so vigorously supported now as it was fifteen years ago; and seeing that we have in New Zealand the records of two distinct periods of refrigeration, the more

– 320 –

reason is there to look for the cause of either in conditions of a general rather than of a local nature. The suggestion I have to offer is that there is a class of evidence which might be usefully sought for, as calculated to finally settle the question of elevation, if found. We can have no direct proofs of great elevation, such as would be sufficient to produce severe climates on our present lowlands. But there might be found proofs that the country was not thus elevated while a severe climate prevailed. The recognition of a raised beach contemporary with the red gravels, or an interstratification of these gravels with marine beds, would give a datum line of first importance. Is any instance of either known? There is none in South Canterbury, for the country here is lower now than it has ever been since the red gravels began to be laid.

The red gravels being completed, there ensued a long period of time after which we find fan-building resumed under the influence of a second age of cold. The climatic history of the interval is not written here in any new deposits, but, and only vaguely, in alteration of the gravels previously laid down. The record tells clearly of a long lapse of time; but whether the climate was on the whole, or ever, as warm as or warmer, say, than our present climate, I know of no evidence to show. Possibly some information on this head could be obtained from the North Island, where, if I have not misunderstood some geological reports, portions of the Wanganui and Hawke's Bay country were submerged after the first cold period, which is registered there, as here, in red gravels. The fossils of the beds of submergence should give some intelligible evidence as to the climate during their deposition. Certainly the climate became warmer, for the process of fan-building ceased.

In this district—and was it not so in many other districts also?—the long warm interval was made more noteworthy by important seismic operations. The shingle-fans here were disturbed in various degrees near the ranges, and subsequently or simultaneously, probably the latter, and as a result of the same subterranean movements, volcanic action was set up at two points, on what are now known as the Timaru and Geraldine downs. At Mount Horrible, in the former case, an explosive volcano burst forth, which discharged dust and grit in such quantity as to bury the country at the present coastline, seven or eight miles from Mount Horrible, to a depth of about 10in. Two miles inland, at the Harbour Quarry, the tuff is 4in. or 5in. deeper. At the south-west corner of Mount Horrible it is 2ft. deep, with no grit so coarse as at the coast. The explosive eruption was followed by an emission of lava, which consolidated into the well-known Timaru dolerite or bluestone. This rock is 100ft. to 150ft. thick at Mount Horrible, where it has been enormously cut away

– 321 –

and some of the streams near the coast are 40ft. to 50ft. thick, whilst others quite thin out there. The volcanic display at Geraldine was on a much smaller scale, and was limited to the emission of lava. At all events, I could detect no tuff beneath the rock where, on opposite sides of the field, the base is exposed. The two lava-fields are about fifteen miles apart at their nearest points.

These two lava sheets have played an important part for the geologist by preserving the land-surfaces of the time of the eruption, while subsequent denudation has provided numerous sections in which those surfaces can be studied. The complete removal of a considerable quantity of rock at the Timaru Harbour Quarry gave an additional and excellent opportunity for further observations on the tuff-covered soil.

With an exception to be mentioned presently, the Timaru dolerite seems everywhere to rest upon the red-gravel formation. At the coast and the Harbour Quarry the rock and its underlying tuff rest upon a dark vegetable soil, containing plant-remains in carbonized matter, and root marks; also worm-borings and bones of large birds. The tuff in places is pierced with blackened tubes, the vestiges of culms of grasses, and black prints of leaves have been seen in it. The subsoil is a gravelly clay, 1ft. to 2ft. deep, derived from decomposition of shingle, and not to be mistaken for a deposited silt, as pebbles are scattered at random through it to the very surface of the soil, these being doubtless the more refractory pebbles of the original shingle-bed. I have nowhere found a soil and subsoil of this description within the red gravels. The marks of persistent land-surfaces, previously mentioned, are in every case in surfaces of beds of clay or silt, deposited as such. (The bed of loam over which water drips, mentioned in Haast's report as occurring in some of the terrace sections, may be a soil of decomposition, a buried soil contemporary with that entombed by the Timaru dolerite.) The soil disclosed by the quarry-work at Timaru contains unmistakable worm-borings; but the fact that pebbles are found in the soil to the very surface seems to indicate that the worms from first to last had not been very numerous, as their operations tend to bury such things. In one spot the Geraldine lava rests upon bare red gravel, as though it here flowed over a watercourse swept clear of recent shingle, and the lower side of the rock is peculiarly splintered, as though it had flowed into water and been suddenly cooled. Elsewhere this lava sheet is seen resting upon a dark soil derived from the red gravels beneath. At one section the underside of the rock shows a semi-cylindrical cavity, presenting the appearance it might do if the lava had flowed over a prostrate log. I have nowhere found any signs of timber beneath the Timaru dolerite, but I obtained, from

– 322 –

amidst the lava, a curious cast of a piece of charcoal from a stick about 2 ½in. in diameter.

The degree of decomposition of the gravels into soil and subsoil, before the lavas overspread them, is proof of long exposure to atmospheric and organic influences. I doubt whether any portion of the grey gravels of the plains has been decomposed into soil and subsoil to a greater depth or more thoroughly. Supposing the degree of decomposition to be equal, and to have been effected under similar conditions, then, whatever age may be assigned to the present plains now, an equal age must be assigned to the red-gravel surface at the time of the volcanic eruption. And this age is not the full measure of the lapse of time between the two cold periods. The dolerite is now covered with loess, which I believe to be one of the earliest products of the second cold age, and the lava was certainly exposed to atmospheric denudation for a long time before any loess was deposited upon it. If we may judge by the roughness of surface of the lower of two streams worked in one of the quarries, the surface of the fresh lava-streams was covered with minor rugosities. The dolerite appears to weather with extreme slowness, yet it was exposed after eruption, and before any loess fell upon it, long enough to permit the whole surface to be smoothed down, and the edges of the rude prisms into which the rock split on cooling to be rounded off, and the cracks in most cases to be widened into little gutters. There has been practically no alteration of the dolerite since. The loess goes hard down upon the rock, or a fine brown mud fills all crevices and forms a level-topped layer over it. It is only in small hollows or pits in the rock, which would early be filled with rain-wash, that the rugged surface of the fresh flow has been at all preserved.

A proof of the great absolute age of the red gravels is the degree to which they are oxidized from top to bottom. They are somewhat less altered beneath the dolerite, I think, yet there they are far more altered than the gravels of the later fans. That the dolerite did afford some protection from further alteration appears to be well shown by one interesting fact, which is also an index of the great age of the formation at the time of the eruption. The Timaru dolerite appears everywhere to rest on red gravels, except at the south-west corner of Mount Horrible. Near this corner one small exposure shows 12ft. or so of loess-like clay, with soil-surface covered by tuff. The clay rests upon rearranged sands, and these on fossiliferous sands, as in the Tengawai terraces. Exactly at the corner, tuff lies upon a few feet of silt and shingle, so loose, so fresh-looking, so unaltered, that they could be matched by river deposits only a few years old. This shingle must have been a river-bed at the time of the volcanic

– 323 –

eruption, or rather at the time of the preliminary upheaval of the gravels, and its appearance has scarcely altered from that day to this. Curiously, this shingle and silt were not burned brick-red by the heat of the rock, as is the shingle elsewhere, to a depth of many feet. I can only suggest that the porosity of the then recent river-gravels allowed the moisture they contained to be easily driven off as the heat of the lava made its way down, so that when the burning heat reached the gravel, this being dry, it did not oxidize the iron in the stones. This river-bed is based upon yellow fossiliferous sands.

From the oxidation of the whole gravel formation, from the decomposition of its surface into soil and subsoil, before the lava overflowed it, and from the weathering of the surface of the lava before the loess of the second glacial period was spread upon it, we must conclude that the “interglacial period” of this hemisphere was of very great duration, measured even by geological standards of time.

So far nothing has been said of the enormous amount of denudation the red gravels have undergone since their disturbance. Previous to disturbance, of course, they would be liable only to terracing by the principal streams flowing over the fans. After disturbance, irregular elevation gave rise to new drainage-lines, and steeper falls gave power to smaller streams to perform excavating work. The red-gravel downs have been enormously cut up, but it is more than probable that the bulk of the work was done during the second glacial period, when these low grounds were undoubtedly glaciated.

If the red gravels are admitted to be the record, in what Professor Geikie calls a “peripheral area,” of a glacial period in the Southern Hemisphere, the climatic experience of New Zealand corresponds with that of North America, where the proofs have been recognised of two great glacial periods, separated by a long interval of time, each of the glacial periods being broken by shorter intervals of ameliorated climate; whilst our experience does not correspond with that of Europe, where some observers recognise three distinct and separate glaciations in the Alps (vide Professor Geikie's presidential address, Brit. Assoc. meeting, 1889, on “Glacial Geology”). The discrepancy between the European and American record can hardly be real, and may arise from a greater value being assigned than should be, by the European observers referred to, to one of the breaks in the earlier or later glacial epochs. Similar breaks are registered in our red gravels, and also in the products of the later cold age, as they are in the North American glacial beds.

I trust that these statements and suggestions respecting the red-gravel, or drift, formation, and its nearer chronographical relatives, will be found interesting and useful. It

– 324 –

may be said that I have given more opinions than facts; the opinions, however, are based upon facts, and I desired to make my paper suggestive rather than descriptive, to provoke further inquiry into the record of a severe, if not truly glacial, climatic period, which will well repay further examination.

Appendix.
[Read 6th November, 1890.]

In the “Handbook of New Zealand Mines,” 1887, page 69, a short account is given, from the pen of Mr. Gordon, of large deposits of auriferous quartz-drift, with leaf-beds and silicified trees, in the St. Bathan's district, Central Otago. The same deposit is referred to as richly auriferous in the “Report on Goldfields,” &c., presented to Parliament this year, pp. 76, 79, and 82. From these descriptions I think this drift must be related, chronologically and in the manner of its production, to the quartz sands and grits on the Tengawai and Hare, described in this paper as produced during the prevalence of a frostless climate in the period succeeding the upheaval of the Pareora marine-beds into dry land. I do not know whether the interior of Otago was submerged by the Pareora sea; if not, then the longer time was given for the production of the quartz-drift. The drift has been much disturbed. A question is, When did the disturbance take place? Was it contemporaneous with the disturbances and volcanic eruptions in South Canterbury which occurred during the interglacial period?