Go to National Library of New Zealand Te Puna Mātauranga o Aotearoa
Volume 43, 1910
This text is also available in PDF
(1 MB) Opens in new window
– 408 –

Art. XXXIX.—The Post-glacial Climate of Canterbury.

[Read before the Philosophical Institute of Canterbury, 5th October, 1910.]

The question of post-glacial changes in climate has attracted so much attention in Europe, and especially in Scandinavia, that it may not be out of place to consider certain indications which point to similar changes occurring in this part of New Zealand. It must be admitted at the outset that the evidence at present available is not strong, and that it is suggestive rather than conclusive; but the author hopes that this paper may serve to attract attention to the importance of the question, and to the desirability of doing our utmost here to make more complete observations in order to see if the same sequence of events followed the recession of our glaciers here as occurred in the British Isles and on the mainland of Europe. In the absence of undoubted evidence, the conclusions herein advanced are tentative in nature, and may have to be modified subsequently, but in disposing of them—if, indeed, that does happen—our knowledge of an important question will certainly be increased. When it is considered that in spite of the large numbers of skilled workers who are studying the question in Europe the differences on major points are very marked, and the conclusions arrived at on smaller ones are often diametrically opposed, the great difficulty of coming to any satisfactory conclusion on the matter in this country in the present state of our knowledge will be readily recognized. The generally accepted opinion as to late Quaternary climates in Europe is that after the recession of the ice, or partly contemporaneously with it, northern Europe enjoyed a climate much milder than at present, and that climate increased subsequently in severity. A school of physiographers, following the lead of James Geikie in Scotland and A. Blytt in Sweden, have urged the occurrence of a succession of maxima and minima of cold, an idea which has been supported by the work of Lewis in Scotland, and specially by that of the Geer-Senander school in Scandinavia, but their conclusions have received strong opposition from Gunnar Andersen and his followers. Seeing this great diversity of opinion among the leaders of thought on the matter, I may perhaps be excused if my own conclusions meet with criticism when the evidence is so much more scanty. I hope, however, that some of the peat-bogs of Southland, Central Otago, and Canterbury may be examined according to the Swedish methods, and an attempt be made to correlate, if possible, the events which followed the recession of the glaciers in this country with those that occurred in Europe. If this can be done our knowledge of the sequence of the general climate of the globe may be much increased.

The question of change of climate in Canterbury was first brought strongly under my notice on the occasion of a recent visit with Dr. L. Cockayne to the head-waters of the Rakaia River, and I am indebted to him throughout this paper for advice and kindly criticism which have been invaluable to me. I owe to him and to others who have helped me with information my sincere thanks for their assistance.

The idea that the climate of New Zealand has undergone marked changes since the retreat of the glaciers is by no means a new one. Captain Hutton

– 409 –

came to this conclusion as a result of his observations on the extinction, of the moa (Trans. N.Z. Inst., vol. 24, 1892). He says, on page 154, “It is also evident that the dead moas could not have been washed into swamps under the present climatic conditions, and the solution of the problem is to be found in the fact that in Pleistocene times, when. these deposits of bones were formed, the climate was very different from what it is now…. As the Pleistocene period passed away the climate no doubt got more equable, and the surviving moas once more increased and multiplied.” His conclusion is based very largely on observations made in Central Otago, and on a study of the conditions under which the bones were found in the swamp at Glenmark, in North Canterbury. A peculiarity in the distribution at the present time of birds allied to the moa may be noted here. Related genera, such as the ostrich, rhea, and emu, now inhabit countries with a dry climate, and it may perhaps be the case that the moas established themselves in the South Island of New Zealand when the conditions were steppe-like in character and the moist climate was responsible for the diminution either on account of their unsuitability for such an environment or because it affected in some way their food-supply.

The moist climate suggested by Captain Hutton was certainly postglacial, though it must be remembered when referring to his writings on the subject that he regarded the older Pliocene as the period of the maximum glacier-extension.

The latest pronouncement on the subject appears to be that of Professor von Lendenfeld, given in the volume published by the International Congress of Geology, 1910. In his article entitled “Das Quartäre Klima von Australien und Neu Seeland” he says, “Die natur der Gletscherzungen (Reisenmoränen, Seitentaler zwischen Gletscher und Talwand) macht durchaus nicht den Eindruck als ob dieser Rückgang der vergletscherung schon zum Stillstand gekommen wäre; sie deuten vielmehr darauf hin, dass er gegenwartig noch fortschrietet, so das ich meinen mochte, dass in der Sudinsel von Neu Seeland, ebenso wie in Sudaustralien, das Klima gegenwärtig wärmer und trockener wird.” The evidence on which this conclusion is based seems somewhat uncertain, but it is in all probability a general statement which would be quite true under any circumstances.

The purely geological evidence of the change in climate since glacial times rests firstly on observations of the behaviour of glaciers, as indicated by Von Lendenfeld; but this evidence is quite inconclusive, as they have been observed for a period too short to furnish data on which to base any well-founded conclusion. There is evidence of a fairly rapid retreat of the glaciers on the eastern side of the Alps; but the cause of this is unknown, although it is probably dependent on climatic changes. On the West Coast the Franz Josef Glacier is now rapidly advancing,* while the Fox Glacier, a near neighbour with very similar surroundings, was, when visited by the author three years ago, showing unmistakable signs of retreat. The Mueller Glacier, too, on the eastern side of the range, is showing signs of advance (vide Lands Report, 1906). It is possible that these advances depend on climatic conditions of a previous time which make themselves evident at the terminal face of glaciers of different lengths and velocities at different

[Footnote] * A recent map of this glacier, made under the direction of Dr. J. Mackintosh Bell, shows that the advance occurs principally on the northern side of the glacier, while the southern side is either stationary or actually retreating.

– 410 –

periods. Only after long-continued observations will it be possible to isolate the various anomalies and refer each to its exact cause. This line of inquiry, therefore, gives at present little satisfactory result except in so far that a recent rapid retreat must be granted, but whether this retreat is periodic or not is quite uncertain.

There is geological evidence of a dry climate obtaining over the region in question either contemporaneously with the extension of the glaciers, or somewhat subsequent to it, in the fact that the loess which so completely mantles the country was a rock-flour carried by wind from the river-beds of glacial rivers. This is the general opinion of its origin, although Captain Hutton maintained that this same deposit was a marine silt. It probably indicates that the conditions over the east coast of this Island resembled somewhat those which occur in Thibet at the present time, only they are not so pronounced. A reference to this probable steppe climate will be made later when mentioning the xerophylly of certain New Zealand plants.

Further conclusions may be based on the examination of the terraces which characterize the valleys of nearly all the rivers of New Zealand. Hutton attributed these almost entirely to a recent rise of the land giving all the rivers increased power of corrasion. In a paper on “The Terrace-development in the Valleys of the Canterbury Rivers” (Trans. N.Z. Inst., vol. 40, 1907) I have given my reason for thinking that, as far as Canterbury is concerned, the major movements have been downwards since glacier times, and that unless the land has been differentially elevated quite recently along an asis almost coincident with the main range, or closely parallel with it, mere elevation cannot account for the characteristic features of the terraces. In that paper I urged the importance of the falling-off of the supply of waste owing to the lowering of the land, thus giving the rivers increased power of corrasion, as the principal cause for their occurrence. I have since seen reason, based on wider observation, to modify this opinion. While admitting the necessity for attaching greater importance to the supply of waste, and recognizing its great influence in the case of the large rivers of Canterbury, there are numerous terraces which cannot be attributed to that cause. Observations made recently on the small fans of detritus in the dumping-grounds of mining claims confirm my opinion that they are built up chiefly when the supply of water and its accompanying load of waste is plentiful, but that, when the supply of water is diminished, terracing of the fan immediately results. The profile of these terraces reproduces exactly those which occur on our large shingle-fans, and also those formed by the rivers which cross the Canterbury Plams. The inference seems, therefore, that the material of our river-terraces was brought down in a pluvial epoch, and terracing commenced actively when the supply of water began to fall off. This course does not affect the principle that supply of waste is also an important factor affecting the formation of terraces. The condition of many of our Canterbury streams at the present time, when deposition is overtaking transportation in the lower portions of their courses, is evidence that a maximum of erosion is past and another cycle of deposition has commenced. This may be due to the lowering of the land, but it may be due to a change in the supply of rain or to alteration in the climatic conditions.

These are the chief lines of evidence of purely geological character which are connected directly with the question, but there are others of biological character which must be considered. The first of these concerns the

– 411 –

character of the land Mollusca, and the others are more or less botanical, and have to do with the peat-bogs and the forests which formerly covered wide tracts on the now treeless or almost treeless regions of Central Otago and Canterbury.

Land Mollusca.

I am indebted to Mr. H. Suter for the following remarks on the land Mollusca of the area. He says that they are certainly of a moist-climate type, that the indigenous fauna is almost entirely confined to the bush, and when this is destroyed or where there is no shelter from rotten logs in moist situations, it disappears entirely. It must, therefore, have established itself in this province when the climate was wetter. This argument standing by itself is not convincing, as the establishment of this snail fauna might date from a time anterior to the glaciation; but taken with other evidence it seems to strengthen the general conclusion that the climate was once moister than it is now.

Evidence from Peat-Bogs.

The evidence afforded by the peat-bogs of this country will, when they have been properly studied, give data from which well-founded conclusions can be drawn. At present their features are comparatively unknown, and the statement still finds currency that they do not contain Sphagnum (J. W. Harshberger: “Bogs, their Nature and Origin”—“The Plant World,” vol. 12, p. 36, 1909), although it really finds an important place among peat-forming plants where the conditions do not allow of good drainage. In some cases where the drainage is bad, and again in other cases where it is good, other plants contribute largely to the formation of the peat. In these cases their ecological conditions are not thoroughly understood at present, but they no doubt depend in some way on climate. Bogs composed of Sphagnum, and also those formed otherwise, occur extensively in both Canterbury and Otago, especially the latter. In the early days of the settlement, peat was regularly cut from the bogs of the central district of Otago and used as firing where wood was scarce. Extensive bogs were found then on the tops of the flat-topped mountains, such as the Rock and Pillar and Rough Ridge; and Dr. Hilgendorf tells me that in the Waipori district twenty-five years ago these bogs were full of totara logs, and other logs lay exposed on the surface of the ground round the heads of gullies in such a way that their distribution could only be explained by supposing them to have been once bog-timber, which had been left stranded as the bogs shrank. These bogs mentioned by Dr. Hilgendorf were in all probability not composed of Sphagnum, but there are others which do contain wood and undoubtedly owe their origin to that moss. Dr. Hilgendorf's statement that the bogs are shrinking in size is a very important one, and if it could be absolutely substantiated it would prove that the climate has undoubtedly changed; but the effect of running stock over bog land tends to consolidate it, and his statement, unless supported by other evidence, would have to be taken with great care, seeing that sheep and cattle have been pastured on these lands. I certainly think that such evidence exists. It must be observed in this connection that the existence of Sphagnum bogs in the dry region of Central Otago, with an average yearly rainfall of about 14 in., and with periods when it has fallen as low as 7 in. per annum, is very striking, since it has been proved that the growth of Sphagnum depends chiefly on the water that it receives from the

– 412 –

atmosphere, and not on ground-water. With reference to this, E. Warming says (“Ecology of Plants, p. 201: Oxford, 1909),” It is erroneous to suppose that Sphagnum sucks up water from the soil; it raises water for an inconsiderable distance. The movement of water in a Sphagnum moor is essentially a descending one.”

The conditions of Central Otago at the present are not favourable to the growth of peat, and those bogs which I have examined do not show any distinct signs of renewal after being dug out for fuel, as they should do if the conditions were favourable for its growth. This seems to bear out the statement that the climate is becoming drier. More careful work will, however, have to be done before this can be definitely established.

Some of the bogs contain abundant remains of the roots and stems of Dacrydium Bidwillii (?). Although no positive evidence could be obtained that this shrub or low tree grew on the bogs, it is nevertheless extremely probable that this was the case, since in many parts of the alpine region of the Southern Alps it is a typical bog-plant, although it will grow also in somewhat dry places. The roots are common in the peat which covers the roches moutonnées of the Upper Waimakariri Valley and other places along the eastern flanks of the range, and it occurs in well-defined layers in the peat-bogs on the line of the Midland Railway near Sloven's Creek, between Broken River and the Cass. These latter may have been swept in by floods when the climate was more rainy than at present, or they may have grown in position on the bog. The presence of well-defined layers of trees and stumps probably points to recurrent periods when the climatic conditions favoured its growth either on the bog or on neighouring land-surfaces. At the present time in some of the bogs a keen struggle for existence is going on between the Sphagnum and the pine. It is likely that in drier conditions the latter would have the advantage, and temporarily extinguish the bog, as has been assumed to be the case with the pines in the old peat-bogs of Scandinavia and Shetland; but this conclusion is open to serious criticism. Whatever the causes controlling the relative growth of these two elements, the presence of the layers of wood and peat points to recurrent conditions, or, rather, to probable alternations of moister and drier climate, as a very slight change in one direction or the other may be a determining factor in the struggle for existence between the Dacrydium and the Sphagnum and other peat-forming plants.

The Presence of Former Forests.

Before the arrival of Europeans, and partially contemporaneous with the early settlement, extensive forests containing trees which flourish in moist situations extended over wide areas to the east of the Southern Alps and over Central Otago which are now almost if not entirely treeless. The existence of this forest is undoubted, and its disappearance is usually put down to fires before the arrival of Europeans or in the very early days of the settlement, an explanation which is extremely questionable and not supported by undoubted evidence. The Rev. J. W. Stack, an authority on the Maori history of this part of New Zealand, includes the tradition of the destruction of these forests by fire among those on which little reliance can be placed, though he remarks that there is no impossibility that they were so destroyed.

The evidence for the existence of this forest is largely based on the observations of the early pioneers and other observers who followed

– 413 –

immediately on them. These men noticed large logs of totara (Podocarpus totara) among the grass and in slips and swamps over very wide areas in Canterbury. The logs were frequently charred, and it was immediately assumed that the former forests were entirely destroyed by fire. The whole question is greatly complicated by the destruction of the evidence by the fires lit by settlers in the early days to clear the country from the rank growth of tussock-grass, so that now there is little proof of absolute value still remaining except in the records and recollection of these settlers. In those cases where I have had to rely on the evidence of such observers I have given their names, and I take the opportunity to thank them for much information of value on the point.

Large totara logs lay plentifully on the foothills along the eastern base of the mountain region of Canterbury. They were found in quantity on the hills around Cheviot at a height of about 1,700ft. (C. J. Westland); on all the country between the Hurunui and the Waipara Rivers, and on the Moeraki Downs, above a height of 800 ft. (P. J. Overton); on Teviotdale, near the coast-line; at Amberley (James Hay); on the Malvern Hills, as mentioned by Lady Barker in “Station Life in New Zealand.” On the arrival of Europeans the present Oxford Forest extended down the Eyre River till it nearly junctioned with the bush then existing near Rangiora; this is proved by the logs formerly lying on that part of the plains. Extensive tracts were covered with, bush on the downs behind Timaru (J. Hardcastle), for large logs of totara are even at the present time found in the small creeks of that now treeless country.

The same is true of other parts of the province near the mountain axis of the Island, as, for example, the Mackenzie country, and the valley of the Cameron River in the basin of the Upper Rakaia. On Banks Peninsula, too, there is evidence for a former greater extent of forest. The top and exposed northern slope of Mount Herbert, on the southern side of Lyttelton Harbour, were bare of trees on the arrival of the earliest settlers, although the gullies held patches of bush; and the country was probably in the same condition when Cook saw it from the ocean half a century before. However, numerous logs of totara are even now found on the very summit of the mountain.

According to my own observations on Banks Peninsula, the points of many of the spurs dividing the bays were absolutely treeless and covered with tussock-grass for a considerable distance from their terminations. This was especially the case on those spurs running north-east near Little Akaloa and Pigeon Bay. It might be urged that salt-laden winds from the sea had killed off the trees in these exposed positions were it not that certain spurs equally exposed were wooded right up to the very edge of the cliffs. According to observers of absolute reliability, in breaking up the open land on such a spur for sowing down in English grasses, buried totara logs of fair size were frequently found far from the edge of virgin bush.

I cannot get any satisfactory evidence that logs lay on the surface of the Canterbury Plains south of the Waimakariri, but large trees of totara have been found buried from 6 ft. to 10 ft. deep in the shingle-beds near Christchurch—as, for instance, at the city waterworks, near the foot of the Port Hills. These were undoubtedly of drift-wood, but the trees were in some cases over 2 ft. in diameter, were perfectly sound, and were used for posts in the fences round the buildings. The trees must have been brought down from some forests that grew either on the plains or further away in the mountains, probably by the Waimakariri River in one of its excursions

– 414 –

away from its present bed across the plains, as only a very large river would be competent to move such enormous logs. The former extension of the bush on the plains is proved by the occurrence of vast quantities of fallen timber and numerous stumps in position throughout the belt of country-stretching from just north of Amberley, through Woodend at the north of the present Waimakariri, and through Marshland on to Christchurch. The present writer has examined such a forest with trees in situ when the excavations were being made for extensions at the Christchurch Hospital. This forest belt extended south through Tai Tapu, round Lake Ellesmere, towards the mouth of the Rakaia, and again on towards Timaru, since extensive deposits of swamp-timber are found near Longbeach, between Ashburton and the sea, and again from the mouth of the Orari River to the termination of the downs just north of Timaru.

On several parts of the coast-line the roots of trees are found in position either submerged below the level of lakes or within the limits of the tide-Such occur in Lake Ellesmere (J. Rennie), at the mouth of the Opihi (J. Hardcastle), near Wash dyke, and again near Pareora (W. Wilson). These show distinctly that the coast-line has been sinking recently, a movement in all probability only a phase of the great sinking movement of the land which, with minor temporary rises, set in during Quaternary times.

This coastal forest was largely of swamp origin, and it was composed chiefly of white-pine, or kahihatea (Podocarpus dacrydioides), and manuka (Leptospermum scoparium), the former a tree which makes a pure association principally on swamps, and the latter a very xerophytic type, but found freely on all classes of soils and in many situations. Remains of black-pine (Podocarpus spicatus) and ribbonwood (probably Plagianthus betulinus) are also found. On the arrival of Europeans relics of this-forest existed in a few localities on the plains—for example, at Riccarton, Papanui, Woodend, Rangiora, and at Temuka. These were mostly of white-pine, black-pine, pokaka (Elaeocarpus Hookerianus), and totara as the timber-trees, in that order of importance, but with little of the last-named. The presence of this swamp-forest was largely determined by the presence of moisture in the form of ground-water, but its disappearance from certain areas may be due to the fact that they had become so swampy that even trees like white-pine could not maintain themselves in the presence of so much water, this tree being really an oxylophyte—that is a xerophytic form which has adapted itself to moist conditions. It is possible, therefore, that the disappearance of this forest may be partly due to altered climate, although the waterlogging of the soil may be also put down to the sinking of the land which has taken place in fairly recent times.

Apart from this coastal forest there were at the beginning of settlement considerable areas of standing bush, containing totara, black-pine, and white-pine, at Mount Peel, Geraldine, Waimate, and specially on Banks Peninsula, as well as in a few other localities in hilly places favoured by a good rainfall and a rich soil. At Mount Peel a considerable area still remains. These were in all probability remnants of a regional forest containing totara which covered extensive areas on the eastern slopes of the main range of the South Island. Excepting that which remains at Mount Peel, its most extensive remnant occurs now in the valley of the Upper Rakaia. Here for miles on the northern bank of the river the slopes of the mountains are covered with a thick forest composed largely of this tree. It occurs also in patches on the southern bank. In this locality the wet westerly winds reach well across the main divide, and the mountains to the

– 415 –

south of the river shelter the upper part of the valley from the cold souther-lies, so that the climatic conditions are eminently favourable for the growth of this tree. Its former extension in the vicinity is proved by the logs which occur now on the south-eastern flanks of Mount Arrowsmith, in the valley of the Cameron River (L. Wood and the author), a district which is now marked by the presence of a xerophytic vegetation.

The extension of this forest into Central Otago was noted by the earliest explorers and settlers, for a noteworthy feature of the land-surface in its original condition was the occurrence of enormous numbers of logs of totara lying on the hillsides and flat-topped mountains, as well as buried in the slips and bogs of that now treeless and steppe-like region.

In his sketch of the “Botany of Otago” (Trans. N.Z. Institute, vol. 1, 1869) Buchanan says, “The general facies of the vegetation of the province on its eastern watershed is grassy, the greater part being open grass land with comparatively small areas of bush along the coast-line and in the gullies of the mountain-ranges, whereas on the western watershed the whole country from the sea to the altitude of 3,000 ft. on the mountains is covered with bush. It is evident that at no distant time the greater part of the province was covered with forest. On many of the grassy ridges may still be seen-the remains of large trees, and over large areas the surface is dotted with little hillocks and corresponding hollows produced from the upturned roots of trees which have been blown over, generally in the line of the prevailing winds, after their destruction by fire, and no doubt there have been many denudations and reproductions of bush. At the beginning of the settlement large tracts of the province were being reclothed with bush, but as the country was opened for cattle and sheep runs this new growth was again burnt off, and a luxuriant growth of native grasses appeared without seeds being sown.” This account was written in the year 1865, and is specially important as showing that a competent observer at that early date was of the opinion that a succession of forests had covered the treeless hills of Otago.

The existence of vast numbers of totara logs in Central Otago is testified to by many other observers, all of whom have acknowledged that the prostrate trees indicate the presence of a forest of wide extent. The latest reference to this appears to be that by Professor Park, in Bulletin No. 5 (n.s.), New Zealand Geological Survey. The author there says, “Forest vegetation is entirely absent, but there is evidence that it was not always so. Above the 2,000 ft. contour-line of Mount Malcolm and Mount Hocken there are still many logs of totara (Podocarpus totara), charred and well preserved, lying on the surface of the ground. The older settlers state that totara logs were at one time common on the Dunstan, Pisa, Carrick, and Remarkable Mountains, and proved of great value to the early pioneers for fuel and fencing purposes. The totara forests apparently flourished above the flood-level of the Pleistocene rivers that filled the old lake-basins. They were probably destroyed by fire.”

The occurrence of the totara in the peat-bogs of Otago is referred to earlier in this paper.

Although the evidence for the wide extent of this forest is conclusive, it cannot be maintained that it covered the face of the country as completely as do the forests on the western slopes of the Alps. All the same, it may have done so. The usual sequence of events in the changes which forest experiences in this region is that when destroyed by fire, wind, or any natural cause it turns into tussock steppe. The transition is at times very rapid,

– 416 –

so that it is very probable that a large part of the tussock country of this Island was once bush-covered. However, some areas must always have been in existence on which there was no bush, where the tussock was established, and from which it could spread to neighbouring tracts when the ecological conditions were favourable for it doing so.

Continual reference has been made previously to the occurrence of totara logs, and the inference will probably be made that the forests were largely, if not wholly, of totara. In all probability such was not the case. At the present time few areas are covered completely by this tree. It occurs widely distributed through many kinds of bush, but it also occurs at times in groves several acres in extent. It was in all probability only a constituent of the former forest, and its predominance among the timber lying on the ground is due to its uncommon power of resisting decay. It is one of the most long-lived of timbers when in damp situations. Other specially resistant timbers, such as broadleaf (Griselinia littoralis), are also found, so that in all probability the woods which grew with it have long since rotted away. It must be noted in this connection that I am not referring here to bog-grown timber, such as the manuka, which is so common in so many swamps.

Since totara was undoubtedly a prominent constituent of this forest, the conditions under which it grows have a very important bearing on the question of the climate which obtained in the area when the forest established itself. Although it is not a rain-forest tree in the same sense as the rimu (Dacrydium cupressoides), and although it is perhaps the most xerophytic of all the New Zealand conifers (vide paper by Miss Griffin, “The Development of some New Zealand Conifer Leaves,” Trans. N.Z. Inst., vol. 40, 1908), it is nevertheless a tree which flourishes under conditions of good drainage with a damp atmosphere, and specially on deep rich soils—it is, in fact, a prominent member of the New Zealand rain forest. It will grow on light pumice and sandy soils if they have plenty of rain, as well as on heavy clays, and even at times on swamps; but the existent patches of bush containing this tree in quantity, and also those which furnished a large part of the totara for timber purposes in former times, were in localities with a good, if not a heavy, rainfall. The “totara areas” of the North Island, according to the report made in 1875 by Major Forrestier Walker, were those tracts along the central mountain axis of the Island and near Lake Taupo which had a rainfall of 40 in. a year and upward. To take, for example, the totara forest on the northern slopes of Tongariro and the Waimarino Forest, which contains in parts a large quantity of that timber: they are both situated in a part of the central plateau which receives a heavy rainfall from the west, while the dry eastern slopes of the Ruapehu-Tongariro ridge are either treeless or dotted with patches of Nothofagus cliffortioides, and the charred fragments of wood in the pumicedrifts which cover that region seem to be from that tree and not from totara.

The distribution of totara is the South Island, occurring as it does in such localities as Banks Peninsula, Peel Forest, at Glenomaru in the Catlin's district, as well as near the east coast about Dunedin, shows that this tree delights in a well-drained soil, with plentiful supply of rain. It also occurs on the excessively moist hills and swamps of Westland, but in the latter it does not thrive. The forest now existing at the head of the Rakaia is merely an extension of the subalpine totara forest of Westland which has

– 417 –

followed the rain across the main divide, and the fact that this tree out of all the rain-forest trees of Westland has done so is really very striking.

Judging from the rainfall records kept at the Bealey, where the usual amount for the year averages about 100 in., and has reached as much as 136 in., the rainfall of the Upper Rakaia Valley, which is exactly similarly placed as regards the main range and the direction of the rain-bearing winds, must certainly approach, if if does not exceed, 100 in. per annum. The climate is decidedly moist and the drainage is good, and under these favourable conditions the totara occurs so plentifully in parts of the district that it virtually excludes other forest-trees.

These facts seem at variance with the xerophytic-adaptation structure exhibited in the leaves of the tree; but this is one of the many cases which have been pointed out by Dr. Cockayne where there is a marked discrepancy between the actual structure and the structure one would expect the plants to show judging from their habitat alone. In his “Notes on the Subalpine Scrub of Mount Fyffe” (Trans. N.Z. Inst., vol. 38, p. 373, 1906) he says, “The amount of xerophylly in many New Zealand plants is by no means a measure of their adaptation to present environment, but is more likely a survival from a former geological period when xerophytic conditions were more widespread.” This opinion is a very important one from a geological as well as from a botanical point of view, and it agrees with that of Dr. L. Diels, quoted in Dr. Cockayne's “Plant Geography of the Waimakariri” (Trans. N.Z. Inst., vol. 32, p. 122, 1900). Dr. Cockayne's statement about the xerophylly of New Zealand plants probably applies to the totara, though it is no doubt possible that the advantage of possessing a moderately xerophytic structure would aid it in its struggle with other plants under increasingly dry conditions, and therefore may explain its importance as a forest-tree in the forests which grew formerly over the somewhat arid regions on the east coast of this Island.

The date of this former forest-extension is fixed as being certainly postglacial, since the remains are commonly found in localities which must have been covered with ice, according to the most conservative opinions as to the extent of our former glaciers; and it must certainly have been posterior to the great glaciation postulated by Professor Park, as the region where these logs now occur in Central Otago was, according to him, covered with a great ice-sheet. It is possible, however, that, as the glaciers retreated from their furthest extension, the forest established itself pari passu on the areas left free by the ice, just as they are now doing in the case of the Franz Josef Glacier. The space of time taken for the gradual extension of such a forest over regions swept bare of soil by the glaciers must have been enormous, and if there has been a succession of forests, as suggested by some observers, then the time must have been very great indeed.

The question of the extent of these forests and the conditions under which the totara grows have been gone into somewhat at length because, in my opinion, it is possible to make the deduction that when the forests established themselves the conditions must have been much different from those obtaining now. It has been just stated that the forests have extended over areas which in Pleistocene times were covered with ice, and were probably dry and steppe-like. Before they could do this the climate must have changed, as the New Zealand rain forest requires a moist climate for its growth. This conclusion seems to agree with those deduced from other observations. I must explain, however, that this conclusion is merely

– 418 –

tentative, and is based on observations of the ecological habits of the totara which are decidedly incomplete.

Reasons for the Disappearance of the Forests.

Although the main point to consider in this paper is the question of the establishment of the forests on an ice-swept country, yet their disappearance has also an important bearing on the question. In several parts of the country it has been noticed that the bush is shrinking even when protected from the interference of man and animals. Far larger areas in the North Island were once covered with kauri forest than those which existed in the memory of man, as is proved by the extensive deposits of subfossil kaurigum far away from growing timber. This is quite apart from the gradual restriction of the kauri forest which has gone on in Tertiary times from its former wide range, proved by the occurrence of fossil kauri-leaves in various Tertiary deposits in Otago. The destruction of the forests containing totara, which once existed in Canterbury and Otago, has usually been put down to the fires lit either accidentally or intentionally by the Maoris, a conclusion largely based on the fact that many charred logs were found by the earliest settlers. There is a Maori tradition, mentioned previously, of the destruction of these forests by extensive fires at one period of Maori history; but the evidence from Maori tradition is almost valueless, and there are indications from Canterbury which certainly point in a contrary direction. Banks Peninsula was thickly peopled by Natives at the time of the arrival of the first settlers and it showed no signs of the ravages by fire (S. C. Farr). If fires had been lighted by the Maoris this is just the place where they should have occurred, yet the hillsides, with the exception of the highest points and some of the headlands, were completely clothed with bush right up to the very settlements of the Natives. The bareness of the headlands seems to have been due to other causes, since from their relative inaccessibility they were not likely to have been swept by fire. These open spaces were probably due to that natural succession of events which turns forest into grass land, but the factor controlling this, change does not seem to be well understood, and it may be a function of the change in climate. At the same time, it must be admitted that continuous and repeated fires in dry seasons, fed with accumulations of dry tussock-grass, would restrict materially the areas covered with bush, and without doubt these fires occurred.

It seems impossible, however, that fire could have destroyed a forest of wide extent and left no patches in sheltered gullies and other places which would have formed centres for its renewal had climatic and other conditions been favourable. If the climate is favourable for the renewal of the rain forest it is difficult to destroy it by burning, even when this is carried on for the express purpose of clearing the land for pastoral purposes. In the case under consideration, the charring of the logs mentioned previously has been caused largely by tussock-fires since the arrival of the white man, although the Maoris certainly carried on a sporadic burning in pre-European times. It is a remarkable fact, however, that the existing patches of bush occur i just those situations in which they might be expected to occur from ecological considerations had a slight desiccation of the climate come about. The bush has disappeared from situations where from exposure to wind, lack of moisture, &c., they would naturally feel first the effects of slightly drier atmosphere. The change required to produce this disappearance is no doubt very slight, as little is required to upset the delicate

– 419 –

balance which exists between the struggling elements of different plant associations. It seems, therefore, that a slight desiccation of the climate can be inferred from the restriction of this forest. The disappearance had in all probability begun before the arrival of the Maori, following on a period of drier climate after a moist period which favoured its growth. The forests would then be specially subject to the action of fire, except in those places favoured by moist conditions. Fire certainly aided in their destruction, but it is improbable that it was wholly responsible for it.

The present conditions are decidedly dry over a large tract of country to the east of the main range; however, this is due not altogether to lack of rain, but to the influence of parching winds. Under the combined influence of these two agents the vegetation is now markedly xerophytic, but when the plants are grown in a moist and still atmosphere they readily revert to leafy forms, as has been proved by the experiments of Dr. Cockayne on the wildirishman (Discaria toumatou) (“The New Phytologist,” vol. 4, p. 79, 1905). Dr. Cockayne has suggested to me that this readiness to revert to the leafy form may be due to the fact that at some period anterior to the present somewhat dry one the prevailing climatic conditions were moist, and that owing to the environment the tendency to become more or less permanently xerophytic was checked, and the ability to respond quickly to a moist atmosphere has been kept latent up to the present time.

I am aware that this evidence is not convincing, and that it is merely sufficient to establish a prima facie case for consideration. If, however, the conclusion that a rainy climate succeeded the last glacier maximum over a wide area to the east and south-east of the Southern Alps be accepted, the reasons for such a change may be briefly considered.

This moist climate may be attributed to one of two main causes—(1) a marked lowering of the land in post-glacial times, with a general climate much the same as that which obtains now; or (2) a change in the climate affecting a wide area, or even the earth as a whole.

Cause of the Change.

A marked lowering of the land in this region of moisture-laden westerly winds would have the effect of making the climate uniformly humid over the whole of the country, instead of being, as now, subject to an abnormally heavy rainfall on the west of the range, while the east is comparatively dry. With lower mountains the total amount of rain intercepted would in all probability be much less than that which would be intercepted if the mountains were higher, but it would be more uniformly distributed. The humidity of the climate depends as much on the number of rainy days as on the number of inches that fall during the course of the year—for example, the climate of the Chatham Islands is markedly humid with a comparatively small annual rainfall. However, before lowering of the land could produce a humid climate in the Mackenzie country or in Central Otago the land would have to be lowered by hundreds, perhaps thousands, of feet below its present level, in order to allow the extension of the moisture-bearing clouds over the ranges near the coast, so that the interior of the country might receive a share of their moisture; and there is no evidence of such a lowering since glacial times. Even if the old beaches found in various parts of the south of this Island were postglacial — which is open to doubt — the amount of sinking that they indicate is totally insufficient to have produced a moist climate over that part of the country which lies along the east of the range. It seems

– 420 –

reasonable, therefore, to attribute the humid climate either to causes which have affected the earth as a whole or to altered meteorological conditions promoted by some change in the distribution of land and water in the Southern Hemisphere.

Conclusion.

The general sequence of events since the glaciation of the South Island in Pleistocene and post-Pleistocene times appears to have been the following :—

(1.)

Glacial conditions, with probable steppe climate existing contemporaneously on the land to the east of the terminations of the glaciers, a condition which probably continued for some time, as the glaciers were retreating.

(2.)

Moist climate over the tract to the east of the main range, during which the forests were established or were widely spread and the rivers built up their fans.

(3.)

Modified steppe conditions over the belt to the east of main range.

The conclusions are, in general, similar to those which have been agreed upon by European geologists as occurring in Europe, although an important school has demanded the existence of a succession of warmer and milder conditions. Of course, it is impossible at present to make any inference as to such a succession of milder and more severe climates in New Zealand, or even to infer that the changes in climate suggested by the various lines of evidence indicated above were contemporaneous. Some of the changes suggested may be due to causes which operated in Tertiary times. However, in view of the general interest in the matter, the author hopes, in spite of obvious deficiencies in the statement of the case, that this paper may serve to draw attention to a problem which has an important bearing on the climate of the world as a whole, and also on the evolution of the vegetation and the plant-associations which exist at present in this country.