The Lake Ellesmere Spit With Map, Sections and Photographs.
[Read before the Philosophical Institute of Canterbury, 2nd October, 1929; received by Editor, 5th February, 1930; issued separately, 29th May, 1930.]
Material of which the Spit is constructed.
Form of the Surface.
Nature of the Beach.
Structure of the Spit.
Origin of the Spit.
Changes in Level of the Land.
References to Literature.
Also appendix by Professor A. Wall, on the Plant Covering of the Spit.
a. Introduction (see Map).
The Lake Ellesmere Spit which divides the lake from the sea presents certain features of interest on account of their bearing on the recent geological history of the country in their vicinity. The spit itself, not taking into account the low-lying land on the southern margin of Lake Ellesmere subject more or less to inundation, stretches from the neighbourhood of Lake Forsyth and the old wave-cut cliffs between Birdling's Flat and the mouth of Price's Valley in a W.S.W. direction to Taumutu, the Maori village near the entrance to Lake Ellesmere. It is some 17 to 18 miles in length, about 2 miles wide at its eastern end near Birdling's Flat, a width which somewhat gradually decreases till at a distance of 10 miles it is about a mile wide, and it narrows to practically a few chains at Taumutu. Its area is approximately 12,000 acres. Its height above sea-level for the greater part of this extent is between 18 and 20 feet, but near a fringe of sandhills on the coast the level is higher, and there is a lowlying area on the lake side, hardly part of the spit proper, which is but a few feet above sea-level and lake-level; in fact when the lake is high portions in its vicinity are flooded. The spit proper is bordered on both sides by strips of land which are higher than the interior portion, and this presents certain features which an account should satisfactorily explain.
Near Taumutu the level of the spit falls and it becomes quite narrow (see Plate 30), so that seas break over it at high tide and render the water of the lake brackish. This is specially the case in
heavy south-east or south-west weather. When the level of the lake rises during the winter it has to be let out by a channel opened through the bank by horse teams and scoops. The rush of water rapidly cuts a path through the loose material, and the lake may remain open to the sea and tidal for a short time, but the first southerly storm blocks it up again, and if the bank is wide and the quantity of shingle thrown up great the level of the lake will rise once more and it will have to be opened again or the low-lying lands round the margin of the lake may be flooded. The water has been known to rise till the Lincoln–Little River railway has been submerged near Kaituna, a rise of approximately three feet. If the bank is narrow the percolation of the water through it, especially at low tide, is sufficient to prevent an undue raising of the water level.
Also there is a similar bank near the eastern end of the spit, ponding back the water of Lake Forsyth. This now forms a permanent barrier, except that in storms the water breaks over the bank and temporarily blocks a passage by the road which now follows along the barrier. However, in the early days of the settlement of the province, it was more commonly open and an old map records the existence of a Maori boat-harbour at the entrance to the lake, and represents it as being permanently open; it is known, too, to have been used by small trading schooners. This indicates that the outlet of this lake has only lately been permanently closed. When the barrier at the outlet is closed by especially heavy seas, which happens at a time when the level of the lake is raised by rains in winter, and by some addition from the sea itself, the low-lying lands at the head of the lake are flooded, and relief works are contemplated which will allow the surplus water to drain away at all times, whatever may be the state of the sea or the width of the barrier.
b. Material of which the Spit is constructed.
The material of the spit is almost entirely of greywacke, a fact which has considerable bearing on its origin, since it can only have come from the south. Banks Peninsula is almost entirely constructed of volcanic rocks—chiefly andesites and basalts—and the exposure of greywacke near the head of Lyttelton Harbour cannot possibly have furnished such material as is necessary to form the spit. There are occasional pieces of limestone and of volcanics, rhyolites an dhypersthene andesites, especially the former, which must have come from the foothills of the Southern Alps between the Malvern Hills and the Rangitata River; the latter cannot have come from Banks Peninsula or from the north, so their occurrence only accentuates the evidence as to the southerly origin of the main material of the spit. One piece of well-rounded basalt, found about a third of a mile from the cliff near Lake Forsyth, resembles somewhat the basalts of Banks Peninsula, and may have come from there. Although several areas of basic rock are drained by the rivers to the south and this piece may have been carried down by them to the coast, the type of rock is somewhat different from any known to occur in connection with these areas.
Included among the greywacke are numerous pebbles of siliceous rock, white, green, and red in colour. These may have come from the siliceous beds which in numerous places are interstratified with the greywacke, or they may have come from the amygdaloids occurring in the andesites near Rakaia Gorge, the Clent Hills, Gawler Downs, and other localities to the south where amygdaloidal andesites form important members of the exposed rocks. Both types, that is, amygdaloidal and jasperoid, are found among the pebbles. Two or three pieces of a hard white sandstone were also picked up resembling concretionary masses occasionally found in the Cretaceous Coal measures.
The possible sources of the material of the spit are two. First of all, there is the gravel brought down by such rivers as the Rangitata, Ashburton and the Rakaia, especially the last, and secondly, there is the material derived from the wear of the coastline. Haast (1879, pp. 399-400) has noticed the erosion of the fringe of the fan formed by the rivers which issued from the Mid-Canterbury glaciers in Pleistocene times, and the products of this erosion find their way north to the vicinity of the Ellesmere Spit. As the ultimate source of this material is the same as of that brought down by the rivers, the relative amount derived from the rivers and from coastal erosion cannot be determined. All the way from the mouth of the Rangitata to the vicinity of the Rakaia the sea-coast is bounded by a cliff formed of gravel. It reaches a maximum height of about 70 feet just north of the mouth of the Ashburton and diminishes in height both north and south from that locality, but the height remains fairly constant for miles in a northerly direction and only falls gradually to nearly sea-level near the mouth of the Rakaia. In the opinion of residents the coast-line is being eroded at present at a rate of between 2 and 3 feet a year. This must have gone on for a long time, since the gullies formed on the surface of the plains in the neighbourhood of Wakanui enter the sea discordantly, indicating that the marine erosion has proceeded so fast that the streams which occupy these gullies cannot lower their beds fast enough to make them accord with the sea-level. It is true that they carry less water than formerly, and probably date their initiation to a time when the climate was wetter, but the amount of the discordance on the sea coast indicates that a strip at least two miles in width must have been removed from the fringe of the fan. This furnishes at the present time a considerable supply of gravel, and in fact it may be primarily responsible for the formation of the spit. The total volume of spoil that the erosion of this fan would furnish is comparable with that contained now in the spit, taking due consideration of its area and the probable thickness of gravel it contains.
The pebbles forming the spit are invariably of the beach-shingle type, flat and well-smoothed. (See Plates 30-31). They range in size from pieces about 9 inches in diameter down to coarse sand grains, but the great majority of the pebbles are between an inch and four inches in diameter. The size of the pebbles appears to be greater along the present sea front and finer in the middle of the spit, and on the side facing the lake, but in places the external beach is made of sand and small pebbles. In the neighbourhood of Lake Forsyth
the quantity of sand appears to be very small, and the material is almost entirely of pebbles, as can be seen not only on the beach but also on the shore of the lake and in the gravel pits that have been opened in various places near the road and railway.
There is no doubt that during the strong winds that accompany southerly storms, large grains of sand and also small stones are easily transported from the coast. There is thus a fair admixture of this material in the surface layer of the spit, so that a poor soil is generally present on which grass and shrubby bushes become established.
c. Form of the Surface.
Although the general surface of the spit may be said to be flat, careful observations show that there are departures from this. In order to determine the cross section accurately, two lines of levels were run across the spit (see Fig. 2) by Mr. W. F. Robinson of the Canterbury College School of Engineering, and I am deeply indebted to him for this and for other help. The first of these lines crosses the spit from the point where the line of railway leaves the lake flats near the Railway Gravel Pit (Plate 27) to the vicinity of the Devil's Knob, the mass of rock terminating the spur immediately west of the Birdling's Flat railway station, and thence onward at right-angles to the coastline—there being thus a bend in the line. Another traverse was made across the spit about 10 miles from the railway.
The first part of Section No. 1 (starting from the shingle pit) shows the presence of a series of undulations, four in number, whose crests are approximately from 16 to 22 feet above sea-level, with hollows between them 7 feet lower. The convexity of these crests is directed towards the E.S.E. Near the road their direction is practically N.-S., but at the railway it has swung round for 15° and on the south side of the line its orientation is N.E.-S.W. and a little further on it is almost E.-W. The crests are reduced in number and become indefinite about 400 yards west of the railway, but a little further on the first continues again as a well-defined ridge for miles and will be referred to later.
The lowest point of this section across the spit is only 9 feet above sea-level and lies immediately west of the most easterly of the four ridges. After this ridge, height 16 feet, the ground is low, 10 to 14 feet above sea-level, and swampy in places for about 15 chains. This low-lying land extends for a long distance towards the W.S.W. down the heart of the spit, and is bounded by one or more ridges on the inner side facing the lake. From about the 16½ mile peg on the railway line the surface gradually rises towards the sea-coast so that where the road down the spit crosses the railway the level is 28.6 feet, and thence on for about 18 chains the level is maintained at about from 31 to 32 feet with minor undulations. Then it rises to 35 feet, a level which continues for about 10 chains, whence it falls to sea-level with two ridges and intervening hollows parallel to the coast. The ridges are at heights of 30 and 24 feet, and the hollows 27.5 and 23.6 feet respectively. The latter of these ridges continues as a well-defined crest right to the outlet of Lake Forsyth, and in its shelter the huts of the settlement there are constructed.
A noteworthy feature is the fact that the top of the spit is at this end of the section constructed of gravel and not of blown material, and since it is unlikely that any sea could build up a spit at a height of 30 to 35 feet above normal sea-level, it implies that the land was lower when the highest point of the spit was constructed, and that it has been raised since then. The two ridges mentioned above no doubt mark successive halts in uplift after the main part of the spit was formed.
The undulatory nature of the surface is quite characteristic, and in addition to the major waves shown on the section there is a succession of minor ones. These occur between the railway line at the Devil's Knob and the sea, and are more clearly seen from the top of the cliff, from which the lines of Wild Irishman (Discaria toumatou), marking successive undulations, are shown most distinctly, the bushes growing in the sheltered hollows between the crests. Between the railway and the sea, about 24 of these crests were thus counted, but it must be remembered that some are indefinite and merge into one another (see Plate 28).
Crossing the line of railway between the Birdling's Flat Station and the point of the hills lying about a mile to the north in the direction of Little River there are fourteen of these crests with a distance between them ranging from 20 yards up to 140 yards, the great majority of the distances being between 35 and 50 yards. It is impossible to give the distances exactly since the summit of the crest is usually indefinite and not determinable within a few yards. The convexity of the curve that these crests exhibit is directed towards the western shore of Lake Forsyth into what must have been an old bay between the site of the present railway station and the hills. From this station the crests get lower towards the hills, in fact the inner ones are from 6 to 10 feet lower than those near the station. The shingle of which they are composed is on the whole smaller in size than that nearer the beach, but there is an admixture of coarser material ranging up to 6 inches in diameter. There are layers of very fine as well as layers of coarser gravel exposed in the pits dug for road metal alongside the road.
A second line of levels was run by Mr. Robinson across the spit at a point about 10 miles from the Devil's Knob. This shows the coastal belt of dunes, and a fairly level surface for the main part of the spit with a marked drop to the level of the flats bordering on the lake. The height of the bank on the inner side of the spit appears to be accordant with that nearer the railway.
An important feature of the inner margin of the spit is a bank forming a line of rampart all along the inner face. This starts in the series of ridges near the railway gravel pit (Plate 27), but these soon peter out and leave a well-defined single ridge extending for about 12 miles down the spit. It is usually about a chain wide, with an outer slope generally of from 5° to 7°, but sometimes as low as 2°, and an inner slope of varying angle, but at times reaching as high as 20°. It rises at times as much as 13 feet above the ground at the back of the spit, but its height is usually from 3 to 6 feet (see Plate 29). Between Mr. Bailey's house, about 11 miles down the spit just beyond Section No. 2, and the late Mr. Willey's house about a mile
further on, its height is in places as much as 18 feet. Occasionally, however, its surface merges into the general level of the spit (see Section No. 2). On the side facing the lake the fall is usually steep for about 6 feet, and then it flattens out and merges into the level and swampy ground facing the lake. In some cases there is a well-defined beach about 7 feet below the level of the crest, and very occasionally the slope is gradual towards the lake. At times, too, the beach merges into the general level of the top of the spit and then a new barrier continues as a switch line in front.
The ridge itself has thus all the characters of a barrier beach, and if that is its origin it must have been formed when the surface of the spit was practically awash in stormy weather, that is, the land must have been some 12 to 15 feet lower than at present. This lower level corresponds with the lower level demanded by the shore platforms cut at the base of the cliffs to be mentioned later.
In some places in front of the ridge there are hollows of considerable size, sometimes over a chain across, like sink holes in appearance whose mode of formation is difficult to explain. The largest of these is five chains long and two chains wide and contains water on the floor. Most of them are dry at present and do not appear to contain water at any period of the year. Spring holes of considerable depth do occur along the flats near the lake and the holes referred to may be due to a similar cause. Some of them are practically at lake- or sea-level, although separated from both by a considerable width of land.
On the inside, i.e., the south side, of the bank the low ground at times extends for long distances, and the depression includes a considerable area of ground. This is notably the case near Mr. Bailey's house, where there are numerous old Maori ovens of large size, indicating prolonged or repeated occupation, the plentiful supplies of fish and birds from the lake, also the quantity of drift wood from the beach, as well as the shelter from the wind afforded by the bank rendering it a favourite camping place. Maori ovens are also found in other parts of the sheltered ground, especially when the bank is somewhat higher than usual.
About 10 miles from the eastern end of the spit the form of the surface becomes more complicated. The bank is broader, up to about a couple of chains, but still well-defined, bare of vegetation as if it had recently been raised. In front are a number of subsidiary ridges marking former levels of the lake in its retreat from the main ridge. These extend for some distance to the west, parallel to the bank. On the inside the well-marked depression behind the bank, referred to previously, tends to disappear and its place is taken by many clearly marked ridges, analogous to those seen near the Birdling's Flat Station, but with a much shorter distance from crest to crest.
At this point the general trend of the barrier beach is about E.-W. but the trend of the crests inside this is first of all N.E.-S.W., but on going towards the west their direction becomes E.N.E.-W.S.W., and further west still they become E.-W., that is parallel to the general trend of the barrier beach in that section. The convexity is thus directed towards the E.S.E. There is a succession of crests, at least
seven in number. They almost block the end of the hollow mentioned as extending from near Section No. 2, past Mr. Bailey's and Mr. Willey's houses, and when followed west they merge into the general surface of the spit.
On the inside of this section, not only where the complex of crests occurs, but also where the inner margin of the spit is more defined, there are projections of gravel extending from the line of the barrier beach, with hollows in between, so that they take the shape of an embayed margin. These are apparently formed by seas breaking across the barrier and carrying material into the lagoon on the inner side, building out blunt tongues into the shallow water, just as may be seen forming now where the sea breaks over the barrier beach into the lagoon north of the Rakaia mouth. These naturally date from the time when the land was lower and the surface of the barrier beach practically awash during the heavy seas that occurred in the open space where the lake now is. The convexity of the crests indicates that the seas before this came from a slightly different direction than when the barrier was formed.
Coastal dunes continue from about two miles south-west of Lake Forsyth right to Taumutu. The belt extends to several chains in width, but narrows towards the western end of the spit. It is well defined and the summits of the dunes present a somewhat accordant summit level, being about 50 feet above sea-level or from 20 to 25 feet above the general level of the spit. The front facing the sea is usually steep, but the inland side merges into the prevailing level of the spit with more stabilised dunes, not so high, lying inside the higher summits. In some cases there is a double row with level ground between. About 8 miles down the spit the dunes appear to be attacked by the sea, while further north the shore appears to be prograding. It has a different form, and drift timber appears half buried in sand a chain or more from the storm-water mark.
Mention should also be made of the old beach which lies across the mouth of Birdling's Valley immediately to the west of the point where the spit ties on to the solid land. Its level is accordant with the top of the spit in its vicinity, and it was evidently constructed when the barrier beach was formed, at a time when the land was lower. It was no doubt due to seas from the west breaking directly into the mouth of the valley, then no doubt a bay facing to the west.
Another interesting occurrence is to be seen at the base of the hills in the embayment north of the road west of the Devil's Knob. Here the following section rests on a base of volcanic rocks:
Loess, passing up into burnt soil, with charcoal fragments, containing angular pieces of burnt rock (basalt), with shell fragments in the upper dark coloured part. Some of these shells are marine and others fresh-water, the latter being of Diplodon menziesii.
Beach gravel, 15 to 18 inches thick, and extending right across the embayment. This is at an estimated height of 12 to 15 feet above the level of the railway across the road. Its height does not equal that of the summit of the spit at its highest point.
There appears to be no reason against the conclusion that the beach gravel lies in position on the top of the burnt material, which is apparently of Maori origin. If this is so, important changes in level have taken place since the occupation of this part of the country by the Maoris. It would mean that the land had sunk so as to allow the old ashes to be covered by marine material, and then raised till it was above the reach of the waves.
d. Nature of the Beach.
In the vicinity of Taumutu the principal constituent of the seabeach is shingle of the true shingle form (Plates 30-31), the size ranging up to about 10 inches in the longer diameter. There is also an admixture of smaller material ranging down to fine sand. As a general rule the part of the beach which is composed chiefly of pebbles of an average size, i.e. between 2 and 3 inches, is that lying between ordinary tide marks. Towards the lower limit of the tidal belt the material becomes finer, while in the direction of the upper limit of the tides the sand increases in amount till it passes into pure dune. In this portion the average size of the occasional pebbles is larger than on any part of the beach (see Plate 30). I take it that these stones have been thrown up by big storm waves and have not experienced the continual attrition which affects pebbles on the belt between tides. It is thus clear that the finer materials of abrasion are removed from the ordinary tidal belt, some are swept downwards and distributed on the sea-bottom by the backwash of the waves, etc., and some are carried away by the winds, the large pebbles referred to above being exceptional.
Further north-east, sand forms the dominant material of the beach, and I have no knowledge of its precise origin. This continues till Lake Forsyth is approached (Plate 31), when the finer constituent becomes relatively unimportant. It may be that by the time this locality is reached in the drift up the coast, all the finer material has either been blown away to form dunes, or been reduced to mud and swept out to sea, the latter in accordance with the views expressed by Marshall as to the effect of attrition on beach pebbles. However, it must be pointed out that the beach between the mouth of the Rakaia and a mile or two south of Taumutu is dominantly of gravel with little fine material, and this is especially true for a few miles north-east of the Rakaia mouth.
Observations of the angle of the beach made in different localities show that it is remarkably uniform, and that the inclination from the foot of the sand-hills down to low-water mark is from 4° to 5°. This is also the case further south-west between the Rakaia and the mouth of the Ashburton.
e. Structure of the Spit.
The structure of the spit is difficult if not impossible to determine as there are no cuttings, natural or artificial, except the beach and the gravel pits near the railway. The first of these gives no indication of structure, but some idea can be obtained from the latter.
The most interesting of the pits is that where the railway meets the spit about a mile on the Christchurch side of the Birdling's Flat Railway Station. At the further end of this excavation the lowest beds are exposed, composed of fine-grained material, well stratified, with layers lying flat or dipping south-east at angles with a maximum of 25°. Over these beds lies well stratified gravel with a uniform dip of 10° to the north-west, which continues uniform for a distance of 15 chains as exposed along the northern face of the pit (see Plate 32). The layers are generally of fine material, the larger pebbles of the fine layers up to an inch in diameter with a general average of half an inch, and with a considerable admixture of coarse sand and finer material. Inter-stratified with this are layers of somewhat coarser material in well defined layers, composed of pebbles generally up to 2 inches in diameter, and with occasional larger stones up to three inches. The stratification of this is definite and clear, and indicates that deposition took place in fairly quiet water, and not on an exposed beach. An angle of dip of the bedding as high as 10°, which may be taken to be the angle of slope of the beach as it was constructed, would indicate according to Marshall (loc. cit.) that it had been laid down in sheltered water—the angle of slope according to this authority being flatter where the seas are heavier, the size of the material of which the beach is constructed being constant.
The break in stratification shown in the section in the shingle pit is definite and clear, and indicates an abrupt change in the conditions at one stage in the formation of the spit. Such a change, not necessarily at the same time, is also indicated in the arrangement of the ridges noted above about 10 miles down the spit.
In the extensive gravel pit at the turn of the railway near the Devil's Knob the stratification is also well developed, the beds dipping seaward at low angles. One method of forming stratified deposits in connection with a spit is shown at the seaward end of Lake Forsyth and near the outlet of Lake Ellesmere. In both these places heavy seas have carried the shingle over the barrier into the lake and sorted it into layers. The regularity of the stratification in the case of the former suggests that the deposition took place in the lake itself, but it is quite possible that the sorting action is due to the rush of the water over the barrier as seems most probable in the case of the outlet of Lake Ellesmere. The stratification in the first-mentioned gravel pit is no doubt due to the beach being formed in the somewhat sheltered waters of the lake by waves coming in from the west. The direction of the convexity of the crests in its vicinity supports this contention. A similar stratification to that occurring in these pits and due to a similar cause is known in the barrier beach between Taumutu and Hart's Creek, to be referred to later.
Little can be said concerning the basement beds of the spit, since there are no natural exposures and no bores with proper logs have been put through the shingle—the only wells being shallow ones and the water from them strongly saline. At low tides a bed of clay is exposed near the outlet of Lake Forsyth, and clay beds are also exposed near the other end of the spit at Taumutu, and blocks of sandy clay, up to 12 inches cube, are found on the beach just south
of the outlet of the lake, which indicates a similar bed to occur in the sea in close proximity. With the evidence at one's disposal it cannot be said whether or not these beds are mere discontinuous lenses or parts of an extensive sheet of clay, continuous with the clay beds beneath Lake Ellesmere.
f. Wave-cut Cliffs. (See Plate 28).
The eastern end of the spit is marked by a definite wave-cut cliff. Near the Devil's Knob it is some 280 feet high, with a height of 400 feet to the east of Lake Forsyth. There is also a pronounced bluff varying in height from 200 to 300 feet to the west of the road from Birdling's Flat Station towards Little River which must have been cut by the sea. On the face of these cliffs are exposed the ends of the layers of volcanic material—lava-flows and beds of agglomerate—of which the Banks Peninsula hills are formed. There is no doubt that these have been wave-cut, and at a time when there was no spit or when the land was submerged so as to allow the waves to break against the ends of the spurs. The upper surface of the ends of the spurs is covered in many places with loess.
Indications of this wave-cut cliff occur to the west, all along the road towards Tai Tapu, but it gradually loses its definiteness when traced to the west, so that the spurs near Tai Tapu and further west still have no cliffed ends, but gradually sink down to plain-level as the spurs of the islands inside coral reefs sink down to the sea. This indicates either that a subsidence of the land took place sufficient to mask the ends of the spurs with alluvial deposits, or that they were not exposed to strong wave action, the latter alternative being the more probable.
When they were formed the height of the shore platforms along the road clearly shows that the land must have been at least 10 feet—perhaps 15 feet—lower than it is now. Of course it is not clear from the form alone whether the cutting was done by waves in a lake or in the sea, but as I will show later there is clear indication that the land was lower at one stage in the history of the spit, that its surface was practically awash, and the cutting might have been done when the sea had access to the area now covered by the lake. If the cutting had been done by lake waves only, then the lake must have been from 12 to 15 feet higher than now relative to the Peninsula, and in view of the low height of the bank of the spit near Taumutu this appears to be very improbable.
On the exposed southern coast of Banks Peninsula in the direction of the entrance to Akaroa Harbour there is little sign of these platforms. In some cases reefs run out which are exposed at low tide, and which may be the remnants of the platform that have escaped destruction by the heavy seas, but in any case it would be surprising if any definite remnants should be visible on this exposed coastline. Any platform occurring may have been determined by hard layers of basalt rather than by the sea operating at a higher level than at present.
g. Origin of the Spit.
From the shape of the pebbles of which the spit is composed, as shown in every exposure, it is clear that it has been formed by wave action and that the material has come from the south; it is also clear that the active transporting agent has been the waves acting on the shore during heavy south-easterly and south-westerly gales. Either of these is sufficient in strength to account for the transport of material along the beach.
The wave action has no doubt been helped at times by the current that flows from the south up the east coast of the island. It is difficult to obtain evidence of the strength and direction of this current. The Admiralty Chart of the coast from Akaroa to Dunedin, revised in 1922, gives no indication of its strength except that in the vicinity of Otago Peninsula it is a 1-knot current, and also that after south-easterly weather there is a set into the Canterbury Bight. As the chart gave no satisfactory evidence I wrote to a competent marine official at Timaru and he gave in a letter very interesting information as to the currents. He said that shipmasters stated that very frequently the set in the Canterbury Bight was to the east and not always as stated on the chart into the bight. Also that very frequently when the Rangitata and Opihi were in flood the water found its way south to Timaru, that barrels washed off ships' decks north of Timaru had been picked up south of that port, and also that only occasionally ships experience a current against them when travelling from Akaroa to Timaru. However, most of the material shed from ships and wharves near Timaru found its way up the Ninety Mile Beach. He was also of the opinion that an eddy occurred in the Canterbury Bight which accounted for the strength and variability of the currents off the coast.
I also wrote to the late Captain Bollons, of the G.S.S. Tutanekai, with regard to this matter and he answered as follows in reference to the longshore currents of the Peninsula and coast south of it: “These are simply surface streams formed by the prevailing weather. In the days of sailing craft several vessels were lost along shore between Ellesmere and Timaru set in by the currents created by south and south-east gales. As the prevailing winds are south-west and south-south-west, the drift is also from that direction. In small sailing vessels we could always beat a passage north but this could not be done bound to the southward. With hard north-east, north, and north-west winds a set occurs round the Peninsula and alongshore past Ouraki and Peraki and into the bight. There is certainly an eddy or an area of no current between Akaroa and Ellesmere during moderate weather conditions, but with any strong winds the surface set runs with the weather. We experience this when at anchor off Akaroa Heads.”
In discussing this matter Mr. Coates, a fisherman at the mouth of Lake Forsyth with many years' experience, said that the drift near the shore was invariably from the south, and that most of the drift wood found near the outlet of the lake was gorse, broom, and lupin sticks such as grow freely in the river-beds to the south. He said also that about three miles out there was a strong current running
towards the south from the direction of Banks Peninsula even in quiet weather, but this was intensified during north-east gales till the current approached 4 knots. Between the two was an eddy where flotsam accumulated.
Thus the main influence which affects the coastal drift of shingle is the waves aiding currents induced by the winds. This coastal drift is prevailingly from the south although there may be an eddy during certain weathers on the south side of the Peninsula which causes a set to the south-west. This is, however, some distance from the shore and does not affect the question. The force of the waves during storms is very great, for example during the great storm in January, 1929, when the Timaru breakwater was damaged, the waves came right up to the top of the bank near the huts on the south side of the entrance to Lake Forsyth, approximately 25 feet above normal sealevel. This is exceptional, and storms of moderate intensity must be looked on as the determining factor in transporting material along the belt where the waves act.
The northerly movement is aided by the fact that the sea deepens very gradually off shore. The charts show a depth of from 10 to 12 fathoms at a distance of from 4 to 5 miles off the coast near the mouth of the Rakaia, 10 fathoms five miles off Taumutu, with a gradual approach of the 10 fathom line to the coast in the direction of the Peninsula. This indicates that the grade of the sea bed off the Rakaia and Taumutu is not appreciably different from the grade of the plains near those places. This means that during storms the waves affect a somewhat broad belt parallel to the coastline.
Although it is manifest that recent aggrading of the coastline is now going on in the vicinity of the outlet of Lake Forsyth, it must not be assumed that it is going on over the whole length of the spit at present. The reach where the sandhills are developed in certain places gives evidence of strong sea attack, in fact it seems possible that were natural agencies left to work out their ends, a break through of the sea might take place some miles to the north of Taumutu.
A somewhat surprising feature of the spit is the absence of a tidal opening alongside the cliff at the present eastern end. If the spit has been formed by drift up the coast then we should expect it to be built up from the west, and following the usual sequence of events the opening should exist as long as possible at the western end, and specially so when at that end there is solid rock along which an open channel can be kept. In the case of Lake Forsyth we can see how such a channel has been kept open along the eastern side of the lake until just recently under circumstances apparently not so favourable as those which might be expected to have obtained at the present root-end of the Ellesmere Spit. Why the entrance should have been closed in the case of the larger lake and kept open in the case of the smaller one, the drift affecting each equally, is not clear, since the drainage area of the former is much larger and contributes much more water than does the latter, and when the Waimakariri, and perhaps the Rakaia also, discharged into the lake the amount of water coming in should have been greater still, and have helped to
initiate and maintain a permanent opening at the Birdling's Flat end.
The position of such openings has been dealt with by Marshall in his Geology of New Zealand (pp. 76-8) and special reference made to the Ellesmere Spit, but I am not quite certain that his explanation is satisfactory as applied to this case, although it is quite reasonable in others.
It is possible that the opening has only recently been closed, although the earliest maps of Canterbury indicate that it was closed for half the year, in the forties of last century, and that this closing may be due to a recent increase in the amount of shingle derived from coastal erosion to the south, but it is remarkable if all the shingle forming the spit has been transmitted along such a slender line as exists now. It suggests that the history of the spit is even more complex than appears to be the case, and that it may only be a remnant of a larger spit which was constructed further seaward when the land was higher, and which is now submerged beneath the belt of shallow water which fringes the spit on the seaward side. This is only a speculation and at present there is no solid evidence on which it may be based, but the peculiar circumstances of the Taumutu end of the spit are not easily accounted for and some such explanation appears to be necessary.
Along the seaward face of the sandhills there are numerous stumps of totara (Podocarpus totara), many of them standing on their roots with the lower part of the trunk vertical, and up to six feet in height. They occur just above high tide mark, and their roots are washed and frayed by drifting sand and gravel, so that all the fine rootlets have disappeared. The top foot or so of the higher trunks is riddled with teredo borings, while the remaining portion is free from them or only shows an occasional hole. This is a point difficult to explain on the assumption that they are entirely drift material, unless for some reason or other the roots and lower part of the stem of totara are not susceptible to the ravages of the ship-worm.
Presuming that the material is drift, and it seems certain that some of it is really so, it can only have come from two districts, (1) Banks Peninsula and (2) the Rakaia Valley. The hillsides of the former were formerly noteworthy for the fine totara trees that grew on them, and they may have been swept down from the Little River Valley or from the Peraki Valley further east, but as the stumps occur beyond Taumutu, this implies a definite current sweeping round the southern side of the Peninsula from the east and continuing southward along the face of the spit, in a direction contrary to the prevailing set of the sea. This may be possible, according to the evidence given in Captain Bollon's letter (p. 159 ante), but if so there should be some admixture of Banks Peninsula volcanics in the material of the spit, and I have found none for certain. In the upper valley of the Rakaia there is a quantity of totara growing, but if the drifted material has come from there, a greater quantity of beech timber should have been associated with it as the beech areas in the valley of that river exceed those where totara grows in extent. However, only an occasional piece of this timber is to be seen, and then
only as fairly small fragments never with roots attached. The stumps, too, are larger than any I have seen in the forest of the Rakaia Valley. So that the drift material has in all probability come from Banks Peninsula.
I have considered carefully the question as to whether or not some of this material may have been due to the remains of a forest that actually grew on the spit. It is admitted that there were no trees there within the memory of the earliest inhabitant, but it is not altogether an unlikely place for a totara forest to occur. As pointed out by Cockayne in his Vegetation of New Zealand (2nd Edition, p. 95) totara is a characteristic feature of the plant covering of dunes, and further I am informed by Mr. F. E. Hutchinson, of the School of Forestry, Canterbury College, that a narrow belt of totara about a chain wide is growing on the sandhills between Hokitika and Ross on the West Coast of this island, in a position exactly analogous to that on the Ellesmere Spit. If so, it demands, first of all, that the spit must have been higher for the trees to be established at all, and then must have been lower than at present to have the tops of the stumps bored by teredo. It is of course possible that these bored stumps which at times stand well back from the beach, are drift wood of a previous epoch, that they were submerged and bored when the land stood at a lower level, and the fact that only the upper portions are bored may be explained by supposing that the lower parts were protected from the teredo by sand and shingle accumulations round the base of the stump. It can hardly be due to the limitation in the range in depth at which the animal works, for the change from bored to unbored wood would not be so abrupt, and it is known that the borer works at depths much beyond what is demanded by the evidence for submergence of the land in this case. As totara is such a durable timber there appears to be no impossibility that these stumps may have been a remnant of old drift wood or of trees growing in position before the submergence of the land. If this explanation holds, then it demands that the land must have been at least 12 feet lower than at present, a conclusion arrived at on entirely different evidence.
No other timber has been discovered on other portions of the spit, and this is decidedly against the contention that a forest may have once occupied it. There should have been some remnant. All the same, the totara which may have formed a fringe of such a forest would survive as logs and stumps long after all the other material had decayed. However, on the opposite side of the lake there is a wide and continuous belt of country where there are the remains of an old forest, mostly kahikatea or white pine (Podocarpus dacrydioides) which was exterminated on the low-lying land before the arrival of the white settlers. And as we know that the land has recently been lower and the surface of the spit awash, entirely apart from the evidence of the teredo borings, it is possible that this white pine forest was destroyed at the same time. Some of the taller stumps of totara occur now at the south-western end of the spit, and if the land were slightly higher then this white pine forest must have extended as far as this and may have joined with a supposed belt of totara growing along the coast.
The correctness of this hypothesis is supported by a statement given to me personally by Mr. A. Dudley Dobson, M.N.Z.S.E., that when investigating a site for a culvert through the spit to drain the lake, he came across stumps of trees in position resting on a bed of clay exposed on the beach and visible at low tide level. The possible extension of the forest is also supported by the presence of submerged standing stumps in the lake about a mile north-east of the mouth of Hart's Creek.
If the circumstances attending the formation of the spit were such as have been outlined then there should be traces of a similar spit on the western side of the lake, formed at a time when the land was at a lower level than at present, and when the sea had access at the Taumutu end. This does certainly exist. Extending from the vicinity of Taumutu along the lake margin in a northerly direction for about two miles is a well defined barrier ridge, composed of beach shingle, well stratified where exposed, and exactly analogous to that on the southern shore of the lake. Its height where it has its greatest elevation is about 17 or 18 feet above the level of the lake, and it is remarkably level on top. The width of the summit is about a chain and it falls gently inland, but presents a fairly steep slope towards the lake. Where the stratification can be observed it is very well defined into layers of coarser and finer material, the largest stones being about three inches in diameter (see Plate 33). On the flank facing the lake the angle is about 6° but on the top there are indications that it flattens out. There is no exposure on the westward face to enable the structure to be determined.
In front of this beach there are at least two others before the level of the lake flats is reached, these are at about 7 feet and at 4 feet above lake-level, and the former of these, especially in more or less sheltered positions, exhibits a well marked barrier form. The topmost beach grades into the complex of beach material on which the township of Taumutu is placed. This must have formed a well marked blunted point or buttress when the land was at a lower level and the sea entered the lake. When traced in the reverse direction towards Hart's Creek, the barrier beach rapidly sinks about two miles from Taumutu, but a lower ridge is continued for another three miles, sometimes very indefinitely developed and then again quite clearly seen. Stretching at right-angles from this ridge out towards the lake are three definite spits, as well as another complex of beach deposits, at a lower level than the main ridge near Taumutu. The most prominent of these is just south of the outfall of Hart's Creek, in the lee of which is a considerable area of sheltered water. They are only a few feet above the lake and at times of high water the lower parts are awash. Behind the main ridge and its extension lies a stretch of exceedingly good land formed from the swamps which once fringed its inner margin, but which have now been effectively drained.
There is thus on the western side of the lake such a ridge as might have been anticipated from the conditions governing the formation of the major spit. Its existence emphasizes the fact that the sea had access to the lake while the barrier beach was being formed, for the ridge just described could only have been formed by the sea.
If it had been formed by the lake it should have been more or less continuous round its western and northern shores, whereas the break at Hart's Creek is what might have been expected had it been formed by the sea as it entered the area now occupied by the lake, and the nature of the pebbles forming the ridge supports this contention since they are almost entirely greywacke. Had the ridge been formed by lake waves alone there should have been a considerable admixture of Banks Peninsula material since the dominant wind on the lake is from the north-east, and this would have ensured that a drift of volcanic material should have come from that direction.
h. Changes in Level of the Land.
The following is a table showing the probable changes in level in the locality:
Higher land when valleys on the peninsula were eroded.
Depression, accompanied by invasion of valley floors by the sea.
Practical still-stand, no major movement when spit was constructed.
Depression till the spit was awash and barrier beach constructed on the lake margin, and the boring of the stumps now exposed on the seaward margin. At this time shore platforms were cut on the ends of the spurs along the Tai Tapu-Little River Road.
Elevation of from 12 to 15 feet, the present level.
It will now be advisable to compare this sequence of events with any chronology of movements of the land indicated in the adjacent districts. To the south there is definite evidence of a downward movement of the land in the neighbourhood of Timaru, the drowned river stream valleys incised in the edge of the Timaru Downs, as well as the presence of submerged forests at Pareora and to the north of Timaru indicating such a movement clearly. Further north the dominant feature of the sea coast, that is, the sea cliff north and south of the mouth of the Ashburton River, and specially between it and the Rakaia, postulates long continued erosion, with little change in level of the land. There may be a slight downward movement, but an upward movement is not ruled out, since the attack on a somewhat easily eroded shore line might rapidly obliterate any shore platform or marine terrace that might have been formed. Between Taumutu and the Rakaia mouth there are lagoons ponded back behind a barrier beach, but these can be explained without any demand for a sinking sea-coast, although some slight movement in that direction may have occurred. These lagoons and other pieces of swampy land may quite well represent shallow salt-water areas that have formed behind a barrier beach, and have been partly filled with detrital matter so that they are slightly above sea-level and are now fresh, if containing water, or have been turned into swampy land.
On the northern side of Banks Peninsula there is evidence of recent uplift of from 12 to 15 feet as suggested by Haast, Speight and Jobberns. This will naturally correlate with the uplift inferred in
connection with the spit. There is also evidence for a higher level of the land from the records of the artesian wells of Christchurch in that peat beds occur at various horizons down to 600 feet beneath the present sea-level. The peat must have been formed on a land surface, so its present level demands depression of the land since it was formed. Inter-stratified with the peat and gravel beds are others containing the molluse Chione stutchburyi, a marine and estuarine form now living chiefly between tides. These shells are also found in the bores put down in places like Lower Riccarton, Ladbrooks and Greenpark, and no doubt they occur also at Halswell and Tai Tapu between these three places, but the wells in the latter districts have been driven with a plug in the pipes, and so no record of the beds passed through is available. The evidence from the first three places in conjunction with that from the Christchurch area proves the former presence of a strait passing round the western margin of Banks Peninsula, as suggested by Haast (1879, p. 400) on the evidence afforded by surface phenomena. This strait must have been a comparatively narrow one as there is no record of shells having been found further out on the plains than the places mentioned and the bores have been mostly through shingle. Thus there is no record of shells in the deep bore at Islington (732 feet) which is about 8 miles north-west of the base of the peninsula hills. However, the interstratification of gravels with marine deposits right up to their foot indicates that at times antecedent to the present the plains tied, or almost tied, the volcanic mass of the peninsula to the mainland. The gravels must have been laid down on a land surface or in immediate proximity to a beach, so that the gravels inter-stratified with marine beds indicate a former higher level of the land, while the sea-shells indicate an invasion of the sea as the land sank.
The shells which are found at a depth of 300 feet in the Christchurch area indicate either a deep sea, or, as is more probable since the molluse is a shallow water form—a progressive lowering of the land, so the alternation of land and sea over the area, evidenced by the inter-stratification of land and marine beds, was repeated several times. Some of the shells brought up from the artesian wells are very stout in substance, and this is a feature shown by shells living on open beaches. So it is probable that the water where they lived was not very well sheltered, and therefore there was no effective protection against the seas sweeping up from the south round the western side of the peninsula. The closing of the southern entrance to this strait by the Ellesmere Spit no doubt took place during a period of relative still-stand of the coast, when the depth of water to the south of the peninsula was such that the transport of material could take place across the strait, and the drift did not follow the line of its western margin. This depth might have been due to change in level of the land or to the progradation of the shore line. It is unlikely that the depth of water at this time exceeded 30 feet, that is, the approximate thickness of the gravel forming the spit.
The fact that the land was much higher than at present within fairly recent times is evidenced by the coastal features of Banks Peninsula. The drowned valleys which mark the whole seaward fringe, and the alluvium-filled valleys facing the plains all demand a
lowering of the land to account for them. No precise determination of the amount of lowering that has taken place can be made from considerations affecting them, except that the minimum amount must be measured by the depths of the seaward end of the valleys. Now the depth of water at the entrance to Akaroa Harbour is 18 fathoms, so there must have been a minimum subsidence of the land amounting to just over 100 feet, but this estimate does not allow for the fact that the harbour may be partly filled with sediment washed from the neighbouring hills, or that the land may have extended some distance further seaward. The only occurrence that I know of where the depth of alluvium has been determined is in the case of the valley behind the town of Sumner where, on boring for water, solid rock was struck at a depth of about 200 feet, so that there is thus indicated a further amount of subsidence. This submergence must have taken place before the spit was constructed. I cannot see any evidence for the additional submergence of 800 feet postulated by Jobberns (1928, pp. 556-7). The presence of a level shelf over half a mile broad on the northerly slope of Mount Herbert, if due to wave action must have taken a long time to form, and the absence of similar shelves in the neighbourhood where they should occur, especially on the southern slopes of the Peninsula, makes me view the conclusion with suspicion. It seems to me far more probable that this flat surface is determined by the hard layer of basaltic rock which forms its floor. Then again if the waves have cut this bench, why is there an entire absence of any trace of benching on the spur on the opposite side of the Charteris Bay Valley. The presence of the ridge on the western side of Gebbies Pass of height somewhat accordant with the shelf on the western side of Charteris Bay appears to me a mere accident. The top of this ridge is formed by the highest exposure of the underlying substratum of greywacke and related rocks, while the steeper ground to the west is formed of the overlying volcanics. There are no benches or beaches at accordant levels on the southern facing of the peninsula such as might have been expected had the land stood still long enough for a shelf to be cut back a distance of over half a mile in the comparatively sheltered waters of the northern side of Mount Herbert. In any case a shore platform should also be accompanied by a train of phenomena and in the absence of any of these, or of all but one, a conclusion as to its existence must be viewed with suspicion. Further, if the land had been depressed till this bench was submerged all the loess which caps the spurs on the southern side of the peninsula should have been removed unless the time of the depression is moved back to pre-glacial, that is, to pre-Pleistocene times, and also there should be signs of long still-stand of the land at the 800 foot level on the Canterbury Plains and on the foot-hills adjacent thereto—but these are entirely absent. Even if the shore platform were cut in pre-glacial times some traces of it should be in existence, especially as the last phase of the glaciation was posterior to the formation of the plains, for moraine lies on top of the gravels at the Rakaia Gorge and in the Ashburton Valley.
I must also express my present disagreement with a suggestion of my own advanced many years ago (1908, p. 32), that the ends of the spurs reaching down on the southern side of Banks Peninsula
owed their flat extremities to marine erosion when the land stood at a lower level. I am sorry that I must apply the same test as given above and conclude that there has been no definite submergence and subsequent uplift of the coast to the extent demanded by the height of the ends of the spurs. I submitted my original contention to Professor W. M. Davis when he was on a visit here, and he said that he saw no evidence of this submergence, and the subsequent uplift, nor does he mention the point in his recent work on Coral Islands, although he takes Banks Peninsula as one of his types of depressed and eroded islands (1928, pp. 151-4). It seems more satisfactory to attribute the flat terminations of the spurs to the existence of a flat or flatter capping of hard rock, such as does occur on the outskirts of the volcanic mass, especially where the lavas are fluid as basalts usually are. There are no beaches on these flat spurs to support the contention that the land form is due to marine erosion. The covering of loess would naturally be deposited when the land was at a higher level and there has been no depression to wash this off since it was blown from the front of the glaciers and from the streams issuing from them in Pleistocene times.
Reference to Literature.
Davis, W. M., 1928. The Coral Reef Problem, pp. 151-4.
Haast, Julius von, 1879. Geology of Canterbury and Westland, pp. 399-401.
Jobberns, Geo., 1928. The Raised Beaches of the North-East Coast of the South Island of New Zealand, Trans. N.Z. Inst., vol. 59, pt. 3, pp. 508-570.
Marshall, P., 1929. Beach Gravels and Sands, Trans. N.Z. Inst., vol. 60, pt. 2, pp. 324-365.
Speight, R., 1908. Terrace Development in the Valleys of the Canterbury Rivers, Trans. N.Z. Inst., vol. 40, pp. 16-43.
—— 1911. Preliminary Account of the Geological Features of the Christchurch Artesian Area, Trans. N.Z. Inst., vol. 43, pp. 420-436.
—— 1917. Geology of Banks Peninsula, Trans. N.Z. Inst., vol. 49, pp. 365-392.
Note.—Since writing the above, Mr. Jobberns has drawn my attention to the existence of greywacke pebbles on the top of a stack on the east side of the outlet of Lake Forsyth. I have examined the spot and also the cliffs in its vicinity, and find that the stack and cliffs are covered with a deposit of typical loess and the pebbles either rest on this or in some few cases are perhaps included in it. The latter cannot be demonstrated for certain, but clear-cut banks of loess show no included pebbles and the few that are apparently included may have been washed down from above and been caught in loess material which has also slumped, thus giving the impression that they are included. The pebbles are of the ordinary beach type of the neighbourhood, and are exclusively of greywacke, quartz, and jasperoid rock. They range in size from three inches downward, but the pebbles at higher levels are usually small. The number is greatest at the lower levels and none occur higher than 200 feet above sea-level. There are at last two Maori ovens with burnt stones in the area looked at.
Now it might, at first sight, be assumed that these pebbles represent an old beach, and that the land was at one time ast least 200 feet lower in level, but there are objections to this conclusion. First of all, there is no definite deposit showing the structure of a beach, nor
is there any eroded shore platform on which it might rest. If this had been the case, the pebbles should have been formed of the country rock, a basalt or basic andesite, and although pieces of this do occur they are not beach-worn at all. There is also the objection, that had the land been 200 feet lower, then all the present spit would have been submerged to a depth of at least 25-30 fathoms, and the sea would have extended over a wide belt of plain to the west. It would then have been impossible for greywacke pebbles of the beach type to have been transported across such a belt of sea, at least 25 miles wide, in the numbers that actually occur, as taking into account their only possible source they must have been. The pebbles should have been of Banks Peninsula origin had they been a beach deposit. On the end of the spur terminating in the Devil's Knob there is a capping of loess, and in this at a height of nearly 300 feet are very small quartz and jasperoid pebbles, never more than an inch in diameter and in very small numbers. These might conceivably have been blown to the position where they occur, or they may have been derived from the heaps of moa gizzard stones, which they most closely resemble, the remains of moa bones being also a characteristic content of the loess. If there had been any definite beach on the hills close to Lake Forsyth, similar deposits should occur on the hills to the west, and all I can find are occasional small pebbles either on or in the loess. The ends of these spurs also show signs of having been occupied by the Maoris as ovens occur frequently. Further west still they are increasingly scarce and finally do not appear at all. So I must conclude that they were not deposited in that situation by the sea, but carried there by adventitious means. All the same I am not prepared to say what this is, except that it is conceivable that heavy seas may have been responsible for some of them near the present coastline, wind for some of the smaller ones; and also, since these spurs were formerly inhabited, that they may have been carried to their present position by man himself.†
If they have been deposited in the position where they now lie by the sea, and some of them are really included in the loess, then it again opens the question as to whether the loess is a silt deposit, as maintained by Hutton, or a land deposit as maintained by most other authorities, including Haast and Heim.