Art. XXVII.—Notes on Wellington Physiography.^*

By C. A. Cotton, Victoria College, Wellington, New Zealand.

direction. In other words, the strike is reasonably constant in direction, being very generally N. 15° E., but varying locally from N. to N. 30° E.

The rocks of the series are of very variable strength, the weakness of some bands being due, apparently, in great part to their shattered nature. The argillites are invariably traversed by innumerable joints, and so also are the greywackes as a rule, but in places they are unjointed except on a large scale. The greywackes with few joints are very strong, forming some of the highest ridges and peaks. They weather spheroidally, and, when broken, present an appearance similar to that of an even-grained igneous rock. The shattered greywackes have in some places been rendered equally strong by the deposition of interlacing quartz veins filling the joint-planes.

Corresponding to the regular strike and the steep dip of the strata and their varying strength, there is a well-marked arrangement of ridges and valleys parallel to the strike. This may easily be recognized on a map, and in the field it is found to be the dominant feature of the topography. In fig. 1 the straight and parallel courses of the Orongorongo and Wainuiomata are especially noticeable. Parallel to these the entrance to Port Nicholson and the Evans Bay — Lyall Bay depression will be noted. On the Wellington Peninsula itself (see also fig. 2) one continuous valley, occupied by the Karori, Makara, and Ohariu Streams, is well marked. The position of another is indicated by the settlements of Karori and Khandallah; it continues southward some distance, and its northward continuation is the Porirua Valley. Even the high bluff of Cape Terawhiti is almost cut off from the neighbouring land by a deep north-south valley.

Fig. 1.—Locality-map of the Wellington District.

Land Features.

to over 1,000 ft., while a few peaks reach to 1,500 ft. and more. The texture of dissection is medium to fine.

Cycles of Erosion.

All the forms recognized appear to be due to stream-action alone. Three sets of forms are recognized, corresponding to different positions of base-level, and it is possible that forms are present corresponding to other pauses in the movements of uplift. It is evident that such pauses did occur, for remnants of coast platforms are found, in places, one above another (see p. 255). Shortness of a period of standstill during which erosive processes work is not in itself a reason why the period should not be dignified with the name of “cycle.” Huntington and Goldthwaite^* have pointed out the analogy between the term “cycle” applied to an erosion period and the term “life” applied to the period of existence of an organism. “Life in one signification is the complete existence of a normal organism during which it passes from infancy, through youth, maturity, and old age to death. The life of man in this sense is seventy years. In another sense life is merely the actual period of existence of any specific organism. An animal whose life in the first sense of the word is fifty years may die the day that it is born, but nevertheless we say that it has finished its life. A cycle in the first sense is ideal and can never be realized, since infinite time would be required to reduce any land-mass to the condition analogous to death—that is, to a plain at absolute base-level. In the second sense any region that is subjected to erosion during a definite period, no matter how short, passes through a cycle and can be described in terms of age and development.” The term “chapter,” proposed by Davis^† for an unfinished cycle, has not come into general use. A number of such brief cycles, corresponding to pauses during the earlier part of the period of uplift in the Wellington district, may have left traces on the topography not yet obliterated, and the number of nearly flat-topped ridges of varying height which lie spread out to the west of Kaukau Peak suggests that in the future, with detailed work and accurate mapping, some at least of them may be recognized with certainty. The writer, however, feels justified at present in grouping the observed features as belonging to only three cycles.

The earliest cycle of which a record is preserved by existing topographic features will be called the Kaukau cycle. Base-level stood perhaps 800 ft. or 900 ft. higher than at present.

To the next, or Tongue Point, cycle belong most of the broad features of the landscape as seen from a high point of view. Base-level stood about 250 ft. higher than at present, and during the cycle the most extensive of the elevated coastal platforms, including that at Tongue Point, were cut (see p. 255).

It was between the Kaukau base-level and the Tongue Point base-level that other pauses occurred that are mentioned above. They were, no doubt, brief, and the Tongue Point cycle itself was of relatively short duration. It is, however, of importance on account of the very recent date at which it was interrupted.

There is, lastly, the present cycle, with present sea-level as base-level.

Forms of the Kaukau Cycle.

Kaukau^* Peak (1,465 ft.) may be taken as a sample of a form belonging to the first, or Kaukau, cycle. An area of about 50 acres at the summit presents the appearance of mature topography, with gently rounded outlines, which abruptly give place to precipitous slopes, rocky crags, and torrent-ravines, forms of the next cycle. The small, gently graded valleys of the summit are transformed within a few yards into torrent-courses with rock beds and steep rock walls. There is no difference of rock-strength to account for the change, but summit and sides alike are composed of the most resistant type of strong greywacke, traversed by few joints.

In fig. 3 the slopes of the summit of Kaukau are seen in the foreground. The surface is littered with blocks of the greywacke, weathering in the manner generally regarded as characteristic of igneous rocks rather than of sedimentary rocks.

Fig. 3.—View of the East Branch of Ohariu Stream, looking Northward from the Summit of Kaukau Peak.
A graded reach, at its lowest point 250 ft. above sea-level, and incised about 50 ft. below the graded valley-floor of the earlier cycle.

Little is left of the surface belonging to the Kaukau cycle, and perhaps the most extensive remnant is a tableland nearly a square mile in extent, standing 950 ft. above sea-level, which exists on the divide west of the Makara Stream. In fig. 2 it is marked P. It has an undulating surface of mature valleys and rounded spurs, appearing from a distance perfectly plane. It is bounded on all sides by the slopes of young ravines eating into it.

Many of the higher ridges show very similar topography, though no other is so nearly plane. From these observations it appears that during the Kaukau cycle the stage of maturity was reached, and that this nearly plane area, P, stood not far above base-level. The height of Kaukau and other peaks above it shows that the relief remained fairly strong.

It is not probable that this cycle was the one which began when the folded range first rose above the sea, for planation might be expected to

be much more complete. The longitudinal drainage corresponding to the strike appears to have been established during the Kaukau cycle and a hypothetical earlier erosion period, for the adjustment of stream-courses to structure which has been preserved in later cycles points to prolonged denudation, and in rocks presenting but slight variation in hardness it is unlikely that anything like complete adjustment could be attained in a period as brief as that occupied by later cycles.

While some adjustments may have been completed in the Tongue Point cycle, there is no doubt that most streams in the initial stage of that cycle followed subsequent courses.

Owing to a peculiar set of circumstances, referred to elsewhere, the captures that have taken place during the present cycle have to some extent destroyed rather than completed the earlier adjustment. It is possible that some of these retrograde changes took place as far back as the Tongue Point cycle.

The Tongue Point Cycle.

The stage reached in the Tongue Point cycle was adolescence or early maturity. In the streams of the Makara-Ohariu system (fig. 2), for example, the stream-courses were graded, and the valley-floors occupied by

Fig. 4.—Graded Reach in the Makara Stream.

On the foreground and on right and left are benches of the flood-plain of the Tongue Point cycle.

broad flood-plains, of which abundant traces remain as benches along the sides of the valleys now trenched by the deep, young valleys of the revived streams, and scored across by the young ravines of insequent tributaries.

Fig. 4 represents the valley of the Makara. The sketch was made from a broad bench of the flood-plain of the Tongue Point cycle. Portions of this are seen also on the other side of the valley. In Plate XVIII, fig. 1, a view is given, looking southward, up the valley from about the same point. It shows the elevated flood-plain of the Tongue Point cycle on the left, and in the centre the later, steeper-grade flood-plain developed by the stream, in a graded reach, in the present cycle. By lateral swinging and migration of meanders on this flood-plain the stream has cut back the valley-slope on the right to a steep scarp.

The Port Nicholson Area.

East of the ridge upon which Kaukau Peak stands there is a complicated topography, the result of subsidence of the Port Nicholson block. The writer is inclined to believe that either the original boundaries of the subsided block were broad flexures rather than faults, or, on the other hand, the original subsidence took place so long ago that topographical evidence of faulting has been obscured by subsequent denudation. The numerous fault-lines suggested by Bell^* run parallel with the strike, and for this reason old faulting along these lines would not be rendered recognizable by revival of erosion.

The north-western portion of Port Nicholson was, however, un doubtedly let down by faulting at a relatively recent date, for along the north-western shore of the harbour there is a clearly defined fault-scarp (see fig. 9). Faulting along this line (the Wellington fault, see p. 257) took place perhaps along with, but more probably after, the submergence of the main portion of the Port Nicholson block. The fault and, in general, the subsidence of the whole block have provided the drainage of the whole area with a much shorter and therefore steeper descent than it formerly had. Two of the active torrents which descended the steep slope quickly succeeded in cutting back so as to tap the drainage of a broad mature valley, the floor of which stood 500 ft. and more above present sea-level (see “Changes in the Drainage of the Karori-Khandallah Valley,” p. 262).

From the above description and from fig. 6 it will be gathered that the coast-line of the downthrown Port Nicholson block is a normal drowned coast, passing through the normal cycle of littoral erosion which has reached the early mature stage. It is thus in strong contrast with the coasts of the neighbouring high-standing blocks described in the next paragraph.

Coast Features.
The Cliffs.

The actual outline of the coast of the high-standing block is the result of marine erosion working back from an earlier coast-line almost certainly bounded by fractures. This seems to be the only view tenable, for the amount of marine erosion necessary to cut back the present coast, with its line of lofty cliffs, from a coast-line of any other form would be enormous, and seems out of the question when a comparison is made with the recently revived condition of the land-drainage. There is no evidence of a slow sinking of the land such as would be required to keep up the activity of wave-action on a receding coast. The depths of hundreds of fathoms recorded within a very few miles of the southern coast indicate that the block to the south has sunk, and the closeness of the hundred-fathoms line to the western coast indicates subsidence in that direction.

The hypothesis of a fracture-bound coast gains further support from the relation of the coast-line to stream-courses on the land-surface. The Silver Stream (see fig. 2) rises at a height of 1,000 ft. only three-quarters of a mile from the southern coast, and flows northward. The western coast also cuts in along a north-easterly line, making an angle with both the strike of the rocks and the stream-courses. The Ohariu Stream, on the north-west, like the Silver Stream on the south, rises almost on the coast, and flows inland.

Cook Strait, which bounds the Wellington Peninsula on the west and south, has been generally regarded as the result of faulting since the time of Hochstetter, whose views were followed by Hutton and more recently by Park. Hochstetter's early view^* was that one island had been thrust laterally past the other—that is, that the movement was of the nature of a “flaw.” As has been pointed out by Suess, however, Hochstetter's later view^† was that Cook Strait owed its origin simply to the subsidence of a mountain block or blocks, and he was aware that the continuation of the North Island ranges is to be found on the same line of strike in the Kaikoura Mountains of the South Island. This relation is brought out by Marshall's^‡ maps of physical features of New Zealand.

The west and south coasts present similar features. The only projecting points are those composed of resistant rock, usually bands that are hardened with interlacing veins of quartz, filling joints. The intervening, less resistant rock bands recede as bays of gentle curvature, bounded by imposing cliffs. The larger streams emerge at beach-level, in gorges revived and steepened by the rapid recession of the coast,

while the smaller ravines are truncated, appearing as notches, hanging at various heights on the cliffs. Wherever the lower reach of a stream makes a small angle with the coast the spur separating it from the sea has the form of a razorback, due to lateral cutting by the stream on one side and the sea on the other.

The coast-line is, therefore, a continuous line of stupendous cliffs, rising in places on the south coast, where the coast-line cuts across the highest ridges, to 700 ft. or 800 ft. In Plate XIX, fig. 1, a portion of the south coast is seen eastward from Sinclair Head. The triangular cliff-facet photographed is 400 ft. or 500 ft. in height. To the west the height of the cliffs increases.

The Coast Platforms.

Along parts of the coast no relics remain of elevated platforms cut by the sea during pauses in the movement of uplift. They have either been completely cut away by the waves or cut off by faulting along new lines of fracture. At other places extensive shelves remain. The most prominent begins at Tongue Point and extends some distance west-ward (see fig. 7, and Plate XVIII, fig. 3). The shoreward edge of this shelf

Fig. 7.—The Euivated Coast Platforms at Tongue Point.

appears to indicate the base-level at the time when the streams of the district developed the greater part of the existing upland topography. For that reason the writer has named that erosion cycle the Tongue Point cycle.

The height of the shelf at its inner edge at Tongue Point is 240 ft. Its slope seaward is at first 10°, but rapidly decreases, and at the end of Tongue Point, where the shelf is half a mile broad, it runs gently out at an angle of 2° or 3°.

The upper surface of the shelf is covered by a veneer. 6 ft. or 8 ft. in thickness, of gravels similar to those of the present beach. They vary irregularly from beds of coarse roughly rounded gravel and boulders, material similar to what is being supplied to-day in large quantities by the smaller streams, to layers of fine flattened discs of beach-shingle varying from the size of a threepenny-piece to that of a penny. A layer of the coarser gravel is seen on the right in Plate XVIII, fig. 3.

The varying height of the outer scarp of this marine terrace as seen from the sea is clearly due mainly to the varying breadth of the portions that have withstood the action of the sea, the seaward slope of the shelf being regarded as nearly constant. At the extremity of Tongue Point it comes down to 170 ft. Beyond the next creek to the west, where there is a well-preserved but narrower remnant, the outer edge bounded by the present scarp is, as might be expected, higher. It is evidently this apparent variation in the height of the shelf that

Lyell,^* affected both sides of Cook Strait. It was, however, a tilt to the west, which depressed the western shore of the strait and elevated the Wellington side as a whole—that is, the area shown in fig. 1—by an amount varying from zero on the north-west to about 9 ft. on the south-east. The raised beaches of the Wellington coast which owe their elevation to that movement have been described and figured by Bell.^† They may be seen also in Plate XVIII, fig. 2, and Plate XXI, fig. 2. Both views are of parts of the eastern shore of Miramar Peninsula.

There is some evidence that this tilt is a continuation of an earlier tilting movement in the same direction, the axis of the movement lying a little to the west of Wellington. On the south-east a series of very fresh raised gravel beaches at Cape Turakirae, the highest being 90 ft. above the sea. are mentioned by Aston.^‡ On the north-west there appears to have been a downward movement of small amount subse-quently to the general movement of elevation the proofs of which have been given. This movement, which has drowned the lower reach of the Porirua Stream, does not appear to have been more than 30 ft. or 40 ft. The stream had previously developed a broad strip of flood-plain, and this has been drowned to a distance of about a mile and a half from the sea. At Porirua there appears to have been little or no movement either up or down in 1855. Raised rock platforms similar to those at Wellington are not found. This agrees with the accounts of eye-witnesses given in substance by Lyell.^§

Fig. 9.—The Scarp Of The Wellington Fault As Seen From Kelburne.
The mouth of the Kawarra is a little to the left of the centre, and the mouth of the Ngahauraga is on the right.

The Wellington Fault.
The Fault-scarp.

be regarded as having extended at least a mile out into the waters of Port Nicholson, enclosing between them the continuations of the present gorges; and the coast must have been cut back to a straight line by wave-action.

The problem may be attacked in two ways: (1.) Search for the rock platforms which should remain to indicate the former extension of the spurs. A glance at fig. 5 shows that these are absent, and that the deepest water of Port Nicholson comes close to this shore Rock platforms, if they existed, ought to have been actually raised above water by the 5 ft. uplift of 1855, but for nearly the whole length of the scarp rocks are not exposed at low water more than 50 yards from the foot of the cliffs. (2.) Comparison with other parts of the coast-line where marine erosion has been more or less effective in cutting back the coast. The coast of the seaward end of Miramar Peninsula (fig. 6) may be considered. Here, indeed, bluffs have been cut back to the extent of a mile, as the exposed rock platform at their base shows, but the coast has by no means been rendered perfectly straight. Moreover, compared with its activity on the outer coast, wave-action within the harbour is extremely feeble. A safe comparison can therefore be made only with another stretch of coast within the harbour. When the eastern shore is examined it is found that wave-action has succeeded only in shaving off the ends of points. Fig. 11 represents the eastern shore as seen from the signal-station on Miramar Peninsula. Its irregular base-line may be noted on the maps, figs. 1 and 5. It should be noted that this side of the harbour is bounded by a strike ridge, and that no spurs of any magnitude run down from it. So a shoreline originally nearly straight has been rendered but little straighter by wave-cutting. Moreover, the increasing height of cliffs towards the harbour - entrance shows that the greater part of the work has been done by waves rolling in from the open sea. The western shore of the harbour, on the other hand, is affected only by waves raised on the harbour itself. The effect of waves raised within the harbour is seen on the shore of Evans Bay (on the left in fig. 6).

The conclusion reached is that the scarp bordering the harbour on the north-west, with its straight base-line, cutting at an angle across the strike both of the rock strata and of the drowned ridges to the south of it, with its faceted spurs and its steep-grade gorges, is the result of recent faulting. Fig. 9 may be compared with the sketches and photographs of the Wasatch Range given by Davis,^* and also with the diagrams illustrating his theoretical discussion of the dissection of the face of a faulted block.^†

Nature of the Movement.

The fact that the portions remaining of the scarp along the faultline are inclined back at an angle of about 55† may indicate that the surface along which movement took place had that inclination. On the other hand, if the fault-plane were steeper the slope would quickly be reduced by slipping along the crest of the high block.

From the absence of slipped material along the base of the Wasatch Range, in Utah, Davis^* argued that the slope of the spur-facets now found there gives the inclination of the plane of faulting. In the case of the Wellington scarp, however, it is uncertain whether a scree of slipped material exists or not beneath the water and silt of the harbour. Nor can the very even slope of the facets throughout the length of the scarp be taken as an indication that they represent the actual plane of faulting. Their slope appears rather to be “the angle of rest for the products of decay” of the material of which they are composed. The writer cannot agree with Bell^† that the slope is steeper than the angle of rest. It is clear that many, if not all, of the clearly defined, sharp-edged facets owe their actual form to waveaction at their bases, the extent to which the scarp has been thus cut back being indicated by a narrow wave-cut platform at its foot. This, however, seldom reaches a width of 40 or 50 yards, and part of it may represent a levelled-off scree of slipped material. It is now almost entirely covered by the railway-embankment along the shore.

Reasons have already been given for believing that the actual movement has been subsidence of the block to the south-east (p. 258). It was assumed by Bell^‡ that the faulting movement was one of block elevation and tilting towards the north-west, and the Porirua Stream was cited as an example of a stream flowing down the tilted back slope of the block. There is no doubt, however, that the Porirua followed its present course before faulting took place. It follows one of the old strike valleys. In the valley there is evidence of recent revival, but not such as would be required by a tilt of the magnitude assumed; it appears to be due solely to the general movement of uplift which has affected the Wellington Peninsula, although perhaps not everywhere by exactly the same amount. The drowning of the lower Porirua may be ascribed to a less-extensive later tilt of a much larger block of country (see p. 257).

Other Faults.

An origin by faulting is implied for some of the longitudinal features of the Wellington Peninsula by Bell,^§ and the line of the Makara Valley is included by McKay^∥ among “active faults and earthquake rents.” The presence of many faults, and particularly of the last mentioned, is revealed in natural sections. The three faults which McKay^¶ describes as “converging on … the capital of New Zealand” can be recognized, although it is difficult to see why they are to be regarded as the continuation of faults in the South Island. The stratigraphy of the district is too little known to allow an estimate to be made of the amount of movement on the fault-planes, and the period at which the main movement took place has not been ascertained. It can be confidently stated, however, for the whole of the area west of the Karori-Khandallah Valley that physiographic evidence of recent faulting is entirely lacking (see pp. 262–64). The boundaries of the subsided Port Nicholson block may next be investigated.

On the map of Port Nicholson given by Bell^* there are indicated, in addition to the Wellington fault, five other fault-lines bounding the

Changes In Drainage Of The Karori-Khandallah Or Long Valley.

This old valley might be called the Karori-Khandallah Valley, from the names of two important settlements in it. For the sake of brevity, it is here called the “Long Valley.” Its line is now followed by the Silver Stream, the Kaiwarra and its tributaries, the upper Ngahauranga, and the Porirua. In fig. 2 the line of the old valley is indicated as a double broken line, and farther north by the line of the Manawatu Railway. Starting at the head of the valley and following it northward, we may note the changes that have taken place. At the head of the Silver Stream, which occupies the southern end of the valley, the divide is now 1,000 ft. above the sea, and the old valley appears to have continued still farther southward, the divide now being rapidly pushed northward by the activity of torrents of the south coast. Two miles and a half from its source the Silver Stream turns very sharply

fault-line until reaching the Long Valley. It has reversed the drainage of the Long V [ unclear: ] alley for a mile and a half. It follows a winding course, but the tapering shape of the spurs on the concave sides of the meander-curves indicates that the winding character is due, at least in part, to lateral cutting that has accompanied the deepening of the gorge. At one point a narrowed and almost cut-off spur is a conspicuous feature in the Ngahauranga Valley. Plate XXI, fig. 1, is a view looking north-east across this spur and up the valley. The height of the narrowed neck above the stream on the downstream side is 200 ft., and on the up-stream side 90 ft. Its breadth is about 100 yards, while the distance roundabout by the course of the stream is three-quarters of a mile. Beyond the divide, 500 ft. above sea-level, at the head of the obsequent Ngahauranga, is the head of the Porirua Stream, which, robbed of two-thirds of its ancient length, still occupies the northern end of the Long Valley. Probably this was the outlet at the close of the Tongue Point cycle.

The cause of most of the captures in the Long Valley is, as has already been indicated, the subsidence of the Port Nicholson block, particularly along the line of the Wellington fault, giving a short descent to sea-level. With regard to the Silver Stream, it seems remarkable that its capture had not taken place earlier and in a less roundabout way than the present outlet to the Karori Stream. A reasonable explanation seems to be that in earlier times, when streams followed the Long Valley and the other main longitudinal valleys of the Wellington Peninsula, the peninsula formed part of a land-area extending to the north-west and to the south far beyond its present limits. Reasons have already been given for the writer's belief that the present coast was determined by fractures after the main lines of the present drainage were established.

Type of Topography.

A consideration of the courses of streams and the elongation of ridges of the Wellington district leads to the conclusion that, apart from local complications due to unequal vertical movement, the topography of the south-western end of the North Island mountain-chain is of the Appalachian type—namely, an old, folded range subjected for a sufficient time to denudation to bring about longitudinal drainage by subsequent streams adjusted to structure, not following original synclinal folds, and afterwards elevated sufficiently to allow dissection by revived streams to produce a surface of strong relief. The analogy with the Appalachian Mountains must not be pushed too far. For example, planation in the earliest cycle seems to have been far from complete, and the absence of transverse streams following antecedent courses is especially noticeable. Their unfortunate absence accounts for the difficulty of railway-construction between Wellington and the western coast. In spite, however, of the obvious differences the remarkable similarity of our range to the Appalachians is brought out by a comparison with Lesley's map of Pennsylvanian topography, repeated by de Lapparent,^* or with the detailed maps of smaller areas given by Salisbury and Attwood.^† It may be noted that the “great Cook Strait river” of Crawford,^‡ it if existed, must have been transverse for part of its course; but reasons have been given above for believing that Cook Strait is not a drowned river-valley.