coast, and to the broad aggraded basin of the upper portion of Wairau Stream, which an the past was dammed and probably diverted by the eruptions of Pupuke volcano. West of Pupuke crater-lake, at the head of Shoal Bay, there is a wide plain at an altitude of approximately 40 ft. above sea-level, which is built of flat-lying consolidated tuffs ejected by the volcano, and is drained by a small stream flowing into Shoal Bay. Northward it passes imperceptibly into a broad southern arm of Wairau plain, whilst to the east its level is continued as a plain-like upland carved in Waitemata beds immediately at the head of Shoal Bay, and to the south by a platform, rising gently to higher levels, which is constituted by debris ejected from the northern of the two craters on the west shores of this bay. It is probable that these co-ordinated surfaces are to be correlated with an erosion-level at about the same altitude around the extreme southern and north-western shores of Waitemata Harbour, and traceable as a prominent terrace on the eastern bank of Lucas Creek.

The relation of a 40 ft. bench in close contiguity to an unfilled caldera at Shoal Bay has distinct interest. Its correspondence of altitude with adjacent erosion surfaces is too great to permit unequivocal acceptance of any supposition that, where underlain, by tuff, this feature represents an accumulation-surface affected since its initial stage only by colluvial processes. In view of this fact, and of the demonstrated widespread distribution around Waitemata Harbour of an erosion level at a comparable height above sea-level, the writers suggest that the bench on the flanks of the Shoal Bay cone is a portion of this extensive surface; as will be shewn later, it was carved during a relatively distant stage in the physiographic development of the Auckland district. From this the deduction is inevitable that some, at least, of the Auckland volcanoes began their activity in far earlier times than others by which have been buried physiographic features which are part and parcel of the modern topography.

Attention must, nevertheless, be drawn to the difficulties of coordination of the various events in the later history of the present district introduced by acceptance of the writers' views. It is shewn in the next Section that if, as is suggested, the volcanic outbursts preceded the formation of the 40 ft. to 60 ft. erosion surface, it is necessary to postulate that an extensive ancestral lowland at least 40 ft. below this erosion level was carved at a stage (3c of the summary at the end of the next Section) when the main movement had long been positive, for, at this horizon, post-Tertiary silts and clays underlie the volcanic debris. Though low-level planation at this stage is indicated by facts elsewhere, it is not shewn to be at all extensive.

On the other hand, if the correspondence of surface level of the volcanic material and erosion surfaces of Shoal Bay is regarded as merely a coincidence, the explanation of the facts is relatively simple (see Section on Later History of Lower Wairau Basin), and the birth of the Shoal Bay-Takapuna volcanoes was by no means so early as is demanded by the alternative hypothesis.

Near the mouth of Kauri Gully, a small stream flowing east from Northcote which has been responsible for the breaching of the southern of the two calderas on the west side of Shoal Bay, there is a

short stretch of surprisingly broad shore-platform, about 200 yds. in maximum width, cut in Tertiary sandstones about 2 ft. below normal high-water level. Its development is abnormal in comparison with adjacent cut-platforms, and cannot he explained as due to recession of sea-cliffs (here about 60 ft. high), because the constituent rocks trend sub-parallel to the shore and shew no apparent differences of hardness along their strike. Furthermore, apparent recession of so large an amount is unknown elsewhere in Waitemata Harbour, unless under conditions similar to those obtaining in the present instance.

The probable explanation of the platform is that at an earlier stage Kauri Gully carved a flood-plain near its mouth, and, in fact, a remnant of a similar plain originally conjoint with the other survives in Sulphur Bay close by. Upon completion of the sub-Recent submergence which has affected the whole of Auckland Peninsula, the shore-line receded rapidly as the veneer of soft unconsolidated alluvium was removed, until stayed by the more resistant and lofty cliffs of Tertiary sandstone beyond the alluvium.

Evolution of Modern Topography.

When viewed from a distant elevation such as Rangitoto, the accordance of general level of the divides of the area now described is so striking that one of the writers (Bartrum, 1922) has elsewhere described the environs of Auckland as a peneplain dissected in consequence of later uplift. Cotton (1922, p. 127, Fig. 130) has similarly applied the term peneplain to the region marginal to the local harbour, but includes as products of the one cycle two erosion levels which are here believed to indicate composite topography. Evidence from elsewhere in Auckland of peneplanation following late Tertiary emergence is clearly afforded by plateaus near North Cape (Bell and Clarke, 1910, p. 614; Bartrum and Turner, 1928, p. 102), whilst Orange (1927, p. 8) describes a peneplain from North Taranaki which is approximately contemporaneous in origin with that of Auckland, and, like the latter, was carved when sea-level was approximately 700 ft. higher than now.

At the close of Waitemata sedimentation, the Kaikoura uplift caused the birth of a broad land-mass which extended both east and west far beyond the limits of the present distribution of Waitemata strata. This may be regarded as the initial stage in the development of the modern composite topography, which clearly has been controlled by post-Waitemata movements of sea-level.

Following this emergence, long-continued erosion developed a wide peneplain which extended many miles east of the present district. Uplift then recommenced and reached a maximum probably not substantially in excess of 600 ft., The modern drainage was initiated during this positive movement, and the land-mass finally reached a stage which usually varies from early to late maturity, mainly under the control of three large stream-systems which probably are tributaries of ancestral consequent streams flowing in a south-east direction, and are themselves partly subsequent upon folds and fractures of the Kaikoura orogeny.

If a suggestion based upon incomplete observations may be proffered, it appears likely that one of the ancestral consequents referred

to followed the lowland depression traceable north-west from Hobsonville, on the north-west shores of Waitemata Harbour, through Kumeu and Kaukapakapa on the northern railway. Its eastern extension is now occupied by the harbour itself.

The three large stream-systems mentioned, with which there is immediate concern, are:—

1. Rangitopuni Stream, which probably joined the early Waitemata River on the low plains east of Kumeu, and shews close approach to N.N.E.-S.S.W. alignment.

2. Okura and Weiti Streams representing the betrunked portions of a dismembered river flowing to the east on the south side of Whangaparaoa Peninsula.

3. Orewa Stream, which enters the sea north of this latter peninsula, is also the betrunked remnant of a much larger stream.

The positive movement which commenced upon attainment of the peneplanation described, was interrupted by several well-marked periods of standstill, during which the streams became graded with respect to appropriate sea-level, and cut wide valley-floors which now constitute terraces and small plateaus at various altitudes. Some of them are very extensive and can be correlated with similar erosion-levels in other parts of Auckland Province. The highest is somewhat poorly demonstrated and may, in fact, be merely a lower portion of the early peneplain described above. It is indicated at a height of about 300 ft. to 350 ft. above sea-level by benches cut on the walls of the valleys of tributaries of Lucas Creek and Okura Stream where they head against Pukeatua, and by the remarkably even summit of a divide which extends for nearly 5 miles south from Tirohanga Hill almost to Pupuke Lake.

At a height of approximately 100 ft. to 120 ft. above sea-level, however, there are extensive uplands near Auckland City on both the northern and southern shores of the harbour, which have broad sometimes plateau-like interfluves locally covered by volcanic accumulations. This erosion-level includes almost the whole of the Takapuna-Devonport Peninsula, and is represented also by small terrace-remnants in tributary-valleys near the head of Lucas Creek. It almost certainly is to be correlated with marine benches noted by Smith (1881, p. 409) at a height of 100 ft. above sea-level at several points near Auckland and around Kaipara Harbour further north.

The next erosion surface is at a height of 40 ft. to 60 ft. above sea-level, and is by far the most extensive ancient flood-plain near Auckland. It is present as terraces in practically all the valleys of the district described in this paper, especially near Albany, in Rangitopuni and Orewa Valleys and near the limit of tidal waters in Okura Stream, and is preserved in wide plain-like remnants on the southern and western shores of Waitemata Harbour approximately west of Kauri Point. Watchman Island is interpreted as a remnant of this plain.

At New Lynn, Hobsonville, and other places, there is demonstration that this 40 ft. to 60 ft. level represents far more than a period of standstill during the major uplift we have discussed. Oscillations occurred prior to this standstill and, as will be shewn, were of considerable magnitude. Near New Lynn and Point Chevalier on the

south-west shores of Waitemata Harbour, the 40 ft. to 60 ft. bench is underlain in part by Tertiary sandstone, and in part by younger highly plastic clays which contain many richly-carbonaceous layers, and often shew stems and roots of plants in the position of growth. Near the abandoned brick-works at Hobsonville there are also some fine-textured gravel beds belonging to this formation. These clays have a maximum thickness, where examined, of over 40 ft., and rest on a slightly-undulating surface of Tertiary sandstones. At New Lynn they have suffered local tilt in common with the surface on which they rest, though it has not been determined whether this is in consequence of folding or fracturing.

These beds are interpreted as deposits formed in lakes or on the swampy floors of the valleys of sluggish streams during the progress of very slow subsidence.

It is thus clearly indicated that prior to the development of the extensive 40 ft. to 60 ft. erosion surface, there was emergence which raised the Auckland area until sea-level relative to adjacent land was much as to-day, or probably a little lower. These conditions allowed the excavation of valleys subsequently infilled to a greater or less extent by more recent clays during a slow negative movement of the strand,^* and it will be shewn in discussion of the history of the Lower Wairau Basin, that the Shoal Bay-Lake Pupuke area possibly was also eroded at this period to form a hollow a little below what is present sea-level. The upward limit of the filling of the valleys or hollows has not been determined, but it is of importance to note that the tilt of the New Lynn beds has been observed in cases distinctly to be opposed to the slope of the surface subsequently carved upon them. This indicates that the flexure or faulting which caused such tilt was approximately concomitant with the period of standstill which closed this phase of minor negative movement.

In addition to the previous examples of erosion-benches, terraces about 15 ft. to 20 ft. above sea-level are common towards the mouths of most of the streams, and should probably be correlated with raised beaches at that level which have been noted at various places around the coasts of Auckland by Smith (1881, pp. 403-410) and others.

Henderson (1914; 1924, p. 580) believes that all shore-terraces not above 120 ft. in elevation above sea-level are due to comparatively recent uplift subsequent to the submergence responsible for the characteristic drowned topography of North Auckland, but from evidence near Auckland it appears that the submergence was one of the most recent events, for although subsequent uplift has probably occurred, yet its magnitude does not exceed 5 ft. in numbers of instances, and it has followed the submergence so closely that the only wave-cut platforms developed and preserved are on exposed locations.

Further facts from other sources are corroborative of this conclusion. In the first place, the logs of bores put down by the Auckland Harbour Board, kindly shewn one of the writers by Mr. Hamer,

formerly Engineer to the Board, shew that broad sub-aqueous benches exist a few feet below low-water level, buried beneath marine filling, near the mouths of the various small streams which debouch on to the central waterfront from the low hills on which Auckland City is built, whilst the former stream-channels are represented by deep narrow infilled troughs. This indicates that flood-plains were carved at this level prior to the submergence, and it is highly probable that these accord in period with that indicated, as has already been described, on the west shores of Shoal Bay. It is believed that the lava-flow constituting the “Black Reef,” near Point Chevalier, which extends over half-way across the harbour towards Kauri Point, followed a trench dissecting one of these now-submerged flood-plains Again, Lucas Creek and numbers of small streams draining the plain-like lowlands near and north-west of Hobsonville are fringed by terraces of the 40 ft. cycle, whilst they have relatively wide deep trenches which are now largely infilled by mangrove swamp and other sedimentary filling. As will be shewn, the same fact applies to Kauri Gully on the west coast of Shoal Bay, and it is inconceivable that such trenches have been excavated by the modern streams unless when the sea was at a much lower level than to-day. The terraces thus do not demonstrate uplift subsequent to submergence.

Other supporting evidence comes from the lowlands fringing Manukau Harbour near Papatoetoe, which are built mainly of pumice-silts deposited in a great early estuary of the Waikato River, and are surmounted here and there by more recent volcanic accumulations. Much of the lowland does not exceed 40 ft. in altitude. It is bordered east of Papatoetoe by hills eroded from Tertiary pocks and drained by streams which often are margined near their debouchure by flights of terraces which are continued on to the lowland areas. Had sub-Recent uplift subsequent to the negative movement now discussed been general, it seems inevitable that marine terraces, in common with others ascribed to such uplift, should be recognizable somewhere in the vicinity of those due to the streams. Yet abandoned sea-cliffs or other signs of shore-line processes have not yet been recognised in that district, although it has been closely examined.

According to the views of the present writers, a considerable and rapid elevation took place as a culmination of the interrupted positive movement just described, and the streams deeply entrenched themselves in their flood-plains. Its rapidity is evidenced by the steep walls of the many narrow valleys drowned by succeeding submergence, and now largely filled up to sea-level by fine-textured sediment. This latter negative movement was apparently of a eustatic kind, for it has affected the greater part of New Zealand, if not the whole, and gave rise to highly-embayed coastlines illustrated by Waitemata Harbour, and the tidal estuaries of Okura, Weiti, Orewa, and other streams of the district described in this paper. As can be judged, from preceding statements, its magnitude was slightly greater than that of the culminating phase of uplift it followed.

It is difficult to make an exact quantitative estimate of the vertical movement involved, for reconstruction of the cross-profile of drowned valleys furnishes inexact data, and indicates only that it may well be of the order of 150 ft. or 200 ft., a figure which Hender-

son (1924) shews is the minimum demanded by facts in other parts of New Zealand. Henderson (1924) summarises many observations dealing with the depth of deposits in various depressed areas, but in many of these instances the filling was initiated, if not completed, long prior to the submergence now under consideration, so that the data have little bearing on the present problem.

Since the submergence was completed waves have been actively cutting back the more exposed portions of the coast, while sheltered bays have been partially filled by the progradation of crescentic pocket-beaches and by the creation of barrier-beaches behind which lagoon-deposits have accumulated. In Shoal Bay barrier-beaches of this type have been built on a perfect though miniature scale, and axe now receding, exposing lagoon-deposits of mud on their seaward face. Curved spits are also prominent near the mouth of Weiti Estuary and other larger embayments.

Cliff-recession has been excessive along the northern coast of Whangaparaoa Peninsula, which is exposed to the full force of northeast gales, as may be guaged from the fact that at the head of Maori Bay the divide has been driven back at least a mile from its original position (estimated by reconstruction of the symmetrical divide usual in areas of Waitemata rocks), until all that remains is a strip of sandstone 10 yds. wide and only about 15 ft. above sea-level. Similarly, in the neighbourhood of Murray and Brown Bays on the east coast of the district, the proximity of the present divide to the shore suggests that here too wave erosion, has been very active.

It has been stated that the final diastrophic movement in the decipherable history of the Takapuna-Silverdale area is a sub-Recent elevation of about 5 ft. This is evidenced by raised beaches at this height above storm-beach level at intervals between Takapuna Beach and Okura South Head, whilst just south of this latter locality cemented shore-conglomerates with rounded boulders of Waitemata sandstone are exposed at the mouth of two small streams at about 3 ft. or 4 ft. above high-water mark. At Black Rock, also, between Takapuna and Milford Beaches, there, is an elevated shore-platform of basaltic lava, associated with raised beach deposits and abandoned sea-cliffs (Fig. 19), whilst similar ancient sea-cliffs of soft Tertiary sandstone occur also at the south end of Long Bay (Fig. 20), six miles further north. These cliffs and platforms have been so little affected by sub-aerial erosion that they testify clearly to the recent date of this final minor elevation.

For convenience of reference the sequence of events which have moulded the present topography may be summarized thus:—

• (1). Main uplift of the Kaikoura orogeny.

• (2). Peneplanation.

• (3). Uplift, varied by at least one minor oscillatory movement in the reverse direction, and punctuated by periods of approximately constant sea-level which are represented by:—

• (a). “350 ft.” erosion surface.

• (b). “100 ft. to 120 ft.” erosion surface.

• (c). “40 ft. to 60 ft.” erosion surface. A slight negative movement of the strand immediately preceded this standstill.

• (d). 20 ft. to (?) -10 ft. erosion surfaces represented by 20 It. terraces and, as a later product, by now-submerged platforms.

• (4). Acute elevation represented by trenches now drowned.

• (5). Submergence followed by small sub-Recent uplift.

The Later History of Wairau Basin.

In the Takapuna-Shoal Bay area there are numerous localities where white silts or clays (often capped by a carbonaceous layer 6 ins. or more in depth) occur usually at a foot or two above high-water level, preserved beneath basaltic flows or tuffs. The chief exposures are on the west shore of Shoal Bay, just east of the southern caldera of that district, and on the outer margins of the Lake Pupuke cone, as along the shore between Takapuna and Milford Beaches (Fig. 18) and in artificial excavations such as a shaft sunk in search of water in Wairau Valley north-west of the lake, and several quarries west and south of it.

The silts generally are very fine-grained and plastic, and contain impressions of sedge-like plants along with stumps and fallen branches of coniferous and other trees up to 10 ins. in diameter. The latter are seen in the position of growth, as they were overwhelmed by the lava, at the north end of Takapuna Beach, and in a quarry at the head of Shoal Bay near the Orphanage. In the latter place what are believed to be the silts are higher above sea-level than elsewhere, reaching an elevation of about 12 ft., and it is possible that here they represent, not bedded silts, but ancient residual clays derived from Waitemata sandstones such as form a small peninsula immediately east of the quarry. In quarries on the west margin of Lake Pupuke these beds have been invaded and disordered by lavas which have baked them into white porcellanite; this generally shews perfect prismatic jointing in addition to less regular shrinkage cracks (Fig. 17). Along the shore between Takapuna and Milford Beaches the silt is usually followed upwards by a thin layer, about 1 ft. in depth, of fragmental volcanic material erupted prior to the outpouring of covering lavas. At one spot, however, it is overlain by a basement conglomerate beneath the lavas consisting of well-rounded pebbles and boulders up to 6 ins. in diameter essentially of basalt, which appears, from its relatively non-vesicular character, to have been worn from a flow, but including also an occasional pebble of greywacke. The conglomerate is at a height of about 1 ft above high-water mark, and represents an ancient beach formed of debris worn from early products of the adjacent eruption. It has especial interest in that it shews that the silts were exposed to vigorous wave-erosion either when volcanic activity commenced, or shortly after this; upon them a very gently-shelving cut-platform must have been developed in the protection of adjacent headlands of recently outpoured resistant volcanic rock.

The greywacke pebble probably represents a fragment from underlying beds brought up by the eruption, though it may conceivably have been carried to the area entangled in drift-wood. Amongst the ejecta of the southern of the two Shoal Bay volcanoes, occasional polished and rounded pebbles of the same Trias-Jura rock are dis-

coverable; the readiest explanation of their presence is that they have been disrupted from underlying Tertiary conglomerates such as outcrop at the base of the Waitemata Series at Motutapu, about 9 miles north-east of Takapuna.

In reconstructing the topographic conditions which obtained when the pre-volcanic silts described were accumulating, it is neceasary first to consider certain features of the present topography. It will be recollected that the aggraded floor of Wairau Valley has a southern arm which is almost continuous with, a plain, built of volcanic tuff, which extends north-west from the head of Shoal Bay at approximately 40 ft. above sea-level, and is separated by an imperceptible divide from water flowing north to Wairau Stream. From this Shoal Bay-Wairau Plain, Wairau Stream turns sharply at right angles, following the contact between the low-lying volcanic accumulations and elevated Tertiary hills north-west of them in a northeast direction to the sea. Half a mile from the open sea it is held up by a barrier of hard basalt, through which its falls have receded a comparatively short distance, whilst its upper reaches are graded with respect to this local base-level.

The various facts suggest that in Pleistocene times, before the volcanic outbursts had begun, and whilst the sea was within a few feet of its present level, the ancestral Wairau Stream flowed south into a wide shallow estuary of which Shoal Bay now represents the diminished remnant. Sedimentation in this sheltered sheet of water built up deposits which locally arose above sea-level and furnished dry land on which coniferous trees (some of which appear to have been kauri) and other vegetation flourished.

The south-east barrier of this estuary is preserved as the Takapuna-Devonport peninsula of Tertiary rocks, but the barrier east of Pupuke crater has disappeared beneath the sea. This may have been an isthmus of Tertiary sandstone destroyed during the vigorous retrogression which hag affected the adjacent coast, or perhaps a wave-built accumulation in the form of a tombolo uniting islands of the Takapuna-Devonport peninsula with the mainland north of Milford, for to-day a similar tombolo unites the former island of Devonport with Takapuna at Narrowneck. Whether isthmus or tombolo, however, the connection has been, shewn to have been destroyed by shoreline retrogression prior to the main period of eruption of Pupuke crater.

Volcanic lavas and ejecta subsequently covered these soft estuarine silts and allowed their preservation; at the same time they temporarily ponded Wairau Stream, and ultimately allowed its diversion to the north-east. The continuation of the level of Wairau Plain as an erosion-level on the north-western margin of Shoal Bay, which has been mentioned earlier, indicates that Wairau Stream continued its ancestral course for some considerable time during a period when sea-level remained approximately 30 ft. higher than now, but was finally captured by a small stream working headward from the eastern coast of the area along the contact between volcanic and Tertiary beds.

Certain evidence points clearly to the fact that the deposition of the estuarine silts and the volcanic eruptions preceded the phase of