Geology of Bombay-Happy Valley Area, Franklin County, Auckland

Transactions and Proceedings of the Royal Society of New Zealand, Volume 65, 1936, Page 386

Geology of Bombay-Happy Valley Area, Franklin County, Auckland By J. A. Bartrum and W. J. Branch. [Read before N.Z. Scienće Congress, May 28, 1935; received by the Editor, June 29, 1935; issued separately, March, 1936.] Introduction. Synopsis of Physiography. General Statement of Stratigraphy. Detailed Physiography. (a) Tertiary Basin north of Whisky Creek, Ramarama. (b) The Main Central Tertiary Area. (c) Tertiary Basin east of Bombay. (d) Happy Valley and the Early History of Wairoa River. Detailed Stratigraphy. (a) Trias-Jura: Hokonui System. (b) Tertiary Beds. (c) Basaltic Lavas and Ejecta (? Pliocene and Pleistocene). References to Literature. Introduction. The present paper gives an account of an area studied at intervals over a number of years by Bartrum and described in 1927 by Branch in a thesis submitted for the Master's degree. Unfortunately, the map presented is inaccurate in regard both to geological boundaries and details of streams and roads; these two last are so imperfectly given in maps published by the Lands and Survey Department that accurate geological mapping would have entailed considerable preliminary topographic work which did not appear to be warranted. Structurally the area described is part of a major earth-block, largely constituted by the Hunua Ranges, upon which various students of Auckland University College have been at work since Dr. Laws first studied the north-west portion of the block in 1924 (Laws, 1931). In addition to Branch's work in 1927, another student, Mr J. Healy, has now completed field-work upon the central portion from Hunua southwards, whilst the southern and eastern fringe has been described by Lyons (1932). Independently of this co-ordinated work, Mead (1930) has made a study of the region, largely from the physiographic viewpoint; apart from this, no noteworthy description of its geology has been published, although Hochstetter (1864) early gave an account of the coal beds at Drury immediately north of the area now described, whilst Cox (1882) and Hutton (1871) also briefly discussed these coal beds. Synopsis of Physiography. Fig. 1 shows diagrammatically the major structure of the whole upland region embraced between the Firth of Thames on the east and the lowlands bordering the eastern shores of Manukau Harbour on the west, and illustrates how the Bombay-Happy Valley area

constitutes the southern part of a major earth-block bounded on the east by a N.N.W.-S.S.E. fault (Wairoa Fault of Laws, 1931), with relative downthrow to the west, which follows the line of Wairoa River until it makes its swing back to the north-east, after which this fault continues for some miles along the eastern margin of Happy Valley. On the west this block is defined by the scarp of another major fault (Papakura-Drury Fault of Laws, 1931), subparallel to the first; this rises steep and rectilinear from bordering flats and lowlands except where its regularity is locally impaired by volcanic extrusions. As on its western margin, so both north and south the block described abuts against plain-like lowlands; in the north against the youthful sediments that fill the fault-angle depression of Papakura Valley (Laws, 1931), and in the south against the lowlands of Pokeno Valley, its southern face being again a steep scarp which is here due to a north-east fracture which may be called Pokeno Fault. If it be permissible to regard the surface of the resistant grey-wackes which constitute the basement rock of the area as substantially that resurrected by erosion from beneath younger covering beds, it may be concluded that the block as a whole has been tilted to the east against Wairoa Fault, for again and again the summit-levels of greywacke ridges decline eastwards, whilst it appears also to have suffered a further tilt which has caused its southern portion to be upraised into a lofty range of hills well over 1000ft. in height, which is breached more or less medially by two southward-flowing streams, whilst a third further east follows south along the line of Wairoa Fault. This southern range of hills forms the boldest relief of the area described. Abruptly terminated on its southern flank by Pokeno Fault, northwards, however, it declines in height more gently, and the region becomes one mainly of closely-dissected subdued hills of moderate relief, varied by one or two cup-like hollows and by an occasional higher elevation such as an isolated hill 1046ft. in height, 5 miles north-east of Bombay, which has here been named Dome Hill. One of the cup-like hollows referred to lies about 3 miles east of Bombay and may be called Bombay Basin; it is an area of Tertiary beds now largely removed by the erosion of the head-water tributaries of Ingaia Stream. At Ararimu, north-east from this basin, there is another low-lying depression united with the last by a narrow but deep air-gap (see Map). Its modern drainage is now to the north by Mangawheau Stream, which escapes from the basin-hollow by a narrow gorge carved in a basalt barrier. A second but broad air-gap occurs in a broad meridional aggraded valley about 3 miles east of Paparata. Still further east there is a remarkable aggraded intermontane hollow, about 420ft. above sea-level, which is called Happy Valley and hugs on its north-east margin the steep scarp due to Wairoa Fault. The chief interest of this paper arises from the problems aroused by Happy Valley and the two air-gaps west of it. The only other area possessing physiographic distinctiveness, apart from the volcanic mass of Bombay Hill, is one lying north of the lower reaches of Ingaia Stream and constituted by Tertiary rocks

of soft character. It is characterised by gentle, broad slopes on which mammillated forms due to slumping of weak shaly members are not infrequent, but its limits are often poorly defined against the almost equally gentle greywacke slopes sometimes existent. General Statement of Stratigraphy. The basement rock of the area is greywacke, which varies in texture from a coarse sandstone to a very fine facies approaching an argillite, and must be referred on lithology, in default of better knowledge of its age, to the Hokonui System (Trias-Jura). It has generally been so deeply weathered at the surface that, over wide areas, it is impossible to recognize it with certainty, so that the boundaries shown on the accompanying map between greywacke and Tertiary beds are often largely conjectural. These latter strata rest with high unconformity upon the grey-wackes; certain facts suggest that, when Tertiary sedimentation began, the basement rocks had been etched into moderate, if not strong, relief, for some of the basal Tertiary strata are locally very coarse in texture. In the present area, as in its northern continuation, these younger beds show considerable variety of facies at different places, even at horizons that must be sensibly the same. The marine facies further north near Papakura, for example, differ very materially from those now described, and, as fossils are infrequent, it is impossible at present to make correlations between the beds of different areas with any degree of assurance. The basal members of the Tertiary sequence in the Ramarama and Bombay districts are fresh-water beds, which appear generally to have been deposited in more or less small isolated basins. Immediately above the greywacke there may be coarse breccia-conglomerates or conglomerates, not known to be in excess of about 10ft. in depth, and then come pale-pinkish-brown or whitish-grey fissile soft shales or fireclays generally rich in dicotyledonous plant remains. Apparently also near the base of the series, there are often white well-consolidated quartz-sandstones perhaps interbedded with argillaceous bands a foot or two in thickness. Occasionally plant material has accumulated in the basal portions of the basins in sufficient quantity to yield brown coals, which are sparsely represented in the Bombay area, though fairly general further north at Drury and Hunua, where they have long been mined in a small way. Though the basal conglomerates, coal and quartz-sandstone may be absent, the fresh-water shales are a widespread unit and may reach not less than 100ft. in thickness. The section in the lower reaches of Ingaia Stream suggests that they are succeeded upwards by greenish marine sandstones, there 150ft. in depth, which are in turn followed by puce-coloured or greyish foraminiferal shales about 100ft. in thickness; but, near the bridge across this stream just below where it leaves its greywacke gorge, the sandstones are eliminated on the south wall of the valley, and on the north are represented by merely a few feet of sandstone with casts and impressions of molluscan shells.

The uppermost member of the Tertiary sequence is a somewhat incoherent dark greenish sandstone of unknown thickness, which outcrops freely in characteristic breakaway scarps north from Ingaia Stream to Ramarama-Ararimu Road. Basalts, probably mainly Pleistocene in date, have been emitted from a number of centres, those of the west being merely a few of the many aligned along Papakura-Drury Fault. The earliest eruptions were subsequent to uplift and erosion of the Tertiary beds, and, in nearly all cases, to cessation of any important movement along the major fractures, for near Bombay extensive extrusions have masked the original scarp of Papakura-Drury Fault, whilst, further east, flows from the hinterland have occupied valleys excavated deeply across the barrier range that is bounded in its southern flanks by the scarp of Pokeno Fault. One centre of eruption, however, namely that north of Whisky Creek, just east of Ramarama, seems to have been active before completion of movement along the adjacent fault-plane, for its lavas end abruptly at a height of about 250ft. above the lowlands nearby, and the small streams at hand would appear to be incapable of breaking up and removing the large columns shed by this mass. Vulcanism has nowhere been recent, for considerable modification of the topography has arisen since activity ceased, with the result, for example, that basaltic caps crown the summits of lofty isolated residuals, such as Puketutu (1229ft.) and Trig. 619 (886ft.). Basaltic material is spread widely over the lowlands west of the uplands, but will not be described, for the problems of the lowlands have no relation to those now discussed. The youngest sediments encountered are referable to the Pleistocene and Recent, and are terrace gravels and other alluvium particularly well represented in the intermontaine basins and the lowland fringes of the area described. A patch of unconsolidated alluvial gravels and clays with pebbles of greywacke occurs at 310ft. above sea-level, perched high on the eroded face of the fault-scarp east of Ramarama, where the road north of Whisky Creek climbs this scarp. They appear to be referable to a period when movement along the fault-plane was still in progress, and are therefore older than any others of these latest sediments. Detailed Physiography. The authors propose to give details of physiography merely as required in connection with the elucidation of problems arising from certain specific topographic features, which are discussed in sequence below. (a) Tertiary Basin north of Whisky Creek, Ramarama. A small hollow about a mile long and half a mile wide, immediately north of the lower part of Whisky Creek, is encircled by a well-elevated rim, constituted mainly by greywacke, and unbroken but for a small deep gap on the west, which allows escape of a small stream (see Fig. 2). Fresh-water shales form the undulating floor

of the hollow, and on the west abut against a steep slope of grey-wacke which is dissected by one or two miniature hanging valleys. On the north and east the passage from Tertiary to greywacke terrain is indefinite. Though the present cup-like form of this area is undoubtedly a consequence of selective erosion, whereby there has been removal of a noteworthy depth of soft Tertiary beds, it seems probable that the present relation of these latter to the greywacke is due to tectonic causes. The steep western greywacke margin of the basin, it is suggested, is a resequent fault-line scarp, developed with reference to a north-north-east fault with a throw to the south-east attaining not less than 150ft. in magnitude, but apparently decreasing rapidly northwards. The only evidence in favour of this suggestion is the apparent absence of coarser clastic beds, such as expectably would occur were the steep greywacke slope the wall of a hollow in which Tertiary deposition took place, whilst the Tertiary shales on the floor of the hollow now dip westward at from 20° to 25° towards the line of the hypothetical fault. It is not clear why the stream draining this small basin did not take a southerly course along the soft Tertiary sediments which apparently provide a ready route to Whisky Creek. Space does not permit full discussion of all explanatory hypotheses possible, so that the authors will content themselves with suggesting that the early consequent drainage at the locality concerned may have taken its course approximately south-west along the fault-angle depression caused by the fault that has been invoked, and was unobstructed by the volcanic mass now in existence, but then not extruded. When the lavas were emitted, however, such a route would be blocked, and the ponded waters would overflow by the lowest gap—here over the western rim. An objection to this hypothesis is based on the fact that, as shown earlier, it appears as if the basaltic lavas concerned were outpoured before completion of major movement along the Papakura-Drury Fault. Thus the north-north-east minor fault postulated would have to be earlier than the major north-north-west fracturing. This conclusion does not accord with reasonable expectations, so that at first it appears by no means unlikely that the outlet stream discussed is antecedent in its nature and was superposed upon the greywacke from overlying Tertiary strata. The basalts, however, were not extruded until these latter sediments were stripped from part of the greywacke of the western rim of the basin, for they there rest upon greywacke, so that, if this explanation of the drainage be accepted, it must again be concluded that the volcanic episode was subsequent to faulting along the north-north-east line, and the same objection as before applies. Other conceivable hypotheses meet with similar difficulties, and it appears most reasonable to the authors to accept the first hypothesis and admit the possibility that the north-north-east faulting invoked preceded that which caused the Papakura-Drury fault-scarp. Such north-north-west fracturing can well explain the low elevation of Tertiary beds both at, and not far south of, the debouchure of Whisky Creek.

(b) The Main Central Tertiary Area. As just explained, it is likely that the Tertiary strata south of Whisky Creek have been downfaulted along a N.N.E.-S.S.W. line, so that the block on which they rest has a downward tilt approximately to the south-west. The southern margin of this area of Tertiary beds follows a sinuous line where it borders the greywacke, which occurs in slopes rising fairly steeply to higher levels. A stream approximately follows the junction for some distance, though locally superposed upon the older rock. There is no topographic evidence of faulting along this east-west direction, and it is impossible to decide whether the steep contact evidenced is a result of down-flexure or fracture, or whether one views in effect the resurrected surface of the greywacke terrain on which Tertiary deposition occurred. In the middle of the area now described, a fault, perfectly demonstrated both by stratigraphic and topographic evidence, separates the area of fresh-water Tertiary beds on the west from marine members higher in the same sequence, but at lower altitude, on the east. A resequent fault-line scarp about 150ft. in average height trends N.E.-S.W. for about half a mile, followed at its base by a small stream, whilst further east the surface rises gradually as an inclined gently-undulating plain-like area furrowed shallowly by several tiny sub-parallel consequent tributaries (see Fig. 3). This plain-like area is underlain by marine sandstones, which are freely exposed in deeper plough-furrows and represent the highest member of the Tertiary sequence in the Bombay district. In the fault-line scarp, greywacke rises locally over 100ft. above the bed of the stream below, and is then covered by fresh-water beds exposed for a depth of 25ft. in an abandoned quarry, but thickening very materially locally where the greywacke surface beneath them declines steeply, for, a little north-east of the quarry, the fresh-water beds descend to not far above the bottom of the scarp. These beds consist of a characteristic, white, fairly compact quartz-sandstone containing leaf-impressions, and occurring in bands about 2ft 6in deep, interbedded with thinner argillaceous layers. On analogy with the best-exhibited sequence of the Tertiary strata available, it is clear that, near the quarry, the downthrow of the fault is of the order of 200ft. to the east. This throw diminishes as the fault is followed north, however, for foraminiferal shales, which are stratigraphically below the uppermost marine sandstones referred to, outcrop not far east of the fault-line in a cutting of Ramarama-Ararimu Road, where they appear to dip gently westward. Though the relatively low general level of the marine beds of this Tertiary area as compared with that of neighbouring greywacke is in part due to the fault described above, it is difficult to decide, as in other localities, whether towards their eastern margin the various members of the Tertiary sequence overlap upon a steeply-rising greywacke basement, or whether the steep greywacke slopes, which rise above the surface of the younger beds to a height of 1046ft. in Dome Hill, represent substantially the plane of a fault. Had the dip of the younger beds shown any close approximation to the slope of the contact, flexure could have furnished a sufficient explanation of the

mutual relations, but the beds dip approximately west at an angle not exceeding 10°, whilst the contact plunges steeply in the same direction at about 25°. In the failure of definite evidence, the authors can only suggest that it is by no means unlikely that a meridional fault with downthrow to the west determines the western margin of Dome Hill, and is partly responsible for the superior altitude of this greywacke mass as compared with that of the Tertiary beds west of it. On the other hand, there is no possible doubt that the greywacke relief adjacent to areas of Tertiary deposition in this locality was considerable and that the land-forms were rugged, for breccia-conglomerates locally of very coarse nature occur at the base of the fresh-water series. A convincing section of these and other beds of the Tertiary sequence can be examined near what will herein be referred to as the mid-Ingaia bridge, about 2 miles south-west of the trig. station on Dome Hill. In a small rill alongside the road climbing the northern wall of the valley from the bridge, a skin of coarse greywacke débris about 3ft. in thickness, with its material little worn by water and including, as well as smaller fragments, blocks 2ft. and more in diameter, veneers a surface of greywacke which can be traced for about 2 chains along its steep south-westward plunge at an angle of about 25°. About 9ft. of similar material is exposed at the bridge and passes rapidly through a small thickness of true conglomerate of less coarse nature into sandstones and other fresh-water strata which do not exhibit the steeply-inclined attitude of the coarser beds, but, after an interval of a very few yards, are found dipping very gently westward. It seems impossible to account for these coarse breccia-conglomerates as other than the deposits of vigorous youthful streams. Their relatively small depth, however, preserved along the direction of dip for not less than 5 chains, indicates that they cannot have been deposited in an alluvial cone, as must necessarily have been the case had the greywacke surface on which they were laid down had originally, as now, an inclination approaching 25°. The conclusion, therefore, follows that this surface has been steepened by flexure or other tectonic movement, and that steep and lofty unreduced greywacke residual elevations existed on the margins of the basin concerned during the deposition of the Tertiary beds therein. It will be demonstrated in the next section that precisely similar conclusions emerge from the evidence available in the locality there described. (c)Tertiary Basin east of Bombay. This basin, which may be called Bombay Basin, is a hollow displaying fairly broad valley-floors composed of modern alluvium alternating with rolling hills of various Tertiary strata. On the north-west the greywacke slopes of Dome Hill (1046ft.) rise very steeply about 500ft. above a valley-floor at their base; on the

west and south-west there is a barrier, provided by the volcanic mass of Bombay Hill, which gives place southwards to the steep greywacke slopes of Puketutu (1229ft.); eastward and south-eastward the greywacke surface beneath the Tertiary beds rises very gently to emerge in low marginal hills. Headwater tributaries of Ingaia Stream drain the basin and unite to leave it by falls cut in basalt and greywacke at its north-west corner. Again, as in the Tertiary area north-west, there is evidence of the existence of highlands bordering the Tertiary lakes in the presence of coarse basal conglomerates of fresh-water nature in the Tertiary sequence. Similarly topographic data suggest that a powerful N.E.-S.W. fault forms the northern boundary of the basin, whilst the abutment of coal beds against greywacke at the base of the steep eastern slopes of Puketutu may be said to augment the evidence in favour of a meridional fault passing near the mid-Ingaia bridge. This displacement near Puketutu, however, would have to be interpreted as involving downthrow to the east, whilst west of Dome Hill such downthrow, if faulting occurred, must have been to the west. Bombay basin, therefore, is interpreted as structural in origin, developed by warp-like down-tilt of a block bounded by faults on its north-west and western margins. Whether these faults are coëval with the major Papakura-Drury fracture further west could not be determined. The present low altitude of the floor of the basin arises from subsequent differential erosion whereby the soft Tertiary beds have largely been removed, whilst the hard greywacke has been relatively little affected. It is suspected that the north-east fault postulated as defining the south-east border of Dome Hill has been almost contemporaneous with the meridional one similarly invoked, and has probably been continued to the south-west by a sag which may have caused the rapid decline in the elevation of the surface of the greywacke from Puketutu, where is is at about 1000ft., northward until, in a distance of about 3 ½ miles, Tertiary beds appear beneath basalt at an elevation of about 600ft., the greywacke basement being stratigraphically about 200ft. lower. Not far from where this happens, the same lavas have also buried greywacke spurs descending westward from Dome Hill, and have so deviated Ingaia Stream, which drains Bombay Basin, that it has turned north and carved for over a mile a deep gorge in greywacke spurs upon which it was superposed. Prior to the outpouring of the lavas, the most northerly of the three present head-water tributaries of Ingaia Stream was reinforced by the drainage of the large intermontane depression at Ararimu which now flows north to Mangawheau Stream. A narrow sinuous through-valley, with a very low saddle at the present air-gap (1 ½ ml. north-east of Dome Hill), now connects this latter depression with the Bombay one. It may be suggested that this early reinforced stream flowed south-west direct to the lowlands near Bombay township, its course continuing the line of the N.E.-S.W. fault that

is believed to delimit the contact between the greywacke of the southern slopes of Dome Hill and the Tertiary beds south of it, and following the axis of the sag that has been postulated as forming an extension to this fault. Whether or not this course is an inherited one consequent upon early tectonic movements is immaterial, for in effect it is now subsequent in the portion that borders the southern slopes of Dome Hill, and undoubtedly long had been so prior to the eruption of the basalts, for, when this latter event took place, the stream had excavated its bed through the Tertiary cover until the greywacke basement was exposed in the north-west corner of Bombay Basin. Later this greywacke was covered by lavas, extruded from Bombay Hill as centre, which effectually masked the westward continuation of the valley in addition to diverting Ingaia Stream and causing it to encircle the eastern and northern margins of the volcanic mass erected. The regular unbroken descent west of the slopes of Bombay Hill as a sheet of lavas masking any pre-existing scarp along the line of Papakura-Drury fault, shows that volcanic extrusion culminated after faulting had ceased. It is thus by no means improbable that the reversal of drainage indicated by the air-gap north-east of Dome Hill was a consequence of warping or tilting connected with faulting, and was in no way a result of the blockage of westward-flowing drainage by lava barriers, but on this point there appears to be no conclusive evidence. Similar reversals have, however, occurred in an air-gap valley about 4 miles east of Paparata and in Happy Valley, but in both these cases the cause appears largely to have been temporary obstruction of southward-flowing streams by floods of lava. (d) Happy Valley and the Early History of Wairoa River. Interesting problems arise in connection with the extraordinary alluviated intermontane hollow called Happy Valley and its modern drainage, whilst intimately associated with them is the early history of Wairoa River. The full history of this river will not be elaborated, but only such parts as are necessary for discussion of the origin of Happy Valley. This latter valley is a broad hollow carved in greywacke; in its northern portion it trends south-south-east for three miles or more and then bends to the south-south-west for a further 2½ miles (see Fig. 5). Its floor has been aggraded to a considerable depth by gravels which are relatively coarse 20 or 30 ft. below the surface, where encountered in wells, though nearer the surface finer detritus is dominant. As a consequence of the alluviation, a broad plain over a mile in width has resulted. In the southern and mid portions its surface is almost horizontal at a level varying from about 400 to 420 ft. above sea-level, but in the north-west it is considerably terraced, the uppermost terraces, however, being at substantially the same height as the undissected plain elsewhere, though rising slightly towards Wairoa River. The greywacke slopes on the western side of the valley are relatively gentle, but on the east wall of its northern arm there is an abrupt, lofty slope, bordered by occasional alluvial fans, which is

interpreted as a scarp developed in connection with the great N.N.W.-S.S.E. Wairoa Fault. There is much to suggest that it is a resequent fault-line scarp; as, however, its actual nature does not greatly affect the main problems of this section, this point will not be discussed. Three streams have their source in the depression, leaving it by diverse routes. Two drain the southern and central portions, and each, in its escape, cleaves the hills forming the south-eastern rim of the basin. The more northerly of these two streams follows the line of Wairoa Fault, here apparently greatly reduced in throw, and after traversing lofty falls eventually reaches Mangatawhiri River just after it has debouched from a deep gash it has carved in the mountainous highlands further east. The other follows the southern arm of Happy Valley and flows west until it joins up with other southward-flowing tributaries to form Paraureroa Stream, which passes south through a mountain barrier over 1000ft. in height by means of a remarkable narrow gorge and finally unites with Mangatawhiri River on the low-lying plains of Pokeno Valley (see Figs. 6 and 7). In the scarp north of the tennis courts in the middle portion of Happy Valley, the third diverging stream has its birth, and pursues its way north-west across the plain, entrenched to an increasing degree below this latter, until it joins Wairoa River where this stream makes its remarkable boat-hook bend, and turns north-west to follow a narrow, but not deep, valley excavated along the line of Wairoa Fault, and eventually flow into the Firth of Thames. Thus some of the drainage of Happy Valley escapes to the east coast of North Island, whilst the rest goes by way of Mangatawhiri River into Waikato River and thus to the west coast. Before its alluviation, Happy Valley was strikingly more widely open than its narrow outlet valleys, and several causes appear to be responsible for this fact. So far as the north-west branch of the valley is concerned, it is probable that its gently-sloping wall represents a pre-Tertiary surface of greywacke which has been tilted towards the east by downthrow along Wairoa Fault, resurrected by erosion, and subsequently modified by such erosion (see Fig. 8). A close comparison is afforded by the fault-angle depression of Papakura Valley (see Laws, 1931). It is likely that the amount of downthrow varied at different points along the line of Wairoa Fault. For example, it almost certainly is greatly reduced near the boat-hook bend of Wairoa River, where two sub-parallel members of the fault-system define a perfect fault-splinter (Fig. 4), whilst topography suggests that the displacement also was reduced as the fault approached the present course of Mangatawhiri River. Another factor that may have operated is a north-west warping or tilting of the major earth-block concerned towards the north-west as an accompaniment of movement along the N.E.-S.W. Pokeno Fault. This would compel any undefeated southerly-flowing stream to carve a deep water-gap in the uprising block, and it is believed that the acute V-shaped pre-basaltic gorge of Paraureroa Stream is an example of such compulsion. On this hypothesis Pokeno Fault

must be of later date than Wairoa Fault. It is true that this interpretation of Paraureroa Stream as antecedent is open to challenge, but no more satisfying explanation of the facts appears to be available. The writers are unable to offer any satisfactory explanation of the wide south-west arm of Happy Valley and can only make the tentative suggestion that it occupies a local downsag. Upon early incomplete examination of the area, the writers formed the tentative conclusion that, prior to taking its present north-west route from the boat-hook bend referred to above, Wairoa River flowed south-east along the northern arm of Happy Valley and escaped from it by the route along the south-east continuation of Wairoa Fault. The substantial uniformity of level of undissected portions of Happy Valley plain and the coarseness and high degree of rounding of some of its materials show clearly that the aggradation evidenced is the work of a far larger stream than now crosses the plain. Such a stream can only have been Wairoa River. With further observation, the writers have come to agree with Mead (1930) that the main outlet of this early Wairoa River was not as stated above, but along the south-west arm of Happy Valley and through the gorge of Paraureroa Stream. After leaving the alluvium, the present waters from Happy Valley following this route cross a narrow barrier of basalt for about 150yds. and then tumble over falls about 100ft in height caused by the basalt, but not exposing its base. This shows that the pre-basaltic valley at this place was excavated in excess of 100ft. below the level of Happy Valley. In contrast, the greywacke barrier in which the more northerly outlet stream has cut its bed even to-day is not more than about 40ft. below the surface of the plain, whilst in the past it must have been at a substantially higher level. It is obvious, therefore, that this outlet cannot have been in existence at the time when the deep pre-basaltic gorge of Paraureroa Stream was in effective operation and that a change of conditions was necessary before this northern spillway was initiated. This change was brought about by the outbreak of volcanic activity from the centre mapped about a mile west of Trig. 619 and possibly also from the hill on which this trig. station is placed, for this bears a cap of basalt. The flood of lava followed down the valley of the stream here named Air-gap Stream into the Paraureroa Valley, continuing well into the outlet gorge, though apparently it did not reach the flats of Pokeno Valley. This is not the only local example of occupancy of stream-valleys by lavas; another occurs about 4 ½ miles further west, where basalts have followed a valley carved deep in the greywacke hills and, emerging on the Pokeno flats, have there spread out in fan-like fashion. The surface of the flow that occupied the Paraureroa valley attained a height of 100ft. or more above the present level of Happy Valley and must have ponded effectively the waters of Wairoa River and those from the air-gap valley. Soon, however, overspilling of the ponded waters took place more or less contemporaneously over the basalt in Paraureroa valley and, further north, along a gap now occupied by the stream following the line of Wairoa Fault to Mangatawhiri River.

The period of maximum ponding must have been relatively short, for it is not represented by any shore-line benches or terraces, so that at first there must have been fairly rapid lowering of the levels of the barriers until they became stabilised upon fresh unweathered greywacke in the north and upon fairly resistant massively-jointed basalt in the Paraureroa, at a height represented by the present level of Happy Valley (400ft.-420ft.). Following this stabilisation came the infilling of the depression to its present level. Whilst in all probability there may have been early spilling of the ponded waters north-west down the course of the modern Wairoa River, such as would have assisted in the capture that later became complete, such capture was not effective until the building of Happy Valley plain was practically finished; this is indicated by the discovery in post-holes just upstream from the basalt barrier at the south-west end of Happy Valley of gravels of far coarser and more rounded nature than could be ascribed to the modern stream there existent. When, however, capture was perfected, the north-west reversal of drainage reflected in the modern streams was speedily effected. There is, however, nothing to indicate whether the considerable recession of falls, whereby the greater part of the basaltic fill of Paraureroa valley has been removed, has been the work of the larger stream that followed this route before the reversal of drainage or that of the small modern stream. This recession has continued until a strip of basalt not more than 150yds. in width intervenes between the falls and the alluvium of Happy Valley. An interesting further example of the influence of the lavas upon the modern drainage pattern is afforded by a small stream west of Air-gap Stream; this has followed for over two miles the contact between basalt on the east and greywacke on the west until it joins Paraureroa Stream. Nothing so far has been mentioned as to the date of the reversal of part of the drainage of the air-gap valley. The aggradational up-building of this valley reached approximately the level of Happy Valley, so that its history may be held to parallel that of this latter area, and to have entered on a new phase subsequently to the blockage of the southern end of the valley by basaltic flows. The stream now leaving the northern side of the air-gap is a reversed one in its head-water portion, and enters Wairoa River near the elbow of capture of this latter. The pirating stream can only have gained sufficient virility to be able to deepen its valley floor and effect rapid headward extension after the captured Wairoa waters had substantially deepened their new valley. It will be noted that, in this section of the present paper, the authors have assumed that no movements of tilting or warping have occurred since the excavation of the original Paraureroa outlet-gorge. The fact that this was carved in pre-basalt times approximately to the level of the Pokeno Valley plains demonstrates that substantially this is true; it is, nevertheless, possible that a slight degree of back-tilting is indicated at the head of Air-gap Stream, for, in spite of the fact that, at an early date, this stream carved a terrace not less than 20ft. in height from the broad alluvial flat occupying the air-gap, to-day there is a broad swamp bordering the terrace (see Fig. 9).

Dr. W. N. Benson very kindly drew the attention of the authors to the similarity of certain of the features of their district to those of the Puy-de-Dôme region of the Auvergne area in Central France, and lent a paper by Glangeaud (1908) and the Clermont sheet (No. 166) of the Service Géologiques des Mines, issued in 1910, which illustrate the geology of this region. On an incomparably grander scale than in the New Zealand area described, Pleistocene flows from volcanoes of the Chaîne des Puys and other centres have travelled many miles along valleys excavated in an elevated plateau of ancient crystalline rocks until they have covered part of Tertiary lowlands further east, which are separated from the ancient rocks by a great N.N.W.-S.S.E. fault. In addition, there is in the Auvergne, as in the Bombay area, remarkably close relationship between the distribution of the eruptive vents and the fracture systems. Detailed Stratigraphy. (a) Trias-Jura: Hokonui System. The basement rocks of the area described are greywackes which are microscopically similar to others general near Auckland, and contain a moderate quantity of feldspar (dominantly various plagio-clases), a variable amount of quartz in angular grains, tiny pellets of fine-grained andesitic lavas and fragments of pyroxenes and of hornblende showing different degrees of chloritisation. They vary in texture to a limited degree and invariably are massively bedded and traversed by numerous irregular joints which are generally occupied by thin seams of quartz, but occasionally, as in a small quarry about a mile north of the tennis courts at Happy Valley, by calcite with rare prehnite. About 6 miles east of Bombay the greywackes contain a deposit of manganese oxide ore which, though small in extent, is probably worth commercial exploitation. Here and there, according to local report, they include minor quartz veins with small and unpayable content of gold. The manganese ore is the product of oxidation of one or perhaps two manganiferous siliceous lodes outcropping about a mile south-west of Trig. 616 on the southern slope of a valley which drains south to the flats near Pokeno. When examined by the authors early in 1934, it appeared as if there were two intersecting veins, but the excavations made were too few to enable this point to be established, and one vein may merely be the fault-displaced continuation of the other. Near creek level a 10ft. body of ore is exposed and is largely hydrate or oxide of manganese with a considerable quantity of reddish haematite-stained siliceous material; it has a strike of 70° and dips at 65° to the north. About two chains further east, up the hill-slope, at an elevation of 700ft., a sub-vertical 10ft. vein, again with masses of reddish quartz included with the ore, is enclosed between walls of shattered greywacke and has a strike of 190°. The same lode is again visible in prospecting pits a chain and a-half further uphill, but is irregular and widens to 25ft. where

exposed in an open cut, whilst a little further on it appears to terminate against an oblique fracture with a strike of 158°, though additional prospecting might locate its continuation. No internal evidence as to age is afforded by the greywackes, so that they are relegated to the comprehensive Trias-Jura Hokonui System on analogy with lithologically similar rocks elsewhere. (b) Tertiary Beds. In the Tertiary succession as determinable there are considerable variations in the nature of the beds pertaining to what appears to be the same horizon in neighbouring localities. Satisfactory palae-ontological evidence for correlation is absent, so that there is no present justification for any attempt to match the various strata with those of more northerly areas east of Auckland where some detailed study already has been made (see Laws, 1931; Firth, 1930), though they may well prove to be basal continuations of the Wai-temata Series. As already shown earlier in this paper, the relief of the grey-wacke terrain when the basal Tertiary beds were being deposited was locally sufficient to give rise to the accumulation of coarse conglomeratic beds on the margins of enclosed basins, whilst in the deeper waters lacustrine shales were laid down, rich in dicotyledonous leaves* Unger (see Hochstetter, 1864) describes several species from similar beds near Drury. and occasionally yielding fresh-water molluscan shells. The conglomerates outcrop in only two of the Tertiary areas described, one of which is east of Bombay, where they may be seen in two small quarries, one three hundred yards east of where Bombay-Paparata Road crosses Ingaia Stream, and the other a quarter of a mile south of this latter point. The other Tertiary area is the main central one, and the place of outcrop is near mid-Ingaia bridge south-west of Dome Hill. where these beds are represented by very coarse breccia-conglomerates, not exceeding 9ft. in maximum thickness, which rest on a greywacke surface inclined west at angle of from 20° to 25°. They grade upwards into a few feet of sandstones with broken plant remains, the dip flattening very rapidly westward; the sandstones are followed in turn by greyish fresh-water shales with well-preserved leaf impressions. In the area east of Bombay only a few feet of conglomerates are exposed, whilst neither upper nor lower surface is visible, so that their thickness is unknown. Neither is there any continuity of exposures, so that the exact sequence of the Tertiary succession there is a little doubtful. Above the conglomerates, in all probability, come medium-grained white quartz-sandstones with plant-impressions; these outcrop in two places, one about half a mile north-west of the crossing of Ingaia Stream by Paparata Road and the other against steep and high greywacke slopes at the north-east corner of this Tertiary basin, about 1 ½ ml. south of the trig. station on Dome Hill. These beds are believed to represent about the same low horizon as coal and associated shales in a small mine on the north flank of Puketutu (see Map), which were also encountered, according to local report,

about 1 ¼ ml. north-north-east from there in a well sunk on the floor of the valley of the middle head-water tributary of Ingaia Stream. At the now-abandoned mine, the base of the coal seam was not exposed when visited by the authors, but 4ft. 6in. of lignite was visible dipping south-south-west at 7° beneath fissile grey shales with leaf-impressions. A few yards east of the collapsed drive of the mine, greywacke was encountered in a short prospecting drive, so that the extent of coal available at this locality must be very small, for the direction of dip will cause the seam soon to abut south against the greywacke of the northern lower slopes of Puketutu, whilst erosion will have removed any northern continuation unless carried to lower levels by change of dip or faulting, such as indeed appears actually to have occurred, if one may credit the reported discovery of coal in a well a mile and a half north-north-east from the abandoned mine. Apparently succeeding the fresh-water shales in the Tertiary succession of Bombay Basin are greenish unfossiliferous sandstones, believed to be marine in origin, outcropping about 150yds. east of where Paparata Road crosses Ingaia Stream, which are poorly exposed for about 30ft. in depth, though their full thickness is not determinable. As their base is below the level of outcrop of the conglomerate 200yds. north of them (645ft.), it may be that, as in the section near mid-Ingaia bridge, this latter bed forms merely a veneer to an inclined surface of greywacke. Above the sandstone there is an eroded cap, not less than 30ft. thick, of foraminiferal shales conformably above the sandstone and appearing to dip with it to the south at 5°. They are well exposed at the road-junction about ¼ ml. east of Ingaia Stream and reappear about 300yds. further along Paparata Road, whilst the greenish sandstones outcrop again in a paddock about 150yds. east-south-east of the road-junction. From the evidence of facts elsewhere it is clear that the marine beds described represent merely a remnant of an extensive sheet of such Tertiary beds now removed by erosion. There is no reason indeed for suspecting that, throughout the whole area described in this paper, there were not the same conditions as led to widespread marine deposition in the Tertiary in the Papakura-Ardmore-Hunua district further north, which was described by Laws (1931). If a cover of younger strata extended broadly over the whole area at the close of Tertiary sedimentation, the present limited extent of such beds is merely a reflection of the degree that the region has suffered faulting or other tectonic movements, which have caused relative uplift of certain areas thereby facilitating the removal from them by erosion of the weak younger beds. In the main central area of Tertiary strata the succession is more complete than in Bombay Basin. As Ingaia Stream is followed from the western lowlands into the upland block studied, it first passes through a basaltic barrier with ejecta on its southern bank and lava on its northern. Very soon, however, fresh-water shales with the siderite concretions common locally in such beds and the usual plant-impressions, which are often wonderfully preserved in the concretions, appear in the stream bed, dipping very gently to the west.

After a few hundreds of yards, however, marine sandstones about 150ft. thick outcrop with a dip of 15° to the south and continue for about 200yds.; in them there are imperfect molluscan shells and occasional shark's teeth and rounded fragments of retinite. They are soon succeeded, apparently conformably, by puce-coloured or greyish foraminiferal shales similar to those of Bombay Basin and fairly well exposed on the south wall of the valley. After a short interval, fresh-water shales reappear upstream dipping south-west at about 5° and here in their true stratigraphic position, and not, as must be the case for their more westerly equivalents downstream, displaced in level by some dislocation. Such displacement is likely to be a consequence of the fault clearly demonstrated and mapped further north as separating the marine from fresh-water members of the Tertiary succession. The full sequence of the beds at mid-Ingaia bridge may now be given, in spite of some repetition of facts already mentioned. The thin basal breccia-conglomerate described earlier, which rests on a steeply plunging surface of greywacke, contains angular blocks of rock as much as 2ft. or more in diameter intermixed with finer material. Its uppermost portion shows a greater proportion of water-rolled pebbles than other parts and is covered by soft argillaceous sandstones with imperfect plant-impressions; these are followed by a 2ft. layer of fairly resistant sandstone with abundant small concretions of marcasite, and these in turn by fissile shales with abundant good carbonaceous imprints of dicotyledonous leaves. Away from the steeply inclined basal conglomerate, which is well exposed adjacent to the road in a tiny rill on the north side of Ingaia Stream, the dip of the series speedily flattens to about 10° or 12° to the north-west. As the road is followed north from the bridge, the fresh-water shales can be traced up to a height of 500ft., their thickness being about 100ft., and they then appear to be overlain by a few feet of glauconitic sandstone, with occasional casts of marine molluscs, which appears in some slipped blocks at the road. The next bed consists of about 100ft. of fine-grained greyish foraminiferal shales with rare molluscan shells. Towards their upper limit these shales become glauconitic and are succeeded at an elevation of 610ft. by poorly consolidated greyish-black somewhat glauconitic sandstones containing occasional thin layers of shale interbedded with them and unfossiliferous, but for a few imperfect casts of molluscan shells. This sandstone is widely exposed at higher levels north of Ingaia Stream and gives rise to highly characteristic breakaway scarps at times as much as 30ft. in height. Attention may be drawn to the noteworthy thinning eastward of the marine sandstone that lies beneath the foraminiferal shales. It is 150ft. thick in the lower reaches of Ingaia Stream and not more than a few feet deep at the road north of the bridge, whilst south of this latter it appears to be absent, for the marine shales there seem directly to overlie the fresh-water ones at an elevation of 470ft. It is to be noted also that the foraminiferal shales are one of the most persistent beds of the area, for they occur not only in the

Ingaia Creek section, but also in Bombay Basin and on Ramarama-Ararimu Road a few hundreds of yards east of the line of the fault that separates the marine from the fresh-water facies of Tertiary beds. Before passing from the Ingaia Stream section, mention should be made of a small isolated outcrop of the white fresh-water sandstone at an elevation of about 550ft. in a small gully about a mile, or less, north-east from the bridge. It shows layers of sandstone about 18in in depth interbedded with more argillaceous layers of the same thickness, and must lie very close to the greywacke basement. Not far distant the uppermost marine sandstones outcrop at little higher elevation, so that, when due consideration is given to the general gentle dip of the series, the mutual relations of the two sets of beds are somewhat corroborative of the suggestion made earlier that the greywacke-Tertiary junction on the west of Dome Hill lies near a fault-line. White quartz-sandstone similar to that just described is exposed for a depth of 20ft. in a small quarry exposed on the scarp of the fault shown on the map as dividing marine Tertiary beds from fresh-water ones. It dips gently to the west and is in layers 2ft. thick alternating with slightly thinner argillaceous ones. A little south of the quarry, greywacke can be detected high up on the scarp a relatively few feet below the level of the floor of the quarry, but, about a hundred yards north of this latter, the sandstone outcrops well below this level, so that its thickness is evidently dependent on the rise or fall of the irregular basement of greywacke on which it has been deposited, and must reach a maximum of about 80ft. West of it fresh-water shales with plant-remains appear in road cuttings at intervals, forming the summit of the high ridge followed by Ramarama Road until superseded by basaltic débris. Similar beds outcrop south of this road in the westward-flowing stream which roughly follows the contact between greywacke and Tertiary strata and have given rise, near where the tongue of basalt marks the debouchure of this stream, to concretions of impure siderite scattered on the stream bed, which, with blocks of the shale, have yielded some excellent valves of large fresh-water molluscs in addition to plants. The molluscs have not been submitted to an expert for identification, though Professor Berry, of John Hopkins University, states in a letter that one closely resembles Anodonta.* Hochstetter (1864) records the occurrence of this genus in similar beds near Drury. In the small Tertiary area north of Whisky Creek, the fresh-water shales are well exposed in the bed of the small stream draining the cup-like hollow there present, and dip west at about 20°, their attitude thus being in conformity with the existence of a fault determining their western boundary, as has been suggested in an earlier section. North of the area studied similar shales are locally conspicuous, especially near the western margin of the upland block constituting what may be called the Hunua Hills structural unit, and in places contain workable coals.

(c) Basaltic Lavas and Ejecta (? Pliocene and Pleistocene). A number of centres of basaltic eruption occur either actually along or near the line of Papakura-Drury Fault from Papakura south, two at least coming within the area now described. The most northerly is near the debouchure of Whisky Creek, and is best shown on the north side of the creek by a lava-flow 150ft. or more in depth opened up in a quarry, 200ft. above the level of the adjacent lowlands, which shows remarkably good columns over 80ft. in length. In the small gorge of Whisky Creek nearby there are bedded deposits resting on Tertiary beds and containing pebbles of greywacke intermixed with far more abundant angular fragments of basaltic ejecta; these deposits evidently largely represent stream-assorted eruptive material. On the south side of the creek there is a cone-like mass built of tuff and consolidated lapilli; a great part of the original cone has, however, been removed by erosion. From this centre of eruption probably came also the small isolated tongue of basalt that outcrops a little over half a mile further south. Basaltic tuffs and coarser ejecta exposed in a fairly narrow gorge near where Ingaia Stream leaves its Tertiary terrain, and columnar lava nearby on the northern side of the valley, are possibly the products of a small local eruption rather than those of the major Bombay Hill centre, which has given rise in the main to flows. Puketutu (1229ft.) is capped by 100ft. or more of basaltic lava, whilst adjacent to it there is the great dome-like mass of Bombay Hill (1000ft.) from which, and possibly other sources south-east of it, considerable floods of basalt have poured westward across the line of Papakura-Drury Fault, completely masking its scarp and burying the underlying terrain whether Tertiary or Mesozoic. It appears to be unlikely that the cap of Puketutu was derived from the same source as gave rise to the Bombay Hill lavas; the date of eruption, however, certainly is so distant from the present that it would be impossible to discuss with any prospect of finality the explanation of this high-level lava-remnant as also of that on the top of Trig. 619 (886ft.). Many considerations make it unlikely that this last cap emanated from the cone-like basaltic mass a little over a mile west of it. This latter does not rise to so great a height, and its lava streams found their way south down the Paraureroa Valley and to-day form a mesa-like divide between two head-water tributaries of Paraureroa Stream. The only other eruptive foci not yet mentioned include a small one marked by coarse ejecta about 5 miles east of Bombay on the south side of Bombay-Paparata Road and another, indicated by ash deposits, about a mile west of Trig. 616. This last probably was the source of lavas which poured south down an adjacent stream valley on to the flats near Pokeno, where they left a fan-like accumulation of lava. A thin sheet of weathered volcanic ash, recognisable in a few road-cuttings near Paparata, has undoubtedly been derived from one or other of the adjacent eruptive centres.

Fig 1.—Diagrammatic representation of topography of area between Firth of Thames and Manukau lowlands.

Fig. 2.—Tertiary basin north of Whisky Creek, Ramarama. Tertiary fresh-water shales underlie grassed slopes bordered by greywacke which is capped in middle background by basalt. Drainage gap on left, with probable resequent fault-line scarp beyond it. Fig. 3.—Resequent fault-line scarp 5 ½ miles east of Ramarama, on south side of Ararimu Road. Fig. 4.—Fault-splinter of Wairoa Fault near boat-hock bend of Wauoa River north-west of Paparimu. Fig. 5.—Happy Valley from basalt barrier at its south-west end. The outlet to Paraureroa Stream is to the right in middle distance. On the left a greywacke hill capped by basalt (Trig. 619).

Fig. 6.—The valley by which Happy Valley drainage joins Paraureroa Stream, which trends from right to left on the near side of distant lofty hills (greywacke). Foreground and bushed terrace across stream, with flat-topped hill on right, are basalt; steeper slopes on left are greywacke. Fig. 7.—Outlet gorge of Paraureroa Stream, looking south. High greywacke hills in distance; basalt platforms astride of stream in middle distance. Fig. 8—Air-Gap Valley from the south. Watei-paiting on broad flat in middle distance. High hills on sky-line form scarp of Wairoa Fault, Note, just below them, the regularity of the sloping greywacke surface (? a resurrected surface). Fig. 9.—Head of Air-Gap Stream (to right, below summit of hill); swampy floor of its valley in foregound. Alluvial terrace on left.

Geological sketch-map of bombay-happy valley area.