The Wellington Fault

Transactions and Proceedings of the Royal Society of New Zealand, Volume 85, 1957-58, Page 633

The Wellington Fault By G. R. Stevens Victoria University of Wellington [Received by the Editor, July 26, 1957] Abstract The Wellington fault is not a plane of movement but a wide ciush zone which passes laterally into a complex zone of anastamosing faults and then into a more open system of minor faults. The fault, which is active and markedly transcurrent, already has a long history of movement and since the beginning of the Pleistocene has probably dislocated Trias-Jura greywackes some 2,000 feet. Recent alluvium is now juxtaposed by overlap and fault contact against the old rocks, but since the latter bear no recognized stratigraphic datums and the former cover so small a time span it has been difficult to assess the patterns, magnitudes, or even the existence of large faults by conventional methods. Geomorphic studies described below allow the recognition of an ancient erosional surface and by using this surface as a datum recent fault movements within the greywackes have been charted. Furthermore it has been possible to study the distribution and map the pattern of ancient faults. Minor as well as major faults of the Wellington Fault System are commonly marked by zones of intense crushing, and these zones are deeply etched by erosion. Thus in the greywackes, which otherwise behave as homogeneous to erosion, faults can be recognized and a distinctive fault pattern has been mapped. Topographic Expression and Historical Notes The Wellington landscape is dominated by two landforms, one tectonic—the Wellington Fault scarp, and one erosional—the Kaukau Surface. The Kaukau Surface (Cotton, 1912: 249; 1955: 14), a senile erosional surface of late Tertiary age, has been ruptured by Pliocene and Pleistocene faulting, and it is the dislocation of this surface which in the first place defines the Wellington Fault. The Wellington Fault downthrows to the south-east and its scarp forms the western margin of the Port Nicholson and Hutt Valley depression (Pl. 43). The general trend of the scarp is north east-south west and the trace of the probable underlying fault is roughly marked by a very prominent line of bluffs, and at the south-west end by fault valleys which run to the southern coast of the Wellington Peninsula at the mouth of the Karori Stream. McKay, in a map published in 1892, indicated the Wellington Fault and showed the Hutt River as being controlled to its headwaters by this fault. He correlated the North Island faults with the then known South Island faults, but correlation is still somewhat in doubt (cf. Cotton, 1954). Bell (1910: 536) recognized the Wellington Fault but thought of the Hutt Valley as a complex graben. Cotton (1912) also described the Wellington Fault, and in 1914 recognized that the Silver Stream, Upper Kaiwarra Stream and Tinakori Stream are aligned along the Wellington Fault shatter belt. Cotton (1914) also noted that the Hutt Valley continues the line of the Wellington Fault north-eastwards, and in 1921 presented his “downwarping” theory to account for the formation of the Port Nicholson—Lower Hutt Basin. Cotton was the first to propose that the basin occupied a fault-angle depression (1921), and he (1951) was also the first to show that the Wellington Fault had an important strike-slip component. The Wellington Fault is now generally accepted as a transcurrent fault with a variable downthrow to the south-east. In the writer's opinion the fault is not simply definable, for it follows an almost mile-wide zone of intensely crushed rock. It is convenient in dis-

cussion to speak of the central part of this zone as the Wellington Fault and of the zone as the Wellington Fault Zone. Although some workers may speak of the fault plane as being exposed in given positions there is nowhere along the fault line any considerable exposure of what might be called the main fault. Instead there are many faults which are sub-parallel to the fault line, and the characters of these may reasonably be attributed to the Wellington Fault as a whole. The Geomorphic Mapping of Faults Most of the evidence for the existence of the Wellington Fault and its associated minor faults is found in the geomorphic features of the surrounding country. The customary tools used in the mapping of faults are lithology and stratigraphy, but in the environment of the very broken, unfossiliferous and uniform greywacke facies of sediments which form the bulk of the Wellington Peninsula, such aids are unavailing and only certain recent deposits contribute to the elucidation of structure. While the use of geomorphology as a tool in the mapping of faults appears at first sight to be a crude device compared with a well defined stratigraphical column there is no question as to its adequacy. There are, for example, few geologists who would fail to recognize the Wellington Fault in Pl. 43, although there may be some who would not appreciate the geomorphic nature of the evidence presented there. An examination of Pl. 43 suggests that a block of resistant rock has been juxtaposed against softer and that is perhaps sufficient to indicate the presence of the fault. That analysis is crude, but by applying and refining the geomorphic tool it is possible to define the Wellington Fault more closely, to determine the amount of throw and sense of lateral displacement on faults, and to recognize small features such as minor faults. Thus in this area where stratigraphic methods have hitherto failed it is possible to map a general fault pattern.Many major faults may be mapped with precision, and it is found that in certain respects geomorphic criteria are superior to stratigraphic. Procedure Photographs of the N.Z. Aerial Mapping Ltd. series (4in to the mile) were used as field sheets, and in the area treated in detail field information was plotted on the based maps (1: 15840). Shatter belts, dislocated Kaukau Surface remnants and alluvial gravels, crest-line and spur jogs and notches were mapped in the field. Trains of minor features as well as straight stream courses have also been mapped in the field as “lineations”. On aerial photographs innumerable additional lineations which are not obvious on the ground can be observed, and these, along with the field observations, conform to the general pattern of Figs. 1 and 2. Criteria of Faulting Miller (1941) working in a similar environment of transcurrent faulting in California was faced with the same problem as the writer, namely, fault recognition in a homogeneous rock terrain lacking cover beds, but possessing a widely developed and dislocated erosion surface. Miller wrote: “When it comes to a consideration of the number of faults in southern California and the evidence for their existence, geologists are more or less divided into two groups. In one group are those who believe that a fault should not be mapped unless the actual fault can be seen nearly everywhere or unless there is positive stratigraphic evidence for it. Many of these workers are not very familiar with fault problems in crystalline rocks. In another group are those who recognize much more faulting on the basis of still other criteria, particularly in areas of crystalline rocks.” (p. 87.) Criteria for the location of faults are given by Miller (1941: 89–97) and the subject of tectonic features is discussed by Cotton (1950) In the following pages the critern

of these authors will be discussed in so far as they apply to the Wellington area. Miller's comment (p. 89) that the criteria are of varying value is still true; a single criterion may be conclusive evidence of faulting, another may be of little value by itself. 1. Fault fractures clearly visible. In the Hutt Valley there is to the writer's knowledge only one instance where offset at a fault plane is visible. This is on Haywards Road, where the Haywards Gravel Formation (Stevens, 1956) has been set against greywacke and a sharp reverse fault produced. In other areas actual fault planes are unrecognizable owing to lack of exposures and prevalence of shattering in the rocks. Miller reported similar problems in southern California (1941: 89). 2. Vertically displaced strata. This criterion can be applied to a few cases where alluvial gravels straddle the Wellington Fault Zone but elsewhere it has no application because of the lack of cover beds and the homogeneity of the terrain. Further, and so far as any but recent faults are concerned, fault-line erosion adjacent to faults has obliterated most primary fault features, including evidence of offsetting of superficial deposits. 3. Recent fault scarplets. These structures, which dislocate the present-day erosional or depositional land surface, are comparable to faults cutting unconformities or bedding. In this context “fault scarplet” is preferable to “fault trace”, as “trace” is a general (geometric) term which should refer to the intersection of a fault plane with any kind of surface, including bedding or fault planes as well as the topographic surface. Recent fault scarplets along the line of the Wellington Fault occur at Emerald Hill (Fig. 2) and southern Wellington. 4. Vertically displaced erosion surfaces. Miller (pp. 90–91) found this criterion to be of the utmost value in the determination of faulting in southern California. He states (p. 90): “. relatively long, straight scarps, separating such old erosion surfaces sharply at different levels, are almost as good as positive stratigraphic evidence of faulting.” Identification of dislocated remnants of an old erosion surface (the Kaukau Surface) provides the key to fault identification in the Hutt Valley. These remnants are readily distinguished by their senile topography and the great depth of weathering present on their surfaces, so that positions and throws of many faults are well defined. In the special circumstances of unfossiliferous greywacke country and Pleistocene faulting, geomorphological mapping of faults is superior to the more customary procedures of fault determination. Conventional methods deal with isolated points in space (i.e, outcrops) and yet allow the joining of such points to represent bedding or fault planes. On the other hand, geomorphological procedures deal with dislocations of a visible landscape surface which can be identified over much of its extent.In the Hutt district the Kaukau Surface is well defined and of considerable area, and dislocation of this surface is revealed by the “stepped” topography on either side of the Wellington Fault and frequently by the presence between “steps” of strong fault lineations. 5. Horizontal displacements. The greatest disadvantage of mapping faults in terms of surfaces is that the method can only be applied to relatively recent faulting. This does, however, aid the study of faulting processes, for it enables the selection of faults of only one period and also allows the study of rates of earth movement. Further advantages lie in the

fact that transcurrent movements of only a few yards can be readily detected and in so far as the surface of reference, the earth's surface, is everywhere exposed, most of the faults over an area can be examined. Allowing for the restriction to faults of one age, the geomorphic method is the most certain of widely applicable methods of measuring transcurrent movement. In areas adjacent to that studied here, horizontal offsets have been recorded by Cotton (1950, 1951) and Adkin (1954), and it is generally accepted that the Wellington Fault and several other faults in the Wellington area are transcurrent. 6. Shatter Zones. In the greywacke terrain of the Wellington Peninsula, erosion is controlled and lineations determined by intimate emposed weaknesses rather than by inherent sedimentary variations in the strength of strata. Despite the segregation of argillites and arenites into rhythmic beds in the Wellington greywackes, the succession is on the large scale homogeneous and of uniform lithology and variation of strength across bedding is slight. Thus the controlling effect of bedding on erosion is much reduced and the landscapes may be described as tending towards isotropy to erosive processes. Furthering this tendency is the intense shattering which pervades almost all the greywacke sandstones and renders the hardest bands as weak to attrition as the softest Still further, in the young topography of Wellington Province the steep gradients favour the transport of large rock particles and allow erosion of shattered hard beds as well as soft. Finally it seems that transcurrent faulting imposes intense local shearing stresses on the rocks so that faults are marked by zones of extreme comminution. In a broadly isotropic environment this cataclastic effect is directly reflected by erosional fault-line features such as stream lineations and notches in crests and spurs. 7.Alignment of saddles. In a full account of geomorphic expressions of faulting, Cotton (1950: 739) maintains: “… jogs and notches, according to the direction of down-throw, remain in the landscape after erosion has eliminated all geomorphic traces of faulting in adjacent valleys; on resistant terrains they may survive into full or late maturity of dissection, whatever the prefaulting form of the surface or its postfaulting history.” To the west of the Wellington Fault notched spurs and aligned saddles are very well developed. Such features are fault-line features sensu stricto, for an “etching-out” of a fault pattern by subsequent erosion has occurred.Faults distinguished by such etching characters may be of any age and should not be confused with recent faults which are generally identified as dislocations of topographic surfaces. 8. Stream lineations. Commonly it is only the river pattern of a greywacke terrain which gives any direct indication of a fault pattern. Valleys bear little relation to the strike of inclined or folded strata and in many places rivers follow zones of obviously fault-shattered rock. The straight stretch of the Tauwharenikau River, Tararua Mountains (N. Z. M.S. 1, N161), provides a good example of deep etching of a shatter zone. Here recent fault traces and fault-controlled saddles have been recognzed to the north-east and south-west of the stream lineation and also in the river valley. Though on a smaller scale, similar stream lineations occur to the west of the Wellington Fault and in particular along the line of the Liverton Fault. In the latter examples shatter belts mapped in the field line up with stream lineations, spur or ridge notches and jogs, and similar features.(Pl. 44, Figs. 1, 2, 3.) Cotton (1950: 746) has discussed the theory of river guidance by “shatter belts” in these words: “In New Zealand the fault-line erosional hypothesis has been called upon to explain alignment of some valleys near Wellington with faults and crush zones.… and generally it has been assumed that these faults zones are ancient.”

Fig. 1 (right).—Fault map of the Hutt Valley. The “Orongorongo Shatter Zone” with associated faults is after Grant-Taylor (1949). Fig. 2 (left).—The Wellington Fault zone north-east from the Upper Hutt Basin. The continuation of the Kaitoke Fault north-east from the Pakuratahi River is after G. J. Lensen.

Cotton points out, however, that some of the fault-valley features, like those along the south-west portion of the Wellington Fault, are primary and of recent age. Allowing for such recent faults, it is still considered that in the area dealt with here the greatest part of the fault pattern is ancient and originated by etching-out along old fault lines. To complicate matters, it seems probable that in many instances recent faulting has followed along earlier-formed faults and so primary and secondary fault features coincide. 9. Zones of excessive jointing with slickensides. On both sides of the Wellington Fault Zone secondary quartz is often present along slickensided joints. Zones of fractures are the extension away from the fault of the fault shatter zones and where fault-line breccias and pugs have been removed from the centre of fault zones by erosion joints provide useful indicators of the presence and trend of the major fault and suggest its hade. Features of the Wellington Fault Zone Recent movement as shown by scarplets Between Wellington City and Upper Hutt the Wellington Fault lies beneath the waters of Port Nicholson and the recent alluvium of the Hutt Valley, and in this area no recent fault traces have been found To the north-east of Upper Hutt a recent fault trace, the Kaitoke Fault (Hall, 1946: 428) has dislocated terraces of the Hutt River.This terrace dislocation has been mapped by H. E. Fyfe and M. T. Te Punga (pers. comm.) at N161. 645456—to the north of Emerald Hill and at N161 626445—to the south of Emerald Hill (Fig. 2). These two workers have found that the amounts of vertical and lateral displacement increase with the age of the displaced terraces and indicate that two lateral movements of 33 feet and 18 feet have occurred with a 1ft to 2ft vertical movement accompanying the former. Considering the Wellington Fault as a zone, movement may be regarded as occurring on different fault planes at different times so that at any one time a single fault within the fault zone may become active.The Kaitoke Fault is a case in point, for this branch has recently been more active than other faults of the system, and its scarp presents an important feature in the topography. The north-eastern continuation of the Wellington Fault zone has been traced by McKenzie (1929) and G. J. Lensen and G. W. Grindley (pers. comm), the latter having found fault traces passing through the low divide between the Pakuratahi and Tauwharenikau Valleys. Breccias of the fault zone At several places along the Wellington Fault Zone brecciation of the greywacke country rock is extreme.In such places it is exceedingly difficult to find an unsheared rock fragment even six inches across within 400 yards of the centre of the zone. At Haywards Road (N160, 520390), half a mile from the Wellington Fault scarp, the alternating sandstone and argillite beds of the greywacke succession, though still retaining their individuality, have been intensely shattered (Pl. 45, Fig. 5). The angle of shattering varies in each bed, and it may be inferred that a large though indeterminate amount of bedding-plane slip has occurred. On the other hand, the greywacke beds exposed at Raroa Road (N164, 318217) astride the Wellington Fault, have been so completely broken that individual strata are unrecognizable. Plate 45, Figs. 1–4, constitutes a series of photographs taken along Raroa Road to typify the degrees of crushing which occur for 400 yards either side of the Wellington Fault. Fig. 1 illustrates a stage of shattering next after that in the Haywards area. The greywacke is intensely shattered and stratification has been completely destroyed. In Fig. 2 the shattered greywacke is overlain by

View south-west along the line of the Wellington Fault Note the “jog” at the mouth of the Korokoro Stream and, north-west of the fault scarp, the Haywards Gravels (flattened areas in foreground) and fault-controlled stream lineations.

Fig.1—Fault-line valley and notched spur developed to the south-west of Hill Road.Belmont(foreground) Fig. 2.—A fault-line valley developed at grid ref.N160. 458327 (west from Pomare Road) along a probable tensional fault.A spur jog maiking a subsidiary fault is also shown. Fig. 3.—Fault-line valley north-east from Hill Road, Belmont (grid ref.N160, 460369). Note the dislocation of the Kaukau Surface, downfaulted to the left (north-west)

Figs. 1–4—Raroa Road Wellington Shattering of gicvwackc close to the Wellington Fault.Note the soliflual debus overlying the shattred greywacke in Fig. 2 The soliflual debris/ greywacke contact is indicated by the arrows Figs. 1.2 width of photo about 4ft Fig. 3 width of photo about 10ft Fig. 4 width of photo about 15ft Fig. 5—Shattered greywacke of the Wellington Fault zone exposed at Haywards Lower Hutt A gorse bush (upper centre) and a cycle (lower centre) indicate scale. All the photograph on the plate are taken looking south—ie. with west to the right.

View north-east from Taita Gorge along the line of the Wellington Fault Note the “jog” at the mouth of the Wakatiket River.

soliflual debris, with a “shaved surface” developed at the contact. There is little difference between the angular soliflual debris and the fault-shattered rock. The next stage in the shattering of the greywacke is the development in the shattered rock of anastomosing zones of even more intense shattering (Fig. 3). In these zones, usually between 3 and 6 inches wide, extreme comminution of the greywacke occurs and fault pug appears—the greywacke having been pulverized to rock flour. Often the small shatter zones shown in Fig. 3 coalesce into a large shatter zone, some 8ft wide, as shown in Fig. 4. The broken pieces of material towards the base of the shatter zone in Fig. 4 are not of solid rock, but are shattered pieces of lithified fault breccia and fault pug. The attenuation of the pieces should be noted. Extent of the Wellington Fault Cotton (1914; 296) was of the opinion that the Hutt River is fault-controlled to its source probably by a continuation of the Wellington Fault. This view is shared by the writer, for the Eastern Hutt River presents a marked lineation (N.S.M.S. 1. N161), and it seems probable that the river flows along a crush zone directly aligned with the Wellington Fault south of Upper Hutt. This does not mean that the Eastern Hutt River Fault is the only continuation of the Wellington Fault, for it is eminently possible for there to be several branches of continuations each at different times. As the fault south of Upper Hutt lies in a broad crush zone it is necessarily difficult to say where individual branches part from the main fault, and because of the complexity of the faulting it must be difficult to single out any branch as the most important continuation. North of Upper Hutt the Fault is poorly defined, but a number of faults can be traced to the north-east by recent fault scarplets. Since the southern part of the Wellington Fault is known to have been active recently these scarplets are taken as evidence of the main continuation of the Wellington Fault. Using this criterion of recency Lensen (Lensen et. al., 1956, p. 131) has traced a continuation of the Wellington Fault into the Tauwharenikau Valley; from there has mapped fault line features as far as Lake Waikaremoana, 180 miles to the north-east, and the whole length of this he names the Eastern Tuara Fault. Transcurrent Nature Cotton (1951) described a fault cicatrice associated with shutter ridges on the south-west portion of the line of the Wellington Fault and estimated a recent dextral transcurrent movement amounting to approximately 200ft.Following up this work, C. M. Laing and G. J. Lensen (N.Z. Geological Survey—pers. comm.) have determined the horizontal displacement of the two shutter ridges present at N164, 268154 as amounting to 210ft and 270ft, and concluded that the Wellington Fault is transcurrent with a relatively recent dextral movement of approximately 250ft and is perhaps reverse. Throw Warping on axes normal to the fault has produced along the eastern side of the Wellington Fault zone a series of basins (see Stevens, 1956) so that vertical displacements must vary along the length of the fault (see Cotton, 1914: 297, Fig. 2; cf. Cotton 1947: 371, Fig. 3). The observed vertical shift across the fault zone is of the order of 2,000ft, but this is distributed over many faults and the maximum throw in the centre of the zone and across the major fault scarp is only of the order of 500–700ft. Along the eastern side of the fault bore holes in Recent sediments in the Lower Hutt—Port Nicholson Basin indicate the depth of the down-faulted greywacke on that side of the fault scarp. At Taita Gorge, the northern boundary of the basin, greywacke lies at only 18ft below the surface, and the throw may be as low as 100–150ft. Farther south in the basin, the Wellington City Council test bore at Wilford (N164/3, 444300)

encountered weathered basement at a depth of 396ft (Stevens, 1956) and as basement greywackes are exposed across the fault at + 100ft, there must be at least 450ft of throw. As the test bore was sited east of the deepest part of the fault-angle depression, 600ft or 750ft of throw probably occurs at Petone and evidence of this kind (Stevens, 1956) extrapolated southwards suggests that throw is at least 1,000ft in the deepest part of the basin (i e., in Port Nicholson). Direction of Hade It is difficult for various reasons to determine the direction of hade of the Wellington Fault. Recent fault movements in basement rocks have a superficial and not a direct expression in the soft alluvial and detrital cover and faces of scarplets and fault traces must rarely represent the attitudes, positions or character of deep seated fault planes. Even where erosion is deep and one might normally expect to detect the inclination of fault planes this is hardly possible because of the intense shattering. Fault lines are frequently obscured by scree from fault or fault-line scarps and even where fault planes are seen their attitudes can have little significance. It may not, for example, be assumed that the normal character of minor faults in the scree or sedimentary cover over a major fault or even in a broad shatter zone of greywacke means that the fault is normal in depth. All such expressions of the fault are in such soft rock that they could well be gravity faults along the fault scarp. Subject to these qualifications, the Raroa Road section, which provides one of the few exposures across the line of Wellington Fault, may be examined. Here the fault zone consists of bands, 1ft to 5ft wide, of fault breccia and the greywacke is very shattered for a considerable distance on either side (Pl. 45). The dips of bands of fault breccia usually approach the vertical but give slight indications that the fault is reverse. Many small faults are exposed in road cuttings from Hutt Road to Korokoro and Belmont, and over all the same reverse tendency of shear planes is shown. The lack of sinuosity of the trace of the Wellington Fault as it intersects the rugged terrain might be taken to indicate that the fault plane is vertical. This need not be so, for the crush zone of the fault is so wide that minor features are obscured and the straightness of the fault can only be observed as a general and large scale phenomenon. There is some suggestion (Anderson, 1942: 55, and Wellman, pers. comm.) that transcurrent faults are straight and usually approach the vertical, and the evidence from the Wellington Fault is not inconsistent with these views. Adkin (1954), on the basis of sections exposed in a tunnel under Thorndon and Wadestown and in a quarry at Grant Road, concluded that the Wellington Fault is normal with a hade of 53° to the south-east (Adkin, 1954, Fig. 5). The evidence for this may, however, have an alternative interpretation, for it is probable that the tunnel section is cut in an old marine-cut cliff and that the “sediment”, “upwarped” against the “fault”, might be soliflual debris and the whole structure a congelifractate slope developed from the marine cliff. Such congelifractate slopes must have developed to a sea level appreciably lower than that of the present day and so could appear at the tunnel level.Congelifractate debris may readily be taken to be dragged strata and the shaved surface below often resembles a fault plane (Stevens, 1957a). This view is supported by the fact that congelifractate material was encountered in test bores for the Wellington Anglican cathedral (N164, 337233), in the Thorndon area, about 400 yards from the base of the cliffs. This material was a stratified deposit and evidently had been transported by small streams from its original position at the base of the marine cliffs. Previous workers in the Hutt Valley—Wellington area, e.g; Cotton (1949, Fig. 176) and Hall (1946) seem to have regarded the Wellington Fault scarp as a primary feature and to consider that the spur facets developed along the scarp represent the actual fault plane This would imply that the fault dips to the downthrow side

Fig. 3.—Restored contour pattern, grid square N160, 390450. Note the effect of the Liverton Fault on the landscape; also how the contours illustrate the retrograded nature of the Liverton Fault scarp.

and is therefore normal. Cotton (1950, p. 737) now considers, and the writer agrees, that these facets bear little relation to the original declivity of the fault scarp. The trace of the principal fault almost certainly lies from 200 to 300 yards in front of the present-day scarp and is covered by the waters of Port Nicholson and the alluvium of the Lower Hutt and Upper Hutt Basins. Erosional Features Acceptance of the thesis that the scarp bounding the western side of Port Nicholson and the Lower Hutt and Upper Hutt Basins is a retrograded fault scarp explains many apparent geomorphic anomalies, such as the development of “jogs” (off-sets in plan) along the scarp line. These features may be interpreted as being of several origins: (i) Offsetting of the principal fault by faults normal to it. (ii) Side stepping—i.e., the offsets may be the expression of an en-echelon series of faults comprising the Wellington Fault Zone. (iii) Differential retreat of the fault scarp. The third or erosional hypothesis is here preferred. An example of differential retreat is displayed on either side of the junction of the Wakatikei River with the Hutt River (Pl. 46). South of this point, the scarp is well developed and presents the fresh sharp-cut aspect of a recent fault scarp; this is deceptive, for in the Taita Gorge the continuation of this scarp is visibly erosional. North of the confluence the character of the scarp is quite different, for it has been dissected into numerous small spurs and in plan a prominent jog separates the two types of scarp. It appears that the scarp south of the confluence is largely erosional; the fault scarp to the north has undergone comparatively little retreat from the original fault-line and has been subjected only to light dissection by small local streams. The reason for this differential retreat is that to the north of the Wakatikei River the Hutt River is deflected towards the eastern side of the Upper Hutt Basin by the Akatarawa River* This river is gorged in the Akatarawa shatter zone, one of the splinter faults of the Wellington Fault, and it thus deflects the larger river (Hutt River) rather than being deflected itself. and does not impinge on the base of the fault scarp. The tendency is for this deflection to become permanent as the river entrenches, and it is only after its first great meander that the river swings round to join the Wakatikei and cut away the Wellington Fault scarp. A similar prominent “jog” at the mouth of the Korokoro Stream (Pl. 43) led Quennell (1938) to propose that a “Korokoro fault” had transcurrent movement normal to the Wellington Fault and that this had offset the main scarp along the harbour. Although the Korokoro fault undoubtedly exists, the jog is more readily attributable to normal erosive processes, and the hypothesis of transcurrent movement seems unnecessary.During the earlier stages of the Lower Hutt-Port Nicholson Basin it is probable that marine erosion was effective along the length of the Wellington Fault scarp, but with the formation of the Hutt delta, the northernmost portion of the scarp (from Korokoro north) was protected from all but river and subaerial erosion. As the tectonic depression of the Hutt Valley widens and deepens towards the harbour, forward growth of the Hutt delta must have fallen off abruptly and at some point the shoreline must have remained relatively stable. Other factors such as tidal sweep in the more open part of the harbour would tend to stabilize the shoreline in its present position. The fault scarp south of the delta (and of the mouth of the Korokoro Stream) must thus have been subject to differential marine erosion for a long period which only ended as a result of tectonic uplifts beginning about 1500 A.D. Since the 1855 uplift the declivity of the marine cliffs has been reduced by subaerial denudation, but the rapidity of marine erosion on the shattered rocks of the fault zone prior to 1855 is still indicated by the presence of numerous “hanging” valleys.

Submarine features Examination of an isobath map of Wellington Harbour (Oceanographic Institute, 1955. by permission of the Superintendent) reveals that the asymmetric profile of the fault angle depression continues below the harbour, the position of the fault trace being indicated by the closing together of the isobaths along the western side of the harbour. On the same map the position of the fault trace beneath the harbour may also be deduced from the intense contortions of the 1 to 8 fathom isobaths, which are taken to mean that there has been slumping of unconsolidated sediments across a low scarp. Subsurface features Bore-hole data provide further indications of the position and retrograded nature of the fault scarp. At the Fletcher Factory bore hole (N164/3, 423/300), 200 yards east of the scarp, solid greywacke was encountered at a depth of only 99ft. Similarly, at Iron Reconditioning Ltd. (N160, 435303), half a mile up-valley and 250 yards from the scarp base, a yellow clay, probably weathered basement, was encountered at a depth of 105ft. These depths to basement may be compared with the 420ft depth recorded from the W.C.C. test bore at Wilford, about 1,000 yards east of the scarp. A similar situation occurs at a number of other points. Grey-wacke outcrops in the river bed to the north of Liverton Road (N160, 502361), and at Moonshine Bridge (N161, 574421) greywacke was encountered 15ft to 20ft below the river bed.Thus along the western margin of the Hutt Valley the alluvial deposits (Hutt Formation, Stevens, 1956) form a thin veneer over the basement greywacke, which has been cut back 200–300 yards. SELECTED FAULTS The following descriptions are not exhaustive of all faults or any fault but are chosen to illustrate typical faults and typical kinds of evidence used in their identification. Only the most prominent faults have been named. Liverton Fault This fault illustrates the effects of vertical displacement of the later-Tertiary Kaukau Surface. In the map (Fig. 3) contours have been reconstructed by ignoring all the gullies cut in the old surface by rejuvenated streams. Such reconstructed contours dispense with the confusing detail of minor-valley dissection. It is apparent on inspection of the map that the topography (in this greywacke setting) is anomalous.The sudden change in the general altitude of the topographic surface is, however, readily explained by invoking a fault along the line of change.This fault may be extrapolated in both north and south directions to explain many similar landscape anomalies and finds confirmation in trains of crest-line and spur jogs and notches and stream lineations. The throw of the fault is observed from the map as probably of the order of 400ft, but there is no evidence as to hade or horizontal shift. Faults such as the Liverton Fault, trending sub-parallel to the Wellington Fault, are here termed subsidiary faults. Towards the Wellington Fault these faults rapidly increase in number, but the evidence for them is not unequivocal On the slopes rising to Belmont no less than five major steps are observable in the topography, but downhill and nearer to the fault where dissection has destroyed the old topography, evidence lies largely in aligned spur notches and stream courses. Shear zones and numerous minor faults with appropriate strikes are observable in valleys like the Ngauranga and in the deep cuttings of Haywards Road. which run across the fault zone. Korokoro Fault Quennell (1938) recognized and named the Korokoro Fault, and in order to explain an apparent offsetting of the Wellington Fault scarp assumed that trans-

current movement had occurred along it. It has been shown above that the offset at Korokoro is probably erosional in origin. In the writer's opinion the Korokoro fault curves into alignment with rather than intersects the Wellington Fault and so produces no offset (i.e., it is a splinter fault). This fault causes a slight vertical displacement of the Kaukau Surface. In the stream floor the crush zone can be observed at several points, and the surface is identifiable at close hand by its deep weathering and undulating topography. Adkin (1951, Fig. 4), on the basis of summit heights to either side of Korokoro Stream proposed that this stream form the north-east border of his “Port Nicholson-Pukerua Sunkland”, and there is no question that the Korokoro Fault downthrows to the south-west. Judging from the topography, the maximum throw on the Korokoro Fault is of the order of 400ft to 600ft. There is neither marked tilting of fault blocks nor sign of horizontal shift on this fault. Akatarawa Fault. The position of the Akatarawa River is controlled by the shatter zone of the Akatarawa Fault. The acute angular relation of the Akatarawa Valley to the Wellington fault line is shown by topographic map N.Z.M.S.1, N161 and the Akatarawa Fault may be regarded as a splinter of the Wellington Fault. Recent fault scarplets (G. I. Lensen, pers. comm.) at grid references N161, 638500, 645517 follow the trend of the now degraded fault, and additional support for fault control of the Akatarawa River is provided by dislocation of the Kaukau Surface along the line of the valley. When the broad profile of the Akatarawa Valley is viewed from the eastern side of the Kaitoke Basin it is obvious that the surface forming the north-eastern side of the valley has been tilted towards the river and the ancient surface to the south-west of the river has been tilted away. The river thus lies in the angle of a fault which downthrows to the north-east, and this throw is inferred from the topography to be of the order of 500ft. This inferred throw is a maximum value, the throw decreasing as the fault passes away from the Wellington Fault. Other Faults Faults resembling the Akatarawa in respect of fault-line valleys and dislocated surfaces are the Wakatikei, Speedy's Stream and Belmont faults.The Wakatikei River is aligned along the Wakatikei Fault and the north-east branch of Speedy's Stream along the Speedy's Stream Fault. The Speedy's Stream Fault is a splinter fault running from the Wellington Fault to the Liverton Fault and conspicuous notched spurs and stream lineations are present along its length. In the case of the Belmont Fault, the Belmont Stream probably follows the fault line for about the first half mile of its course and then leaves it, apparently to return once again about one mile from its source.Variations in the degree of shattering of the bedrock are clearly shown in the stream profiles (Stevens, 1957b). Quennell (1938) recognized several of the faults associated with the Wellington Fault; the “Normandale fault” (part of the writer's Liverton Fault) and “Quail fault”, which are two of the subsidiary faults, as well as the “Belmont” and “Hayward” faults. Hall (1946: 428, 429, Fig. 1) also considered that the Haywards Valley was probably controlled by an ancient shatter zone. In the broader picture Quennell envisaged several simple parallel faults continuing for the length of the western Hutt Valley; in contrast, the writer's view is that a wide fault zone occurs within which there is great complexity and an anastomosing pattern of minor faults.

Barbed Tributaries of the Hutt River Living to the east of the Hutt River the Silverstream, Mangaroa, Pinehaven and Stokes Valley Streams are all peculiar in that they join the river at an acute angle which points upstream. Such confluences have been called “barbed” by Cotton (1914, p. 296) who considers that they commonly have some special structural explanation. Two explanations could apply here. (1) Since the regional strike of the greywackes in this area is broadly north-south the tributaries may have been adjusted to the strike direction. In contrast the Hutt River is controlled by a fault which intersects the strike of beds at a low angle and it becomes possible for barbed junctions to develop.Against this thesis, the homogeneity which elsewhere characterises the greywackes appears to apply to the eastern valley-side also and marked rock control of minor streams appears to be lacking. (2) Alternatively, the eastern tributaries may be controlled by splinter faults passing out from the Wellington Fault. This view is strongly supported by the mapping by Grant-Taylor (1949) of a shatter zone in the Mangaroa Valley and by the presence of dislocated surfaces along the hills to the west of the Mangaroa Stream. The general lack of a pronounced pattern on the eastern slopes of the Hutt Valley is atributable to the gentler slopes and lighter dissection there and to the probability that the most intensely faulted part of the area east of the Wellington Fault is buried under alluvium. Those lineations which do occur parallel to the Wellington Fault become less numerous and weaker to the east.In general the shearing and jointing of the greywackes die out away from the fault, and it seems likely that splinter faults like the Mangaroa also die out in that direction. References Adkin, G. L., 1951. Geology of the Paekakariki area of the coastal lowland of western. Wellington Trans. Roy. Soc. N. Z., 79 (2): 157–176. —— 1954. Late Pleistocene and Recent faulting in the Otaki-Porirua and Dalefield-Waipoua districts of south Wellington. Trans. Roy. Soc. N. Z., 81 (4): 501–519. Anderson, E. M., 1942. The Dynamics of Faulting. Oliver and Boyd Ltd., Edinburgh. Bell, J. M., 1910. The physiography of Wellington Harbour. Trans. N. Z. Inst, 42: 534–540. Cotton, C. A., 1912. Notes on Wellington physiography. Trans. N.Z. Inst., 44: 245–265. —— 1914. Supplementary notes on Wellington physiography. Trans. N. Z. Inst, 46: 294–298. —— 1921. The warped land surface on the south-eastern side of the Port Nicholson depression, Wellington, N.Z. Trans. N.Z. Inst., 53: 131–143. —— 1947. The Alpine Fault of the South Island of New Zealand from the air. Trans. Roy. Soc. N.Z., 76 (3): 369–372. —— 1949. Geomorphology—an Introduction to the Study of Landforms. Fifth edition, Whitcombe and Tombs Ltd., Wellington. —— 1950. Tectonic scarps and fault valleys. Bull. Geol. Soc. Amer., 61: 717–750. —— 1951. Fault valleys and shutter ridges at Wellington. N. Z. Geographer, 7 (1): 62–68. —— 1954. Submergence in the lower Wairau Valley. N.Z. Jour. Sci Tech., 35b (4): 364–369. —— 1955. New Zealand Geomorphology. Reprints of Selected Papers, 1912–1925. N. Z. University Press. Davis, W. M., 1913. Fault-line scarps in central Sweden. Bull. Geol Soc. Amer, 45 518–520. Hall, W. E., 1946. Notes on the later geologic history of Wellington Peninsula. N. Z. Jour. Sci. Tech, 27b (6): 427–430. Lensen, G. J., Stevens, G. R., Wellman, H. W., 1956. The Earthquake risk in the Wellington District. N. Z. Sci. Rev, 14: 131–135. McKay, A, 1892. On the geology of Marlborough and south-east Nelson. N. Z. Geol. Surv. Rep. Geol. Explor, 21: 1–28.

McKenzie, D. W., 1929. The continuation of the Wellington Fault into the Hutt Valley. Unpublished M.Sc. thesis. Victoria University Library. Miller, W. J., 1941. Recognition of faults in southern California. Jour. Geol, 49: 87–100. Quennell, A. M., 1938. The physiography and structure of the Porirua district. Unpublished M.Sc. thesis, Victoria University Library. Stevens, G. R., 1956. The Stratigraphy of the Hutt Valley region, Wellington, N.Z. N. Z. Jour. Sci. Tech., 38 B. —— 1957a. Solifluxion in the Hutt Valley. N.Z. Jour. Sci. Tech., 38 B: 279–296. —— 1957b. Geomorphology of the Hutt Valley. N.Z. Jour. Sci Tech., 38 B: 297–387. Grant-Taylor, T., 1949. The Geology of Whiteman's Valley. Unpublished M.Sc. thesis, Victoria University Library. G. R. Stevens, New Zealand Geological Survey, Lower Hutt, N.Z.