The Geology of the West Coast from Abut Head to Milford Sound—Part 1.

Transactions and Proceedings of the Royal Society of New Zealand, Volume 71, 1942, Page 282

The Geology of the West Coast from Abut Head to Milford Sound—Part 1. By H. W. Wellman and R. W. Willett. [Read before the Wellington Branch, October 9, 1941; received by the Editor, November 10, 1941; issued separately, March, 1942.] Abstract. A map is presented showing the geology of the West Coast from Abut Head to Milford Sound. The geology is described, the Tertiary beds being discussed in greater detail than the older rocks. Attention is drawn to a major fault that extends along the west of the Southern Alps and to the direct and indirect effect of this fault on the physiography. A late Tertiary peneplain is considered to extend along the west side of the Alps and the relation between this peneplain and the present drainage is discussed. An interpretation of the geological history is presented, and it is considered that the Alps since the Hokonui orogeny have been twice base-levelled and re-elevated. The description of the Pleistocene deposits will be presented in a later paper. Introduction. The district described extends along the west coast of the South Island from Abut Head to Milford Sound and has an area of 2,000 sq. miles. The northern and southern parts are popular tourist resorts, world famed for forest and alpine scenery, but the middle part is less spectacular and has been far less visited. The field work, favoured by fine weather, occupied six weeks during August and September, 1941. The time available did not permit large scale mapping and only the Pleistocene and Tertiary rocks were observed in any detail. The following sources of information were used to supplement the authors' field observations in the preparation of the geological map which accompanies this account: Turner (1930b), Physiographic Features of the Lower Cascade Plateau; Healy (1938), The Geology of the Coastal Strip from Big Bay to Professor Creek; Benson and Holloway (1940), Notes on the Geography and Rocks of the Ranges between the Pyke and Matukituki Rivers. Some information was also obtained about the Paringa River and Bullock Creek areas from unpublished reports of prospecting carried out for the Labour Department by W. J. Bolitho and H. J. Evans. The alluvial and Pleistocene boundaries are largely those of C. E. Douglas and have been taken from a geological map compiled by him and now deposited in the Hokitika office of the Lands and Survey Department. The authors wish to express their thanks to the Director of the New Zealand Geological Survey for permission to publish this paper. Vegetation and Culture. On the West Coast, growth of vegetation is favoured by heavy, well distributed rainfall, associated with little wind or frost. Forest as impenetrable as tropical jungle covers much of the lowland, and a less dense growth clothes the mountain sides below the timber line. In the northern part of the area forest has been cleared from about

half the alluvial flats, but in the southern part where settlements are fewer only the more lightly forested river flats have been cleared and grassed. Total destruction of the forest is unlikely here, for the forest quickly advances on the man-made clearings to compete successfully with the introduced plants. This heavy forest cover makes geological observation difficult, practically confining rock outcrops to the coast, small streams, and mountain tops. Small topographic features are often masked by forest growth. The main south road which connects the settlements of Weheka, Waiho, Wataroa and Harihari with Hokitika has recently been completed to Paringa River, eight miles south of the Bruce Bay timber milling settlement. The road ends at Paringa River and communication is continued south to the Haast settlement by a pack track about forty miles in length. A road formed close to the coast connects this settlement with the scattered one at Okuru and is at present being continued east of Mount McLean to connect with the completed road between Jackson Bay and the junction of the Arawata and Jackson rivers. A pack track extends up the Jackson River and passes over the Martyr Saddle to give access to the lower Cascade Valley and to the coast at Barn Bay, the present southern limit of horse traffic. The old Barn Bay-Big Bay track is now overgrown and impassable for pack horses. Big Bay is connected with Martin Bay by a coastal pack track and with Marion on the Eglinton Valley-Milford Sound road by a track along the Pyke and Hollyford Rivers. Several tracks enable horse traffic to follow the main rivers toward their headwaters, but the Haast River track is the only one which gives access to the eastern side of the Alps. Within recent years air travel has brought the numerous airdromes within a few hours of Hokitika and has greatly reduced the district's previous isolation. Previous Geological Work. In 1863 Hector published an account of his geological expedition to the west coast of Otago, with a map showing the coast line from Awarua Point to Milford Sound. In this report, Hector noted the presence of Tertiary rocks in the Martin Bay-Big Bay area. The same writer in 1868 published some notes on the coast and harbours from Milford Sound to Hokitika, in which mention was made of the rocks observed along the coast line, special mention being made of morainic accumulations. Hacket (1869) examined the coast line from Cook River to Hokitika and accurately described many geomorphic and geological features, paying particular attention to the moraines. In 1877 McFarlane outlined the position of the Red Hills ultra-basic intrusion, and noted the presence of a conglomerate in the lower Cascade Valley and on the Cascade Plateau. In 1877 Cox and McKay examined the Tertiary rocks outcropping along the coast at Jackson Bay and Abbey Rocks. From Abbey Rocks they followed the coast northwards, making several trips up the rivers to the mountains. The results of these examinations were published in two papers by Cox (1877a, 1877b). The first deals

with the geology of the Westland district as far north as Reefton, and the second with the coal measures at Jackson Bay. Cox mapped the continuous belt of schist that extends along the western side of the Southern Alps, and correlated the less metamorphosed rocks to the west with the Maitai Series. Granite intrusions were shown extending as far south as Paringa River and lying to the west of the schist belt. The Tertiary rocks at Abbey Rocks were discussed, and the considerable deformation which they have suffered noted. Cox also mentioned the great accumulation of moraines formed by glaciers which have retreated and are now represented by the glaciers of the present day. In 1879 Haast summarised the result of many years' work in Canterbury and Westland, and published an excellent geological map of these areas. The older rocks of Westland were divided into three main formations, namely, from west to east:—Westland formation, Gneiss granite formation and the Waihoa formation. Further east the Alpine greywackes were classed as the Mount Torlesse formation. The Tertiary rocks that form a coastal strip in Westland were accurately mapped, as were the morainic accumulations that extend from Lake Brunner to Bruce Bay. Haast discussed the various formations in some detail, paying particular attention to the moraines. Park (1887) described and mapped the area extending from the head of Lake Wakatipu to Big Bay, noting the Tertiary rocks at Awarua Point and Long Reef Point. He considered the rocks of Sara Hills and McKenzie Range to belong to the Kakanui Series. Park also mapped the Te Anau Series, the ultra-basic rocks in the Olivine River and the Red Hills, and noted the presence of a patch of Te Anau rocks at the western end of the Sara Hills which has since been shown (Healy, 1938, p. 90b) to belong to the early Tertiary lamprophyre intrusions. Bell and Fraser (1906) and Morgan (1908) in the Hokitika and Mikonui districts respectively described in detail areas similar in many respects to South Westland, throwing light on the relations between the older rocks which continue for many miles to the south. The physiography of the lower Cascade Valley was discussed by Turner (1930a), who showed that the drainage pattern on the Cascade Plateau was due to control by lateral moraines. In 1930b and 1933 he published accounts of detailed petrological work on the metamorphic and intrusive rocks of South Westland and mapped the northern part of the Red Hills ultra-basic intrusive mass. Healy (1938) showed that the Tertiary beds previously mapped by Park at Awarua Point continued southward as a narrow strip to Madagascar Beach. He also discussed the older rocks and the physiography. In a paper discussing land forms in southern New Zealand, Benson (1935) published a map showing the geology and major fault lines of southern New Zealand. Among the faults mapped, he included a hypothetical fault extending along the Fiordland coast from Dusky Sound to Cascade Point, and a fault separating Palaeozoic schists from alluvium extending from Haast River to Arawata

River to join a fault along the north-east face of the Burmeister Tops. Benson and Holloway (1940) in a paper on the geography and rocks of north-west Otago published a generalised geological map of south-western New Zealand showing the Tertiary beds extending northwards from Madagascar Beach to a point halfway between the Gorge River and Cascade Point, together with the metamorphic zones of the Te Anau and lower Palaeozoic rocks. Physiography. 1. General. The West Coast from Abut Head to Milford Sound lies to the north-west of the highest part of the Southern Alps. Although the Alps are only from twenty to thirty miles from the sea, they do not slope evenly towards it, but are terminated by a major fault, the Alpine Fault. No extensive areas of high land comparable to the Alps occur on the north-west of this fault and the mountains, which rarely rise above 2,000 ft., are separated by wide areas of low relief. It is convenient to consider the Alpine Fault as forming the boundary of the Alps proper and the coastal strip which lies between this fault and the sea. The Mahitahi, Waita, and Arawata Rivers flow across this coastal strip and divide it into four districts, each with distinct geomorphic features. The most northerly of these districts extends northwards beyond the area described herein, nearly to the township of Ross. In it, morainic hills form most of the high land and older rocks rarely outcrop. Except for the lakes, of which Mapourika and Wahapo are the largest, the low land between the moraines is covered with recent gravel plains which slope seaward from the foot of the mountains. The Wataroa, Waitangi, Waiho, Cook, Karangarua, Makawhio and Mahitahi Rivers issue from glaciated valleys in the mountains and flow over the surface of the gravel plains to the sea. The largest of these rivers, the Wataroa, Waiho, Cook and Karangarua, are particularly heavily loaded with detritus and for most of the year they are discoloured by rock flour produced by the active glaciers in their headwaters. All the rivers mentioned above are aggrading their beds and spread widely over the gravel plains in numerous branches which often change in position and size after heavy floods. The rivers are not confined by definite banks, and, except in floods, are only a few feet deep, their rate of flow depending on the grade of the gravel plains which are steepest opposite the highest part of the Southern Alps, the more southerly rivers, the Mahitahi and Makawhio, being less rapid. Two smaller rivers, the Omoeroa and the glacier fed Waikukupa, both have headwaters in the Alps and flow in more confined courses between steep morainic hills. It is along the coast of this northern district that the morainic bluffs are best exposed. These bluffs are connected by even beaches behind which many of the rivers flow for some distance parallel to the coast before entering the sea.

In the next district to the south, that between the Waita and Mahitahi Rivers, the morainic hills are replaced by higher land composed of older rocks through which the Paringa and Moeraki (or Blue) Rivers flow in wide glaciated valleys. Lake Paringa lies between these rivers and occupies the bed of an old glacier which once flowed on both sides of Fish Hill to rejoin before reaching the coast at Abbey Rocks. The lake is now drained by the Hall River, a tributary of the Paringa River, the old glacial course to the sea being filled with alluvium. The coast between Waita and Paringa Rivers is very different from that to the north, the beaches being smaller and cliffs often extending to low water mark; islands are common off the coast and show that Tertiary rocks once extended further seaward. A sea cut bench about fifty feet above sea level extends for much of the distance between Paringa and Blue Rivers and is represented further south by a similar feature, Sardine Terrace. Such high level, marine cut benches have not been observed in the morainic bluffs or in the glacially modified Tertiary rocks further south. This coast line resembles the geologically analogous stretch from Martin Bay to Yate Point described by Healy (1938). Low land extends for much of the twenty miles between Waita River and the northern boundary of the next high block south of Arawata River. This low-lying area of alluvial flats between the sea and the Alpine Fault is little above sea level and has an average width of four miles. The flats are not unbroken, but instead of the morainic hills which interrupt the gravel plains of the northern part of the coastal strip, we have here isolated, rounded, granitic hills, very similar in shape, but varying in height from a hundred or so to a maximum, in Mount McLean and Mosquito Hill, of over 2,000 feet. These isolated, rounded hills and other similar hills further north are enumerated later when discussing granite outcrops. Their origin is uncertain, and it is not known if they represent granite intrusions from which the surrounding country rock has been eroded or if they are the characteristic weathering of a granitic belt which extends under the gravel surrounding these isolated hills. It is certain, however, that they have been modified by the alpine glaciers which issued from the mountains only a few miles to the east. Rivers draining ninety miles of alpine slopes converge on this district and flow across it to the sea. The largest of these, the Haast and Arawata Rivers, are similar to those of the northern part of the coastal strip and have wide beds with ill-defined banks. The three smaller rivers, the Okuru, Turnbull, and Waitoto, are less rapid and leave their glaciated alpine valleys to cross the coastal strip by slightly entrenched, open meanders. Large areas of swamp lie between the raised banks of these rivers. From Waita River to Jackson Bay the coast is low lying and, except for a small hillock of greywacke at Mussel Point, composed of sand and gravel. The hills which extend along the southern part of the coastal strip from Arawata River to Yate Point are broken at Cascade, Awarua and Hollyford Rivers by wide alluvial flats. At Awarua River the flats extend from the head of Big Bay along the valley of Pyke River to Hollyford River and separate the Sara Hills from

the other high land. The only other large river is the Gorge which flows in a narrow rocky valley cut through both Greenland and Tertiary rocks. Benson, Bartrum and King (1934) and Benson (1935) considered that the sloping moraine-covered Cascade Plateau is quite comparable with the coastal plateau described at Preservation Inlet, and later Benson and Holloway (1940) published two photographs, one of which shows part of the Sara Hills and the other a part of the coastal strip near the lower Gorge River. They considered that these two districts also represent part of the coastal plateau. In a later part of this paper it will be shown that the Cascade plateau is not only covered by moraine as was suggested by Turner (1930), but that Turner's Conglomerate Series is also moraine and cannot be separated from the overlying material. Healy (1938, p. 83b, Fig. 4) published a photograph of the Sara Hills showing what appears to be an even sloping surface forming the seaward part of these hills. This apparent surface appears to end near the centre of the photograph, whereas actually it extends landward and merges into the northern face of the Sara Hills. It is not at all certain that there are any flattish areas on top of these seaward sloping points, for the apparent flatness may be the result of viewing a convex ridge in profile. They may be as convex in cross-section as the similar looking moraines are known to be further north. In the same way, as on the south side of Big Bay the seaward-sloping Awarua Point is continued east by a bench cut along the north side of Big Bay. We consider that all these features have been caused by glacial erosion and deposition and differ from similar features exhibited by the northern moraines only in that the morainic covering is thinner and more of the underlying rock is exposed. 2. Alpine Fault. The fault which forms the western boundary of the Alps in Westland and south Nelson has already been mapped and discussed by Morgan (1908) and Henderson (1937). In the latter publication (Fig. 1) it has been called the “Alpine Fault,” a name which has been adopted in preference to “Gregory Valley” used by Morgan (1908, p. 72) to describe the valley eroded on the line of this fault in north Westland. Although it has been mapped in different localities as far north as Lake Rotoroa, its continuity and southern extension have not been recognised and in this account the writers endeavour to show that the fault is continuous in south Nelson and Westland, and is the same as that which forms the even Fiordland coast from Milford Sound to Breaksea Sound. In south Nelson the fault has been mapped by Henderson and Fyfe in the published geological maps of Arnauld, Matakitaki, Una, Travers, and Lewis survey districts, and by Morgan (1908) in Toharoa, Totara, Whitcombe Pass, Mount Bonar and Poerua survey districts. The southern extension of the fault is no less definite than the northern, for it can be traced inland through south Westland and western Otago to Milford Sound, and

continues along the Fiordland coast, where a fault has been successively figured by Park (1910), Henderson (1928, Fault Map of New Zealand, and 1937), and Benson (1935). In all these figures except Henderson's (1937) the fault has been shown to extend not only along the even Fiordland coast but also off the coast northwards past Milford Sound to Cascade Point. For this northward extension there seems to be little evidence, for it will be seen from Fig. 1 that the coast is far less even than south of Milford Sound and that the projecting Cascade Point is composed of moraine and has no structural significance. The following criteria have been used to map the fault:— (1) The presence of a scarp or a sudden but regular change in summit height. (2) Wide crush zones with slips and rapid erosion. (3) Subsequent rivers flowing along co-linear courses parallel to the trend of the Alps, connected by low passes. (4) Change in rock type. (5) Offsetting of river courses. (1) The presence of a scarp or a sudden but regular change in summit height. Although the eastern side of the fault is usually marked by a steep slope, this cannot be regarded as having been directly caused by vertical movement, but is more likely due to erosion of the crushed material along the fault line, for along the northern part of the fault the summit heights are about the same on either side. On the eastern side the summit heights are sufficiently regular to afford a surface of reference; but to the west, rounded granite hills mentioned above stand above alluvial gravels. It is not clear whether the granite hills stood above the late Tertiary peneplain (Benson, 1935) as monadnocks and had the same relation to the general surface as they have now, or whether glacial and post glacial erosion has removed the surrounding land. As there is greater lithological uniformity along the line of the Alps than there is along the coastal strip to the west, where granite intrusions have caused considerable local variations, it is to be expected that summit concordance will be less on the western side. There is, however, a considerable difference in summit height on either side of the fault from Waitaha River south to Cascade River, as the following heights taken from the New Zealand Four Mile Map will show. The heights are from north to south, and those on the west of the fault are shown on the left. Mount Bonar 3,527 ft. Mount Ashmore 4,812 ft. Omoeroa Hill 2,237 ft. Mount Price 3,394 ft. Near Paringa River 2,024 ft. Argentum 4,057 ft. Mount McLean 2,400 ft. Mount Watney 4,823 ft. Colin Hill 2,712 ft. Mount Jackson 3,796 ft. The hills on the western side are not continuous as on the eastern side of the fault where the high land is broken only by the glaciated river valleys, but are separated by wide stretches of alluvium little above sea level. The scarp is conspicuous where it faces this alluvium,

as it does from Cook River Flats to Moeraki River and from Haast River to Arawata River. It is also well shown by the north-western face of the flat-topped Mataketake Range which extends unbroken by large rivers for 20 miles. South of Arawata River the throw of the fault decreases and the westward sloping Hope-Blue River Range continues the slopes of the Alps seawards without much apparent break. Here late movements on the fault may have been only horizontal. On either side of Lake McKerrow where the fault was next observed a considerable downthrow on the west was again visible, it being particularly clear on the south side of the lake opposite Hokuri Creek. Benson, Bartrum and King (1934, Fig. 5) have shown a similar feature in a sketch of the coast 15 miles south of Lake McKerrow as seen from the entrance of Milford Sound, in which the even seaward slope of Yate Point can be seen to be much lower than the land to the east. Healy (1938, p. 84B, Fig. 6) has shown the same feature seen from the north. This coastal strip of relatively low land lies to the west of the Alpine Fault and is continuous with that seen from Lake McKerrow. Benson, Bartrum and King (1934) have correlated this coastal strip with the Coastal Plateau at Preservation Inlet which they did not consider to be caused by down-faulting. From Milford Sound to Breaksea Sound the general line of the coast is an even curve continuing the line of the fault from the north, and over this part deep soundings recorded close to the coast indicate that the fault has a considerable down-throw to the west. (2) Wide crush zones with slips and rapid erosion. Large slips are particularly common on the eastern side of the Alpine Fault. These have been mapped by Bell (1906) in the head-waters of Waihinihini Creek and east of the south end of Lake Kanieri. The large slip in the head waters of Macgregor Creek, a tributary of Waitaha River, was considered by Morgan (1908, p. 70) to be due to crushing caused by “a series of faults more or less parallel with the Gregory Valley overthrust.” Crushing was also observed by Morgan (idem. p. 72) on the same line at Cowhide Creek and on the south side of Wanganui River. From this river south to Waiho River moraine and recent material cover much of the fault line. Between the Waiho and Haast Rivers the fault crosses higher country, both on the Waiho-Weheka road where crushed rock is exposed on many of the road cuttings and on the formed track between Paringa and the Haast a few miles south of the Maori Saddle. The distinct notch between the moraine and the schist through which the Weheka-Waiho road now passes, figured by Hackett (1869, p. 12), is almost certainly due to rapid erosion of the crushed schist by tributaries of Waikukupa and Clearwater Rivers, the overlying harder morainic material being undermined. A similar notch is visible on the line of the fault a few miles south of the Karangarua River and to the south of Lake McKerrow; these are probably also due to erosion along crush zones and represent the early stage in the development of the co-linear river valleys.

The crush zone is particularly well exposed on the Paringa-Haast track, and there compact micaceous fault pug is flanked by a wide zone of shattered material, which still preserves the schistosity planes but is divided into small fragments by closely spaced joints. (3) Subsequent rivers flowing along co-linear courses parallel to the trend of the Alps connected by low passes. The subsequent rivers flowing along co-linear courses form a more impressive feature than do the differences in height on either side of the fault, for such streams flow along the fault for one-third of the 300 miles of its length between Milford Sound and Lake Rotoroa. In north Westland the depression formed by these streams was recognised by McKay (1893, p. 136), and Morgan (1908, p. 72) considered it a line of weakness caused by faulting, and termed it the “Gregory Valley.” The following list of streams which flow along it for some part of their course are given from north to south and serve to fix the position of the fault:— South Nelson: Bull Creek and lower part of Durville River; Alfred River and Lake Daniels. North Westland: Upper Grey River; Tass and Nancy Rivers; Evans River; South-westerly flowing part of the Taramakau River; Rough Wainihinihi River; Mount Brown Creek; North-easterly flowing part of Styx River; South-westerly flowing part of Hokitika River; Canada Creek; Head of Doctors Creek; South-westerly flowing part of Mikonui River; Pollock Creek; Douglas Creek. South Westland: Harold Creek; North-easterly flowing part of Waitangi River; Head waters of Blackwater River: The Windbag; Head of Wakapohai River; Head of Waita River; Jacksons River; Cascade River. Western Otago: Head of Pyke River; Hokouri Creek. In practically all cases the streams listed above are connected by low passes and those within the Mikonui Subdivision have been described by Morgan (1908, p. 45). Similar passes over which the road has been constructed connect the Waiho River, Omoeroa River, Waikukupa River, and Cook River flats. The Maori Saddle over which the Paringa-Haast track has been constructed lies on the same line. (4) Change in Rock Type. From south Nelson to Jacksons River in south Westland the Alpine Fault forms the boundary between Greenland rocks intruded by granites and gneiss on the western side and schist on the eastern. The granite rocks appear to have transgressed this fault only between Paringa and Haast Rivers, where mica-bearing pegmatite dykes injected along the schistosity planes, extend east for several miles within the schist belt. From Jackson River, the schist belt swings to the south away from the coast, and the fault forms the boundary between the Greenland rocks to the west and the ultra-basic, the Te Anau, and the Crystalline Complex which lie successively on the eastern side to the south. The relation between this fault and the

pre-Tertiary rocks shows that it was an important feature in pre-Tertiary times; but as the relative ages of the Greenland Series and the schist are unknown, little can be inferred as to the direction of pre-Tertiary movements. (5) Offsetting of river courses. Many of the rivers which have rock bound courses on either side of the Alpine Fault show a displacement to the north-east at the west side of the fault line. The following list shows the amount of this deflection as measured on the Four Mile Map:— Miles South side Lake McKerrow 0.8 North side Lake McKerrow 0.6 Martyr River 0.4 Head of Jacksons River—Lascelles Creek 1.1 Turnley Creek—Cauldron Creek 1.2 Carl Creek—Lake Ellery 1.1 Arawata River 0.8 Moeraki River 0.8 Paringa River 0.8 Waikukupa River 0.8 Tuke River 0.6 Kakapotahi River 0.6 This is more regular than would be expected from coincidence, and is interpreted as being due to lateral movement along the fault having displaced a series of sub-parallel streams. The amount of lateral movement would be expected to differ gradually along the fault, but unless other major faults are present it should be constant over short distances. Fig. 2 has been prepared to show the hypothetical development of the drainage in the Jackson River district where the streams are better defined than in most other areas along the fault. On the north-west side of the fault it is fairly clear that Lake Ellery once flowed north-westward along Smoothwater River through the low saddle which now separates these streams, and that in the same way Cauldron Creek was once part of Stafford River. The changes in the direction of flow of Lake Ellery and Cauldron Creek are due to captures by Jackson River, which eroded rapidly headward along the fault zone. A small stream (which flows south-east to Jackson River), the head of Lascelles Creek, and Carmichiel Creek probably represent a further stage in the dismemberment of a more southerly old stream. It is likely that these three old streams once headed back beyond Jackson River, but now no streams join Jackson River opposite the junction of either Lake Ellery or Cauldron Creek. Joseph Branch, Turley Creek and Carl Creek, however, all flow north-west in the same direction as the old streams, and appear to be the headward portions which have all been equally displaced north-west for slightly over a mile relative to the lower part of their courses prior to capture by Jackson River.

Attitude of Alpine Fault. Morgan (1908, p. 72) considered, largely on theoretical grounds, that the Southern Alps are bounded on the west by an overthrust fault or faults. Actual observations of the dip of the fault plane have, however, rarely been made, but Henderson (1937, p. 79) records that the Wairau Fault (Alpine Fault of Fig. 1 of the same publication) is practically vertical, and the regular trace of this fault further south also indicates that this is likely to be the case. As, however, this fault either forms or is close to and parallel to the boundary between very different Palaeozoic rocks, early movements probably took place along the same line at different times in the past and these may not have been similar to the last movement. Movement along this fault has probably continued until recent times for Morgan (1908, p. 72) mentions that schist has been thrust over recent gravel at Hendes Ferry on the south side of Wanganui River and we observed on the west side of the fault Recent or Pleistocene gravels and silt dipping at moderate angles at Paringa and Moeraki Rivers. At Paringa, these gravels are well exposed along the north bank of the river and extend from the road for ten chains upstream. The beds dip at 25° to the north-west and are composed of gravel, sand and fine silt, mostly derived from a schist terrain. It is not likely that this dip is an original depositional feature for it is uniform for the whole distance over which the beds were observed and the beds show little evidence of current bedding. On the Paringa-Blue River track about five chains south of the Blue River bridge fine silts dip south-east at 25°, these beds are overlain by moraine and appear to be similar to the fossiliferous Pleistocene silts at Pug Creek, to be described in a later paper. For the whole of the 400 miles over which the fault has been mapped, from Resolution Island to Lake Rotoroa, it shows no minor irregularities, being a curve convex to the west at the southern end, and with a radius of curvature increasing towards the north, becoming practically a straight line from Milford Sound to Grey River. Farther north it becomes more irregular but possibly continues as the Waimea Fault along the east of the Nelson lowlands. Its regularity and association with the main axis of the South Island suggests that it has played an important part in the formation of the Southern Alps. Its position and relation to the geology in South Westland is shown by Fig. 1 and its relation to the main drainage divide of the South Island by Fig. 3. 3. Drainage Development. The main divide of the South Island extends from Cameron Mountains in Southern Fiordland to D'Urville Island and shows only minor irregularities over the whole of this distance. It is likely that at least some of these irregularities are due to capture by the more rapid western rivers. This is substantiated by the low passes at the head of the Haast and Hollyford, two rivers which extend

further east than the average; and these passes are assumed to represent air gaps over which the captured head waters of the eastern rivers once flowed. From the Fiordland drainage divide which extends from Mount Tutoko to the Cameron Mountains the rivers and fiords of Fiordland flow north-west, west, south-west, and south almost at right angles to the coast line. On the eastern side they flow east and south-east to meet the Waiau Depression at the same angle. It will be seen from Benson's diagram (1935, p. 4, Fig. 2) that the course of the rivers on the western side is essentially at right angles to the contours of the late Tertiary Peneplain. On the eastern side, however, the contours as drawn do not bear this relation to the rivers. There is some evidence that the eastern contours should be further west, as Mount Soaker, approximately 6,080 ft., is not included within the 6,000 ft. contour and the regular westward rise of the Keplar Mountains between lakes Manapouri and Te Anau is not expressed by the contours. If the contours be adjusted to these heights they then become normal to the drainage, as on the west. Such a relation suggests that the drainage was initiated by and is consequent on the folding of the late Tertiary peneplain and that the highest part of this surface should closely agree with the drainage divide. This is found to be the case. In the same way the main drainage divide further north may represent the approximate position of the anticlinal axis along which the late Tertiary peneplain was folded, there being a similar relation between summit height and drainage in the Mount Cook region as in Fiordland. This is illustrated by Fig. 4, in which the relation of the main rivers and glaciers to the summit heights, generalised by contours, is shown. The surface represented by these contours is probably well below the late Tertiary peneplain in the northern and highest part of Fig. 4 where any flat surfaces have been destroyed by erosion, and summit heights are now controlled by the slope of the valley sides and the height of the valley bottoms. Further south flat-topped residuals still remain and are conspicuous as forming parts of the top of the Mataketake Range. It will be seen from Fig. 4 that south-east of the Alpine Fault and west of the main divide the rivers and glaciers form a regular drainage pattern which becomes increasingly modified toward the north-east and is approximately normal to the summit height contours. On the eastern side of the divide the drainage bears no such direct relation to the summit heights. In the southern part of Fig. 4 the north-westerly flowing streams can be seen to be interrupted by a system flowing north-east or south-west parallel to the strike of the schistosity planes. These are considered to be subsequent streams which have developed later than the north-westerly flowing consequent streams, their development being favoured by easier erosion along schistosity planes or parallel faults. The modification of a north-westerly flowing stream system near Jackson River similar to the above has been previously described while discussing the Alpine Fault.

General Geology. Greenland Series. Morgan used the term Greenland Series for a group of north-westerly striking, closely folded greywackes and argillites well developed at Mount Greenland and the adjoining highland east of the township of Ross. Towards the eastern schist belt these rocks are increasingly intruded by granites. Slates, observed by Cox (1877a, p. 77) east of Lake Mapourika and in McDonald's Creek, were correlated by him with those at Mount Greenland and with other rocks near Reefton and classed as the Auriferous Series. Along the south bank of the Waiho River, and at scattered localities in Omoeroa River, predominantly north-west striking grey-wackes and argillites are exposed beneath the moraines and gravels, the same rocks probably extend east to Omoeroa Hill. Chocolate-brown weathering is a conspicuous feature of these rocks. Somewhat similar rocks in the Paringa and Jackson Bay district were classed by Cox (1877a) with the mica schist east of the Alpine Fault; but the authors consider that, in spite of the rocks having suffered greater metamorphism than those further north, they still resemble the Greenland rocks more than they resemble the foliated schist east of the Alpine Fault. These rocks outcrop over a large area in the Paringa-Waita River district, and are exposed on the coast south of Okuru at Mussel Point, where they were described by Turner (1933, p. 191) as biotite hornfels. Turner (1930b and 1933) has described the hornfels and gneisses of the Jackson River-lower Cascade River district in considerable detail, and has mapped (1933, p. 236) rocks of the oligoclase zone as extending west to the coast at Jackson Bay; this is unlikely to be the case, for the more coastward rocks near the Tertiary contact are far less altered than those examined by Turner further east. Still further south in the Big Bay district Healy (1938, p. 86B) discusses previous correlations of the basement rock made by Park (1887) and Turner (1930b and 1933) and concludes from information supplied by Turner, who examined sections cut from rocks between Big Bay and Wolf River, that the basement belongs to the biotite zone of the Maniototo Series. Healy (1938, p. 85B) states that the degree of schistosity appears to increase inland, as was observed by the authors to be the case also at Jackson Bay, and he suggests that this may be due to a subjacent granite intrusion. Judging from field observation, the writers consider that although there may be some slight schistosity developed in the Greenland rocks south of Jackson Bay, the bulk of the alteration is due to contact metamorphism by the granite intrusions and that the rocks west of the Alpine Fault should be separated from the foliated schist to the east of the fault in the same way as they have been separated further north by Morgan (1908), Bell (1906) and Henderson (1937). Granites and Gneisses. Discontinuous granite intrusions occur towards the eastern side of the belt of Greenland rocks, at least as far south as the Jackson

River. On the west side of the granites the contact with the intruded Greenland greywackes and argillites is an intrusive contact, hornfels and spotted schists being developed. On the eastern side the relation of the granite intrusions to the alpine schists is less clear, for except between the Paringa and Haast Rivers where pegmatite sills intrude the schists, no granite-schist contacts have been observed as the granite becomes gneissic and appears to grade almost imperceptibly into high grade schist and mica schist of probable sedimentary origin. This schist forms a part of the Arahura Series of Bell (1906) and continues south to form the schist belt of Central Otago. From the Lewis Pass to the Mount Cook region, this belt is only a few miles wide, being flanked to the east by greywackes and argillites and further south it widens and extends east of the main divide. Immediately east of the Alpine Fault, the gneisses and schists are remarkably parallel to this fault and to the main line of the Alps, but become more irregular further east. There appears to be no purely granitic intrusions as large as those of the Ross district further south, and many of the areas mapped as granite on Fig. 1 are in part Greenland rocks containing a large proportion of gneiss and granitic dykes. Topographically, the granite and gneiss belt is characterised by isolated, beehive shaped mountains. The rather remarkable isolation of these mountains was mentioned by Bell (1906, p. 41) and the shape is particularly well shown in the Haast district where the lower parts of the hills are covered with alluvium. The following is a list of these mountains given from north to south:— Turiwhate Range 4,482 ft. Island Hill 3,275 ft. Bell and Fraser (1906) Mount Tuhua 3,688 ft. Mount Graham 2,716 ft. The Doughboy 1,967 ft. Mount Misery 3,050 ft. Doctor's Hill 2,230 ft. Morgan (1908) Bald Hill 3,802 ft. Fraser Peak 3,822 ft. Unnamed knobs at Wataroa. Carnavans Knob. Unnamed Knobs at Paringa. Mosquito Hill (Haast, 1879) Mount Brown. Unnamed knob between the Okura and Turnbull Rivers. Gill Hill. Mount McLean 2,400 ft. Quartz Hill. The Knoll, Big Bay.

Tertiary. Historical. The Tertiary rocks near Martin Bay were first described by Hector (1863), who correlated the limestone, sandstone and conglomerate occurring on the headland between Martin Bay and Big Bay with similar rocks at Coal Island and Preservation Inlet. The following sequence at Abbey Rocks was described by Cox (1877a), who considered the coal beds to be identical with those at the Brunner Mine. Fine grained quartz sandstone. Limestone. Grey sandstone. Green sands. Black shale. Limestone. Fine grained brecciated volcanic rock and tufa. Coal beds. Cox (1877b, p. 94) described the Tertiary rocks at Jackson Bay and Smoothwater River and stated that “along the coast line from the heads to Smoothwater River the following sequence prevails”:— 1. Limestone (highly flaky and traversed by veins and calcspar). 2. Marls (highly calcareous). 3. Sandstones and marls. 4. Conglomerates. 5. Marls. 6. Sandstones. A more complete description of the rocks was given by Haast (1879) and in the accompanying geological map he has shown the Tertiary rocks as Cretaceo-Tertiary and Middle Tertiary, extending as a coastal strip from a point between the Paringa and Mahitahi rivers to the Waita River and from Jackson Head southward to the boundary of the map in substantially the same position as in Plate 48, Fig. 1, of this account. Haast, however, did not describe the Tertiary rocks forming this coastal strip south of Jackson Head in any great detail, but correlated them with the Grey Coal Measures (p. 293) of the Waipara Series. He states (p. 293) that “for a great distance along the coast all rocks belonging to this formation are hidden below morainic accumulations till we approach Paringa River, where several miles north of this estuary it is again largely developed, forming a strip several miles broad, and reaching to the Waita River. After a considerable interval we meet it again at Jackson's Bay, whence we can follow it to the southern boundary of the Province, always exhibiting the characteristic feature of littoral deposits.” The examination of this area by the writers served to emphasise the remarkable accuracy of Haast's map. In 1887 Park described the Tertiary rocks that form the extremities of the headlands enclosing Big Bay. He stated (idem, p. 131),

that “they dip west at high angles, and are deeply involved in the old rocks. Their sequence, reading downwards, is:— 1. Coarse conglomerate and sandstones. 2. Hydraulic limestone interbedded with sandstone. 3. Gritty green and grey sandstones. 4. Slightly indurated marly clays with more calcareous layers, which weather out as ridges.” Healy (1938) mapped the Tertiary beds at Big Bay in considerable detail and showed that they extend as a narrow strip as far south as Madagascar Beach. He has shown that although the Tertiary beds are much faulted, the contact with the underlying schist is not a fault contact, either at the south head of Martin Bay or on the south side of Big Bay, where conglomerates rest unconformably on schist. At Madagascar Beach, however, he found evidence of considerable deformation, and stated that (idem, p. 87b) “the beds have been subjected to considerable shearing stress at the time of warping, and near the contact of limestone and sandstone thin streaks of each have been inserted into the other, and the whole contorted.” Healy (idem, p. 89) correlated the Tertiary beds at Martin Bay with those described by Benson (1933) at Chalky Island and by Cox (1877b) at Jackson Bay, and suggested that they are of Lower Miocene age. Benson and Holloway (1940, p. 12) in a footnote suggested a northward extension of the Tertiary sediments mapped by Park (1887) at Awarua Point, Big Bay. Distribution and Structure. At Jackson Bay, the Tertiary rocks that form Jackson Head extend eastward beyond the narrow, low-lying neck of land at the present settlement for a distance of about ten chains and are separated from a more easterly infaulted strip of Tertiary beds by a band of schist about three chains wide. This is well exposed in the coastal road cutting just east of the Jackson Bay settlement. Although the actual contacts were not observed, it is almost certain that the Tertiary rocks are not resting on the schist in their original sequence, but have been complexly involved with the basement rocks. We do consider that between Jackson Bay and Milford Sound the present position of the Tertiary beds can be safely taken as the order of deposition, for while the beds usually dip seaward at angles of 30° to 60°, there are considerable variations. At Long Ridge Point the beds dip to the south-east, away from the sea, over a considerable area; and Healy (1938, p. 80b) showed the Tertiary beds dipping to the east at Madagascar Beach and Smokey Rock. Moreover the lack of lithologic continuity along the strike is more than can be reasonably considered as the result of deposition variations, and the occurrence of lithologically similar beds at various parts of the apparent sequence is very suggestive of repetition. Nevertheless the total thickness of lithologically different beds along an east-west section through Long Reef Point at Martin Bay is about 7,000 ft. This thickness is made up as follows:— 1,400 ft. limestone. 2,000 ft. conglomerate with minor limestone. 3,500 ft. of alternating marls and limestones.

On the whole the beds present a deceptive appearance of regularity which is dispelled by close observation; and it appears that much work will have to be done before the true age and sequence can be interpreted. The writers traced the Tertiary rocks from Jackson Bay to Big Bay, but south of Jackson Bay did not see the actual contact between the Tertiary and the schist, although limestone was observed about a chain north-west of an outcrop of schist at the mouth of Limestone Creek, a tributary of Dee Creek. From Barn Bay to within half a mile of Big Reef Point no outcrops of Tertiary rock were seen; but from the latter point southward to the strip of Tertiary mapped by Healy (1938) the outcrop is continuous except where obscured by gravels. The lithology and structure over this part of the coast are similar to that at Jackson Bay and Big Bay, the beds consisting of limestones, conglomerates, sandstones and mudstones much faulted and indurated. There can be little doubt of the correlation of the whole group, and it is almost certain that they continue under Pleistocene and Recent material in the Lower Cascade Valley to form a continuous strip, from Jackson Head to Madagascar Beach. Evidence that the Tertiary beds continue between Jackson Head and Arnott Point is provided by a sample of limestone kindly collected from Open Bay Island by Mr. Champion, of the Public Works Department. This material is very similar to the limestone at Jackson Bay, being sheared, slickensided and traversed by small veins of calcite. It is likely that the patches of rock and shallow water known to exist between Jackson Head and Arnott Point represent similar material that has been eroded below sea-level. The Tertiary rocks between the Waita and Mahitahi Rivers were not observed by the authors on their trip in 1941, but one of the authors (H. W. W.) has previously examined the Tertiary rocks near the Waita River during geophysical work, and it is known that the basic volcanic rocks reported near Sardine Terrace continue, under recent gravels for several miles south of the last outcrop observed on the beach. It is also known that the Tertiary rocks extend continuously along the coast to the mouth of the Paringa River, forming cliffs and islands similar to those between Long Ridge Point and Big Bay. The eastern boundary of the Tertiary beds along this strip does not appear to be as regular as that of the strip further to the south, and H. J. Evans (unpublished report) has shown that the Tertiary beds do not form a continuous cover right to the eastern boundary, but that outcrops of greywacke and argillite occur in the middle part of Bullock Creek about a mile from the coast. Along the coast the beds dip seaward in the same way as at Tititira Point and Jackson Head. Except for an outcrop of upper Tertiary beds reported by Cox (1877a, pp. 70, 83) from the Waikukupa River and not observed by the authors, no Tertiary rocks have been reported between Mahitahi River and the township of Ross. What was probably a continuation of the Waikukupa upper Tertiary sediments was seen on the north bank of the Omoeroa River about half a mile above the junction of Gibb Creek. The “papa bottom” reported by miners in the lower

part of Sandfly Creek, a small stream between the Waikukupa and Omoeroa Rivers, is probably another outcrop of the same material. The Omoeroa outcrop is unfortunately not a good one, for only a few feet of mudstone close to river level is exposed and a dense growth of vegetation separates this from the greywackes, showing a few chains further upstream. A continuation of the Tertiary coastal strip southward under moraines is probable, for Cox (1877a, p. 86) reported the discovery of volcanic rock similar to that at Abbey Rocks in the moraines north from Paringa, and we observed similar material as well as a boulder of carbonaceous sandstone at Otorokua Point, between the outcrops at Omoeroa and those at Paringa. Outcrops of conglomerate, sandstone and limestone were observed at Homers Saddle, over which the Hollyford track passes between Hidden Falls and the Pyke-Hollyford junction. The beds outcrop at intervals on the track and are largely overlain by morainic material. The relation between these beds and the basement was not observed; but as they were striking north-north-east and dipping at high angles they probably represent a continuation of the deeply involved beds found by Turner (Benson, 1935, p. 9) at Lake Fergus at the head of the Eglinton Valley, which they resemble both in lithology and degree of induration. The conglomerate shown by Benson (1940, Fig. B) at this place may perhaps be some of these Tertiary beds. Age and Correlation of Tertiary Beds. The relation of the Omoeroa River beds to the other Tertiary beds is unknown, the outcrop being an isolated one, but the mudstone is less indurated than the Tertiary beds south of the Mahitahi River, and more closely resembles the lower Blue Bottom siltstones of Hokitika district, with which it has been correlated, than do the samples from Jackson Bay. The Jackson Bay sample was collected close to the basement rock on the coast road east of Jackson Bay settlement, and is an extremely well indurated calcareous mudstone, the associated beds being limestone and sandstone. The relation of this calcareous mudstone to the other beds is not clear, for the beds are faulted and may even be overturned. If, however, there has been less deformation and the beds are substantially as deposited, then several thousand feet of limestone, sandstone, conglomerate and similar mudstone which extend west to Jackson Head overlie it. Dr. H. J. Finlay has kindly examined samples of mudstone from Omoeroa River and from Jackson Bay and prepared the following notes. Likely material from Homers Saddle was also submitted to Dr. H. J. Finlay for examination, but proved to be barren of foraminifera. It may be significant that the nearest marine beds on the eastern side of the Alps are those at Lake Fergus and at Moonlight Fault, which are also considered to be mid-Tertiary (Benson, 1935, p. 9, footnote), Hutton (1939, pp. 85, 86). This may mean that the Cobden Limestone and Kaiata Mudstone, which underlie the Blue Bottom in the Greymouth-Hokitika District, are absent in South Westland; but until further work is done this cannot be taken for granted, for it is likely that the several thousand feet of sediment, much of which is fine grained marl and limestone, represents more than one Tertiary stage, and may in part be older than the samples examined.

Report on Microfauna. Omoeroa River Mudstone. By H. J. Finlay. The sample from Omoeroa River washed down with some difficulty to a residue containing very little sand but an enormous number of crushed tests of foraminifera, principally Globigerina. All the specimens have the surface details obscured and rendered sub-hyaline by alteration, and practically all are very considerably squeezed and battered out of shape by great pressure. An abundant fauna was evidently present, but the state of preservation makes it difficult to extract in anything like entirety. Sorting was continued until but few additional forms were being found, but doubtless a long examination would add 10–20 more. The following 60 species were recognised:— Rhabdammina cf. abyssorum Carp. Ammolagena clavata (J. and P.) Ammodiscus incertus (d'Orb.) Ammodiscus archimedis (Stache) Glomospira cf. charoides (J. and P.) Trochamminoides cf. proteus (Karrer) Haplophragmoides cf. glomeratum (Brady) Cyclammina incisa Stache Vulvulina pennatula (Batsch) ?Semivulvulina capitata (Stache) Karreriella bradyi (Cush.) Karrerulina n.sp. Listerella weymouthi Fin. Martinottiella aff. communis (d'Orb.) Tritaxilina zelandica Fin. Migros medwayensis (Parr) Sigmoilina aff. asperula (Karr.) Robulus dicampylus Planularia (s. s.) small n.sp. “Planularia” n.sp. (as in Mokau beds) Marginulinopsis n.sp. (as in Clifden 2) Dentalina soluta (Reuss) Nodosaria longiscata d'Orb. Chrysalogonium cf. globifera (Batsch) Chrysalogonium cf. obliquata (Batsch) Siphonodosaria globulifera (Kreuz.) Lagena aff. longispina Brady. Guttulina problema d'Orb. Plectofrondicularia whangaroica (Stache) Bolivina lapsus Fin. Bolivina cf. anastomosa Fin. Uvigerina cf. miozea Fin. Uvigerina cf. canariensis d'Orb. Hopkinsina n.sp. aff. parkeri (Karr.) Siphogenerina ongleyi Fin. Siphogenerina vesca Fin. Bulimina bremneri Fin. ?Bulimina scobinata Fin. Nodosarella subnodosa (Guppy) Pleurostomella alternans Schwag. Pleurostomella brevis Schwag. Cassidulina subglobosa Brady. Siphonina australis Cush. Pullenia sphaeroides (d'Orb.) Pullenia quinqueloba (Reuss) Pleurostomella alternans Schwag. Eponides neosoldani Brotzen Cibicides ihungia Fin. Cibicides n.sp. aff. tholus Fin. Cibicides catillus Fin. Planulina n.sp. Discorbis bertheloti (d'Orb.) Discorbis scopos Fin. Laticarinina halophora (Stache) Parvicarinina altocamerata (H. A. and E.) Globorotalia miozea Fin. Globorotalia dehiscens Fin. Globigerina aff. apertura Cush. ?Globigerina triloba Reuss Globigerina bulloides d'Orb. Two-thirds of this list consists of long ranging forms which do no more than indicate the facies background; the remaining 20 species are useful as age indicators. It is obvious at once that the fauna is a “Blue Bottom” one, with the addition of certain elements (Hopkinsina parkeri, Bulimina bremneri, etc.) characteristic of the Ihungia. The association of the latter with Listerella weymouthi, Sigmoilina aff. asperula, Siphogenerina ongleyi and aff. rerensis, Bulimina senta, Cibicides ihungia, aff. tholus, and catillus, Planulina n.sp., and Globorotalia miozea and dehiscens points definitely to that part of the Blue Bottom overlying the basal zones (with Pseudogaudryina anachrons) and underlying the zone where Orbulina first

enters. The whole of this sequence is of course older than the Taranakian Blue Bottom, and belongs in a wide sense to the Lower Miocene, round about Aquitanian. The Blue Bottom that lies between the zones mentioned is again divisible into two parts, and though the allocation of the present sample to these is less definite, the upper is more likely. This period is correlated on micro faunas with (1) that part of the standard Te Uri Stream (Hawke's Bay) section lying directly above the hard alternating siltstones (roughly Waitakian) which cover the Wanstead greasy marls; (2) the lower part of the Ihungia beds of Muddy Creek (Poverty Bay); (3) the Mahoenui beds (Taranaki); and (4) that part of the Clifden section (Southland) approximately comprising beds 4–6a (Park, Finlay). The abundance of arenaceous and Lagenid genera indicates still water and off-shore deposition. The commonest species in the fauna are Listerella weymouthi, Siphogenerina ongleyi, Tritaxilina zelandica, and Cyclammina incisa, and these would seem to indicate water of fair or considerable depth. I have seen no Recent New Zealand faunas from below 60–70 fathoms, but to this depth all are characterised by a quite different and much shallower appearing assemblage. (Textularia, Miliolids, Robulus, Ehrenbergina, Cassidulina, Cancris, Cibicides, Dyocibicides, etc.) Vulvulina, Glomospira, Listerella, Migros, etc., prefer to live between 500–1,000 fathoms; while Siphogenerina ongleyi is very closely related to S. seriata (Cushman and Jarvis), a deep water Globigerina ooze species from the Trinidad Tertiary, and less so to S. irregularis (Bagg) at present living in 400 fathoms off Hawaii. A very close match in faunal make-up and general species resemblance is provided by the Lower Green Clay (no. 13 of the Cipero Section) of Trinidad, a definitely deep water fauna. Recent discussions by various authors (Palmer, Hedberg, Norton, etc.) of somewhat similar Cuban Venezuela, and West Indian assemblages render it probable that such faunas as the Awamoan of All Day Bay and Pukeuri are from over 100 fathoms depth, probably 200–500 fathoms, while the Trinidad and Omoeroa beds are deeper still, and more likely to be around the 7–800 fathom mark. Jackson Bay Mudstone. The Jackson Bay mudstone was so indurated that the extraction of a fauna was much more difficult. It had to be pounded up in a mortar and the final washed residue still consisted mostly of lumps of unbroken matrix. Nevertheless, about 40 specimens were obtained representing about two dozen species. Identification would have been difficult without the Omoeroa fauna for comparison, for it is fairly evident that much the same fauna is represented. The presence of the following 9 species enables a correlation to be made with considerable probability:— Haplophragmoides cf. glomeratum (Brady). Cyclammina incisa Stache. Dorothia cf. arenata Cush. Dentalina soluta (Reuss). Hopkinsina n.sp. aff. parkeri (Karr.). Siphogenerina ongleyi Fin.

Bulimina cf. pupula Stache. Gyroidina zelandica Fin. Planulina n.sp. Geological History. Little can be said of the pre-Tertiary geological history until more is known of the age of the pre-Tertiary rocks. It is known though that the old rocks must have been strongly folded, intruded by granites, and considerably eroded before the deposition of the Tertiary beds. Little is also known of the surface upon which the Tertiary beds were deposited, for no stripped Tertiary surfaces have been reported close to the Southern Alps and the Tertiary beds are usually so deformed that detailed study of the lower beds is impossible. Benson (1933, p. 428) considered that the Tertiary sediments at Preservation Inlet were derived from an area of considerable relief, and the numerous conglomerate bands in the West Coast Tertiary sediments between Yate Point and Paringa River indicate that this may have been the case there also. It is unfortunate that possible deformation makes uncertain the position of these conglomerate bands in the sequence, and we do not know if they continue right through the sequence and represent intermittent uplift of adjoining high land or if they are confined to the base and are merely a basal conglomerate upon which the finer beds were deposited and which is now repeated by faulting. The basal West Coast coal measures are overlain by marine beds of different ages and probably represent deltaic conditions prior to marine transgression. The earliest marine beds are near Greymouth, where the Island Sandstone of Bortonian age (Finlay and Marwick, 1940, p. 106) overlies a thick series of coal measures. Successively overlying this is the fine grained and more extensive Kaiata Mudstone and the Cobden limestone. Above the Cobden limestone is the lower Blue Bottom of Hutchinsonian age (idem, p. 117) with which Dr. H. J. Finlay has correlated both the mudstone at Omoeroa River and the more indurated material at Jackson Bay. Of about the same age are the beds at Preservation Inlet (Benson, 1933, p. 427) and the in-faulted beds at Lake Fergus and Moonlight Fault (Benson, 1935, p. 9 footnote). The distribution of these mid-Tertiary beds and the absence of lower Tertiary beds suggest that in mid-Tertiary time the lower Tertiary sea transgressed both along the southern part of the West Coast and eastward over western Otago and parts of adjoining Southland. This sea may not have extended over what is now the higher part of the Southern Alps for although the fine grained lower Blue Bottom at Omoeroa and Greymouth-Hokitika districts indicates less relief than at present, the Alps have nevertheless since been elevated along a pre-Tertiary and probably Mesozoic trendline. The schist belt which lies along this line may represent a positive unit which has been intermittently uplifted in a similar manner at other times in the past.

Geology of part of the West Coast, South Island

It is uncertain how long deposition continued, for in South Westland and Preservation Inlet any softer younger beds would probably have been removed by erosion, and at Lake Fergus and Moonlight Fault the whole sequence may not have been in-faulted and preserved. In the Greymouth-Hokitika district marine deposition continued without major unconformity till Waitotaran time (Bell and Fraser, 1906, p. 86; Finlay and Marwick, 1940, p. 126). The upper Blue Bottom grades upward into sand and gravel (Bell and Fraser, 1906, p. 85) and this is the first indication of the deformation to follow. The olivine basalt at Koiterangi Hill (idem, p. 84) and the basalt, volcanic breccia and conglomerate interbedded with the marine sediments from Paringa to Waita River show that marine deposition was associated with volcanic activity. The continuous deposition in the Greymouth-Hokitika district makes it likely that no major deformation took place in the adjoining South Westland district until after Waitotaran times. This deformation appears to have been most intense along the line of the present Southern Alps, for the sediments which have been preserved in or close to the Alps at Homers Saddle, Lake Fergus, Moonlight Fault and along the South Westland coastal strip south of Paringa are more deformed than the more distant Tertiary beds at Preservation Inlet, Waiau Depression and along the West Coast north of Paringa River. In the same way further north along the Alps the beds at Murchison and Nelson are more deformed than beds of the same age further west in north-west Nelson. The induration of the Tertiary beds appears to be closely associated with the amount of deformation they have suffered, for the infaulted beds at Moonlight Fault, Lake Fergus, Homer Saddle and the less steeply dipping Tertiary beds between Yate Point and Jackson Bay are so indurated as to make confusion with the pre-Tertiary beds possible. Further north near the Mikonui River highly indurated Tertiary beds have been classed as Tertiary greywackes by Morgan (1908, p. 110). In north-west Nelson, the mountains formed by the folding of the Tertiary sediments still exist and largely control the present drainage; this is not the case in the more closely folded areas nearer the Alps where Benson has shown (1935a) that the major relief features which must have been formed by this folding have been truncated by a late Tertiary erosion surface, the late Tertiary peneplain. Further evidence for this erosion surface was found at Homers Saddle and along the coastal Tertiary strip. In neither of these places are the Tertiary folds represented in the present topography. Further north this surface is represented by the summit height concordance illustrated by Fig. 4. The formation of this now greatly modified surface must have taken place after the folding of the Tertiary beds and before or during the deposition of the pre-Pleistocene Moutere Gravels; for in Nelson and Westland these gravels rest on the truncated edges of the Tertiary beds (Henderson and Grange, 1926). If the folding of the Tertiary beds is post Waitotaran then this surface must have been formed in Nukumaruan or Castlecliffian time.

In South Westland and in the higher part of the Alps where covering beds are absent, the original relief of this late Tertiary peneplain can possibly be inferred by a study of the drainage pattern initiated on it by later warping, a regular pattern of parallel streams suggesting that there was little relief to interfere with the direct course of the rivers. Such a regular drainage pattern has been shown to exist in the southern part of the Mount Cook region while discussing the origin of the drainage there. It was considered that the arching of the late Tertiary peneplain along the present drainage divide initiated the regular system of consequent streams which now flow north-westward from the divide. This hypothesis involves elevation of about 10,000 feet along the Alpine axis in late Tertiary time. The consequent stream system flowed north-westward across the grain of the country and was probably quickly modified by subsequent streams eroding along the strike of the schist. Cotton (1916, p. 245) states that “in the study of the major relief features of New Zealand the trend of the Mesozoic folds may be almost entirely disregarded, for these features have been blocked out by much later orogenic movements”; and it is true that over most of New Zealand there is no exact correspondence between the direction of the pre-Tertiary and the post Tertiary fold axes; but this is not the case with the Southern Alps over that part where it is best defined. Hutton (1885, p. 195) recognised that over much of the Alps the schist belt is parallel to the axis of elevation, but considered that the mountain ranges formed by the Mesozoic folding still survive and have not been base levelled since their formation. It is likely that the Alps were reduced to low relief not only during the depositions of the Tertiary beds, as was suggested by Cotton (1916, p. 247) but also later by the formation of the late Tertiary peneplain after they had been re-elevated during the folding of the Tertiary sediments. It is therefore likely that the parallelism between the schist belt and the present topographic axis is not accidental but represents persistent uplift of the old fold axes. The marine bench now from 50 to 400 feet above sea level was probably cut before the main advance of the ice. The later history of the area is that of the advance and retreat of the ice, and will be discussed in the second part of this paper. It is however included in the following tabulated summary of the geological history. Recent Local overthrusting of schist over recent gravels and tilting of recent gravels along line of Alpine Fault. Formation of alluvial plains. Modification of glacial eroded topography and moraines. Pleistocene Melting of piedmont ice and gradual retreat of alpine glaciers. Advance of glaciers carrying schist. Possible further elevation of Mount Cook region to expose schist belt to erosion. Possible movement along Alpine Fault. Advance of glaciers over low lands. Marine terraces cut and marine beach gravel deposited within a few miles of present coast line.

Pliocene-Pleistocene Orogeny Initiation of present drainage at Fiordland and on western side of Alps. Arching of late Tertiary peneplain along axis close to present drainage divide, axis clevated 7,000 to 12,000 feet. Origin of present Alps. Probable movement along Alpine Fault. Murchison district and north-west Nelson elevated 2,000 to 5,000 feet. ?Castlecliffian (Upper Pliocene) Formation of matureland (late Tertiary pene-plain) over Fiordland, western side of Alps and west Nelson. Moutere Gravels deposited at Nelson and north Westland. Mid-Pliocene Orogeny ?Nukumaruan Present mountains formed and drainage initiated in less deformed areas from Greymouth west of Murchison to north-west Nelson. High mountains probably formed along line of Alps. Intense deformation of Tertiary sediments along line of Alps, strongly deformed sediments now preserved at Homers Saddle, Lake Fergus, Moon-light Fault, Yate Point to Paringa River, Mikonui River, and from Murchison to Nelson. Less intense at lower Waiau Depression, Preservation Inlet, and along West Coast from Paringa River to north-west Nelson. Waitotaran Opoitian (Lower Pliocene) urenuian tongaporutuan(Upper Miocene) Fairly continuous deposition of marine sediments in Greymouth-Hokitika district, upper beds composed of sand and gravel. Retreat of sea from Central Otago. Possible continuous deposition of marine sediments in western Southland. Awamoan Hutchinsonian (Lower and Middle Miocene) Marine sediments deposited in north Westland, South Westland, western Otago, and western Southland. Waitakian Duntroonian (Upper Oligocene) Transgression of sea along southern part of West Coast and eastward over western Otago and western Southland. Whaingaroan Kaiatan (Lower Oligocene) Probable continuous deposition in Greymouth-Hokitika district, the sea transgressing north, east and south over coal measure deltas. Tahuian Bortonian (Eocene) Island Sandstone deposited near Greymouth. References. Bell., J. M., and Fraser, C., 1906. The Geology of the Hokitika Sheet, North Westland, N.Z.G.S. Bull., 1. Benson, W. N., 1933. The Geology of the Region about Preservation and Chalky Inlets, South-west Fiordland, New Zealand, Trans. N.Z. Inst., vol. 63, pp. 393–432. — 1935a. Some Land Forms in Southern New Zealand, Australian Geographer, vol. 2, pp. 3–22. — Bartrum, J. A., and King, L. C., 1934. The Geology of the Region about Preservation and Chalky Inlets, Southern Fiordland, New Zealand, Trans. N.Z. Inst., vol. 64, pp. 51–85. — and Holloway, J. T., 1940. Notes on the Geography and Rocks of the Ranges between the Pyke and Matukituki Rivers, North-west Otago, Trans. Roy. Soc. N.Z., vol. 70, pp. 1–24.

Bolitho, W. J., 1937. Report on Prospecting Operations at Paringa, South Westland, Unpublished Report, N.Z. Geological Survey. Cotton, C. A., 1916. The Structure and Later Geological History of New Zealand, Geol. Mag., Dec. 6, vol. 3, pp. 243–249. Cox, S. H., 1877a. Report on the Westland District, Rept. Geol. Explorations during 1874–76, no. 9, pp. 63–93. — 1877b. Report on Coal Measures at Jackson's Bay, Rept. Geol. Explorations during 1874–76, no. 9, pp. 94–5. Evans, H. J., 1937. Report on Prospecting Operations in the Haast Area, South Westland, Unpublished Report, N.Z. Geological Survey. Finlay, H. J. and Marwick, J., 1940. The Divisions of the Upper Cretaceous and Tertiary in New Zealand, Trans. Roy. Soc. N.Z., vol. 70, pp. 77–135. Haast, J. von, 1879. Geology of the Provinces of Canterbury and Westland, New Zealand. The Times, Christchurch. Hacket, T. R., 1869. Geology of the Okarita District. Rept. Geol. Explorations during 1868–69, no. 5, pp. 8–15. Hector, J., 1863. Geological Expedition to the West Coast of Otago. Otago Provincial Gazette, no. 274, Nov., 1863. — 1868. Notes on the Coast and Harbours from Milford Sound to Hokitika. Rept. Geol. Explorations during 1866–67, no. 4, pp. 42–48. Healy, J., 1938. The Geology of the Coastal strip from Big Bay to Professor Creek, North-west Otago, N.Z. Journ. Sci. Technology, vol. 20, pp. 80b–94b. Henderson, J., 1937. The West Nelson Earthquakes of 1929, Dept. Sci. and Ind. Research, Bull. 55. — and Grange, L. I., 1926. Motueka Subdivision, 20th Ann. Rept. Geol. Survey Branch, Mines Dept., pp. 4–5. Hutton, C. O., 1939. The Bob's Cove Tertiary Beds and the Moonlight Thrust-fault, Trans. Roy. Soc. N.Z., vol. 69, pp. 73–88. Hutton, F. W., 1885. Sketch of the Geology of New Zealand, Quart. Jour. Geol. Soc., vol. 41, pp. 191–220. McKay, A., 1893. Geological Explorations of the Northern Part of Westland, Mines Statement, N.Z. Parl. Paper, C-S, pp. 132–186. MacFarlane, D., 1877. Notes on the Geology of the Jackson and Cascade Valleys, Rept. Geol. Explorations during 1876–7, no. 10, pp. 27–30. Morgan, P. G., 1908. The Geology of the Mikonui Subdivision, North Westland, N.Z.G.S. Bull. 6. Park, J., 1887. On the District between the Dart and Big Bay, Rept. Geol. Explorations during 1886–87, no. 18, pp. 121–137. Park, J., 1910. The Geology of New Zealand, Christchurch, Whitcombe and Tombs. Turner, F. J., 1930a. The Metamorphic and Ultrabasic Rocks of the Lower Cascade Valley, South Westland, Trans. N.Z. Inst., vol. 61, pp. 170–201. — 1930b. Physiographic Features of the Lower Cascade Valley and the Cascade Plateau, South Westland, Trans. N.Z. Inst., vol. 61, pp. 524–535. — 1933. The Metamorphic and Intrusive Rocks of Southern Westland, Trans. N.Z. Inst., vol. 63, pp. 178–284.