Notes on the Geography and Rocks of the Ranges Between the Pyke and Matukituki Rivers, North-west Otago.

Transactions and Proceedings of the Royal Society of New Zealand, Volume 70, 1940-41, Page 1

Notes on the Geography and Rocks of the Ranges Between the Pyke and Matukituki Rivers, North-west Otago. By W. N. Benson and J. T. Holloway. (With Plates 1–3.) [Read before the Otago Branch, June 13, 1939; received by the Editor, June 27, 1939; issued separately, June, 1940.] Introduction and Acknowledgments. The region here discussed comprises the southern ends of the Red Hills and Olivine Ranges, the northern end of the Humboldt Range, and the Barrier Range which extends eastward perpendicularly to the other three. Brief comment is made concerning areas to the west and east of these ranges. Our region contains the widest areas of high land and the most extensive glaciers and snowfields in the Otago Alps. The work of the early explorers, such as Hector (1863) and Alabaster (1863) or of Douglas (1885), and of surveyors such as McKerrow (1863), Wilmot (1883-4) and Mueller (1883–5), though fixing the positions of the main rivers, and, by resection, those of a number of the peaks, permitted only a general sketching of the intervening topographic detail, which the circumstances of McKay's (1879) and Park's (1886) arduous geological explorations did not allow them to rectify. During the last few years, however, the photographs obtained during many mountain ascents and traverses by members of the Otago Section of the New Zealand Alpine Club and others, have made it possible to supplement and modify many of the details on the official Lands and Survey Department map. (Latest 1938 Edition of the New Zealand Four-Mile Sheet No. 28). Such a modified map is presented herewith. (Plate 1). The details as to the positions and names of the peaks, passes, tributary streams and glaciers between the Hollyford, Pyke and Dart Rivers are based on the observations and photographs secured by one of us (Holloway) during four summers' mountaineering, 1935–8. (See Jackson, 1935, Holloway, 1936, 1937, 1938). Details for the Forbes Range are based in part on Sim's (1931) map supplemented by the information derived from Mr. A. R. Craigie. The new features shown on the Richardson Range were reduced from official air photographs by permission of the Director of the Geological Survey and with the help of Mr. R. W. Willett. Details shown between the upper Dart and Arawata Rivers and especially the Snow-white Glacier are based on a sketch map, photographs and other information generously communicated by Mr. A. R. Craigie. The names shown for certain of the features are provisional and subject to the approval of the Honorary Geographic Board. The estimated heights of peaks (rarely derived trigonometrical determination) are, where possible, those on the Lands and

Survey Department's map. Nearly all others cited were first published in the New Zealand Alpine Journal, which, where alternative estimates are available, gives figures often higher than the official estimates, if the assumed peak-identifications are correct.* Peak. L. and S.D. N.Z.A.J. Peak. L. and S.D. N.Z.A.J. Ark 7190 8000 Irvine 7750 7200 Betty 7361 8000 Johansen 6605 7000 Bride 6850 7200 Stefanssen 6554 7500 Cosmos 7340 8000 Watkins 6657 7500 David 7287 8000 Welshman 5490 6000 Hedin 6740 7500 The panorama illustrating most of the region discussed herein was enlarged from a series of photographs taken by Mr. G. L. Edwards from the East Peak of Mount Earnslaw, and was drawn by Mr. Eric Miller, F.R.I.B.A., formerly chairman of the Otago Section of the New Zealand Alpine Club, to whom our thanks are especially due. Other members of this club, by supplying localised rock-specimens, photographs and verbal information have added to our knowledge of the region. Our thanks are due also to the Public Works Department for indicating the course of the Hollyford Road now under construction. We owe to the courtesy of Mr. V. C. Browne permission to publish his copyright air views of the Hollyford and Pyke Valleys on Plate 2 herewith. Thanks also are due to Drs. Turner and Hutton for the general metamorphic conceptions utilised in this paper, and for discussion of certain points that have arisen; to Dr. Turner for the systematic collection from various sources of the specimens and information cited concerning areas beyond the north-western limits of region; and to Dr. Hutton for the sketch of the metamorphic grade of rocks around Lake Wakatipu. The indebtedness of one of us (Holloway) to his climbing companions on his four expeditions is gratefully acknowledged. The geological portion of this work, so far as concerns petrographic detail, is based on a series of specimens obtained from twenty-six localities by J. T. Holloway, supplemented by material obtained by Mr. I. L. Burr and Mr. J. H. Christie near the Lower Hollyford, by Messrs. Ian and Peter McKellar near the Upper Dart, by Mr. A. R. Craigie near Mount Tyndall and at Stargazer, by Messrs. G. Simpson and G. Thomson in the Upper Matukituki Valley, and by Mr. Eric Miller on Mount French. Comparative examination has also been made of material described by Drs. Turner and Hutton (1936) from the Richardson, Forbes and Humboldt Ranges, and by Dr. Marshall (1905) from the vicinity of the Cow Saddle. Brief reference is made also to material obtained north-west of our area by Mr. J. Healy (1938), and also that collected by Messrs. J. H. Christie, E. James and C. Yunge, now in the Museum of the Geological Department of the Otago University. In general our petrological studies may be considered as an extension of the investigations made by Turner and Hutton in neighbouring areas.

Geological Structure And Topography. The most striking feature of the general topography as seen from any high peak within this region is the general accordance of the summit level (see Plate 1 and Plate 3, Fig. a.). Though early recognised by Andrews (1905) as resulting from peneplanation, this accordance has since been discussed in greater detail (Benson, Bartrum and King, 1935, Benson, 1935a), the whole Fiordland region being described in explanatory terms as an elevated subdued matureland, which has been thoroughly dissected and subsequently heavily glaciated. That the region was not in a very advanced state of peneplanation immediately prior to the uplift which made possible the deep dissection, seems to be indicated by the forms of the higher mountain-shoulders, the presence of a rather broad and shallow matured upland valley at the head of the Seaforth River between Lake Manapouri and Dusky Sound (described and illustrated by Wilmot, 1897), and the degree of irregularity among the summit-levels. The last factor is, however, of less significance, being influenced by the headward erosion of streams in the new cycle. There is not sufficient evidence to indicate whether the initial unevenness of surface was the result of mature erosion following an uplift of a few hundred feet only of what had been a surface produced by advanced peneplanation—a suggestion which seems to accord with the features of the Upper Seaforth River—or merely an incomplete single-cycle peneplanation occurring during Late Tertiary times after the Mid-Tertiary (Miocene ?) crustal dislocation (cf. Benson, 1935a). Long prior to such dislocation, however, the Fiordland crust-block had been broken by a series of intersecting master-joints, fault-zones and shatter-belts, and such lines of weakness clearly guided the direction of valleys cut both during and after the Late Tertiary peneplanation. Moreover, they were revived during the differential crust-movements, formation of horsts and graben—or fault-angle troughs—which accompanied the Late Tertiary-Pleistocene elevation. (Benson, 1933, 1935). Such ancient “lineaments” (cf. Hobbs, 1911) are very distinct in the Preservation Inlet area, 140 miles south-south-west of the present region, and some of those therein represented are in evidence in the region here discussed. Thus the Lower Hollyford Valley and Lake McKerrow, the Right Branch of the Routeburn, the Upper Rockburn, the Lower Beansburn, the Dredgeburn, most of the Lower Dart Valley, the Earnslaw Burn, the Lower Rees Valley, the Upper Shotover, and the upper part of the West Branch of the Matukituki Valley appear to follow zones of structural weakness which trend in a direction slightly west of north (cf. the Acheron Passage in Dusky Sound). Valleys trending north-north-east to north-east (compare Edwardson and Long Sounds and the general trend of the coast of Fiordland) are represented by Lake Alabaster, the Pyke Valley, the peridotite intrusion, the Upper Olivine and Hidden Falls Valleys, the Upper Beansburn, the Middle Dart Valley, the Upper Rees Valley, upper part of Snowy Creek, the Dart and Snow-white Glaciers, the Upper Waipara, Joe, Williamson and Arawata Rivers, and the Ice Plateau on the flanks of

the Olivine Range. Zones of structural weakness trending nearly east and west (cf. Cunaris and Dusky Sounds), are represented by the Lower Hidden Falls, Lower Olivine, Forgotten, Diorite, Barrier and Red Pyke Valleys, the Barrier Range, the Upper Dart Valley, the highest portion of the Rees (which is continued through a low col into Snowy Creek), the lower part of the West Branch of the Matukituki, and perhaps also the highest part of the Arawata River. There are hints of the topographic influence of joints or zones of structural weakness running in a south-easterly direction (cf. Doubtful Sound), but they have only slight expression. The region is divided into major geomorphological units by the principal valleys, and these may be described briefly as follows:— West of the Hollyford and Pyke Valleys are the Darran, Skippers and Sara Ranges divided by the Lower Hollyford Valley and Lake McKerrow. The first consists chiefly of more or less gneissic granodiorite with a western belt of biotite-schist fringed by older (?) Cainozoic sediments (Healy, 1938); it is not yet known whether any schist occurs on the eastern flanks of this range. Healy's (1938) account and Browne's air-photographs show that a dissected coastal plateau four miles wide sloping up to a height of over 1800 feet (comparable with that near Preservation Inlet described by Benson, Bartrum, and King, 1935), lies between the sea and the main Darran Range. The regularity of the summit-level of the range is indicated by several illustrations (e.g. Benson, 1935b, Plate 6; 1935a; Fig. 5 and Plate 1 herewith). Mounts Tukoko, Madeleine and Christina, which are a thousand feet or more higher than the general summit-level, are probably remnants of former monadnocks on the old matureland surface. The spurs of the Darran Range slope gently eastward till they turn sharply down to the Hollyford Valley. The marked shoulder thus produced in the valley sides decreases in height seawards from an initial elevation of 5000–6000 feet. North-east of Lake McKerrow, the coastal plateau rises to a height of 2200 feet in the Sara Hills, and consists of semi-schist and schist invaded by porphyrite, with a marginal Tertiary cover (Healy, 1938). North-east again of Big Bay, the McKenzie Range and the gravel-covered Gorge River Plateau, (Tertiary sediments on a schist foundation fide J. H. Christie and C. Yunge), is a further extension of this plateau. See Plate 2, figs. (b) and (a), left hand skyline). The Skippers Range rises directly east of the Sara Hills to an estimated height of 6000 feet (Park, 1887). The rather undulating, once heavily planed upper surface (Plate 2, fig. a) is probably a modification of the old matureland surface, and contrasts markedly with its steep flanks. The range was first ascended by Hector in 1863, but neither he nor any other explorer has given any detailed account of its form. It is cut off to the north by the deep and probably fault-bounded depression filled by gravel and moraines stretching from Big Bay to the Pyke River. The Pyke and Hollyford Valleys, as recognised by McKay (1891), have been cut along an important fault-zone which continues southwards into the upper Eglinton Valley, wherein Turner (1936)

found a strip of marine Tertiary sediments thrown down nearly five thousand feet below the adjacent general level of the Fiordland matureland surface (cf. Benson, 1935b). The form of these valleys can be gathered from the above-mentioned illustrations. Beneath the higher shoulders are more or less marked traces of a shoulder about 2500–2000 feet above the present valley floors, perhaps indicating vaguely the base-level attained at the time of formation of the coastal plateau. The northward decrease in height of this shoulder along the Pyke Valley suggested by Plate 2, fig. (a), may have resulted from spur-truncation by the branch of the Hollyford Glacier which moved up the lower Pyke Valley to discharge into the Big Bay depression, and on its retreat left moraines (e.g. at the Knoll), which deflect the northern tributaries of the Pyke from their former direct seaward course, so that they now discharge through Lake Alabaster (Park, 1887, Marshall, 1906). The major tributaries entering this valley from the east, High Falls Creek, Hidden Falls Creek, the Olivine River, and Barrier River, though entering at present at grade, drop over noteworthy falls or high cascades a short distance back from their mouths, possibly the result of headward erosion of streams rejuvenated by uplift during an inter-glacial period followed by step-making during the last glaciation, in which, perhaps, the oversteepening of the lowest (500–300 feet) portion of the valley-sides was effected. No data are available for discussing the relation (if any) of such an uplift with the raised beaches and terraces, 100–300 feet high, that fringe the shore (Hector, 1863, Park, 1887), or the occurrence of uncemented conglomerate noted along the eastern side of the Lake Alabaster and the Pyke Valley (Christie, priv. com.). The branch of the Hollyford Glacier which was deflected by the Skippers Range into Lake McKerrow failed to remove all of the spur over and around which it passed, and has left a strongly glaciated spur-residual. Post-glacial dissection of this and other remnants of spurs have etched out a series of shatter-zones running parallel to the general direction of the valley. (Plate 2, fig. (b).) During the last glaciation the rock-floor of the main valley may have been eroded to a depth below the present sea-level, but has since been partially covered by fluvioglacial deposits forming the present broad and swampy valley-bottom. The Bryneira Range east of this consists chiefly of semi-schistose or phyllitic breccia and tuffs, quartzites and limestones, with almost vertical dip and perhaps a general anticlinal arrangement. It was described by Park (1887) as being “remarkable for the uniformity of its outline and the equality of height exhibited by its higher peaks* Probably this is all that was intended by his later (1921) reference to the “flat-topped” character of the Bryneira Range. many of which are over 6000 feet high. The effect of glaciation seems to be remarkably vivid. It consists on its western side of a number of square-shouldered spurs separated by narrow ravines. The spurs terminate abruptly, presenting steep faces to the Pyke Valley. They are mostly devoid of vegetation, and consist of great

billowy masses, with finely rounded outlines, clearly showing the marks of long continued ice-erosion” (Park, 1887). (See Plate 2, fig. (b).) The eastern side of the range rises steeply, but is recessed to a moderate extent by various cirques separated by truncated spurs. The range itself is notched by Alabaster Pass at an elevation of about 4500 feet. This pass was strongly eroded (Marshall, 1906) by an overflow from the glacier which filled the valleys divided by the Cow Saddle. Few details are available concerning the northward continuation of the Bryneira Range west of the Red Hills, save the occurrence in it of intrusive masses of diorite. East again of this range lies a belt of ultrabasic and basic rocks having a variety of topographic expression. Where it consists of schistose serpentine only, as along the narrower or marginal portions of the belt, or where the ultrabasic rocks fail to appear and only the fault-zone (into the lower portions of which they were injected) is present, weathering and erosive processes have been rapid, and a succession of subsequent collinear valleys has resulted, namely, the Upper Hidden Falls and Olivine Valleys and those of tributaries of the Diorite, Barrier and Red Pyke Rivers. The broad (600 yards) grassy flat of Cow Saddle, situated at an elevation of about 3000* Estimates vary between 2514 feet and 3650 feet. feet and increased in area by alluvium held up by small moraines, lies on schistose serpentine and sheared gabbros (Marshall, 1906). So also does the flat between the lower and upper gorges of the Barrier River. The “Olivine Ledge” (Plate 3, fig. b.) forming the prominent shoulder on the east side of the upper Olivine Valley probably consists of more massive peridotite than that into which the river has cut.† The stream here occupies a gutter forty feet deep cut into the glacial and scree-rubble in the bottom of the glaciated valley. In general the massive peridotite, pyroxenite and serpentine, lying between the schistose margins of the broader portions of the intrusion, resist weathering and rise into striking prominences, such as Fiery Peak near Cow Saddle, Little Red Hill a mile north of the Barrier River and, six miles further north, the larger mass of Red Hill proper which gives its name to the range. These all have the ochreous colour and lack of vegetation characteristic of ultrabasic rock-masses, which, associated with ruggedness and high relief, make extremely conspicuous peaks, as has been noted by all who have referred to this region. The Humboldt and Olivine Ranges which, with the Barrier Range, make up the crust-block immediately to the east of this, comprise a greater width (about six miles) of country above a height of 5500–6000 feet than exists in any other part of Otago. The whole mountain mass consists of chloritic, sericitic or quartzitic schists with an approximately meridional strike and a westward dip at very high angles on the western side of the block, though dipping less steeply easterly or more often westerly on the eastern side of the block. They are cut by strongly marked joints, the combined topographic effect of these structural features being the production of markedly asymmetric

ridges. (See Plate 3, fig. a, and compare Healy and Willett, 1938, p. 13.) The western margin of the block is not nearly so continuous as the steep slightly recessed eastward slopes of the Bryneira Range on the further side of the relatively low country occupied by the ultrabasic rocks, but is deeply trenched by numerous valleys extending back into cirques, some of which contain small glaciers. The shouldered spurs and ridges above the 5000–5500 level lead back to a highland area, on which lie nearly twenty square miles of glaciers, some of which descend with a comparatively gentle slope from the peaks and arětes, before discharging into steeply sloping hanging or valley-glaciers. The Ice Plateau discharging eastward by the Andy Glacier affords a notable example of these. The topography repeats on a broader scale the features of Skippers Range, the higher parts of which were moulded by the headward erosion of glaciers on the portion of a matureland surface remaining above the heads of youthful gorges, which were rapidly cutting back into the recently elevated Fiordland region just before glacial conditions became established. That the greater extension of ice should be on the eastern or lee-side rather than the western side of the main divide accords with the general experience in other lands. (e.g. Gilbert, 1904.) Attention was first called to it by McKerrow (1863) when discussing the sudden floods among these mountains. “Change of temperature is the necessary and immediate cause, but … a flood may occur without any perceptible increase of temperature for the wind by transporting the snow to a lower altitude occasions the same effect as a rise of temperature … The snow-line on the north-west (windy) side of the mountain range was higher than on the south-east, the sheltered side, showing the wind is a very decided cause in producing the effects now under consideration.” In addition to this the familiar contrast between the heating effects of the morning and afternoon sun may be recalled (cf. Garwood, 1910). It was among the peaks, arětes and glaciers of this portion of the region herein discussed, that the bulk of the exploratory work of one of us (Holloway) was carried out. East of this block the Dart Valley is cut to a depth of nearly 6000 feet below the summit-level, and probably follows a shatter-zone, which continued southwards, determined the western side of Lake Wakatipu, while northward it is followed by the Beansburn. Above the junction with this stream, the Dart follows a north-north-east to south-south-west direction. Up to the junction with the Brideburn, the valley has a generally alluviated floor and a fall that averages about 20 feet per mile, but above this it narrows and rises nearly 500 feet within two miles before opening out into Cattle Flat. East of the Lower and Middle Dart Valley is the Forbes Range, the topography of which has been described by Sims (1931), Henderson (1937), Healy and Willett (1938). It consists of foliated quartz-albite-sericite schists, often chloritic or rich in epidote, possessing a general dip towards the west at angles which near the summit of Mount Earnslaw, may decrease to about 20 degrees. The mountain rises a thousand feet above the general summit-level, and though it

may well be merely the remnant of a former monadnock, there is a possibility that a horst-like uplift of this range between the bounding, probably fault-guided, Dart and Rees Valleys. The features of the Rees Valley and the Richardson Range are being investigated by Willett for the Geological Survey; a preliminary account thereof has recently appeared (Willett, 1939). Near the head of the Rees River the strongly glaciated valley follows a north-east to south-west course, but its highest portion runs directly west from the pass (4750 feet) leading into the collinear middle portion of the Snowy Creek Valley. This east-to-west strike characterises most of the Upper Dart, which as an open, in part alluviated, valley rises gently for six miles, receiving the discharge from small tributary glaciers on the Barrier Range on the north. The largest of these, the Whitburn glacier, also has a western course before it bends southwards towards the Dart Valley, terminating in a rather gently sloping valley, a mile above the steep descent into the Dart Valley. From the head of this glacier the north-eastern end of the Barrier Range is formed by the Mount Edward-Moriori-Maori-Liverpool Ridge, with peaks 8680–8040 feet high, east of which the Dart Glacier, five miles in length, follows a south-westerly direction. Though tributary glaciers enter it at grade from the east, its western side is marked by an almost continuous strip of shelf-glacier in the lee of the above-mentioned ridge which, widening into its recesses, forms the Penck, Park, Marshall, and Hess Glaciers (see Knowles, 1931). The whole system of glaciers in this valley illustrates clearly the dismemberment by shrinking of the great Pleistocene, dendritic, Dart Valley Glacier. North-west of the Edwards-Maori Ridge the Snow-white Glacier, recently explored by Mr. A. R. Craigie and party, moves north and north-east, ending on a platform high above the deeply incised head of the Arawata River. At several points before its termination, the glacier overtops the strongly eroded ridge which encloses it on the north, thus discharging by more than one stream into the Arawata Valley. The Barrier Range between here and its western extremity presents an almost unbroken wall between the Dart and Arawata Valleys, the lowest point (5800 feet) being O'Leary's Pass, leading from the Pass Burn to the valley draining from the Victoria Glacier into the Joe River. The information available to the writers is insufficient to determine whether the great amphitheatre between the uppermost Arawata and Joe Rivers has resulted solely from erosional processes as a magnificent “trough's end.” The north-easterly direction of the valleys in this region is suggestive of marked structural control. Williamson's Flat, at a height of about 1200 feet above the sea, is described as “about four miles long by two miles wide, bounded on the south side by magnificent sheer cliffs a thousand feet in height, surmounted in turn by lesser cliffs and terraces hewn out by glacial action, rising in gradually retreating stages to the great white masses of the Snowball Glaciers.” (Miller, 1925.) An air-photograph taken by Miss Theomin from the north shows that these terraces

are determined by the planes of schistosity which here dip gently to the west, while the cliffs follow joint-planes almost perpendicular thereto. To the east of this Mount Aspiring (9957 feet), the “New Zealand Matterhorn,” probably the remnant of a great monadnock, rises 2000 feet above the general summit-level. Geology and Petrology. The Bryneira Range. Park (1887) described this range as a syncline composed chiefly of Maitai and Te Anau rocks, the former being grey, blue, and red fissile slates and impersistent masses of limestone with interlaminated sandstones, striking meridionally with almost vertical dip, the latter consisting of massive sandstones of all shades of green and aphanitic breccias in contact with the serpentine near the head of the Hidden Falls Creek, but separated from them at the low saddle by an indurated grey pyritous rock. They are everywhere intersected by dykes of grey “syenite and granite.” The western foot of the range he describes as “Kakanui” rocks like those of the Skippers Range—hard, grey, micaceous sandstones, blue calcareous schists, soft micaschist with a peculiar wavy or corrugated structure. It may be doubted, however, whether there is good reason for the distinction of three sedimentary series. Microscopical investigations show that the rocks on the western foot of the range (north-eastern shore of Lake Alabaster, Localities 2 and 3, are very fine-grained pale apple-green, jointed but not markedly schistose (2480, 4731), consisting of quartz and albite in grains rarely more than 0.005 mm. in diameter clouded with dust-like grains of epidote (?) among which are small (0.01 mm.) plates of sericite often, but not always, elongated in the direction of incipient schistosity. Occasionally small unoriented plates of colourless chlorite are present. In other rocks from the same neighbourhood (4842, 4843) the grain-size (0.08 mm.) is larger, the schistosity more pronounced and the feldspar grains a little elongated. A chloritic pseudomorph after augite may occasionally be recognised in rock which seems to have been originally tuffaceous. A much less altered phase of such rocks (4861) was found at Homer's Saddle (Locality 4) six miles further south, a mineralogically reconstituted crystal-tuff (albite, chlorite and epidote) in which are abundant larger feldspar grains 0.2–0.5 mm. in diameter. This rock shows only slight traces of incipient schistosity, but is associated with a strongly sheared, minutely granulated phyllonite (4860) in which thin bands of quartz-albite granules (0.01–0.02 mm.) and some sericite alternate with bands rich in sericite, dusty epidote and occasionally finely divided carbonaceous matter. Between these two localities near Locality 3 on the north-east shore of Lake Alabaster, Burr obtained apparently in situ, but possibly from a massive landslip, a rather finely granular keratophyre (4847), with but little sign of schistosity. It contains phenocrysts of albite (Ab94) up to 3 m.m.* This is merely a pass over a spur projecting into the Lower Hollyford Valley; not the well-known saddle at its head.

long, with a few leucoxenic and chloritic pseudomorphs set in a matrix of finely (0.05 mm.) lathy albite, intersected by narrow shatter-zones containing a little quartz and epidote (dust-like or coarsely prismatic), with sometimes a narrow selvedge of chlorite. About Mount Frenchman (Localities 8 and 7) more or less schistose grey limestones are well developed (4758, 4759), and contain small irregular grains of quartz and albite with occasionally thin laminae of chlorite and a little epidote. Here also are examples of more or less altered Te Anau Breccia. A coarsely clastic rock (from Locality 9) contains fragments 2–5 cms. in diameter set in an epidote matrix, the individual fragments being composed of finegrained keratophyre (the albite showing faintly its original trachytic structure), with shreds of epidote, actinolite and sphene replacing the original pyroxene and ilmenite. In (4732) from Locality 7 a slightly schistose rock contains fragments of chloritised andesite and keratophyre in a ground-mass of finely granular quartz and albite, with scattered calcite and interstitial chlorite and haematite. A comparable example (4841) from the Hidden Falls is a crystal-tuff, with fragments of keratophyre and spilite in a completely reconstituted matrix. In other rocks, e.g. (4733), the rock-fragments are not present, and the slightly schistose rock is composed of the above-mentioned minerals only. The green and red semi-schistose phyllitic rocks noted by Park prove to be more or less schistose spilite-keratophyre tuffs. One of these (4735), not found in situ, consists of a minutely granular (0.02 mm.) mosaic of quartz and albite with rare apatite prisms, scattered sericite flakes, and parallel plates of strongly pleochroic yellow-green chlorite in plates (0.01 mm. long), alternating with bands consisting chiefly of chlorite with a subordinate amount of ferruginous epidote and dusty magnetite, and others in which sericite, haematite, and dust-like or idioblastic epidote predominate. In Sunny Creek, which enters the Hollyford ten miles south of the Hidden Falls Creek, an area of the Chl. 3 metamorphic grade according to Turner (1938), Dr. Marshall obtained a related rock (Za 8). It is almost mylonitic in its degree of shearing, but the movement seems to have been followed by some crumpling and recrystallisation, with the development of a less finely granular mosaic (0.1 mm.) in streaks and irregular patches. Adjacent to the peridotite at Cow Saddle is what may be the grey pyritous rock stated by Park (1887) to occur along the fault-zone into which the peridotite was injected. It (Ya 156 = a slide kindly lent by Dr. Marshall), also has been intensely mylonitised, the quartz grains are sometimes ribbon-like, though usually equant and oriented, so that the optic axes make high angles with the plane of shearing. Irregular, but sometimes lenticular, albite grains occur among them, and are only occasionally elongated in the general direction of schistosity. The fretted edges, sieve-like inclusions, and separation by narrow irregular bands of matrix, suggest that a considerable amount of recrystallation has occurred since shearing ceased. Narrow tortuous veinlets of minute fibro-granular epidote traverse the general mosaic, and are in turn intersected by veinlets of

unoriented quartz and albite, into which project glassy clear outgrowths from such large residual grains of albite as they may traverse. A specimen (1913) from Deadman's in the Hollyford Valley, eight miles south of the Hidden Falls, differs from the above chiefly in the abundance of the large (0.5–1.0 mm.) albite grains, and of laminae of sericite and haematite, and in the absence of the veinlets. It is probable that these Bryneira Range rocks are the more altered northward extension of the Te Anau rocks described by Turner (1936) near the head of the Routeburn Valley, and in other regions west of Lake Wakatipu. In limestone among them Hector (1891) noted traces of unnamed fossils near Lake Harris which he thought to be Permian, and Park (1921) recorded the occurrence of casts of Productus, Spirifer, Spiriferina, a Turbo-like gasteropod and two corals in the northern end of the Livingstone Range. Probably the same series of Te Anau breccias, etc., are represented further south on the shore between Orepuki and Riverton, a spilitic aggregate among which H. Service (pers. com.) recognised pillowlava, which Benson also saw later near Riverton. Traced further south-east these volcanic rocks are represented in more altered form in the spilite-keratophyre tuffs, etc., of the Greenhills, Bluff Peninsula, in which a zaphrentoid coral has been found (Service, 1938). In the Bryneira Range all occurrences of the breccias and associated rock show the effects of regional metamorphism in varying degree. Turner (1938) has usefully summarised the criteria on which he and Hutton have established four subzones, Chl. 1–Chl. 4 of increasing grade of metamorphism within the general chlorite zone. Of the rocks described above 4861 perhaps 4732 might be considered as showing metamorphism no higher than Chl. 1, but the remainder seem more typically of the Chl. 2 grade, the more schistose examples approaching Chl. 3. The Skippers Range and Neighbourhood. Little information is available concerning the rocks of the Skippers Range. Hector (1863) seems to have been the only geologist to ascend this range (from the eastern side), and describes the rocks as “showing in the lower parts porphyritic greenstone like that near Milford Sound,” (see below) “overlain by granite-gneiss consisting of laminae of feldspar and quartz with plates of pearly mica irregularly dispersed. The rock sometimes passes into rose-tinted” [haematite ? piedmontite ?] “schist, or white quartzite with layers of mica and feldspar arranged in very fine laminae. They are the same rocks as are met with on Black Peak near Wanaka Lake.” His manuscript map of 1864 (now in the Geology Department of Otago University) shows the formations here grouped with “contorted and foliated schists.” Park (1887), while indicating the presence of crystalline schists on the north-east side of Lake McKerrow chiefly on Hector's authority, classes the bulk of the rocks in the Skippers Range as grey micaceous sandstones, blue calcareous schists and soft mica schists with a peculiar wavy or corrugated appearance, apparently comparable with the rocks on the eastern shores of Lake Alabaster. In the McKenzie Range, to the north-east of Big Bay,

are coarse mica-schists and hornfelsic mica-schists in the watershed of the Gorge River as shown by specimens collected by Mr. Charles Yunge, who states, however, that between these and the Red Hills are slates and greywackes with small tongues of serpentine. West of the Skippers Range and south-west of Martin's Bay, the basement rocks are dark grey north-easterly striking schists, with quartz lenticles apparently decreasing in metamorphic grade towards the coasts. In Turner's view they may be quartz-biotite-muscovite schists, resulting from the thermal metamorphism of rocks in the chlorite zone. (Healy, 1938.)* It may be recorded here on the evidence of specimens and information supplied by Messrs. J. H. Christie and C. Yunge that the Tertiary sediments noted by Park north of Big Bay extend over much of the western slopes of the McKenzie Range, rising to a height of nearly 2000 feet, and across the lower portion of the Gorge River Valley and to the north thereof. Tertiary limestone occurs here and there along the coast for at least ten miles north-east of Big Bay. (See Generalised Map, Plate 1) The Older Intrusive Igneous Rocks. The region of these biotite-bearing schists lies immediately north of the end of the Fiordland granodiorite which makes up much of the Darran Range. Pegmatites, aplites and porphyrite invade the schists between Martin's and Big Bay. A gneissic epidiorite pebble (not in situ) at the head of Lake McKerrow (4846) is typical of many other rocks from the Fiordland batholith being composed of large slightly zoned andesine tabulae, and schillerised diallage passing marginally into pale green hornblende fraying out into poikilitic intergrowth with biotite, the last mineral occurring also in the cleavage planes of the feldspars. The darker minerals tend to be aggregated and drawn out into streaks wherein also are ilmenite passing into sphene, crushed plagioclase, quartz and a little clinozoisite. Apatite is a rare accessory. The occurrence of a pebble of hornblendic hypersthene-gabbro (1912) at Deadman's (see above), on the western side of the Hollyford River, may also be noted. The amphibole occurs in but small amount only, moulded on or partly replacing the diopside, and some replacement of hypersthene by talc has occurred. The rock is generally similar to the mica-norites of the Darran Range (Marshall, 1907), but contains very little biotite. Contrasting with this there occurs in the upper part of the Kaipo River a poikilitic, hornblende, epidote, garnet-gneiss (1914) with small grains of deep red-brown rutile, but no feldspar or quartz, differing in this respect from the various hornblende-garnet-gneisses described from Milford (Speight, 1910). Intrusive masses of diorite, such as were stated by Park to be common in the Bryneira Range, were observed by Holloway to abound in the continuation of the range north of the Olivine River at least as far as Mount Annetta. The specimen he collected (4736, Locality 6) is rather crushed, and the plagioclase has been converted into albite, seamcd or partially replaced by epidote.† Several specimens of Fiordland granodiorite (1420, 1421, 1423) occurring as boulders in the Routeburn Valley have their plagioclases similarly altered. The subpoikilitic

brown hornblende is schillerised and the ilmenite converted to iron-stained leucoxene. Narrow shatter-zones filled with finely granular albite, quartz and epidote with a little sericite traverse the rock. Other examples may be afforded by a saussuritised uralitic “diorite” at the “Bryneira” (Alabaster ?) Pass (Marshall, 1906), and a pebble from near the Hidden Falls (4845), a coarsely granular rock containing partially saussuritised basic oligoclase and pale brown poikilitic hornblende marginally chloritised and including small amounts of colourless diopside, ilmenite and apatite. In contact with this rock is the margin of a dyke of diorite aplite or malchite, an equigranular rock composed of prisms of turbid feldspar (0.2 mm.) on which pale brown hornblende grains are moulded, and accessory ilmenite. From the southern end of the Skippers Range, a quarter of a mile from the “chair” over the Pyke River (Locality 1), Burr obtained (4840) an augite porphyrite in which those phenocrysts which were probably orthopyroxene have been completely replaced by talc and dusty carbonates. The augite prisms (3 mm. long) are in part still fresh, pale greenish and slightly zoned, but there are also patches of chlorite and chalcedony possibly replacing pyroxene. The groundmass is basaltic containing lathy labradorite, granular augite and magnetite. The above described igneous rocks may be considered to be all coeval with the Fiordland batholith, and possibly of very late Palaeozoic or early Mesozoic age. They are almost certainly older than the late Mesozoic peridotites now to be discussed (cf. Turner, 1934). Ultrabasic and Basic Igneous Rocks. The Te Anau (Bryneira Range) rocks are cut off to the east by a strongly-marked fault into which, as Park showed, has been injected a series of ultrabasic and very subordinate basic plutonic rocks, which have not everywhere risen to the same height in the earth's crust, and are now exposed in a rather discontinuous series of intrusions traceable southwards from the Red Hills. Those rocks are easily recognisable by their characteristically barren outcrops. They are well developed between the Red Hill proper and the Barrier River, and may be seen from a distance exposed on the pass between this valley and that of the Diorite Stream, and on the pass between the Diorite Stream and the gorge of the Olivine River. From an air-view it would seem that both branches of the Diorite Stream lie on these rocks. There are frequent exposures on the mountain slopes on either side of the Upper Olivine River, and a very large development on Fiery Peak (Locality 10) and the western slopes of Mount Niobe (Locality 11). The adjacent floor of the Hidden Falls Valley is mantled with peridotite-debris, and there is another outcrop on the slopes of Mount Frenchman (Locality 9) where, as noted by Park, it comes into contact with the Te Anau Breccia. To the south there is a gap of several miles in the continuity of the peridotite belt before coming to the small outcrops on the Hollyford slopes of the Lake

Harris mountains, the presence of which was reported by McKay (1881) and verbally by Butement (See Ulrich 1890). It is possible that other small outcrops occur between Lake Harris and the Lower Routeburn and Caples Valleys, as peridotite pebbles were found in the Routeburn Valley (1425, 1443, 1452), but no evidence of the extension of the main fault-zone which the peridotites invade has yet been noted immediately south of the Hidden Falls Valley, and the stretch between this and the serpentines traced by Hutton (1937) southwards from the east branch of the Eglinton Valley deserves detailed study. A general account of the petrology of these rocks in the Red Hills region was given by Ulrich (op. cit.), and in much greater detail by Turner (1930). Marshall's (1906) account of these by the Cow Saddle is the only source of petrological information concerning those within the present area. The few rocks which were collected during the present exploration do not permit us to add much to this. Marshall showed that at the Saddle there lies on the east a wide mass of Iherzolite with vertically dipping bands of more or less serpentinised Iherzolite and of diorite. Immediately west of this is a thin band of pyroxenite, followed by one of gabbro, merging into diorite which abuts against the Te Anau rocks of the Bryneira Range. Our most massive specimen (4737) of peridotite from Fiery Peak, though probably part of Marshall's Iherzolite, contains but little diopside, and approaches more closely to the harzbergites. The enstatite has locally passed into talc or less often bastite, but there are very few veinlets of normal chrysotile serpentine. The spinellid mineral is a chromite-picotite, but in addition thereto are a few crystals and irregular grains of an opaque mineral with metallic lustre which is possibly awaruite. Pebbles of normal schistose chrysotile serpentine were obtained on the saddle in the Olivine Valley three miles further north-north-east (Locality 23), and again on the eastern slope of Mount Frenchman (Locality 9), where it is associated with a gabbro that is but little different from that at Cow Saddle. In Marshall's slide (E 4) of this the plagioclase is entirely decomposed and replaced by cloudy translucent saussurite. The subophitic diopside is either quite fresh and in regular intergrowth with pale brown amphibole, or is replaced by very finely flaky antigorite often fringed with pale green actinolite, which may extend with complete crystallographic continuity into the brown amphibole. A few grains of pyrites are present. Serpentines of various types derived from the ultrabasic rocks in and beyond the northern margin of our area have been collected by Mr. C. Yunge in the Gorge (Jerry), Durward, and Red Pyke-Rivers (1495–9, 2475–6), and these, with a score of specimens obtained by Dr. G. Moir from Red Hill, a mile north of the boundary of our region (1238, 1258–61, 1403–1417) described by Dr. Turner (1930, 1934), and others collected by Mr. R. F. Landreth, illustrate a variety of altered peridotites, serpentines, gabbros, and their peculiar associated vein-rocks.

The Schists of the Humboldt, Southern Olivine And Barrier Ranges And The Upper Dart-Matukituki Region. These ranges from which the bulk of our rock specimens were collected consist almost entirely of schists, the macroscopic features of which were well described by Park, who distinguished green schist, grey pyritic quartz-schists and softer micaceous schists distributed throughout the series at irregular intervals, the whole complex showing many variations in mineral composition. He noted that they have been subjected to intense lateral pressure and locally show a strong plication. The strike of the schistosity is almost uniformly meridional, and the dip on the western side of the Humboldt Range is nearly vertical or even very steeply inclined to the West, where it is cut off against the peridotites by an almost vertical fault. On the eastern side of the range the dip of the schistosity is much gentler, and is usually directed westerly at angles less than 40°, or even 10°. The same gentle westerly dip is seen in the Forbes Range further east (Healy and Willett, 1938). In the main range between the Dart and the Matukituki, the rocks still strike meridionally and are granular chloritic schists containing magnetite still dipping west and overlying the soft mica-schist further east. At the far source of the latter river a thick band of compact green rock dipping west at an angle of 45° is overlain by soft mica schists, which latter weathering rapidly leave the green rock as a smooth sloping surface on one side of the approach to Hector's Col (Locality 42). Using the criteria stated by Hutton and Turner (1936) and Turner (1938) for the recognition of the metamorphic grade, the rocks of Mount Nox in the Humboldt Range appear to belong to subzone Chl. 2, while from Turner's (1935) descriptions those near Mounts Momus and Somnus appear to be transitional between the subzones Chl. 2 and Chl. 1. The limit of subzone Chl. 2 probably extends to about Mount Chaos. The remainder of the Humboldt Range, the Forbes Range and eastern slopes of the Richardson Range are in subzone Chl. 3, while the rocks of the Olivine and Barrier Ranges, the main portion of the Richardson Range and the regions further north and east (cf. Hutton and Turner, 1936) are characteristic of subzone Chl. 4. Representative Rock Types: Subzones Chl. 1–2. The rocks described by Turner (1936) include the “blue slate” (1989) and partially sheared keratophyre-spilite tuff (1988) from near Mount Somnus, and the phyllite (2077) [extremely like that (4860) from Homer's Saddle, Lower Hollyford described above], the slightly sheared greywackes (1997–8, 2060–2) and the fine-grained quartz-albite-calcite-semischist (2065). The rocks on Mount Nox (Locality 30) include (4754) a grey unfoliated semi-schist in which the quartz-albite mosaic has a grain-size rarely larger than 0.01 mm. though containing larger strained relict grains. Pale chlorite and sericite not markedly oriented are present—sometimes clouded with dusty epidote and rarely prisms of tourmaline up to 0.2 mm. long. A still more finely granular example (4755) shows an alternation of

the quartz-albite layers with those composed of matted sericite and subordinate chlorite. In a dark olivine-green plicated, but still more finely granular rock (4760) the flaky minerals are much more abundant, and are associated with a little finely granular epidote and occasional small tourmaline prisms. Sub-zone Chl. 3. In the schists, both of this sub-zone and Chl. 4, the foliae are often wrinkled and irregularly discontinuous, thickening or fraying out into broken or anastomising laminae as if the rocks had been to some extent “telescoped” in the folding and shearing processes. Sometimes there appears to have been repeated shearing, the later movements occurring in a direction oblique to the earlier. Occasionally the quartz and, to a less extent, the albite, grains have been extended into ribbon-structures more or less parallel to the axial planes of the small corrugating folds or obliquely to the foliation. Owing to the fact that our specimens are all unoriented rock chips, the structures present can be described in general terms only. (Cf. Knopf and Ingerson, 1938, pp. 157–162 and Plate 17.) Characteristic Mineral Assemblages. Quartz-albite-chlorite-sericite-haematite-schists. The examples of these rocks within the limits of the Chl. 3 subzone are (4756, 4739) from Localities 27 and 31 respectively. The first is very finely granular, the grain-size being rarely greater than 0.02 mm., while the second on the opposite margin of the subzone has a grain-size almost 0.2 mm. Foliation is fairly well expressed, layers of sericite and very finely divided haematite alternating with layers consisting of the other three minerals with a little scattered albite. Ribbon-structure of the quartz in a direction oblique to the foliation is well developed in 4739. Quartz-albite-sericite-epidote-stilpnomelane-schists. Phyllitic nearly homogeneous rocks at Localities 19 (4743) and 25 (4742) are characterised by a quartz-albite mosaic with grain-size (0.005 mm.) less than that usual for this subzone. They contain small flakes of sericite and prisms of clinozoisite generally oriented in a common direction, with late-formed stilpnomelane either enveloping the quartz and albite poikilitically or forming small flakes elongated in any direction. It is of the yellow-brown ferric type (cf. Hutton, 1938). Quartz-albite-chlorite-epidote-(haematite)-schists. These include the less finely granular schists at Localities 21 (4748) 26 (4747), and 28 (4746). The laminae of quartz and albite grains (0.02–0.10 mm.) associated with a little epidote, chlorite and sometimes calcite, alternate with bands of chlorite and epidote fringing or fringed by streaks of finely divided haematite. In these the epidote occurs either in close-packed aggregates of minute (0.02 mm.) granules or in idioblastic prisms 0.2 mm. long. It is a ferruginous strongly pleochroic and birefringent type.

Epidote-chlorite-actinolite-schists. The sole example studied from subzone Chl. 3 is a pale applegreen, rather foliated rock (4749) from Locality 22. It consists chiefly of actinolite in ragged prisms (0.2–0.05 mm.) mostly trending in a general direction, though frequently oblique thereto, set in a matrix of more finely fibrous actinolite, very pale almost isotropic prochlorite and minutely granular epidote. A comparable rock (1907), differing from the latter in the scarcity of actinolite and presence of irregular streaks of quartz-albite mosaic (grain-size 0.03 mm.), was obtained in the Red Pyke Valley, and was derived probably from the northern extension of the Bryneira Range. Sub-zone Chl. 4. The bulk of the schists examined may be included in this group, and include rocks with mineral assemblages and structures similar to the above, but differing from them in their greater grain-size and more marked foliation, together with others possessing different mineral-assemblages and structures. In general the grain-size increases towards the eastern margin of our area, where also occur rocks with the most distinctive structures. Characteristic Mineral Assemblages and Rock Types. Quartz-albite-chlorite-sericite-(haematite)-schists. These are represented by rocks from Localities 12 (4740), 16 (4741), 32 (4844), 33 (4858), 34 (4851), 35 (4855), 37 (4852), 38 (4857), 39 (4863), 41 (4859), 43 (1304, 1305, 1467), and a rock (4864) from the summit of Mount Stargazer two miles north of Mount Aspiring. The foliation is expressed chiefly by the occurrence of layers of sericite with or without finely divided haematite or graphite alternating layers of quartz and albite. The chlorite, a prochlorite, may be in either association. Accessory epidote and rarely apatite and calcite are occasionally present. The original lamination, best shown by the opaque material, has been torn into lenticles and greatly contorted into open or close-packed isoclinal folds, broken or dragged out by shearing. Ribbons of quartz and albite running more or less parallel to the axial planes of the corrugations meet their limbs obliquely (4741, 4844, 4851, 4852), but may occur parallel to the sheared-out laminae (4857, 4858, 4863), or follow conformably the corrugations, or showing intersecting directions (4855), the result possibly of repeated shearing during folding. In some cases the effect results from the growth of elongated quartz grains across tensional fractures which lie oblique to the general schistosity (4851). The rock (4859) from the Shotover Saddle contains fairly well developed crystals of sphene and of pyrites in addition to the above-mentioned minerals. Perpendicularly to one or more (but rarely to all) of the sides of the latter, ribbon-like quartz grains may extend 0.2–0.6 mm. out into the general matrix in which there are indications of two directions of shearing. The rock (4864) from Mount Stargazer, beyond the north-eastern corner of our map, is characterised by the strongly marked development of albite into equant porphyroblasts 0.5–1.0 mm. in diameter, with straight lines of inclusions in varying

directions indicative of rotation of the albite grains since their formation. The subsequent recrystallisation of the groundmass is indicated by the manner in which the bands of platy minerals bend and become wider or narrower as they wind between the porphyroblasts; quartz is present in small amount and calcite is abundant. Crystalloblastic growth of albite, with inclusion of platy chlorite and streaked graphitic material is displayed also by coarsely granular quartzose schists (on Mount French, Locality 43) (1305), and the ridge extending eastward from it (1304) (Turner, 1934). Quartz-albite-chlorite-epidote-schists. These were obtained from Localities 18 (4848), 33 (4849), 36 (4850), 37 (4854), 40 (4862), and 42 (1471). The first of these is transitional from the group last described. There is some elongation of quartz and albite grains in a general direction across the corrugated folding of the dark lamellae, the direction of which is preserved by the inclusions in the larger colourless grains. Sericite, calcite and finely divided magnetite occur in some amount, and occasional scattered or grouped prisms of tourmaline 0.2 mm. long. In the next two the colourless minerals chiefly albite form broadly lenticular grains up to 0.8 mm. in diameter, usually without inclusions in the closely matted chlorite and epidote. The general grain-size of 1471, the “compact green rock” of “Hector's Col” (MacKay, 1882, p. 82) is smaller, while that of 4854 is still smaller, and the colourless minerals are very subordinate. A little actinolite is present. The magnetite-bearing chlorite-schists of the Upper Matukituki Valley are represented by (4862), a more coarsely granular rock from the eastern slopes of Mount Tyndall, in which the faces of the black octahedra are sometimes covered by a thin film of calcite. Epidote-chlorite-actinolite-schists. The rocks of this group may or may not contain subordinate amounts of quartz, albite, sericite or sphene. They are well represented in the northern part of our area (Localities 13 (4753), 14 (4751), 15 (4752), and 18 (4750). In 4752 the matrix of the rock is chlorite, containing a little actinolite and minute “insect-egg” grains of sphene. In this are set idioblastic brightly pleochroic prisms (0.2 mm.) of epidote elongated in the general direction of schistosity. Through the matrix extend lenticular streaks of actinolite and albite (0.1 mm.), the latter containing actinolite fibres either elongated in the general direction or curved as in “snowball garnets,” indicating rotation of the host-mineral during its crystallisation. Twining is developed in these albites to an extent that is unusual among the metamorphic rocks of this area. Quartz-albite-sericite-epidote or actinolite-stilpnomelane-schists. These are represented at Localities 18 (4744) and 35 (4856). In the first the colourless minerals are dominant, and are more or less elongated parallel to the general schistosity, the thin, dark layers consisting of sericite, clinozoisite and graphitic dust. The stilpnomelane has developed late and may envelop grains of quartz and albite, or form bunches of sub-radiating flakes up to 0.5 mm.

long, lying at any angle to the schistosity. It is of the dark olivine green type with but little brownish tinge, and is therefore rich in ferrous oxide (cf. Hutton, 1938). A little chlorite, actinolite and pyrites is present. In (4856) actinolite is abundant, and the brownish stilpnomelane still more so, while clinozoisite occurs only in small grains in the thin turbid sericitic streaks. Both coloured minerals are elongated in the general direction of schistosity. Quartz-stilpnomelane-spessartite-schist. This has so far been obtained only at Locality 17 from Mount Blockade, at the head of the Forgotten River near the transition between the Chl. 3 and Chl. 4 sub-zones. It (4745) is a dark glossy brown slightly foliated schist. The quartz grains (about 0.1 mm. in average diameter) are aggregated into laminae about 0.5 mm. thick, alternating with thinner layers composed chiefly of stilpnomelane flakes about 0.5 mm. long, lying almost parallel to the schistosity. Flakes of this mineral also occur in subradiating aggregates or singly, lying obliquely across the quartz laminae. Associated with the dark layers are streaked aggregates of innumerable minute (0.01 mm.), almost colourless, idioblastic garnets, which are also scattered sparingly through the quartz layers. Dr. Hutton has kindly examined this slide and remarks (priv. com.) that it is a normal type comparable with rocks found in the Chl. 3 and Chl. 4 sub-zones in the Lake Wakatipu region, and may be compared with one (3202) from Staircase Creek near Kingston which he has illustrated (Hutton, 1938, p. 175, fig. 3b). The tiny garnets forming strings and dense clusters are a common feature in schists such as this; they contain very high MnO and may correctly be called spessartite. The stilpnomelane has an olive-green to olivine-brown colour, and therefore must contain appreciable FeO, and as spessartite is abundant, the parsettensite molecule may be present in considerable amount in the stilpnomelane mineral. He agrees that it is probable that the few small (0.015 mm.) idiomorphic almost colourless doubly terminated prisms may well be the almost colourless tourmaline elbaite. Some General Considerations. Certain major features of tectonic interest are suggested, not only by the distribution and characters of the various grades of metamorphic rocks, but by their relation to the line of ultrabasic intrusion to which attention will be first directed. The elongation of the intrusive masses of ultrabasic and basic igneous rocks along the general strike of the Southern Alps has been generally recognised, and was summarised by one of us (Benson, 1926), but as a result of the present mapping, and more particularly of the work of C. O. Hutton (1937), who has traced the extension of these rocks for sixty miles further to the south, it is now possible to recognise that it swings in a westerly convex arc a hundred miles in length. This long arcuate line of intrusions exemplifies excellently Suess's dictum (1909) that such intrusive masses “form sills in dislocated mountains

that sometimes follow the bedding planes and sometimes the planes of movement.” The boundaries of the intrusive peridotites, etc. are sharply marked faults separating them from the schists of the Humboldt and Olivine Ranges to the east, and from the semi-schist of the Bryneira Ranges to the west, a point first noted by McKay (1881) and emphasised by the more detailed examinations of Park (1887). Further north, Turner (1930) has shown that the peridotites, etc., occupy a fault-zone between the schists of the Olivine Range and the much more strongly metamorphosed schists of the Hope-Blue River Range to the west. In the south, though he did not investigate in detail the character of the formations bounding the serpentine on either side beyond commenting that they comprised chiefly the greywackes of the Te Anau series and their slightly metamorphosed equivalents, Hutton (1937) recognised that “the actual contacts on both sides are marked by zones of intense shearing along which fault-breccias have been developed.” It is, however, noteworthy that throughout its extent the peridotites, etc., of this long belt like those of the Great Serpentine Belt of New South Wales (Benson, 1926, p. 39, fig. 11) have been injected into an almost vertical fault-zone, so that active horizontal thrusting had ceased at the time of their intrusion, which in our present case is believed to have been at the close of the Late-Jurassic—Early-Cretaceous (?) orogeny. A detailed interpretation of the structures and textures of the above described metamorphic rocks must await petrofabric investigation of oriented specimens. That they have suffered intense lateral compression, shearing and “external rotation” seems apparent, and the general west-south-westerly convexity of the boundaries of the several metamorphic sub-zones accords with the evidence derived from Manapouri (Turner, 1937-8), and Eastern Otago (Turner, 1938), that the compressive stresses acted across N.N.W.-N. axes and may have been connected with the westerly to south-westerly superficial thrusting near Preservation Inlet. Further, it seems, from the distribution of the various grades of metamorphism, that a late phase of the crust-movements was a flexure on a broad scale resulting in an uplift in the regions north and east of the Lake Wakatipu regions and a southward pitching syncline to the west where the rocks of the Chl. 1 sub-zone are bounded to east and west by those of Chl. 2 and 3 [Hutton and Turner, 1936, confirming McKay's' (1881) earlier studies]. The northern limit of this major syncline is recognisable in the southern margin of our present area. It is possible that this broad flexuring occurred in connection with the intrusion of the peridotites, etc. Certainly it was completed and had been followed by prolonged erosion before deposition of the Cainozoic marine sediments, for the latter rest directly on almost unmetamorphosed or slightly altered Te Anau rocks west of the Livingstone Range (Hutton, 1937), and near the head of the Hollyford River (Hutton and Turner, 1936), but on rocks of the Chl. 4 sub-zone in the inverted sequence at Bob's Cove, Lake Wakatipu (Park, 1909).

List of Specimens Described. Register No. Otago Univ. Geol. Dept. Map. Loc. No. Locality. Collector. 1304 Summit Range E. of Mt. French B. Lazarus 1305 43 Summit of Mt. French 2 m. sth. Mt. Aspiring E. Miller 1420-31 Boulders on South side of Mt. Momus F. J. Turner 1467 43 Mt. French E. Miller 1471 42 Hector's Col—Matukituki Saddle, head of Left Branch G. Thomson & G. Simpson 1495-9 Headwaters of Jerry R. (tributary of the Gorge River) C. Yunge 1898–1910 The Red Pyke and Durward Rivers " 1912 Deadman's, Hollyford Valley E. James 1913 " " " " 1914 Upper Kaipo River " 1988-9 South-west slopes of Mt. Somnus F. J. Turner 2475-6 Durward Creek and Upper Pyke J. H. Christie 2480 2 East side of Lake Alabaster " 4731 3 " " " " " 4732 7 Low Peak, Mt. Frenchman Holloway 4733 8 Summit, Mt. Frenchman " 4734 9 East Slope, Mt. Frenchman " 4735 11 Pebble, Cow Saddle " 4736 6 North of Olivine R., west of Peridotite " 4737 10 Fiery Peak " 4738 9 East Slope, Mt. Frenchman " 4739 31 Summit, Mt. Hedin " 4740 12 Summit, Mt. Gyrae " 4741 16 Summit, Mt. Intervention " 4742 25 Third Peak, Mt. Tantalus " 4743 19 Summit, Mt. Mallory " 4744 18 West Slope, Mt Gates " 4745 17 Mt. Blockade " 4746 28 Col between Mts. Niobe and Tantulus " 4747 26 Second Peak, Mt. Tantalus " 4748 21 West Slope, Mt. Brenda " 4749 22 Fohn Saddle " 4750 18 West Slope. Mt. Gates " 4751 14 Foot, Intervention Saddle " 4752 15 Top, Intervention Saddle " 4753 13 Summit, Pic d'Argent " 4754 30 Summit, Mt. Nox " 4755 30 " " " 4756 27 Summit, First Peak, Mt. Tantalus " 4757 20 Summit, Bride Peak " 4758 8 Summit, Mt. Frenchman " 4759 7 Low Peak, Mt. Frenchman " 4760 30 Summit, Mt. Nox " 4839 18 East Side, Mt. Gates (Head of Joe River) " 4840 1 West Side, Pyke River near Chair L. Burr 4841 5 Pebble, Hidden Falls River " 4842 2 North End Lake Alabaster " 4843 2 " " " " 4844 32 Mt. Stefanssen Holloway 4845 5 Pebble, Hidden Falls River L. Burr 4846 — Pebble, Head of Lake McKerrow "

4847 3 N.E. of Lake Alabaster L. Burr 4848 18 Mt. Gates Holloway 4849 33 Whitburn Glacier, Median Moraine I. C. McKellar 4850 36 Quarter Mile S.W. end of Dart Glacier " 4851 34 West side, Whitburn Glacier " 4852 37 South-east side, Dart Glacier " 4853 37 " " " " " 4854 37 " " " " " 4855 35 East side, Whitburn Glacier " 4856 35 " " " " " 4857 38 South-east side, Dart Glacier " 4858 33 Median Moraine, Whitburn Glacier " 4859 41 Shotover Saddle A. R. Craigie 4864 — Mt. Stargazer, 2 miles N. Mt. Aspiring " 4860 4 Homer's Saddle, Lower Hollyford Valley L. Burr 4861 4 " " " " " " 4862 40 Lower E. slope, Mt. Tyndall A. R. Craigie 4863 39 Summit, Mt. Tyndall " E4 39 Cow Saddle P. Marshall Ya 156 39 " " " Za 8 — Sunny Creek, Hollyford Valley " Postscript added April 22, 1940. Dr. V. I. E. Whitehead has recently explored the portion of the Olivine Range extending along the margin of the region we have mapped and for six miles north thereof, and has found the need for slight modification of certain geographic details in the map here presented. Thanks to his collecting, we can now record the presence in this newly visited region, of quartz-albite-chlorite-(sericite-epidote)-schists, with subordinate or abundant chlorite and either poikiloblastic structure with rotated albite grains (5042, 5046, the latter from near the snout of Trinity Glacier), or with ribbon-like elongation of the quartz grains (5044, 5045); also of quartz-albite-sericite-(chlorite-haematite)-schist with sigmoid ribbon-like quartz grains probably resulting from recrystallisation during shearing (5043). These rocks show the structure characteristic of the Chl. 4 metamorphic sub-zone, which has thus been traced to within six miles of the southern limit of Dr. Turner's pioneering petrological study of the Olivine Range in 1929–1934. Dr. Whitehead has also provided us with further specimens from the belt of ultrabasic intrusions, namely saussurite gabbro from the vicinity of Beresford Peak (5047) and from Little Red Hill (5040) on the opposite side of the Barrier River, together with a partially serpentinised lherzolite (5041) from the last-named locality.

Literature Cited. Andrews, E. C., 1905. Some interesting facts concerning the glaciation of south-western New Zealand, Rept. Aust. Assoc. Adv. Sci., vol. 10, pp. 189–205. Benson, W. N., 1926. Tectonic conditions accompanying the Intrusion of Basic and Ultrabasic Igneous Rocks, Memoirs National Academy of Sciences. Washington, vol. 19, pp. 37–9, 43–4. — 1931. The Geology of the Region about Preservation and Chalky Inlets, South-west Fiordland, New Zealand, Part 1, Trans. N.Z., Inst., vol. 63, pp. 393–432. — 1935. (with Bartrum, J. A., and King, L. C.) Ibid., Pt. 2, Ibid., vol. 64, pp. 51–85. — 1935a. Some Land Forms in Southern New Zealand, Australian Geographer, vol. 2, pp. 3–23. — 1935b. Notes on the Geographic Features of South-western New Zealand, Geographical Journal, vol. 86, pp. 393–401. Garwood, E. J., 1910. Features of Alpine Scenery due to Glacial Protections, Geog. Journal, vol. 36, pp. 310–336. Gilbert, G. K., 1904. Systematic Asymmetry in the High Sierras of California, Journ. of Geology, vol. 12, pp. 570–586. Hector, J., 1863. Geological Expedition to the West Coast of Otago, Otago Provincial Gazette, No. 274, Nov. 5, p. 435. — 1891. Progress Report, 1890–91, Report Geological Explorations, p. xlv. Henderson, J., 1937. Glenorchy District, 31st Ann. Rept. Geol. Surv. Branch, Dept. Sci. and Ind. Research, pp. 16–22. Hobbs, W. H., 1911. Repeating Patterns in the Relief and in the Structure of the Land, Bull. Geol. Soc. America, vol. 22, pp. 123–176. 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. Healy, J., and Willett, R. W., 1938. Glenorchy Subdivision, 32nd Ann. Rept. Geol. Survey Branch, Dept. Sci. and Ind. Research, pp. 12–16. Holloway, J. T., 1936. From the Middle Dart to the Back of Beyond, N.Z. Alpine Journal, vol. 6, pp. 289–299 (with plates). — 1937. Olivine Bound, Ibid., vol. 7, pp. 50–58 (with panoramas). — 1938. Further Olivine Peaks and Passes, Ibid., pp. 204–213 (with panorama and map). Hutton, C. O. and Turner, F. J., 1936. Metamorphic Zones in North-west Otago, Trans. Roy. Soc. N.Z., vol 65, pp. 405–6 and map. — 1937. Preliminary Note on the Occurrence of an Ultrabasic Intrusion in the Livingstone Range, Western Otago, Ibid., vol. 66, pp. 349–50. — 1938. The Stilpnomelane Group of Minerals, Mineralogical Magazine, vol. 25, pp. 172–206. Jackson, A. J., 1935. Peaks of the Middle Dart, N.Z. Alpine Journ., vol. 6, pp. 36–44 (with plates and map). Knopf, E. B., and Ingerson, E., 1938. Structural Petrology, Geol. Soc. America, Memoir 6. Knowles, J. D., 1931. Ascents of Mounts Maori and Edward from the Dart, N.Z. Alpine Journ., vol. 4, pp. 128–133 (with plates). Marshall, P., 1906. Geological Notes on the Country North-west of Lake Wakatipu, Trans. N.Z. Inst., vol. 38, pp. 561–568. — 1907. Geological Notes on the South-west of Otago, Ibid., vol. 39, pp. 496–503.

Mckay, A., 1881. On the District West and North of Lake Wakatipu, Rept. Geol. Explorations, 1879–80, pp. 118–147 (with map). — 1882. The Geology of the Waitaki Valley and parts of Vincent and Lake Counties, Ibid., 1881, pp. 56–92 (with map and figs.). — 1892. On the Geology of Marlborough and South-east Nelson, Part ii, Ibid., 1890–91, p. 23. Mckerrow, J., 1863. Report of the Chief Surveyor, Otago Provincial Gazette, p. 381. — 1871. On the Physical Geography of the Lake Districts of Otago, Trans. N.Z. Inst., vol. 3, pp. 254–263. Miller, E., 1925. Mt. Aspiring and Matukituki Valley, N.Z. Alpine Journal, vol. 6, p. 20. Mueller, G., 1884. Report on the West Coast between the Cascade Plateau and Jackson's Bay on the North, and Lake McKerrow and the Cascade Valley on the South, Ann. Rept. Dept. Lands and Survey, N.Z. Park, J., 1887. On the District between the Dart and Big Bay, Rept. Geol. Explorations, 1885–6, pp. 121–137 with map. — 1921. The Geology of Western Southland, N.Z. Geol. Survey, Bull. 23, p. 14. Service, H., 1937. An Intrusion of Norite and its accompanying Contact Metamorphism at Bluff, New Zealand, Trans. Roy. Soc. N.Z., vol. 67, pp. 185–217. Sim, J. A., 1931. Pluto Peak and Mt. Sir William, N.Z. Alpine Journal, vol. 4, pp. 191–205 and map. Speight, R., 1910. Notes on the Geology of the West Coast Sounds, Trans. N.Z. Inst., vol. 42, pp. 255–267. Suess, E., 1909. The Face of the Earth (Sollas Translation), vol. 4, p. 564. Turner, F. J., 1930. The Metamorphic and Ultrabasic Rocks of the Lower Cascade Valley, South Westland, Trans. N.Z. Inst., vol. 61, pp. 170–201. — 1930a. Physiographic Features of the Lower Cascade Plateau, South Westland, Ibid., pp. 524–535. — 1934. Metamorphic and Intrusive Rocks of Southern Westland, Ibid., vol. 63, pp. 179–284. — 1935. Schists from the Forbes Range and adjacent Country, Western Otago, Ibid., vol. 64, pp. 161–174. — 1936. Metamorphism of the Te Anau Series in the Region North-west of Lake Wakatipu, Ibid., vol. 65, pp. 329–349. — 1938. Progressive Regional Metamorphism in Southern New Zealand, Geol. Magazine, vol. 75, pp. 160–174. Ulrich, G. H. F., 1890. On the Discovery, Mode of Occurrence and Distribution of the Nickel-Iron Alloy, Awaruite, on the West Coast of the South Island of New Zealand, Quart. Journ. Geol. Soc., vol. 46, pp. 619–632. Willett, R. W., 1939. Glenorchy Subdivision, 33rd Ann. Rept. Geol. Survey Branch, Dept. Sci. and Ind. Research, pp. 7–9. Wilmot, E. H., 1897. Report on the Region between Lake Manapouri and Dusky Sound, Ann. Rept. Dept. Lands and Survey, N.Z.

Fig. (a)—Generalised Geological Map of South-western New Zealand. Fig. (b)—Detailed map of the region between the Hollyford, Pyke, and Matukituki Rivers based on Sheet 28 of the official Four Mile Map (1938) supplemented by information derived from many sources. Heights shown are approximate only, and certain place-names are provisional and subject to the approval of the Honorary Geographical Board. Fig. (c)—Panorama looking S.W.-N.E. from East Peak Mount Earnslaw, drawn by Mr. E. Miller from photographs by Mr G. L. Edwards. Mount Cook, occasionally visible from this point, is 100 miles distant. To face page 24

Fig. (a)—Lake Alabaster and the pvke River Valley. The Red Hills in the middle distance, the North Bivnena Peaks with the summit-level of the Olivine Range on right. Fig. (b)—The Lower Hollyford River Valley The Skippers Range in middle the mouth of the Pyke River and Lake McKenow beyond which appears a portion of the dissected coastal plateau (Sara Hills).—Copyright An Photographs by Mr. V. C. Browne

Fig. (a)—Looking north along the strike of the schists from Pic d'Argent (Locality 13). Mt Darkness in foreground. Red Hills proper on the right skyline, the Hope-Blue River Range on the left. Showing the co-ordination of the higher summit level and its contrast with that of the Lower Gorge River (coastal) Plateau on the extreme left Schistosiv-planes dip steeply to the west joint-planes to the east. Fig. (b)—Looking along the Olivine River from the summit of Fiery Peak. “The Olivine Ledge” in the middle foreground ending in a spur of Mt. Sunsct. Mts. Ark and Intervention beyond with the Red Hills to the left separated from the former by the South Branch of the Diorite Stream. Mt. Tantalus under the cloud.—Photographs by J. T. Holloway.