Go to National Library of New Zealand Te Puna Mātauranga o Aotearoa
Volume 67, 1938
This text is also available in PDF
(3 MB) Opens in new window
– 185 –

An Intrusion of Norite and Its Accompanying Contact Metamorphism at Bluff, New Zealand

[Read before the Otago Branch, April 13, 1937; received by the Editor, April 14, 1937; issued separately, September, 1937.]


  • Introduction.

  • Outline of Geology.

  • Metamorphic Series.






    Greenhills group.


    Tewaewae Point group.


    Foreshore group.


    Mutual Relationships.


    Nature of Metamorphism.



  • Intrusive Series.







  • Conclusion.

  • Table of Rock Analyses.

  • Literature Cited.


Bluff, the southernmost major port in New Zealand, lies at the end of an undulating range of hills connected with the mainland by a low-lying gravel plain. While the gravels all lie on the eastward, or leeward, side of the range, the westward side is exposed to the prevailing south-west winds and sea-currents of Foveaux Strait. The range, about 9 miles long in a N.W.S.E. direction and averaging some 1 ½ miles in width, includes Bluff Hill and the Greenhills Range, separating which is a narrow dune-covered isthmus at Ocean Beach. Bluff Harbour, a tidal lagoon on the eastern side of the range, has a narrow entrance bounded on the one side by Bluff Hill, and on the other by the rock-defended promontory of Tewaewae Point Peninsula. A smaller tidal lagoon, Mokomoko Inlet, lying to the north of Bluff Harbour, drains northwards into the estuary of New River. Throughout this paper, “Bluff Hill” will refer only to that portion of the range to the south-east of Ocean Beach, whereas “Bluff Peninsula” will be considered as including also the Greenhills Range.

As early as 1855, Charles Forbes, a surgeon on H.M.S. “Acheron,” which was at the time the base-ship for a survey of the West Coast Sounds and Stewart Island, noted that Steward Island, Ruapuke Island (12 miles south-east of Bluff) and Bluff consisted of “dense blue-coloured rock containing hornblende in place of mica” (Forbes, 1855, pp. 521–522). In 1863 James Hector observed at Bluff “compact greenstones,” “siliceous felstones,”

– 186 –

“syenite gneiss,” “felspathic gneiss,” and “granitic rock,” noting too their abundance of hornblende and poverty of micas (Hector, 1863, p. 441). In 1872 F. W. Hutton made a rapid reconnaissance over most of the province of Southland, and concluded that the Bluff-Greenhills area was composed of a “broad dioritic dyke” intrusive into Upper Palaeozoic slates which he considered to be the equivalents of Hector's Te Anau Series (Hutton, 1872, p. 102). In 1875, however, he called the intrusive rock a “syenite” (Hutton, 1875, p. 41). Some years later W. S. Hamilton (1877, p. 452), while confirming Hutton's views upon the intrusive nature of the “syenite,” suggested that the “dyke” extended from Ruapuke Island through Bluff Hill to New River Heads.

In the following year Professor James Park made a geological survey from Bluff to Green Hills. He regarded the invading rock as “hornblende syenite” penetrating “sandstones and slates” of the Te Anau Series. Especially interesting, since most of the metamorphic rocks are now known to be of pyroclastic origin, was his recognition of “diabasic ash breccia” in the Te Anau rocks exposed in a quarry at Green Hills (Park, 1888, pp. 72–73). Shortly afterwards, Hutton (1889, pp. 128–129) published petrographical notes upon some of the rocks of the district, and later added descriptions of a further collection, which included the types “hornblende diorite,” “enstatite diorite,” and “olivine gabbro” from Bluff Hill, with “hornblende porphyrite” and “greenstone ash” from Greenhills (Hutton, 1891, pp. 354–355): the last-named is apparently the same as Park's “diabasic ash breccia.” His microslides were subsequently examined by Baron, who informed him that the term “enstatite diorite” should be replaced by “norite” since the hornblende was secondary (vide Hutton, 1899, p. 484). This was the first recognition of the noritic nature of the Bluff intrusive mass.

Eleven years later, Thomson examined microscopically a few specimens collected by him from the coastline immediately to the south-east of the Bluff wharves, i.e., the area referred to in the present paper as the “Foreshore” (Thomson, 1910, pp. 36–39). He noted the occurrence of bedded hornblende-schists (previously referred to as “slates of the Te Anau Series”) and concluded that they were probably of sedimentary origin.

L. J. Wild (1911, pp. 317–339) discussed carefully the microscopical and chemical relationships of the “rocks that outcrop on the portion of the foreshore of Bluff Harbour lying between the wharves and Starling Point, together with the related rocks of Tewaewae Point.” He concluded that the hornblende-schists of the former locality were the more strongly metamorphosed equivalents of the “porphyries” of the latter, and that metamorphism was due to the intrusion of the norite forming Bluff Hill.

With the intention of extending the work started by Wild, the present writer examined the district during December, 1932. The major portion of the microscopic investigation of the material was carried out at the University of Otago, New Zealand, during 1933, and the study was continued for a brief period during 1934 at the Royal School of Mines, London.

Picture icon

Bluff Peninsula
New Zealand

– 187 –

The writer is indebted to Professor W. N. Benson, of the University of Otago, for his advice and material aid in the work, and to Dr. F. J. Turner for helpful discussion throughout the investigation, and for reading the manuscript and correcting proofs; to Mr. D. E. S. Mason, Bluff Harbour Board Engineer, for assistance in the field; to Professor P. G. H. Boswell and Professor A. Brammall, of the Imperial College of Science and Technology; and to Dr. A. W. Groves for chemical analyses. Acknowledgment is also made of the opportunities for research provided by a Duffus Lubecki Research Scholarship (University of Otago) and later an 1851 Exhibition Science Research Scholarship; also of monetary grants from the Royal Society, London, for chemical analyses, and the Royal Society of New Zealand for field expenses.

Outline of Geology.

Bluff Peninsula is mainly composed of a lenticular intrusion of norite (approximately 9 miles × 1 mile) which extends along a N.W.S.E. line from Bluff Hill to the mouth of New River (see Pl. 36). Exposed at intervals along the eastern margin of the intrusion are the oldest rocks of the district, the Metamorphic Series (basic or semi-basic tuffs with dominant spilitic affinities, intrusive rocks, and a few true lavas, which have undergone varying degrees of contact metamorphism). These are developed in three chief localities, Greenhills, Tewaewae Point, and the Foreshore, * which will be dealt with individually in detail. Evidence will be adduced to indicate that the rocks of each group are closely connected in age and origin, and that all have suffered contact metamorphism consequent upon the intrusion of the Bluff norite.

A few minor intrusive rocks of younger age than the norite and ranging in composition from ultrabasic to acid are considered, together with the norite, as members of an Intrusive Series. On the eroded surface of the intrusive and metamorphic rocks lie Pleistocene or Recent gravels and sands of which no further mention will be made.

Though definite age cannot be assigned to the different groups of rocks, it is concluded, after comparing them with those of other parts of Southland and Otago, that the Metamorphic Series may be correlated with the (?) late Palaeozoic Te Anau Series of Western Otago, and that instrusion and metamorphism were effected probably in late Palaeozoic times, and at least before the mid-Triassic marine transgression.

Metamorphic Series.

1. Distribution.

(a) Greenhills Group.

Rocks of the Metamorphic Series occur in the Greenhills area in a band running from the mouth of Mokomoko Inlet to Bluff Harbour and passing thence through Colyer's Island down the

[Footnote] * The term “Foreshore” is here used to denote the portion of the western shore of the harbour extending for a quarter of a mile to the south-east from the foot of Henderson Street.

– 188 –

western shore of the harbour to Green Point. Although the other islands in the harbour were not visited, it is probable that they too are composed of the same rocks. The best exposures are in two quarries near Greenhills railway-station, where beds of partially-recrystallised spilitic tuff are worked for road-metal. Half a mile along a road running west from the railway-station, fine-grained hornfelsic rocks outerop for about three-quarters of a mile; their western boundary is marked by intrusions of pyroxenite and peridotite. Although the north-eastern boundary of the group is clearly marked against the gravels of the Southland Plains, the western boundary is not yet definitely known, since the writer had not the opportunity to examine either the greater part of the Greenhills Range, or the coastline stretching west from Mokomoko Inlet, past Barracouta Point to a spot two miles south-east of the latter. At the last locality, however, a peridotite is exposed as a shoal rock at low tide; a quarter of a mile to the south-east of this, are granulite-hornfelses and schists of the Metamorphic Series. These can be followed southwards for half a mile, after which they are replaced by a small intrusion of pyroxenite, which in turn gives way to the Bluff norite; the latter continues round the coast to Ocean Beach and Bluff Hill. The boundaries shown on the map between intrusive and metamorphic rocks in the Greenhills Range are therefore only approximate, and conform essentially with those given by Park (1888).

The tuffs, especially those along the shores of Mokomoko Inlet and Bluff Harbour, are well-bedded, with a general north-west strike and a steep dip to the north-east; the dip increases steadily towards the south-west. The grade of metamorphism increases also from north-east to south-west. Thus the least altered rock, collected at the mouth of Mokomoko Inlet, contains abundant fragments of its original constituents; the corresponding rocks from the Greenhills quarries have lost almost all trace of these; while the hornfelses found along the road running westwards from Greenhills station, though certainly of finer grain-size, are completely reconstituted. Though little is known of the rocks occupying the area between these slightly altered tuffs and the high-grade granulites exposed on the sea-coast to the west of them, the latter are considered to belong to the same sedimentary series since they are well-bedded, have the same general dip and strike as the tuffs to the east and north-east, and chemically are closely similar to these (see analyses nos. 52b and 209).

Intersecting the tuffs at the mouth of Mokomoko Inlet is a small dyke of trachybasalt. It has suffered minor lateral displacement and jointing along with the surrounding rocks, and has been subjected to the same low grade of contact metamorphism.

(b) Tewaewae Point Group.

The metamorphic rocks of Tewaewae Point Peninsula are exposed as shoal-water reefs at close intervals from a point 250 yards north-east of the wharf, round past Tewaewae Point, to the promontory 1100 yards to the south-south-east; no outerops other than those along the shore were found since the peninsula itself is covered by sand-dunes. Though exposed over a distance of only one mile in all, seven distinct rock types have been recognised—epidiorite, albitic

– 189 –

metadolerite, keratophyre (?), quartz-keratophyre, garnet-bearing quartz-keratophyre, albite-actinolite-schist, and metabasalt. A sketch map of Tewaewae Point Peninsula is appended, showing the localities of the specimens collected, and the probable continuation of the rocks beneath the sand dunes (Text-fig. 1).

Picture icon

Text-Fig. 1.—Sketch Map of Tewaewae Point Peninsula showing localities of rock specimens collected, and the probable continuation of the rock-masses beneath the sand-dunes.

  • E—Epidiorite

  • A—Albitic Metadolerite

  • S—Keratophyre (?)

  • K—Quartz-keratophyre

  • K—Garnet-bearing Quartz-keratophyre

  • H—Albite-actinolite-schist

  • M—Metabasalt

  • G—Granite

The oldest members of the group appear to be massive epidiorites which form the promontory 1100 yards south-south-east of Tewaewae Point, occupy the coastline for 250 yards midway between here and Tewaewae Point, and outcrop again at the wharf.

Exposures of massive albitic-metadolerite occur on the promontory 250 yards north-east of the wharf (No. 184), and again on the shore to the south-east across the peninsula, at a point 700 yards from Tewaewae Point (No. 191). It is possible that both occurrences are part of a single dyke, striking north-west and south-east, and intersecting the epidiorites.

Massive quartz-keratophyres occupy the greater part of the remainder of the coastline in the peninsula—from the wharf to Tewaewae Point and thence for 700 yards to the south-south-east, where they adjoin albitic metadolerite. About 200 yards east of this point a small dyke of quartz-keratophyre intersects epidiorites. This is of interest since on the Foreshore, where the metamorphic grade is much higher, acid granulites somewhat comparable with the quartz-keratophyres

– 190 –

are seen to invade hornblende-schists comparable with the epidiorites. A dyke of garnet-bearing quartz-keratophyre 8 feet in thickness, with a strike of N. 27° W. and dip of 43° to the south-west, intersects epidiorite 850 yards south-south-east of Tewaewae Point.

Albite-actinolite-schist was noted at the contact of quartz-keratophyres and epidiorites about 400 yards south-south-east of Tewaewae Point. The schist is in a band 2 feet wide, striking N. 13° W. and dipping at 75° to the west.

At a point 600 yards south-south-east of Tewaewae Point, a rock tentatively described as “keratophyre” occurs as a dyke, about 6 feet thick, cutting epidiorite. The strike is N. 14° W. and dip 15° to the west. The final type represented on the peninsula, metabasalt, is present in two dykes: the first (No. 188), about 3 feet thick and with north-north-west strike and steep dip to the south-west cuts through epidiorite on the promontory 1000 yards south-south-east of Tewaewae Point; and the second (No. 182), 20 feet thick with strike W. 40° N. and steep dip to the south-west, intersects quartz-keratophyre 250 yards south-south-east of Tewaewae Point.

All the above rocks show the effects of low-grade thermal metamorphism. The relative ages of the various members cannot be determined completely, but it is almost certain that each was in place prior to the major intrusion of the Bluff norite, and has suffered partial reconstitution as a result of it.

(c) Foreshore Group.

Exposed at low tide along the Foreshore is a strongly-banded series of dark-green schists and hornfelses, striking about 15° north of west and dipping steeply southwards at angles between 78° and 90°. The banding which is due to alteration of dark-green richly hornblendic rocks and lighter yellowish-green feldspathic members, lends to the group the superficial aspect of an indurated series of bedded sediments; most of the banding, however, appears to be due more to metamorphic differentiation and diffusion than to sedimentation.

The chief petrographic types represented are horneblende-andesine-schists and hornblende-oligoclase-schists, with minor bands of hornblende-pyroxene-schists. Metamorphism has obliterated any signs of unconformity between the bands, and it is therefore impossible to speculate upon the relative ages of the different types. In three places, respectively 500, 550 and 700 yards south-east of Henderson Street, are outcrops of an acid granulite. In the first and third the granulites appear to be bedded among the green schists, but their intrusive nature is shown in the second, where a lenticular patch of granulite cuts obliquely across the foliation-planes of a hornblende-biotite-schist. The granulites are similar chemically to the quartz-keratophyres of Tewaewae Point. At intervals along the Foreshore, and forming dykes cutting both the hornblende-schists and the acid granulite, is a coarse-grained amphibolite which appears to have been originally a coarse dolerite. No low-grade equivalent of this rock was found anywhere in the area.

– 191 –

On their southern margin the metamorphic rocks of the Foreshore are overlain by a boulder-strewn beach which unfortunately obscures their contact with the Bluff norite. On their western margin also the contact is hidden beneath a raised beach-platform which stretches for some 300 yards from the water-front to the foot of Bluff Hill.

On the harbour shores between the wharves and Ocean Beach railway-station are a few small exposures of green schists and hornfelses, which, on account of their limited extent and similarity to the rocks of the Foreshore, are classed with the latter in preference to assigning them to a distinct additional group.

2. Petrology.

(a) Greenhills Group.

(i) Partially-reconstituted tuffs. The majority of the specimens examined are relatively coarse-grained non-schistose rocks showing traces of their original fragmental structure (e.g., Nos. 52a, 52b* … to 52i, 111–122, 124). The size of the individual fragments varies usually between 0.2 mm. and 2.0 mm; occasionally, however, clastic fragments up to 6.0 mm. in diameter have been observed (e.g., in Nos. 52h and 52i). The small amount of original material includes abundant fragments of altered igneous rocks, acid plagioclase, a little pale-green augite and purplish titanaugite, and accessory apetite and iron-ore. Though the volcanic fragments are partially reconstituted, their original structures, such as trachytic arrangement of the feldspar lath (e.g., Nos. 52d, 124) porphyritic, vesicular and amygdaloidal structures (e.g., Nos. 52h, 124) are still visible. The clastic components are set in a completely recrystallized base of acid plagioclase, epidote, chlorite, actinolite, and greenish-blue hornblende, accompanied in most slices by accessory sphene and pyrite; in a few instances there are small amounts of secondary quartz. It is frequently possible to see direct replacement of one or other of the clastic constituents by crystalloblastic products. Commonly observed changes of this kind are: augite to chlorite and epidote; augite to actinolite; ilmenite to sphene.

From the above general description it is clear that the rocks under discussion, though of detrital origin, have been derived from basic or semi-basic igneous rocks, and are therefore either tuffs or greywackes. That they are tuffs rather than greywackes is suggested by the apparently complete absence of any clastic grains of quartz such as are abundant and persistent throughout the greywackes from the older formations of the South Island (ef. Turner, 1933a, pp. 215, 231; 1935a, p. 335). It is certain, however, that some at least were deposited under marine conditions, for a single specimen of a coral was found by a quarryman in one of the Greenhills quarries and identified by Dr. R. S. Allan, of Canterbury College, as being probably a Zaphrentid). The regular stratification over wide areas points to

[Footnote] * For analysis see p. 192.

[Footnote] † Thanks to the courtesy of the Curator of the Southland Museum, Dr. Benson has examined the matrix of this specimen microscopically: it is a fine-grained altered spilitic tuff composed essentially of albite, dusty or finely granular epidote, abundant minute partially chloritised flakes of biotite and uralitic tremolite.—ED.

– 192 –

deposition in water, and it is therefore possible that the rocks have originated, as F. C. Phillips considers the Green Beds of the Scottish Dalradian to have done, through “fairly direct derivation from basic igneous rocks and deposition without much admixed material” (Phillips, 1930, p. 240).

The basaltic nature of the material composing the tuffs is quite normal except for the fact that there is no sign of original basic plagioclase. Actually, the few recognisably allogenic feldspars all have the composition of basic albite (averaging Ab93 An7) while the crystalloblastic plagioclase is slightly more sodic (averaging Ab95 An5). This prominence of albite immediately suggests spilitic affinities for the tuffs, a forecast which is borne out by chemical analysis. Table I shows analyses of a representative tuff from Greenhills, an average basalt and an average spilite. Bearing in mind variation among the individual analyses of spilites, the most apparent departure from the average in the analyses of tuff here considered is the low percentage of TiO2; this deficiency might, if necessary, be explained as being due to local impoverishment in ilmenite, titanaugite, etc., by gravitational sorting of the settling tuff-material or by gentle current action during deposition.

[The section below cannot be correctly rendered as it contains complex formatting. See the image of the page for a more accurate rendering.]

Table I.
(a) (b) (c)
SiO2 53.56 49.06 51.22
TiO2 0.46 1.36 3.32
Al2O3 18.66 15.70 13.66
Fe2O8 3.21 5.38 2.84
FeO 5.97 6.37 9.20
MnO 0.14 0.31 0.25
MgO 3.54 6.17 4.55
CaO 6.56 8.95 6.89
Na2O 6.57 3.11 4.93
K2O 0.76 1.52 0.75
H2O 0.72 1.62 1.88
P2O5 0.17 0.45 0.29
CO2 0.94
——— ——— ———
100.32 100.00 100.72

Tuff (52b), Greenhills, Bluff Peninsula, N.Z. Anal. by H. Service.


Basalt. Average of 198 analyses (Daly, 1933, p. 17).


Spilite. Average analysis (Sundius, 1930, p. 9).

(ii) Andesine-actinolite-hornfels. Two specimens of this type of rock (Nos. 53, * 55) were collected at points on the roadside 1100 and 1300 yards, respectively, west from Greenhills railwaystation. In hand-specimen both are fine-grained, dark greenish-grey rocks, with no trace of schistosity or foliation. In thin section, they are almost identical, differing only in grain-size. The coarser rock (No. 55) is a completely recrystallized hornfels, containing about 40% of interlocking idioblastic prisms of pale-green actinolite, and about 10% of brownish-green biotite developed as small ragged flakes moulded around prisms of actinolite. There are a few porphyroblastic fibrous or prismatic crystals of an amphibole of slightly deeper colour than the actinolite; these are frequently

[Footnote] * For analyses see table at end of paper.

[Footnote] † Following definitions of Harker (1933, p. 203).

– 193 –

concentrated along the cleavage-planes in association with flakes of biotite. The base of the rock is made up of finely-granoblastic, rarely-twinned plagioclase with a composition near to medium andesine. Spongy grains of iron-ore and yellow epidote are minor accessories. The fine even grain of the hornfelses and their completely reconstituted state indicate that the parent rocks were themselves fine-grained, probably either basic ash or sediment of corresponding composition.

(iii) Oligoclase-trachybasalt. A dyke of oligoclase-trachybasalt intersects the bedded tuffs at the eastern side of the mouth of Mokomoko Inlet (No. 123). It is a fine-grained, dark-grey homogeneous rock consisting of a felted mass of short laths of original plagioclase (basic oligoclase), with interstitial minor secondary chlorite, epidote and albite-oligoclase, and accessory iron-ore, sphene, haematite, perovskite and rutile. The original plagioclase has undergone no alterations other than the development of minute dust-like inclusions, and incipient marginal recrystallization to a mosaic of clear granules of albite-oligoclase. Moulded around the plagioclase laths, and often projecting into their granular margins, are flakes of bright green chlorite associated with which are small grains of greenish-yellow epidote. Sphene is an abundant accessory as small grains often seen replacing grains of original ilmenite. The assemblage of reconstituted minerals shows that the rock has been subjected to low-grade metamorphism. (Compare the tuffs described above.)

(iv) Pyroxene-hornblende-biotite-albite-granulite and Garnet-calcite-pyroxene-albite-granulite. A banded specimen containing these two types (No. 209) * was collected from the bedded metamorphic rocks outcropping on the west coast of the Greenhills Range two miles south-east of Barracouta Point. The major portion is a fine-grained, dark green, non-schistose rock showing a trace of foliation; bounding this along one face is a narrow band, about half an inch in thickness, of light-brownish material containing visible crystals of a colourless garnet. The garnetiferous band lies parallel to the foliation in the green portion and to the dip of the whole series of metamorphic rocks exposed in the neighbourhood. A thin section cut perpendicularly to the plane of junction of the two bands shows the dark part to be a banded pyroxene-hornblende-biotite-albite-granulite, whereas the light part is a granulite composed predominantly of a pale-green garnet with small amounts of pyroxene, calcite and albite.

The dark-green granulite is a holoblastic, fine-grained rock containing bands respectively rich in pyroxene, hornblende and biotite. The dark minerals are set in a granoblastic base of water-clear rarely-twinned albite-oligoclase together with a very small amount of quartz. Pale-green slightly pleochroic pyroxene (diopside rich in heden-bergite) occurs in globular grains averaging 0.015 mm. in diameter. The hornblende bands are characterized by roughly parallel, subidioblastic prisms of hornblende with the following pleochroism: X and Y = brownish-green; Z = bottle-green, X < Y < Z; extinction angle, Z to c = 26°. Many of the hornblendes are sieve-like

[Footnote] * For analysis see Table II, p. 195.

– 194 –

with small included grains of plagioclase. Reddish-brown biotite is present to a minor extent as small xenoblastic flakes in both hornblendic and pyroxenic bands, but is concentrated in some areas to the almost complete exclusion of the other two minerals.

The garnetiferous band contains about 90% of a clear, faintly-green garnet which, though frequently occurring as porphyroblasts up to 0.5 mm. in diameter, is usually in grains of about one-tenth that size; the porphyroblasts contain a few small inclusions of pyroxene and plagioclase, while the smaller grains of garnet are associated in a finely granoblastic aggregate with grains of calcite, pale-green pyroxene and acid plagioclase. Qualitative chemical tests upon the garnet prove that it is a lime-rich type containing a little manganese. The pyroxene has properties agreeing with those of a fairly pure diopside. Calcite is common along with albite (Ab94. An6) as small xenoblastic grains moulded against the pyroxene and garnet; in a few areas large pools of calcite and plagioclase, which are clear in their central portions, contain scattered granules of pyroxene and garnet towards their margins. Embedded in places in the calcite and along cracks in the garnet are very fine fibres and short prisms of a brownish-pink mineral referred to thulite. The fibres, which frequently mass together as small spherulites, have parallel extinction, refringence slightly higher than that of the ordinary ray in the calcite (about 1.7), low birefringence and very faint pleochroism in shades of brownish-pink. The abundance of calcite in this band. and the lime-rich varieties of pyroxene and garnet, suggest that the original rock was a calcareous band interbedded in the tuffs.

The pyroxene-hornblende-biotite-albite-granulite is of interest on account of its uncommon mineralogical constitution. It is well established that in basic igneous rocks undergoing either thermal or regional metamorphism of low grade, pyroxenes are unstable, and that they appear as crystalloblastic minerals only at high grades (e.g. see Harker, 1932, pp. 108, 110, 284). The stable plagioclase in such rocks, when adjusted to high-grade conditions, is a basic variety, typically labradorite (e.g. see Harker, 1932, pp. 93, 284; Tilley, 1924, pp. 65, 66, etc.). The Greenhills granulite, essentially basaltic in composition, must have been metamorphosed at high grade, judging from its high content of crystalloblastic pyroxene and its completely recrystallised condition (all traces of original igneous texture having been obliterated). The plagioclase (albiteoligoclase) is, however, much less calcic than is usual in such rocks, and the same is true for two specimens of adjacent and interbedded granulite (Nos. 210, 211). This mineralogical indication of spilitic origin is confirmed by chemical analysis of the granulite from Greenhills (Table II). Further, field evidence suggests that the granulites in question belong to the same series as the tuffs described from the north-eastern flank of the range and shown by chemical analysis to be spilitic in composition. Thus, although the writer is unaware of any previous record of similar pyroxene-granulites containing albite or oligoclase in place of the usual labradorite, it seems clear in the present instance that the Greenhills pyroxene-granulites are high-grade equivalents of the low-grade spilitic tuffs further north-east,

– 195 –

and that the high percentage of soda in the initial rock is responsible for crystallisation of acid plagioclase even at high metamorphic grade.

[The section below cannot be correctly rendered as it contains complex formatting. See the image of the page for a more accurate rendering.]

Table II.
SiO2 53.31 53.56
TiO2 0.98 0.46
Ai2O3 16.43 18.66
Fe2O2 2.12 3.21
FeO 6.29 5.97
MnO 0.18 0.14
MgO 4.26 3.54
CaO 9.16 6.56
Na2O 5.19 6.57
K2O 0.88 0.76
H2O 1.73 0.72
P2O5 0.14 0.17
——— ———
100.67 100.32

Pyroxene-hornblende-biotite-albite-granulite (No. 52b), Greenhills, N.Z. Anal. A.W. Groves.


Spilitic Tuff (No. 209), Greenhills, Bluff Peninsula, N.Z. Anal. H. Service.

Except for the composition of the feldspar there is some similarity between the granulites described above and the pyroxene-granulite-hornfelses and beerbachites recently reviewed by A. G. MacGregor (1931b, pp. 506–521), who shows that these rocks are metamorphic derivatives of basic igneous rocks. He notes, however, the possibility that beerbachites might also be produced by metamorphism of “diabasic or basaltic ash, deposited in water, and quite uncontaminated by normal sedimentary detritus.” The Greenhills tuffs appear to satisfy this latter condition in that they were laid down, in part at least, under subaqueous conditions with little or no admixture of quartz.

(v) Hornblende-pyroxene-oligoclase-garnet-granulite. A single specimen of this type of rock, No. 210, was collected from the bedded metamorphic rocks exposed on the coastline, 2 miles south-east of Barracouta Point. It is a fine-grained dark-green rock with a trace of foliation. In a section cut perpendicularly to the foliation-planes it consists predominantly of elongated parallel lenticles of finely granular hornblende separated by oligoclase-rich bands. The hornblende, which forms about 50% of the rock, has the following pleochroism: X = pale greenish-yellow, Y = brownish-green, Z = dark bottle-green; X < Y < Z; Z to c = 22°. The hornblendic areas contain abundant small grains of pale-green pyroxene, a pink garnet, and minor small grains of haematite opaque iron-ore and cloudy sphene. The garnet, which is restricted to hornblendic areas, occurs as porphyroblasts reaching 0.5 mm. in diameter, which include sievelike small grains of hornblende, and by force of crystallisation have distorted the adjacent foliae with consequent development of an augen-like structure. Separating the hornblende-rich bands are areas of finely granoblastic rarely twinned plagioclase (acid oligoclase) which contain only small amounts of dark minerals. A few small porphyroblasts of plagioclase have clear margins surrounding cores packed with inclusions of hornblende and frequently clinozoisite:

– 196 –

no difference in composition between core and margin of the plagioclase is discernible. A single crystal of pyroxene 0.4 mm. long appears to be a relict of igneous origin: its central portion, yellowish-brown, is surrounded by a narrow zone charged with a fine dusty iron-ore and this again is enveloped by a zone of green hornblende containing coarse iron-ore dust.

(vi) Hornblende-plagioclase-biotite-schist. This rock (No. 213) occurs as a small intrusion cutting irregularly through the bedded hornblende-granulites described above from the western coast of the Greenhills range. It is a more basic rock than the granulites—one which might originally have been a normal basalt or dolerite, in contrast with the spilitic nature of the rocks from which the granulites have been derived. It is slightly foliated; the main constituents are bluish-green hornblende and plagioclase (basic andesine); less abundant minerals are biotite, pale-green pyroxene, magnetite, haematite, rutile and quartz. Hornblende (about 30%) is in subidioblastic, roughly parallel prisms ranging up to 0.5 mm. in length; the larger prisms, which are usually concentrated in clusters, are evidently the product of direct replacement of large crystals of original pyroxene. The hornblende is similar to that in the schists of the Foreshore: X = pale greenish-yellow, Y = grass-green, Z = bluish-green, X < Y < Z; to c = 20°. The enclosing matrix (60%) consists of clear xenoblastic grains of basic andesine, usually untwinned and free from inclusions. Small flakes of biotite (X = pale greenish-yellow, Z = dark brownish-green) occur sparingly in most foliae, but locally are relatively plentiful.

(b) Tewaewae Point Group.

(i) Epidiorite. The epidiorites (Nos. 177–180, 190) are composed dominantly of coarsely crystalline basic plagioclase and secondary hornblende; the former is usually in excess of the latter, but many of the darker types contain up to 50% of hornblende. Pale-yellow epidote is usually present to the extent of only about 5% (rarely 30%) and in one rock (190) is almost absent. There is about 5% of quartz in all specimens. Iron-ores, apatite, biotite and sphene are found as minor constituents.

A typical specimen (190) * is a coarse-grained, dark-greenish grey massive rock weathering to a steely black. It is much altered, but still shows an original gabbroid or gabbro-porphyritic texture. It consists mainly of stout prismatic crystals of basic plagioclase (75%) and irregular, stumpy frayed prisms of deep bluish-green hornblende, with minor amounts of quartz, biotite, epidote, sphene and iron-ore. The plagioclase frequently exhibits normal zoning; the central zones of andesine-labradorite being crowded with small unoriented prisms of hornblende, grains of iron-ore, sphene and sometimes cloudy clinozoistic epidote, while the outer less basic portions are corroded and granulated. Furthermore, whereas the inner zones are clouded with minute opaque dust-like inclusions lying in more or less well-defined lines parallel to the traces of the albite-twinning lamellae, the outer zones are almost free from them. About 20% of the rock is composed of blue-green hornblende of two distinct

[Footnote] * For analysis see table at end of paper.

– 197 –

habits: firstly, broad prisms up to 0.75 mm. in length, which have irregular, frayed margins and are of the same textural order as the plagioclases; and, secondly, fine prismatic or acicular crystals, forming a fringe to the larger prisms, or completely isolated in the groundmass, or penetrating into the feldspars often as interlocking, subradiate tufts, enclosed within the calcic cores. The first variety, seen by its frequently developed octagonal cross-sections to be pseudomorphous after pyroxene, is strongly pleochroic, with X = greenish-yellow, Y = brownish-green, Z = grass-green with a tinge of blue; X < Y < Z; Z to c = 22°. Along some of the cleavages and fracture-planes, the mineral is irregularly blotched with dark-brown and shows strong pleochroism as a result of incipient replacement by biotite. The acicular hornblende on the other hand has a deep greenish-blue colour for the Z direction and is thus perhaps a more sodic variety. Granular quartz occurs in the groundmass as small clear patches enclosing long needles of hornblende and apatite. Iron-ores are present both as spongy grains of clouded ilmenite associated with the hornblende pseudomorphs and as granules in the groundmass. Also in the groundmass are numerous small areas of a pseudomyrmeketic intergrowth of quartz and sodic plagioclase. A few small flakes of greenish-brown biotite are developed partly round the crystals of hornblende and partly as fringes to some of the larger grains of iron-ore. In one specimen (178) in which traces of original texture are masked by the abundance of hornblende there is 10% of pale yellow epidote, while another rock (179) * contains about 40% of epidote.

In all the specimens examined the plagioclase ranges in composition between basic andesine and basic labradorite, and in every case the crystalloblastic plagioclase is only slightly more acid than the original feldspar. The abundance of epidote in some rocks and its almost complete absence in others is attributed to migration during metamorphism. Similar mobility on the part of epidote has often been noted, e.g., by Sundius, in spilites from the Kiruna district (1930, p. 3). The clouding of the plagioclase, already briefly referred to, calls for further comment. Such clouding has frequently been recorded, e.g., Williams (1886, p. 21) in the hypersthene-gabbros of Baltimore, Tilley (1921, p. 86) in the Lincoln gneisses of South Australia, Joplin (1933, pp. 152–156) in the gabbros of the Hartley district, New South Wales. Recently, MacGregor (1931a, pp. 524–538) has studied in detail the significance of this same feature with special reference to his own observations in Ayrshire and Aberdeenshire, and after consideration of some score of examples cited by other authors, has concluded that in most cases it is an effect of contact thermal metamorphism. Several instances are given by MacGregor in which the clouding affects only the more basic portions of zoned crystals, and he also notes that the most common orientation of the strings of inclusions is parallel to the traces of the albite-twinning lamellae. In both these respects the clouding in the feldspars of the rocks here described is similar to that attributed by MacGregor to contact metamorphism.

[Footnote] * For analysis see table at end of paper.

– 198 –

The following explanation of the present composition and structure of the epidiorites is offered: (a) The epidiorites were originally gabbroidal rocks consisting of large zoned crystals of basic plagioclase and of aluminous pyroxene. (b) Conditions of low-grade contact metamorphism then followed, with resultant conversion of the pyroxene to hornblende (with liberation of much epidote), clouding of the feldspar, and recrystallization of the groundmass. (c) Following the contact metamorphism, there appears to have been a minor hydrothermal phase during which the partially hornfelsed and still heated rocks were permeated by weakly sodic solutions. The results of this latter phase were (1) almost complete obliteration of hornfelsic texture in the groundmass; (2) formation of a little clinozoisitic epidote in the large feldspars, with partial destruction of their clouding; (3) growth of acicular sodic hornblende crystals, not only round the margins of the pseudomorphous hornblendes, but also within the large feldspars and in the groundmass; (4) segregation of areas of clear, granular quartz in the groundmass; (5) formation of pseudomyrmekitic intergrowths of quartz and acid plagioclase; (6) alteration of ilmenite to cloudy or granular sphene. This interpretation conforms with the metamorphic history of the district as a whole, since every rock older than the Bluff norite has suffered contact metamorphism accompanying intrusion, while the occurrence of sodic veins traversing the norite of Bluff Hill shows that sodic solutions following intrusion have been active elsewhere (see p. 210).

The penetration of the plagioclase by acicular hornblende is believed to afford an example of metamorphic diffusion as described by Stillwell (1918), Eskola (1932) and others. Since the hornblende is of the same sodic species as that forming fringes to the pseudomorphous hornblendes, its entry into the plagioclase is considered to have been accomplished during the final hydrothermal phase.

(ii) Albitic Metadolerite. The two specimens (Nos. 184, 191) are medium to coarse-grained massive rocks consisting of albite (80%) and hornblende (20%) with accessory epidote, iron-ore, biotite, sphene, apatite and quartz. In 184* the albite (Ab92 An8) is in the form of broad subhedral prisms ranging up to 2.5 mm. in length ophitically enclosed in coarsely crystalline hornblende. In a few crystals there are signs of incipient marginal recrystallization, while inclusions of granular clinozoisite and dusty iron-ore are plentiful, and secondary slight kaolinization and development of iron-stain are fairly general. The remainder of the rock is composed chiefly of hornblende, which belongs, as in the epidiorites, to two generations, the first represented by pseudomorphs after original pyroxenes, and the second by more sodic fine acicular crystals which fringe the pseudomorphs, penetrate the plagioclases, and form an integral part of the reconstituted base. The first is strongly pleochroic with X = yellowish-green, Y = olive-green, Z = green with a bluish tinge, while the second shows similar colours for X and Y, but is a deep greenish-blue for the Z direction. Iron-ore includes long skeletal crystals of ilmenite slightly altered to sphene, and small compact

[Footnote] * For analysis see table at end of paper.

[Footnote] † From the present composition of the rock it is obvious that the original feldspar prior to metamorphism must also have been albite.

– 199 –

grains which are probably magnetite: Around the margins of many of the grains of iron-ore, and lying along the cleavages of some of the larger crystals of hornblende, are small flakes of greenish-brown biotite. Apatite and sphene occur sparingly as accessory fine needles and minute grains respectively. The small amount of groundmass in the rock is completely reconstituted to a granular aggregate of quartz, albite and hornblende. There are also many small pseudomyrmekitic intergrowths of quartz and albite, reminiscent of those noted above in the epidiorites.

(iii) Quartz-keratophyre. Quartz-keratophyres are represented by two specimens collected respectively at Tewaewae Point (189), * and on the coast 400 yards to the south-south-east. They are coarse-grained, yellowish-green rocks with conspicuous phenocrysts of feldspar and small dark crystals of hornblende. No. 189 contains about 20% of subhedral phenocrysts of orthoclase and albite (Ab93 An7) reaching 1.5 mm. in length. Infrequency of twinning in the albite makes estimation of its abundance difficult, but the high percentage of soda indicates that albite predominates over orthoclase, or that the latter is a soda-bearing type. The phenocrysts are set in a hornfelsic groundmass consisting mainly of orthoclase and albite, together with quartz and a little hornblende (Z = bluish-green) and biotite. Small greenish-brown xenoblastic flakes of biotite are common in the groundmass. The other minor constituents are grains of pale green epidote, sphene, apatite, opaque iron-ore and haematite. The second specimen (No. 187) is a finer-grained more completely hornfelsic rock, a few phenocrysts of quartz and albite alone being preserved unchanged. The analysis confirms the microscopic evidence that these rocks are the same as Wild's “porphyries” (1911, p. 327).

(iv) Garnet-bearing quartz-keratophyres. Specimen No. 183, from a dyke cutting the epidiorites half a mile east of Tewaewae Point, is a fine-grained light-grey rock showing numerous small pink garnets. It contains subhedral prismatic phenocrysts of acid plagioclase up to 1.5 mm. in length, and rounded grains of garnet averaging 0.3 mm. in diameter. These are set in a partially recrystallized, orthophyric groundmass of acid plagioclase (80%), biotite and hornblende (5%–10%), and clear interstitial quartz (10%). The phenocrysts of plagioclase (? medium oligoclase) contain inclusions of dusty iron-ore and pale-yellow epidote, and show incipient marginal recrystallization to clear granules of albite-oligoclase. The garnet occurs in somewhat spongy grains which have usually developed in the groundmass, or less commonly within the large plagioclases. The plagioclase of the groundmass (albite-oligoclase) is in small stout prisms clouded with minute dusty inclusions and partly recrystallized around their margins. Dark constituents include small flakes of greenish-brown biotite, prisms of bluish-green hornblende, and a little bright-green chlorite. Accessories are slender prisms of apatite and small grains of iron-ore.

From the total recrystallization of the mafic constituents, and the incipient recrystallization of the plagioclase, it is clear that this rock has undergone the same low-grade contact metamorphism as the epidiorites that it intersects. The original orthophyric texture

[Footnote] * For analysis see table at end of paper.

– 200 –

of the groundmass can still be discerned, while the plagioclase phenocrysts have not been lost. To ascertain the nature of the garnet, a qualitative blowpipe test was carried out, and a strong reaction for manganese obtained. The stability of pink garnet rich in spessartite (in contrast with the instability of pure almandine) in contact metamorphism has been demonstrated by Tilley (1923, pp. 190, 198; 1926, pp. 47–50). In the Bluff rock the garnets that are surrounded by minerals of the groundmass are typically porphyroblastic, but their growth within the original phenocrysts of plagioclase is unusual. Similar crystalloblastic growth of garnet within feldspars has been recorded by Wiseman (1934, p. 388) in epidiorites from the South-West Highlands of Scotland.

(v) Keratophyre (?). The single specimen (No. 185) is a fine-grained, light-green rock showing in section a few pseudomorphs after original phenocrysts of pyroxene and plagioclase, set in a recrystallized base of acid plagioclase, quartz, acicular hornblende, red-brown biotite, chlorite and epidote, together with accessory magnetite and perovskite. The pseudomorphs after pyroxene are composed essentially of granular iron-rich epidote and small flakes of deep-green chlorite. Ghosts of the original albite twinning in the plagioclase phenocrysts, which are largely altered to a pale variety of epidote, permits determination as medium oligoclase. Fully 90% of the base consists of granoblastic, rarely twinned plagioclase (medium oligoclase) in which traces of an original trachytic texture can be observed. This rock might perhaps be considered a very feldspathic keratophyre.

(vi) Albite-actinolite-schist. No. 186 is a very fine-grained, light-green, somewhat schistose rock, consisting of a granoblastic aggregate of albite, actinolitic hornblende, magnetite and biotite. Two-thirds of the rock is composed of xenoblastic usually untwinned grains of albite surrounding stout prisms of actinolite and ragged flakes of dark greenish-brown biotite. Magnetite occurs in very small grains which tend to form streaky aggregates and make the rock somewhat schistose. There is no trace of epidote, and quartz was not recognised with certainty.

The field-relations of the rock are not clear, but it is probably a sheared keratophyre.

(vii) Metabasalts. No. 188 is a dark-greenish-grey massive rock, with abundant sub-parallel phenocrysts of hornblende reaching 10.0 mm. in length, enclosed by a fine-grained trachytic groundmass rich in hornblende. The hornblende phenocrysts often possess the octagonal cross-sections and the schiller-and salite-structures of pyroxene, after which they are evidently pseudomorphous. As in the epidiorites described above the phenocrysts possess an inner, compact zone surrounded by a frayed and fibrous border. The compact hornblende exhibits the following pleochroism: X = greenish-yellow, Y = olive-green. Z = brownish-green; X < Y < Z. The amphibole of the outer zones and also that of the groundmass differs in showing a deep greenish-blue colour, for the Z direction. There are a few stout subhedral phenocrysts of basic plagioclase (Ab50 An50) often zoned, and not infrequently clouded with a fine brown dust or partially replaced by small grains of colourless epidote. Traces of original trachytic

– 201 –

texture are preserved in the largely reconstituted groundmass—a plagioclase-hornblende-hornfels containing small amounts of quartz, biotite, sphene, clinozoisite and hematite. Granular patches of clear quartz, containing small grains of hornblende, biotite and clinozoisite, are present in many parts of the slide.

(c) Foreshore Group.

(i) Hornblende-andesine-schists. * Nos. 35, 36, 38–41, 45, 48, 50c, 52, 61, and 65 are dark greenish-grey rocks, often schistose, but generally compact and tough. They are mainly fine-grained, and on the weathered surface phenocrysts of feldspar ranging up to about 3.0 mm. in diameter are commonly left projecting from the base. Typical specimens contain blastophenocrysts of plagioclase (usually about 5%, occasionally 20%) set in a schistose base of plagioclase and hornblende. These large relict crystals constitute augen around which are moulded the finer mineral of the base; clear, granular quartz has usually crystallized in wedge-shaped areas at the ends of these augen. The base is made up of a granoblastic aggregate of water-clear crystals (0.02 mm.) of andesine, in which are set less abundant, sub-parallel slender prisms of green hornblende. Larger porphyroblastic bluish-green hornblendes (up to 2.0 mm.) frequently are found with their long axes lying parallel to the direction of schistosity, and dense masses of the smaller prisms doubtless take the place of the original ferromagnesian minerals in the rock. Quartz which is normally present in the groundmass to about 2% or 3% may take up locally as much as 40% in thin quartzo-feldspathic bands. Small grains of iron-ore and sphene are constant accessories, and pale-yellow granular epidote is abundant in the blastophenocrysts of plagioclase in some slides (e.g., 48a). In other slides (35, 39) are small crystals of (?) perovskite and in No. 65 a few fine prisms of apatite. Biotite (up to about 10%) occurs as small flakes clustered round the edges of the hornblende prisms, and shows strong pleochroism from pale straw colour to deep greenish brown. Evidence that biotite has formed in part at least at the expense of hornblende is afforded by its presence along cleavage-planes in some of the hornblende phenocrysts (41, 50c). In some sections (e.g., No. 45) there is a concentration of biotite around the larger grains of iron-ore, especially where the latter abut against hornblende.

The blastophenocrysts of plagioclase are usually too much altered to allow accurate determination of their composition, but the albite-twinning preserved in several rocks indicates andesine-labradorite, while in one case (No. 65) they are unaltered and are a basic labradorite. The chief alternation-product which obscures the twinning is a colourless weakly birefringent clinozoisite, but direct recrystallization has sometimes transformed the original phenocrysts into finely granular masses of clear secondary plagioclase. In one rock (No. 48a) fresh phenocrysts of plagioclase contain strings of minute, rod-like opaque inclusions, arranged in lines parallel to the traces of the albite-twinning lamellae (ef. Tewaewae Point epidiorite). In this slide, and also in No. 65, the feldspars show normal zoning, and while the less basic zones are free from inclusions, the cores of basic labradorite

[Footnote] * For analysis of 48a see table at end of paper.

– 202 –

are packed with small grains of clinozoisite. In No. 36 some of the phenocrysts are pierced by fine needles of pale greenish-blue hornblende, a feature again reminiscent of the Tewaewae Point epidiorites. The recrystallized plagioclase of the base is free from inclusions, and generally untwinned. Its composition remains fairly constantly about medium andesine (Ab60 An40, and in all cases it is slightly more acid than the phenocrysts.

(ii) Hornblende-acid oligoclase-schists. (Nos. 42, 44, 49, 51, 62b). In the field it is impossible to differentiate between these rocks and the hornblende-andesine-schists just described, and in thin sections both types exhibit identical textural features. However, in the rocks now under discussion both the relict and recrystallized plagioclase are acid oligoclase. A typical specimen (No. 51)* contains alternating coarse quartzo-feldspathic and finer-grained hornblende bands, the former consisting essentially of quartz 50%, acid oligoclase 40% and hornblende 10%, and the latter of acid oligoclase 60% and hornblende 40% with relatively plentiful idioblastic sphene. In the hornblendic bands the texture is strongly schistose owing to the parallelism of the short prisms of blue-green hornblende, but the quartzo-feldspathic layers are more granulose, and the hornblende here occurs as large sieve-like porphyroblasts with no common orientation. Small blastophenocrysts of acid oligoclase (Ab83 An17), crowded (as in the hornblende-andesine-schists) with needles of pale greenish-blue hornblende and opaque dust-like inclusions, are present in some of the dark-coloured bands. The clear, untwinned recrystallized feldspar of the base appear from its low refractive index to be slightly more acid than the relict plagioclase (cf. the quartzo-feldspathic bands where the composition is Ab87 An13). The characteristic abundant bluish-green hornblende occurs mainly as parallel, subidioblastic prisms in the base. Associated with some of the larger prisms are small grains of iron-ore. Apatite and sphene are minor accessories. As in the hornblende-andesine-schists, biotite may make up as much as 10% of some rocks: it is the same greenish-brown type as in the more basic schists, and often appears to be developing at the expense of hornblende.

The similarity in character of the blastophenocrysts of plagioclase in these schists to the plagioclase crystals in the albitic metadolerites of Tewaewae Point suggests that the former are higher grade metamorphic equivalents of the latter. Comparison of analyses strengthens this view (see table of analyses, p. 215).

(iii) Hornblende-pyroxene-labradorite-schists. (Nos. 37, 50a, 50b, 60 (ii), 160, 163, 166). These schists, characterised by the presence of crystalloblastic pale-green pyroxene, are found at intervals on the Foreshore along the harbour-front between Bluff wharves and Ocean Beach. Interbedded among the hornblende-schists, they show in the field no distinguishing features. In hand-specimen gradations are seen from very fine-grained dark-green hornfelsic rocks with conchoidal fracture (No. 37), through markedly fissile somewhat coarse-grained rocks showing no banding (Nos. 50a, 160, 166), to strongly schistose rocks consisting of alternating dark-green and light-yellow bands (Nos. 50b, 60 (ii), 163).

[Footnote] * For analysis see table at end of paper.

– 203 –

No. 163* is a holoblastic fine-grained rock in which segregation has resulted in the development of parallel bands of three main types, viz., quartzo-feldspathic, hornblendic and pyroxenic. The first consists of granoblastic aggregates of quartz and clouded plagioclase (Ab40 An60) in about equal amount, as grains averaging 0.3 in diameter. The hornblendic bands contain 70% dark-olive-green hornblende as subidioblastic prisms averaging 0.1 mm. in length, set in a base of rarely-twinned plagioclase (about Ab47 An53), minor quartz and sometimes a little pyroxene. There are gradations from the dominantly hornblendic bands, through those containing about equal amounts of hornblende and pyroxene, to bands composed dominantly of the latter mineral. One such contains pyroxene (70%), plagioclase (20%), iron-ore (5%), and quartz (5%). The pyroxene is in rounded poikilitic grains averaging 0.5 mm. in diameter, which often completely surround the grains of iron-ore. It has a faint greenish tinge with detectable pleochroism, birefringence 0.028, maximum extinction angle (Z to c) = 45°, positive sign, and a rather low optic-axial angle. These properties agree with those of diopside containing between 15% and 25% of hedenbergite. Iron-ore (sometimes bordered with clear sphene) is much more abundant in the pyroxenic bands than in those rich in hornblende, and may result from the release of iron during conversion of hornblende to pyroxene. A few flakes of brown biotite are sometimes seen adjacent to grains of the ore mineral in both the pyroxenic and hornblendic bands. Schists closely similar to the above are Nos. 160 and 60 (ii); the former has about equal amounts of hornblende and pyroxene, while in the latter the ferromagnesian mineral is almost wholly of pyroxene.

The initial stages in the formation of pyroxene is seen in No. 166, a totally recrystallized rock with nematoblastic texture. About 40% of this rock is composed of subidioblastic prisms of hornblende of all sizes up to 0.4 mm. long which are arranged along parallel lines in a base consisting of equal amounts of finely granular basic plagioclase and quartz. Throughout hornblende and base alike are small round grains of pyroxene.

No. 37 is a fine-grained granulitic type with no trace of schistosity. The granoblastic base (80% of the section) consists of equal amounts of quartz and untwinned basic plagioclase. Pale-green pyroxene constitutes fully 15% of the rock as small, isolated grains, and as deeply indented sieve-like porphyroblasts reaching 0.2 mm. in diameter. Unoriented small prisms of green hornblende and small grains of iron-ore are minor constituents (1% each).

(iv) Acid Granulites. The acid granulites of the Foreshore (60, * 62a, 63) are slightly gneissic, fine-grained quartzo-feldspathic rocks consisting essentially of a crystalloblastic aggregate of elongated grains of quartz (40%) and alkali-feldspar. The feldspar includes orthoclase and oligoclase-andesine, the latter in clear subidioblastic grains surrounded by xenoblastic, sometimes slightly sericitized grains of orthoclase. A few interstitial grains of microcline are present, and myrmekitic intergrowths are frequently developed at the orthoclase boundaries. Though the fine grain-size of the rock makes micrometric

[Footnote] * For analysis see table at end of paper.

[Footnote] * For analysis see table at end of paper.

– 204 –

measurements difficult, the preponderance of the potash over soda shown by the analysis proves that plagioclase is subordinate to orthoclase and microcline. The dark-coloured constitutents (about 2% or 3%) include numerous small wedge-shaped grains of clear sphene associated with small grains of iron-ore, a few ragged crystals of reddish-brown biotite, and a little bluish-green hornblende. The analysis agrees well with Daly's “average granite.”

(v) Amphibolite. Rocks classed as amphibolites, apparently derived from porphyritic dolerites, include Nos. 43, 46, 48, 50, 64 from the Foreshore, and Nos. 132, 159, 164 from between the wharves and Ocean Beach railway-station. A typical specimen (50) from a dyke 350 yards south-east of Henderson Street is a dark-green rock, seen microscopically to contain blastophenocrysts of hornblende averaging 2.5 mm. (20%) and porphyroblasts of andesine (5%) set in a schistose fine-grained base of hornblende, biotite and plagioclase. The large crystals of grass-green hornblende are of a compact, non-fibrous variety, and frequently have small flakes of deep-brown biotite developing along their cleavages. Many of the larger crystals still retain cores of colourless aluminous pyroxene with large optic-axial angle and positive sign. Where small isolated patches of hornblende occur within the augite, they, too, are in optical continuity with the main rim of hornblende. The porphyroblasts of plagioclase reach 1.5 mm. in length, and are crowded with small prisms of hornblende, flakes of biotite, and granules of iron-poor epidote and iron-ore. Their composition is not certain, but does not greatly differ from the basic andesine of the groundmass. The schistose base is composed of subparallel prisms of hornblende, flakes of greenish-brown biotite, and grains of water-clear, twinned plagioclase with the composition of basic andesine (10%). Small grains of sphene are especially abundant within the hornblende surrounding the crystals of original augite. Iron-ore and apatite are minor constituents.

Picture icon

Text-Fig. 2.—Field sketch showing the intrusion of a dyke of amphibolite (originally a porphyritic dolerite and still showing cores of augite within the hornblendes) into bedded hornblende schists. Black areas are veins of epidote and clear spaces quartz (Q). Locality: 375 yards south-east from foot of Henderson Street, Bluff Foreshore.

– 205 –

A field-sketch of this dyke is given in Text-Fig. 2. The main point of interest is the abundance of pale-green epidote lining its walls and veining the invaded hornblende-schist. As concentrations of epidotic veins such as this are rare on other parts of the Foreshore, it appears that the epidote has migrated from the amphibolite during metamorphism.

3. Mutual Relationships.

The Tewaewae Point rocks include two main types, both apparently intrusive: epidiorites of more or less normal gabbroidal or basaltic composition, and albitic-metadolerites more closely comparable with rocks of the spilitic suite. Both these types texturally are closely similar, and neither has suffered metamorphism of a sufficiently high grade to cause obliteration of original texture.

On the Foreshore also the green schists can be classed in two main groups, viz., hornblende-andesine-schists and hornblende-oligoclase-schists. Petrographic and chemical evidence indicates that these andesine-and oligoclase-bearing schists are comparable with the Tewaewae Point epidiorites and albitic metadolerites respectively. Owing to the reconstituted state of the schists of the Foreshore, it has not been found possible to determine whether they were originally intrusive rocks, lavas or derived sediments. The writer favours a purely igneous origin and ascribes their apparent “bedding” to migration of various bases during metamorphism.

The slightly altered rocks from the north-east side of the Greenhills range are definitely tuffaceous and markedly spilitic. High-grade albite-bearing pyroxene-granulites exposed on the coast on the south-west side of the range are likewise believed to be metamorphosed spilitic tuffs.

It is therefore clear that in the three localities the bulk of the metamorphic rocks were originally intrusive rocks, lavas, tuffs or derived sediments of closely allied composition. Rocks with gabbroidal or basaltic composition are abundant in the first two localities, while spilitic types are common in all three. Furthermore, there is evidence on Tewaewae Point that, before the metamorphism was effected, the main mass of epidiorites and albitic metadolerites was invaded by keratophyres; prior to metamorphism green schists of the Foreshore were also cut by acid igneous rocks now represented by leucocratic granulites. All these rocks have been affected to greater or less extent by the intrusive Bluff norite, and all therefore probably belong to a single pre-norite cycle of igneous activity.

4. Nature of Metamorphism.

It is now well established that the two most characteristic equilibrium assemblages resulting from low-grade regional metamorphism upon rocks of basic igneous composition are chlorite-albite-epidote or actinolite-epidote-albite-chlorite (e.g., Tilley, 1923; Vogt, 1927; Sugi, 1931; Phillips, 1930; Turner, 1933a, 1935, 1935b; Wiseman, 1934). It is also recognised that the assemblage green hornblende-medium plagioclase is attained in such rocks only at high grades of regional metamorphism corresponding to the zones of almandine, kyanite and sillimanite as defined for associated pelitic schists

– 206 –

(Phillips, 1930). Furthermore, crystalloblastic pyroxene is rarely formed even at the highest grades in such rocks (Harker, 1932, p. 284).

On the other hand, in pure contact metamorphism one of the first changes at relatively low grades is the conversion of pyroxene to green hornblende, followed by recrystallization of the basic plagioclase to andesine or andesine-labradorite (Harker, 1932, pp. 108–110). At higher grades, the hornblende in turn is partially or completely replaced by pyroxenes, and the composition of the plagioclase becomes more calcic, while epidote and zoisite disappear (e.g., Tilley, 1921, 1924; Stillwell, 1923; Browne, 1927; Joplin, 1933; Turner, 1933b). Thus the assemblage green hornblende-medium plagioclase, which is stable under conditions of high-grade dynamothermal metamorphism, is also equally characteristic to low to moderate grades of contact metamorphism.

To return to the case under consideration, it is obvious from the incompletely reconstituted state of many of the rocks of Tewaewae Point and Greenhills that they have suffered metamorphism of only low grade. Yet the most characteristic crystalloblastic mineral assemblages are (a) green hornblende+labradorite in the epidiorites, and (b) green hornblende+albite (or albite-oligoclase) in the spilitic rocks. In the more completely recrystallized schists of the Foreshore where the grade of metamorphism is much higher, the characteristic corresponding assemblages are (a) hornblende+andesine and (b) hornblende+acid oligoclase. Again among the high-grade rocks of both the Foreshore and the south-west of the Greenhills range minor bands are found containing pale-green pyroxene; on the Foreshore where the rocks correspond to epidiorite, the plagioclase is labradorite, but in the Greenhills area, where the low-grade equivalents are spilitic tuffs, the plagioclase is albite-oligoclase. It is to be noted too that the lowest grade of metamorphism was reached around the Greenhills railway station and on Tewaewae Point (localities at least half a mile distant from outcrops of norite), while the highest grades were attained relatively close to the norite. It is thus clear that metamorphism has been almost purely thermal, and that it accompanied and was caused by intrusion of the Bluff norite. The development of clouding of feldspars and the general absence of strain-structures in rocks of low grade in this district accord with this conclusion.

5. Age.

Though no precise conclusion can be reached as to the age of the rocks of the Metamorphic Series, there are strong reasons for correlating them with the Te Anau Series (probably late Palaeozoic) of Western Otago. The main evidence may be summarised as follows: (a) The Te Anau Series which is composed mainly of basic igneous material—greywackes, breccias, tuffs and some true lavas—is the only known older formation of this type in Western Otago. The Te Anau rocks continue southwards in an almost unbroken line from their type locality near Lake Te Anau to the Riverton-Orepuki District, some 20 miles north-west of Bluff, where they have been invaded by a mass of norite comparable with the Bluff intrusion. (b) The rocks of the Te Anau Series have recently been shown to

– 207 –

have been affected by the widespread regional metamorphism with which the principal phase of plutonic intrusion in the Fiordland region is believed to be coeval (Turner, 1935a). The Bluff norite and the varied plutonic rocks of Orepuki, Riverton, etc., probably are to be correlated with this Fiordland intrusive series. * (c) In a quarry at Greenhills a coral was found which has been identified by Dr. R. S. Allen, of Canterbury College, as a “Zaphrentid.” Specimens of Zaphrentis have also been collected in Upper Palaeozoic (Permian ?) greywackes at Clinton, Eastern Otago (Marwick, 1925, pp. 362–363), and in Permian or Permo-Carboniferous grey-wackes in the Nelson District (Benson, 1921, p. 19). These rocks tentatively are considered to belong to the same major system as the Te Anau Series. (d) During the last two years albitic rocks with spilitic affinities have been recorded from Carnic and Jurassic conglomerates at Nugget Point (Mackie, 1935, p. 289) and Kawhia (Bartrum, 1935, pp. 98–100) respectively, showing that in mid-Mesozoic times such rocks were exposed to erosion in both the North and South Islands of New Zealand. Keratophyres have also been recorded from the breccias of the Te Anau Series in the district between Lake Wakatipu and Lake Te Anau (Turner, 1935, p. 333).

Intrusive Series.

I. Distribution.

Rocks of the intrusive series include the main Bluff norite, and all intrusive rocks which are known to be of younger age than the norite, viz., peridotite, pyroxenite, quartz-diorite, and granite.

The norite, which is the only member to attain wide areal extent, stretches in an almost unbroken line for nine miles from Bluff Hill to the mouth of New River. The whole of Bluff Hill, except for the narrow strip of metamorphic rocks outcropping along the harbour-front, consists of this rock, and it is probable that it also occupies the greater part of the Greenhills range. The only locality where the norite was actually seen in contact with the metamorphic rocks is at a point 1500 yards west of the wharves, where hornblende-schists are cut by irregular veins of norite. The frequent presence of dykes of gneissic gabbro intersecting the main mass indicates that intrusion was not confined to one place.

Small outcrops of partially serpentinized peridotite were observed in two localities: one, on the roadside one and a half miles west of Greenhills railway-station, and the other in a reef exposed at low tide on the sea-coast three-quarters of a mile to the south. Since both outcrops are isolated, the field-relations of the rocks are uncertain. Two hundred yards east of the first-mentioned outcrop of peridotite are exposures of a coarse-grained pyroxenite which extends eastward for about another two hundred yards. The same rock is again exposed on the coast to the south-east of the second exposure of peridotite. The presence of wide sand-dunes between the road and the coast renders it difficult to determine the relations of the peridotites and pyroxenites, but probably both are younger than the norite.

[Footnote] * From a study of the basic igneous intrusions in the Manapouri district Dr. Turner has recently come to a similar conclusion.

– 208 –

Two occurrences of quartz-diorite have been found, and in both instances the rock is intrusive into norite. The first is to be seen one mile south-west of Starling Point, where numerous thin veins of quartz-diorites intersect dykes of gneissic gabbro. Again, on the harbour-front half-way between the wharves and Ocean Beach railway-station, many circular patches of quartz-diorites invade massive norite.

Possibly the youngest igneous rock in the district is a biotite-muscovite-granite, which also was noted in situ in only two places: a small patch 9 inches in diameter cutting epidiorite some 1000 yards east-south-east of Tewaewae Point, and a dyke intersecting norite on the coast-line two miles north-west from Ocean Beach.

II. Petrology.

(i) Norites. The following description is a general summary of the main characters of the norite as determined from examination of some fifty slices of rocks collected on Bluff Hill and the Greenhills range. In hand-specimen they are typically somewhat gneissic, medium-grained, greyish-green rocks with recognisable plagioclase, pyroxene, hornblende, and in some cases conspicuous flakes of black biotite. In most specimens, the grains are about 2.0 mm. in diameter, but a richly-hornblendic coarse pegmatitic rock often occurs as small nodules and veins in the main rock, while other types are strongly-gneissic, fine-grained and dense black in colour.

The texture varies from allotriomorphic granular to subidiomorphic granular, while in a few cases ophitic texture is seen. Slight parallelism of the prisms of plagioclase is apparent in most rocks, but is developed especially in dykes of younger norite which cut through the main mass. The chief constituent minerals are basic plagioclase 60–80% (average of 50 specimens, 65%), pyroxene 0–35% (20%), hornblende 0–35% (12%) and iron-ore 1–5% (3%), while apatite is a constant accessory. Less common additional minerals are olivine (up to 15%), biotite (up to 5%), quartz (up to 5%), epidote, prehnite, green spinel, a zeolite, tremolite and chlorite.

The plagioclase varies between acid labradorite and labradorite-bytownite, but in the great majority of the specimens examined it is basic labradorite (about Ab36 An64). Albite twinning is always developed, and is sometimes accompanied by twinning after the pericline and Carlsbad laws. The crystals often exhibit zoning, with inner zones as basic as labradorite-bytownite and outer zones as sodic as acid labradorite. While the outer zones of these crystals are clear and uncracked, the basic cores usually show irregular cracks which are filled with an indeterminate dark “stain.” It is thus suggested that there was movement in the magma after the crystallization of the more basic plagioclase, and that this was followed by crystallization of the outer zones while the magma was at rest. Alteration to pale-yellow iron-poor epidote is sometimes found, especially in the cores of the zones crystals.

Pyroxene is the predominant femic mineral, and occurs as idiomorphic to subidiomorphic crystals with broad prismatic habit, or less commonly as reaction rims about grains of olivine. It is pink

– 209 –

in colour and faintly pleochroic. On account of the frequency of sections showing inclined extinction, the mineral was at first thought to be clinohypersthene and enstatite-augite (Service, 1934). Mr. C. O. Hutton, * however, has examined sections with a universal stage and concludes that the mineral is normal hypersthene with some clino-pyroxene intergrown in lamellar fashion so as to simulate twinning. The axial angle of 55° and negative sign as determined by Mr. Hutton agree with hypersthene containing equal amounts of FeSiO3 and MgSiO3. In none of the sections examined by the writer has been detected common colourless augite, with positive optical sign, large optic axial angle, and large extinction angle, though Wild (1911, p. 321) records its presence. The pyroxene is frequently clear and free from inclusions (except where alteration to hornblende is in progress), but it is also often packed with schiller inclusions, and may show a patchy green clouding due to some alteration-product of indeterminate composition.

Slightly less abundant than the pyroxene is compact green hornblende, which typically shows the following pleochroism: X = greenish-yellow, Y = brownish-green, Z = olive-green, X < Y < Z. Types with grass-green colour for the Y and Z directions are not uncommon; but none has the deep bluish-green tint found in the hornblende of the metamorphic rocks. In the majority of the rocks the hornblende can be seen both as reaction-rims about the pyroxene (or olivine) and as isolated homogeneous subidiomorphic prisms. In the former cases the junction between pyroxene and green hornblende is usually sharp, but in some slides the minerals are separated by a narrow zone of faintly coloured hornblende, a feature noted by Wild (1911, p. 321). The faintly-coloured intermediate zone is sometimes to be seen between hornblende and olivine when reaction-rims of this type also are found. In all sections where hornblende occurs, both as reaction-rims about pyroxene or olivine and as isolated prisms, its optical properties are the same. In many slides compact hornblende may also be seen to have developed as a reaction-product between iron-ore and plagioclase. On the above grounds it is concluded that all the hornblende in the norite is of primary magnetic origin, and is derived by reaction between the magma and suspended crystals of olivine, pyroxene or iron-ore. This question will be discussed more fully in a later paragraph. In one specimen (No. 2) from a boulder half a mile north of Starling Point the norite contains, besides the normal compact hornblende, a felted patch of finely-fibrous lighter-coloured amphibole (uralite) surrounding a spongy mass of iron-ore and mingled with pale-green chlorite. This rock is close to the hornblende-schist, and may possibly be a hybrid. Also in No. 169a, where the norite is threaded with siliceous prehnite-bearing veins, fibrous uralite is developed along with the normal compact hornblende.

Iron-ore, which appears to have separated in two generations, is present both as small grains enclosed within the feldspars and mafic minerals and as large irregular masses which are moulded around these. It also occurs as a fine dust along cracks in the rare crystals of olivine, and in symplectitic intergrowth with pyroxene in

[Footnote] * Personal communication.

– 210 –

reaction-rims to olivine. Thomson (1916, p. 37) considered the iron-ore to be probably ilmenite, but Hamilton (1887, p. 453) noted the influence of the rocks upon the magnetic needle. The writer found, by testing with a weak magnet after crushing and panning, that about 95% of the concentrate was magnetite.

Olivine was noted in only two rocks (Nos. 167 and 169c), both from a cliff-face 1 mile south-west of Starling Point. In No. 167 the cracks in the olivine are lined only with iron-ore, and the grains are surrounded by reaction-rims of pyroxene and hornblende. In the latter specimen, however, where the rock is in contact with prehnite-bearing veins, the olivine, besides showing the above two features, is also altered in places to a mass of fibrous tremolite and grains of iron-ore.

Biotite, if present, is invariably developed in the neighbourhood of iron-ore, e.g., in No. 108 (i), either where the latter adjoins pyroxene or hornblende, or often where it is completely isolated from them. The flakes are usually small, but in No. 108 (i) they reach 1.5 mm. in diameter. The pleochroism is X = pale greenish-yellow, Y = Z = dark reddish-brown. Quartz, though usually absent, may be present as small interstitial grains. Apatite, which is enclosed by feldspars, pyroxenes and hornblende, may occur as rounded grains, as broad short prisms, or rarely as long fine needles. Secondary iron-poor epidote is present in small amount in the basic cores of some of the zoned crystals of plagioclase.

Prehnite was noted in one specimen of norite (No. 95), but was also recognised in sodic veins (e.g., Nos. 170a–e) traversing an olivine-bearing norite. In No. 95 it occurs in narrow, parallel veins as small ragged prisms associated with stilbite. No. 170 is from a vein consisting mainly of acid oligoclase, quartz and prehnite, with minor amounts of calcite, chlorite, biotite, sphene (the only occurrence of sphene noted in the gabbros and associated veins), apatite and iron-ore. The prehnite is seen threading through the grains of quartz and penetrating into and fraying out from the ends of the crystals of feldspar. In one slide (No. 170d) the prisms of prehnite are embedded in a wedge-shaped mass of calcite.

Green spinel is present in a number of slides (e.g., 110, 169c) in vermicular intergrowth with green hornblende, where the latter forms reaction-rims to olivine or pyroxene. A few small, round grains of the same mineral are also found either enclosed within or abutting against large grains of iron-ore. Examples showing a similar close association of green spinel and iron ore are given by Holland (1900, p. 168) and J. H. L. Vogt (1909, pp. 91–2).

In two specimens (Nos. 95, 173) thin veins traversing the norite contain a fibrous zeolite whose properties agree most closely with those of stilbite, and which in No. 90 is associated with prehnite. Tremolite occurs only in one specimen (No. 169c), where it appears to be developed at the expense of olivine as fine radiating needles intermingled with dusty grains of iron-ore. The only other mineral recognised in the norite is chlorite, which has been found in one rock (No. 167). Here an aggregate of finely flaky pale-green chlorite and grains of iron-ore is completely surrounded by grains of pyroxene.

– 211 –

In considering the causes of crystallization of hornblende two possibilities are to be noted: firstly, that the hornblende has been formed from the enclosed grains of pyroxene, etc., after consolidation of the magma by subsequent metamorphic action, or, secondly, that it is a product of normal magmatic reaction between early-formed crystals and magma. The following points are opposed to the first alternative:—


The rocks do not show the effects of stress after consolidation and therefore cannot have undergone any form of dynamic metamorphism. The typical plutonic texture and the high-temperature mineral assemblages also preclude this possibility.


If it be assumed that hydrothermal metamorphism caused the change, there is a marked absence of metamorphic change in the other minerals in the rocks (except in a very few specimens, e.g., where prehnite-and zeolite-bearing veins traverse the gabbro).


The hornblende has not the fibrous or acicular habit typical of uralite, nor is there any saussuritization of plagioclase such as typically accompanies development of uralite.

The alternative hypothesis of magmatic reaction meets with none of these difficulties, and there is additional mineralogical evidence in support of it. Points in favour of origin by magmatic reaction are:—


The typical unaltered plutonic texture of the rocks.


The reaction-phenomena shown by minerals other than hornblende (summarised in the following section).


The compact, greenish-brown character of the hornblende.

The available evidence therefore favours derivation of the hornblende from pyroxene or olivine by the normal processes of magmatic reaction as outlined by Bowen (1928, pp. 54–91), and there seems no necessity to postulate subsequent metamorphism in order to explain all the phenomena observed.

The various types of reaction-rims and coronas shown by the minerals in the norites may be summarised as follows:—

  • Olivine → pyroxene.

  • Olivine → pyroxene → hornblende.

  • Olivine → hornblende.

  • Pyroxene → hornblende.

  • Intergrowth of hornblende and green spinel.

  • Intergrowth of pyroxene and iron-ore.

  • Development of hornblende adjacent to iron-ore.

  • Development of biotite adjacent to iron-ore.

The boundary between olivine and its reaction-rims of pyroxene is always sharp, but where green hornblende is seen replacing the olivine, the former occurs as an irregular fringe which is commonly separated from the olivine by a narrow zone of pale amphibole (e.g., No. 167). This is not a universal feature, for in a single slide the colourless zone is recognisable in some reaction-rims but not in others. The colourless hornblende is also often developed between grains of

– 212 –

pyroxene and rims of green hornblende. Reaction-rims of pyroxene around olivine, and of hornblende around olivine or pyroxene, occur only along boundaries where the original mineral is in contact with plagioclase, or has been so at an earlier stage. The phenomena of reaction-rims have been discussed fully by Sederholm (1916), who cites many examples in which green hornblende forms coronas about olivine or hypesthene, from which it is separated by a zone of colourless amphibole.

In a few rocks (e.g. No. 169c) where pyroxene adjoins plagioclase and has developed a rim of green hornblende, the latter is packed with small grains of green spinel arranged roughly perpendicularly to the line of contact of hornblende and plagioclase. Examples of vermicular intergrowths containing spinel, other than those given by Sederholm (1916, pp. 25–6), are recorded by W. N. Benson (1910, pp. 518–529) from volcanic rocks of Dundas near Sydney (in this case the intergrowth is with pyroxene), and T. Vogt (1927, pp. 519–520) in the Sulitelma district, Norway.

In the olivine-bearing rocks (Nos. 167 and 169c), iron-ore is intergrown with pyroxene where the latter borders plagioclase-Intergrowths of this type have been described by Duparc and Pearce (1905, p. 456), who state that it is a result of primary magmatic reaction.

(ii) Peridotite. In hand-specimen, the peridotite (No. 58) from the motor-road one and a-half miles west of Greenhills railway-station is a medium-grained, light-green rock with uneven fracture. Under the microscope it is seen to consist of olivine (30%) which has been largely replaced by chrysotile serpentine. The olivine has the large optic axial angle and positive sign of the magnesian variety. Along the junction between the original grains and in the cracks of the olivine, the serpentine consists of colourless, parallel fibres associated with which are long streaks of spongy iron-ore. Pseudomorphous after the crystals of olivine and enclosed within the earlier mesh-serpentine are felted masses of a pale-green finely-fibrous variety which is almost isotropic, and which is nearly free from iron-ore (cf. Harker, 1923, p. 76). The olivine shows a peculiar “schiller” effect due to the separation along its cleavage planes of minute deep-brown prismatic inclusions which always lie in well-defined lines parallel to the vertical axis.

A similar partially serpentinized peridotite (208) from the coastline three-quarters of a mile to the south contains a few large grains of pale-green augite, the cleavages of which are penetrated by strings of colourless serpentine derived from the olivine.

These occurrences of chrysotile-serpentine are unusual for southern New Zealand, where antigoritic types usually prevail (e.g. Turner, 1930, p. 197, and 1933, p. 260), but a section of pyroxene-bearing peridotite in the collections of the Department of Geology, Otago University, labelled “Southland,” shows the same type of alteration.

(iii) Pyroxenite. The pyroxenite outcropping on the road running west from Greenhills railway-station is a coarse-grained, massive, dark-green rock consisting mainly of colourless augite (95%) and

– 213 –

olivine (5%). The augite is in strongly-schillerised crystals reaching 8.0 mm. in length and altering along the cleavages to common green hornblende. Within the augite and at the junction of different crystals are small nests of coarsely-fibrous pale-green chrysotile. Olivine, as rounded grains averaging 1.0 mm. in diameter often completely enclosed by augite, shows two types of alteration, viz., to deep brown pseudomorphous iddingsite (with separation of iron oxide along the cracks) and, to long often radial needles of tremolite mixed with grains of iron-oxide and a little fibrous tale. The other specimen of pyroxenite (215) is essentially the same, but is cut by a narrow vein of fresh brown hornblende (Z to c = 19°).

(iv) Quartz-diorite. Two closely similar specimens (Nos. 26 and 161 * of quartz-diorite have been collected. No. 26 (1 mile south-west of Starling Point) is a medium-grained slightly gneissic rock, with slender prisms of hornblende averaging 5.0 mm. in length set in a clear base of quartz and feldspar. It is granitoid in texture and consists essentially of plagioclase (75%), hornblende (15%), and quartz (10%). The plagioclase, medium-andesine (Ab56 An44), occurs in subidiomorphic stout prisms showing albite, Carlsbad and pericline twinning. When zoning is evident the cores of acid labradorite are commonly obscured by a dirty greenish indeterminate decomposition product. Hornblende forms subidiomorphic slender prisms which have the following pleochroism: X = greenish-yellow, Y = brownish-green, Z = grass-green, X < Y < Z. Quartz occurs interstitially in clear grains up to 1.0 mm. in diameter. Accessory constituents are small crystals of apatite, grains of clouded sphene and rarely of iron-ore. No. 161 is a rather more gneissic rock containing fully 20% of quartz and less than 5% of hornblende, but otherwise resembles the rock just described.

(v) Biotite-muscovite-granite. Specimen No. 181 * from the coast 1000 yards east of Tewaewae Point is a coarse-grained rock, containing about 50% of quartz, 25% of orthoclase, 15% acid plagioclase, 10% of biotite, a little microcline, and accessory muscovite, magnetite and sphene. Orthoclase, which rarely shows Carlsbad twinning, occurs in subidiomorphic crystals of about 1.0 mm. diameter, usually slightly clouded with kaolin. The plagioclase is acid oligoclase (Ab87 An13); usually a much sericitised inner portion is surrounded by a clear rim, but there is no indication of magmatic zoning. Slightly chloritized biotite is present in rounded plates, averaging 1.2 mm. in diameter, and strongly pleochroic (deep reddish-brown to pale honey-yellow). Quartz is present interstitially in clear grains reaching 3.0 mm. in diameter. There are one or two small plates of primary muscovite, and grains of microcline which appear to have crystallized late along with the quartz. Iron-ore occurs only in the chlorite, and apatite is a minor accessory.

The other specimen, No. 205, collected from a dyke on the sea-coast two miles north-west of Ocean Beach, is closely similar.

[Footnote] * For analysis see table at end of paper.

[Footnote] * For analysis see table at end of paper.

– 214 –

Recently Dr. F. J. Turner (1930, 1933a, 1935a) has brought forward evidence indicating that there were two periods of major intrusion in South Westland and Otago. The first of these (possibly late Palaeozoic), was accompanied by the uprise of extensive intrusions of granite, diorite and gabbro and is believed to belong to the same orogeny as was responsible for progressive metamorphism of the schists of Central Otago and Westland. The second period, which appears to have been coeval with the early Cretaceous or late Jurassic Hokonui Orogeny, was characterized by the intrusion of the great peridotite masses of Nelson, Westland and Western Otago. Although at Bluff the main intrusive rock is norite, accompanied by dioritic and granitic rocks in only minor amount, granite, according to the writer's observations, is more in evidence on Ruapuke Island, while on Stewart Island it is most extensively developed (Williams, 1934). Gabbros and norites are found at intervals (e.g., at Orepuki) between Bluff and Fiordland where the granite-diorite-gabbro series reaches its maximum (Benson, 1933).

Thus, it is concluded that the Bluff norite was intruded during the same orogenic period as the plutonic rocks of Fiordland and Stewart Island. As stated above, this was probably in the late Palaeozoic.


The metamorphic rocks of Bluff afford an example of contact metamorphism of basic, semi-basic and semi-acid intrusive rocks, lavas, tuffs and derived sediments. Metamorphism has resulted from the intrusion of a mass of norite which forms the backbone of the Bluff Peninsula. As the intrusion is approached, stages can be traced from slightly altered intrusive and tuffaceous rocks through more completely recrystallized hornblende-schists into high-grade pyroxene-granulite-hornfelses. Two main chemical divisions in the metamorphosed rocks are recognisable—firstly, rocks of more or less normal doleritic or basaltic composition, and, secondly, rocks showing marked affinities with the spilitic suite. Of most interest perhaps is the presence of acid plagioclase instead of the usual basic variety in high-grade pyroxene-granulites from Greenhills (derivatives of spilitic tuffs).

The norite typically contains hypersthene and abundant green hornblende which is believed to be a product of primary magmatic reaction rather than of post-magmatic metamorphism. Cutting the norite are a few small intrusions of peridotite, pyroxenite, quartz-diorite and granite.

There is evidence for correlating the metamorphic rocks with the (?) late Palaeozoic Te Anau Series of Western Otago, and for considering that the norite is of the same age as extensive series of plutonic rocks developed throughout Fiordland; the date of intrusion is therefore possibly late Palaeozoic.

– 215 –

[The section below cannot be correctly rendered as it contains complex formatting. See the image of the page for a more accurate rendering.]

Table of Rock-Analyses.
No. 51 52b 209 184 48a 65 163 190 179
SiO2 59.13 53.56. 53.31 50.46 51.50 45.25 54.73 50.93 49.20
TiO2 0.74 0.46 0.98 1.36 0.33 1.61 0.80 0.36 1.21
Al2O3 16.43 18.66 16.43 17.67 19.50 18.68 15.24 18.37 15.42
Fe2O3 1.72 3.21 2.12 4.65 2.30 3.48 3.71 4.70 5.27
FeO 5.92 5.97 6.29 7.73 5.19 9.00 5.31 6.24 7.44
MnO 0.17 0.14 0.18 0.21 0.09 0.28 0.19 0.14 0.23
MgO 4.01 3.54 4.26 4.05 5.80 6.00 5.21 4.24 4.43
CaO 5.52 6.56 9.16 7.89 10.76 11.65 12.15 9.74 14.68
Na2O 5.28 6.57 5.19 4.25 3.56 2.02 1.56 3.57 0.60
K2O 0.23 0.76 0.88 0.35 0.31 0.24 0.33 0.49 0.20
H2O+ 0.64 0.65 1.57 1.29 0.93 1.37 0.60 0.70 1.17
H2O− 0.06 0.07 0.16 0.11 0.16 0.13 0.10 0.25 0.21
P2O3 0.31 0.17 0.14 0.11 0.15 0.20 0.13 0.70 0.20
——— ——— ——— ——— ——— ——— ——— ——— ———
100.16 100.32 100.67 100.13 100.58 99.91 100.06 100.43 100.26

[The section below cannot be correctly rendered as it contains complex formatting. See the image of the page for a more accurate rendering.]

No. 189 60 53 108 103 70 161 181
SiO2 69.82 70.50 55.72 51.46 47.25 44.43 74.02 72.55
TiO2 0.50 0.38 1.11 1.51 1.36 0.40 0.30 0.28
Al2O3 13.78 14.71 16.27 15.43 17.49 24.16 12.39 13.88
Fe2O3 3.11 1.78 1.46 3.92 4.73 1.19 0.87 0.23
FeO 1.44 0.50 8.39 8.16 6.88 4.91 1.02 1.57
MnO 0.09 0.03 0.18 0.24 0.27 0.10 0.04 0.05
MgO 0.91 0.36 4.85 4.53 5.64 5.67 0.65 0.43
CaO 2.02 2.02 6.20 8.98 12.20 16.36 5.13 1.35
Na2O 5.74 3.66 3.37 3.41 2.30 0.91 3.23 3.46
K2O 1.30 5.40 1.41 0.93 0.07 0.07 0.38 4.99
H2O+ 1.00 0.28 0.76 0.67 0.44 0.67 0.51 0.69
H2O− 0.10 0.16 0.19 0.07 0.09 0.06 nil 0.13
P2O5 0.04 0.13 0.35 1.22 1.42 1.60 1.28 0.67
——— ——— ——— ——— ——— ——— ——— ———
99.85 99.91 100.26 100.53 100.14 100.53 99.82 100.28
  • 51: Hornblende-oligoclase-schist, Foreshore. Anal. H. Service.

  • 52b: Spilitic tuff, Greenhills quarry. Anal. H. Service.

  • 209: Pyroxene-hornblende-biotite-albite-granulite, Greenh'lls. Anal. A.W. Groves.

  • 184: Albite-metadolerite, Tewaewae Point. Anal. H. Service.

  • 48a: Hornblende-andesine-schist, Foreshore. Anal. H. Service.

  • 65: Hornblende-andesine-schist, Foreshore. Anal. A.W. Groves.

  • 163: Hornblende-pyroxene-labradorite-schist, Foreshore. Anal. A.W. Groves.

  • 190: Epidiorite, Tewaewae Point. Anal. H. Service.

  • 179: Epidiorite rich in epidote, Tewaewae Point. Anal. A.W. Groves.

  • 189: Quartz-keratophyre, Tewaewae Point. Anal. A.W. Groves.

  • 60: Acid granulite, Foreshore. Anal. A.W. Groves.

  • 53: Andesine-actinolite-hornfels, Greenhills. Anal. A.W. Groves.

  • 108: Typical norite, Bluff Hill. Anal. A.W. Groves.

  • 103: Hornblendic norite, Bluff Hill. Anal. A.W. Groves.

  • 70: Olivine-bearing norite, Bluff Hill. Anal. A.W. Groves.

  • 161: Quartz-diorite, Bluff Hill. Anal. A.W. Groves.

  • 181: Biotite-muscovite-granite, Tewaewae Point. Anal. A.W. Groves.

– 216 –
Picture icon

Text-Fig. 3.—Diagram to illustrate the spilitic nature of rocks 189, 52b, 209, and 51, and the more normal doleritic composition of rocks 190, 48a, 163. (The analysis of an albite metadolerite, No. 184, does not fall as near the spilitic field as was anticipated. This may be accounted for by reason of the fact that only a few grammes of rock, chipped from a small handspecimen, were available for analysis.)

Literature Cited.

Bartrum, J. A., 1935. Metamorphic Rocks and Albite-rich Igneous Rocks from Jurassic Conglomerates at Kawhia, Trans. Roy. Soc. N.Z., vol. 65, pt. 2, pp. 95–107.

Benson, W. N., 1910. The Volcanic Rocks of Hornsby and Dundas near Sydney, Proc. Roy. Soc. N.S.W., vol. 44, pp. 495–555.

— 1921. Recent Advances in New Zealand Geology, Rep. Austr. Ass. Adv. Sci., Section C, Presidential Address.

— 1933 The Geology of the Region about Preservation and Chalky Inlets, South-west Fiordland, New Zealand, Part I, Trans. N.Z. Inst., vol. 63, pp. 393–432.

Bowen, N. L., 1928. The Evolution of the Igneous Rocks, Princeton University Press.

Browne, W. R., 1927. On some Metamorphosed Dolerites from Broken Hill. Proc. Roy. Soc. N.S.W., vol. 61, pp. 383–400.

Daly, R. A., 1933. Igneous Rocks and the Depths of the Earth, McGraw-Hill.

Duparc, L., and Pearce, F., 1905. Recherches geologiques et petrographiques sur l'Cural du Nord, II, Mem. Soc. Phys. et Hist. Nat. du Geneve, vol. 34, pp. 383–602.

Eskola, P., 1932. On the Principles of Metamorphic Differentiation, Compt. Rend. Geol. Soc. Fin., no. 5, pp. 68–77.

Forbes, C., 1855. On the Geology of New Zealand, Q.J.G.S., vol. ii, pp. 521–530.

Hamilton, W. S., 1887. Notes on the Geology of the Bluff District, Trans. N.Z. Inst., vol. 19, 1886, pp. 452–455.

Harker, A., 1923. Petrology for Students, Cambridge University Press.

— 1932. Metamorphism, Methuen, London.

Hector, J., 1863. A Geological Expedition to the West Coast Sounds, Otago Prov. Govt. Gazette, 5th November, pp. 435–468.

Holland, T. H., 1900. The Charnockite Series, Mem. Geol. Surv. India, vol. 28, pt. 2, pp. 119–249.

Hutton, F. W., 1872. On the Geology of the District of Southland, in the Province of Otago, Rep. Geol. Explor., 1871–72, pp. 96–112.

— 1875. The Geology of Otago, Mills, Dick & Co., Dunedin.

— 1889. The Eruptive Rocks of New Zealand, Proc. Roy. Soc. N.S.W., vol. 23, pp. 102–156.

— 1891. Note on the Eruptive Rocks of the Bluff Peninsula, Southland, Trans. N.Z. Inst., vol. 23, 1890, pp. 353–355.

Joplin, G. A., 1933. Petrology of the Hartley District, II, Proc. Linn. Soc. N.S.W., vol. 58, pts. 3–4, pp. 125–158.

Macgregor, A. G., 1931A. Clouded Feldspars and Thermal Metamorphism, Min. Mag., vol. 22, no. 133, pp. 524–438.

— 1931b. Scottish Pyroxene Granulite Hornfelses and Odenwald Beerbachites, Geol. Mag., vol. 68, no. 11, pp. 506–521.

Mackie, J. B., 1935. The Geology of the Glenomaru Survey District, New Zealand, Trans. Roy. Soc. N.Z., vol. 64, pp. 275–302.

– 217 –

Marwick, J., 1925. Upper Palaeozoic (Permian) Fossils at Clinton, N.Z. Jour. Sci. and Tech., vol. 7, no. 6, pp. 362–364.

Park, James, 1888. On the Geology of Bluff Peninsula, Rep. Goel. Explor., 1887–88, pp. 72–74.

Phillips, F. C., 1930. Some mineralogical and chemical changes induced by progressive metamorphism in the Green Bed group of the Scottish Dalradian, Min. Mag., vol. 22, no. 129, pp. 239–256.

Read, H. H., 1931. The Geology of Central Sutherland, Mem. Geol. Surv. Scotland.

Sederholm, J. J., 1916. On Synantectic Minerals, Bulll. Comm. Geol. Finlande, no. 48.

Service, H., 1934. Note on the Occurrence of Clinohypersthene and Enstatite-Augite in the “Norite” from Bluff, New Zealand, Trans. Roy. Soc. N.Z., vol. 64, pt. 2, pp. 147–150.

Stillwell, F. L., 1918. The Metamorphic Rocks of Adelie Land, Rept. Austr. Antarctic Exped., Ser. A, pt. 1.

— 1923. Amphibolites and Related Rocks from the Moraines, Cape Denison, Adelie Land, Rep. Austr. Antarctic Exped., Ser. A, vol. 3, pt. 4.

Sugi, K., 1931. On the Metamorphic Facies of the Misaka Series in the Vicinity of Nakagawa, Rov. Sagami, Jap. Jour. Geol. and Geogr., no. 2, pp. 87–142.

Sundlus, N., 1930. On the Spilitic Rocks, Geol. Mag., vol., 67, no. 1, pp. 1–17.

Thomson, J. A., 1910. Note on some Rocks from Parapara, Bluff Hill and Waikawa, Trans. N.Z. Inst., vol. 49, 1909, pp. 33–39.

Tilley, C. E., 1921. The Granite-Gneisses of Southern Eyre Peninsula (South Australia) and their Associated Amphibolites, Q.J.G.S., vol. 77, no. 306, pp. 75–134.

— 1923. The Petrology of the Metamorphosed Rocks of the Start Area (South Devon), Q.J.G.S., vol. 79, no. 314, pp. 172–204.

— 1924. Contact Metamorphism in the Comrie Area of the Perthshire Highlands, Q.J.G.S., vol. 80, no. 317, pp. 22–71.

— 1926. On Garnet in Pelitic Contact Zones, Min. Mag., vol. 21, no. 113, pp. 47–50.

Turner, F. J., 1930. Metamorphic and Ultrabasic Rocks of the Lower Cascade Valley, South Westland, Trans. N.Z. Inst., vol. 61, pp. 170–201.

— 1933a. The Metamorphic and Intrusive Rocks of Southern Westland, Trans. N.Z. Inst., vol. 63, pp. 178–284.

— 1933b. The Genesis of Oligoclase in Certain Schists. Geol. Mag., vol. 70, pp. 529–541.

— 1935a. Metamorphism of the Te Anau Series in the Region Northwest of Lake Wakatipu, Trans. Roy. Soc. N.Z., vol. 65, pp. 329–349.

— 1935b. Contribution to the Interpretation of Mineral Facies in Metamorphic Rocks, Amer. Jour. Sci., vol. 29, pp. 409–421.

Vogt, J. H. L., 1909. On Labradorite-Norite with Porphyritic Labradorite Crystals, Q.J.G.S., vol. 65, no. 258, pp. 81, 103.

Vogt, T., 1927. Sulitelmafeltets geologi og petragraphi, Norges Geologiske Undsersökelse, no. 121, Oslo (English Summary, pp. 449–560).

Wild, L. J., 1911. The Geology of the Bluff, New Zealand, Trans. N.Z. Inst., vol. 44, pp. 317–339.

Williams, G. H., 1886. The Gabbros and Associated Hornblendic Rocks in the neighbourhood of Baltimore, Bull. U.S.G.S., no. 28.

Williams, G. J., 1934. A Granite-Schist Contact in Stewart Island, New Zealand, Q.J.G.S., vol. 90, pt. 3, pp. 322–353.

Wiseman, J. D. H., 1934. The Central and South-west Highland Epidiorites: a Study in Progressive Metamorphism, Q.J.G.S., vol. 90, pt. 3, pp. 354–416.