Go to National Library of New Zealand Te Puna Mātauranga o Aotearoa
Volume 66, 1937
This text is also available in PDF
(4 MB) Opens in new window
– 231 –

Basic and Ultrabasic Rocks in North-west Otago

[Read before the Otago Institute, November, 1935; received by the Editor, November 15, 1935; issued separately, December, 1936.]



Previous Work

Distribution and Occurrence of the Basic and Ultrabasic Rocks


  • Gabbroidal rocks

  • Nephrites and tremolite-rocks

  • Serpentinites

  • Talc-schists

  • Miscellaneous rocks associated with the ultrabasic rocks

Summary of Alteration Phenomena

Causes of Alteration


List of Specimens Examined

List of Literature Cited


The subject matter of this paper is the result of investigation of small basic and ultrabasic intrusions in four separate localities in Western Otago, viz., the Caples, Greenstone, Routeburn, and Springburn Valleys. The first three valleys are situated in mountainous country west of Lake Wakatipu, into which the waters of the Routeburn, Greenstone, and Caples Rivers eventually flow, while the Springburn Creek, lying about 18 miles east of Queenstown, is a tributary of Gentle Annie Stream, itself a tributary of the Kawarau River.

Collecting and geological mapping were carried out as thoroughly as possible, but the rugged and heavily forested nature of the country west of Lake Wakatipu greatly restricted the tracing of outcrops in that region.

Previous Work.

The earliest mention of the occurrence of ultrabasic or basic rocks west or north-west of Lake Wakatipu is contained in S. H. Cox's (1879) report to Hector on the Wakatipu and Greenstone District, in which he records the occurrence of black serpentine in thin veins in dark-coloured jointed slates in the Greenstone Valley. A year later, McKay (1881) mentioned that serpentine associated with the rocks of the Te Anau Series cuts across the Greenstone Valley, while Park (1887) recorded the occurrence of ultrabasic rocks in a region north of the Routeburn Valley. Still later Marshall (1906) examined ultrabasic rocks at the headwaters of Olivine Creek and elsewhere.

During his survey of the Cromwell Subdivision, Park (1908) investigated the outcrop of serpentine in Springburn Creek, and subsequently Henderson (1923) observed there outcrops of chloritetalc-schists in addition to the serpentine mass described by Park. No

– 232 –

further work appears to have been done until 1932, when the writer explored the entire Caples Valley and located there six outcrops. In 1933 Dr F. J. Turner examined a serpentinous mass in the lower Routeburn Valley and collected from it many specimens, a few of which were described by Dr Turner himself (1935), while the bulk of the material was handed over to the present writer for examination. In 1935 the writer revisited the Caples and Greenstone Valleys and spent several days at the Springburn outcrops.

Distribution and Occurrence of Basic and Ultrabasic Rocks.

Caples Valley.

Basic and ultrabasic rocks here occur in six localities. The most northerly outcrop is a roughly circular, dome-shaped mass, approximately five chains in diameter, rising with steep or vertical sides to a height of 75ft. above the surrounding alluvium; the rocks of which this “Knob” is composed have been smoothed off and striated by ice. The “Knob” is made up entirely of basic rocks, and in every way appears closely similar to the isolated structures described by Read (1931, p. 77) from central Sutherlandshire. The second outcrop, which occurs some four chains to the south-east of the “Knob,” is an oval mass of dark-green serpentine covering an area of approximately 40yds × 20yds., and striking about north and south. It is bounded on the west by contorted greywackes and on the east by sub-Recent alluvial gravels. Rather more than 30yds. south-west from this last occurrence there occurs a narrow band, approximately 4ft. wide, of pale greyish to green nephritic rocks; its extension in a southerly direction could not be traced on account of dense vegetation. This outcrop is bounded by steeply dipping greywackes and at the western boundary the component rock is nodular and strongly slickensided. Proceeding further south the next outcrop is that of a dark-green serpentine exposed for a distance of about 300yds. along the precipitous bed of Talc Creek. It appears first at an altitude of 1390ft., but cannot be traced beyond an altitude of 1650ft. in the creek bed. At this point, where it adjoins the greywackes on the west, there are six or seven bands of pale-green talc, one to three inches wide, their strike being approximately 15°–25° west of north. In several parts of the serpentine mass zones of slickensiding may be seen, and these are filled with calcite or with crushed serpentine cemented by calcite, and generally strike about north and south. About eight chains to the south of this locality four more outcrops occur in the bed of a steep watercourse, Beech Creek, the largest being about one chain in length. Finally, some twelve chains further south two small outcrops of dark-green and greyish ultrabasic rocks occur. Dense intervening forest-growth rendered it impossible to observe if the isolated outcrops described above were continuous across the dividing ridges.

Greenstone Valley.

Ultrabasic rocks were found approximately 4 ½ miles up the valley from its junction with Caples Valley. They are a series of dark-green and greyish-green serpentines with talcose bands, but

– 233 –

their relation to the greywackes was obscured by a great slip which had come down from the adjacent mountain-side.

Routeburn Valley.

The intrusion in this locality is marked by a slip just where the pack-track crosses it, about 2 ½ miles above the junction of the Routeburn with the Dart River. It is a mass of slickensided greyish-green tremolitic rock together with some serpentine, invading sheared slates. A tributary joining the Routeburn a little to the east contained many boulders of ultrabasic rocks indicating the presence upstream of a second intrusion.

Springburn Valley.

The outcrop of igneous rocks occurs at a point about 5 miles from the junction of Gentle Annie Stream with the Kawarau River. It is an oval mass occupying an area of about half an acre, with the major axis lying about 40°–50° west of north, parallel with what appears to be a fault. It is mainly a massive dark-green serpentine with narrow veins of amphibole asbestos, bands of similar but much coarser asbestos 3in. Wide, and veins of carbonates which are often slickensided.


Gabbroidal Rocks.

The Caples Valley “Knob” is composed entirely of gabbroidal rocks described below as serpentinised olivine-gabbros, augite-picrites, pyroxenites, saussurite-gabbros and epidiorites. The serpentinised olivine-gabbros are coarse-grained rocks relatively unsheared and composed essentially of saussuritised plagioclase, ferruginous mesh-serpentine and large schillered plates of diallage. With increase of pyroxene, these rocks grade into augite-picrites and eventually into pyroxenites. The pyroxenites have suffered much crushing and typically contain granulated and uralitised diallage, with sphene and epidote, set in a matrix of chlorite. The saussurite-gabbros are even more sheared and chemically reconstituted, and as a result are usually fine-grained. In them unaltered plagioclase is rarely observed, while diallage occurs as granulated and altered porphyroclasts spread throughout a saussuritic matrix derived from the original feldspar. Finally the epidiorites are sheared, originally rather feldspathic rocks, which though preserving the original gabbroidal texture are largely reconstituted to a mixture of albite, epidote, uralitic hornblende, and chlorite. In them the mechanical effects of shearing are less marked than in the saussurite-gabbros, and the rocks themselves were originally more feldspathic than is usual with the latter.

Careful field and petrographic examination of the “Knob” showed rapid variation in mineralogical composition and textural features of the component rocks, for blocks of almost unsheared augite-picrite appear to be embedded in intensely sheared saussurite-gabbros, while pyroxenites are surrounded by epidiorites. There are two changes common to all these rocks: (1) they have all suffered to a greater or less extent from shearing; (2) they are all derivatives of

– 234 –

mineralogically related basic types. Therefore it is suggested that the original rock-mass was a heterogeneous gabbro with local segregations of pyroxenite and sometimes picrite, which was later subjected to intense but local shearing accompanied by considerable mechanical disruption and displacement along closely-spaced zones, thus producing a block arrangement of the component rock-types, and accentuating the initially heterogeneous condition of the mass as a whole.

Serpentinised olivine-gabbros.

The serpentinised olivine-gabbros occur in the Caples Valley “Knob,” where they make up several rectangular blocks. No. 2560 in the hand-specimen is a coarsely crystalline rock with crystals of bronze-coloured pyroxene reaching 30mm. in length. Microscopically it is holocrystalline and contains the following minerals: saussurite 50%, serpentine 20%, diallage 25%, and accessories 5%. Original feldspar is now completely altered to a saussuritic mass with which is associated numerous flakelets of a clear, colourless mineral with medium birefringence and positive elongation; this may be sericite or prehnite, but is too fine-grained for precise determination. The pyroxene is a strongly schillered diallage with an average length of 2.0mm. Simple orthopinacoidal twinning is frequent, while a good diallage parting is usual; the extinction angle, Z to c, is 42°–44°. Magmatic reactions have produced partial rims of brown, poorly pleochroic hornblende, while in other places recrystallisation has occurred with the production of fringes of colourless tremolite. Large areas of fibrous serpentine (up to 4.0mm. in diameter) intersected by wavy lines composed of fibres in parallel arrangement, reproduce a type of mesh-structure. The fibres in these areas may lie in parallel position or else haphazardly, in which case they appear almost isotropie when viewed between crossed nicols. The mineral is pale yellowish-green with faint pleochroism. The fibres have positive elongation and a maximum birefringence of about 0.01–0.012 as determined by means of a quartz wedge, while the cores of the meshes often have an anomolous interference tint. This mineral, a ferruginous chrysotile, has here been derived from olivine. A second and similar serpentine forms what appear to be pseudomorphs after green hornblende, and though its optical properties tally with those given above, it does not reproduce the mesh-structure so well. Enclosed relicts of green hornblende are present, while its cleavage has been preserved in the serpentinous pseudomorphs. Coarse tremolite, 0.2–0.4mm. in length, fringes the diallage and occurs as prisms embedded in or stabbing the serpentine. Green hornblende appears to be entirely secondary and is present as pale-green poorly pleochroic plates up to 0.75mm. in length with variable birefringence never exceeding about 0.02. Iron-ores are rare, and when present are much altered; rarely there is a little leucoxene. In sections cut from two other specimens the amount of diallage reaches 60–70%. Following Hatch and Wells (1926, p. 256) these two latter rocks have been termed augite-picrites.

– 235 –

Altered pyroxenites.

No. 2537 is a medium-grained massive rock containing large crystals of pyroxene; it weathers to a yellow clay. In thin section the rock is holocrystalline and shows effects due to shearing. The component minerals are diallage 50%, hornblende 10%, chlorite 30%, calcite 7%, epidote, sphene, etc., 3%. The pyroxene is diallage in fractured, granulated and twisted grains averaging 0.85mm. in diameter, and is set in a chloritic matrix. Schiller inclusions are often dense, though frequently only locally developed in any particular crystal-grain, and undulose extinction common. The pyroxene is much altered to chlorite or less frequently to tremolite. Primary brown hornblende, which occurs as reaction-rims to the pyroxene, as peculiar flecks in that mineral, and as separate crystals, is often somewhat bleached with consequent loss of schiller inclusions. The hornblende may be altered to chlorite, calcite and granular sphene, while less frequently it may be fringed with colourless tremolite, or rarely a pale-green hornblende. Two varieties of chlorite occur. The first is a pale-green poorly pleochroic type occurring in a felted mass of tiny platelets in which are often embedded relicts of diallage and hornblende. The fibres have positive elongation and a deep-blue anomolous interference tint. A second more ferruginous type, also replacing pyroxene, has positive elongation but stronger birefringence (0.012–0.014) and a weak pleochroism, yellow (X) to yellowish-green (Y = Z). Calcite in irregular grains, 0.4mm. in diameter, is usually secondarily twinned, while rare iron-ores are much altered to leucoxene.

No. 2538 is macroscopically similar to No. 2537, but not so strongly sheared. The component minerals are diallage 40%, chlorite 25%, epidote 15%, stilpnomelane 10%, calcite 5%, and albite 5%. Diallage occurs in clear striated grains, 0.8mm. to 1.0mm. in diameter, often broken and granulated, and quite free from schiller inclusions, while undulose extinction is strong. It is little altered, though some serpentine, stilpnomelane and epidote may have arisen from it. Rounded areas of “bowlingitic” serpentine optically similar to that described above occasionally show a type of mesh-structure. Pale bluish-green strongly pleochroic epidote in a peculiar granular form, and stilpnomelane in well-developed highly birefringent platelets, often radially arranged, are always closely associated. Granular sphene, rare iron-ore, secondarily twinned calcite and albite with clonozoisitic-epidote inclusions complete the mineral assemblage.


(Nos. 2526, 2527, 2528, 2530, 2544). The saussurite-gabbros are green to greenish-black speckled rocks of very variable size of grain, which is dependent upon the amount of shearing they have suffered. In thin section cataclastic effects are shown in all specimens, while occasionally bands of rock (No. 2528) have become completely recrystallised. Diallage is usually the most abundant constituent, but falls to about 35% in the most altered rocks (No. 2526). Diallage-lamination, sometimes confined to only a portion of a grain, is well developed, while schiller-structure is present in less altered types (No. 2528). In other more altered specimens the schiller inclusions

– 236 –

have probably been absorbed (see Flett and Hill, 1912, p. 88). Occasional grains show orthopinacoidal twinning. The plates seldom exceed 4.0mm. in length and may be fractured, bent, granulated or intensely crushed, the porphyroclasts being spread throughout a saussuritic matrix. Stretching of grains (No. 2544; see Harker, 1932, p. 157) and undulose extinction were noted. In the less altered rocks (No. 2528) the pyroxene is altered to green hornblende with a bluish-green tint for the Z vibration direction, and chlorite; in other cases to chlorite alone, chlorite and tremolite, or to a mixture of pennine, epidote, carbonate, possibly prehnite (No. 2530) and a little tremolite. Ilmenite or sphene are often associates (Nos. 2527, 2526 and 2544), the former having sometimes separated out along the pyroxene cleavages. In No. 2530 dusty material, probably finely divided epidote, appears similarly to have separated, while rarely incipient replacement by talc and carbonate was noted. Rare pale-brown hornblende altering to green hornblende or tremolite occurs as a reaction border to diallage, or as small flecks, 0.3mm. in length, embedded in and elongated along the cleavage-lines of the pyroxene (No. 2528). Green hornblende is always secondary and rarely (No. 2528) may occur in one-half of the rock (to the exclusion of pyroxene) as streaked-out grains embedded in saussurite, imparting local linear schistosity. In most rocks the feldspar is now represented by saussurite, often with sericite (No. 2527), but sometimes (Nos. 2530 and 2544) large twinned cracked and somewhat granulated crystals of albite may occur. Recrystallisation has taken place along the cracks, and in the crystalloblastic grains patchy secondary lamellar twinning may be developed. Dense masses of clinozoisitic-epidote and perhaps prehnite are associates. Chlorite occurs in clear green pools (Nos. 2544 and 2526), often associated with tremolite and perhaps zoisite (No. 2544), while uncommon iron-ores include magnetite, replaced by leucoxene in the more altered types, leucoxenised ilmenite (No. 2527) and pyrites (No. 2528). Twinned calcite was noted in Nos. 2530 and 2526, where it constitutes 5% of the rocks.


(Nos. 2524, 2531, 2533, 2535, 2536, 2540, 2549). Macroscopically the epidiorites are black, fine-grained, massive rocks passing into greyish-green types when strongly sheared. In the less altered types green hornblende is the dominant coloured silicate, its place being taken to a large extent by chlorite in the more altered rocks. The amphibole commonly occurs in clear frayed grains, somewhat schillered immediately adjacent to rare relicts of augite (No. 2549); in less altered types twinning was noted (No. 2549). The Z vibration direction frequently has a bluish-green to olive-green tint, while sometimes the edges of a grain may assume a deep-blue tint (No. 2549). The hornblende shows several types of alteration in different rocks and even in different parts of the same rock. The changes noted were to chlorite alone (Nos. 2531 and 2536), or together with epidote and sphene (No. 2549); to tremolite and chlorite (No. 2531), or to tremolite alone (Nos. 2524, 2531 and 2540). Augite is rare (Nos. 2549 and 2531) and is always surrounded by green hornblende. In the less crushed rocks feldspar occurs as lathy or tabular twinned crystals approximating to nearly pure albite and associated with

– 237 –

needles of clinozoisite (Nos. 2533 and 2549), or clinozoisitic epidote and minor sericite (No. 2531). Rarely the epidote mineral may be rather ferruginous (No. 2535). Increase in shearing stress has caused granulation of feldspar so that in highly altered types a fine-grained saussuritic matrix forms much of the rocks. Stilpnomelane, in well-developed platelets, occurs in Nos. 2533, 2535 and 2549. In two cases (Nos. 2533 and 2535) it has developed from chlorite, while in No. 2549 it is closely associated with iron-ore or with hornblende. In this rock it may also be concentrated round clusters of epidote grains. Chlorite varies in mode of occurrence and in composition. A pale-green negative pennine, comparable to the delessite of Winchell (1933, p. 282) is common in irregular or occasionally (No. 2533) radially arranged plates, and may occur alone or with other varieties. In No. 2535 a second type forms peculiarly shaped aggregates of fibres, almost leaf-like in form; their elongation is negative, birefringence approximately 0.008–0.010, while a strong pleochroism follows the scheme:

X = Y = bluish-green.
Z = colourless.
X = Y > Z

A positive pennine may also be present in some rocks. Rarely chlorite is lacking (No. 2524). Clusters of clinozoisite needles are sometimes embedded in pools of chlorite (Nos. 2531 and 2535). Noteworthy in one rock (No. 2524) is the presence of ovoid patches composed entirely of water-clear xenoblastic grains of albite and idioblastic prisms of clinozoisite, rarely exceeding 0.4mm. in length; veinlets of clinozoisite, 0.3mm. wide by 3.0mm. long, occur with the prisms of the mineral oriented perpendicularly to the vein walls. It would appear that these segregations of clinozoisite may represent products of metamorphic differentiation (see J. S. Flett and J. B. Hill, 1912, p. 50; J. A, Dunn, 1929; P. Eskola, 1932; cf. also metamorphic differentiation of quartz-kyanite-rocks, H. H. Read, 1933, pp. 323–324).

Accessory constituents include magnetite (Nos. 2535, 2540, 2549), leucoxenised magnetite (Nos. 2531, 2536), skeletal crystals of ilmenite, often altered (Nos. 2533, 2535, 2549), sphene (Nos. 2531, 2536, 2540), calcite (Nos. 2535, 2536) and apatite (Nos. 2533 and 2549).

Nephrites and Tremolite-Rocks.

Included here are rocks composed entirely or almost so of tremolite. The writer follows Dr. Turner's classification, which is founded essentially on microstructure (Turner, 1935, pp. 189–191). The rocks grade from typical nephrites consisting of uniformly small oriented bundles of fibres which are twisted and interwoven with one another (nephritic structure) to tremolite-schists composed of fine to coarse prisms of tremolite. Intermediate between these extreme types are the semi-nephrites which consist partly of fine fibres and partly of coarser prismatic crystals often arranged in sheaves. Also to be described under this caption are the tremolite-phyllonites, rocks which have suffered extreme shearing. Finally there are three groups in which important amounts of chlorite, talc, or talc + carbonate respectively enter into the composition.

– 238 –

Non-schistose Nephrites.

No. 2529 is macroscopically a non-schistose, pale greyish-green nephrite; it has the characteristic splintery fracture and a hardness of approximately 5 on a smooth surface. It is composed entirely of a mat of twisted and interfelted prisms of tremolite, varying in length from 0.01mm. to 0.2mm. Occasional bunches of fibres in parallel or sheaf-like arrangement are set in the finely nephritic base. The extinction angle (Z ∧ c) of the amphibole is 20°. A vein of tremolite, 0.4mm. wide, cuts the rock, while rare platelets of pale-green negative chlorite and a few clusters of chromite grains complete the association of minerals.

No. 2542 is similar to No. 2529, the nephritic structure being very well developed. Iron-ores are lacking. Nos. 1802 and 1804 described by Turner (1935, p. 189) are closely similar to No. 2542, though distinctly harder in hand-specimen.


(Nos. 2552, 2567). No. 2567 is a pale-green rock with poor linear schistosity, having a splintery fracture and a hardness of about 4.5–5. Microscopically the rock is closely similar to No. 1829 described by Turner (1935, p. 190). The tremolite has an extinction angle (Z to c) of 16–20°, and a negative, biaxial figure, the optic sign ruling out any possibility of the mineral being cummingtonite. In addition, monomineralic patches of two varieties of chlorites occur, one being a delessite and the other a positive type with deep-brown tints between crossed nicols. They are both almost colourless and do not appear to be closely interleaved.

No. 2552 is similar to No. 2567, but contains abundant chlorite, again in two varieties as before; these appear to have crystallised simultaneously under similar physical conditions. Noteworthy are peculiar “clots” (0.1–0.15mm. in diameter), usually embedded in chloritic patches and also scattered abundantly throughout the rock, composed of curiously twisted and entwinned hair-like fibres of tremolite. A little chromite was noted.


The tremolite-schists are greenish-grey to pale-green, poorly fissile rocks, often crushed, contorted and slickensided, with a hardness on a polished surface varying from 3–5.5.

No. 2565 is typical of six rocks (Nos. 2565, 2566, 2568, 2571, 2577, 2580) which possess plane-schistosity. It is composed almost entirely of fibrous tremolite, not exceeding 0.1mm. in length (though in a few specimens, Nos. 2577 and 2580, the maximum length is 0.25mm.), which, while showing no regular orientation in a section cut parallel to the fissility, also lack the twisted felted disposition that characterises nephritic structure. Accessories in these rocks include chlorite (optically positive—Nos. 2565, 2571, 2577), or associated positive and negative chlorites (No. 2568), relict green and brown hornblende (Nos. 2577, 2580), chromite (No. 2565), sphene (all six), and magnetite (Nos. 2568, 2571, 2577).

– 239 –

No. 2534 is somewhat similar to No. 1845 described by Turner (loc. cit., p. 191) in that it indicates transition between the tremoliteschists with plane-schistosity on the one hand and the semi-nephrites on the other. Rare accessory constituents include granular yellowish garnet, which, following Turner (loc. cit., p. 193) is a member of the grossularite-uvarovite series, and magnetite altering to haematite.


(Nos. 2543, 2554, 2587.) This group includes rocks characterised by an abundance of both tremolite and chlorite, but essentially they do not differ from the previous group into which they pass with decrease of chlorite.

No. 2554 is a dark-green, poorly schistose, strongly slickensided rock. Poorly oriented bunches of parallel acicular prisms of tremolite make up half the rock, and in addition tremolite may occur as “clots” embedded in chlorite, or forming narrow veinlets. Two chlorites, one optically positive and colourless, and the other negative and pale-green, occur interfelted with tremolite or with each other. Rare chromite forms the nucleus of chloritic patches, a feature recalling that described by Turner (loc. cit., p. 192).

No. 2587 is a pale greenish-grey poorly fissile rock. Coarse distinct prisms of tremolite, rarely oriented, average 0.15mm. in length, but may measure 0.4mm. by 0.2mm. The rest (40%) of the rock is composed of two chlorites. The properties of the amphibole and chlorite may be quoted in full as being typical of the chief constituents of the whole group of nephritic and allied rocks.

The amphibole is colourless, optically positive and biaxial; Z to c is 22°. The refractive indices are as follows:—

α = 1.620 ± 0.001
γ = 1.632 ± 0.001
γ − α = 0.012 ± 0.001

The more abundant of the two chlorites is negative like the delessite of Winchell, with a pleochroism which follows the scheme:—

X = yellow.
Y = Z = green.
X < Y = Z

The refractive index β = 1.595 ± 0.002. The second type is optically positive and deeper-green in colour.

No. 2543 indicates a transition towards the chlorite-schists. The bulk of the rock is composed of positive pennine together with minor delessite. Isolated fibre-like prisms of tremolite occur in short stumpy groups and as sprays fringing hornblende relicts. Sphene is plentiful, while iron-ores are rare. A few veinlets of tremolite, tremolite with chlorites or chlorites alone were noted.


Only one phyllonite, No. 2541, was obtained. Macroscopically it is a soft, grey, strongly fissile, slickensided rock, rather phyllitic or slate-like in appearance. The bulk of the rock is composed of tiny interleaved platelets of both optically positive and negative colourless

– 240 –

pennines, the former being the more abundant. Tremolite (25%) occurs in unoriented distinct prismatic crystals and also as fringes to relict hornblende (green and brown) and rare riebeckite. Narrow veinlets of tremolite, finely granular iron-ore, epidote and sphene complete the mineral association.

The schistosity and phyllitic appearance of this rock are the result of mechanical degradation of an original coarsely crystalline tremolite-chlorite-schist, and it has now all the features characteristic of a phyllonite as defined by E. B. Knopf (1931, p. 14). This rock, however, is not a product of retrogressive metamorphism, that is, it is not diaphthoretic.


Here are placed tremolite-schists in which talc ranks as an essential constituent.

No. 2520 is a massive green, highly contorted, non-schistose, strongly slickensided rock with a hardness of about 2–3. Tremolite occurs in unoriented prismatic crystals with occasional local development of nephritic structure. Abundant fine-flaky talc is scattered among the tremolite fibres or aggregated into dense patches. Chromiferous garnet occurs as strings and groups of greenish-yellow, dusty, xenoblastic grains which frequently surround a nucleus of opaque iron-ore, perhaps chromite. Pale-green serpentine is rare.

Macroscopically No. 2548 is a massive green, rather talcose rock, strongly slickensided and contorted. Half the rock is composed of acicular crystals of tremolite, usually with parallel orientation, crossed by narrow strain-slip bands; a little semi-nephritic base also occurs. Abundant fine-flaky talc is intimately associated with the tremolite and may occur in elongated lensoidal areas (the result of shearing) fringed by sub-parallel fibres of tremolite oriented approximately parallel to the margins. A little weathered iron-ore completes the rock assemblage.

No. 2570 is a pale-green rock with strong linear schistosity. Markedly parallel acicular prisms of tremolite sometimes showing strain-slip bands are intimately associated with plentiful pale-green almost isotropic serpentine. Talc flakelets, aggregated into patches or scattered among tremolite or serpentine, appear to be developing from the latter. A little positive pennine, and uncommon grains of magnetite and sphene are accessory constituents.


Tremolite-talc-calcite-schists are to be found as narrow zones within the Springburn intrusive, and No. 2589 is typical. It is a coarsely fibrous greenish rock with a strong linear schistosity, resulting from the parallel development of very abundant amphibole. This mineral is biaxial and negative, and the extinction angle, Z to c, is 18°. Refractive indices are as follows:—

α = 1.610 ± 0.001
γ = 1.625 ± 0.001
γ—α = 0.015 ± 0.001
– 241 –

Plentiful talc occurs throughout, while carbonate is abundant in irregular untwinned grains and often in perfect rhombohedra, averaging 0.3–0.4mm., which seem to have crystallised after shearing ceased. A little pale-green serpentine was noted, with a low birefringence and positive elongation. No undoubted instance was observed of talc originating from this mineral, though replacement of serpentine by talc is seen in other rocks described later. It is interesting to note here that this rock is closely comparable with No. 1840 described by Turner (loc. cit., p. 191) and obtained from a Maori camp-site in the Shag River Valley, North Otago. A Maori route passed within three miles of this locality.


The serpentinites described below have been divided into four groups: (1) mesh-serpentinites with bastite; (2) the sheared equivalents of group one; (3) serpentinites composed essentially of antigorite; (4) a group characterised by abundant talc in addition to serpentine, representing a transition towards the talc-rocks described in a later section.

Optical Properties of the Serpentine-minerals.

In view of the variation in the optical properties of the serpentine-minerals, it is considered advisable to define the terms used in literature and in the descriptions to follow. It appears that only two types, viz., chrysotile and antigorite have definite distinctive properties enabling them to be determined with certainty by microscopic means. The former is the fibrous variety with comparatively small optic axial angle and positive optic sign, while the elongation of the fibres is always positive (i.e., parallel to the acute bisectrix); this mineral appears to crystallise under static conditions only. Antigorite, which on the other hand is a typical “stress mineral,” has positive elongation (perpendicular to the acute bisectrix), with a biaxial and optically negative character, and characteristically crystallises in colourless micaceous flakes, tabular parallel to the 001 cleavage. This orientation differs from that in a figure given by Winchell (1933, p. 281, fig. 204), which incorrectly shows the elongation as parallel to the negative acute bisectrix X, although in the accompanying text it is stated that the elongation of lamellae is positive. However, the writer has had the opportunity of examining sections of antigorites from type localities such as the Swiss Alps, the Mikonui area of New Zealand, the Great Serpentine Belt of New South Wales, and less-known regions such as the Olivine Range of south-west New Zealand, and in every case the elongation of lamellae was positive and the optic sign negative.

The term bastite has been limited to include serpentine pseudomorphous after non-aluminous pyroxene and showing parallel fibrous structure. In the rocks described it would seem that even in the mineral thus defined the optical properties are not constant. It

[Footnote] * The term serpentinite has been proposed by Lodochnikow (1933, p. 145) for rocks consisting almost exclusively of minerals of the serpentine group, and is here used in the same sense.

– 242 –

has low polarisation tints and orthopyroxene cleavage planes are usually clearly retained, though Bonney and Raisin (1905, p. 694) state that these become “as it were, soldered up.” It varies from a non-ferruginous colourless variety, presumably after enstatite, to a green ferruginous type, possibly pseudomorphous after bronzite or perhaps hypersthene. In the latter case it then shows distinct though sometimes slight, pleochroism, being pale-green for Z and yellow or greeenish-yellow for X. The elongation of the bastite fibres is constant, being parallel to Z, but the sign and optic axial angle vary. The non-ferruginous type appears to be constantly negative and biaxial, 2V being small, while the ferruginous type is often positive, with 2V variable. All varieties appear to be devoid of inclusions of iron-ores. Therefore the evidence obtained from the study of the serpentines described below is not in support of Winchell's (1933, p. 281) contention that bastite is a coarse variety of antigorite, particularly when regard is paid to the character of the optic sign and the fibrous crystalline form.

In addition to bastite, deep-green, pleochroic, serpentine with rather different properties occurring as parallel fibrous aggregates or plates is recorded. Firstly, this serpentine mineral has a moderate birefringence, showing red of the first order, when bastite gives only greys and white; and, secondly, it has a rather dusty appearance in contrast to the clearness of bastite. Following Bonney and Raisin (1905, p. 696–7) it is believed that this mineral may represent serpentinised augite, perhaps a poorly aluminous diallage.

Serpophite is a term suggested by Lodochnikow (1933, p. 145) for serpentines which either are microscopically textureless or possess an indefinite texture which is variable within but small microscopic areas. Following Lodochnikow, therefore, the writer has adopted this term for extremely finely crystalline, textureless serpentine which lacks the typically fibrous habit of chrysotile or the flaky or micaceous form of antigorite. Macroscopically massive serpentine-rocks are often composed largely of serpophite.

Finally, there remain the dark-green serpentines with mesh-structure, and the pale-green fibrous types which often crystallise as a result of the shearing of the former. The birefringence of the former variety is low to medium, but that of the latter is usually very low; the elongation of the crystals in both cases is positive (perpendicular to the acute bisectrix) and while the optic sign is always negative (contrast with chrysotile) the axial angle varies. Such minerals, though they are in some respects optically similar to antigorite, have not developed the micaceous habit typical of that mineral, and are therefore merely classed as fibrous negative serpentines.


(Nos. 2521, 2546, 2551, 2558, 2561, 2555). The mesh-serpentinites include rocks composed mainly of serpentine with well-developed mesh-structure, with or without bastite. In the hand-specimen they are dark-green, dull or sometimes enamel-like, massive rocks, often exhibiting lustrous plates of bastite. Frequently they are strongly slickensided and traversed by veins of fibrous serpentine or a carbonate.

– 243 –

No. 2546 is typical and may be described in detail. The bulk of the rock is composed of pale yellowish-green mesh-serpentine, the individual fibres averaging 0.5mm. in length. The birefringence is not greater than 0.014, and there is strong pleochroism according to the scheme:—

X = yellowish-green.
Z = green.
X < Z

The cores of the meshes are frequently occupied by colourless serpentine, perhaps serpophite, together with finely granulated iron-ore. Pale-green sometimes colourless plates of bastite, up to 2.0mm. in diameter, are faintly pleochroic, slightly biaxial and negative, and have retained the original cleavage of pyroxene. Occasional pseudomorphous plates of pale-green, dusty serpentine which do not differ essentially from the bastite except that they are a deeper green and more strongly pleochroic, also occur. Occasional indistinctly bounded areas are composed of fine flaky antigorite, granular carbonate and a little iron-ore. Primary iron-ores are quite abundant as clusters of large rounded grains up to 2.5mm., but averaging 0.75mm. in diameter. They are quite opaque, like magnetite, and in this respect are strikingly similar to the spheroidal chromite described and analysed by Benson (1914, p. 681). A few cubes and octahedra are associated with the large grains, but in only one was there any sign of a brown colour due to transmitted light. On comparing Dr. Benson's section from Paling Yard with No. 2546, the writer prefers to identify the black ore as chromite rather than magnetite. Irregular or occasionally cuboidal grains of secondary magnetite and rare pyrites are scattered throughout the rock. Finally, this latter is cut by narrow veinlets of colourless fibrous material with a moderately high birefringence and a positive, slightly biaxial figure; with these properties it seems best to refer this mineral to brucite.

Nos. 2555, 2561 and 2551 are similar types; in the last, bastite plates are often fractured, the cracks being filled with fibrous serpentine. No. 2551 contains as much as 25% of bastite and serpentinous pseudomorphs, while No. 2521 is almost devoid of these constituents. The latter rock contains numerous ramifying veins of slightly biaxial and optically positive fibrous chrysotile. The last two rocks (Nos. 2551, 2521) contain a little secondary iron-ore. Antigorite-carbonate aggregates were noted in Nos. 2555, 2551, 2561, and 2558, and rare brucite in Nos. 2551 and 2561.

Sheared Serpentinites.

Under this caption are described two serpentinites, Nos. 2547 and 2562, which, in the hand-specimen, are both dark-green, massive, contorted rocks, the former containing numerous narrow veinlets of fibrous material.

In thin section No. 2547 appears to have been a mesh-serpentine similar to No. 2546, but it is now much altered as a result of shearing. Several shear-zones filled with torn and recrystallised serpentine are present. Original mesh-serpentine is rarely seen, the mineral being now represented by a finely fibrous pale-green type. Large,

– 244 –

sometimes dusty, non-pleochroic serpentinous pseudomorphs, twisted and fractured, have slightly higher birefringence and deeper colour than the above type. The fibres have positive elongation, while an aggregate interference figure is uniaxial and positive. Occasional fragments of bastite occur. Primary iron-ore is similar to that described earlier in discussing mesh-serpentinites. Several semiopaque pale brownish masses, 0.85mm. in diameter, which appear white in reflected light, are believed to be alteration products derived from aluminous pyroxene (see Benson, 1914, p. 671; Bartrum and Turner, 1929, p. 121; and Turner, 1930, p. 187). As a result of shearing, the rock has become partially recrystallised to a colourless chlorite with low birefringence and a uniaxial positive figure. It occurs in wide monomineralic patches, along shear-zones and as borders to grains of primary iron-ore. In the latter case the fibres of chlorite are usually unoriented, but rarely a radial structure can be made out. Associated with this chlorite is a little delessite. Secondary iron-ore occurs as clouds thrown out from the serpentine during shearing. Veinlets consist of talc or serpentine.

No. 2562, also sheared, has arisen from a serpentine similar to No. 2546. The structure is cataclastic with sheared fragments of the original rock embedded in recrystallised matrix, which is a pale-green, finely fibrous and optically negative serpentine with a lower refractive index and birefringence than the original mineral; elongation is positive. It is devoid of the clouds of iron-ore particles present in the original serpentine. Chromite identical with that described earlier for mesh-serpentinites is a minor constituent.


Antigorite-serpentinites are represented by two rocks. The first, No. 2519, is massive, dark-green and traversed by narrow veinlets of carbonate. In thin section the dominant constituent is antigorite in colourless blades, arranged haphazardly or in sheaf-like bundles radially disposed to a nucleus; also in parallel groupings. The flakes vary in size from the dimensions of microantigorite of Lodochnikow (1933, p. 145) to 0.3mm. in length. There are large fissured plates of pleochroic green serpentine with birefringence of about 0.018, in which replacement by antigorite is clearly shown, the flakes lying at random within the relicts, but more frequently spreading or fraying out from them. Fine flaky talc is embedded in and is replacing this green, serpentine. It is also abundant as aggregates of minute platelets from which there frequently radiate blades of antigorite, or as sinuous veinlets which pass alike through relicts of serpentine and aggregates of antigorite. It is definitely of later formation than the antigorite, and in numerous instances this mineral is itself flecked with platelets of talc indicating a replacement of the former by the latter. Of simultaneous or perhaps even later formation are numerous veinlets of twinned carbonate averaging 0.5mm. in width, down the centre of which there frequently runs a narrow band of clear-green, finely fibrous, rather highly birefringent chrysotile. Iron-ores include large grains of chromite and small granules and cubes of secondary magnetite and pyrites. A feature of note in connection with the primary iron-ore is the

– 245 –

presence of an enclosing border of tiny platelets of chlorites, sometimes radially but more often haphazardly arranged. Two varieties occur (positive pennine and delessite), both of which may directly surround the chromite, but occasionally the delessite type may rim the other variety.

No. 2532 is a clear-green strongly slickensided rock with an enamel-like surface. In thin section the structure is cataclastic. Blades of antigorite (75%) with typical thorn-structure have in some cases clearly formed at the expense of fibrous green serpentine, wisps and fragments of the latter being scattered among the antigorite blades. Occasional large undefined areas occur where very fine antigorite passes into a form with parallel structure. Again, other areas are composed of pale-green serpophite in which the fibres (positively elongated) are sometimes arranged in a poorly knitted fashion or occasionally radially. Carbonate occurs in veinlets and in irregularly-shaped multiply-twinned grains; frequently the twin lamellae are bent and broken while gliding has occurred along a series of parallel planes. In shear-zones a rather ferruginous serpentine (probably a chrysotile) of later formation than the antigorite has crystallised. Accessory constituents are rare and include grains of iron-ore and platelets of a positive chlorite. In a duplicate section, No. 2532A, these accessories are more plentiful, but carbonates are wanting; the rock, however, is more evenly textured.

Talcose Serpentinites.

The talcose serpentinites are characterised by an abundance of both talc and serpentine minerals, the former developing from the latter. They may therefore be considered as representing an intermediate stage in the transition from serpentinites to talc-rocks.

No. 2575 is a massive, dark-green, non-fissile rock with an enamel-like lustre. Ragged relicts of mesh-serpentine make up 30–40% of the rock, with occasional plates of green bastite. The serpentine is everywhere being replaced by talc, which is developed irregularly within plates of serpentine, or is aggregated into dense masses enclosing relicts of that mineral. Primary iron-ore occurs in irregular or rounded grains and a little secondary pyrites is also scattered throughout the rock. Two other sections (Nos. 1439, 1444) reproduce similar features.

In the hand-specimen, No. 2578 is almost identical with No. 2575, but microscopically shows pronounced mechanical effects of shearing. Twisted fibres of pale-green serpentine make up most of the rock. Recrystallisation, however, has produced locally a very pale to colourless serpentine, which in patches has the micaceous habit of typical antigorite, while in adjacent areas the texture is fibrous rather than micaceous. Both the green and colourless serpentine minerals may be flecked with talc, and numerous veinlets of this mineral ramify throughout the rock. Remnants of sheared veinlets of chrysotile were noted, while scattered throughout the rock and often located along veinlets of talc there are flakes of a bowlingitic serpentine. Occasional pools of almost isotropic pale-green chlorite

– 246 –

occur, often with an opaque spinellid nucleus and showing the same zonary arrangements as in No. 2519 (see earlier under antigorite-serpentines).

No. 2555 is a dark-green, slightly fissile, lustre-mottled, veined rock. In thin section the rock has an inherited poecilitic structure, in which sharply bounded aggregates of minerals are set in a finely fibrous base of pale-green, poorly pleochroic, serpophitic serpentine with a very low refringence. Closely associated with the serpentine is a negatively elongated, strongly pleochroic, yellowish-green chlorite. The poecilitically enclosed patches referred to above are of three types:—


Masses rarely exceeding 1.0mm. in diameter composed entirely of irregular grains and idioblastic rhombs of (?) magnesite.


Rounded though usually sharply defined aggregates composed of tiny flakes of talc and iron-ore wrapped round with dark-green, moderately birefringent serpentine. Occasionally tremolite or (?) magnesite are present, while the latter mineral is sometimes almost the sole constituent. Within some of these spots the serpentine may form sinuous lines reproducing a mesh-effect.


Larger and usually poorly defined areas up to 5.0mm. in diameter consisting of relicts of coarse tremolite often with magnesite embedded in a matrix of flaky talc which clearly has formed at the expense of the amphibole.

In addition to these areas, occasional prismatic crystals of tremolite always altering to talc, lie at random within the serpentine matrix, while groups of needles and hair-like forms of iron-ore were also noted. Iron-ores include magnetite and pyrites, while there are also rare prismatic crystals of clinozoisite.

This rock bears a rather close resemblance to some of the lustre-mottled carbonated serpentines of the Kalgoorlie goldfields, described by Thomson (1913, p. 635–637).

No. 2581 contains the following minerals: serpentine 40%, chlorites 30%, talc 15%, carbonates 10%, iron-ores 5%. Pale-green serpophite makes up structureless areas often intimately associated with chlorites, of which two varieties may be recognised. In these two types, however, there is considerable variation in optical properties. Firstly there is an optically negative, nearly uniaxial pennine, ranging from a deep-green (Z) and almost isotropic type to a pale variety with deep-blue anomolous interference tints. The second is pale-green to colourless; it is nearly uniaxial and optically positive with deep-brown interference tints. These two chlorites are always closely interleaved with one another. Talc is plentiful as patches concentrated in the serpophite, or lying as scattered flakes within the chlorites and serpentine from which it is developing. Iron-ores include ragged grains of magnetite and pyrites, while veinlets of talc and twinned carbonate cut the section.

– 247 –

The talc-schists, which constitute a minor group associated with the serpentinites, are mineralogically composed of fine flaky talc together with accessories.

The talc-schists may be subdivided into those rocks (a) without accessories (Nos. 2522, 2525, 2550, 2564) and (b) with accessories (Nos. 2545, 2553, 2569).

As typical of group (a) No. 2550 may be described. Macroscopically it is a very pale-green, strongly slickensided, schistose rock containing embedded within it unaltered relicts of dark-green serpentine. Microscopically the entire section is composed of flakes of talc averaging 0.06mm. in length which in a section cut parallel to the schistosity are completely unoriented. In the rocks placed in group (b) talc occurs as above, but in addition there are present various accessory minerals. Sheared-out, fractured and often lensoidal groups of chromite grains are present in Nos. 2545, 2553 and 2569, while pyrite is rare (No. 2569). Twisted and sheared relicts of bowlingitic serpentine have been observed (Nos. 2545, 2553), with rarely a little (?) antigorite (No. 2545); sphene is uncommon (Nos. 2553, 2569). Garnet occurs in No. 2545, where it forms tiny yellowish-green grains, or rarely dodecahedra; the garnet may form a granular rim to chromite and is probably a member of the grossularite-uvarovite series. Tremolite was recognised with certainty in No. 2553 only, while chlorites were noted in two rocks (Nos. 2545, 2569).

Miscellaneous Rocks associated with the Ultrabasics.

Under this heading are included those rocks more or less intimately associated with the ultrabasics, and which cannot be placed with any of the preceding groups.


This rock (No. 2588) was found in association with the Spring-burn intrusive mass, but its relation thereto is not clear. In the hand-specimen it is a greenish-grey, slightly fissile rock with abundant lustrous plates of talc. The rock contains the following component minerals: chlorite 60%, epidote mineral 15%, talc 15%, sphene 5%, calcite 5%. The chlorite is an unusual type occurring in small rather stout colourless plates averaging 0.08–0.1mm. in diameter. It is optically positive and uniaxial, with a birefringence of about 0.006 and may be classed as pennine. The epidote mineral occurs in colourless to very pale-yellow xenoblastic grains. The iron content, and hence the birefringence, varies greatly even within the limits of a single crystal-grain, the maximum observed value for γ − α being approximately 0.030–0.032, corresponding to about 10% of Fe2O3. Other grains, however, and even borders of highly birefringent grains, have a deep-blue anomolous interference tint, but the wide optic axial angle precludes identification as zoisite; they are therefore termed clinozoisite. The habit of the talc is unusual; it forms large sieved porphyroblastic plates averaging 1.3mm. in diameter which may be repeatedly twinned parallel to 001. The optic axial angle is unusually high for this mineral, 2E

– 248 –

measured by the Becke method being approximately 40°. Sphene forms clusters of semi-opaque grains, while magnesite (?) in ragged grains and rhombs is also present. Iron-ores are rare.

Serpentinite-calcite Contact.

A vein of calcite 10mm. wide and slickensided on one side cuts a mass of clear-green serpentinite. In a thin section (No. 2539), cut across the serpentinite-calcite contact, the vein in seen to be composed of irregularly shaped, multiply-twinned grains of calcite averaging 3–4mm. in diameter, which decrease in size to about 0.5mm. where they form ramifying veinlets within the optically negative serpentine. A little positive pennine was noted. A duplicate section (No. 2539A) cut from the serpentinite adjacent to the vein shows that rock to be similar to No. 2546, but with in addition numerous shear-zones and veinlets filled with calcite which is twinned and the lamellae often bent. The serpentine is largely recrystallised to the variety in No. 2539. A little bowlingitic serpentine was noted.

Calcite Vein in Serpentinite.

No. 2583, obtained from a calcite-vein in the Springburn serpentinite mass, is described because it encloses coarse plates of a black opaque mineral at present referred to as ilmenite, which shows up prominently in the hand-specimen. The hardness of this latter mineral is about 5, though it is difficult to estimate the value accurately, as it flakes rather readily under pressure of a knife-blade; the streak is black, and there is a good cleavage parallel to the principal faces of the tabular crystal. On one side the vein is strongly slickensided. In thin section irregular grains of twinned calcite averaging 4.0mm. in diameter form 90% of the vein. Twin lamellae are bent and the grains somewhat granulated. Negative pennine forms interstitial patches between, or pools within, the grains of calcite. The only other constituent is the black ore in long bars which average 2.5mm. in length and 0.75mm. in width. Occasionally they show well-developed terminal planes and frequently have a partial border of granular sphene.

Summary of the alteration Phenomena.

The changes which have been observed in the basic and ultra-basic rocks may be considered separately under two headings:—





It must be pointed out, however, that it is impossible to separate these processes completely, as, in most cases, cataclasis has been accompanied by mineralogical change.

(a) Mechanical Effects. The remarkable and rapid variation of the rock-types present within the Caples Valley “Knob” is believed by the writer to have been brought about by intense but

Picture icon

Basic Intrusions in N.W. Otago

Picture icon

Fig. 1.—A large fissured relict of green serpentine being replaced by antigorite, in antigorite-serpentinite, No 2519. Magnification, × 22.
Fig. 2.—A tremolite-talc-carbonate-schist (No. 2589) showing oriented prismatic crystals of tremolite with interstitial talc (T). A few idioblastic rhombs of carbonate are also present Magnification, × 46
Fig. 3.—A carbonated talcose-serpentinite (No. 2585), showing pseudomorphs of talc and serpentine after olivine, partially replaced by carbonate, and set in a base of finely fibrous serpentine. A little magnetite and tremolite are also seen. Magnification, × 46.
Fig. 4.—Mesh-serpentinite (No. 2546) showing clear, colourless plates of bastite and black grains of chromite set in a base of dark-green mesh-serpentine. Magnification, × 42.

Picture icon

Fig. 5.—A calcite [ unclear: ] (No. 2583) showing coarse bars of (?) [ unclear: ] and grains of twinned calcite. Magnification, × 18.
Fig. 6.—A cluster of grains of chromite set in mesh-serpentine (No. 2546). Magnification, × 28.
Fig. 7.—A view of the “Knob,” Caples Valley, Western Otago.

– 249 –

local shearing within this small basic intrusion, with the result that rectangular blocks of varying textural and mineralogical composition have been brought into juxtaposition by mechanical displacement. Microscopically observable effects of shearing are universal though variable in the gabbroidal rocks which make up this mass. In the serpentinised olivine-gabbros and augite-picrites there is little sign of pure cataclasis; but in the pyroxenites and saussurite-gabbros, on the other hand, the plates of diallage are bent, fractured, or in the more altered specimens completely granulated. Operation of shearing-stress on the plagioclase of the saussurite-gabbros has caused the development of undulose extinction and bending of twin-lamellae; with increasing stress fracture and then granulation occur, these changes being accompanied by considerable mineralogical rearrangement. Later a renewal of stress has sometimes brought about the appearance of patchy, secondary lamellar twinning in the now chemically reconstituted plagioclase (albite). Development of poly-synthetic twinning in calcite is universal wherever this mineral is present, and may be accompanied by gliding along the twinning-plane or followed by distortion of the lamellae themselves. As shearing increases in intensity chemical change plays an increasingly important role and obscures the purely mechanical effects, so that the most sheared rocks of the “Knob,” though of very fine grain, appear to be almost completely crystalloblastic. No macroscopic schistosity has been developed in any of the Caples Valley “Knob” rocks, though local linear microstructure is often apparent.

The effect of directed pressure is especially evident in certain groups of the tremolite-rocks. In most cases stress has been great enough to produce a marked schistosity observable both in the hand-specimen and in thin section. Again, surfaces of strong slickensiding, especially noteworthy in the Routeburn rocks, have often been produced parallel to the schistosity and indicate considerable mechanical displacement within the mass. Other evidence of shearing is seen in the development of microscopic strain-slip bands in many sections, and in the manner in which tremolite fibres enwrap eyes of talc in some specimens. Extreme shearing and mechanical degradation of an original tremolite-chlorite-schist are believed by the writer to have been the mechanism by which the very fine-grained tremolite-chlorite-phyllonites have originated.

Among the serpentinites purely mechanical effects are not so readily observed owing to the fact that serpentine will itself readily recrystallise on being subjected to directed pressure. However, in a few examples, torn and twisted relicts of dark-green mesh-serpentine set in recrystallised matrix were observed, while a series of faint strain-slip bands were noted crossing fine parallel serpentine fibres.

In the talc-schists the only mechanical effects recorded are the granulation and streaking out of grains of chromite into lenses and also the streaking out of “ghosts” of relict hornblende.

– 250 –

(b) Chemical Effects. In the basic and ultrabasic rocks the observed mineralogical changes involved in the alteration may be summarised as follows:—


In the basic rocks of the Caples-Greenstone-Routeburn area.

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

Schillered diallage clear diallage.
Diallage green hornblende tremolite.
Diallage tremolite.
Diallage chlorite.
Diallage tremolite + chlorite.
Diallage chlorite + sphene.
Diallage chlorite + epidote (or carbonate) + (?) prehnite.
Diallage serpentine + epidote.
Diallage talc + carbonate.
Schillered brown hornblende partially bleached amphibole.
Brown hornblende green hornblende tremolite.
Brown hornblende chlorite + calcite + sphene.
Green hornblende serpentine.
Green hornblende tremolite.
Calcic plagioclase saussurite.
Calcic plagioclase albite + epidote + minor sericite (or) prehnite.
Chlorite stilpnomelane.
Chlorite talc.
Iron-ores leucoxene.


In the ultrabasic rocks of the Caples-Greenstone-Routeburn area.

Relict green hornblende tremolite or chlorite (or both).
Relict brown hornblende tremolite or chlorite (or both).
Riebeckite tremolite.
Ferruginous serpentine pale green finely fibrous negative serpentine.
Ferruginous serpentine positive pennine.
Serpentine antigorite-talc.
Serpentine talc.
Chromite garnet.

In the foregoing mineralogical alterations it is clear that there has been considerable migration of molecules in order to effect the changes observed. For example, the change chromite → garnet involves the transference of much lime, alumina and silica towards the spinellid mineral; similarly with many other reactions.


In the ultrabasic rocks of Springburn Creek.

In the case of the Springburn serpentinite mass the mineralogical changes are rather complex and somewhat different from those observed in the rocks of the Caples, Greenstone and Routeburn Valleys. The rock forming the Springburn intrusion is believed to have been derived from a harzburgite with poecilitic structure, and composed of olivine, enstatite and minor hornblende. No relicts of these minerals, however, remain. The initial change has been the serpentinisation of the pyroxene (with concomitant development

– 251 –
  • of minor chlorite) and olivine, accompanied by alteration of hornblende to colourless tremolite. Subsequent replacement of these secondary minerals by talc is clearly seen, but such alteration has been selective in that the serpentine derived from olivine has now been replaced almost completely, while the dark-green serpentine pseudomorphous after pyroxene shows only the initial stage of conversion to talc. Simultaneously with this reaction steatitization of tremolite has also occurred and numerous relicts of that mineral remain embedded in a matrix of talc flakes and show all stages of replacement. The last change in the sequence is general replacement of talc by a carbonate, probably magnesite, accompanied by the introduction of a little pyrites.

The mineralogical changes actually observed in these rocks are thus:—

Serpentine (after pyroxene) talc.
Serpentine (after olivine) talc magnesite.
Chlorite talc.
Tremolite talc magnesite.

Causes of Alteration.

(a) Basic Rocks. It is at first sight rather surprising to find that the basic rocks of the Caples Valley “Knob,” which have suffered so greatly from shearing and have in some instances become recrystallised, are associated in the field with a series of sedimentary rocks that show only faint effects of metamorphism of very low grade. This contrast in the degrees of metamorphism displayed respectively by the basic intrusions and the rocks they invade may be due to four causes:—


The well-known susceptibility of basic and ultrabasic rocks to dynamic metamorphism (e.g., see S. Foslie, 1931, p. 227).


As is so frequently the case in other parts of the world and more especially in New Zealand, the basic and ultrabasic intrusives are located along zones of thrusting where shearing is particularly intense (cf. W. N. Benson, 1926; F. J. Turner, 1933, p. 276).


In the intrusive masses under consideration many of the metamorphic changes may well have been effected when the temperature was still somewhat higher than normal, i.e., during the cooling phase following intrusion.


Within the recently injected masses abundant residual magmatic water was doubtless available to assist the chemical reconstitution initiated by shearing (cf. A. Harker, 1933; pp. 14–18).

(b) Ultrabasic Rocks. Owing to the limited extent of the outcrops little field evidence of the causes of alteration is available. It is certain, however, that the water which effected serpentinisation of the peridotites must have been contained within the ultrabasic intrusions themselves and could not have been derived from any younger acid intrusives in the manner postulated by various writers

– 252 –

for serpentines in other parts of the world, for the granites in the South Island of New Zealand antedate the ultrabasic rocks (see F. J. Turner, 1933, p. 277).

The origin of the talc is a more complex problem, but six points which bear on the subject may be given here:—


The talc-rocks are always strongly schistose.


They occur in localised bands along which there has been intense movement.


In other zones of shearing and slickensiding within the serpentinites there is no talc.


The primary chromite present in the talc-schists is always crushed and streaked out.


Serpentinisation is always earlier than the formation of talc.


The typical “stress mineral”, antigorite, is present in some of the serpentinites.

Gillson (1927, p. 287), Hess (1933a, pp. 406–407; 1933b, p. 646) and Rietz (1935, p. 257) all believe that talc-rocks have originated by the hydrothermal alteration of serpentine and other minerals by solutions from younger acid intrusions, and that in most cases stress was of little or no direct importance. The solutions that assisted in the formation of talc in the present instances, like the serpentinising waters mentioned above, must also have been derived from the basic magma itself, but whether such change was purely hydrothermal or was dependent also upon the operation of shearing-stress is not clear. In view of items 1, 2 and 4 above, the writer favours this second alternative, which agrees with Read's (1934, p. 666) suggested origin of talc in tear-bands and at the sole of an overthrust serpentine block in Unst, Shetland Islands. On the other hand it is equally clear that shearing of serpentine may also occur without obvious chemical reconstitution or may result merely in the formation of antigorite. Variable conditions which might account for these differences in the behaviour of the ultrabasics when subjected to shearing are (1) temperature at which shearing occurs and (2) availability of the additional silica required to allow the transition serpentine → talc.

The occurrence in one of the Springburn rocks of coarsely porphyroblastic talc, which has clearly formed under static conditions, indicates in this one case a purely hydrothermal origin under the actions of solutions derived from the serpentine itself (autometamorphism). Magmatic solutions have been particularly active in this locality, for Park (1908, p. 80), quoting analyses by Maclaurin, shows that the schists immediately adjacent to the serpentine have become very much enriched in soda.


The writer wishes to express his gratitude to Dr. W. N. Benson and Dr. F. J. Turner, of the Geology Department, University of Otago, for their assistance and helpful criticism so freely given during the writing of this paper.

– 253 –

List of Specimens Examined.

List of Specimens Examined.
Specimen No. (Geology Dept., Otago University) Locality.
1439 Slip, 300 yards above Bridge, Routeburn.
1444 "
2519 Greenstone River.
2520 "
2521 "
2522 "
2523 "
2524 “Knob,” Caples Valley.
2525 Talc Creek, Caples Valley.
2526 “Knob,” Caples Valley.
2527 "
2528 "
2529 Intrusion 3, four chains south of “Knob,” Caples Valley.
2530 “Knob,” Caples Valley.
2531 "
2532 First waterfall, Talc Creek, Caples Valley.
2533 “Knob,” Caples Valley.
2534 Intrusion 2, four chains south of “Knob,” Caples Valley.
2535 “Knob,” Caples Valley.
2536 "
2537 "
2538 "
2539 Two chains from highest outcrop, Talc Creek, Caples Valley.
2540 “Knob,” Caples Valley.
2541 Lower intrusion, Beech Creek, Caples Valley.
2542 Intrusion 3, Caples Valley.
2543 Lower intrusion, Beech Creek, Caples Valley.
2544 “Knob,” Caples Valley.
2545 Highest intrusion, Beech Creek, Caples Valley.
2546 Intrusion 2, Caples Valley.
2547 Lower part of intrusion, Talc Creek, Caples Valley.
2548 Upper part of intrusion, Talc Creek, Caples Valley.
2549 “Knob,” Caples Valley.
2550 Upper part of intrusion, Talc Creek, Caples Valley.
2551 Talc Creek, Caples Valley.
2552 Intrusion 3, Caples Valley.
2553 "
2554 "
2555 Intrusion 2, Caples Valley.
2558 "
2560 “Knob,” Caples Valley”.
2561 Intrusion 2, Caples Valley.
2562 "
2563 Intrusion 3, Caples Valley.
2564 Upper part of intrusion, Talc Creek, Caples Valley.
2565 Slip, 300 yards above Bridge, Routeburn.
2566 "
2567 "
2568 "
2569 "
2570 "
2571 "
2575 "
2577 "
2578 "
2579 "
2580 Creek, Routeburn, two miles above Dart River.
2581 "
2583 Outcrop near headwaters of Springburn Creek.
2585 "
2586 "
2587 "
2588 "
2589 "
– 254 –

List of Literature Cited.

Bartrum, J. A., and Turner, F. J., 1929. Pillow-lavas, Peridotites, and Associated Rocks of Northernmost New Zealand, Trans. N.Z. Inst., vol. 59, pp. 98–138.

Benson, W. N., 1914. The Geology and Petrology of the Great Serpentine Belt of New South Wales; pt. 3, Petrology, P.L.S., N.S.W., vol. 38, p. 4.

——1926. The Tectonic Conditions Accompanying the Intrusion of Basic and Ultrabasic Igneous Rocks, Mem. Nat. Acad. Sci., vol. 19, no. 1, Washington.

Bonney, T. G., and Raisin, C., 1905. The Microscopic Structure of Minerals forming Serpentine, Q.J.G.S., vol. 61, pp. 690–714.

Cox, S. H., 1879. The Wakatipu and Greenstone District, Repts. of Geol. Explor. N.Z., during 1878–1879, pp. 53–55.

Dunn, J. A., 1929. The Geology of North Singhbhum including parts of Ranchi and Manbhum Districts, Mem. Geol. Surv. Ind., vol. 54.

Eskola, P., 1932. On the Principles of Metamorphic Differentiation, Compt. Rend. Geol. Soc. Finlande, No. 5.

Flett, J. S., and Hill, R. N., 1912. On the Geology of the Lizard and Meneage, Mem. Geol. Surv., England and Wales.

Foslie, S., 1931. On Antigorite-Serpentines from Ofoten with Fibrous and Columnar Vein Minerals, Norsk. Geol. Tids., B. XII.

Gillson, J. L., 1927. Origin of the Vermont Talc Deposits, Econ. Geol., vol. 22, pp. 246–287.

Harker, A., 1932. Metamorphism, London, Methuen & Co.

Hatch, F. H., and Wells, A. K., 1926. The Petrology of the Igneous Rocks, 8th ed., London, G. Allen & Unwin, Ltd.

Henderson, J., 1923. Talcose Schist in Springburn Valley, Kawarau Survey District, Otago, 17th An. Repts. Geol. Surv. of N.Z. (n.s.), p. 12.

Hess, H. H., 1933a. Hydrothermal Metamorphism of an Ultrabasic Intrusive at Schuyler, Virginia, Amer. Jour. Sci., vol. 26, pp. 377–408.

——1933b. The Problem of Serpentinisation and the Origin of Certain Chrysotile Asbestos, Talc, and Soapstone Deposits. Econ. Geol., vol. 28, pp. 634–657.

Knopf, E. B., 1931. Retrogressive Metamorphism and Phyllonitisation, Pt. 1, Amer. Jour. Sci., vol. 21, pp. 1–27.

Lodochnikow, W., 1933. Serpentines and Serpentinites and the Petrological Problems connected with them, Problems of Soviet Geology, No. 5.

Marshall, P., 1906. Geological Notes on the Country North-West of Lake Wakatipu, Trans. N.Z. Inst., vol. 38, pp. 560–567.

McKay, A., 1881. District West and North of Lake Wakatipu, Repts. Geol. Explor. of N.Z., during 1879–1880, pp. 118–147.

Park, J., 1887. On the District between the Dart and Big Bay, Repts. Geol. Explor of N.Z., during 1886–1887, pp. 121–137.

——1908. The Geology of the Cromwell Subdivision, N.Z. Geol. Surv. Bull., No. 5.

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

——1933. On Quartz-Kyanite-Rocks in Unst, Shetland Islands, and their bearing on Metamorphic Differentiation, Min. Mag., vol. 23, No. 140, pp. 317–328.

——1934. The Metamorphic Geology of Unst in the Shetland Islands, Q.J.G.S., vol. 90, No. 360, pp. 637–688.

Rietz, T. Du., 1935. Peridotites, Serpentines, and Soapstones of Northern Sweden, Geol. Foren. I. Stockholm Forhand, No. 78, pp. 133–260.

Thomson, J. A., 1914. On the Petrology of the Kalgoorlie Goldfield (Western Australia), Q.J.G.S., vol. 69, No. 276, pp. 621–677.

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

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

——1935. Geological Investigation of the Nephrites, Serpentines, and Related “Greenstones” used by the Maoris of Otago and South Canterbury, Trans. Roy. Soc. N.Z., vol. 65, pt. 2, pp. 187–210.

Winchell, A. N., 1933. Elements of Optical Mineralogy, pt. 2, New York, John Wiley & Sons.