Go to National Library of New Zealand Te Puna Mātauranga o Aotearoa
Volume 71, 1942
This text is also available in PDF
(3 MB) Opens in new window
– 307 –

Structural Petrology of Quartzose Veins in the Schists of Eastern Otago—Part I.

[Read before the Otago Branch, July 8, 1941; received by the Editor, July 16, 1941; issued separately, March, 1942.]

Material Studied.

Work on the structural petrology of the schists of East Otago has till now been confined mainly to the coarsely crystalline schists of Subzone Chl. 4 of the Chlorite Zone (e.g., Turner, 1940; Turner and Hutton, 1941). The less completely reconstituted finer-grained schists and semischists of Subzone Chl. 3 and Chl. 2 are less suited to petrofabric analysis, for in these rocks the newly crystallised grains of quartz chlorite and muscovite are in the main too small to allow accurate measurement of their orientation. However, many of these low-grade rocks are traversed by quartz-rich laminae and vein-lets of sufficiently coarse grain to be investigated readily with a universal stage. Representative specimens from the south-eastern border of the Otago metamorphic belt, between the Milton-Lawrence railway and the Maungatua Range, have been selected for fabric analysis, and the first series of results obtained is presented in this paper.

The rocks in question (Nos. 6231 and 6231a in the collections of the geology department, University of Otago) are schistose grey-wackes or “semischists” of Chl. 2 collected from points twenty yards apart in a steep cutting on the north-east side of the Milton-Lawrence road, south-east of Mt. Stuart railway station. Deformed clastic grains of plagioclase in all stages of replacement by albite, sericite, clinozoisite and pumpellyite, together with relict grains of quartz are set in a fine-grained matrix of newly crystallised albite, epidote, stilpnomelane, sericite, chlorite, and quartz. Finely granular pyrrhotite is disseminated fairly plentifully through the whole rock.

Megascopic Scructure.

The structural features observable in the field and in hand-specimens are as follows:—

The principal schistosity S2 (selected as the ab fabric plane) is well defined in No. 6231a, less so in No. 6231; it strikes N. 50° E. and dips 25° to S.E. A second set of s-surfaces S1, visible in micro-sections and on suitably oriented polished surfaces of the hand-specimen, intersects S1 at 30°–45° in lines parallel to the lineation, and dips E.S.E. at angles of approximately 50°–60° (Fig. 1C). The linear structure in S2 (= b axis of the fabric) is faintly defined in the hand-specimen No. 6231a, where it has a trend of N. 10° E., and is almost invisible in No. 6231. In both rocks, however, it is defined microscopically by streaking and elongation of partially reconstituted mineral grains as seen in sections parallel to S2; some of these also show dark crenulated discontinuous streaks transverse to b, possibly representing traces of original bedding.

– 308 –

Like all rocks observed by the writer in this vicinity, both specimens are intricately veined with narrow quartz-rich veinlets which form the subject of the present study. For convenience of description four series of veinlets, designated V1 to V4 in chronological order, are distinguished, though in addition there are a few inconspicuous streaks of intermediate orientations.

  • V1 Series: black (h O I) veins 0·2 to 4 mm. wide, dipping eastward subparallel to S1 and cutting the main schistosity S2 at angles of 30°–45° (rarely 70°).

  • V2 Series: veinlets often 5 mm. in width, approximately parallel to S2 and cutting sharply across those of the V1 Series; narrower veinlets are black, but the most prominent have a light central zone margined on each side by black.

  • V3 Series: a single clear-cut white vein (12 in 6231; Fig. 1B) differing only slightly in orientation from the V1 Series, but mineralogically different from and of later origin than veins of the other two groups just described.

  • V4 Series: prominent white veins, often 1 mm. to 5 mm. wide, approximating to (O k l) of the fabric, and usually sub-parallel to the ac fabric plane or dipping northward at 70°–90° to S2. Veinlets perpendicular to S2 but oblique to b are also included in this group, although they intersect the schistosity in a line trending N. 75°–80° E. instead of N. 100° E. as in typical V3 veins, from which, however, they branch directly in some instances. (See Veins 1 and 2, Fig. 1A.)

The mutual relationships of typical veins of all four series are indicated in Fig. 1, which includes sketches of blocks sawn from the two hand-specimens. Individual veins are numbered (not in chronological sequence), for reference in fabric analysis. As in previous work on Otago schists, the positive end of the c fabric axis emerges from the upper surface of the block, and the positive end of b is directed approximately southward.

General Characteristics of Growth Fabrics in Veins.

The veined “semischists” of Mt. Stuart were selected as the starting point for investigation for two reasons. Firstly, if the general development of fabric in the laminated schists of eastern and central Otago has been influenced partly by post-tectonic growth of crystals (growth fabric) as well as by paratectonic crystallisation (tectonite fabric), then the clear-cut veins just described should show the type and degree of preferred orientation to be expected in such growth fabrics, and will furnish a standard with which the fabrics of more intensely metamorphosed schists may be compared. In the second place fabric analysis of the vein materials should show to what extent deformation has been active subsequently to growth of veins, and in this way should afford indications as to whether the process of vein formation was connected with or subsequent to deformational metamorphism (cf. Turner, 1941, p. 11).

It is appropriate, therefore, first to review the general characteristics of growth fabrics developed in veins, as described by Sander (1930, pp. 156–162) and subsequently discussed by other writers (e.g., Knopf and Ingerson, 1938, pp. 67, 126–128).

– 309 –
Picture icon

Fig. 1—A, B: Sketches of section-blocks of specimens 6231a (A) and 6231 (B) showing the principal veins in relation to schistosity S2 (= ab) and lineation (lin. = b). Veins of V1 Series full black (dip steep), V2 stippled (minor veinlets black), V3 blank, V4 shaded. Thickness of all veins exaggerated. C: Stereographic projection (of the b hemisphere) upon the ac plane, showing the relation of S1, S2 (= ab) and the geographic horizontal plane (broken line).


Growth of quartz crystals in a vein traversing a rock that is not undergoing deformation is governed by a blastetrix, i.e., by a surface perpendicularly to which growth of crystals encounters minimum resistance (Sander, 1930, p. 159) and which typically coincides with the vein wall. If the blastetrix directly influences the space-lattice orientation of the growing crystals it is said to be anisotropic; if there is no such influence it is isotropic.


The most obvious effect of an isotropic blastetrix upon the growth fabrics is the development of a dimensional orientation of the quartz grains with their long axes perpendicular to the blastetrix. A high degree of orientation of this type implies a marked difference between the maximum velocity of growth (at right angles to the

– 310 –

blastetrix) and the minimal velocities for less favoured directions in the fabric. H. W. Fairbairn (1937, p. 118) implies recognition of this principle when he states that “this orientation is usually sharp in filled fractures, but may be lacking altogether in replacement fabrics.”


In the case of isotropic minerals, or minerals which are capable of growing with nearly equal velocity in all crystallographic directions, the dimensional orientation referred to under (2) is not accompanied by orientation according to space-lattice structure. However, quartz is strongly anisotropic as regards velocity of crystal growth, and for such minerals preferred crystallographic orientation tends to accompany orientation according to dimension. The dominating crystallographic orientation rule for quartz in the examples cited by Sander (1930, p. 342, D225, 226) is for the optic axis to grow at right angles to the blastetrix (or foundation wall of the vein), but a weaker tendency for the optic axis to lie parallel to the vein wall has also been noted. Since the crystallographic orientation is achieved during growth of the vein, by a process of growth-selection (Sander, 1930, p. 160) involving suppression of unfavourably oriented crystals and complementary expansion of those oriented according to the controlling rule, the degree of preferred orientation increases with distance from the foundation wall where crystallisation commenced (Knopf and Ingerson, 1938, pp. 126, 127).


Orientation according to crystal lattice, accompanied again by dimensional elongation of grains perpendicularly to the vein walls may also arise from initial preferred orienting of the “seed crystals” under the influence of such physical conditions as the surface tension between growing crystals and the enveloping parent solution (Sander, 1930, p. 161).


Lattice orientation of the vein mineral may sometimes be traced to the influence of an anistropic blastetrix, i.e., to the direct influence of pre-existing oriented grains in the material of the vein walls. This is most commonly encountered in, though not confined to, veins originating by a process of replacement (Fairbairn, 1937, pp. 119–121).


Fabric diagrams constructed for any mineral whose growth has been controlled solely by a blastetrix should be symmetrical about the axis normal to the blastetrix. They should show either (a) a point maximum coinciding with the pole of the blastetrix, or (b) a girdle with the pole of the blastetrix as centre, or (c) no significant preferred lattice orientation of mineral grains.

Petrography and Fabric of the V4 Series.

The principal constituent of all veins in the V4 Series is quartz, always accompanied by some albite, which in Vein 9 is equal in abundance to quartz. The quartz is coarser than in veins of the three earlier series. In the wider veins the individual grains may reach dimensions as great as 2–3 mm. × 0·3–0·5 mm. and often show a tendency toward elongation perpendicular to the vein-walls; the larger grains almost invariably show undulose extinction and locally pass into aggregates of granules apparently the product of ruptural deformation of larger individuals. Albite occurs in three distinct habits, all of which may be represented in any one vein, viz.,

– 311 –

(a) large grains sparsely interspersed among the coarser quartzes; (b) small sometimes idioblastic grains growing within and enclosed by the larger grains of quartz, (c) minute granules associated with chlorite, mica and opaque carbon in discontinuous undulating strings and streaks parallel to the vein-walls. A few rounded or sharply crystallised grains of apatite are present in every section, while in several instances (especially Vein 3) scattered tufts of actinolite needles trend roughly at right angles to the walls. Yellow sulphides mainly pyrrhotite , which are plentiful in the surrounding rock, are also represented by rare larger granules (ranging up to 0·4 mm.) within the veins themselves (e.g., ab section of Vein 9). For the most part, however, the pyrrhotite is located along the borders of the veins (e.g., 2, 3 and 4), while microscopic fractures which cut all the V4 veins parallel to (h O l) of the fabric, are filled with pyrrhotite where they continue into the adjacent rock-matrix and more rarely within the vein as well (Vein 2). Introduction of metallic sulphides must, therefore, have been subsequent to the period of vein growth, though the vein walls appear to have afforded surfaces of percolation for the mineralising solutions.

Picture icon

Fig. 2—Sketches showing outlines of quartz grains in typical sections of veins of V4 Series: A, Vein 4; B, Vein 5; C, Vein 2; D, Vein 9 at junction with Vein 13 (V1 Series).

[Footnote] † Qualitative chemical tests show the presence of nickel in samples of magnetically separated pyrrhotite from 6231 and 6231a.

– 312 –

Specially distinctive of veins of the V4 Series is the absence of stilpnomelane minerals and iron-stained carbonate , both of which are usually abundant in veins belonging to the other series. The only exceptions are seen at intersections with earlier veins (e.g., Vein 9 at its intersection with 6 [V2] and 12 [V3]), where streaks of carbonate in the earlier veins continue without interruption across the later. Carbonate of this type is thought, therefore, to be relict material left unreplaced by quartz during growth of the veins of the V4 Series.

The clearest example of dimensional orientation is that furnished by Vein 4, which is divided longitudinally into three distinct zones (x, y, z in Fig. 2A), viz., an asymmetrically situated narrow inner zone of small nearly equidimensional granules, flanked by marginal zones of coarser grains with a strong tendency for elongation at right angles to the vein wall. Two almost continuous and especially distinct lines of inclusions, along which the hand-specimen fractures readily, mark the boundaries between the three zones. The degree of preferred orientation of the quartz space-lattice is illustrated in Figs. 3–7.

Fig. 3 is based on measurement of optic axes in 132 large grains situated in the outer and middle portions of the two marginal zones (x and z), and shows a strong tendency for the quartz crystals to lie with the optic axis nearly perpendicular or steeply inclined to the walls of the vein (cf. Sander, 1930, D225, D226). Certain minor differences of fabric in zones x and z are brought out when the two corresponding elemental diagrams (Figs. 4 and 5 respectively) are compared. In Fig. 5 the axial maxima are located close to the pole of the vein wall (= b), in accordance with the usual rule for growth orientation in quartz, but Fig. 4 (zone x) shows the maxima lying on a small circle (radius 35°) with V, a point 20° distant from b, as centre. The discrepancy between Figs. 4 and 5 may perhaps be attributed to the small number of grains available for measurement; or on the other hand it is just possible that the fabric of zone x has been influenced by a hitherto unrecorded orientation rule. According to current opinion (cf. Eskola, 1939, p. 278) the “force of crystallisation” of a crystal growing in a solid medium may be correlated with denseness of packing of atoms in the crystal-lattice. It might be expected, therefore, that growth of quartz grains in a replacement vein should proceed most rapidly in the three known directions of closest packing of silicon atoms. These are [00*1] the vertical crystal axis, [01*0] the edge r:m, and [11*1] the edge r:z. The dominating orientation rule for vein-quartz as recorded by Sander—c crystal axis perpendicular to the vein wall—establishes a tendency for crystals of quartz to grow most rapidly parallel to the c

[Footnote] ‡ Probably calcite stained with iron from oxidation of pyrite (cf. Turner and Hutton, 1941).

[Footnote] † In this and all subsequent diagrams the positive end of a fabric axis is marked a, b, or c; the negative end, a, b, or c where necessary

[Footnote] ¶ Zonal symbols and lettering of crystal faces follow the standard schemes employed in the fourth edition of E. S. Dana's Text-book of Mineralogy, pp. 133, 470. Corresponding zonal symbols used by other writers are [0001], [2110] and [2113] respectively.

– 313 –

axis, while the second recorded rule—c crystal axis parallel to vein wall—could be explained (among other possibilities) by a tendency for crystals to grow parallel to the unit prism-rhombohedron edge [01*0]. If the preferred direction of growth were the third direction of close packing (r:z), then the optic axes of the grains concerned would lie on a small circle of radius 41° about the pole of the blastetrix. While not wishing to attach undue importance to a single fabric diagram based on only 52 measurements, the writer would nevertheless tentatively draw attention to the similarity between the orientation pattern of Fig. 4 and that demanded by the third possible rule just discussed.

Origin of Vein 4 and other members of the V4 Series by simple filling of fissures is suggested by the clear-cut plane walls of the veins and their regular orientation (cf. Fig. 1). Certain features of the microstructure on the other hand favour replacement, e.g., the universal presence of linear streaks and threads of wall material passing unbroken through adjacent grains of quartz, and the total lack of correspondence between the larger broken grains of quartz, feldspar, etc., on opposite walls of the veins. Further the central zone y is seen under high magnification to be sharply separated from the adjacent marginal zones x and z, so that individual grains are seldom shared by adjacent zones. This necessitates assumption that growth of Vein 4 involved at least two distinct stages.

The fabric diagram Fig. 7 illustrates the orientation of all available grains of quartz (244) in the central zone y of Vein 4. The majority of the measured grains are roughly equidimensional and are much smaller (0·3 mm. diameter) than those in the marginal zones. A small proportion of grains (23 in all), elongated more or less perpendicularly to the vein wall were measured separately, but failed to show either significant difference in orientation from that of the predominating equidimensional type or any relationship between crystal orientation and the plane of the vein wall. The degree of preferred orientation shown in Fig. 7 is low, but since the same general pattern and almost identical maxima appear in the two elemental diagrams prepared for two sectors of the same thin section, the diagram is considered to be significant. Fig. 7 cannot be interpreted as representing a simple growth fabric, for there is no obvious relation in symmetry between the orientation pattern there depicted and the plane of the vein wall, and the dimensional orientation observed is only rudimentary. On the other hand the maxima and sub-maxima approximate to a poorly developed cleft girdle the axis(b′) of which lies in the ab plane and is inclined at about 25° to b. The most reasonable conclusions to be drawn are that the fabric depicted in Fig. 7 is a poorly developed tectonite fabric which is not shared by the two marginal zones of the same vein, and that development of the central zone y preceded growth of the two later marginal zones x and z. It is difficult to visualise how this last could be accomplished except by replacement proceeding outward from the boundaries between y and x or z.

Further evidence in support of this hypothesis is supplied by a diagram (Fig. 6) prepared from measurements of 86 small grains

– 314 –

within the marginal zone, x, immediately adjacent to the central zone y; groups of such grains are interspersed among the ends of the large strongly elongated grains whose orientation is recorded in Fig. 4 (cf. Fig. 2A). There is no resemblance between Fig. 6 and Figs. 3 and 4, but the diagram apparently represents a fabric of almost unoriented grains possibly showing a slight degree of influence by the fabric of the adjacent zone (Fig. 7). The known general tendency for preferred orientation to increase in strength as the vein grows outward from the foundation wall has already been commented upon (p. 310). If this principle holds good for zone x of Vein 4, then the two marginal zones x and z must have-grown outward, not inward.

Most veins of the V4 Series are less regular in structure and fabric than is Vein 4. The tendency for elongation of quartz grains perpendicular to the vein wall is much weaker, though still distinct (cf. Fig. 2B). Small single granules, and aggregates not infrequently showing the characters of superindividuals; are more plentiful than in the marginal zones of Vein 4, while the larger grains of quartz invariably show most pronounced undulose extinction and sometimes the effects of ruptural granulation.

In Fig. 8 the mean orientation of the optic axis has been plotted for each of 69 large undulose grains (1 mm. to 3 mm. long) in two thin sections of Vein 3 (44 grains) and one of Vein 1 (25), taking the projection of the vein wall as the common datum (broken line in Fig. 8). This brings out a degree of preferred orientation comparable with that illustrated for Vein 4 in Figs. 4 and 5, with the quartz axes tending to lie at high angles to the plane of the vein wall. When measurements of 48 medium-sized grains (0·2 mm. to 0·8 mm.) grains are plotted with those of 25 coarser individuals associated with them in the same thin section of Vein 1, the degree of preferred orientation is much lower (Fig. 9). Fig. 10 is based upon a plot of 170 small grains (0·05 mm. to 0·2 mm.) in the same thin section of Vein 1, and shows no significant preferred orientation of the optic axis in relation to either a blastetrix or the tectonic axis b of the enclosing schist. A similar condition holds too for Fig. 11, representing total quartz (200 grains of all sizes) in one section of Vein 3. Thus in both these veins recognisable preferred orientation governed by growth is confined to the larger grains, which, though numerically few, make up a high proportion (30% to 80%) of the total area of a given thin section. The smaller grains most probably include two types, namely those resulting from rupture of originally coarse grains, and others whose small size is due to slow growth from nuclei unfavourably oriented for growth influenced by the controlling blastetrix.

In Veins 2 and 5 the degree of dimensional orientation displayed by even the coarser grains of quartz is only slight (Fig. 2B, C) and corresponding selective fabric diagrams, not reproduced in this paper, fail to bring out preferred lattice-orientation of any obvious significance.

Vein 9, a sharply defined regular narrow veinlet slightly less than 1 mm. wide, is typical of the smaller veins of the V4 Series in that dimensional orientation of grains of both quartz and equally

– 315 –

plentiful albite is generally lacking. The quartz is for the most part in clear unstrained grains 0·2 mm. to 1 mm. in diameter, while the albite has a similar texture but is usually clouded with swarms of minute brownish dusty inclusions. The optic axes of 40 grains of quartz were measured in a bc section of Vein 9 and replotted on an ab projection together with those of ·50 additional grains measured in a section cut parallel to ab. The resultant fabric diagram is shown in Fig. 12. Though the maxima are strongly defined (8% per 1% area at L) they bear no clear relation to a visible blastetrix such as the plane of the vein walls; nor is it likely, in the absence of any tendency toward uniform elongation of the grains, that growth from a blastetrix could have produced any significant orientation pattern in the fabric. A clear indication that the maxima of Fig. 12 have no significance is supplied by measuring and plotting the axes of an additional 27 grains of quartz in another section of the same vein, cut parallel to ab; of these only one plotted point falls within the areas (dotted) enclosed by the 3·3% contours of Fig. 12, while fifteen lie within the unoccupied area outside the minimal 1·1% contour.

In marked contrast with the general condition described above for Vein 9 is a strong tendency for dimensional orientation of quartz grains within this vein where the latter intersects veins of one or other of the earlier series, e.g., at junctions with veins 6, 7, 12, 13 (cf. Fig. 2D). Here the grains of quartz are prismatic in habit with mean dimensions 0·5–1 mm. × 0·1–0·2 mm. and are invariably clouded with minute dusty inclusions, in contrast with the typically water-clear condition of quartz grains elsewhere in Vein 9. In sections cut parallel to ab of the fabric the direction of elongation is parallel to the trace of the intersected veins (12 and 13) and therefore approximately but not exactly transverse to Vein 9 (cf. Figs. 1B, 2D, 13). In bc sections the grains are again elongated parallel to the trace of the veins intersected by Vein 9 (6 and other members of the V2 Series), but an ac section shows no appreciable elongation of grains situated at the junction of Veins 9 and 6. In all the observed instances, then, the quartz grains of these intersection fabrics tend conspicuously toward elongation in directions which lie somewhere between the b fabric axis and the normal to Vein 9. It is probable that the blastetrix differs slightly for the different intersected veins. Maximum ease of growth for directions parallel to the intersected vein, combined with minimum ease of growth parallel to the plane of Vein 9,-could result in elongation of grains in that direction within the intersected vein which makes the greatest possible angle with the plane of Vein 9. This direction, T in Fig. 13, is almost the same for Veins 12 and 13, in which case it is inclined at about 82° to the plane of Vein 9.

Attempts were made to estimate the degree of lattice orientation accompanying the strong dimensional orientation of quartz grains in these intersection fabrics. In Fig. 13 are plotted the poles of 35 quartz axes measured in ab sections across the intersections of Vein 9 with Veins 12 (25 grains) and 13 (10 grains). The only symmetrical feature of this diagram is the minimum surrounding T the pole of the assumed blastetrix; in none of the grains measured is the optic axis

– 316 –

inclined at less than 40° to the normal to the blastetrix. However, when Fig. 13 is compared with Fig. 15 which represents the normal quartz fabric for Vein 12, it will be seen that b, the axis of the quartz girdle and the centre of the minimal area of Fig. 15, coincides approximately with T of Fig. 13. It seems probable therefore that such preferred, lattice orientation as is displayed in Fig. 13 has been inherited with little change from the original fabric of the intersected veins (Veins 12 and 13).

In Fig. 18 the poles of 36 quartz axes from the intersection of Veins 9 and 6, measured in a bc section and subsequently rotated into ab, have been added to the 35 poles plotted in Fig. 13. The collective diagram shows no significant preferred orientation, such as might be expected if the lattice orientation of the grains were governed only by growth from a blastetrix of constant orientation throughout the rock.

Specially instructive is Fig. 19 which represents the axes of 140 typical dusty grains of quartz from an intersection between Veins 9 and 6 as seen in a section cut parallel to ac. The diagram shows a strong minimum centred about the c fabric axis. Across a broad equatorial zone normal to c the distribution of maxima and minima corresponds in detail with that shown by the equivalent zone of Fig. 20, which depicts the general orientation of quartz grains in Vein 6 as determined from two other thin sections. The influence of the original preferred orientation of quartz grains in Vein 6 upon the subsequently developed “intersection fabric” is obvious.

Picture icon

Fig 14—A, Vein 9 (V1 Series) cutting velnlet of V1 Series without offset. B, Vein 12 (V3 Series) cutting Vein 6 (dotted) and smaller veinlets of V2 Series. C. Vein 11 (V1 Series) intersected by Vein 6 (dotted) and other veinlets of V2 Series.

It is the writer's opinion that Vein 9, like other members of the V4 series, developed by a process of metasomatic replacement proceeding outward from an initial fracture. Where this fracture cut across pre-existing quartz veins conditions favoured the growth of greatly elongated grains of quartz parallel to the wall of the older vein and at as high an angle as possible to the plane of the growing vein. This process must have involved elimination of many preexisting grains at the intersections, but the effect of this selective

– 317 –

growth upon the lattice orientation of the fabric seems to have been inconsiderable, except at the junction with Vein 6. Here the features of the old fabric seem to have been retained in detail, except that all grains whose axes were inclined at low angles to c of the fabric have been eliminated.

Applying the geometrical criteria for emplacement of igneous dykes by a process of replacement, as recently put forward by G. E. Goodspeed (1940), further convincing evidence of replacement origin for Vein 9 is furnished by its relation to intersected veins of the other series. Oblique intersections with Veins 8, 13, and other members of the V1 Series and with Vein 12 of the V3 Series, as seen in sections and surfaces parallel to ab or ac of the fabric, show no offsetting of the intersected vein (Fig. 14A) such as must have occurred if Vein 9 had originated by dilation. This is the only available instance where a vein of the V4 series obliquely intersects a continuous older structure capable of being traced with certainty on both sides of the vein.

Petrography and Fabric of the V3 Series.

The single representative of the V3 Series is a yellowish vein (12) 2 mm. wide, differing little in orientation from some veins of the V1 Series. Unlike the latter, however, Vein 12 is certainly later in origin than the numerous V2 veins which it intersects. At intersections with Vein 12 in ac sections, most veins of the V2 Series have been offset by 2 mm. to 3 mm., a distance considerably greater than the displacement that would result from simple dilation of the vein walls during growth (cf. Goodspeed, 1940, p. 190, Fig. 16). In marked contrast with this general condition is the complete lack of offset of Vein 6 where it is cut by Vein 12 (Fig. 14B). It is therefore concluded that the fracture occupied by Vein 12 formed after the growth of most veins of the V2 Series, and that differential movement on this fracture caused the displacements noted above; Vein 6 developed later, while crystallisation of quartz and associated minerals in Vein 12 was later still. There is no conclusive evidence as to whether the vein developed by replacement or by dilation.

The constituents of Vein 12 are quartz (dominant), albite (10%–20%) and iron-stained carbonate (10%) with which pyrite is associated in one thin section; mica is absent, and stilpnomelane is restricted to occasional small wisps near the vein walls. The grains of quartz and albite are irregular and equidimensional in outline, and average 0·1 mm. to 0·2 mm., occasionally reading 1 mm. in diameter. Undulose extinction is often faintly, never prominently, shown by the quartz. Most of the albite is water-clear, but a few of the larger grains, sometimes showing a tendency to exert their crystal outlines against the quartz, are clouded with brownish dust.

The quartz fabric of Vein 12 is illustrated in Fig. 15 , a fabric diagram based upon measurement of total quartz (350 grains) in a

[Footnote] † In Fig. 15 and all other ac diagrams in this paper except Fig. 19 the original projection has been rotated through 180° about a so as to conform with the orientation of diagrams in previous papers by the writer. This must be borne in mind if comparison is made between diagrams such as Fig. 15, and Figs. 1, 2 and 14 which depict the geometrical relationship of the veins as they actually appear in sections and polished specimens.

– 318 –

section cut parallel to ac. This shows a rather weak but nevertheless definite cleft girdle the axis (b′) of which is inclined at 15°–20° to b of the megascopic fabric. A second diagram (Fig. 16) was prepared from measurements of 300 grains in a section cut parallel to ab, at a distance of 4 mm. from the first thin section. Fig. 17 represents the diagram of Fig. 16 rotated into the ac plane. It shows the same cleft girdle as in Fig. 15, with corresponding principle maximum M and submaxima N, P and R. In Vein 12 there is thus no trace of preferred orientation governed by growth, but instead a tectonite fabric of the well-known cleft-girdle type, with a tectonic axis b′ making an angle of 15°–20° with the megascopic b and therefore trending 5°–10° W. of N.

When Fig. 16 is compared with Fig. 7 (quartz in middle zone of Vein 4), exact correspondence as regards general pattern and the position of the girdle axis is apparent, while there is also a general similarity as to distribution of maxima within the girdle. The latest phase of deformation recorded in the fabric of the rocks has thus left nearly identical imprints upon two widely separated veins of totally different orientation. It is difficult to visualise how this deformation, resulting as it has done in but little visible distortion or displacement of the veins whose fabric it has determined, could involve external rotation of the component grains or intragranular slip on any but the smallest scale. The conclusion seems inescapable that a tectonite fabric such as that depicted in Fig. 7 or Fig. 16 can develop during deformation of only slight intensity. It would also appear not unlikely that recrystallisation, involving selective growth of favourably oriented grains under the influence of a stress-system symmetrically related to the axis b′, was responsible for the evolution of the fabrics in question rather than actual movements of rotation or slip.

Attention is also drawn to the general similarity between Figs. 15 and 17 and fabric diagrams of the cleft-girdle type recently published for quartz in horizontal veins (parallel to ab) in more completely metamorphosed schists from near Waipori, some 13 miles due north of the locality discussed in this paper (Turner, 1940a, p. 144, Figs. 11, 12). In this latter case too the girdle axis has a bearing of 5°–10° W. of N. (Turner, 1938, p. 108).

Petrography and Fabric of the V2 Series.

These are regular dark-coloured veins lying parallel to the principal schistosity S2 and hence also to the ab fabric plane. As seen in ac sections the smaller members of the series have been offset by several millimetres at intersections with Vein 12 (Fig. 14B), while further complications in their orientation are introduced by refraction of the vein walls wherever they cut the earlier-formed veins of the V1 Series (Fig. 14C).

The main constituent, quartz, is accompanied by abundant deep green stilpnomelane which imparts a dark green to black colour to the hand-specimen, and which in the wider veins (e.g., 6) is especially concentrated along the vein walls. Minor albite and iron-stained carbonate are typically present, while in Vein 6 there are also sparsely scattered sharply crystallised flakelets of colourless mica.

Picture icon

Fig. 3—Vein 4, Quartz. 132 large grains in marginal zones of vein (x and z, Fig. 2A). Contours 6, 4, 2, 0.8%; maximum concentration 6%. Broken line indicates the plane of Vein 4. Fig. 4—Vein 4. Quartz, 52 large grains in marginal zone x. Contours 6, 4, 2%. Broken ares show projection of small circle with centre V and radius 35°. Fig. 5—Vein 4, Quartz, 80 large grains in marginal zone z. Contours 6, 4, 1.2%. Fig. 6—Vein 4, Quartz, 85 small grains in marginal zone x, lying immediately adjacent to medial zone y. Contours 5, 2.4, 1.2%; maximum concentration, 6%. Fig. 7—Vein 4, Quartz, 244 small grains in medial zone y. Contours 3, 2, 1, 0.4%; maximum concentration, 4%. Fig 8—Veins 1 and 3, Quartz. 59 large grains. Contours 6, 3, 1.4%; maximum concentration 7%. Broken line indicates trend of vein-wall; v nearly coincides with the positive end of the a fabric axis.

Picture icon

Fig. 9—Vein 1, Quartz, 73 large and medium-sized grains including 25 plotted in Fig. 8. Contours 5, 3, 1.3%; maximum concentration, 5%. Broken line indicates trend of vein-wall. Fig. 10—Vein 1, Quartz, 170 small grains. Contours 3, 1.8, 0.6%; maximum concentration, 3.5%. Fig. 11—Vein 3, Quartz, 200 grains of all sizes. Contours 3, 1.5, 0.5%; maximum concentration, 5%. Fig. 12—Vein 9, Quartz, 90 grains; 50 measured in section parallel to ab, 40 measured in section parallel to bc and rotated into ab about b. Contours 5.5, 3.3, 1.1%; maximum concentration, 8% at L. Broken line indicates trend of Vein 9. Fig. 13—Vein 9, Quartz, elongated grains at intersections with Veins 12 (dots, 25 poles) and 13 (triangles, 10 poles). The small circle is drawn at 35° angular distance from T the pole of the assumed blastetrix. Other broken lines show the positions of Veins 9, 12 and 13 upon the projection. Fig. 15—Vein 12, Quartz, 350 grains measured in ac section. Contours 3.2, 2, 1.2, 0.3%; maximum concentration, 4% at M. Broken arc indicates trend of Vein 12.

Picture icon

Fig. 16—Vein 12, Quartz, 300 grains measured in ab section. Contours 3, 2, 1, 0.3% : maximum concentration, 4% at M. Fig. 17—Vein 12, Quartz. Fabric diagram of Fig. 16 rotated 90° about a. Fig. 18—Vein 9, Quartz, 71 elongated grains at intersection with Veins 6 (36 grains, measured in bc section), 12 (25 grains) and 13 (10 grains). Contours 4, 3, 1.5%. Fig. 19—Vein 9, Quartz, 140 grains at intersection with Vein 6, measured in section parallel to ac. Contours 5, 3.5, 2, 0.7%. Fig. 20—Vein 6, Quartz, 400 grains. Contours 3, 2, 1, 0.25%; maximum concentration, (K), 4%. Broken line indicates the trend of Vein 6. Fig. 21—Vein 6, Quartz, 400 grains measured in section parallel to bc; resultant diagram subsequently rotated through 90° into the plane ac. Contours 2.5, 1.5, 0.5%; maximum concentration, 3% at K. Broken line indicates the trend of Vein 6.

Picture icon

Fig. 22—Vein 6, Stilpnomelane. 94 poles to (001) cleavage, measured in ac section (central portion of vein). Fig. 23—Vein 6, Stilpnomelane, 45 poles to (001) cleavage, measured in bc section (central portion of vein). Fig. 24—Vein 11, Quartz, 300 grains. Contours 3, 2, 1.3, 0.3%: maximum concentration, 4% (K and L). Broken line indicates the trend of Vein 11. Fig. 25—Vein 8, Quartz, 230 grains. Contours 3.5, 2.6, 1.3, 0.4%. Broken arc shows the mean trend of Vein 11. Fig. 26—Vein 8, Quartz, 200 grains measured in section approximately parallel to bc. Contours 6, 4, 2, 0,5%; maximum concentration (K), 9%. Broken arc shows the local trend of Vein 11 within the measured section; e e = elongation of quartz grains. Fig. 27—Vein 8, Quartz. Fig. 26 (with the 0.5% contour simplified) rotated 90° about c into the ac plane.

– 319 –

The grains of quartz, sometimes showing faintly undulose extinction, are approximately equant in outline and average 0·1 mm. to 0·2 mm. in diameter in Vein 6. In a section of the same vein cut parallel to be superindividuals elongated transversely to the trend of the vein wall (i.e., parallel to c) and composed of ten or twelve subindividuals with a faint tendency towards elongation parallel to b were noted.

The quartz fabric of Vein 6 as determined by measuring 400 grains in two thin sections cut parallel to ac is represented in Fig. 20. Since the same pattern appeared in three elemental diagrams, prepared respectively for (a) 125 grains adjacent to the upper wall, (b) 175 grains from the medial zone of the vein and (c) 100 grains from traverses across the full width of the vein, it was concluded that the quartz fabric is essentially homogeneous for the vein in question. This conclusion is borne out by comparison of Fig. 20 with a second diagram (Fig. 21) that was constructed by plotting 400 measurements made in a bc section and subsequently rotating the resultant diagram through 90° into the ac plane. Coincidence of maxima, submaxima and the minimal area around b is almost exact. Fig. 20 resembles Fig. 15 in showing a cleft girdle with b′, a line inclined at about 30° to b, as axis. There is also a general similarity, not amounting to identity, as regards location of maxima within the girdle. The logical conclusion is that Fig. 20 represents a tectonite fabric developed during the deformation that also led to the preferred orientation shown by quartz in Vein 12. No recognisable trace of any growth orientation that might previously have existed has survived this deformation.

The two diagrams (Figs. 22 and 23) illustrating the orientation of stilpnomelane in Vein 6 were constructed from measurements of 94 flakes in a section parallel to ac and 45 flakes in a bc section respectively. In each case measurements include all flakes encountered in traverses along the central portion of the vein. The densely matted stilpnomelane that has crystallised along the vein walls is unsuitable for accurate measurement and was therefore excluded. If Figs. 22 and 23 are interpreted in conjunction, according to the principles discussed in a previous paper (Turner and Hutton, 1941, p. 231), the apparently sharply defined girdle of Fig. 22 proves, from the scattered disposition of the poles in Fig. 23, to be in reality a rather poorly developed girdle, the axis of which approximates to b of the megascopic fabric. The tendency, shown in Fig. 22, for the flakes to lie with (001) inclined at low angles to the vein wall (= ab) might equally well be interpreted as the result of growth orientation or as an indication of a concealed set of s-surfaces sub-parallel to ab. The latter alternative is favoured by comparison with Fig. 28 (representing stilpnomelane in Vein 8).

Measurement of 70 sharply crystallised flakelets of muscovite in ac sections and 28 similar crystals in a bc section revealed an almost complete lack of preferred orientation for this mineral, apart from a faint minimum about b.

Petrography and Fabric of the V1 Series.

The veins grouped in the V1 series vary considerably in orientation, but in general approximate to (hOl), intersect the main

– 320 –

schistosity ab at 30°–45° (rarely 70°), and hence commonly lie sub-parallel to the less well defined schistosity S1 (cf. Fig. 1B). Petrographically they resemble the veins of the V2 Series. Where they cut obliquely across earlier structures the trend of the latter continues across the intersection without offset , a fact indicative of origin of V1 veins by replacement rather than dilation. The fabric of Veins 8 and 11 was investigated in some detail, but with somewhat ambiguous results as far as the significance of the preferred orientation of the quartz grains is concerned.

Axes of 300 clear almost non-undulose equant grains of quartz measured in an ac section of Vein 11 are represented in Fig. 24. There is little trace of the cleft-girdle pattern recorded in Figs. 15, 20 and 21, but the two well-defined principal maxima K and L in Fig. 24, like those of the three other diagrams cited, are situated eccentrically with regard to b, but approximately symmetrically on each side of the plane of the vein in question (broken lines). However, K and L of Fig. 24 are closer to the plane of Vein 11 than are corresponding maxima of Figs. 15, 20 or 21 to the planes of Veins 12 or 6 respectively. In Fig. 24 there is no symmetry about the pole of the vein wall such as would be present if the fabric were dominated by growth orientation, nor is it likely that a growth fabric could have survived in Vein 11 but not in the younger veins such as 6 and 12.

Still more difficult to interpret is Fig. 25, which depicts the orientation of 230 grains of quartz in an ac section of Vein 8. This vein dips more steeply than others of the V1 Series, and is less regular in form, apparently as a result of disturbance by slip on s-surfaces which appear in thin sections to be inclined at angles of 10° or 20° to ab. Lack of any simple pattern in the quartz diagram Fig. 25 may perhaps be correlated with the deformed condition of the vein. Stil-pnomelane from the central portion of Vein 8 shows a more pronounced preferred orientation than in the other case discussed above.

Picture icon

Fig. 28—Vein 8, Stilpnomelane, 100 poles to (001) cleavage, measured in ac section (central portion of vein). Fig. 29—Vein 8, Stilpnomelane, 50 poles to (001) cleavage, measured in section (central portion of vein) approximately parallel to bc.

[Footnote] † For example, Vein 8 in section 6231D.

– 321 –

The results of plotting the pole of the (001) cleavage in 100 flakes in an ac section and 50 flakes in a bc section are given separately in Figs. 28 and 29. Strong concentration of poles adjacent to c in Fig. 29 shows that the peripheral concentration in Fig. 28 represents an actual girdle, the axis of which (b”) is inclined at 20° to b.

A second diagram for quartz in Vein 8 (Fig. 26) was based on 200 axes measured in a section subparallel to bc; the same diagram, with the 0·5% contour simplified, was rotated into the ac plane (Fig. 27) for comparison with Figs. 24 and 25. They are simpler in pattern than is Fig. 25, and represent a higher degree of preferred orientation (9% per 1% area at K) than is recorded in any other tectonite fabric described in this paper. Furthermore, the grains in question, though no coarser than in other sections of V1 or V2 veins, are noticeably elongated in a direction (e e) oblique to the schistosity but approximately parallel to b” the axis of the stilpnomelane girdle. The elongation is, therefore, regarded as of deformational origin, and the accompanying lattice orientation must likewise be attributed to tectonic causes. The pattern of Figs. 26 and 27 is dominated by two mutually perpendicular girdles with axes respectively parallel to e e (=b”) and subparallel to a of the megascopic fabric. Maximum L is located in the first of these, K in the second, while in each case the maximal area is elongated along the girdle in which it lies. When Fig. 25 is compared with Fig. 27 it will be found that the maxima L and K, as well as the minimal areas, correspond very well in the two diagrams and it is therefore concluded that the distribution of maxima and minima in Fig. 25, though far from regular, is not fortuitous. Obviously however, the quartz fabric of Vein 8 is not completely homogeneous.

Interpretation of Maxima in the Quartz Diagrams.

In view of the prevailing lack of unanimity of opinion as to (1) the mechanism by which preferred orientation of quartz grains develops in tectonites, and (2) interpretation of maxima in quartz diagrams (especially when these lie on small circles of the projection), the writer prefers not to speculate upon the possible significance of the maxima recorded in the diagrams that accompany this paper. Not only is it generally recognised that a number of distinct types of maxima are theoretically possible for S-tectonites (cf. Sander, 1930, D61; Griggs and Bell, 1938, pp. 1741–1744), but in the rocks here described there are at least two sets of schistosity-surfaces, S1 and S2, and the trend of the tectonic axis (b, b′, b”) varies through a fairly wide range. It is therefore not surprising to find some agreement between the positions of individual observed maxima and those of maxima which might be expected assuming that the quartz has been oriented in S1 or S2 according to one of the recognised theoretically possible rules. For example, in Fig. 15, assuming the plane of Vein 12 (= S1) as the controlling slip-plane and b′ as the corresponding tectonic axis, maxima N, M and R might be interpreted as equivalent to Type VI of Griggs and Bell (loc. cit., pp. 1742, 1743) while P would correspond to Type I. Again in Fig. 27 the concentration of axes at L could have developed by alinement of (0001) in

– 322 –

s-planes parallel to the girdle axis b” and inclined at low angles to ab. On the other hand, since the growth of all veins in these rocks took place after metamorphie deformation had passed its greatest intensity, it is only to be expected that in some diagrams there will be (as actually is the case) maxima which, while certainly not fortuitous, are difficult to reconcile with orientation on any visible set of s-sur-faces governed by any of the theoretically possible rules at present recognised.

Summary and Conclusions.

The following facts have been revealed by the present petro-fabric investigation of quartzose veins in sixteen thin sections cut from two large blocks of “semischist” from near Mt. Stuart railway station, south-east Otago:—


Within the field of a large hand-specimen the quartz fabric for the rock as a whole is not homogeneous. Veins that are differently oriented with respect to a, b and c of the megascopic fabric frequently though not invariably show different patterns of preferred orientation of the component quartz grains. Usually the quartz fabric of any individual vein is approximately homogeneous, except for veins in which growth fabrics predominate (i.e., most veins of the V4 Series).


In the three earliest sets of veins, viz., the V1, V2, and V3 Series, respectively approximating to (h O l) (= S1), ab (= S2) and (h O l) of the fabric, no recognisable trace of either dimensional or space-lattice orientation controlled by growth is retained in the quartz fabric, except for superindividuals elongated perpendicularly to the vein wall in a bc section of Vein 6 (V2). The most general pattern of the fabric diagrams is a cleft girdle about a subhorizontal axis b′ which is inclined at 20°–30° to the megascopic b and trends N. 5°–20° W. This is comparable with the previously described pattern and field orientation of the quartz fabric for ab veins in a schist from near Waipori, 13 miles distant. The quartz fabric of the earliest V1 veins does not conform to this simple pattern and is difficult to interpret apart from recognising its tectonic origin. In one instance (Fig. 26) a B B′ pattern is clearly indicated (B = a of the megascopic fabric; B′ is a line in the bc plane inclined at 20° to b and parallel to b” the axis of the stilpnomelane girdle of the same vein). In no case was the preferred orientation, even when of ambiguous significance, found to be fortuitous.


The characteristics of the quartz fabric in veins of the latest series V4, subparallel to ac or to steep (O k l) surfaces are summarised thus:—


In narrow veinlets less than 1 mm. wide there is no recog-nisable preferred orientation of quartz according either to dimension or to space-lattice structure (e.g., Vein 9).


In wider veins, 3–5 mm. in width, preferred orientation of varying intensity has been controlled by a blastetrix parallel or nearly so to the vein wall. Strength of lattice orientation can be correlated directly with degree of dimensional orientation in these veins.

[Footnote] † The presence of such s-planes is indicated by the stilpnomelane fabric of the vein in question.

– 323 –

In many of these wider veins only the coarser grains of quartz show recognisable preferred orientation. It is completely lacking in the finer-grained quartz, some of which has been produced by ruptural breaking down of larger individuals.


In one case (Vein 4) a central zone composed of small grains of quartz, having a tectonite fabric with a weakly defined girdle about b′ is flanked on either side by a lateral zone of elongated grains the fabric of which has been determined by growth from a blastetrix.


Generally speaking the preferred orientation of both quartz and stilpnomelane in the investigated veins is much less perfectly developed than that in the quartz-rich bands of coarse schists of subzone Chl. 4 in Eastern Otago described elsewhere.


By analogy with the present instance the quartz fabrics for veined schists in general throughout eastern and central Otago can be expected to be controlled mainly by deformation. Growth fabrics should be represented only in late ac and (O k l) veins.

From a consideration of petrographic, petrofabric and other structural data brought together in this paper, the writer offers the following conclusions as to the history of the investigated rocks (compare with conclusions reached in Turner, 1940, pp. 189, 190; Turner and Hutton, 1941, p. 239) :—


Quartzose veins, which also contain other minerals present in the adjacent semischists, have developed in the rocks of Subzone Chl. 2 of the Chlorite Zone after metamorphic deformation of the latter had passed its climax. The earliest veins came into existence after the two macroscopic schistosities had been established in the enclosing semischist, but deformation certainly continued after development of the three earlier, and at least one member of the fourth and latest series of veins. Vein formation is therefore visualised as a metamorphic process spread over several distinct stages in the later phase of regional metamorphism in this area.


As in other parts of Eastern Otago the tectonic axis of this later phase of deformation departed somewhat from that connected with the earlier phases.


In chronological sequence, veins tended to form more or less parallel to the following structural planes: S1 [= (h O l)]; S2 (= ab); S1; ac and steep (O k l) surfaces.


There is no evidence that the bulk of the materials of which the veins are composed has been introduced from external sources such as igneous intrusions, but rather from the discontinuous nature of the veins themselves and from the presence in them of abundant albite, stilpnomelane, etc., origin by metamorphic differentiation may be inferred (cf. Turner, 1941). This does not exclude the probability that some minor constituents, in this case the sulphides, have been introduced from an external magmatic source. The writer pictures slow development of veins from pore-solutions, concentrated along fractures and shear planes, and containing at a given time such dissolved material as would be required for maintenance of equilibrium between solution and minerals in the surrounding schist under the local temperature-pressure conditions. With waning metamorphism

– 324 –

and simultaneously decreasing temperature, the following sequence of mineral assemblages, as shown by the petrography of the veins, has crystallised from the pore solutions:—


quartz, albite, stilpnomelane, chlorite, muscovite, epidote (and/or pumpellyite); in enclosing schist only.


quartz, albite, stilpnomelane, calcite (minor muscovite, apatite); V1 and V2 Series.


quartz, albite, calcite (minor apatite); V3 Series.


quartz, albite, actinolite (minor apatite); V4 Series.


Pyrrhotite and pyrite, introduced after full growth and subsequent fracturing of veins of V4 Series.


There is good evidence that certain veins of the V1, V2 and V4 Series grew by metasomatic replacement extending outward in each case from an initial hair-crack. There is no certain evidence that any of the veins described originated by infilling or dilation of initial cracks or fissures. It is, therefore, considered that origin by replacement is likely in all cases.


Tectonite fabrics such as those depicted in Figs. 15 and 20 can develop in veins, apparently by a process of selective recrystallisation, during deformation which brings about only slight strain in the rock as a whole.


The gentle south-easterly dip of the main schistosity S1 appears to have no connectioin with any phase of metamorphic deformation discussed above, and by analogy with other instances in Otago is thought to be a result of much later local tilting (cf. Turner, 1940, p. 189).


The thanks of the writer are due to the Council of the Royal Society of New Zealand, who defrayed the cost of making oriented rock sections with a research grant from the Hutton Memorial Fund.

Literature Cited.

Eskola, P., 1939. In Die Entstchung der Gesteine (T. F. Barth, C. W. Correns, P. Eskola), Berlin, J. Springer.

Fairbairn, H. W., 1937. Structural Petrology, Queen's University, Kingston, Canada.

Goodspeed, G. E., 1940. Dilation and Replacement Dykes, Jour. Geol., vol. 48, pp. 175–195.

Griggs D., and Bell, J. F., 1938. Experiments bearing on the Orientation of Quartz in Deformed Rocks, Bull. Geol. Soc. Amer., vol. 49, pp. 1723–1746.

Knopf, E. B., and Ingerson, E., 1938. Structural Petrology, Mem. Geol. Soc. Amer., no. 6.

Sander, B., 1930. Gefügekunde der Gesteine, Vienna, J. Springer.

Turner, F. J., 1938. Petrofabric Investigations of Otago Schists, no. 2, Trans. Roy. Soc. N.Z., vol. 68, pt. 1, pp. 107–121.

— 1940. Structural Petrology of the Schists of Eastern Otago, New Zealand, Amer. Jour. Sci., vol. 238, pp. 73–106, 153–191.

— 1940a. Applicability and Limitations of Petrofabric Analysis without use of a Universal Stage, Trans. Roy. Soc. N.Z. vol. 70, pt. 2, pp. 140–149.

— 1941. The Development of Pseudostratification by Metamorphic Differentiation in the Schists of Otago, New Zealand, Amer. Jour. Sci., vol. 239, pp. 1–16.

Turner, F. J., and Hutton, C. O., 1941. Some Porphyroblastic albite-schists from Waikouaiti River (South Branch), Otago, Trans. Roy. Soc. N.Z., vol. 71, pt. 3, pp. 223–240.