Go to National Library of New Zealand Te Puna Mātauranga o Aotearoa
Volume 66, 1937
This text is also available in PDF
(2 MB) Opens in new window
– 201 –

Interpretation of Schistosity in the Rocks of Otago, New Zealand

[Read before the Otago Branch, May 19, 1936; received by the Editor, May 20, 1936; issued separately, September, 1936.]



The Significance of Schistosity.

  • Definitions.

  • Theories of Sharpe and Sorby.

  • Leith's Theory.

  • Becker's Theory.

  • Sander's Theory.

Development of Schistosity in Otago Schists.

Some Petrofabric Data for New Zealand Schists.

  • Summary of General Orientation Rules.

  • General Indications in New Zealand Schists.

  • Partial Analysis of Two Quartz Fabrics.

  • Petrofabric Evidence as to Course of Metamorphism.


Literature Cited.


No problem is more important in metamorphic geology than the interpretation of those structures to which the term schistosity is generally applied. In the preface to his account of experiments upon the origin of slaty cleavage G. Becker (1904, pp. 10, 11) emphasised this fact in these words;—“Scarcely a mountain range exists, or has existed, along the course of which belts of slaty rock are not found, the dip of the cleavage usually approaching verticality. Are these slate beds equivalent to minutely distributed step faults of great total throw or do they indicate compression per pendicular to the cleavage without attendant relative dislocation? Evidently the answer to this question is of first importance in the interpretation of orogenic phenomena.” Nevertheless, in most textbooks written in English the interpretation of schistosity with reference to pre-existing rock-structures and to the deformation accompanying its development is inadequately treated.

Recent statistical investigations of grain-orientation in dynamically deformed rocks by a number of European workers, headed by Professors Bruno Sander, of Innsbruck and Walter Schmidt, of Berlin, have yielded a mass of valuable data bearing upon these and related problems, and new conceptions of the significance of schistosity have arisen (e.g., see Sander, 1930, 1934, 1934a; Knopf, 1933). The explanation of schistosity thus reached by European petrologists in some respects differs radically from those hitherto held by most British and American workers. At the same time certain principles embodied in older hypotheses (notably, that of

– 202 –

G. F. Becker) play an important part in the modern theories advanced by Sander and Schmidt. For a complete understanding of the latter it is therefore necessary to consider also the views of earlier workers on the subject of schistosity.

In the first part of this paper a brief review is given of some of the more important of the older conceptions of schistosity, especially those of G. F. Becker and C. K. Leith, and the essentials of Sander's theory are outlined for comparison therewith. The writer himself makes, in this section, no new contribution to the problem, but merely presents a summary for the use of other workers in New Zealand and to serve as a basis for discussion of the data recorded in the second part of the paper. In the latter the available facts concerning the nature and mode of origin of schistosity in the schists of Central and Western Otago are summarised and examined, incomplete and even fragmentary though they are. Satisfactory solution of local problems, however, cannot be possible until detailed analyses of grain orientation have been made and the results correlated with data obtained in the field.

The Significance of Schistosity.


The term schistosity is used in different senses by different writers. Most English petrologists (e.g., Harker, 1932, pp. 194, 197; Tyrell, 1932, pp. 272, 273) use the term schistosity to denote dimensional parallelism of the constituent mineral grains in a rock resulting from crystallisation under the influence of directed pressure; slaty cleavage, on the other hand, is applied to rocks in which the dimensional parallelism of the grains is believed to be due in part to mechanical processes, such as rotation and “flow ” (Harker, 1932, pp. 153–155; Tyrell, 1930, pp. 283, 284). American writers, on the other hand, do not make a distinction between schistosity and slaty cleavage. C. K. Leith and his associates use the term flow cleavage to cover both, but, as recently pointed out by Mrs Knopf (1931, p. 17), the expression is unfortunate, as it implies a condition of genesis now known not to be universal. G. F. Becker (1904, p. 9) recognised the identity of schistosity and slaty cleavage, and distinguished the latter only as “the most regular and extreme form of cleavability or schistosity.” In contrast with Leith and other petrologists of the Wisconsin school, Becker held that the fundamental condition causing schistosity was weakened cohesion arising from shearing, and that the dimensional parallelism of the constituent grains was incidental and not an essential condition to schistosity.

The recent investigations of Sander, Schmidt and others have now shown clearly that there are several distinct types of schistosity arising in a number of different ways. A broad and essentially non-genetic definition is therefore necessary. Throughout this paper the term schistosity will be used to denote the property by virtue of which rocks cleave along surfaces (not necessarily plane) determined by crystallisation or mechanical deformation of the rock forming material in the solid state under the influence of stress or

– 203 –

high temperature. Schistosity as thus defined is a metamorphic phenomenon, and does not include parallel fabrics or fissility resulting from flow in partially molten masses.

The term foliation also requires accurate definition. Some writers (e.g., Fairbairn, 1935) use it loosely as being equivalent to schistosity. Others, such as Becker (1904, pp. 9, 10), apply the word foliated to the rather complex type of schistosity in which two or more directions of fissility intersect at low angles. The usage followed in this paper will be that employed by Harker (1932, p. 203), who states that “foliation consists in a more or less pronounced aggregation of particular constituent minerals of the metamorphosed rock into lenticles or streaks or inconstant bands….” The necessity for some such term is obvious, since many schistose rocks of low metamorphic grade are non-foliated in the above sense of the word, while hornfelses which completely lack schistosity may show well-marked foliation due to accentuation of differences in composition of adjacent layers in the initially banded parent rock (Turner, 1933, p. 197).

Theories of Sharpe and Sorby.

Sharpe, Sorby and other early workers on dynamically metamorphosed rocks held the view that schistosity develops in a direction perpendicular to that of the external force which brings it into existence. It is not necessary to discuss the exact nature of the processes by which this condition is imagined to be brought about, as full details are easily accessible elsewhere (see Becker, 1904, pp. 11, 12; Harker, 1932, pp. 153, 154). It is interesting to note, however, that the explanation advanced by Harker (1932, pp. 154, 155, 193–195) is also based upon the supposition that schistosity always lies at right angles to the direction of the deforming force.

Leith's Theory.

The theory of schistosity advanced by Leith, Van Hise, Mead and their associates has been discussed exhaustively by Leith in particular (e.g., Leith, 1905; Leith and Mead, 1915, pp. 169–182), and is usually associated with his name. It is based largely upon the results of field-studies on the Precambrian rocks of North America, and has been widely accepted by American and English geologists. The essential features of Leith's hypothesis may be summarised thus:—


Schistosity is the result of rock-flowage involving two distinct processes, both of which induce dimensional parallelism of the component mineral grains in the deformed rock. These are recrystallisation of grains with their long axes in parallel position, and mechanical rotation of preexisting grains.

– 204 –

The direction of schistosity is always parallel to the direction of elongation of the deformed rock mass, i.e., normal to the C (shortest) axis of the strain ellopsoid.*


In cases of non-rotational strain, when the deforming force acts perpendicularly to the plane of resistance (cf. compression between the jaws of a vice), the resultant schistosity is normal to the direction of the deforming force (Fig. 2a).


When the rock-mass undergoes rotational strain, i.e., when the deforming force is inclined to the plane of resistance, the positions of the principal axes of the strain ellipsoid change continuously with reference to the direction of the casual force. The rotational element involved brings the schistosity into a position inclined obliquely to the direction of greatest external stress, though at any instant during deformation the schistosity tends to develop normal to that direction. The extreme case of rotational strain is pure scission (cf. displacement of a pack of cards). Leith's conception of the orientation of schistosity resulting from scission is illustrated in Fig. 2b.


Strain-slip cleavage (fracture cleavage) is a secondary structure superposed upon schistosity under a new set of stress conditions. It is interpreted as a shear structure comparable with Becker's conception of schistosity (see below).

Becker's Theory.

An important contribution to the elucidation of the schistosity problem was made in 1893 by G. F. Becker, who later brought forward convincing experimental evidence in support of his view (Becker, 1904) and destructively criticised the theory of Leith (Becker, 1907). Becker's theory is based upon mathematical analysis of stress and strain in a homogeneous body undergoing plastic deformation without change in volume, under the influence ofsimple external forces. His conclusions, though lucidly and convincingly presented, were until recently largely neglected as being supposedly inapplicable to schistosity developed under natural conditions. Petrofahric investigations in recent years have shown, however, that Becker's mathematically deduced conception of schistosity is of fundamental importance, if allowance is made for the discrepancy

[Footnote] * The strain ellipsoid is the ellipsoidal body assumed by an initially spherical portion of the rock-mass during its deformation. The three principal axes in order of decreasing length are denoted by A, B and C. Becker was the first to explain the orientation of schistosity in deformed rock masses in terms of the strain ellipsoid.

– 205 –

that exists between the ideal case assumed by him and the more complex conditions which prevail in the deformation of heterogeneous rocks. The following is a summary of Becker's theory:—

Picture icon

Fig. 1.—AC section of strain ellipsoid showing relation between directions of schistosity S, strain-slip S′ and the deforming force (indicated by arrow), according to the theory of G. F. Becker: (a) for nonrotational strain, (b) for pure scission (a rotational strain).

Picture icon

Fig. 2.—The same according to the theory of C. K. Leith.


Deformation of the type under consideration is accomplished by differential slipping along two sets of parallel planes of maximum slide. These are also planes of no distortion (circular sections) in the strain ellipsoid, since they retain their circular outline unchanged during the process of distortion of the initial sphere. Both sets of slip-planes are parallel to (and therefore intersect along) the B axis of the strain ellipsoid, and the direction of slip is perpendicular to B.


When this differential movement exceeds the elastic limit, but falls short of the breaking strain, a permanent weakening of cohesion between the planes of no distortion results. Schistosity is the tangible expression of this weakened cohesion, and the schistosity-planes are thus parallel to the circular sections of the strain ellipsoid.

– 206 –

Parallel orientation of minerals with their long axes in the plane of schistosity results from mechanical arrangement during flowage, and is thus the effect rather than the cause of schistosity (cf. Knopf, 1933). At the same time mechanically generated heat facilitates mineral reconstruction, and the resultant growth of new crystals in the direction of least resistance along the planes of slip may somewhat intensify schistosity arising from flowage (Becker, 1904, p. 22).


In cases of non-rotational strain, schistosity develops in two directions symmetrically inclined at an angle between 45° and 90° to the short or C axis of the strain ellipsoid; i.e., to the direction of the external deforming force (Fig. la).*


In cases of rotational strain (much more frequent in nature than non-rotational strain) the B axis of the ellipsoid is the axis of internal rotation or axis of shear. The two sets of planes of maximum shearing-stress are thus parallel to the axis of internal rotation. During strain their positions change continuously. One set of planes sweeps through the strained mass very much more rapidly than the second. set. Viscosity is sufficiently high to prevent flowage on the fast-moving set, and here the only possible appreciable effect is rupture, which sometimes does actually occur, giving rise to strain-slip cleavage. On the slow-moving set shearing-stress acts long enough in a given direction for viscosity-effects to be reduced to a small value, and here flowage takes place and schistosity results. In the extreme case of pure scission (approached but never actually attained in nature) the causal force is parallel to the resultant schistosity (Fig. 1b). In other instances of rotational strain there is no simple relaxation between the direction of the casual force, the axes of the strain ellipsoid and the schistosity-planes.

Becker's theory has been criticised severely on the grounds, first, that the case postulated (homogeneous strain of isotropic bodies) is not analogous to natural deformation of rock-masses, and, secondly, that the deduced relationships between directions of shearing-planes, causal force and the axes of the ellipsoid are based upon strain within the elastic limit, whereas schistosity is developed under conditions involying strain beyond the elastic limit. Further, the fundamental assumptions that the planes of no distortion are surfaces of maximum slide. and that rupture takes place along such surfaces have been questioned (e.g., Griggs, 1935).

Nevertheless, Becker, working from an ideal simple case, arrived at results which can be applied profitably to the more complex cases involved in deformation of rocks. Mrs E. B. Knopf (1933, p. 447) has recently stated that “the most notable feature of Becker's

[Footnote] *For the argument upon which this conclusion is based the reader is referred to Becker's papers (cf. also Harker, 1932, p. 140).

– 207 –

theory is that deformation takes place by slipping of material along planes predetermined by their relation to the stress distribution in the body undergoing strain.” That schistosity very commonly does develop parallel to these planes of shearing is fully borne out by petrofabric investigations. This now fully established conclusion is in direct opposition to the frequently asserted views that schistosity develops normal to the direction of the deforming force, or parallel to the AB plane of the strain ellipsoid.

A further important conclusion arising from Becker's work is that the presence in a rock of intersecting sets of schistosity-planes, or of schistosity crossed obliquely by strain-slip cleavage, does not necessarily indicate repeated metamorphism. Such multiple structures are to be expected as the result of a single deformation, if schistosity is regarded as a shear structure. Indeed, on this interpretation of schistosity it is the frequent development of only one set of schistosity-planes that requires explanation. Such has been supplied by Becker for cases of rotational strain as outlined under (5) above.

Sander's Theory.

Sander's theory of schistosity is based upon the results of petrofabric analyses—especially statistical optical investigation of grain-orientation—of schistose rocks (Sander, 1930). There is an extensive German literature on the subject, much of which is not available in New Zealand. A summary in English of the principles upon which the method is based and the results achieved in the field of metamorphic petrology has recently been given by Mrs E. B. Knopf (1933), while instructive illustrations of the theory have appeared in papers by Gilluly (1934), Sander (1934), Knopf (1935) and Fairbairn (1935, 1935a).*

In addition a useful and clearly presented introduction (in English) to the whole subject of petrofabric analyses has recently been published by H. W. Fair-bairn (1935b), of Queen's University, Canada.

An outline of Sander's conclusions as to the origin and significance of schistosity is given below, the petrofabric evidence upon which his theory is based being as far as possible omitted:—


The first permanent deformation in a rock mass is accomplished by differential movement along parallel slip-surfaces. This movement involves rotation of mineral grains and gliding of individual grains along crystallographiec gliding-planes, and in this way for each mineral a regular crystallo graphic (space-lattice) orientation of the component grains is achieved. The study of the type of orientation-patterns and the degrees of orientation thus produced in the various minerals is an important part of petrofabric analysis, and yields valuable evidence bearing upon problems of schistosity.


The schistosity thus initiated typically is parallel to the slip-surfaces (cf. Becker). In some instances, however, schistosity is determined by elongation of mineral grains

[Footnote] *See also E. Ingerson, Am. Jour. Sci., vol. xxxi, pp. 161–187, 1936.

– 208 –

(e.g., quartz or calcite) in a plane (“plaiting-surface ”) bisecting the acute angle between two intersecting slip-surfaces, and hence is parallel to the AB plane of the strain ellipsoid (cf. Leith). Fairbairn (1935, 1935b) has suggested that schistosity parallel to plaiting-surfaces may prove to be more widely developed than has hitherto been recognised.


The slip-planes may be mechanically-induced surfaces arising subparallel to the circular sections of the strain ellipsoid. This condition usually holds in the case of relatively isotropie rocks such as shales, so that the resultant slaty cleavage of the deformed rocks has no relation to the original bedding.


In markedly anisotropic rocks the slip-planes typically are pre-existing surfaces of weak cohesion such as planes of bedding, flow, foliation or previously developed schistosity, all of which are included by Sander in a single category as s-planes.- During deformation of a mass consisting of alternating beds of variable competency, slipping along the intervening s-planes is often accompanied by development of flexures in the competent beds. The incompetent beds simultaneously yield by slipping, giving rise to minor drag-folds which are usually essentially slip-folds* (Knopf, 1935). When deformation is complete the original s-planes have been transposed into a new direction (Knopf, 1931, pp. 16–18) but still retain their identity. The schistosity so developed is parallel to the transposed s-planes and has been termed “transposition cleavage,” in contrast with “slaty cleavage ” which cuts across the original s-planes. This “transposition eleavage ” may come to lie in a direction normal to that of the external causal force (cf. Leith's view). Strain-slip cleavage may be regarded as an intermediate stage in the development of. “transposition cleavage ” and is therefore a feature of the early stages of deformation of anisotropic bedded rocks. It must be noted, however, that when an initially isotropic shale has acquired schistosity of the “slaty cleavage ” type, it has become markedly anisotropic; further deformation may result in isoelinal folding of the slaty cleavage, giving first strain-slip cleavage and perhaps ultimately a new schistosity of the “type. In slaty rocks of initially relatively isotropie character the development of strain-slip cleavage, therefore, is characteristic of the late stages of deformation, or may indicate repeated metamorphism (Knopf, 1935)


Schistose rocks which owe their structure to differential movement of the constituent grains during deformation belong to the major group of rocks which Sander terms

[Footnote] * As explained clearly by Mrs Knopf (Knopf, 1933, p. 464) slip-folds are structures resulting from pure non-homogeneous shearing along planes parallel to the axial plane of the slip-fold. Their presence therefore does not imply the existence of a compressive force acting normal to the fold axis.

– 209 –

tectonites. Two main divisions of the class are recognised on the basis of the type of motion involved in the deformation, as indicated by petrofabric analysis. In S-tectonites the motion is dominantly one of crystal gliding in the shear-planes (cf. the lateral displacement of a pack of cards); in B-tectonites the motion is essentially an external rotation of the component particles about the B axis of the strain ellipsoid, comparable with the movement of balls in a ball-bearing. The two classes of tectonite are not sharply separable: some degree of gliding of grains in the slip-planes in one or more directions perpendicular to B is always indicated in B-tectonites, while orientation-diagrams of S-tectonites invariably show traces of minor effects of external rotation. In many tectonites, especially B-tectonites, a linear element in the schistosity is developed at right angles to the direction of translatory movement and parallel to the axis of rotation (i.e., to B of the strain ellipsoid.) This linear direction has thus the same tectonic significance as the direction of fold axes. It is usually clearly marked in the hand-specimen, and may be due to (a) intersection of two slip-planes, or (b) minute corrugations and folds, or (c) elongation of mineral grains (e.g., quartz, calcite, amphibole, etc.) parallel to the B axis of the strain ellipsoid.*


In many schists, especially those of coarse grain, the structure is the result of crystallisation after cessation of tectonic movement. The schistosity is here due to growth of crystals with their long axes in the direction of greatest ease of growth, i.e., parallel to the best-developed s-planes. Crystallisation of this type, whereby a pre-existing anis-totropy is preserved and sometimes even accentuated in the recrystallised rock, is termed by Sander (1930, p. 172) Abbildungskristallisation (“mimetic crystallisation” of F. E. Suess and E. B. Knopf, “portrait crystallisation ” of H. W. Fairbairn). Two cases are to be distinguished:—


The space-lattice orientation-pattern of the initial tectonite is preserved by continuation of growth of orientated “seed-crystals.” At the same time the grains show marked dimensional orientation.


When recrystallisation is dominated by growth of new crystals, the orientation is purely dimensional and the tectonite orientation-pattern is lost.

In porphyroblastic schists large crystals of albite, mica, hornblende, etc., with purely dimensional orientation may be enclosed in a tectonite matrix having space-lattice orientation. “Crystallisation-schistosity ” resulting from mimetic crystallisation has hitherto often been explained as due to the operation of Riecke's principle governing recrystallisation. Petrofabric analysis shows this explanation to be incorrect.

[Footnote] *For discussion on this latter point see Fairbairn, 1935b, pp. 46–48.

– 210 –

Any theory of schistosity should account for the fact that there is a single predominant direction of cleavage in most schistose rocks. Leith's hypothesis that cleavage is developed parallel to the AB plane of the strain ellipsoid readily explains this phenomenon; but if schistosity-planes are interpreted as shear-planes, as in the hypotheses advanced by Becker, Schmidt and Sander, the problem is more difficult to solve. Sander (1930, especially pp. 101–103) has discussed fully this question of Einscharigkeit. Petrofabric analyses, while not yet proving the truth or otherwise of Becker's ingenious explanation of the development of single schistosity by rotational strain (see under 5, p. 11), has thrown much light upon the question as a whole. Sander believes that the grain-fabric in most tectonites indicates the existence of at least two sets of s-planes. When one of these is more strongly marked than the others it usually defines the schistosity. When there are two equally well-developed sets of s-planes the result may be either a single macroscopic schistosity (plaiting-surface) bisecting the acute angle of intersection, or alternatively a symmetrically developed double schistosity. In cases where the original rock was markedly anisotropic, e.g., in a well-bedded sediment, shearing along only one set of s-planes (the original s in the undeformed rock) is to be expected.

From the above discussion it is clear that Sander's conception of schistosity is much broader than-previous hypotheses. Several distinct types of schistosity are recognised :—-


Schistosity parallel to shear-planes (s-planes), whether developed mechanically for the first time or moulded upon s-surfaces of a pre-existing anisotropy. Note that though the greatest importance is attached by Schmidt and Sander to Becker's hypothesis, it is clearly recognised that the latter is not a universal theory of s-planes nor does it explain all those structures to which the term schistosity is applied (Sander, 1930, p. 99).


Schistosity parallel to plaiting-surfaces, believed to be parallel to the AB plane of the strain ellipsoid (cf. Leith).


Schistosity developed parallel to pre-existing s-planes by mimetic crystallisation (cf. Leith). This includes post-tectonic crystallisation in the late stages of metamorphism, or non-tectonic erystallisation governed-by the s-planes of an undeformed anisotropie rock.


Tension joints perpendieular to B typically are well developed in tectonites. Sometimes these may be so closely spaced that they impart to the rock a pronounced fissility or schistosity which is thus parallel to the AC plane of the strain ellipsoid (Sander, 1930, p. 219).

It is also clear that the direction of the external deforming force can seldom be deduced from the schistosity of a rock. On the other hand, the “tectonic axis” parallel to B of the strain ellipsoid, and the orientation of shear-planes can usually be determined with considerable precision.

– 211 –

Development of Schistosity in Otago Schists.

The mineralogical changes which accompany the transition from greywacke and slate to schists of the Central Otago type, and the passage of these in turn into high-grade schists such as those of South Westland, have been described previously (Turner, 1933. 1934, 1935). An attempt is now made to summarise the evidence bearing upon the mechanical evolution of the schistose structure in the schists of Otago. Though admittedly this field has as yet been little more than touched upon, it is nevertheless possible to suggest at least partial answers to questions as to the cause, mechanism and degree of complexity of metamorphism experienced by these rocks.

The first stages in the process of development of schistosity may be inferred from examination of the sheared greywackes, semischists, slates and phyllites from the outer part of the Chlorite Zone—Subzones Chl. 1 and Chl. 2 as defined by Turner (1935) in Northwest Otago. In derivatives of greywacke the schistosity arises as a result of (1) mechanical shearing-out of clastic grains of quartz, plagioclase, pyroxene, etc., into parallel elongated lenticles and streaks of recrystallised or reconstituted material, and (2) simultaneous growth of minute crystals of chlorite, actinolite, epidote, sericite and haematite with their long axes in the plane of schistosity. At the same time the average grain-size of the rock is greatly reduced, the coase clastic quartz becoming finely recrystallised, while other clastic minerals are gradually replaced by fine-grained aggregates of reconstitution-products. Rocks of typically phyllitic appearance and structure may in some instances be the ultimate product of this breaking-down process (Turner, 1934, pp. 161, 172), the only elues as to their origin from more coarse-grained greywacke being the persistence of a few undestroyed relict grains, and their mineral composition, especially the absence or rarity of sericitic mica. This latter is, however, by no means an infallible indication of greywacke-parentage (Turner, 1933, pp. 183, 184).

Judging from the rocks on the outskirts of the Chlorite Zone, the parent greywackes were resistant massive rocks with poorly developed bedding, except on a major scale as when beds of greywacke and argillite alternate. Nevertheless, development of schistosity parallel to the directions of bedding—whenever the latter can be determined in the field—is almost universal. In rare instances the schistosity is definitely transverse to the bedding, as, for example, in certain slates from the Chl. 1 Subzone in the extreme north-west of Otago.

The changes which follow obliteration of clastic structure and establishment of schistosity are shown by the typical Chlorite Zone schists as developed in Subzones Chl. 3 and Chl. 4. Foliation parallel to the schistosity is initiated, and at the same time the average grain-size rapidly increases and the schistosity becomes even more prominent. This applies both to pelitic rocks and to' derivatives of greywacke, though in the former case the changes commence at a distinctly lower grade of metamorphism than in the latter. In Otago the average size of the quartz and albite grains in imperfectly

– 212 –

foliated schists of Subzone Chl. 3 is 0.02 mm. to 0.2 mm. for derivatives of greywacke, while corresponding figures for well-foliated schists of Chl. 4 are 0.25 mm. to 2 mm.

It is difficult to decide to what extent this rapid growth of the component grains is due to mimetic crystallisation following cessation or diminution of tectonic activity (cf. Sander, 1930, p. 172). The presence in many schists of large porphyroblasts of actinolite, hornblende, biotite, stilpnomelane, etc., dimensionally oriented with their long axes lying in any direction in the plane of schistosity appears to indicate the effectiveness of post-tectonic erystallisation in these rocks. It is especially noticeable in some of the fine-grained phyllitic rocks, such as those of the Haast Gorge (e.g., No. 1290*), in which the dimensions of the porphyroblasts may be twenty or thirty times as great as those of the grains in the enclosing ground-mass. Again, arrangement of porphyroblasts (especially of stilpnomelane minerals or amphiboles) in sheaves or coarsely radial aggregates in the schistosity-plane often affords certain evidence of post-tectonic crystallisation.

On the other hand, there are unmistakable and widespread indications that shearing often remained active during the later stages of metamorpic crystallisation. The structure may be further complicated, however, by belated minor cataclasis accompanying orogenic movements subsequent to the main metamorphism. Evidence of shearing during the final phases of the principal metamorphism is summarised below:—


In coarse albite-schists of the Chlorite Zone in many parts of Otago and South Westland the albite porphyroblasts enclose strings of inclusions which represent a schistosity developed early in the metamorphism. Typically the lines of inclusions are twisted into the well-known S-shaped form (e.g., see Turner, 1933, pp. 229, 230), which affords clear evidence of mechanical rotation of the host-crystals during growth (cf. Knopf, 1933 p. 461; Fairbairn, 1935, pp. 50, 51). As the porphyroblasts themselves appear to belong to the later stages of reconstitution, it follows that differential movement of grains must have continued till crystallisation ceased. [Of different significance are certain schists containing coarse albites which have crystallised under static conditions after cessation of movement. Thus schists from the middle part of the South Waikouaiti Valley, near Dunedin, contain abundant large porphyroblasts of albite which have grown across and sometimes almost obliterated the macroscopic schistosity. These enclose irregularly contorted lines of dark inclusions, the orientations of which seems to have been determined by pre-crystalline deformation (helicitie structure)].


Strain-slip cleavage is extremely prevalent among the schistose rocks of Otago. Its presence in fine-grained slates and phyllites is especially obvious and has been commented

[Footnote] *Specimen and section in collections of the Geology Department, University of Otago.

– 213 –

upon by earlier observers, but in the writer's experience it is equally common even in the coarse-grained schists. Typically the strain-slip appears to represent a second and less distinct cleavage formed simultaneously with the principal schistosity, or in some cases is definitely a late superposed structure; but incompletely metamorphosed grey-wackes and semischists, e.g., in the Upper Routeburn Valley, may possess a strain-slip cleavage which appears to repre-sent an early stage in the development of schistosity (cf. Knopf, 1935).


Contortion and buckling of schists subsequent to growth of sehistosity and foliation are sometimes found, as in the rocks described by Professor James Park (Park, 1906, p. 15, plates facing p. 16; 1909, pl. xxvi) from the Alexandra and Queenstown Subdivisions, and in quartz-albite-chlorite-schists collected by the writer from the far south of West-land. In such rocks close isoclinal folding combined with strain-slip resulting from rupture along the limbs of the folded laminae may give rise to structures exactly comparable with those figured by Knopf (1931, pp. 17, 18) as intermediate stages in the development of transposition-eleavage. In the low-grade poorly foliated sericite-schists derived from the Te Anau Series in the district north-west of Lake Wakatipu corrugation on a more minute scale gives rise to a type of linear schistosity (Turner, 1935, p. 342). In other rocks from the same region corrugation of the flexible micaceous foliae has been accompanied by rupture of the more brittle epidotic laminae, with subsequent crystallisation of microscopic closely spaced veinlets of quartz and amphibole along the surfaces of fracture. While it is here obvious that corrugation occurred subsequently to the growth of sehistosity and foliation, it is equally certain that the process was accompanied by some recrystallisation, especially of quartz in the contorted quartzose laminae.


In most Otago schists the individual mineral grains show at least minor effects of cataclasis following on crystallisation; common strain-effects are undulose extinction in-quartz and sericite, bending of cleavage-laminae in micas and chlorites and of twin-lamellae in calcite, and fracturing of prismatic minerals such as amphiboles, epidote minerals and tourmaline.


Segregation-veins composed chiefly of quartz and albite sometimes accompanied by epidote, chlorite or calcite are very characteristic of the schists of Otago. Though obviously of later origin than the schistosity and foliation, the veins often trend parallel to these latter structures; but veins cutting across the schistosity are almost equally common. Under the microscope the vein-minerals them selves usually show minor cataclastic effects.

– 214 –

The high-grade schists of South Westland locally show convincing mineralogical evidence of the operation of retrogressive metamorphism under the influence of earth-movements belonging to a later epoch of tectonic activity than that with which the main metamorphism is associated (Turner, 1933, pp. 249, 250).

The above summary will indicate that metamorphism of the Otago schists was not a simple process. The evolution of the present structure—and hence of the present mineral assemblages which constitute the schists—has been governed throughout by the operation of shearing-stress. The intensity of this influence has obviously fluctuated considerably during the metamorphic cycle, but it seems to have continued to operate even to the final stages of crystallisation and reconstitution. This is consistent with the conclusion reached by earlier workers that the metamorphism in Otago was essentially dynamic in character, even though it has since been proved that in Westland and Fiordland dynamic and contact influences combined to give metamorphism of a still more complex type than that experienced further east.

The formation of segregation-veins, the partial shearing of these and the development of minor strain-effects in the schists themselves may all be due to tectonic activity of a distinct and later epoch, such as is believed by the writer to be responsible for the retrogressive mineralogical changes shown by high-grade schists in South Westland. Local buckling and crumpling of schists (e.g., adjacent to relatively young faults) is also certainly a post-metamorphic effect, and must be distinguished from the extensive corrugation and small-scale isoclinal folding which have developed over wide areas during the later stages of metamorphism.

Some Petrofabric Data for New Zealand Schists.

Summary of General Orientation Rules.

The known rules for space-lattice orientation of different minerals as determined from petrofabric investigations are given by Sander (1930, pp. 173–217) and summarised by Fairbairn (1935b, pp. 34–38). For each mineral certain crystallographic planes or directions tend to lie parallel to the slip-planes (s-planes of shearing) of the deformed rock. The more important cases so far established are as follows: for calcite e (0112) or less commonly r (1011); for quartz the optic axis, which at the same time tends to lie perpendicular to B (other rules for quartz are also known); for amphiboles (100) (with the c crystallographic axis either parallel or perpendicular to B); for micas (001); for albite (010) (with the c axis either parallel or perpendicular to B). The degree of each orientation in a particular rock is indicated by the percentage of grains oriented approximately in the manner demanded by the appropriate rule for the mineral in question.

– 215 –

General Indications in New Zealand Schists.

In the absence of the necessary apparatus, statistical analysis of grain-orientation by means of the universal stage has not yet been attempted in New Zealand. However, there are plentiful indications, clearly observable with the ordinary petrographic microscope, that space-lattice orientation is almost universal in Otago schists. Further, this general evidence supports the view that schistosity has developed along shear-surfaces in these rocks.

The following data drawn from the writer's observations hold for schists in the Otago-South Westland region. Regular orientation of quartz grains is not usually obvious unless a fairly high degree is attained, in which case gypsum-plate tests on a section cut parallel to the schistosity show a preponderance of slow (Z) vibration-directions at right angles to the linear schistosity. Orientation according to one of the two brachypinacoid rules is very frequently shown by albite, so that sections perpendicular to Z (the acute bisectrix) are commonly met with in rock-sections cut parallel to this schistosity. This incidentally renders it easy to determine the composition of the plagioclase with considerable accuracy, by measurement of the extinction angle X to 001, and by comparison of α and β with ω for quartz and μ for Canada balsam. Orientation of muscovite and chlorite is always well marked, but it is usually difficult to determine whether it is based upon space-lattice or on purely dimensional rules, as in each case 001 tends to lie parallel to the schistosity. In actinolite and tremolite there is often a tendency for sections in the schistosity-plane to show low extinction-angles and interference-tints of a relatively low order, indicating orientation on one of the orthopinacoid rules. Sphene commonly attains a high degree of orientation with the acute bisectrix approximately normal to the schistosity-plane, so that the grains frequently give centred bisectrix interference-figures in rock-sections cut parallel to the schistosity. This last case would appear to be a space-lattice orientation, for the grains in question more often than not are xenoblastic rounded drops of equidimensional granules. The mechanism of orientation remains obscure. Gliding on the cleavage (110) or either of the twinning planes (100) or (221) could hardly account for the observed orientation, but gliding on (102), which though not a recognised gliding-plane is nevertheless approximately normal to the acute bisectrix Z, is possibly the correct explanation.

Partial Analysis of Two Quartz Fabrics.

Using an ordinary petrographic microscope and simple accessory apparatus (a mechanical stage and a gypsum plate), it is possible to make partial statistical analyses of grain-orientation for minerals such as quartz in which the orientation rule refers to a simple and easily determined optical direction (the optic axis). The results yielded by partial analysis are incomplete compared with the accurate details determinable by use of the universal stage, and in cases of ill-defined or complex fabric may be of little value. However, provided the fabric is relatively simple and the orientation is well-defined, data of considerable tectonic significance may be obtained by this method, which further has the advantage of rapidity.

– 216 –

Two examples are described below in detail in the hope that other investigators in New Zealand may be encouraged to carry out similar work.

(a) Quartz-haematite-schist, South Westland (No. 1317).

The specimen, a highly quartzose quartz-haematite-magnetite-schist from the Jackson Valley (see Turner, 1933, p. 221, pl. 28, fig. 12) was selected on account of its simple mineral composition, the abundance of quartz and the existence of a well-defined linear element in the schistosity (a common condition in Otago and West-land schists). Following the usual convention, the fabric is described with reference to three mutually perpendicular axes, a, b and c, as in crystallography. The ab plane is taken as parallel to the schistosity, with b parallel to the linear element; c is thus normal to the schistosity and to the linear direction therein. (It will be remembered that the b axis thus coincides with the tectonic axis of the deformed rock, i.e., with B of the strain ellipsoid, so that the α axis gives the direction of movement during deformation.)

Sections were cut parallel to ab, bc and ca respectively. In the plane ab the dominant mineral, quartz, occurs in rather large irregular nearly equidimensional grains with a tendency to elongation parallel to b; the linear element in the schistosity arises from the parallel orientation of linear streaks of iron-ore grains and of slender parallel prisms of actinolite. In bc there is similar alternation of quartzose foliae and streaks of iron-ore, but the quartzes are definitely elongated parallel to b (the trace of the schistosity-plane), and the grains of haematite are notably flattened in the same direction. In ca the sehistosity is poorly marked compared with bc; the streaks of iron-ore are somewhat discontinuous and tend to curve between the quartzose layers; the quartz grains are somewhat elongated parallel to a.

The section to be examined is held in a mechanical stage so that while rotation of the stage through any angle is permitted, movement of the section with reference to the stage is possible only in two mutually perpendicular directions, rotational differential movement being impossible. Measurements are made upon a series of quartz grains selected at random as the section is moved backward and forward across the field of view. It was found sufficient to measure between 175 and 300 grains in each section, though a larger number may be necessary for schists in which the fabric is less well defined. The angle between the trace of the optic axis of each grain and a selected line of reference (b in ab and bc, a in ca) is now measured by noting the stage-reading when the slow (Z′) direction of each grain is parallel to a selected cross-wire, and later taking a single reading when the line of reference is in a similar position.

– 217 –
Picture icon

Fig. 3.—A: Curves showing orientation of Z′ (the trace of the optic axis) for quartz grains in the three principal sections ab. bc and ac of quartz-haematite-schist No. 1317. B: Diagrammatic representation of the relation between* the axes a. b, and c and the schistosity in No. 1317.

For each section the degree of orientation of the quartz grains is brought out by plotting the percentage of quartz grains whose slow directions (Z′) fall within a given angular interval (5° to 20°), against angular distance from the line of reference. In the present instance a large interval (20°) was selected, and this was found to give smooth curves without greatly reducing the sensitiveness of

– 218 –

the method. The probable percentage of slow quartz axes to be expected within an interval of 20° in the case of totally unoriented fabrics is 20/180 × 100, i.e., 11%. Values greater than twice or less than half this figure (i.e., > 22% or < 5.5%) may therefore safely be considered as significant in interpreting the curves. The significance of values between 5.5% and 22% should be judged with more caution.

The curve for the ab plane (Fig. 3) is simple and regular, rising to a strong maximum of about 50% at a point corresponding to a direction inclined at 83° to b; i.e., 50% of the measured grains in the ab plane have their Z′ directions lying at angles between 73° and 93° to b the direction of linear schistosity. Two interpretations of this curve are possible:—


The quartz axes Z may be concentrated in a direction making an angle of 83° with b and more or less parallel to a.

Picture icon

Fig. 4.—Quartz-fabric for No. 1317 shown on stereographic projections upon the ab, bc and ac planes: a, b and c axes shown by broken lines; quartz-girdle shown by full lines; sectors of concentration of quartz axes within girdle shown in black, actual maxima m being indicated by small circles.

– 219 –

The quartz axes Z may be concentrated in a plane steeply inclined to the ab plane and intersecting the latter in a line making an angle of 83° with b.

Comparison with the curve for bc (Fig. 3) shows clearly that the second of these alternatives is the correct one, for in the bc curve there is a strong orientation of the axial traces in a direction making

Picture icon

Fig. 5.—Orientation curve for quartz grains in duplicate section cut parallel to bc plane in quartz-haematite-schist No. 1317.

– 220 –

an angle of 65° with b. As would be expected, the curve for the ac plane (Fig. 3) indicates a much wider spreading of the quartz axes than in the other two planes measured. There are nevertheless two distinct subequal maxima corresponding to directions respecively parallel to a and inclined at 15° to c.

The orientation of the plane of concentration of quartz axes, i.e., the quartz girdle, is shown in stereograms (Fig. 4, A, B) constructed from the curves for the ab and bc planes. Since the girdle is not exactly perpendicular to either of these planes, the curves are asymmetrical, the degree of asymmetry being greater for the ab curve than for bc. The former curve is steeper in the obtuse angle between b and the girdle trace than in the supplementary acute angle (see Fig. 3) from which it may be deduced that the girdle is inclined as shown in Fig. 4A. Similarly the orientation of the girdle as depicted in Fig. 4B corresponds to the asymmetry shown in the bc curve. In Fig. 4C the position of the girdle is as deduced from the two previous figures, while the maxima within the girdle are plotted from the ac curve.

To check the accuracy of the method, a second section was cut approximately parallel to bc and the orientation of the grains was measured as before. The resulting curve Fig. 5 agrees strikingly with the corresponding curve Fig. 4B, even to the small maximum rising between the 90° and 130° points. Attention is drawn to the fact that in the first curve the maximum is located at 65° from b while in the curve for the check section it lies at 73° from b. This discrepancy may be partly due to lack of sensitiveness in the method employed, but more likely is to be attributed to inaccuracy of orientation in cutting the second section. The second figure (73°) has therefore been discarded in favour of the value (65°) indicated by the first curve.

The quartz fabric described above is interpreted as follows:—


The pattern is that of a B-tectonite in that the optic axes (Z) of the quartz grains tend to lie in a girdle, rather than to be concentrated parallel to a single direction as in S-tectonites (cf. Knopf, 1933, p. 454).


The maximum parallel to a, as shown by the curve for the ac plane, is a feature frequently observed in tectonites of both classes, and indicates that the schistosity-plane ab is a plane of shearing.


Whereas in typical B-tectonites the quartz girdle is normal to the b axis, in the rock here described it is inclined at only about 65° to b. The orientation is as if a girdle originally approximately perpendicular to b had been rotated through an angle of 25° about a, so that a triclinic symmetry is initiated in place of the monoclinie symmetry typical of the normal B-tectonite pattern. Trielinic orientation-patterns of this type are well known, and are interpreted as resulting from crossed strain involving external rotation of particles about a as well as b (Knopf, 1933, pp. 457, 458; Fairbairn, 1935b, pp. 81–87). Without discussing the matter at length it may be stated briefly

– 221 –

that the crossed strain may have alternated rythmically with the principal strain as a result of fluctuating conditions of stress during a single deformation; or alternately it may have been superposed upon the principal strain during a second deformation. On this second assumption the line chosen as the b axis coincides with B of the strain ellipsoid for the initial and most important phase of deformation in which the most obvious features of the fabric for the rock as a whole were determined. Subsequently, however, during a later deformation for which the mean axis B′ of the new ellipsoid was inclined to the original B, the quartz fabric has been partially readjusted to new conditions, so that the quartz girdle is now distinctly oblique to b (B). This latter possibility is in keeping with the writer's conception of the metamorphic history of the Otago schists.


The direction of linear schistosity b, clearly established without statistical analysis of the grain-fabric, indicates the orientation of the tectonic axis corresponding to the first and principal phase of deformation. Movement during deformation has been in the s-planes, at right angles to b. Since the rock in question was not collected specially for petrofabric analysis its field orientation is unknown. However, for specimens which have been marked in the field before being broken from the outerop, determination of b is of great importance, since it not only gives the direction of the tectonic axis but also makes it possible to say whether the b (B) axis is of the subhorizontal or the steeply-dipping type as recognised by Sander (1934, pp, 43, 44).

(b) Quartz-piedmontite-schist, Shotover Valley, Western Otago (No. 1489).

The rock is a highly schistose rather coarse-grained rock consisting of quartz, muscovite, piedmontite and subordinate garnet. The grains of quartz are flattened parallel to the schistosity plane ab but are distinctly elongated parallel to the direction of linear schistosity b. The muscovite is in coarse flakes oriented parallel to ab, this parallelism being more marked in the bc than in the ac section. The linear element in the schistosity is due largely to approximate parallelism of the slender piedmontite crystals in this direction, though their exact orientation varies through about 10°. Dense clouds of minute dusty inclusions are developed within the quartz grains on surfaces which appear to lie approximately parallel to the bc plane of the rock; in sections parallel to ac and ab they are therefore arranged as grey lines more or less parallel to c and b respectively, but in the bc section they merely give rise to indefinite clouding of the grains.

– 222 –
Picture icon

Fig. 6.—Quartz fabric in quartz piedmontite-schist No. 1489: A, B, C show stereographic representation of fabric (cf. Fig. 4); D indicates position of a, b and c axes with reference to schistosity; E shows orientation curves for-quartz-grains in sections ab, bc and ac (cf. Fig. 3).

The three curves for orientation of Z′ in quartz grains in sections cut parallel to ab, bc and ac are shown in Fig. 6E. As in the case of the South Westland rock previously described, the ab and bc curves indicate the existence of a girdle inclined at a high angle (about 70°) to b. Here also the trace of the quartz girdle on the ab plane makes an angle of about 85° with b. However, the ac curves for the two rocks show important differences as well as similarities. While the maximum subparallel to c is well marked in the piedmontite-schist, there is a second maximum in the ac plane corresponding with a direction about halfway between a and c, while the direction of the a axis is marked by a strong minimum. These features are borne out by the relative heights of the maxima in the ab and bc curves.

– 223 –

The triclinic symmetry of the fabric pattern is thus more pronounced than in the previous example. The development of a minimum for a in addition to that parallel to b is specially significant and indicates that rotation about a played an important part in the deformation. The same general type of complex deformation appears to have been responsible for the development of the two fabrics considered above. But the influence of the crossed strain has been more pronounced in the case of No. 1489 than in the South Westland rock No. 1317.

Petrofabric Evidence as to Cause of Metamorphism.

The hypothesis of crystallisation according to Riecke's principle under the direct influence of the vertically directed pressure of overlying rocks (load metamorphism) has been advanced by Daly to explain the regional development of schists with subhorizontal or uniformly tilted schistosity. The schists of Central Otago, as exposed, for example, between Alexandra and Roxburgh on the one hand or Ida Valley on the other, exemplify this condition well, though in some districts such as north-western portions of the province, the dip approaches the vertical over wide areas. In examining a case of supposed load metamorphism in North America Gilully (1934) recently pointed out that the fabric in such rock. should be symmetrical about a line perpendicular to the schistosity. Most Otago schists, however, show a pronounced linear element in the schistosity suggestive of symmetry about a line lying within or inclined at a low angle to the schistosity. The statistical investigation described above bears this out for the two specimens concerned, and further shows conclusively that the rocks in question are tectonites whose rather complex fabrics have been determined by shearing. The purely petrographic evidence summarised in a previous section points in the same direction. The possibility of load metamorphism being responsible for the present condition of the schists of Otago and Westland must therefore be excluded.

The meagre petrofabric data so far collected are consistent with the hypothesis that metamorphism accompanied deformation governed by low-angle thrusting (cf. Benson, 1921, p. 30), and was followed by minor deformation during a subsequent tectonic cycle. If the schistosity is regarded as having developed by shearing parallel to the original bedding—and to the writer this appears to be the most probable explanation—the hypothesis of metamorphism controlled by low-angle thrusting is further strengthened.


The writer wishes to record his indebtedness to Mrs B. Monheimer, of Dunedin, for assistance in translating German literature on petrofabric analysis.

Literature Cited

Beokek, G. F., 1893. Finite homogeneous Strain Flow and Rupture of Rocks, Bull. Geol. Soc. Am., vol 4.

——1904. Experiments on Schistosity and Slaty Cleavage, U.S. Geol. Surv. Bull, no. 241.

——1907. Current Theories of Slaty Cleavage, Am. Journ. Sci., ser. 4, vol. xxiv, pp. 1–17.

– 224 –

Benson, W. N., 1921. Recent Advances in New Zealand Geology, Rept. Austr. A88. Adv. Sci., Section C.

Fairbairn, H. W., 1935. Notes on the Mechanics of Rock Foliation, Journal Geol., vol. xliii, no. 6, pp. 591–608.

——1935a. Petrofabric Analysis and some possible Applications, Canadian Mining Journal, July, 1935.

——1935(b). Introduction to Petrofabrio Analysis, Dept. of Geology, Queen's University, Kingston, Canada.

Gilluly, J., 1934. Mineral Orientation in some Rocks of the Shuswap Terrane as a Clue to their Metamorphism, Am. Jour. Sci., vol. xxviii, pp. 182–201.

Griggs, D. T., 1935. The Strain Ellipsoid as a Theory of Rupture, Am. Jour. Sci., vol. xxx, pp. 121–137.

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

Hutton, C. O. and Turner, F. J., 1936. Metamorphic Zones in North-west Otago, Trans. Roy. Soc. N.Z., vol. 65, pt. 4, pp. 405, 406.

Knopf, E. B., 1931. Retrogressive Metamorphism and Phyllonitisation, Am. Jour. Sci., vol. xxi, pp. 1–27.

——1933. Petrotectonics, Am. Jour. Sci., vol. xxv, pp. 433–470.

——1935. Recognition of Overthrusts in Metamorphic Terranes, Am-Jour. Sci., vol. xxx, pp. 198–209.

Leith, C. K., 1905. Rock Cleavage, U.S. Geol. Surv. Bull., no. 239.

Leith, C. K., and Mead, W. J., 1915, Metamorphic Geology, New York, H. Holt.

Park, J., 1906. The Geology of the area covered by the Alexandra Sheet, N.Z. Geol. Surv. Bull., no. 2.

——1909. The Geology of the Queenstown Subdivision. N.Z. Geol. Surv. Bull, no. 7.

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

——1934. Petrofabries (Gefügekunde der Gesteine) and Orogenesis, Am. Jour. Sci., vol. xxviii, pp. 37–50.

—— 1934 (a). Typisierung von deformierten Tonschiefern mit optischen und röntgenoptischen Mitteln, Zeits. Krist., (A), 89, pp. 97–124.

Turner, F. J., 1933. The Metamorphic and Intrusive Rocks of Southern West-land, Trans. N.Z. Inst., vol. 63, pp. 178–284.

——1934. Schists from the Forbes Range and Adjacent Country, Western Otago, Trans Roy. Soc. N.Z., vol. 64, pp. 161–174.

——1935. Metamorphism of the Te Anau Series in the Region North-west of Lake Wakatipu, Trans Roy. Soc. N.Z., vol. 65, pt. 3, pp. 329–349.

Tyrrell, G. W., 1930. The Principles of Petrology, 2nd ed., London, Methuen and Co.