Current Views on the Origin and Tectonic Significance of Schistosity.

Transactions and Proceedings of the Royal Society of New Zealand, Volume 72, 1942-43, Page 119

Current Views on the Origin and Tectonic Significance of Schistosity. By Francis J. Turner University of Otago. [Read before the Otago Branch, October 14, 1941; received by the Editor, October 17, 1941; issued separately, September, 1942.] Introduction. Current views on schistosity cannot be summarised without discussion of older hypotheses embodied in classical papers, selected examples of which are therefore cited. The number of recent publications containing statements that bear upon the subject is so large that only a limited number dealing with the more important aspects of the subject are referred to. Discussion is confined to the following topics:— (a) Definition and nature of schistosity. (b) Processes contributing to the development of schistosity. (c) Relation of schistosity to deforming movements and forces. Definition and Nature of Schistosity. There is considerable confusion in the prevalent usage of terms such as schistosity, foliation, flow cleavage, fracture cleavage, etc., and much of this confusion arises from employing genetic terms (e.g., flow cleavage) for structures whose origin cannot usually be demonstrated with certainty from ordinary field data and petrographic observations. Sander (1930, pp. 97, 98, 99) recommends the employment of a purely descriptive non-genetic term to cover the parallel fabrics of metamorphic rocks; he includes all such structures (“sets of planes of mechanical inhomogeneity”) in his term “s-surfaces” (“s-planes”). The present writer (Turner, 1936, pp. 202, 203) has defined schistosity, as far as possible in non-genetic terms, 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 under the influence of stress or high temperature.” This includes only metamorphic structures (or pre-metamorphic structures accentuated by metamorphism) and excludes parallel fabrics developed by flow in partially molten masses. It is emphasised at the outset that schistosity may originate in more than one way, and that observations in the field, on the hand-specimen and with ordinary petrographic methods are frequently insufficient to diagnose the mode of origin for any particular case. It is here that the results of fabric analysis, interpreted kinematically in conjunction with the other structural data, are of greatest assistance.

The structural features that may be associated with schistosity are summarised briefly thus:— (1) The most obvious and widely developed character of schistose rocks is a tendency for platy or prismatic crystals of minerals such as micas, chlorites, amphiboles and epidotes to show dimensional parallelism. (2) There is also a tendency in most schistose rocks for crystal cleavages to lie parallel to the schistosity (as in micas and chlorites) or to a line in the schistosity (as in amphiboles). C. K. Leith (1913, pp. 78, 79) discusses this point and concludes that “the dimensional parallelism of mineral particles is the controlling factor in rock cleavage; (and) that to this control is due the mutual parallelism of mineral cleavages of mica or hornblende cleavages.” [This conclusion must be modified in the light of petrofabric evidence—see below under (4).] (3) J. Tyndall (1856, pp. 44, 45) experimentally produced perfect cleavage in wax by subjecting it to pressure, and concluded from this that schistosity is independent of preferred dimensional orientation of the mineral grains. [Sorby (1856), however, considered that parallelism of acicular crystals of wax was responsible for the cleavage produced by Tyndall.] This brings out a further possibility, namely, that schistosity may sometimes be determined not so much by the presence of dimensionally parallel crystals as by the development of planes of mechanical weakness during deformation. According to G. F. Becker's theory schistosity (or slaty cleavage) “is due to a weakening of cohesion along planes of maximum tangential strain” developed during deformation (Becker, 1904, p. 11). He further implied that the cohesion referred to was at least partly intermolecular (loc. cit., p. 22). Daubrée from experimental work on the production of schistosity had come to similar conclusions many years previously. (4) Sander has shown that in almost all schistose rocks there is a greater or less degree of preferred crystallographic orientation of the component mineral grains, even for minerals such as quartz, calcite or feldspar which usually occur in equidimensional grains and therefore lack dimensional orientation. Orientation of grains according to the space-lattice structure would thus appear to be more important than orientation according to crystal form. (5) Schistose rocks frequently cleave along surfaces of rupture. The term fracture cleavage was used by Leith to cover such structures (e.g. Leith, 1913, pp. 61–64) which have been termed false cleavage, strain-slip cleavage, etc., by various other writers (cf. Harker, 1932, pp. 157, 158). Such surfaces of rupture may originate in more than one way and may later be modified or intensified by post-tectonic crystallisation of new mineral grains. In this modified state they may be indistinguishable from schistosity of the type described by Leith as flow cleavage (i.e. schistosity resulting from plastic flow of solid rock). It has been stated (e.g. Swanson, 1941, pp. 1247, 1252) that rocks with “flow cleavage” can part along any plane parallel to the schistosity, whereas “fracture cleavage” can occur only upon a limited number of surfaces of mechanical weakness (slip- or rupture-surfaces). In the writer's opinion this distinction is invalid. Even

when preferred orientation is strong, crystallographic parallelism of the component grains is merely approached, and so any one surface of “flow cleavage” is determined in detail by the size, shape, orientation, and mechanical properties of the crystals (usually of several minerals) encountered. In a coarse rock with preferred orientation of a low order, the planes of “flow cleavage” may be more widely spaced than the “fracture cleavage” surfaces of a fine slate. (6) When schistosity has developed during metamorphism at high temperatures, or when crystallisation accompanying low-grade metamorphism has been sufficiently prolonged, a laminated structure [foliation, according to the usage of Harker (1932, p. 203)] involving concentration of particular mineral assemblages in alternating layers parallel to the schistosity has commonly been produced as a result of metamorphic differentiation. On the other hand this is far from being a universal feature of schists of low metamorphic grade even when schistosity is very perfectly developed. The origin of this laminated condition and its relation to schistosity and bedding have recently been discussed by the writer (Turner, 1941). The writer does not agree with the statement of Swanson (1941, p. 1246) that laminated structure of this sort is typically absent from rocks having “flow cleavage.” Schistosity as a Product of Mechanical and Crystalloblastic Processes. The processes concerned in development of schistosity have been classed in three groups (cf. Grubenmann and Niggli, 1924, pp. 234, 235; Tyrrell, 1930, pp. 303, 304):— (a) Cataclastic (rupture, rotation and differential movement of rock components). (b) Plastic (deformation and flow of crystals or rocks without destruction of their continuity). (c) Crystalloblastic (growth of crystals). The first two groups are essentially mechanical (kinematic) processes, while crystalloblastic growth of crystals by contrast is of a chemical nature. E. B. Knopf (1931, p. 17) has suggested on this basis that the terms cataclastic and crystalloblastic cleavage (i.e., schistosity) might prove useful in distinguishing between parallel fabrics dominated by mechanical processes and crystalloblastic growth respectively. Chemical reconstitution and recrystallisation of the rock-forming minerals, when active, will play an important part in determining the nature of the ultimate fabric of the rock. Sander (1930, pp. 262, 263; Knopf, 1933, pp. 460–462) recognises three possibilities as to the relation between crystallisation and deformation. (a) Mechanical deformation alone; any recognisable crystallisation is pretectonic (prekinematic). (b) Mechanical deformation, accompanied (and assisted) by growth of crystals; crystallisation is paratectonic (parakinematic). (c) Mechanical deformation, followed by growth of crystals; crystallisation is post-tectonic (postkinematic).

Sander draws attention to the difficulty of distinguishing between fabrics resulting from paratectonic crystallisation that has continued slightly after cessation of deformation, and those governed by purely post-tectonic crystallisation. Furthermore the sequence of events in deformational metamorphism may include alternating phases dominated now by mechanical processes, now by crystal growth (cf. Turner, 1940, pp. 188–190). Again deformation may be accomplished largely by recrystallisation alone. It is thus clear that it will be impossible to classify all examples of schistosity as either cataclastic or crystallo-blastic, even if these terms are used in a broad sense. While it is recognised that several or all of the processes cited above may take part in a given instance of metamorphism, it is nevertheless helpful to consider each of the schist-forming processes separately and to note the importance attached to each in current theories of schistosity. The aspects considered below are as follows:— (1) Development of slip-surfaces in rocks during plastic deformation. (2) Ruptural deformation of rocks in relation to schistosity. (3) Rotation of rigid grains in a plastically flowing matrix. (4) Plastic deformation of individual grains. (5) Rupture of grains. (6) Paratectonic crystallisation as a factor in rock deformation. (7) Post-tectonic crystallisation and its effects upon the rock fabric. Development of Slip-surfaces During Plastic Deformation of a Rock. The condition necessary for plastic deformation of materials such as rocks which possess a high breaking strength is high confining pressure such as obtains at depths of a few miles within the earth's crust. Such conditions have been artificially reproduced in the laboratory, and an interesting series of investigations upon the plastic deformation of rocks and minerals has been commenced by D. Griggs (1936). G. F. Becker (1893, 1904, 1907), from a mathematical analysis of stress and strain in a homogeneous body subjected to plastic deformation without change in volume, concluded that change in shape of the mass is accomplished by differential slip along two sets of planes of maximum slide, with which he identified the schistosity of naturally deformed rocks. He further correlated the planes of maximum slide with the circular sections of the strain ellipsoid (the ellipsoidal form assumed by an initially spherical body during deformation). He showed, too, that whereas differential movement would be active on both sets of slip-surfaces if deformation were brought about by simple compression (as in a vice), it would probably be restricted to one set of slip-surfaces in the much more frequent case of rotational strain (as when a plastic mass on a rigid surface is deformed by an obliquely directed force).* The reader unfamiliar with the terminology and principles of deformation is referred to E. S. Hills (1940), pp. 20–43.

Certain details of Becker's theory have been severely criticised: there is obviously a considerable difference between the homogeneous body considered by Becker and the heterogeneous mass which constitutes a rock; further, certain fundamental assumptions made in applying the strain ellipsoid to the case in question have been shown to be incorrect (e.g., Griggs, 1935). Nevertheless Becker's results, arrived at from examination of an ideal simple case, can profitably be applied (though not in a mathematically exact sense) to the complex case of rock deformation. “The most notable feature of Becker's 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. This fundamental concept retains its value in spite of the fact that experimental work has indicated that the planes of maximum shearing stress in elastic strain are not the planes upon which slip takes place above the elastic yield point.” (Knopf, 1933, p. 447). Sander and Schmidt have reached conclusions that are in general agreement with Becker's thesis from a different line of approach, namely, a kinematic interpretation of the structures of deformed rocks with reference in particular to the orientation of mineral grains by space-lattice structure as revealed by petrofabric analysis. They attach great importance to Becker's views and emphasise the role of slip movements on surfaces of shearing (a type of s-surface) which, like Becker, they correlate in many cases with the schistosity of the deformed rock (e.g. Sander, 1930, pp. 97–101, 218–219). In two fundamental respects, however, they differ from Becker:— (1) According to Sander and Schmidt, schistosity originating parallel to shear-planes, though very important, is not the only type of schistosity commonly occurring in deformed rocks. (2) When a relatively homogeneous (mechanically isotropic) rock such as a shale is plastically deformed, the slip-planes that develop have a purely mechanical origin, much as pictured by Becker, i.e., the rock yields along surfaces of maximum shearing-stress. Much slaty cleavage transverse to the bedding originates thus. More usually, however, the rock is not originally homogeneous, and prior to deformation already possesses surfaces of weak cohesion such as planes of bedding, flow, foliation or preexisting schistosity. Under these circumstances slip will tend to occur along the mechanically weak s-surfaces already present, rather than on planes of higher shearing-stress for which the strength of the rock is also greater. This is supported by commonly observed (but by no means universal) coincidence between bedding and schistosity. The writer elsewhere has given a more detailed summary of the views of Becker and of Sander and Schmidt on the relation between schistosity and slip movements accompanying deformation (Turner, 1936, pp. 202, 210). Ruptural Deformation of Bocks in Relation to Schistosity. At atmospheric pressures rocks and rock-forming minerals are brittle; under increasing stress (e.g., in a simple compression) they deform elastically until the elastic limit is reached, after which they

fail by rupture. Under high confining pressures, however, rocks deform plastically after the elastic limit is exceeded. That is, the shape of the mass is permanently altered without loss of its essential continuity. Nevertheless, plastic deformation of rocks, even under these conditions, will not proceed indefinitely but ultimately ends in rupture (cf. Griggs, 1936). Therefore the generally recognised distinction between zones of fracture and flowage in the earth's crust is valid in only a broad sense, and it must be understood that fracture can be and often is associated with plastic deformation in the zone of rock flow (cf. Sander, 1930, p. 91). “Earlier it was generally assumed that ruptures of any kind would have been formed during a late phase of mountain-building, after plastic deformation, as when the rocks no longer had the same plasticity as during the principal deformation of the orogeny. But according to the results of more recent investigations upon fabric, it appears still more likely that cracks and joints can originate at any time during the mechanical deformation and in conjunction with the crystalloblastic process.” Eskola (1939, p. 310). Local rupturing of individual crystals, though common enough in plastic deformation of rocks, will not be considered here, but only rupture on a larger scale independent of the orientation of the grains concerned and sufficient to affect the continuity of the rock. These fractures are of two kinds, viz., shear-cracks and tension-joints. Their relation to the fabric of the deformed rock and to the deforming forces has been discussed by Sander (1930, pp. 91–97) Eskola (1939, pp. 308–312) and others. The following essential points may be mentioned here:— (1) The commonest shear-cracks are those parallel to the tectonic axis of the deforming movement (= b axis of the fabric) and hence denoted by the general symbol (hOl) with reference to the fabric axes.* The fabric axes a b and c are three mutually perpendicular lines of reference employed in describing features of a rock fabric. The c axis is normal to the plane of most conspicuous schistosity (ab), while b is the direction of lineation in the schistosity surface. It is parallel to the axis of folds developed simultaneously with schistosity. These (hOl) cracks may be inclined to the prevailing schistosity (ab) at any angle, but angles of less than 45° are usual, while in the case of steeply dipping schistosity in fold-mountain ranges (schistosity due to “flattening”) the shear cracks may often be approximately parallel to the schistosity itself (cf. Eskola, 1939, p. 309). When several sets of shear-cracks intersect in b the rock tends to split into elongated rod-like fragments parallel to b (pencil structure). Closely spaced (hOl) cracks giving rise to a “fracture cleavage”) that intersects the schistosity (“slaty cleavage”) obliquely have long been recognised as commonly occurring in slates. They have been attributed by many writers (e.g., Leith, 1905; Harker, 1932, pp. 158–159) to superposition of a second deformation upon that responsible for the main schistosity. Becker (1907) on the other hand offered an ingenious explanation for the simultaneous development of (a) schistosity resulting from plastic flow on one set of planes of maximum shearing stress and (b) fracture cleavage originating by rupture on the second set, during deformation in-

volving rotational strain. Modern students of rock fabric (e.g., Sander, 1930, pp. 100–103, and Eskola, 1939, p. 288) attach great importance to Becker's theory. (2) Tension-cracks genetically connected with schistosity are usually perpendicular both to the schistosity plane (ab) and the linear structure (b). These ac joints (cross-joints or Q-joints) range from microscopic cracks that may be restricted to the more brittle layers (e.g., Sander, 1930, Fig. 104), to large-scale fractures that may give rise to conspicuous surface features such as the rectangular “schist tors” of the schist areas of Central Otago. Sander (1930, p. 219) cites instances when closely spaced ac tension-cracks impart to the rock a fissility that can itself be classed as schistosity (cf. also Eskola, 1939, p. 309). (3) Other sets of fractures than the (hOl) and ac cracks described above are well-known in schistose rocks and may be due to either tensional or shearing stress. They cannot themselves give rise to a schistose structure and therefore are outside the scope of this paper. (4) Of widespread occurrence, especially in laminated rocks, is the type of schistosity termed strain-slip cleavage, fracture cleavage, etc., which originates by complete or incipient rupture of the rock along closely spaced parallel planes (e.g., see Leith, 1913, pp. 61–64). The evolution of schistosity (Umfaltungsclivage of Sander) by small-scale close folding in laminated rocks followed by rupture along the stretched folds was described by Heim in his classic work on the Tessin gneiss of Switzerland, and has recently been discussed in several easily accessible papers by Mrs E. B. Knopf (Knopf, 1931, pp. 14–20; 1935, pp. 204–207; 1938, pp. 189, 190, pl. 19). Elsewhere (Turner, 1936, p. 208) the present writer has summarised the process thus:— “In markedly anisotropic rocks the slip-planes typically are preexisting 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 or alternating beds [or laminae] of variable competency, slipping along the intervening s-planes is often accompanied by development of flexures in the competent beds [or layers]. The incompetent beds simultaneously yield by slipping, giving rise to minor drag-folds which are 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 may be termed ‘transposition cleavage’ [Umfaltungsclivage] in contrast with the ‘slaty cleavage’ which cuts across the original s-planes…. 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 isoclinal folding of the slaty cleavage, giving first strain-slip cleavage and perhaps ultimately a new

schistosity of the ‘transposition cleavage’ type. In slaty rocks of initially relatively isotropic character the development of strain-slip cleavage therefore is characteristic of the late stages of deformation, or may indicate repeated metamorphism (Knopf, 1935.)” Rotation of Rigid Grains in a Plastically Flowing Matrix. In 1853 Sorby, from his early microscopic studies of slate and artificially compressed mixtures of clay and haematite, attributed the observed dimensional parallelism of the mineral flakes to mechanical rotation by which their longest dimensions were brought into a plane (the schistosity or cleavage) perpendicular to the compressing force. The geologists of the Wisconsin School (e.g., Leith and Mead, 1915, pp. 173–176) included mechanical rotation of the original grains and ruptured fragments, together with recrystallisation and growth of new grains, as the main processes concerned in rock flowage by which schistosity was presumed to originate. In fact rotation as pictured by Sorby has long been generally accepted as one of the mechanisms by which prismatic and platy minerals may achieve preferred dimensional orientation in metamorphic rocks. It has been discussed from the view-point of fabric analysis by Sander (e.g.; 1930, 149–150), Knopf (1938, pp. 133–135) and Eskola (1939, p. 291) who agree that rigid inequidimensional components of a rock may become oriented according to their external form during deformation of the rock. Where the components in question are crystals with a pronounced prismatic or tabular crystallographic habit, the resultant preferred orientation will be according to crystal space-lattice as well as external form. Sander (1930, p. 150, Fig. 59) has figured an example of preferred orientation of small, rigid rods (with their long axes in the AB plane of the strain ellipsoid) in plasteline, induced by artificial compression and consequent plastic flow of the heterogeneous mixture. He also records instances where elongated grains of quartz, enclosed by calcite that has flowed plastically during deformation of the rock, have developed preferred orientation with their long axes (= c crystal axis) subparallel to the B axis of the fabric, i.e., to the axis of rotation (Sander, 1930, p. 201, D 175, 186). The following remarks by Mrs Knopf (Knopf, 1938, pp. 134, 135) may be cited in connection with the role of mechanical rotation in contributing to dimensional parallelism of crystals. “It is possible that this orienting by grain shape has played a larger part in the development of preferred orientation than has hitherto been suspected…. Sander has always made a clear distinction between the orienting mechanisms of gliding and those of movement of rigid bodies in response to flow. But he has always stressed the fact that a preferred orientation that is now recognised by the position of the crystal lattice in the grains where there is no orienting by shape may, nevertheless, have been assumed in a stage of seed orienting when the shape of the minute grains determined their position. Later grain growth destroyed the fabric habit of the smaller grains and thus destroyed the evidence of any relation between shape and position, but preserved the preferred orientation of the crystal lattice.”

It is probable nevertheless that dimensional orientation of crystals in the AB plane of the strain ellipsoid as a result of mechanical rotation alone would not reach a high degree of perfection unless deformation of the rock concerned was extreme. G. F. Becker (1904, p. 11) long ago considered mathematically the case of preferred orientation induced in mica scales originally oriented at random in an enclosing mass of mud subject to compression. For the scales to attain an average inclination of 2° to a given plane (preferred orientation of a high order) it would be necessary to compress the whole mass until it was reduced to 5½% of the initial thickness. So far the discussion has concerned only rotation of rigid in-equidimensional crystals or aggregates of grains in a plastically deformed matrix. Petrofabric analysis of deformed rocks shows that grains of a given mineral very commonly tend to be oriented, regardless of their shape, with some mechanically significant crystallographic line or pole (e.g., the optic axis in quartz, the pole of 001 in micas, the pole of the rhombohedral twin lamellae in calcite) in a plane perpendicular to a line (the b axis) in the fabric of the rock. The corresponding fabric diagram shows a girdle of concentration of the poles in question, with b as the girdle axis. Orientation patterns of this type have been explained as resulting in many cases from a complex process of intergranular movement and grain deformation, in which rotation of the individual grains about b brings them into a position favouring deformation by such means as intragranular slip. W. Schmidt (1932, pp. 172–175) has discussed this aspect of the case in some detail, and points out that perfect girdle fabrics may develop as a result of deformation involving but slight rotation of grains. Nor is this by any means the only mechanism by which girdle fabrics may originate (cf. Knopf, 1938, 153–155). According to Schmidt's theory, girdle fabrics with space-lattice orientation controlled in part by rotation of grains can be produced only if the minerals concerned are mechanically anisotropic, i.e., there must be crystallographic planes and lines of maximum ability to deform by gliding. He speaks of minerals such as garnet, which possess no such planes or directions as being “condemned to continual rotation” (loc. cit., p. 172), as evidenced by the well-known spiral arrangement of inclusions in “snowball” porphyroblasts of garnet. Plastic Deformation of Rock Components. As early as 1849 Sharpe suggested that flattening of the component grains at right angles to a compressive force was responsible for the parallel fabric of slates (cf. Harker, 1932, p. 154). Geologists of the Wisconsin School (e.g., Leith and Mead, 1915, pp. 177, 178) also attached some importance to mechanical elongation of rock components parallel to the schistosity (as evidenced by deformed crystals, pebbles and fossils). Examples of undoubted flattening, bending and general distortion of grains have been described and figured in many subsequent works on “dynamic” metamorphism (e.g. Sander, 1930, Fig. 66; Harker, 1932, Fig. 184 B).

During the last dozen years the role of deformation of crystals in metamorphism has been critically examined in the light of a mass of data accruing from petrofabric analysis of naturally deformed rocks and from experimental investigation of the behaviour of rocks in the laboratory under high shearing stress and confining pressures. The literature is too extensive to be reviewed in this account,* The reader will find numerous references in the accounts by Sander (1930, pp. 147–156) and Knopf (1938, especially pp. 86–99, 163–178) and in Eskola's recent critical summary (Eskola, 1939, pp. 297–308). but some conclusions regarding mechanism of grain deformation are summarised below; others are discussed in a later section dealing with the relation of schistosity to the directions of deforming forces and movements. When a rock is being deformed its component grains are subjected to differential stresses the intensity of which varies in different planes and directions within each grain, as determined by the position of the grain in relation to the stress system affecting the rock body. A grain may react to these imposed stresses in one of two ways, according to whether the stress exceeds on the one hand the frictional resistance to intergranular movement along boundaries between adjacent grains, or on the other the elastic limit of the mineral for a particular crystallographic direction. In the former case intergranular rotation is possible. In the latter case the crystal will deform by gliding parallel to that crystallographic plane and direction for which the elastic limit has been exceeded. Most rock-forming minerals possess crystallographic planes and directions of potential gliding [whether by twinning or simple translation (cf. Knopf, 1938, pp. 88, 89)], and when these coincide approximately with the planes of maximum shearing-stress for the deforming rock body, conditions most favourable to grain deformation by gliding are realised. Neglecting the complicating factors discussed by Schmidt and Eskola in the works cited above, it will readily be understood that intergranular rotation, combined with gliding of grains that have attained a favourable orientation, might give rise to a high degree of preferred orientation according to space-lattice structure for the mineral concerned. The resultant schistosity (or schistosities) might be parallel to the planes of slip in which the crystal glide-planes have become oriented throughout the rock, or it might be governed by a parallel dimensional orientation of the elongated grains (not necessarily parallel to the slip-planes of the rock-body) produced as a secondary result of grain deformation (see later, p. 137). A programme of experimental work on deformation of rocks and crystals at high confining pressures was initiated a few years ago by D. Griggs and important results have already come to hand (Griggs, 1936, 1940; Griggs and Bell, 1938). Marble cylinders, jacketed to exclude water from the intergranular pores, were compressed under confining pressures of 8,000–10,000 atmospheres with resultant shortening of 4–25% (Griggs, 1936, pp. 566–569). Micro-sections of the deformed rock showed no effects of rupture nor any marked change in size or shape of grain. Lamellar twinning on

(10·12) was developed to a very high degree showing that flow was accomplished by twin-gliding. Furthermore, the component grains had reached a high degree of preferred orientation according to space-lattice structure, with the twinning planes inclined at high angles to the direction of the compressing force—a pattern familiar enough in fabric diagrams of rocks that have been deformed by Sander's process of flattening (see below). Nevertheless Griggs (1940) urges exercise of great caution in applying the results of high-pressure experiments to the interpretation of natural rock fabrics. The high confining pressures employed greatly increased the strength of the materials tested; in one case a five-fold increase of strength was reported for a confining pressure equivalent to a depth of 22 miles in the earth's crust. It is thought very unlikely that differential stresses sufficient to overcome this increased strength are in fact developed within the earth's crust. Experiments on single crystals of quartz failed, even at the highest pressures available, to produce plastic flow by translation or twinning (Griggs and Bell, 1938). Further experiments are being carried out to determine the effects of time, temperature and the presence of solutions in the intergranular pores in deformation by flow. The first results available (Griggs, 1940) indicate that the presence of solutions is of great importance in reducing the strength of marble. Test pieces immersed in carbonated water were found to flow readily at stresses of one-fifth the magnitude recorded for the previous experiments in the absence of water. Deformation was accomplished mainly by recrystallisation, however, and no marked effects of twin-gliding were recorded. Rupture of Grains. When a rock is deformed beyond the elastic limit under low confining pressures and if recrystallisation is unimportant, catalastic rupture and breaking down of the original grains is the mechanism involved. Neither schistosity nor any type of preferred orientation of grains need result from pure cataclasis, though actually simultaneous rolling out and rotation of the fractured material usually gives rise to a preferred orientation governed by external form of the fragments (cf. Harker, 1932, pp. 168, 169). The reader is also reminded that rupture of individual grains also may accompany plastic deformation of the rock as a whole, for “although the individual elements may be ruptured, the rock itself retains its continuity in space, because indirectly related crystallising movements either work pari passu with the ruptural or plastic deformation, or follow the rupture before the continuity is lost” (Knopf, 1938, p. 38). Bearing in mind this complex interplay between rupture, plastic deformation and crystallisation of the rock components during deformation which may be termed plastic with reference to the rock-mass as a whole, it will be readily understood that fracturing of grains may possibly play an important part in the evolution of preferred orientation of grains according to space-lattice structure,

provided that the form of the fragments produced is related to the space-lattice. Recent petrofabric studies and the results of high-pressure deformation of quartz crystals while not conclusive nevertheless suggest strongly that the preferred orientation of quartz in deformed rocks may well be controlled by fracturing of crystals along planes of weakness in the lattice. The most significant conclusions arrived at with regard to quartz are as follows:— (1) Sander (1930, pp. 173–178) believes that the predominating process of orientation of quartz in deformed rocks is one by which the grains of quartz break down by fracture into needles elongated parallel to the c crystal axis. These needles are then assumed to become oriented by the shearing movements so as to lie in the slip-planes of the rock mass, with their long axes parallel to the direction of movement. It should be noted that this is merely a hypothesis based upon observations within large undulose grains of quartz, and the results of many fabric analyses of quartz in a variety of rock types. (2) Eskola (1939, pp. 303, 304) following Schmidt's conception of plastic deformation of quartz grains by translation-gliding on the basal plane, unit prism and unit rhombohedron, and taking into account the studies of Hietanen (1938) upon undulose quartz grains in Finnish quartzites, concluded that rupture along surfaces parallel to the vertical crystal axis combined with flexure-gliding in the basal plane is one of the main orienting mechanisms. (3) Griggs and Bell (1938) found that at extreme confining pressures quartz will not deform plastically but behaves elastically up to the point of rupture. Moreover, the differential stress necessary to cause rupture under high confining pressures was found to be very great indeed. However, in the presence of water and Na2CO3, though again there was no plastic deformation, rupture was effected at much lower differential stresses—the magnitude of which is reconcilable with values obtaining in actual deformation in the outer part of the earth's crust. The quartz under these circumstances showed a strong tendency to fracture into needle-like fragments elongated parallel to the c axis, the unit rhombohedron or the basal plane. If these fragments now become dimensionally oriented in the manner assumed by Sander, the several possible resultant lattice orientations would correspond well with the principal types of quartz orientation generally recognised by Sander and others in actual fabric diagrams. This same thesis has been elaborated somewhat by Fairbairn (1939). Paratectonic Crystallisation. Crystallisation proceeding simultaneously with deformation has long been recognised as a most important factor in connection with the development of schistosity. There has, however, been a good deal of confusion and lack of agreement as to the relative effectiveness of several mechanisms of crystallisation. (1) Riecke's Principle: Schistosity governed by strong preferred dimensional orientation of mineral grains, supposedly in a plane perpendicular to a simple compressive force, has been

attributed by many writers to recrystallisation controlled by “Riecke's principle.” Riecke's papers are not available to the present writer, but more than one process of orienting appears to be concerned in “Riecke's principle” as described in current literature. Johnston and Niggli (1913, pp. 610, 611) summarise the principle thus: “… when a crystal is strained, the solubility of the strained face is increased. Consequently material tends to dissolve off the strained faces of a crystal in contact with a saturated solution in any solvent, and to be redeposited where the strain is less. This does not necessarily imply that the mineral be redeposited on the unstrained faces of the original crystals; it may go to form new crystals. In very many cases, of course, a reaction or transformation will occur, so that the mineral will not be redeposited in its original form.” The same authors then go on to deduce that both the original and the newly growing grains will tend to be elongated in the plane perpendicular to the direction of greatest stress. The two complementary mechanisms involved are deformation (flattening) of the pre-existing grains and simultaneous growth of new grains in unstressed positions (cf. also Harker, 1932, p. 145; Eskola, 1939, pp. 282, 283). The former could give rise only to a dimensional orientation of grains which would still retain their initial random lattice orientation. But the newly growing crystals could assume a preferred orientation according to space-lattice, provided the mineral in question is anisotropic and has the capacity to grow more rapidly in one crystallographic direction than in others. This aspect has been stated thus by Eskola (loc. cit., p. 283): “In the growth of the crystals a selection takes place: the crystals situated in the correct position grow further in the schistosity-planes, while the crystals situated transversely to this direction remain smaller and shorter, as is frequently found with the ‘cross biotites’ or ‘cross hornblendes’ of the crystalline schists. If the mineral possesses a favoured direction of growth (e.g. mica or amphibole) an orientation thus comes into existence—crystallisation-schistosity.” The effectiveness of Riecke's principle in metamorphism of rocks has been questioned by several writers in the light of petro-fabric evidence (e.g. Knopf, 1933, pp. 460, 461, 470; Gilluly, 1934, p. 191; Fairbairn, 1937, pp. 61, 62); nor does Sander (1930) attach any weight to it. On the other hand Niggli (Grubermann and Niggli, 1924, p. 464) and Eskola (1939, pp. 294, 295) believe that the Riecke principle can be effective, acting in conjunction with other mechanisms, in cases where the direction of the compressive force is approximately perpendicular to the schistosity. This condition is believed to hold for those steeply dipping schists of fold-mountain regions, the deformation of which Sander describes as “flattening” often without much tectonic transport. The results of fabric analysis show clearly that Riecke's principle is much less generally applicable as an explanation of crystallisation-schistosity than was once thought to be the case. As far as

the present writer is aware no instance is known of a fully investigated schist fabric that may be attributed wholly and without doubt to its influence. Moreover, if orientation by space-lattice is taken into account in conjunction with the more obvious preferred orientation according to grain-form, it is difficult to visualise how the operation of Riecke's principle could explain any of the following very commonly occurring cases:— (a) Fabrics in which mineral grains with strong lattice orientation related to the schistosity show no obvious preferred dimensional orientation. The quartz and calcite fabrics of many rocks possessing crystallisation schistosity belong to this category. It seems reasonable to infer that dimensional orientation of platy or prismatic minerals in the same rock is secondary and depends primarily upon lattice orientation of the grains in question. (b) Fabrics with strongly developed girdle patterns such as are common in B-tectonites. The fabric is symmetrical about a line in the schistosity plane, not the normal to the schistosity as demanded by Riecke's explanation. (c) Fabrics in which strong dimensional orientation with elongation of grains in the plane of schistosity is accompanied by a lattice orientation that cannot be correlated with the schistosity. For example, in many of the rocks where schistosity has developed by “flattening” the grains of quartz, noticeably elongated in the plane of schistosity, have their optic axes in one or other of two planes that intersect the schistosity-plane obliquely. (2) Sander's Views on Paratectonic Crystallisation: Sander (1930, pp. 115, 269–275) distinguishes between two types of componental movement by which deformation is achieved in meta-morphism: direct componental movements including mechanical deformation and rotation or sliding of grains, and indirect componental movements involving transport of atoms or atomic groups by such means as solution and re-deposition, diffusion through pore solutions, convection governed by temperature gradients, and migration of solutions under the influence of pressure gradients, insofar as all these movements are related to the process of deformation. The velocity of movement of individual grains is always extremely slow, but for a deformation of constant velocity is pro-portional to the size of the grains concerned. Solution and redeposition of minerals such as quartz or calcite is also a slow process and requires a minimum time in which to be effected. If, then, the velocity of mechanical movement of grains exceeds a limiting value, solution and crystallisation of mobile material are impossible. On the other hand, if mechanical deformation is sufficiently slow and the rock sufficiently fine in grain paratectonic crystallisation can occur. Increased temperature, too, will speed up chemical reactions and will therefore favour crystallisation (cf. Turner, 1941, pp. 12–14). Crystallisation is specially active on the boundaries of larger pre-existing grains and along surfaces of discontinuity due to shearing or tensional rupture, and by healing the small-scale fractures considerably assists in preserving the continuity of the rock during deformation (cf. Knopf, 1938, p. 38).

Paratectonic crystallisation as pictured by Sander* In a recent discussion on the origin of “flow cleavage,” C. O. Swanson (1941, p. 1249) first defines as “hypotheses of movement” those that attribute preferred orientation purely to mechanical movements (direct componental movements in Sander's terminology). Himself an advocate of the hypothesis of recrystallisation of mineral grains with their cleavages and longest dimensions perpendicular to the direction of maximum stress, Swanson (loc. cit., p. 1252) then states, in criticism of Sander's views, that “… the Sander school attributes to movement all the oriented fabrics of deformed rocks.” This hardly conveys an adequate picture of the complex interplay of crystallisation and mechanical movement visualised by Sander. According to Swanson rock flowage is brought about by indirect componental movement exclusively; Sander pictures indirect and direct componental movements acting together, not the latter alone as implied in the above quotation. usually leads to establishment of some type of preferred orientation of the growing grains. Three possibilities have been recognised from analysis of fabric:— (a) By recrystallisation without any rotation of grains a lattice-orientation may develop similar to that arising from direct componental movements (plastic deformation, rotation of grains, etc.). An instance of partial recrystallisation of quartz illustrating this principle was examined in great detail by Sander and has been cited by Mrs Knopf (Knopf, 1938, pp. 148, 149, Fig. 40; p. 171) who states that this is one of the most important explanations of the development of preferred orientation in tectonites. (b) During the early stages of deformation the small “seed-crystals” assume a tectonite orientation (according to space-lattice) by combined operation of the intergranular and intragranular movements described in earlier sections of this paper. When in the later stages of deformation paratectonic crystallisation becomes effective the earlier preferred orientation related to the still effective movement-plan is preserved in the recrystallised rock. (c) Sander attaches considerable importance to the principle of growth of crystals with their long axes in the direction of the greatest ease of growth (Wegsamkeit), thus giving rise to dimensional and lattice orientation of grains (cf. Eskola, 1939, p. 294). Crystallisation of this type may be either paratectonic or post-tectonic. It is therefore discussed more fully below. (3) Experimental Results: The work of Griggs (1938) on the plastic deformation of marble and limestone at high confining pressures has shown that recrystallisation, made possible in his experiments by the presence of Na2CO3 in solution, is probably of great importance in natural deformation of rocks in the accessible portion of the earth's crust. While it was impossible to produce plastic deformation of dry quartz even at enormous confining pressures, preliminary results suggest that quartz, too, becomes mobile in the presence of solutions and can deform plastically by recrystallisation. Post-Tectonic Crystallisation. From fabric studies it seems probable that crystallisation processes commonly continue to operate after active deformation of the rock has ceased and while the temperature of metamorphism still

remains high. Apart from change in size of grain this post-tectonic crystallisation may affect the fabric in any of three ways:— (a) Where the rock fabric is already anisotropic as a result of strong penetrative movements during the deformational phase of metamorphism, the fabric planes upon which slipping occurred are now surfaces of readiest migration of pore solutions, i.e., surfaces of greatest ease of crystal growth. “Those crystals that have their directions of most rapid growth in the fabric in the schistosity plane become elongated in this direction, as for example the amphibole pencils of garbenschiefer. In rocks with a purely plane schistosity, the amphibole prisms lie unoriented in the s-planes; in rocks with linear schistosity … the c axes of the amphibole lie subparallel to the axis of lineation, no matter whether this represents the tectonic axis b = B (as usually in overthrust sheets) or whether it is parallel to the glide direction a (as in most fusion tectonites with fluidal fabric and in shear-surface mylonites)” (Eskola, 1939, p. 294). The resultant grain orientation is essentially dimensional, and the schistosity is a type of crystallisation-schistosity. Growth of crystals has not caused the schistosity but has merely preserved and emphasised a set of s-planes already in existence (Knopf, 1938, p. 68). It must be borne in mind that the original s-planes may in the first place have originated in a number of different ways, e.g., as shear-surfaces, fracture cleavages, etc., and even need not be connected with deformation of any sort. Thus s-planes of purely sedimentary origin in a shale may give rise, on recrystallisation and reconstitution of the rock in the absence of differential stress, to a well-defined schistosity; the rock in question though schistose, is not a tectonite. Schistosity parallel to bedding is probably more often due, however, to development of slip-surfaces parallel to bedding in the manner described in an earlier section (cf. Eskola, 1939, p. 295). (b) Post-tectonic mimetic crystallisation may also preserve an earlier preferred orientation according to space-lattice, that has developed during a pre-crystalline or paracrystalline deformational phase of metamorphism. This process may be compared with annealing of cold-worked metals, and seems from results of fabric analysis to play a notable part in the development of coarse-grained schists with lattice-orientation. (c) Under other circumstances, recrystallisation of a particular mineral after deformation has ceased may destroy the original lattice orientation. This mechanism has been suggested by Sahama (1935, p. 20) and Eskola (1939, p. 295) as a possible explanation for the almost unoriented condition of the quartz in certain granulites from Greenland and Finland. While noting that cases of “unorienting” must be common in rocks that recrystallise or are chemically reconstituted at high temperatures, Eskola (loc. cit.) states that anisotrophy, once attained by orienting of grains in a rock mass, persists for a very long time and exerts a powerful control over its response to differential stresses in later deformations.

Kinds of Schistosity in Relation to Movements and Forces of Deformation. From the above discussion it is clear that schistosity is seldom a product of any one simple process, but rather of several processes operating simultaneously or in sequence, in accordance with a definite symmetrical movement-plan, which in turn is controlled by a system of forces acting upon the rock mass. Whereas it is usually possible by combining field and laboratory observations upon rock fabrics (including the elaborate investigations involved in petrofabric analysis), to synthesise the movement-plan in some detail, and even to trace changes in its pattern when such occur, it is always more difficult and often impossible to reconstruct an accurate picture of the force system that governed the movement. Until Sander's work became known, most writers on schistosity attempted to correlate it with some simple system of deforming forces, and many made use of the strain ellipsoid in this connection. It was common practice, too, to generalise too widely from particular cases and to admit only one general type of correlation between schistosity and deforming forces as being possible. Prevalent views, some of which are still found in current text-books, include the following:— (1) Schistosity always perpendicular to a compressive force (e.g., Sharpe, Sorby, Harker), (cf. Harker, 1932, pp. 154, 155, 193–195; Swanson, 1941, pp. 1249, 1250, 1259). (2) Schistosity involving preferred dimensional orientation of crystals, always parallel to the AB plane of the strain ellipsoid, i.e., perpendicular to the direction of maximum shortening of the mass. For a simple compression as between the jaws of a vice, the schistosity would be at right angles to the compressing force (Leith and Mead and followers of the Wisconsin School) (cf. Leith and Mead, 1915, pp. 169–182; Turner, 1936, pp. 203, 204; Mead, 1940, p. 1010). Fracture cleavage was interpreted as a shear-phenomenon of independent (later) origin. (3) Schistosity parallel to shear surfaces (Becker) (cf. Becker, 1907; Turner, 1936, pp. 204, 205). Two possibilities were recognised: (a) Two equal intersecting schistosities with the direction of the compressive force bisecting the obtuse angle of intersection. (b) A single well-defined schistosity (the usual case) crossed by simultaneously developed fracture cleavage; the major axis of the strain ellipsoid bisects the acute angle of intersection (Becker, 1907, Fig. 6, p. 13; Sander, 1930, Fig. 44, p. 101); the deforming force, exact direction unknown, is oblique to the schistosity. The work of Sander, Schmidt and their co-workers in petrofabric analysis can be applied to the problem of schistosity in relation to deformation if it be assumed that the features of a tectonite fabric conform to the symmetry of the movement-plan under whose influence the fabric developed. That such is essentially the case cannot be doubted by anyone familiar with the quantity of petrofabric data now available. These modern investigators of fabric recognise that none of the above three theories of schistosity is either

wholly wrong or universally applicable; but it seems likely that Becker's long-neglected theory of schistosity as a shear phenomenon, when modified to some extent, furnishes the most generally applicable explanation of schistosity. Workers on petrofabrics correlate s-surfaces of all kinds with deforming movements, but do not attempt anything more than a very broad correlation with deforming forces. They have shown that several sets of schistosity-surfaces may originate in a single deformation (as Becker insisted), though it is also possible for a second schistosity to be superposed upon s-planes of an earlier deformation. Interpretation of schistosity, therefore, requires great care and should be based upon as full data as possible. The following possibilities are recognised:— (1) In most schistose rocks that have been deformed by penetrative movement affecting the component grains the tectonic axis of the deformation is parallel to the trend of the lineation (b fabric axis) in the principal plane of schistosity. The tectonic axis is also parallel to the axis of folds developed at the same time as meta-morphism and is therefore the most important recognisable direction in the fabric—more useful in working out the tectonics of the area than is the schistosity itself. It is perpendicular to the direction of slip movement in shear-planes and parallel to the axis of rotational movements connected with deformation. Lineation in a schist may be due to one or more of four structural characters: (a) intersection of two contemporaneously developed sets of slip-surfaces parallel to (hOl) of the fabric;* Lineation due to intersection of two sects of s-surfaces belonging to two different deformations has no such significance and is not a tectonic axis. (b) microfolding of alternating micaceous and quartzo-feldspathic layers; (c) growth of pencils of minerals such as quartz and feldspar in a micaceous or chlorite matrix; (d) elongation of mineral grains parallel to the lineation. Sander (e.g. 1934, pp. 42, 43) has distinguished two cases corresponding to two distinct types of deformation; in one the tectonic axis is nearly horizontal; in the other it is approximately vertical. In either case fabric analysis typically yields orientation diagrams with well developed girdles the axis of which is the lineation b. A special case where the macroscopic lineation is not parallel to the tectonic axis is described below under (3). For a good example of the use of regional mapping of linear structure in a tectonic study, see Phillips, 1937, pp. 591–597. (2) The most widely occurring type of schistosity is that which develops parallel to surfaces of slip that commonly though by no means always coincide with pre-existing s-planes (especially bedding planes). Possible distinctive criteria are: lineation parallel to the tectonic axis b; presence of several sets of visible schistosity surfaces (sometimes including surfaces of strain-slip or fracture cleavage) intersecting in b; rotational structures visible in thin sections and polished sections cut at right angles to b; pencil structure parallel to b; tension joints perpendicular to b; girdle patterns in fabric diagrams. When fabrics of this type characterise large areas of rocks in which the schistosity is sub-horizontal except for the effects of post-metamorphic dislocations, the schistosity is probably due to slip

accompanying regional alpine thrusting movements, e.g., the Shuswap terraine of British Columbia (Gilluly, 1934); the Central Otago schists of southern New Zealand (Turner, 1940). This is especially likely if the trend of the lineation is fairly constant over wide areas. In the past such areas have been cited as instances of static recrystallisation under vertically directed load, but this explanation cannot be reconciled with the movement-plan deduced from the fabric (see especially Gilluly, 1934). (3) In local zones of mylonitisation and on slickenside surfaces the fabric is the result of more intense though localised deformation and much more rapid translatory movement than accompanied deformation of the schists described in the preceding paragraph. The linear structure is parallel, not perpendicular, to the direction of movement, and is not the tectonic axis of the deformation. Corresponding fabric diagrams usually fail to show a girdle pattern or if a girdle is present its axis is perpendicular to the linear structure. Similar lineations parallel to the direction of flow are, of course, well known in gneissic igneous rocks the fabric of which has been determined by movement in a partially molten mass. (4) A type of schistosity prevalent especially in rocks metamorphosed in deep-seated zones now exposed in certain Precambrian terraines, is one which usually dips steeply and appears to have developed at right angles to a simple compressive force. Many of the rocks of the Lake Superior region described by Leith, Mead and others as having “axial-plane foliation,” and the Finnish granulites studied by Eskola, Sahama, Hietanen and co-workers belong to this category. Sander applied the term “flattening” (Plättung* Translated as “plaiting” by some authors.) to this kind of deformation, for details of which the reader is referred to Sander (1930, pp. 219, 220), Knopf (1938, pp. 69, 145–149), and especially Eskola (1939, pp. 286, 300–302). The essential mechanism of deformation, as described by Eskola, is as follows:—In a rock mass slip commences on two planes of maximum shearing stress originally inclined at 45° to the compressing force. As slipping movement continues these planes rotate in opposite directions about their line of intersection (b fabric axis) and at last approach mutual parallelism when they come to be at high angles to the compressive force. Any crystal that becomes oriented in either major slip-plane, whether dimensionally or so as to allow translation of the individual grain on a crystallographic plane of potential gliding or twinning, is by this means slowly rotated about b until at last both the direction of elongation of the grain and its plane of translation (if present) approach perpendicularity to the direction of compression. In this way a surface of schistosity (ab of the fabric) comes into existence, which is determined mainly by the flattened outlines of the grains; if the two original major slip-planes (now nearly parallel) remain distinctly differentiated, the megascopic schistosity is a compromise between the two. The schistosity may be emphasised by post-tectonic mimetic crystallisation, and Eskola (loc. cit.) believes that Reicke's principle may be a contributing factor in development of the characteristic elongated form of the grains. Now although any

major slip-plane in the rock mass, once developed, commences to rotate towards the ab position, the maximum shearing-stress still continues to operate in planes, fixed in space, at 45° to the direction of compression. In the case of quartz and calcite, which adapt their space-lattice orientation to a new stress system much more readily than micas or hornblende, many of the grains that have been greatly flattened in ab as described above have adjusted their internal structure to the stress system still operating in the closing stages of deformation, i.e., with their crystallographic translation-plane parallel to one of the planes of maximum shearing stress. For such minerals there is no relation between space-lattice and dimensional orientation in rocks deformed by flattening. Characteristic features in the fabric of flattened rocks are:—Strong dimensional orientation of mineral grains with pronounced flattening parallel to the schistosity and elongation parallel to the lineation (this applies to quartz and calcite as well as to minerals with flaky or prismatic habit); micas and chlorites lie with the basal plane subparallel to the schistosity and corresponding fabric diagrams show a single maximum somewhat drawn out transversely to the lineation (submaxima are absent, and girdles are incomplete); quartz and calcite diagrams show maxima symmetrically developed on either side of the schistosity plane, or else a series of submaxima symmetrically developed about the pole of the schistosity and representing positions of concealed slip-planes in various stages of rotation about b; tension joints parallel to ac (i.e. perpendicular to the lineation) are usually well developed and may be closely spaced in the quartzose layers as seen microscopically. (5) When compression has produced only a relatively slight displacement of the flattening type and subsequent or simultaneous crystallisation of mica, etc., preserves and emphasises the two sets of fabric planes on which slipping took place, the result is a pair of microscopically distinct equally developed schistosity surfaces, the obtuse angle of which is bisected by the compressing force. Rotational structures are absent and the lineation is due purely to the intersection of the two sets of schistosity surfaces. Corresponding fabric diagrams show an orthorhombic symmetry and the two sets of schistosity should be represented clearly by maxima for both micaceous minerals and quartz or calcite. There is complete graduation from fabrics of this type, through fabrics in which the visible slip-planes intersect at progressively more acute angles, to typical flattened fabrics such as those discussed above under (4). (6) Schistosity may also be a product of simple recrystallisation unconnected with deformation but preserving an earlier set of surfaces of non-tectonic origin (e.g., bedding). Distinguishing criteria of such fabrics are: absence of lineation; absence of (hOl) surfaces, ac joints, etc.; development of dimensional and corresponding space-lattice orientation of minerals with pronounced crystallographic habit such as micas or hornblende; for minerals like garnet, quartz or calcite, which do not develop a pronounced crystallographic habit under conditions of metamorphic growth, there may or may not be a dimensional orientation, but space-lattice orientation is never developed.

(7) Rarely, tension fractures parallel to the ac fabric plane are so closely spaced as to give rise to a parallel parting that can be termed schistosity (Sander, 1930, p. 219; Eskola, 1930, p. 309). Schistosity of this type should be easily distinguishable by its relation to other schistosity surfaces and epecially by its orientation at right angles to the lineation, i.e., to the axis of the mica girdle (if present). (8) The presence of two intersecting sets of schistosity surfaces (one of which may often be a strain-slip cleavage) may be due to the superposition (“overprinting”) of one deformation upon another, or to a change in the movement-plan in the later stages of a long-continued deformation (cf. Ingerson, 1936; Knopf, 1938, p. 149; Turner, 1940, p. 189). The presence of two intersecting lineations (one of which is a tectonic axis) in the principal schistosity is a certain magascopic indication of this condition. In corresponding fabric diagrams the axes of the mica and quartz girdles usually fail to coincide, since in most cases the quartz fabric alone bears the imprint of the last deformation. Sometimes it is also possible from the texture alone to determine that crystallisation has been post-tectonic (with reference to the last deformation) for some constituents, pre-tectonic for others. (9) In folds originating through flexural-slip, the slip-surfaces and the corresponding schistosity of the type described under (2) above lie parallel to the bedding. On the other hand, in pure slip-folds schistosity originating by slip trends parallel to the axial planes of the folds. Also parallel to the axial planes is schistosity developed by flattening [cf. (4) above] when this is present in folded rocks. Literature Cited. Becker, G. F., 1893. Finite Homogeneous Strain Flow and Rupture in Rocks, Bull. Geol. Soc. Amer., vol. 4, pp. 13–90. —– 1904. Experiments on Schistosity and Slaty Cleavage, U.S. Geol. Surv. Bull., No. 241. —– 1907. Current Theories of Slaty Cleavage, Amer. Jour. Sci., Ser. 4, vol. 24, pp. 1–17. Daubree, A., 1876. Experiences sur la schistosité des roches et sur les déformations de fossiles correlatives de ce phénomène, Bull. Soc. Geol. France, 3rd ser., vol. 4, pp. 529, etc. Eskola, P., 1939. Die metamorphen Gesteine, Die Entstehung der Gesteine, (T. F. Barth, C. W. Correns, P. Eskola), Berlin, Springer. Fairbairn, H. W., 1937. Structural Petrology, Queen's University, Kingston, Canada. —– 1939. Hypotheses of Quartz Orientation in Tectonics, Bull. Geol. Soc. Amer., vol. 50, pp. 1475–1492. Gilluly, J., 1934. Mineral Orientation in some rocks from the Shuswap Terraine as a clue to their Metamorphism, Amer. Jour. Sci., vol. 28, pp. 182–201. Griggs, D., 1935. The Strain Ellipsoid as a Theory of Rupture, Amer. Jour. Sci. vol. 30, pp. 121–137. —– 1936. Deformation of Rocks under High Confining Pressures, Jour. Geol., vol. 44, pp. 541–577. —– 1940. Experimental Flow of Rocks under conditions Favouring Recrystallisation, Bull. Geol. Soc. Amer., vol. 51, pp. 1001–1022. 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.

Grubenmann, U., and Niggli, P., 1924. Die Gesteinsmetamorphose, I, Berlin, Borntraeger. Harker, A., 1932. Metamorphism, London, Methuen. Hietanen, A., 1938. On the Petrology of the Finnish Quartzites, Bull. Com. geol. Finlande, No. 122. Hills, E. S., 1940. Outlines of Structural Geology, London, Methuen. Ingerson, E., 1936. Fabric Analysis of a Coarsely Crystalline Poly-metamorphic Tectonite, Amer. Jour. Sci., vol. 31, pp. 161–187. Knopf, E. B., 1931. Retrogressive Metamorphism and Phyllonitisation, Amer. Jour. Sci., vol. 21, pp. 1–27. —– 1933. Petrotectonics, Amer. Jour. Sci., vol. 25, pp. 433–470. —– 1935. Recognition of Overthrusts in Metamorphic Terraines, Amer. Jour. Sci., vol. 30, pp. 190–209. —– 1938. Principles of Structural Petrology, pp. 1–208 in Structural Petrology, Mem. Geol. Soc. Amer., No. 6. (Knopf, E. B. and Ingerson E.). Leith, C. K., 1905. Rock Cleavage, Bull. U.S. Geol. Surv., No. 239. —– 1913. Structural Geology, New York, Holt & Co. Leith, C. K. and Mead, W. J., 1915. Metamorphic Geology, New York, Holt & Co. Mead, W. J., 1940. Studies for Students: Folding, Rock Flowage and Foliate Structures, Jour. Geol., vol. 48, pp. 1007–1020. Phillips, F. C., 1937. A Fabric Study of some Moine Schists and Associated Rocks, Q. J. G. S., vol. 92, pp. 581–620. Sahama (Sahlstein), T. G., 1935. Zur Regelung der Gesteine im Kristallin von Liverpool-land, Meddelelser om Gronland Kom. Vidensk. Undersogelser Gronland, Bd. 95, Nr. 6. Sander, B., 1930. Gefügekunde der Gesteine, Vienna, Springer. —– 1934. Petrofabrics (Gefügekunde der Gesteine) and Orogeneses, Amer. Jour. Sci., vol. 28, pp. 37–50. Schmidt, W., 1932. Tektonik und Verformungslehre, Berlin, Borntraeger. Sharpe, D., 1849. On Slaty Cleavage (second Communication), Q. J. G. S., vol. 5, pp. 111–115. Sorby, H. C., 1853. On the Origin of Slaty Cleavage, Edinburgh New Phil. Jour., vol. 10, pp. 137–147. —– 1856. On the Theory of Slaty Cleavage, Phil. Mag., 4th Ser., vol. 12, pp. 127–129. (With observations by J. Tyndall, pp. 129–135.) Swanson, C. O., 1941. Flow Cleavage in Folded Beds, Bull. Geol. Soc. Amer., vol. 22, pp. 1245–1264. Turner, F. J., 1936. Interpretation of Schistosity in the Rocks of Otago, New Zealand, Trans. Roy. Soc. N.Z., vol. 66, Pt. 2, pp. 201–224. —– 1940. Structural Petrology of the Schists of Eastern Otago, New Zealand, Amer. Jour. Sci., vol. 238, pp. 73–106, 153–191. —– 1941. The Development of Pseudostratification by Metamorphic Differentiation in the Schists of Otago, New Zealand, Amer. Jour. Sci., vol. 239, pp. 1–16. Tyndall, J., 1856. Comparative View of the Cleavage of Crystals and Slate Rocks, Phil. Mag., 4th ser., vol. 12, pp. 35–48. Tyrrell, G. W., 1930. The Principles of Petrology, 2nd Ed., London, Methuen.