Petrofabric Investigations of Otago Schists No. 1—Quartz and Mica Fabrics of Three Schists from Patearoa.
(Read before the Otago Branch, September 14, 1937; received by the Editor, September 15, 1937; issued separately, March, 1938.)
Contents.

Literature Cited.

Tectonic Considerations.

The Quartz Fabrics.

The Mica Fabric.

Macroscopic and General Characters.

Method.

Introduction.
Introduction.
During the past ten years petrographic and field investigations of the schists of Otago (Maniototo or Wanaka Series) have considerably increased our knowledge as to the origin and metamorphism of these rocks. Subdivision of the Chlorite Zone into subzones of progressive metamorphism has been attempted. The importance of shearing as a factor in the metamorphism, and the improbability of pure loadmetamorphism being responsible for the condition of the Otago schists are now recognised. Further there is now a considerable body of evidence indicating that metamorphism took place in more than one stage, the principal of which may have been as early as Permian or early Triassic. As yet, however, no definite and complete solutions to the following problems have emerged:—
(1) 
The relation between the Maniototo Series (coarse schists), the Kakanui Series (sheared greywackes, semischists and finegrained schists) and the sheared fossiliferous Triassic sediments of Mt. St. Mary. 
(2) 
The number of stages involved in the metamorphism of the Maniototo Series. 
(3) 
The dates to which such metamorphic stages should be assigned. 
It should be possible to obtain reliable evidence bearing upon these three problems by applying the methods of modern petrofabric analysis to the various rocks concerned. With this object the writer has commenced a series of petrofabric investigations upon suitable material from different parts of Otago. At first the results will be recorded in full as the work progresses, until enough data have been collected to allow generalisation upon the fabric of Otago schists as a whole.
The present paper embodies the results obtained from examination of three specimens of schist collected from within Subzone Chl. 4^{*} about threequarters of a mile northeast of Patearoa Hotel just east of the Waipiata Road.
Method.
Before the specimen is finally broken from the outcrop, its field orientation is marked upon it. Where, as in the present instance, there is a wellmarked planeschistosity with obvious lineation, it is sufficient to mark the top north end of the slab, at the same time noting strike and dip of schistosity and direction and dip of the lineation. The three principal directions in the specimenblock are desgnated a, b and c as in crystallography: c = the normal to the plane of schistosity; a and b are mutually perpendicular directions in the schistosityplane, b being parallel to the direction of lineation. Three mutually perpendicular thin sections are now cut parallel to the ab, bc and ac planes respectively, and mounted with their cut surfaces against the glass slide. Later calculations are simplified if a uniform procedure is adopted in orienting a, b and c with respect to field directions, and in cutting sections from particular ends of the three axes. In the rocks now under discussion the scheme shown in Fig. 1 was adopted; if, however, it was found necessary in particular cases to cut the section from the opposite end of the axis in question, the fabric was later plotted in reverse so as to conform with the standard scheme of Fig. 1.
Fig. 1.—Diagram to show orientation of sections with reference to schistosity (ab), lineation (b) and the reference axes (a, b and c). The geographic orientation is given by the northwardpointing arrow. Sections were cut along broken lines.
In the absence of a universal stage, the main features of the quartz and mica fabrics in the three sections were determined by
[Footnote] * For definition of the four subdivisions of the Chlorite Zone in Otago see Turner (1935) and Hutton and Turner (1936).
means of an ordinary petrographic miscroscope, fitted with a simple mechanical stage by which the possibility of rotational movement of slide with respect to stage is eliminated. For two or three hundred grains in each section the angle is measured between a crystallographic line of reference (Z′ in quartz; trace of 001 cleavage in mica) and one of the sectional axes (b in ab and bc sections; a in ac section). A partial example of such a determination, made for quartz in a section parallel to ab, is given below:—
[The section below cannot be correctly rendered as it contains complex formatting. See the image of the page for a more accurate rendering.]
Reading with Z′ parallel to E—W crosswire.  No. of grains.  Percentage of grains.  Percentage of grains per interval of 10°.  Angular distance Z′ to b. 

135°139°  16  6.4  (140–95)=45°  
140°144°  13  5.2  11.6  
145°149°  23  9.2  14.4  50° 
150°154°  16  6.4  15.6  55° 
155°159°  32  12.8  19.2  60° 
160°164°  13  5.2  18.0  65° 
165°169°  10  4  9.2  70° 
170°174°  6  2.4  6.4  75° 
175°179°  3  1.2  3.6  80° 
The second column shows the number of quartz grains whose slow vibrationdirections (Z′) lie within the angular distance of 5° indicated by stage readings in the first column. In the third column this number is shown as a percentage of the total number of grains measured. The fourth column is derived from the third by adding pairs of consecutive figures in the latter. Thus 11.6% of the grains measured lie within 5° on either side of a direction inclined at 45° to the reference direction b.
The principal features of the fabric in each section are best shown by curves in which the angular distance between the crystallographic reference line (e.g., Z′ in quartz grains) and b or a as the case may be, is plotted against percentage of grains falling within overlapping intervals of 10°. As pointed out by Ingerson (1936), it may be difficult or impossible to correlate the features shown by independent thin sections of different orientation if the fabric is not homogeneous. However in the rocks investigated by the writer correlation of the principal maxima and minima developed on the curves for ab, bc and ac sections has usually proved possible. The quartz curves are in most cases far from simple. Therefore any threedimensional orientation of maxima on a stereographic projection that will simultaneously satisfy the observed maxima on the three curves (ab, bc and ac), may safely be assumed as approximately correct. Examples of the application of this principle will appear later.
Macroscopic and General Characters.
The rocks forming the subject of this paper were collected from the northern end of an isolated hill of schist surrounding Trig. N, ¾ mile northeast of Patearoa, Upper Taieri Survey District (see map accompanying Williamson, 1934). They are coarse schists consisting of albite, quartz, muscovite, epidote and minor stilpnomelane and ironore. Schistosity and foliation (cf. Turner, 1936, pp. 202, 203) are well developed and subhorizontal; locally they may dip at angles ranging up to 10° in a southeast to southsoutheast direction. Lineation in the schistosity is well defined and lies between 155° and 165°, thus approximating to the N.W.S.E. to N.N.W.S.S.E. trend that the writer has constantly observed for this structure in schists throughout Eastern Otago. It is determined partly by orientation of the constituent grains (especially mica and epidote) with their long axes in parallel position, and partly by segregation of minerals into pencillike aggregates.
Veinlets of quartz from 1 mm. to 10 mm. in width are common. Many of these, especially the larger veins, show little regularity and appear to be of late origin; but in two of the three specimens selected for examination there are narrow veinlets cutting perpendicularly across the schistosity and showing strong corrugation as seen in sections parallel to the ac plane.
As in other parts of Otago the schists of Patearoa are strongly jointed. The principal joints are widely spaced, nearly vertical and trend approximately at right angles to the linear element in the schistosity.
The Mica Fabric.
Both the macroscopic schistosity and the lineation parallel to b are determined largely by the mica fabric. In sections parallel to ab (the schistosityplane) many of the mica flakes show loworder interferencetints indicating that (001) is substantially parallel to the plane of the section; an even greater number of flakes however are inclined to ab and tend to lie with their long axes parallel to b. In the bc section the crystals of mica are nearly all edgeon and their orientation parallel to b is very perfect. The ac section presents a marked contrast: the flakes are all edgeon to the observer, but the direction of schistosity (a) is comparatively poorly marked as far as the micas are concerned. Statistical analysis of the mica fabric shown in ac sections of the three rocks, shows that the majority of the crystals lie with their long axes inclined at less than 30° to a. The corresponding curves show two principal maxima, one at 20° and the other at 170° angular distance from a. In two sections (4456, 4457) the position of a is marked by a definite drop in the curve, but in the third (4458) there is a distinct maximum at a comparable with the other two noted above (see Fig. 2).
Fig. 2.—Orientation of mica in ac sections of the three rocks; the angular distance is measured between the trace of the (001) cleavage and the positive end of the fabric axis a. in an anticlockwise direction from the latter. Stereographic projections upon ac show the Splanes S1, S2 and S3 corresponding to the maxima on the curves.
It is thus obvious that the flakes of mica are especially concentrated in two planes (S1 and S3) intersecting in b and inclined at low angles (20° and—10°) to the plane of schistosity (S2). The latter divides asymmetrically the acute angle between S1 and S3 and is itself marked (especially in 4458) by continuous slightly undulating sheets of subparallel micas that not infrequently show undulose extinction and effects of bending, in contrast with the clearcut undeformed flakes that define S1 and S3. The relative positions of S1, S2 and S3 in the three ac sections is shown in stereographic projections in Fig. 2.
An interesting feature shown identically by the ab curves for both 4456 and 4458 is the definite submaximum at 20° angular distance from b (Fig. 3). This in each case is approximately at right angles to one of the principal directions of concentration of Z′ for quartz grains in the same sections, and thus appears to have a tectonic significance which will be discussed later. In 4457 (which was collected at a distance of only a few feet from 4458) there is no trace of this submaximum, but there is a marked concentration of the mica flakes within a 15° interval on the opposite side of b, i.e. between 165° and 180° on the corresponding curve (Fig. 3). This, however, has no apparent relation to quartz maxima in the same section.
The Quartz Fabric.
Two difficulties attend investigation of the quartz fabrics, viz. abundant development of quartose veins, the fabric of which has been determined partly by growth and partly by subsequent shearing, and predominance of albite over quartz in the main portion of the rock. The albite is for the most part untwinned, and in micaceous rocks where each grain is surrounded by a frame of mica, it, may be impossible to distinguish rapidly between quartz and albite since comparative refractive index determinations are then impossible. The work was therefore confined to examination of (a) quartz grains throughout the main portions of 4456 (sections parallel to ab, bc and ac) and 4458 (bc section only), and (b) quartz grains in veinlets in all three sections for 4457 and 4458.
Quartz Orientation in Main Portion of Rock.
Specimen 4456 is free from transgressive veinlets. The grains of quartz are relatively small (0.05 mm. to 0.3 mm.), almost invariably show undulose extinction, and occur as nests lying between the larger grains of albite and mica. The albites are irregular in shape but show a distinct tendency toward elongation parallel to b; this is especially noticeable in the bc section where an incipient augen structure is also brought out by slight flexing of the micas where they border the elongated rather lenticular albites.
Orientation curves for quartz in the three sections of 4456 are shown in Fig. 4. There are two distinct maxima on each curve. It is highly probable that these three pairs of maxima are the projections of a single pair of directions in which the Z′ directions of the quartz grains tend to be concentrated, and these may temporarily be assumed to lie in girdles sensibly perpendicular to ab (as appears likely from what is so far known as to the orientation of quartz girdles in general). On this supposition the three curves may now be correlated as follows:—
In Fig. 5 stereographic projections showing the observed quartz maxima in the ab and ac sections are given, together with a projection upon bc showing the positions of maxima as calculated from the ab and ac projections. The maxima must lie on PQ and RS in Fig. 5a,
Fig. 5.—Stereographic projections upon the principal fabric planes ab, ac and bc, showing positions of the quartz maxima A and B as deduced from the curves for ac and ab sections (Fig. 4). A and A′ are opposite poles of the same diameter of the sphere of projection; similarly for B and B′. Construction lines are broken. North is indicated in ab by the arrow.
and on LM and NK in Fig. 5b. For maxima to be developed in the top right and bottom left quadrants of Fig. 5c, they must be in the two righthand quadrants of Fig. 5a. In the ac projection the poles representing the two maxima (A′ and B′)^{*} are fixed at the intersections of (1) LM and the circle RS, and (2) NK and the circle PQ.
[Footnote] * In the projection upon ab (Fig. 5a) the upper poles A and B of the directions of concentration of quartz axes are shown. The corresponding poles on the lower half of the sphere would be A′ and B′.
By graphic construction A′ and B′ lie on small circles of the sphere distant 47° and 38° respectively from the pole of the c axis. Transferring to the bc projection, the theoretical maxima A′ and B are fixed by the intersections of these small circles and the projections of the great circles RS and PQ; their angular positions with reference to b are thus estimated as 57° and 77°. The observed maxima (see curve in Fig. 4) are 55°60° and 70°75°. The above correlation is therefore warranted, especially since the original curves cannot be regarded as accurate to less than 5°.
The most noticeable features in the fabric are:—
(1) 
A welldefined minimum parallel to b (cf. curves for ab and bc, Fig. 4). 
(2) 
A distinct minimum parallel to a, as shown especially in the ac curve, Fig. 4. 
(3) 
Two maxima A and B lying in planes perpendicular to ab and inclined at 50° and 110° respectively to b; the maxima themselves make angles of 47° and 38° respectively with c. 
(4) 
A considerable proportion of the quartz axes are subparallel to c (cf.ac and bc curves of Fig. 4). 
(5) 
Apart from the minimum parallel to b, the quartz fabric is apparently unrelated to the b axis, the principal direction in the mica fabric. 
(6) 
One quartz maximum in the ab section (110°) is perpendicular to the submaximum (20°) in the ab mica curve. This feature is also shown by one of the other rocks investigated, and indicates a quartz girdle perpendicular to a secondary mica axes b′. 
The positions of the quartz maxima are not readily explained by assuming gliding in S1, S2 or S3 according to either the unit prism rule or the Boehm lamella rule (cf. Fairbairn, 1935, pp. 35, 36). Until other fabrics are investigated, discussion of the possible origin of that just described is better postponed; in the meantime it may be noted that the original fabric corresponding to the present mica orientation has been obliterated except for the minimum parallel to b, and that a complex fabric has subsequently developed.
Of the six sections cut from specimens 4457 and 4458, only the bc section of the latter was suitable for measuring accurately the orientation of quartz grains distributed through the rock. The resultant curve shows a minimum at b and two maxima inclined at 75° and 100° to b. These might perhaps be developed by gliding on the Boehm lamellae or by twinning on (1011) in the ab plane, movement being in a direction strongly inclined to b. There is little resemblance to the bc curve for 4456, apart from the minimum parallel to b.
Quartz Orientation in Veinlets.
Transgressive veinlets of quartz accompanied by a little stilpnomelane and mica are well shown in all three sections of both specimens 4457 and 4458. The veins as seen in the bc and ab sections trend parallel to b with great regularity; in ac they are strongly contorted as shown in Figs. 6, 11 and 12. The vein fabric taken as a whole is inhomogeneous, for while alternate limbs have similar
fabrics the orientation in adjacent limbs of a single fold are dissimilar. This is obvious even from rapid inspection with a gypsum plate. At the actual crests there is no difference either in grain or in orientation between the fabrics of the inner and outer portions
Fig. 6.—(a) Camera lucida drawing of part of one of the quartz veins measured in ac section of 4458; obvious crossfractures indicated. (b) Sketch showing contorted Veinlet in ac section of 4457 (diagrammatic only).
of the foldhinge, such as has been described by Sander (1930, p. 258) in veins of rather similar appearance. Inhomogeneity of the veinfabric is borne out by the occurrence of two distinct types of orientationcurve for veins exposed in ac sections of the rocks.
In Figs. 7 and 8 the curves representing orientation of veinquartz in specimens 4457 and 4458 respectively are shown. The full lines in the ac diagram represent the fabrics of limbs of type p (Fig. 6) composed mainly of coarse grains of quartz with strongly undulose extinction, while the broken curves for ac correspond to limbs of type q made up principally of finegrained quartz (4457), sometimes enclosing aggregates of coarse undulose grains (4458). Limbs of this second type often show obvious planes of shearing parallel to a, along which streaks of mica continuous with that of the enclosing matrix may be drawn out sometimes completely breaking the vein; this is particularly the case in specimen 4457 where shearing of the q limbs is more advanced than in 4458.
In limbs of both types flakes of stilpnomelane and coarse undeformed muscovite, both apparently products of late crystallisation, are present. Their long axes as seen in the ab section are oriented parallel to b. In the ac section there is a marked tendency for flakes of both minerals to lie with the long axes parallel to a in the sheared limbs of type q, and perpendicularly to the vein wall in the less sheared limbs p.
Fig. 7.—Orientation of quartz In veinlets, specimen 4457. In the ac diagram, the full line corresponds to limbs of type p (Fig. 6), and the broken line to limbs of type q, The full and broken curves In the bc diagram represent measurements upon two distinct veinlets (or possibly two sectors of the same vein).
Fig. 8.—Orientation of quartz in veinlets, specimen 4458. In the ac diagram the full curve corresponds to limbs of type p, while the broken line represents orientation in limbs of type q (cf. Fig. 7).
Correlation of curves for quartz in the three principal sections in each rock is complicated by uncertainty as to whether particular veinlets measured in the ab and bc sections correspond to limbs of the p or q type as developed in the ac section. From careful comparison however the following facts emerge:—
(1) In both rocks the ac curves for limbs of type p show a strong maximum at A and a submaximum X in corresponding positions. In 4457 the former was tentatively correlated with the single maximum A shown by the ab curve and the position of the corresponding maximum in bc was thus calculated graphically in the stereographic projections of Fig. 9 as 118°. The observed maximum of 115°–120° in bc curves for 4457 confirms this correlation, and justifies selection of the point A as corresponding maximum in the ab curve of 4458 (where two equal maxima A and D are shown). Similarly correlation of A in the ab and ac curves for 4458 (Fig. 10) gives a graphically calculated maximum at 111° in bc as compared with the observed maximum of 110°.
(2) Taking the same set of curves in 4458, the submaxima shown at X in ab and ac may be correlated to give a graphically determined point at 102° from b on the bc curve (cf. Fig. 10). This is consistent with the form of the latter. Turning again to the curves for 4457,
Fig. 9.—Stereographic projections of quartz fabric in veins cutting specimen 4457; the maxima in all three projections are deduced from the curves for ac and ab sections (Fig. 7). North is indicated by the arrow in the ab projection.
it is probable by analogy that the inconspicuous submaxima X (Fig. 7) have similar significance, for they appear at points corresponding closely to X in Fig. 8. Calculation from the position of X on the ac and ab curves gives a position at 95° from b in the bc section of 4457; actually the bc curve shows a break at 100°.
(3) In 4458 correlation on similar lines is possible for submaximum D which is especially prominent in the ab curve. There seems to be no corresponding point in 4457.
(4) The curve representing limbs of type q in the ac section of 4457 has two maxima B and G at 60° and 105° respectively from the a axis. Correlation with the principal maximum in the ab curve using appropriate stereographic projections (Fig. 9) gives calculated maxima at 73° and 98° in the bc section. The observed maxima on the broken curve for bc are in fair agreement (70° and 110°) considering the fact that the bc and ac sections actually cut different veins.
(5) In corresponding curves for 4458 (broken curve for ac in Fig. 8) two similar points L and N are separated by a third maximum M approximately parallel to c. These two points might equally well
Fig. 10.—Stereographic projections of quartz fabric in veins cutting specimen 4458; the maxima in all three projections are deduced from the curves for ac and ab sections (Fig. 8). North is indicated by the arrow in the ab projection.
lie on the same girdle as A or on the girdle containing D. The first alternative would be consistent with the fabric of similar limbs of the folded veins of 4457, while the second might help to account for the presence of a distinct maximum at 95° to 100° from b in each of two veins measured in the ab section of 4458. In the stereograms of Fig. 10 the former alternative is indicated.
Significance of the Vein Fabrics.
The veining just described in specimens 4457 and 4458 typifies one of the several systems commonly developed in Otago schists, and its origin and tectonic significance are therefore worth careful consideration. Several possibilities present themselves in this connection:
(1) Growth of zigzag veins with simple orientation of the quartz with Z either perpendicular or parallel to the veinwalls (Fairbairn, 1935, p. 117). Maxima to be expected in the ac curve according to the first rule would be 60°–65° from a in the limbs of type p, and 135° in limbs of type q.^{*} Assuming the second orientation rule the expectable maxima would be 150°–155° for p, and 45° for q. The actual curves for 4457 and 4458 show no trace of the pairs of maxima to be expected on either assumption, and the hypothesis of simple growth without later deformation is thus eliminated.
(2) The fabric might have developed by shearing or flexurefolding of veins in which the grains were initially oriented with Z perpendicular to the veinwall. In this case there should still be some trace of the initial maximum parallel to a, particularly in the less deformed sectors p; in sectors q this maximum would perhaps be accentuated by gliding on the unit prism in ab during shearing of the rock. The actual curves for ac sections of both rocks show a strong minimum parallel to a, so that this hypothesis also is inconsistent with the facts observed.
(3) There still remains the possibility that veins that originally formed parallel to the bc plane and were composed of grains oriented with Z parallel to the veinwall, have subsequently been deformed by shearing or flexure. In support of this view, the persistent minima parallel to a and b in vertical sections strongly suggest an initial orientation of grains with Z perpendicular to ab. Further, the main features of the fabrics developed in the sectors p and q in both rocks may be explained on the assumption that deformation was accomplished by horizontal shearing in alternate fold limbs (q), with longitudinal shearing parallel to the walls in the intervening series of limbs (p).^{†}
If deformation of limbs p involved external rotation of the whole limb from the vertical to its present position, combined with longitudinal gliding of the grains on the (1010) prism, there should be a maximum at 150°—155° from a in the ac curve. The measured
[Footnote] * The angle p ∧ b = 150°–155°; q ∧ b = 45° (cf. Fig. 6).
[Footnote] † In view of the uncertainty that still exists as to the mechanisms by which orientation of quartz is accomplished in tectonites, the following paragraphs must be regarded merely as a tentative explanation of the observed fabric data. The data themselves and the tectonic conclusions put forward in the next section are not dependent upon the validity of this hypothesis.
maxima are 135° in 4457 and 132° in 4458. The difference of 20° is interpreted as a lag inherited from the original orientation with Z at 90° to a; in other words the rotation of the whole sector through 60°–65° was accompanied by rotation of the individual component grains through only 40°–45°.
In 4458 the fabric of limbs q is believed to be the result of partial granulation of quartz by shearing in the plane of schistosity ab, brought about by gliding of grains on the Boehm lamellae. The maxima L and N separated by an angle of about 45° (reduced to 40° in the projection upon ac) agree perfectly with this supposition, and are symmetrically placed at 65° to the plane of shearing as theoretically required by the crystallographic position of the Boehm lamellae in quartz. The intervening maximum M is the result of local preservation of aggregates of coarse grains, with their initial orientation (Z parallel to c) unimpaired; such indeed are clearly observable in limbs of this type as seen in the ac section.
The sectors q in 4457 are more completely sheared and show only the two maxima (B and G of Fig. 7) corresponding to L and N in 4458. These are separated by the requisite angular distance of 40°–45°, but are asymmetrically placed with reference to ab (60° and 105° from a, in the ac section). This would agree with origin by gliding on the Boehm lamellae in S_{3} rather than S_{2} (ab). It is also possible, however, that the asymmetry should be attributed to distortion of fabric accompanying rotation of the main quartz girdle into its present position during a later phase of metamorphism (see below).
The maxima discussed above (A, G and B in 4457; A, N, M and L in 4458) all lie on or close to girdles making an angle of 115° (4458) to 120° (4457) with b as seen in the ab sections. These girdles appear to be the tangible result of movements in ab acting at 115° to 120° to b, i.e. in a N.E.S.W. direction across a N.W.S.E. tectonic axis b′. As will be shown in the next section the tectonic axis of the main deformation in which the veins were first contorted and in which certain features of the present fabric were established trends about 15° to 25° W. of N. The submaximum D in the curve for limbs p in 4458 would on this assumption have been inherited from this earlier deformation which was also responsible for development of the present schistosity and lineation. The submaximum X in the curves for limbs p represents yet another disturbance of the quartz fabric probably during the last phase of metamorphism.
Late crystallisation^{*} of mica and stilpnomelane after cessation of the main shearing about b and b′ has mimetically preserved two distinct structures, viz., (1) the ab plane of shearing in the sheared limbs of the folds (q), and (2) the original transverse structure of the vein in the less sheared limbs of type p.
[Footnote] * Probably by growth of “seedcrystals” oriented during deformation, accompanied by some crystallisation of new flakes in the directions of greatest case of growth.
Tectonic Considerations.
It is generally recognised that quartz becomes adjusted to the stress conditions set up during a deformation much more readily than mica. Consequently when, as in the present instance, nonagreement of quartz and mica fabrics is attributed to superposition of a later upon an earlier deformation, it may be assumed that the mica fabric was established first. In the Patearoa rocks then the following sequence of events must hold: (1) establishment of the main mica fabric about b; (2) development of the quartz girdle about b′, with simultaneous slight disturbance of the mica; (3) late disturbance of the quartz fabric, mica being unaffected.
In further reconstructing the tectonic history of the Patearoa schists from the evidence put forward in the preceding pages, two contrasted interpretations of the latter are still possible, according to whether the corrugated quartz veins antedate or are younger than the schistosity. As seen upon the ac surface of the handspecimen or in the corresponding microsection, the general appearance of the veins suggests formation at a relatively late stage after the schistosity had developed. Nevertheless it appears more probable that growth of the veins parallel to bc antedates the schistosity. Contortion of the veins and development of schistosity would on this assumption be manifestations of a single deformation. The principal points in support of this hypothesis are as follows:—
(1) It is difficult, on the assumption that the veins are later than the schistosity, to account for their orientation. Late growth of veins parallel to the cleavageplane ab, or filling the ac joints that so commonly form during a deformation, might reasonably be expected. Where schistosity has arisen by plaiting (Fairbairn, 1935) tension joints may also occur parallel to bc, but only locally in stretched pebbles or in especially competent thin beds. The veins cutting the Patearoa schists are certainly not of this type.
(2) A considerable differential movement in the schistosityplane must be assumed to account for the present contorted condition of the veins. Displacement of this magnitude appears to the writer to be sufficient cause for development of the schistosity.
(3) Some at least of the coarse mica has crystallised subsequently to deformation of the veins, e.g. where the walls are locally bordered by clear undeformed flakes obviously of late origin; also the mica and stilpnomelane of the veins themselves.
(4) The writer is at present investigating a series of schists from Waipori, Eastern Otago, in which the present schistosity cuts perpendicularly across contorted Splanes (probably the original bedding) oriented similarly to the quartz veins of the Patearoa schists. A faint indication of a similar structure is preserved in a section of one of the rocks forming the subject of the present paper (4456).
Assuming then that the veins were formed prior to development of schistosity, the following summary of the tectonic history of the Patearoa schists is put forward:—
(1) Growth of quartz veins parallel to an early set of approximately vertical Splanes—perhaps the original bedding. The present trend of the veins, approximately parallel to b as seen in the ab section, indicates that the tectonic axis of the initial Splanes must have approximated to the present b, i.e. 15° to 25° W. of N.
(2) Development of the present schistosity by subhorizontal shearing in ab, with simultaneous contortion of the quartz veins as described on p. 459. The lineation parallel to b must have been determined during this, the principal phase of metamorphism, and the corresponding tectonic axis was thus 15° to 25° W. of N. as in the first phase of deformation. In other words the present schistosity is the result of horizontal shearing directly across the strike of the initial vertical Splanes. Orientation of quartz about the b axis is no longer obvious except that the quartz curves for ab sections of all three rocks show submaxima perpendicular to b and minima parallel to b, while the maximum D in No. 4458 (Figs. 8 and 10) is nearly perpendicular to b. However the approximate positions of the quartz maxima within the main girdle may well have been determined during this phase of deformation, and indeed must have been so if the tentative explanation of the fabric given on p. 459 is correct.
(3)Deformation across a N.W.–S.E. tectonic axis with horizontal shearing in the already established schistosity ab now distorted the quartz fabric to its present form, giving the main quartz girdle a N.E.–S.W. trend. This applies both to veins and to quartz distributed throughout the rock. By simultaneous gliding of mica flakes in the plane of schistosity, the original mica fabric was disturbed and a submaximum parallel to the new axis was initiated (as seen in the ab section).
(4) A late disturbance of the quartz fabric is indicated by the submaxima X in diagrams corresponding to the leastsheared limbs (p) of the contorted veins (4457, 4458), and in the equivalent maximum A in curves representing the quartz fabric of the enclosing schist (4456). The micas remained unaffected, thus indicating the relative lateness of this movement. The trend of the tectonic axis during this phase of deformation was 25° E. of N.
Mica lying in the schistosityplane (ab = S_{2}) includes both sharply defined and distorted crystals. Crystallisation of the latter must at least be earlier than the third phase of deformation as outlined above. On the other hand final crystallisation of the sharply defined undeformed flakes that lie in S_{1} and S_{3} is posttectonic with respect to this phase. The significance of these surfaces S_{1} and S_{3}, retained only by late mimetic crystallisation of the micas, is uncertain. Possibly they represent the contorted remnants of the original Splanes parallel to which the veins of quartz first developed.
While the writer favours the general scheme described above, it must be borne in mind that the quartz veins may possibly be younger than the schistosity. On this hypothesis the sequence of events would be :—
(1) 
Horizontal shearing resulting in simultaneous development of S_{1}, S_{2} and S_{3}, intersecting in the tectonic axis b (15°–25° W. of N.) 
(2) 
Growth of quartz veins parallel to bc. 
(3) 
Reshearing in ab about a tectonic axis b′ (N.W.), with simultaneous contortion of quartz veins and development of the main features of the quartz fabric. 
4. 
Late minor shearing about an axis trending 25° E. of N. 
It is still too early to generalise as to the geological age of the various phases of metamorphism described above. Attention may however be drawn to the fact that deformation appears to have taken place in at least four distinct stages, in which the respective directions of the corresponding tectonic axes were (1) and (2) 15°–25° W. of N., (3) N.W., and (4) 15°–25° E. of N. The third of these corresponds with the strike of rocks folded during the late Jurassic Hokonui orogeny, while the last may be related to the general N.N.E. to N.E. trend of Tertiary faultlines in this part of Otago (cf. map, Williamson, 1934). Both represent late directions superposed upon the fabric that developed during the main metamorphism about the N.N.W. axis.
Literature Cited.
Fairbairn, H. W., 1935. Introduction to Petrofabric Analysis, Dept. of Geology, Queen's University, Kingston, Canada.
Hutton, C. O., and Turner, F. J., 1936. Metamorphic Zones in Northwest Otago, Trans. Roy. Soc. N.Z., vol. 65, pt. 4, pp. 405, 406.
Ingerson, E., 1936. Fabric Analysis of a Coarsely Crystalline Polymetamorphic Tectonite, Am. Jour. Sci., vol. xxxi pp. 161–187.
Sander, B., 1930. Gefugekunde der Gesteine, Vienna, J. Springer.
Turner, F. J., 1935. Metamorphism of the Te Anau Series in the Region Northwest of Lake Wakatipu, Trans. Roy. Soc. N. Z., vol. 65, pt. 3, pp. 329–349.
— 1936. Interpretation of Schistosity in the Rocks of Otago, New Zealand, Trans. Roy, Soc. N.Z., vol. 66, pt. 2, pp. 201–224.
Williamson, J., 1934. Naseby Subdivision, Ann. Rept. Geol. Surv. N.Z., no. 28, 1933–1934, pp. 6–10.