Myron G. Best. Igneous and metamorphic 2003 Blackwell Science

Подождите немного. Документ загружается.

three dimensional geometry of porphyroblast–matrix

relations because the pattern of inclusions can depend

on where the section cuts through the porphyroblast—

in its center or rim—and on its orientation relative to

the surrounding matrix foliation. To overcome incom-

plete and misleading information related to this “cut

effect,” parallel serial slices cut through the rock,

preferrably in three orthogonal orientations, may be

required to elucidate the true three-dimensional fabric

of the internal fabric in the poikiloblast and the exter-

nal fabric in the surrounding matrix. These two fabrics

are sometimes designated as S

and S

, respectively

(Section 15.1.1).

Prekinematic Fabrics. Mineral grains existing in the

rock prior to deformation are bent, kinked, and exhibit

ribbon texture, deformation lamellae, and undulatory

extinction in thin section under cross-polarized light

(Figures 17.12, 17.14, 17.17, 17.25b). Other prekine-

matic fabrics include fractured grains (Figure 8.7) and

boudins (Figure 15.10).

Prekinematic poikiloblasts hosted in a foliated ma-

trix can have randomly arrayed inclusions or straight

inclusion trails oblique to the matrix foliation. The

former happens where the poikiloblasts overgrew an

isotropic groundmass; one possible scenario is a hornfels

that contains cordierite and andalusite poikiloblasts

(Figure 14.25c) and later develops an overprinting

foliation under nonhydrostatic stress but otherwise

similar metamorphic conditions, while the poikilo-

blasts remain essentially undeformed. Poikiloblasts

with straight S

oblique to S

might have overgrown an

inherited bedding in a metasedimentary rock and a

subsequent deformation contorted or rotated the

matrix relative to the poikiloblast, or vice versa (Fig-

ure 17.28). Passchier and Trouw (1996, p. 161) recog-

nize an “intertectonic” fabric of the same character in

which a pre-existing foliation imprinted by an early

episode of deformation is overgrown by the poikilo-

blasts and then a later deformation overprints a second

foliation in the matrix while rotating the poikiloblast.

Synkinematic Fabrics. These are perhaps the most com-

monly encountered in rocks formed in an essentially

single-cycle metamorphism where deformation and

mineral growth are simultaneous processes. Many

synkinematic fabrics have no particularly speciﬁc fea-

tures indicative of pre- or post-kinematic grain growth

and are simply a texturally equilibrated tectonite fab-

ric, such as in Figures 15.8 and 17.29a. The latter ﬁgure

illustrates a common fabric that can be interpreted to

be synkinematic: folia of micas are deﬂected in the

vicinity of porphyroblasts, bowing over them. Contro-

versy has arisen whether the growing porphyroblasts

exerted a “force of crystallization” that physically

bowed out the adjacent foliation or whether the

wrapped foliation was deformed in the vicinity of the

porphyroblasts, such as by some sort of pressure solu-

tion process. If porphyroblasts grow along grain

boundaries, enveloping unreacted matrix grains (Fig-

ure 17.2), or entirely consuming them, then there can

be little force of crystallization. On the other hand, if

the foliation is deformed by ductile processes at the

more rigid porphyroblasts, in an analogous manner as

Evolution of Imposed Metamorphic Fabrics: Processes and Kinetics

547

(a)

Microlithon

(b)

(c)

17.28 Schematic models of development of prekinematic fabric

where a pre-existing poikiloblast (shaded circle) is rotated

relative to the surrounding matrix during homogeneous defor-

mation. The external foliation S

in the matrix, which could be

relict bedding in a metasedimentary rock, becomes oblique to

the internal foliation S

in the poikiloblast (deﬁned by a trail of

inclusions incorporated from the matrix during growth). Note

the importance of specifying a frame of reference when de-

scribing the rotation of a grain in a rock. Also note that the

deformation paths differ in these three models but the ﬁnal

relation betwen poikiloblast and its host matrix is similar.

Redrawn from Passchier and Trouw (1996). (a) Model is sub-

jected to right-lateral (dextral) simple shear; the poikiloblast

experiences a clockwise rotation with respect to the matrix.

(b) Model is subjected to coaxial pure shear; poikiloblast

remained stationary but S

experienced signiﬁcant rotation

relative to it and an external frame of reference, such as the

surface of the Earth. (c) Model is subjected to left-lateral

(sinistral) simple shear. S

experienced signiﬁcant rotation

relative to the poikiloblast and to an external frame of refer-

ence, while the poikiloblast remained stationary because it

is embedded within and coupled to a nonrotating, non-

deforming microlithon (shaded; see Section 15.1.1).

compaction of the viscous glassy matrix around phe-

nocrysts in a welded tuff (Figure 7.34c), then a prekine-

matic timing cannot be excluded.

Mineral aggregates in strain shadows (Figure 15.10b

and Section 15.1.2) extending along the foliation be-

side boudins or porphyroblasts imply growth during

deformation.

A common synkinematic poikiloblast–matrix rela-

tion develops where mineral growth and noncoaxial

ductile ﬂow were essentially concurrent and continu-

ous at similar rates, resulting in a relative rotation

between the growing poikiloblast and the surrounding

matrix foliation. The most common product of this

process is depicted in Figure 17.29b where curved in-

clusion trails that deﬁne S

in the rotated poikiloblast

are continuous with and merge in the perimeter into

the external matrix S

. Deciding which was the actively

rotating fabric element—the poikiloblast or its matrix

—and how to deﬁne its rotation with respect to refer-

ence frames has been a subject of considerable debate.

This ambiguity is depicted in Figure 17.28 (although

drawn for a prekinematic fabric the principles are the

same). Garnet porphyroblasts in some schists show

substantial rotation, say, 150°, creating spiral inclu-

sion trails, which most geologists (e.g. Williams and

Jiang, 1999) have interpreted to have formed as the

spherical garnet rotated during growth in a rock sub-

jected to simple shear, hence the designation “snow-

ball” garnet (Figure 17.30). But this traditional view

has been challenged by T. H. Bell, S. E. Johnson, and

548 Igneous and Metamorphic Petrology

Muscovite

Biotite

Albite

(a)

(b)

Epidote

Actinolite  chlorite

Garnet

Quartz

0 0.5mm

Albite

17.29 Synkinematic fabrics. (a) Schist in which mica-rich folia bow around porphyroblasts of garnet. (b) Greenschist-facies metabasite. Palona

Schist, San Gabriel Mountains near Los Angeles, California. Thin section from collection of William D. Carlson, University of Texas at

Austin. The albite poikiloblast grew during dextral simple shear that is manifest in the sigmoidal (curved) trails of inclusions (S

) and the

inclusion-free albites in the asymmetric strain shadow (see also Figure 17.30c). Rotation of growing poikiloblast was clockwise with

respect to the external matrix foliation, S

co-workers (e.g. Johnson and Bell, 1996), who argue

that S-shaped inclusion trails result from progressive

growth over crenulated matrix foliation; although the

porphyroblasts may rotate relative to some reference

frame, this is not essential to their formation.

It may be more difﬁcult to accurately interpret

poikiloblast–matrix relations in synkinematic fabrics

that developed where the period of mineral growth

was longer than a brief pulse of deformation at the

beginning or end of the growth (see Figure 17.32b).

Similar difﬁculties arise where the deformation was the

longer of the two processes.

Postkinematic Fabrics. These are represented in many

cases by modiﬁcations of, or grain overgrowths onto,

pre-existing anisotropic fabrics. In Figure 17.31a and

b, depicting postkinematic fabrics, strain-free grains

that lack undulatory extinction under cross-polarized

light have mimicked and partially taken the place of

fabrics produced by deformation, respectively. Figure

17.31c is a crenulated foliation that has been partly

overgrown by a poikiloblastic garnet whose undulating

faithfully preserves the form of the S

. In detail, the

largely reacted phyllosilicate-rich folia might be repre-

sented by trains of small phyllosilicate ﬂakes within the

garnet and the quartz-rich layers by perhaps larger in-

clusions of quartz. There might be a possibility of con-

fusing the internally curved inclusion trails

with a synkinematic fabric but the clearly overprinted

crenulated matrix in Figure 17.31c is in contrast to the

straight S

in Figure 17.30. Large biotites in the foliated

chlorite-rich matrix of Figure 17.31d reveal no S

that

might provide a hint of an overgrown matrix foliation,

yet their random orientation and undeformed nature is

consistent with postkinematic growth.

17.3.4 Polymetamorphism

The single orogenic cycle presumed in foregoing dis-

cussions is compounded in most long-lived polymeta-

morphic orogens during successive docking of island

arcs and plateaus and locally culminating continent–

continent collisions in the history of plate convergence.

Multiple episodes of deformation, heating–burial and

uplift–cooling take place in many regional tectonite

terranes.

Accurate interpretation of polymetamorphic fabrics

is based on the principles discussed above for the

timing of recrystallization and deformation; however,

the effects of multiple overprinting must be dealt with.

Thus, a particular episode of grain growth might be

postkinematic relative to one foliation-forming event

but prekinematic relative to a subsequent one (Figure

17.32). Once again, poikiloblast–matrix relations can

be of critical importance because poikiloblasts can pre-

serve vestiges of matrix fabrics that have been erased

in the present surrounding matrix by subsequent

Evolution of Imposed Metamorphic Fabrics: Processes and Kinetics

549

Time

(d)

(c)

(b)

75°

180°

0°

17.30 Synkinematic “snowball” garnet poikiloblast in weakly foliated plagioclase–biotite–quartz gneiss. (a) Sample from near the ductile shear

zone that separates the Proterozoic Grenville and Archean Superior Provinces, near Sudbury, Ontario (see Figure 19.31). The spiral tex-

ture within the poikiloblast in lower center is expressed by a train of felsic inclusions incorporated from the asymmetric strain-shadow

zone into the growing garnet while it was being rolled, like a snowball, as shown schematically in (b)–(d) after the model of Schoneveld

(1977). Angles denote amount of rotation between S

and S

(a)

overprinting mineral reactions and/or deformation

(Special Interest Box 17.2 and Figure 17.34). In this

respect, poikiloblasts are like photographs of past

events that can be viewed at later times.

Three guidelines in dealing with polymetamorphic

fabrics can be kept in mind:

1. Apparent overprinting in a rock fabric may be the

result of only one continuous episode of deforma-

tion. Transposition of a pre-existing foliation forms

a new foliation that subsequently may be folded

and itself become transposed during continuation

of the same simple shearing process (Figure 17.33).

This happens in some ductile shear zones.

2. The converse of 1 also happens, namely, a second

episode of deformation merely enhances the effects

of a ﬁrst without producing a distinct overprint. A

rock that possesses a ﬂattening foliation created by

pure shear and expressed by parallelism of mica

ﬂakes might be subjected to a distinct later episode

of deformation of the same style and orientation

550 Igneous and Metamorphic Petrology

(a) (b)

(d)

(c)

17.31 Postkinematic fabrics produced by recrystallization following after deformation, as seen in thin section. (a) Individual ﬂakes in the

folia of micas deﬁning the fold are not bent or otherwise strained because each has a uniform optical extinction under cross-polarized

light. They can be interpreted to have mimicked a previously folded layer. (b) Strained plagioclase that shows bent twin lamellae and

undulatory extinction under cross-polarized light has been partially replaced by small, lower energy, strain-free polygonal grains. At least

two interpretations are possible. First, the texture is postkinematic and was produced by static annealing after stress-induced deforma-

tion. Second, the small grains may be a product of dynamic recrystallization during dislocation creep (compare Figures 17.26 and 15.15).

(c) Garnet poikiloblast contains wavy trains of inclusions in continuity with the crenulated folia in the matrix. This is consistent with a

postkinematic relation because, in addition to the continuity between S

and S

, the inclusion trains mimic the amplitude and wavelength

of the crenulated matrix and other poikiloblasts show a similar relationship. They are not likely to be synkinematically rotated garnets as

in Figure 17.30. (d) Oriented chlorite ﬂakes are transected by randomly oriented biotites that are interpreted to have formed during a

late episode of recrystallization in a hydrostatic stress ﬁeld.

under similar metamorphic conditions that only

strengthens the expression of the original foliation.

3. Multiple episodes of deformation can be diachro-

nous—occurring at different times in different

places within the orogen. One way this happens is

during oblique convergence of an oceanic plateau

as it rams at different times along strike against the

overriding continental margin. This diachronism

can present challenges in correlating deformation

and recrystallization events throughout a large oro-

gen, especially where it might have been dismem-

bered and parts displaced large distances during

later faulting.

17.3.5 Shear-Sense Indicators

All three components of deformation—translation,

rotation, and distortion—can be manifest in tectonites,

such as in mylonitic rocks formed in ductile shear

zones in the deep crust where there has been differen-

tial movement between adjacent rock masses. (Deep-

crustal, ductile shear zones are the counterpart of

brittle faults in the shallow cooler crust, as shown in

Figure 8.9.) The translation component is revealed in

offset markers, while distortion, or strain, is manifest

by changes in shape of various markers, such as cob-

bles and lava pillows in metamorphosed conglomerate

(Figure 14.13) and subaqueous lava ﬂows (Figure

14.20), respectively. A rotational component can be

seen in the fabric in Figure 17.30. Additional rotational

fabric elements that serve as kinematic indicators of the

sense of shear are depicted in Figure 17.35 (see also

Barker, 1990, pp. 89–103; Passchier and Trouw, 1996,

pp. 109–29; Snoke et al., 1998, pp. 305–67). In Figure

17.35, the direction of shear can be imprinted on shear

planes as some sort of parallel lines, which may be a

stretching lineation (Section 15.1.1) or aligned elongate

mineral grains. However, this lineation gives no actual

indication of the sense of shear, or which of the two

Evolution of Imposed Metamorphic Fabrics: Processes and Kinetics

551

(c)(b)(a)

(d)

Time

Phyllosilicates

Garnet

Quartz

Staurolite

17.32 Schematic poikiloblastic growth and deformation that preserves stages of foliation development during polymetamorphism. (a) Garnet

poikiloblast (shown as a circle for simplicity) encompasses inclusions aligned parallel to foliation S

. (b) Deformation of S

and creation

of a new crenulation foliation S

. Continued growth of garnet poikiloblast, incorporating into its rim curved trains of inclusions from

crenulated foliation. Note core that preserves relics of undeformed but now rotated S

. A new garnet nucleates and as it grows also

incorporates curved trains of inclusions from crenulated foliation. (c) S

is completely obliterated by creation of the new foliation S

except for preserved relicts in garnet poikiloblasts. Rectangular staurolite poikiloblast grows, incorporating inclusions mimicking S

but

not earlier S

that had been destroyed in the matrix. Redrawn from Johnson (1999). (d) Chronologic relations between mineral growth,

deformation, D, and development of foliations S

and S

552 Igneous and Metamorphic Petrology

Special Interest Box 17.2 Unraveling episodic

poikiloblast growth and folding: a case history

An example of how poikiloblast–matrix relations can

be used to determine the complex history of a poly-

metamorphic terrane is provided by a study of the

Fleur de Lys Supergroup, northern Newfoundland

(Stallard, 1998). This upper-greenschist to lower-

amphibolite facies terrane is believed to represent a

late Proterozoic–early Ordovician continental margin

developed on Grenville basement rocks (Section

19.5.1) overriden by thrust sheets during the Taconic

Orogeny. In outcrop, the dominant fabric is a folia-

tion, designated S

, expressed by compositional layer-

ing and parallel preferred orientation of mica ﬂakes;

it is parallel to axial planes of tight to isoclinal folds

and has been overprinted by later crenulation folia-

tions, S

and S

, and associated folds. The matrix

surrounding the garnet poikiloblasts is composed

of quartz, muscovite, biotite, ilmenite, garnet, albite,

epidote, chlorite, carbonate, and graphite; their

growth obliterated foliations earlier than S

Examination of numerous oriented thin sections

reveals a complex history of episodic poikiloblast

growth incorporating a succession of multiple crenu-

(a) (b)

(d)

(c)

Time

17.33 Schematic development of overprinting transposition folia-

tions and asymmetric folds by progressive simple shear. (a)

Body that contains a foliation S

is subjected to simple shear.

(b) Asymmetric folding and development of an incipient

axial plane foliation S

. (c) S

is transposed (Figure 14.19)

along S

shear surfaces; note relict fold hinges from (b).

(d) Further simple shear creates asymmetric folds in S

and

parallel remnants of S

; a new incipient axial plane folia-

tion S

appears. Only a few relict fold hinges from (b) are still

evident.

lation folds of evolving foliations (Figure 17.34).

Curved inclusion trails in poikiloblasts do not appear

to have resulted from rotation during syntectonic

growth. Foliations evolved by growth of matrix grains

punctuated by episodes of crenulation folding and

transposition. Four distinct growth phases of garnet

poikiloblasts, G

–G

, are recognized, but not in every

poikiloblast. In contrast, albite poikiloblasts grew

during only one late episode. Evidence for each

growth phase and their timing is based on inclusion

trail mineralogy, texture, geometry, and poikiloblast–

matrix relations. In garnet poikiloblasts, inclusion

trails of S

and S

have counterparts in the surround-

ing matrix but trails of S

, S

, and S

do not.

Each phase of garnet poikiloblast growth seems to

have begun early in a spurt of foliation development

and resulted in overgrowth on earlier foliation(s)

adjacent to the poikiloblast. As growth was arrested,

the foliation was subsequently intensiﬁed against the

rigid garnets, which grew mostly within quartz-rich

domains in the rock rather than across high-strain,

mica-rich domains of the actively developing folia-

tion. The micaceous folia might have functioned as

some sort of barrier for access of nutrient ions for

garnet growth.

possible motions produced the lineation. If the overly-

ing rock moved to the right relative to the underlying

rock in Figure 17.35 the sense of shear is right-lateral,

or dextral. If to the left, sinistral. To accurately deter-

mine the sense of shear we must examine a surface of

the tectonite that is oriented perpendicular to the shear

plane, which might be a foliation, and parallel to the

lineation (the direction of shear). In the absence of a

natural so-oriented examination surface, such as a

joint, it is necessary to collect an oriented hand sample

from the outcrop and then cut an examination surface

with a diamond saw. In any case, preparation of several

differently oriented surfaces, and usually thin sections,

may be necessary to verify the true kinematic pattern.

For example, Figure 17.35 shows that on some oblique

surfaces an offset marker gives the opposite, and,

therefore, false, sense of shear. Thus, an observation

made on a single oblique section can be misleading.

Determination of the direction and sense of shear

accompanying differential movement of rock masses,

often from microscopic textural features in tectonites,

can be of fundamental relevance in understanding the

tectonic evolution of metamorphic terranes and the evolu-

tion of continental-scale orogens where they develop.

This is yet another example of a recurring principle in

petrology, namely, attention to the smallest of details

can provide signiﬁcant information regarding large-scale

(a)

(c)

Crenulation

foliation

development

Poikiloblast

growth

Garnet

Albite

0mm1

(b)

17.34 Complex inclusion trails and chemical zoning in a poikiloblast from the Fleur de Lys Supergroup, Newfoundland record episodic growth

and deformation. (a) Photomicrograph under plane-polarized light. (b) Annotated drawing showing successive foliations (thin double

lines) and growth zones G

(contains 6–8 wt.% MnO), G

(2–6 MnO) that constitutes the bulk of the grain, and G

(2 MnO) as fringes

on left and right. Subtle sparse quartz and rutile inclusion trails (oriented up and down ﬁgure, not obvious) in growth zone G

may

represent relict foliation S

. Quartzose strain shadow around relict grain G

was crenulated and overgrown during G

, preserving relict

as curved inclusion trails of quartz. S

inside poikiloblast is continuous with S

in upper left and truncated by S

in lower right. S

and

are evident in matrix. Black grains are Fe–Ti oxides. C, retrograde chlorite replacing garnet. Photomicrograph and data courtesy

of Aaron Stallard. Reproduced with permission from Stallard A. 1998. Episodic porphyroblast growth in the Fluer de Lys Supergroup,

1998 by Blackwell Science.

(a)

True dextral

sense of shear

Apparent

sinistral offset

Shear direction

Shear

plane

(b)

(c)

(e)(d)

(g)

(h)

(f) S–C

fabric

17.35 Schematic shear-sense indicators and related features in non-

coaxial simple shear deformation. These kinematic indicators,

shown here for a dextral sense of shear, are most accurately

determined on an examination surface (light shaded) oriented

perpendicular to the shear planes (dark shaded), or tectonite

foliation, and parallel to the shear direction, manifest by pen-

etrative lineations. (a) Offset planar marker, such as a dike or

vein, that is oblique to the shear surfaces. Note that an oppo-

site, false sense of offset is apparent in a improperly oriented

examination surface. (b) Asymmetric crenulation cleavage

(Section 15.1.1 and Figure 15.3; see also Passchier and Trouw,

1996, p. 112 for distinguishing between cleavages formed

by coaxial versus noncoaxial shear). (c) Asymmetric folds.

(d) Fragmented porphyroclast that is stepped by an antithetic

(sinistral) sense of shear. (e) Lenticular mica “ﬁsh” showing

inclined (001) cleavage. (f) S–C fabric shown in Figure 17.36a.

(g) and (h) Mantled porphyroclasts typical of mylonites and

commonly of feldspar in quartzo-feldspathic protoliths. The

mantle surrounding the intact but highly strained porphyro-

clast is composed of a ﬁne-grained aggregate of the same phase

that was sheared off during ductile ﬂow of the surrounding

matrix and concentrated into the strain shadows on each side

of the porphyroclast. But because of its rotation in the non-

coaxially ﬂowing matrix, the mantle was drawn out along the

shear surface, forming asymmetric tails that are like either of

the Greek letters sigma () or delta ().

(a)

02cm

17.36 S–C fabric consisting of two simultaneously developed foliations serves as a indicator of sense of shear during noncoaxial deformation in

a ductile shear zone. (a) Polished slab of Santa Rosa mylonite, Palm Canyon, California, in a late Cretaceous–early Tertiary thrust zone

affecting granitoids of the Southern California Batholith. Mylonitization occurred at about 500°C and was created by a “general” shear

that consisted of a simple shear plus a component of pure shear normal to the compund foliation. Photograph provided by Carol Simp-

son. Reproduced from Simpson C. 1986. Determination of movement sense in mylonites. J. Geol. Ed., 34: 246–252, with the permission of

the National Association of Geology Teachers. (b) “Mica ﬁsh” in S–C fabric. The (001) basal cleavage deﬁnes S. (c) Possible explanation

of development of S–C fabric according to Takagi (1992).

geologic features. It can be of crucial importance to

decide, for example, whether a particular shear zone

resulted from convergent-thrust or divergent-extensional

motion between adjacent terranes, or whether the

motion was not dip slip but strike slip, and if so

whether it was right- or left-lateral, or oblique-slip?

S C Fabric. Compound foliations are another im-

portant kinematic fabric element, in addition to those

depicted schematically in Figure 17.35, that indicate

sense of shear. They develop simultaneously in shear

zones and are especially evident in quartzo-feldspathic

protoliths (Figure 17.36a). The two foliations, ﬁrst

described by Berthé et al. (1979), are designated as S-

surfaces (from the French schistosité for the plane of

ﬂattening; compare German Schiefer for schist) and C-

surfaces (French cisaillement for the plane of shearing).

S is deﬁned by aligned mica ﬂakes, ribbonlike quartz,

and feldspar aggregates. The oblique (45°) C-surfaces

are parallel to the overall shear zone and consist of

spaced, thin, discrete zones made of ﬁner, more highly

strained grains. The sigmoidal form of S between C-

surfaces as well as the aligned mica ﬂakes (“mica ﬁsh”

in Figure 17.36b, c) along S indicate the sense of sim-

ple shear on them. A stretching lineation is common on

the C-foliation. Although S–C fabric resembles crenu-

lation foliation found in low-grade phyllosilicate-rich

rocks (Figures 14.18 and 15.3; Section 15.1.1) and

transposition foliation (Figure 14.19), these two folia-

tions are superposed on distinctly earlier surfaces of

anisotropy, whereas the two foliations in S–C fabric

develop simultaneously.

17.3.6 Patterns of Deformation and Flow:

Tectonic Signiﬁcance of Fabric Geometry

Interpretations of the tectonic signiﬁcance of anisotropic

fabrics are a concern of the structural geologist (e.g.

Twiss and Moores, 1992), yet this cannot be ignored by

petrologists seeking some understanding of the evolu-

tion of metamorphic terranes. One basic question that

554 Igneous and Metamorphic Petrology

(b)

Mica ﬁsh

(001)

(c)

Initial

undeformed

state

Distributed

continuous simple

shear; S-surface

parallel to plane

of ﬂattening

Simple shear along

spaced discrete

C-surfaces; composite

S–C fabric

might be posed is how the orientation of the L- and

S-fabric elements—the lineation and foliation in the

tectonite—relate to the principal axes of stress in the

nonhydrostatic stress ﬁeld that caused the deformation

(Section 8.1). Accurate and complete evaluation of all

three components of deformation is generally im-

possible in real rocks because of the lack of suitable

markers. Furthermore, the observed fabric of a rock

reﬂects a complex strain history, or deformation path,

that is the sum of all the increments of deformation it

has experienced throughout its entire history, from the

initial to the ﬁnal state (Box 17.1). If the initial fabric

of the undeformed rock was perfectly homogeneous and

isotropic and the deformation was perfectly coaxial,

then the resulting strain can be homogeneous. But real

rock bodies are almost never isotropic and they are

certainly not homogeneous on most scales; all sorts of

linear and planar discontinuities are mechanically and

rheologically signiﬁcant in stressed rock. Furthermore,

strictly coaxial deformation is generally deemed to be

uncommon in metamorphic rock systems. Thus, it is a

long interpretive jump from the observed pattern of

the fabric to the state of stress, little of which can be

inferred unambiguously. Principal axes of stress do

not necessarily coincide with the axes of strain ellip-

soids in deformed rocks.

In many instances, the dominant foliation in a meta-

morphic rock is parallel to the plane of ﬂattening in the

ellipsoid (Figures 14.12 and 17.37). Simple ﬂattening

yields only a foliation, as in S-tectonites, and simple

extension (constrictional strain) yields only a lineation,

as in rare L-tectonites. At least two possibilities exist

for widespread L–S tectonites. Foliated and lineated

fabrics that have no evidence for rotational movement,

or have shear-sense indicators in opposite directions,

were possibly produced by pure shear. If, on the other

hand, the L–S tectonite shows a consistent sense of

shear (Figure 17.35), then deformation by simple shear

is likely. In this pattern of ductile ﬂow, the symmetry of

the overall fabric is monoclinic, with the single sym-

metry plane perpendicular to the foliation and to the

axis of rotated crystals and/or parallel to a stretching

lineation, in contrast to the orthorhombic symmetry of

the pure shear fabric.

Only in cases of extreme strain do shear planes

approach parallelism with the plane of ﬂattening in the

strain ellipsoid (Figure 17.37). Foliations are not always

planes of shear, one exception being the C-plane in C–S

fabrics (Figure 17.36).

17.4 ORIGIN OF ANISOTROPIC FABRIC

IN METAMORPHIC TECTONITES

The problematic origin of anisotropic tectonite fabric

has intrigued geologists since the infancy of the science

of geology. The problem is not only petrological but

structural as well because, for example, foliation is

commonly parallel to the axial plane of associated folds

and lineation parallel to hinge lines.

Field and laboratory investigations have shown that

imposed anisotropic fabrics in tectonites result from

penetrative ductile ﬂow in nonhydrostatic stress ﬁelds.

Variations in the details of tectonite fabrics imply a

variety of factors affecting their origin, including the

character of the applied nonhydrostatic stress, the

magnitude of the overall strain, the rate of strain, T,

the inherited fabric of the body (whether anisotropic

or isotropic), its grain size, and mineral and chemical

composition, especially the availability of water or

other ﬂuids. Moreover, imposed fabrics can be inho-

mogeneous on different scales, probably reﬂecting

inhomogeneities in one or more of these factors.

Tectonite fabrics can have any one or more of the

following properties:

1. A preferred dimensional orientation of typically in-

equant mineral grains, such as oriented amphiboles

(Figure 15.8) and oriented mica ﬂakes (Figure

17.29a), as well as ﬂattened and elongate grains of

minerals that are typically equant, such as quartz

and feldspar (Figures 17.14 and 15.17).

2. A preferred orientation of crystallographic axes

and planes, that is, the crystal lattice. In typically

inequant mineral grains, such as amphiboles and

phyllosilicates, a shape orientation as in property 1

automatically, in most cases, ensures a lattice orien-

tation. However, typically equant minerals low in

the crystalloblastic series such as quartz, carbon-

ates, and feldspars can possess strong lattice pre-

ferred orientation without shape orientation, and

vice versa.

3. A compositional layering, expressed usually by

planar variations in concentration of felsic and maﬁc

minerals (Figure 15.2, see also Figure 19.24a) or of

felsic and phyllosilicate minerals (Figure 15.3).

Current thinking regarding the origin of these three

types of anisotropy is discussed next.

Evolution of Imposed Metamorphic Fabrics: Processes and Kinetics

555

17.37 Progressive deformation by simple shear on surfaces parallel

to the plane of ﬂattening of an initial (large) ﬁnite strain ellipse.

The plane of ﬂattening, represented by the maximum axis of

the ellipse, rotates away from the shear surfaces but toward the

plane of ﬂattening of the smaller reference ellipse that shows

the progress of strain in the experiment. Redrawn from Twiss

and Moores (1992).

17.4.1 Preferred Dimensional Orientation of

Mineral Grains

There are four mechanisms that produce shape

anisotropy in grain aggregates, depending on the

mineralogical composition of the rock and the meta-

morphic conditions.

1. Bodily rotation of inequant grains. Relatively rigid

inequant grains can rotate in a more ductile matrix

toward the plane of ﬂattening by either pure (Fig-

ure 17.38) or simple shear. The grains do not

necessarily change shape during rotation. Ductility

in the matrix can be a consequence of constant-

volume plastic deformation of weaker grains or

volume-reducing pressure solution of more soluble

minerals. Very large strains are necessary to pro-

duce strong preferred orientations solely by rota-

tion if the inequant grains were initially random.

Grains unfavorably oriented with respect to the

strain pattern do not rotate.

2. Intracrystalline plasticity. Plastic slip and twinning

are veriﬁable mechanisms of producing changes in

grain shape and preferred orientation of grain

shapes in aggregates where none existed prior to

deformation. Several hundred percent extension is

possible in simple shear by slip in a conﬁned grain

(Figure 17.15) but the strain is more limited in

twinning and kinking. Not only does the external

form of the grain change shape, preferred orienta-

tion of grain lattices also develops by internal rota-

tion of crystallographic planes and axes (discussed

later). Development of preferred orientation in a

grain aggregate is modeled in Figure 17.19.

3. Differential pressure solution and precipitation.

In the presence of an intergranular ﬂuid, mineral

grains can be dissolved in highly stressed parts and

the solute precipitated elsewhere, either on the

same grains in strain shadows or at some distant

site, such as in veins. Neighboring relatively less

soluble grains that are inequant may be bodily

rotated as the matrix around them collapses during

volume loss. In crenulation foliation and other

types of spaced cleavages, differential solution

transfer may remove silica from the limbs of the

crenulations, leaving higher concentrations of less

soluble minerals that become aligned in the newly

imposed foliation (Figure 15.3). It is expected that

this sort of solution activity would occur in ﬁne-

grained rocks, such as slates and phyllites where

diffusion paths are short. Nonetheless, crenula-

tions and apparent associated differential solution

is evident in some coarser grained, higher grade

schists.

4. Recrystallization. During recrystallization, grains

advantageously oriented in the prevailing nonhy-

drostatic stress ﬁeld may grow, whereas others less

favorably oriented may be selectively eliminated. In

this competitive growth system, elongate mineral

grains tend to grow their longest dimension in the

direction of least work against the least compressive

stress, 

, and their shortest dimension in the 

direction. In mimetic recrystallization, new grains

nucleate and grow on former, perhaps smaller, ori-

ented grains of similar atomic structure, consuming

them in the process, thus enhancing the expression

of preferred orientation. Mimetic recrystallization

may be involved in a simple prograde sequence at

increasing T, turning shale into slate, then phyllite,

and ﬁnally schist. At each stage, ﬁne-grained, oriented

phyllosilicates are replaced by coarser phyllosilicate

ﬂakes of similar orientation (Section 14.1.1).

17.4.2 Preferred Orientation of Crystal Lattices

in Tectonites

For aggregates made of shape-oriented amphiboles,

phyllosilicates, and other minerals that are inherently

inequant, lattices also possess a preferred orientation.

Thus, in mica-rich schists, c-axes of micas are inher-

ently oriented mostly perpendicular to the foliation.

But our concern here is with the origin of lattice

orientation in aggregates of typically equant felsic and

carbonate minerals low in the crystalloblastic series

with weak anisotropy in their surface energies. A dra-

matic illustration of lattice preferred orientation can

be seen in thin sections of quartz-rich ultramylonites

under cross-polarized light with a gypsum or mica

retardation plate inserted; in alternate 90° orientations

the aggregate is mostly green/blue or mostly yellow/

orange, virtually as if it were a single crystal of quartz.

556 Igneous and Metamorphic Petrology

17.38 Bodily rotation of randomly oriented inequant grains embed-

ded in a ductile matrix deformed by homogeneous pure shear.

The grains rotate toward the plane of ﬂattening and deﬁne a

foliation. Note that grains initially oriented at a high angle to

the plane rotate less.