Myron G. Best. Igneous and metamorphic 2003 Blackwell Science

Подождите немного. Документ загружается.

Metamorphic Rocks and Metamorphism: An Overview

407

at the solidus is dependent on rock composition,

regimes of metamorphism and partial melting in het-

erogeneous rock bodies overlap hundreds of degrees.

Such overlapping conditions are thought by many

petrologists to be represented in migmatites that con-

tain leucocratic granite intimately mingled with more

refractory maﬁc metamorphic rock (Figure 11.24). At

the lower limit, the distinction is equally blurred be-

tween incipient metamorphism and diagenetic changes

in sedimentary and volcaniclastic deposits. In this tran-

sition, the typically very ﬁne grained rocks are generally

not amenable to study using thin sections but require

other laboratory techniques, such as X-ray diffractom-

etry, transmission electron microscopy, and reﬂected

light microscopy (Frey and Robinson, 1999). The trans-

itional continuum between diagenesis and metamor-

phism can be seen in thick sedimentary and volcanic

sequences in arc-related depositional basins where P

and T increase with depth of burial. The next two

paragraphs describe the nature of this continuum in

volcaniclastic and clay deposits.

Volcanic glass is metastable from the moment of

its formation but can persist for very long periods of

time, especially in dry climates. However, vitroclasts

deposited in saline lakes can be diagenetically altered

to zeolites in a few thousand years. In addition, during

burial at progressively increasing P and T, glass in

vitric tuffs can be replaced by zeolites and analcite

(Figure 14.2). Stabilization of analcite with quartz, or

of the zeolites heulandite and clinoptilolite with quartz,

is generally considered to be indicative of the onset of

metamorphism resulting from burial. Large variations

in the depth of burial necessary for incipient meta-

morphic equilibration have been found by drilling into

active geothermal areas and from stratigraphic rela-

tions in exhumed basin deposits. In geothermal areas,

such as the Salton Sea in southern California and

the Wairakei-Broadlands in northern New Zealand,

Table 14.1. (Continued).

Orthopyroxene Opx (Mg,Fe)SiO

Paragonite* Pg Na

(Si

)(OH)

Periclase Per MgO

Phengite Phe K

(Al,Mg,Fe)

6–7

2–1

(OH,F)

Phlogopite* Phl K

(OH)

Plagioclase Pl (Ca,Na)Al

2–1

2–3

Prehnite Prh Ca

Al(AlSi

)(OH)

Pumpellyite Pmp Ca

(Mg,Fe

2

,Mn)(Al,Fe

3

,Ti)

O(OH)

(Si

)

(SiO

)

·2H

Pyrite Py FeS

Pyrope* Prp Mg

Pyrophyllite Prl Al

(OH)

Pyrrhotite Po Fe

1–x

S where x  0–0.125

Quartz Qtz SiO

Rutile Rt TiO

Sapphirine Spr (Mg,Fe

2

,Fe

3

,Al)

(Al,Si)

Scapolite Scp (K,Na,Ca)

(Al,Si)

(Cl,CO

,SO

)

Serpentine Srp Mg

(OH)

Siderite Sd FeCO

Sillimanite Sil Al

SiO

Spessartine* Sps Mn

Spinel* Spl MgAl

Staurolite St (Fe

2

,Mn,Zn)

(Al,Fe

3

,Ti)

[(Si,Al)O

]

(O,OH)

Stilpnomelane Stp (K,Na,Ca)

0.6

(Mg,Fe

2

,Fe

3

)

Al(O,OH)

·2–4H

Talc Tlc Mg

(OH)

Titanite (sphene) Ttn CaTiSiO

Tourmaline Tur complex boron silicate

Tremolite* Tr Ca

(OH,F)

Vesuvianite (idocrase) Ves Ca

(Al,Fe)

(Mg,Fe)

(Si

)

(SiO

)

(O,OH,F)

Wairakite Wa CaAlSi

·H

Wollastonite Wo CaSiO

Zoisite* Zo Ca

O·AlOH(Si

)(SiO

)

Chemical compositions from Deer et al. (1997). Select chemical analyses listed in Appendix A. Abbreviations from Bucher and Frey (1994).

Some minerals(*) are end-members of solid solutions that can also occur in nearly pure form. Clay minerals are in italics.

increases during burial to more than 100°C, smectites

are gradually replaced by more stable, less water-rich

illites, another complex clay mineral group. Interlay-

ered clay particles (0.0039 mm in diameter) of illite

and smectite are common. The expulsion of water—

both physically entrapped between clay grains during

deposition and chemically bound in the clay minerals

—and conversion to higher T minerals is one manifes-

tation of diagenesis, making mud into shale. (Some

geologists prefer to use the term mudrock for any rock

that is made mostly of clay minerals and to restrict the

name shale for laminated or platy mud rocks. The inter-

action between clay minerals and organic material is

explored in Special Interest Box 14.1.) Still higher

temperatures, to perhaps as much as 300°C, convert

mixed-layer clays into chlorite, while illites are recon-

stituted into sericite. This is a ﬁne-grained white mica

which may include muscovite, paragonite, and pyro-

phyllite but is usually phengite, a high-silica musco-

vite in which there is coupled substitution of Si and

2

or Mg for 2Al so that the atomic ratio Si/Al 3.0

(Table 14.1; see also Dempster, 1992). These chlorites

and white micas that are produced at the onset of

metamorphism are harder, less hydrous, and generally

slightly coarser grained than the clay minerals in shale

and are typical of the aphanitic platy metamorphic

rock called slate. Water liberated during diagenesis

and incipient metamorphism carries dissolved ions

leached from the unstable recrystallizing clays and

other phyllosilicates. Thus, these changes at the onset

of metamorphism are not strictly isochemical.

In magmatic rocks where the water fugacity is very

low, primary high-T minerals, such as feldspars and

pyroxenes, can persist indeﬁnitely over a fairly wide

range of metamorphic pressures and temperatures.

However, where the water fugacity is sufﬁciently high

under subsolidus conditions, hydrous metamorphic

minerals are stabilized and can partially to completely

replace primary minerals, depending on the availability

of water. Water not only stabilizes new phases but also

catalyzes mineral reactions by enhancing kinetic rates

of atomic diffusion and crystal growth.

The water required for this metamorphism (or altera-

tion, as some geologists prefer to call it) can have

different sources. In closed magma systems, juvenile

water exsolved from the crystallizing melt reacts at

subsolidus temperatures with the primary magmatic

minerals so that the rock effectively “stews in its own

juices” during autometamorphism or deuteric alteration.

This is particularly common in granitic intrusions

where primary minerals are partially re-equilibrated to

lower-T hydrous minerals (Plate IV). In open magma

systems, water may be drawn into a cooling dike from

water-bearing wallrock or water may enter a cooling

lava ﬂow emplaced in a lake or the ocean, as at the

global spreading ridge system. Tensile fractures created

408 Igneous and Metamorphic Petrology

reconstituted tuffs and sedimentary rocks occur at

depths as little as 1–2 km, based on drilling. In basins

with lower geothermal gradients, reconstitution occurs

at depths 8 km. Hence, T appears to be a more signi-

ﬁcant factor in reconstitution than P. The presence of a

separate aqueous ﬂuid phase is also necessary because

zeolites and analcite contain more water (10 and

about 8 wt.% H



, respectively; Appendix A) than is

typically found in silicic glass. Also, the concentration

of CO

in the ﬂuid phase cannot be very large or car-

bonate minerals are stabilized instead of zeolites and

analcite. Figure 14.2 shows delicate glass shards that

have been perfectly preserved through the conversion

into analcite. Growth of the ﬁne-grained metamorphic

mineral grains has not erased the original vitroclastic

texture of the tuff that is preserved as a relict fabric.

Kaolinites and more widespread smectites, of which

montmorillonite is a principal variety, are complex clay

minerals formed by weathering of feldspars and other

alumino-silicates and are a major constituent of soils

and of mud deposited in sedimentary basins. Chemic-

ally, clay minerals are typiﬁed by their relatively high

contents of H

O and Al (Tables 14.1 and A.2 in

Appendix A). Rocks made largely of clays, or their

metamorphic equivalents, are called pelites. As T

Feldspar

phenocrysts

10mm

Relict

glass shards

replaced by analcime

14.2 Relict vitroclastic texture in a vitric–crystal tuff subjected to

incipient burial metamorphism. Compare this photomicro-

graph under plane polarized light of a rock from New South

Wales, Australia (Wilkinson and Whetten, 1964) with Figure

7.32b. Shards of glass have been entirely replaced by analcime

(analcite). Fine-grained quartz, alkali feldspar, heulandite

zeolite, chlorite, and montmorillonite occur in the matrix

around the relict shards and phenocrysts. Black circles are air

bubbles in the thin section cement. Thin section provided

courtesy of John Whetten.

Metamorphic Rocks and Metamorphism: An Overview

409

Special Interest Box 14.1 Low-grade meta-

morphism of mudrock and the origin and

migration of oil

Oil is created from incompletely oxidized organic

debris, chieﬂy plant material, that is deposited with

clay and other ﬁne-grained sediments in sediment-

ary basins. With time, the mud and organic debris

become buried to greater depths within the basin as

they are covered by younger sediments. As burial

depth increases, T also increases. As T rises to about

50°C, the organic debris converts to a dark, solid,

complex hydrocarbon called kerogen. With further

burial, as T rises from about 50 to 100°C, the solid

kerogen transforms into the liquid hydrocarbon

called crude oil.

At this point, crude oil is a sticky, viscous liquid

that is ﬁnely dispersed among and stuck to the

grains of clay and other sedimentary particles in the

basin. An oil well drilled into this material would

recover no oil because the crude oil is too widely

dispersed and the molecules adhere to the sediment-

ary grains. Therefore, the oil cannot ﬂow through

the rock to the well. For the oil to become recover-

able, it must ﬁrst migrate out of this source rock and

concentrate in reservoir rock. What causes the oil to

move from the clay-rich source rock and eventually

concentrate elsewhere?

A clay mineral called smectite is the most abund-

ant clay formed by weathering and thus the most

abundant clay deposited in sedimentary basins.

Smectite converts metamorphically into another clay

called illite as the T rises from about 50 to 100°C

during burial in sedimentary basins. Therefore, illite

is the most abundant constituent of shales.

Smectite contains about 40% water by volume;

illite contains only a few percent. Thus, as smectite

converts to illite, large quantities of water are for-

cibly expelled from the rock. This expulsion of water

during low-grade metamorphism of clay occurs in

the same T range in which kerogen transforms to

liquid oil. The forcibly expelled water then ﬂushes

the oil from the clay-rich source rock and causes

it to migrate into reservoirs. [Hydraulic fractures

likely play an important role in this migration.]

Calculations of the amount of energy required to

convert kerogen to oil in sedimentary basins show

that, in most basins, there is not enough thermal

energy available to form liquid oil in the T range

50–100°C. The conversion should not occur until

much higher temperatures are attained. Yet 50–

100°C is the T range of observed oil formation in

most sedimentary basins.

Recent experiments and theoretical work show

that the reaction in which smectite converts to illite

during cooling and contraction facilitate advective

penetration of water into dikes and lava ﬂows.

In hydrating changes, exact stoichiometrically bal-

anced reactions are difﬁcult to write because of their

complexity. The following serve only to indicate in

a general manner the fate of the high-T, primary

magmatic mineral(s) on the left that convert into a

lower-T, secondary, more hydrated subsolidus mineral

or mineral assemblage on the right. Note the liberated

elements that can be removed as soluble ions in an

aqueous ﬂuid.

Primary magmatic → Secondary subsolidus

mineral(s)  water mineral(s)

biotite  water → chlorite  rutile

(or titanite)  K  Si

hornblende  water → chlorite  rutile

(or titanite)  Si  Ca

calcic clinopyroxene → actinolite or epidote

 water

olivine/orthopyroxene → serpentine  Fe-oxides

 water

plagioclase  Ca  Fe → epidote

 water

feldspars  water → sericite  Si  K

(higher T )

feldspars  water → clay minerals  Si

(lower T )  Ca  Na

Quartz remains stable under a wide range of geologic

conditions.

acts as a catalyst that causes transformation of kero-

gen to oil at much lower temperatures than would

be necessary otherwise. A catalyst is a substance

or process that causes or accelerates a chemical

reaction without being permanently changed by the

reaction.

This research on clay mineral reactions in sedi-

mentary basins suggests that, in many of the world’s

great oil-producing basins, low-grade burial meta-

morphism of clay minerals is important to both the

formation of oil and the migration of oil from

source rocks to reservoirs.

Excerpt from Modern Physical Geology by

Graham R. Thompson and Jonathan Turk, copyright

by permission of the publisher.

410 Igneous and Metamorphic Petrology

In a thin section of a metamorphosed andesite (Fig-

ure 14.3), the formation of new, more stable hydrous

minerals has not obliterated the original magmatic

aphanitic–porphyritic texture. Because of this pre-

served relict magmatic texture, perhaps fortiﬁed by

ﬁeld relations, there is little doubt of the ancestry of the

rock. Hence, the metamorphic rock may accurately be

called a meta-andesite.

Note the important contrast in the nature of the

equilibrating changes in the foregoing examples. In

the shale, reconstitution was driven by increasing T

and the ensuing endothermic mineral reactions liber-

ated water while producing less hydrous minerals. On

the other hand, in initially “dry” magmatic protoliths,

water was added to stabilize hydrous minerals at

subsolidus temperatures (Figure 5.31).

14.1.2 Recrystallization under Hydrostatic

Conditions: Newly Imposed Granoblastic Fabric

In most metamorphic equilibration processes, new

fabrics are progressively overprinted over protolith

fabrics, obliterating them. Only in ﬁner-grained, less

metamorphosed rocks are relict features preserved.

The newly imposed solid-state growth fabrics are,

for the most part, distinctly different from magmatic

fabrics and from the texture of clastic sedimentary

rocks. We ﬁrst examine metamorphic fabrics created

under hydrostatic states of stress. The inﬂuence of non-

hydrostatic states is considered in Section 14.1.3.

Recrystallization as used in this textbook refers to

solid-state production of new mineral grains from

pre-existing ones. Two distinct processes can be

recognized:

1. Recrystallization sensu stricto. Boundaries of exist-

ing grains are texturally modiﬁed in some way. No

new phases are created. Static heating can coarsen

grain size whereas recrystallization of strained

grains can yield smaller grains, the stored “strain

energy” derived from the work of deformation

being an important driving force for grain bound-

ary modiﬁcation.

2. Solid-state crystallization. Nucleation and growth

of crystalline grains of a new phase or phases are

stabilized by changing metamorphic conditions, such

as formation of white mica and chlorite from illite

and smectite in the shale to slate transition cited

above. Once nuclei of the new phase or phases are

viable, requisite ions for grain growth diffuse from

nearby decomposing unstable, or reacting, mineral

grains. Some sort of ﬂuid is commonly involved in

the mineral reaction.

Increase in Grain Size without Changes in Constituent

Phases. Increasing T, or an elevated T maintained for

a signiﬁcant period of time, can modify grain size and

shape as grain boundaries adjust to a minimal surface

energy (Section 6.6.2). Aggregates experience Ostwald

ripening whereby smaller, higher energy grains are

(a)

Plagioclase

Pyroxene

(b)

Relict

plagioclase

phenocryst

Fe–Ti

oxide

mm01 mm01

Relict

pyroxene

phenocryst

14.3 Thin section views of andesite and meta-andesite. (a) Andesite made of phenocrysts of clinopyroxene and plagioclase in a felty matrix of

glass, feldspar microlites, Fe–Ti oxides (black), and pyroxene. (b) Low-grade meta-andesite that has a relict aphanitic–porphyritic

texture. Note similarity in shapes of relict pyroxene and plagioclase phenocrysts to fresh phenocrysts in (a) even though they are replaced

(pseudomorphically) by secondary minerals, including epidote after clinopyroxene and epidote, albite, and sericite after plagioclase.

Metamorphic Rocks and Metamorphism: An Overview

411

consumed at the expense of larger, more stable ones.

An example of grain coarsening is seen in the con-

version of limestone to marble (Figure 14.4), where

protolith features such as fossils deﬁned by ﬁne-

grained arrays are completely erased. A familiar analog

of grain coarsening is found in aggregates of soap bub-

bles in which, over time, bubbles increase in size while

decreasing in number (Figure 14.5). Grain boundary

adjustments can also change grain shapes. For aggreg-

ates of common felsic and carbonate minerals, the

equilibrium shape, as viewed in two-dimensional sec-

tions, is ﬁve-, six-, or seven-sided polygons with triple

grain boundary junctions forming angles near 120°

(Figures 6.23 and 6.24).

Marble and the olivine rock in Figure 6.24 illustrate

granoblastic texture that results from grain boundary

equilibration in the solid state. This texture, also

seen in polyphase rocks (Figure 14.6c, d), consists of

an isotropic aggregate of polygonal grains of more

or less similar size, perhaps within an order of magni-

tude. Inequant grains, such as micas, are randomly

oriented.

The blastic sufﬁx on some names for metamorphic

fabrics is derived from the Greek blast, meaning lump.

11 minutes 49.5 156 225

(a) (b)

4 cm

3 cm

14.4 Contrasting textures at about the same scale of (a) ﬁne-grained, gray fossiliferous limestone and (b) coarse-grained granoblastic white

marble. Textural features in gray limestones are deﬁned by aggregates of minute black carbonaceous particles that are eliminated by

reaction with water during metamorphism, releasing CO

and CH

from the rock (see Special Interest Box 16.2). Thorough recrystal-

lization by grain boundary adjustments of calcites and growth of larger grains at the expense of smaller ones has obliterated the texture

of the original limestone. The end product is a coarse white marble.

14.5 Increase in size of soap bubbles in a ﬂat cell as a function of time. Numbers are minutes after creation of bubbles by agitation. Note the

120° triple junctions that mimic the grain boundary conﬁguration in granoblastic texture (compare Figure 6.24). The evolution of the

bubbles serves as a model for Ostwald ripening, or grain coarsening, during recrystallization in rocks that can perhaps require millions

of years instead of minutes. Redrawn from Smith (1954).

(a)

Chlorite

Calcite

Fe–Ti oxide with

rim of sphene

(b)

Chlorite

Plagioclase

Actinolite

Sphene

Relict

plagioclase

0.5 mm 0.5 mm

Actinolite

Epidote

Calcite

(c)

Plagioclase

Hornblende

(d)

Plagioclase

Fe–Ti

oxide

Fe–Ti

oxide

0.5 mm 0.5 mm

Pyroxene

14.6 Prograde thermal metamorphism of diabase under essentially hydrostatic stress conditions. (a) Weakly metamorphosed diabase (meta-

diabase) or greenstone. Relict ophitic fabric (Figures 7.15 and 7.22) is well preserved because recrystallization has not produced large

enough grains to obliterate original grain boundaries. Magmatic pyroxenes have been replaced by aggregates of randomly oriented,

chemically zoned actinolites. Embedded within the actinolite aggregate are patches of ﬁne-grained chlorite and tiny epidotes. Ti-bearing

oxides have partially reacted with mobile Ca and Si to produce rims of sphene (titanite). Original lath-shaped calcic plagioclases

(labradorite) are partly replaced by aggregates of minute epidote (high relief ) and white mica grains; remaining plagioclase is more

albitic. Local large calcite grains represent relict amygdules. (b) Greenstone or ﬁne-grained amphibolite. An isotropic aggregate of the

same minerals comprising (a), except here grains formed by solid-state growth are larger with simpler, cleaner outlines. The original

magmatic fabric is virtually obliterated; only vaguely deﬁned rectangular areas of untwinned albite and epidote aggregates suggest the

former existence of magmatic plagioclases. (c) Amphibolite. Coarser, well developed granoblastic fabric. All vestiges of the original

magmatic fabric have been erased. (d) Granoblastic plagioclase–pyroxene granofels. Mineral phases in this high-grade rock have

compositions of a near-solidus gabbro but the texture is clearly metamorphic (compare Figure 7.16a).

412 Igneous and Metamorphic Petrology

Crystallization Yielding New Minerals and New Fabrics.

Subsolidus equilibration generally yields new mineral

phases as well as new imposed fabrics, both stable

under new conditions. This is well illustrated by the

progressive metamorphism at increasing T under high

water fugacities of an ophitic diabase in Figure 14.6.

In the initial stages of subsolidus crystallization at

low temperatures, primary magmatic labradorites are

mutual competition for space brings into play the

differing stabilities of grain boundaries (Section 6.6.2),

rather than growth velocities in different crystallo-

graphic directions as occurs in an isolated crystal freely

suspended in a melt.

There is, therefore, a signiﬁcant contrast between

the metamorphic fabric of the granoblastic amphibo-

lite and plagioclase–pyroxene granofels (Figure 14.6c

and d, respectively) and the mineralogically similar

magmatic rocks shown in Figures 7.15 and 7.16. This

contrast centers on plagioclase—the dominant mineral

in magmatic rocks as well as many metamorphic rocks.

In magmatic rocks, plagioclases are typically euhedral

to subhedral tablets—their characteristic habit result-

ing from uninhibited growth in a melt. Polysynthetic

twinning and compositional zoning (Figures 5.14d and

6.17) developed during growth are present in virtually

all magmatic plagioclases. In metamorphic rocks, on

the other hand, equilibrium plagioclase shapes are dic-

tated by the demands of minimal surface energy dur-

ing solid-state recrystallization, typically resulting in

anhedral polygonal outlines. Twinning tends to be less

conspicuous in plagioclases grown wholly in the solid

state. Zoning, if present, is generally more subtle and

oscillatory zoning is absent.

In some recrystallizing rocks, limited nucleation pro-

duces grains of a phase that are substantially larger than

grains of other more readily nucleating phases in the

rock. Although this fabric resembles the phenocryst-

matrix relation in porphyritic magmatic rocks, its sub-

solidus origin is signiﬁcantly different and warrants a

different name—porphyroblastic. The large, generally

euhedral to subhedral crystals called porphyroblasts

are typically some type of metamorphic alumino-

silicate such as garnet, staurolite, andalusite, or kyanite

(Figures 14.8–14.10). Garnet and andalusite are rare

in magmatic rocks and kyanite and staurolite non-

existent. Whereas quartz is a common phenocryst in

magmatic rocks, it is nonexistent as a porphyroblast

in metamorphic rocks. Feldspars, micas, amphiboles,

pyroxenes, and olivine are common phenocrysts but

only locally are they porphyroblasts.

Growth of isolated porphyroblasts that differ in

composition from the pre-existing protolith volume

which they now occupy can be considered as a local

grain-scale metamorphic differentiation. Thus, the

volume of rock occupied by the large kyanites in the

metamorphosed shale (Figure 14.9) lacks the K, Na,

Fe, Ca, and H

O occurring in the surrounding matrix,

which has less Al than the kyanite. There was substan-

tial movement of different ions into and away from the

growing kyanite porphyroblasts.

Porphyroblasts commonly contain inclusions of

other minerals (Figure 14.10; see also Figure 14.25c), in

which case they are called poikiloblasts. They can

furnish signiﬁcant insight into the metamorphic history

Metamorphic Rocks and Metamorphism: An Overview

413

replaced by a sodic plagioclase, commonly albite. This

albitization liberates Ca and lesser Al which allow

formation of phases such as calcite (if CO

fugacity is

also appropriately high) as well as hydrous Ca–Al sili-

cates including prehnite, pumpellyite, epidote, and

zoisite (the Al analog of Fe-bearing epidote). Fe–Ti

oxides can be converted into titanite. Primary clino-

pyroxene is replaced by actinolite (Figure 14.7) and

possibly chlorite. At higher temperatures, the hydrous

Ca–Al silicates, albite, and aluminous chlorite react

together, creating calcic plagioclase and aluminous

amphibole (hornblende). At still higher temperatures,

water-liberating reactions and recrystallization yield an

anhydrous assemblage of pyroxene and plagioclase that

is stable as the system approaches the basalt solidus.

Once all of the relict magmatic grains have been

consumed to form the granoblastic aggregate of new

phases the fabric is distinctly metamorphic.

Although metamorphic fabrics produced by sub-

solidus recrystallization may superﬁcially resemble

magmatic fabrics produced by crystallization of a melt,

especially in rocks of similar modal and mineralogical

composition, they are really quite distinct. In the

former, growing minerals compete with one another

for common space in the solid aggregate, whereas only

in the closing stages of crystallization of magmas where

melt fractions are small does this competition occur. In

wholly solid-state recrystallization, the ever-prevailing

14.7 Overgrowth and replacement of primary magmatic clinopy-

roxene by a bundle of subparallel amphibole needles. Devel-

opment of this speciﬁc texture is called uralitization but is a

type of epitaxy or epitaxial growth in which a secondary phase

nucleates and grows on a crystalline substrate that has a sim-

ilar atomic structure and inﬂuences the orientation of the

overgrowth. In this case, amphibole and clinopyroxene are

both chain silicates.

414 Igneous and Metamorphic Petrology

of a rock. For example, some included minerals may

not be the same as those in the surrounding matrix and

might be armoured relics of former reactant phases

incorporated into the growing poikiloblast that were

eliminated in the matrix through progressive mineral

reactions. Other poikiloblasts have encompassed fabric

features, such as microfold hinges (see Figure 17.34),

developed during an earlier episode of deformation

that were erased elsewhere in the rock during progres-

sive recrystallization. Thus, poikiloblasts can preserve

records of past mineral reactions and deformation, in

the same way that photographs provide records of

bygone events.

14.1.3 Recrystallization under Nonhydrostatic States

of Stress: Tectonite Fabric

The rheologic response of rocks to nonhydrostatic

stress (Sections 8.1 and 8.2) is such that rocks under

near-surface conditions of low P and T behave in a brit-

tle manner at high strain rates. In shallow crustal faults

(Figure 8.9; see also Twiss and Moores, 1992, Chapter

4), brittle rocks are broken, crushed, and pulverized

by cataclasis to form dilatant, unconsolidated fault

breccia or ﬁne-grained clay-rich gouge. Subsequent

percolation of mineral-laden groundwater along the

fault may cement the loose particles into a coherent

cataclasite rock (Figure 8.7a). The clasts in usually

Andalusite

10mm

14.10 Poikiloblasts of andalusite in slate, Sierra Nevada Foothills,

California. Photomicrograph in plane-polarized light. Minute

graphite inclusions in the form of a cross in the tetragonal

prisms of andalusite deﬁne the variety called chiastolite. These

poikiloblasts lie in a very ﬁne-grained matrix of biotite 

chlorite  white mica  quartz but are absent in the lighter-

colored, relict sedimentary bed made mostly of quartz that had

insufﬁcient Al to stabilize the Al

SiO

polymorph.

14.9 Porphyroblastic kyanite–mica (pelitic) schist north of Kaladar,

Ontario. Porphyroblasts of kyanite lie in a ﬁne phaneritic

matrix of white mica, biotite, and quartz.

14.8 Porphyroblastic garnet amphibolite, Barton Mine, Gore

Mountain, near North Creek, New York. Porphyroblasts of

almandine-rich garnet, some the size of soccerballs (30 cm dia-

mter), are surrounded by a rim of virtually pure hornblende

(black hornblendite) several cm thick. Amphibolite matrix is

made of white plagioclase and hornblende. Camera lens cap

for scale.

Metamorphic Rocks and Metamorphism: An Overview

415

randomized (isotropic) cataclastic fabric have sharp,

angular shapes and many are polygranular. Through-

going fractures cut across many grains. Very locally,

brittle comminution (pulverization) in concert with

frictional melting at high strain rates may create

pseudotachylite (Special Interest Box 11.1).

At greater depths in the crust at elevated P and T,

metamorphic rocks respond to nonhydrostatic stress,

especially at low strain rates, by a more continuous

and homogeneously distributed solid-state ductile

deformation or, as it is frequently called, ductile ﬂow,

because of its overall resemblance to viscous ﬂow. Duc-

tility is accomplished by straining (changing the shape

of ) individual grains in the rock through movement of

individual ions or packets of ions in a systematic way,

including grain boundary migration, recrystallization,

and plastic slip. During ductile ﬂow the rock retains its

cohesion—it does not break apart as in brittle deform-

ation. Downslope movement of glaciers is an example

of solid-state ductile ﬂow (Special Interest Box 14.2

and Figure 14.11).

Ductile deformation resulting from nonhydrostatic

stress is responsible for the development of imposed

anisotropic fabrics in metamorphic rock bodies. (It

may be recalled from Section 7.9 that an anisotropic

fabric has different attributes in different directions in

the rock body.) Such fabrics are often claimed to be the

hallmark of metamorphism, serving as a distinguishing

criterion from the effects of diagenesis and alteration.

The most common anisotropic fabric is a planar foli-

ation, a term derived from the Latin folium, which

means “leaf.” Schist is a typical metamorphic rock

that possesses foliation. In some rocks, a linear fabric

feature, or lineation, lies on foliation surfaces.

Special Interest Box 14.2 Glaciers as

metamorphic bodies

Ice is a mineral, and a solid aggregate of ice crystals

is a rock, large bodies of which are called glaciers.

Although requiring cold climatic conditions to

form, glaciers otherwise satisfy the criteria of a

metamorphic rock. They deform during downslope

movement under their own weight by solid-state

ductile ﬂow, producing folds in dirty layers within

glaciers (Figure 14.11b) that resemble folds in deep

crustal metamorphic bodies (Figure 8.4). To permit

ﬂow, individual ice crystals experience internal

plastic deformation and recrystallization, not unlike

that which occurs in tectonites. Thin sections of

glacier ice show well developed granoblastic fabric

characteristic of metamorphic rocks (Figure 14.11a).

As in other metamorphic rocks, melting occurs in

glacier ice if temperatures become sufﬁciently high.

14.11 Glaciers as metamorphic rock bodies. (a) Photomicrograph under cross-polarized light of granoblastic glacier ice. Each grid in this thin

section is 1.0 cm. Small black spots are air bubbles in the ice. Photograph courtesy of A. J. Grow, US Army Cold Regions Research and

Engineering Laboratory. (b) Aerial photograph of the southern margin of the Barnes Ice Cap, Bafﬁn Island, Canada showing ductile-ﬂow

folds in dirty layers. Photograph courtesy of Canadian National Air Photo Library.

(a) (b)

416 Igneous and Metamorphic Petrology

In many metamorphic rocks, ductile deformation

is accompanied by concurrent recrystallization; com-

monly one enhances the other. The elevated temperat-

ures necessary for ductile ﬂow also promote mineral

reactions and grain boundary adjustments. However,

some metamorphic rocks experience deformation after

recrystallization or vice versa; in still other polymeta-

morphic rocks multiple episodes of each process are

recorded. Because deformation and recrystallization

typically occur during contractional tectonism in oro-

genic belts, the Alpine geologist Bruno Sander in the

mid-1900s introduced the term tectonite for rocks

possessing anisotropic tectonite fabric (Turner, 1981).

Because tectonites are exposed over vast areas of

deeply eroded ancient orogens the question immedi-

ately arises whether there is a consistent geometric

relation between the orientation of the prominent

foliation, and lineation if present, and the directions of

the principal axes of nonhydrostatic stress (Section 8.1)

that produced it. Whether such a relation exists has

been one of the enduring controversies in geology, un-

like the simple relation that exists for magmatic dikes

intruded into hydraulic extensional fractures that

formed perpendicular to 

(Figure 8.2). The question

of the geometric relation between fabric orientations

and stress directions is examined in Section 17.3.6. For

now, we make the provisional working assumption that

foliations are more or less parallel to the plane of ﬂat-

tening (Twiss and Moores, 1992, p. 331), or the plane

that parallels the maximum and intermediate axes of

the triaxial strain ellipsoid (Figure 14.12). Metamor-

phic lineations can have different geometric relations

to the pattern of deformation depending on their par-

ticular nature and origin. One common possibility is

that the lineation parallels the direction of extension,

or the maximum axis of the strain ellipsoid.

It was once believed that nonhydrostatic states of

stress stabilized certain minerals, such as kyanite, that

were created during concurrent solid state crystalliza-

tion. But all such “stress minerals” have been shown to

grow in the laboratory under hydrostatic conditions, so

the concept has been abandoned. Nonetheless, defor-

mation can exert a catalytic effect on recrystallization

because the work of straining grains imparts additional

energy that can overcome activation energy barriers,

hastening the kinetics of mineral reactions and elimin-

ating otherwise metastable phases.

Examples of tectonite development are considered

next.

Foliated and Lineated Metaconglomerate. One of the

most obvious compound effects of nonhydrostatic

stress and recrystallization is found in a metaconglom-

erate that contains ductilely deformed relict cobbles in

a quartz-mica matrix. In the original conglomerate

protolith, quartzite cobbles approximated spheres, or

at least nearly equidimensional ellipsoids whose ratio

of major to minor axial dimension was no more than

two or three. In the metaconglomerate tectonite, they

are intensely ﬂattened and elongated ellipsoids whose

ratio is as much as 30 (Figure 14.13). Examination in

thin section of the quartz grains making up the relict

cobbles reveals a similar ﬂattened and elongate shape

that resulted from recrystallization and ductile strain in

a nonhydrostatic stress ﬁeld. The integrated effect of

these ellipsoidal grains that make up individual relict

cobbles and the integrated effect of their common

orientation through the metaconglomerate rock body

expresses a scale-independent foliation and lineation.

In other words, the tectonite body possesses the same

foliated and lineated tectonite fabric on all scales, from

the outcrop to the hand sample to the thin section. It is

not unusual to ﬁnd similar fractal-like scale invariance

extending to entire mountain ranges in terranes of

metamorphic tectonites (Figure 14.14).

Mylonite. Mylonite was ﬁrst described in the late 1800s

by Charles Lapworth from the Moine thrust of north-

ern Scotland (see Figure 14.30). Since then, the origin

of this type of tectonite has been controversial (e.g.,

Snoke et al., 1998). Relative to its protolith, mylonite

is much ﬁner grained and has enhanced anisotropic

fabric. Mylonite occurs as sheets as thin as a cm or less

to broad zones as much as 10 km thick. The common

Trace of

TRUE FOLIATION

Trace of

TRUE LINEATION

foliation

14.12 Block diagram of foliation and lineation in a hypothetical tec-

tonite related to strain ellipsoid. Redrawn from Twiss and

Moores (1992). The strain ellipsoid (upper left) represents the

result of homogeneous deformation of a reference sphere in

the initial undeformed state. Foliation (shaded) is parallel to

plane of ﬂattening in the strain ellipsoid; this plane is deﬁned

by the longest axis a and the intermediate axis b and is per-

pendicular to the shortest axis c. Lineation that lies within

the foliation is parallel to the direction of elongation, or the

longest ellipsoid axis a. On randomly oriented rock surfaces

the obliquely intersecting foliation appears as line traces that

are not a true lineation.