1.1 Introduction
I wrote this book to help you to interpret what you see when you look at thin and polished sections of rocks with the microscope. I say ‘help’, rather than ‘teach’, because I do not want to give the impression that every microstructure can be easily and unambiguously interpreted in terms of processes that produced the rock. Many can be, but in many other instances, conventional interpretations are ambiguous or poorly understood. So I intend the book to be only a guide, and I present alternative ideas where appropriate. A healthy scepticism should be maintained when interpreting rock microstructures yourself and when reading the interpretations of others.
1.2 History of the Examination of Rocks with the Microscope
Rocks in natural outcrops, in samples knocked off these outcrops and in drill cores, are beautiful and instructive. We can see different minerals, and identify many of them with the aid of a hand lens. We can also see some of the more obvious structures in the rocks. However, cutting a slice (section) through a rock with a diamond-impregnated circular saw and polishing the sawn surface shows us the various minerals alongside each other, rather than piled confusingly all around each other. This reveals the structure even more clearly, as can be seen in the polished facing slabs on many buildings and bench tops.
But we always want to see more. So, when D. Brewster, in 1817, and William Nicol, in about 1830, showed how to make a slice of crystalline material thin enough to transmit light (0.03 mm is the standard thickness) and stuck it to a glass microscope slide (Shand, Reference Shand1950, p. 6; Loewinson-Lessing, Reference Loewinson-Lessing1954), it was not surprising that a curious person, such as Henry Sorby, should start looking at these thin sections of rocks (Sorby, Reference Sorby1851, Reference Sorby1853, Reference Sorby1856, Reference Sorby1858, Reference Sorby1870, Reference Sorby1877, Reference Sorby1879, Reference Sorby1908). Sorby learnt the technique of making thin sections from W. C. Williamson in 1848 (Judd, Reference Judd1908; Folk, Reference Folk1965) and made the first rock thin section in 1849 (Judd, Reference Judd1908). Sorby was the first to look seriously at rock sections with the microscope, beginning with a study of chert, a siliceous sedimentary rock that was a very appropriate choice for microscopic investigation in view of its very fine grain size. He described and suggested a mechanical origin for slaty cleavage (Sorby, Reference Sorby1853, Reference Sorby1856); noticed many of the basic features of igneous and metamorphic rocks; made many important observations on sedimentary rocks, including carbonate rocks (Sorby, Reference Sorby1879); investigated pressure solution (using fossil crinoids); described meteorites; and published the first papers on the examination of polished sections of metals with the microscope (Sorby, Reference Sorby1864, Reference Sorby1887). Thus, Sorby is not only the founder of petrography (the description of rocks), but also the founder of metallography as well (Smith, Reference Smith1960). In 1858 he investigated fluid inclusions in minerals, heating crystals to watch the gas bubbles disappear, in order to get an estimate of the temperature of crystallization of the mineral (Folk, Reference Folk1965).
Sorby was followed soon after by many others, as discussed by Johannsen (Reference Johannsen1939) and Loewinson-Lessing (Reference Loewinson-Lessing1954). Prominent among them were Zirkel (Reference Zirkel1863, Reference Zirkel1866, Reference Zirkel1876), who learnt the technique of making thin sections from Sorby, Vogelsang (Reference Vogelsang1867), Fouqué & Michel-Lévy (Reference Fouqué and Michel-Lévy1879), Rosenbusch (Reference Rosenbusch1873, Reference Rosenbusch1877), Allport (Reference Allport1874) and Teall (Reference Teall1885, Reference Teall1886, 1888). Since those days, the light microscope has become the main tool for identifying minerals and examining their microstructures, though it has been augmented by many modern techniques (Section 1.6).
1.3 How Relevant Is the Microscope Today?
Many petrologists concentrate on the mineralogical and chemical aspects of rocks, without spending much time looking at rocks with the microscope. In fact, in these days of marvellous techniques for the chemical and isotopic analysis of minerals, some people feel that simply looking at and measuring the shapes and arrangements of crystals in rocks with the microscope is a little out of date. However, carrying out detailed chemical and isotopic analyses of minerals when you do not understand the relationships of these minerals to other minerals in the rock is a waste of expensive resources, at the very least.
On the other hand, many structural geologists look at the physical or structural aspects of minerals and rocks, especially from the viewpoint of deformation processes and preferred orientations of grains, without being concerned about the chemical aspects of these processes. Both approaches are valuable, of course, but their interrelationships can be particularly illuminating. Fortunately, many researchers are attempting to integrate the chemical and physical approaches, and the study of rocks with the microscope provides a link between them. In fact, the detailed study of processes in rocks at the microscopic scale is now a major area of research, especially among younger people, in many universities and other research institutions. Moreover, new observational techniques are being developed and used, as discussed in Section 1.6.
Research microscopes commonly have both transmitted and reflected light facilities. An excellent example of the simultaneous use of transmitted and reflected light microscopy is the study of Columbia River basalts by Long & Wood (Reference Long and Wood1986), in which reflected and transmitted light photos are arranged side by side, clearly revealing the dendritic shapes of the opaque Fe–Ti oxide minerals and their relationships to the transparent and translucent silicate minerals. Some leading books and review articles on minerals in reflected light, with emphasis on microstructures, are those of Edwards (Reference Edwards1947, Reference Edwards1952), Bastin (Reference Bastin1950), Cameron (Reference Cameron1961), Ramdohr (Reference Ramdohr1969), Stanton (Reference Stanton1972), Craig (Reference Craig, Jambor and Vaughan1990a, Reference Craig, Jambor and Vaughan1990b), and Craig & Vaughan (Reference Craig and Vaughan1994).
1.4 Mineral Identification
Learning to identify minerals takes time and practice, and is outside the scope of this book. Close teaching in a laboratory situation is the best way to learn about the optical properties of minerals, using textbooks specifically written for the purpose (e.g., Fleischer et al., Reference Fleischer, Wilcox and Matzko1984; Shelley, Reference Shelley1985a; Gribble & Hall, Reference Gribble and Hall1992; Deer et al., Reference Deer, Howie and Zussman1992). Ideally, this should go hand-in-hand with learning about microstructures.
1.5 The Concept of a Section
Thin and polished sections are two-dimensional sections through three-dimensional objects, and this must always be kept in mind, as explained in some detail by Hibbard (Reference Hibbard1995). Mineral grains can have unexpectedly complex three-dimensional shapes (e.g., Rigsby, Reference Rigsby1968; Byron et al., Reference Byron, Atherton, Cheadle and Hunter1994, Reference Byron, Atherton and Hunter1995, Reference Byron, Atherton, Cheadle and Hunter1996). Two or even three orthogonal sections may be necessary to reveal the structure of structurally anisotropic rocks, and some recent detailed microstructural studies have used (1) serial sectioning (e.g., Byron et al., Reference Byron, Atherton, Cheadle and Hunter1994, Reference Byron, Atherton and Hunter1995, Reference Byron, Atherton, Cheadle and Hunter1996; Johnson & Moore, Reference Johnson and Moore1996), coupled with image analysis by computer, in order to construct a three-dimensional image of the microstructure and (2) computed X-ray tomography (Section 1.6), to reveal the three-dimensional distribution of large crystals (porphyroblasts) in metamorphic rocks (Denison & Carlson, Reference Chernoff and Carlson1997), plagioclase chains in basalts (Philpotts et al., Reference Philpotts, Brustman, Shi, Carlson and Denison1999) and former melted rock (leucosome) in migmatites (Brown et al., Reference Brown2002).
1.6 Newer Techniques
This book deals mainly with microfabrics visible in the optical (light) microscope, either in thin or polished section, using polarized light. However, some newer techniques are also very useful for revealing features not apparent or less clearly shown in polarized light as outlined. Several examples of photos taken using these techniques will be presented in the book. The new techniques underline the fact that the microscopic study of rocks is a dynamic, progressive field of research.
1.6.1 Cathodoluminescence
Cathodoluminescence (CL) is a technique that can reveal internal microfabrics of grains of some minerals, for example, compositional zoning, microcracking and replacement veining in quartz, calcite, dolomite, zircon, plagioclase, K-feldspar, diamond and apatite (Sippel & Glover, Reference Sippel and Glover1965; Smith & Stenstrom, Reference 409Smith and Stenstrom1965; Sprunt, Reference Sprunt1978, Reference Sprunt1981; Zinkerngel, Reference Zinkerngel1978; Field, Reference Field1979; Sprunt & Nur, Reference Sprunt and Nur1979; Hanchar & Miller, 1984; Matter & Ramseyer, Reference Matter, Ramseyer and Zuffa1985; Owen & Carozzi, Reference Owen and Carozzi1986; Reeder & Prosky, Reference Reeder and Prosky1986; Marshall, Reference Marshall1988; Morrison & Valley, Reference Morrison and Valley1988; Ramseyer et al., Reference Ramseyer, Baumann, Matter and Mullis1988; Yardley & Lloyd, Reference Yardley and Lloyd1989; Hopson & Ramseyer, Reference Hopson and Ramseyer1990; Barker & Kopp, Reference Barker and Kopp1991; Shimamoto et al., Reference Shimamoto, Kanaori and Asai1991; Mora & Ramseyer, Reference Mora and Ramseyer1992; Williams et al., Reference Williams, Buick and Cartwright1996; D’Lemos et al., Reference D’Lemos, Kearsley, Pembroke, Watt and Wright1997; Watt et al., Reference Watt, Wright, Galloway and McLean1997, Reference Watt, Oliver and Griffin2002; Hayward, Reference Hayward and Jambor1998; Ahn & Cho, Reference Ahn and Cho2000; Janousek et al., Reference Janousek, Bowes, Braithwaite and Rogers2000; Müller et al., Reference Müller, Seltmann and Behr2000; Pagel et al., Reference Pagel, Barbin, Blanc and Ohnenstetter2000; Rubatto & Gebauer, Reference Rubatto, Gebauer, Pagel, Barbin, Blanc and Ohnenstetter2000; Barbarand & Pagel, Reference Barbarand and Pagel2001; Hermann et al., Reference Hermann, Rubatto, Korasakov and Shatsky2001; Penniston-Dorland, Reference Penniston-Dorland2001; Peppard et al., Reference Peppard, Steele, Davis, Wallace and Anderson2001; Rubatto et al., Reference Rubatto, Williams and Buick2001; Rougvie & Sorensen, Reference Rougvie and Sorensen2002; Rusk & Reed, Reference Rusk and Reed2002; Viljoen, Reference Viljoen2002). CL is especially useful for revealing microstructural details in minerals that are colourless in the light microscope, for example, calcite, quartz and feldspar. It can be used with the light microscope or the scanning electron microscope, and some applications are discussed in Sections 3.12.7, 3.12.9 and 5.9.3. CL is combined with X-ray topography to reveal the internal structure of diamonds (e.g., Field, Reference Field1979).
CL is caused by defect structures in the crystal lattice, such as impurity atoms, vacancies and dislocations produced during formation and/or deformation of the mineral, which therefore reflect conditions of crystallization, deformation and alteration. The technique involves coating a polished thin section with carbon and bombarding it with electrons in a vacuum, which produces light from substitutional atoms in an excited state.
1.6.2 Laser-Interference Microscopy
This is a relatively new optical technique that detects small differences in refractive index, and so can reveal in great detail subtle compositional differences (on which refractive index depends), for example, in zoned plagioclase (Chao, Reference Chao1976; Pearce, Reference Pearce1984a, Reference Pearce1984b; Pearce et al., Reference Pearce, Griffin and Kolisnik1987a, Reference Pearce, Russell and Wolfson1987b).
1.6.3 Scanning Electron Microscopy
Scanning electron microscopy (e.g., Lloyd, Reference Lloyd1987) is capable of revealing sharp microstructural details in shades of grey, though arbitrary colours may also be assigned, to form a false-colour image. It involves backscattered and forescattered imaging in the scanning electron microscope (SEM). This is particularly useful for (1) revealing the detailed microstructure of small grains and fine-grained aggregates and intergrowths (e.g., Vernon & Pooley, Reference Vernon and Pooley1981; Wirth & Voll, Reference Wirth and Voll1987; Cashman, Reference Cashman1988; Simpson & Wintsch, Reference Simpson and Wintsch1989; Swanson et al., Reference Swanson, Naney, Westrich and Eichelberger1989; Johnson & Carlson, Reference Johnson and Carlson1990; van der Voo et al., Reference van der Voo, Fang, Wang, Suk, Peacor and Liang1993; Brodie, Reference Brodie1995; Harlov & Wirth, Reference Harlov and Wirth2000; Blundy & Cashman, Reference Blundy and Cashman2001; Drüppel et al., Reference Drüppel, von Seckendorff and Okrusch2001; Rickers et al., Reference Rickers, Raith and Dasgupta2001; de Haas et al., Reference de Haas, Nijland, Valbracht, Maijer, Verschure and Andersen2002; Schieber, Reference Schieber2002), (2) identifying very fine-grained minerals (e.g., Prior et al., Reference Prior, Boyle, Brenker, Cheadle, Day, Lopez, Peruzzo, Potts, Reddy, Spiess, Timms, Trimby, Wheeler and Zetterström1999), (3) revealing fine-scale compositional zoning in minerals (e.g., Yardley et al., Reference Yardley, Rochelle, Barnicoat and Lloyd1991; Müller et al., Reference Müller, Seltmann and Behr2000; Piccoli et al., Reference Piccoli, Candela and Rivers2000; Alexandrov, Reference Alexandrov2001; Hermann et al., Reference Hermann, Rubatto, Korasakov and Shatsky2001; Kuritani, Reference Kuritani2001; Rubatto et al., Reference Rubatto, Williams and Buick2001; Ginibre et al., Reference Ginibre, Kronz and Wörner2002a, Reference Ginibre, Wörner and Kronz2002b; Lentz, Reference Lentz2002), (4) measuring orientation differences between grains and subgrains as small as 1 μm across (Prior et al., Reference Prior, Trimby, Weber and Dingley1996; Lloyd et al., Reference Lloyd, Farmer and Mainprice1997; Trimby et al., Reference Trimby, Prior and Wheeler1998;Prior et al., Reference Prior, Boyle, Brenker, Cheadle, Day, Lopez, Peruzzo, Potts, Reddy, Spiess, Timms, Trimby, Wheeler and Zetterström1999; Wheeler et al., Reference Wheeler, Prior, Jiang, Speiss and Trimby2001) and (5) revealing domains of different orientation in optically isotropic minerals, such as garnet (Prior et al., Reference Prior, Wheeler, Brenker, Harte and Matthews2000, Reference Prior, Wheeler, Peruzzo, Speiss and Storey2002; Spiess et al., Reference Spiess, Peruzzo, Prior and Wheeler2001) and pyrite (Boyle et al., Reference Boyle, Prior, Banham and Timms1998).
1.6.4 Transmission Electron Microscopy
The interpretation of some optical microstructures can be ambiguous, for example, some recovery features in deformed quartz (Section 5.4) and fine exsolution lamellae (Section 4.9). In such instances the transmission electron microscope (TEM) can provide more reliable information. The principles and some applications have been reviewed by Champness (Reference Champness1977), Putnis and McConnell (Reference Putnis and McConnell1980), McLaren (Reference McLaren1991) and Putnis (Reference Putnis1992). Transmission electron microscope resolves much smaller objects, such as very fine to submicroscopic intergrowths, and can reveal the arrangement of defects (including dislocations, discussed in Section 5.3.2) in the atomic structure of individual grains of both optically transparent and opaque minerals (e.g., McLaren et al., Reference McLaren, Retchford and Griggs1967; McLaren & Retchford, Reference McLaren and Retchford1969; Green, Reference Green, Radcliffe, Heard, Borg, Carter and Raleigh1972; Phakey et al., Reference Phakey, Dollinger, Christie, Heard, Borg, Carter and Raleigh1972; McLaren & Hobbs, Reference McLaren, Hobbs, Heard, Borg, Carter and Raleigh1972; McLaren, Reference McLaren, Mackenzie and Zussman1974, Reference McLaren1991; Champness & Lorimer, Reference Champness, Lorimer and Wenk1976; McLaren & Etheridge, Reference McLaren and Etheridge1976; Champness, Reference Champness1977; Zeuch & Green, Reference Zeuch and Green1984; Doukhan et al., Reference Doukhan, Doukhan, Koch and Christie1985; Allen et al., Reference Allen, Smith and Buseck1987; Cox, Reference Cox1987a; Couderc & Hennig-Michaeli, Reference Couderc and Hennig-Michaeli1989; Hennig-Michaeli & Couderc, Reference Hennig-Michaeli and Couderc1989; Green, Reference Green and Buseck1992; Ando et al., Reference Ando, Fujino and Takeshita1993; Doukhan et al., Reference Doukhan, Sautter and Doukhan1994; Vogelé et al., Reference Vogelé, Cordier, Sautter, Sharp, Lardeaux and Marques1998).
1.6.5 X-ray Tomography
A more recent development in the study of rock microstructure is the use of high-resolution computed X-ray tomography. This technique maps the variation of X-ray attenuation within solid objects, the attenuation varying with each mineral present. A source of X-rays and a set of detectors revolve around the rock sample, producing images in layers or cross-sections. The series of two-dimensional images can be computed into a three-dimensional representation of the grains and aggregates in the rock, which gives a clearer picture of spatial relationships and crystal size distributions (e.g., Carlson & Denison, Reference Carlson and Denison1992; Carlson et al., Reference Carlson, Denison and Ketcham1995, Reference Carlson, Denison and Ketcham1999; Carlson & Denison, Reference Denison and Carlson1997; Denison et al., Reference Denison and Carlson1997; Brown et al., Reference Brown, Brown, Carlson and Denison1999; Philpotts et al., Reference Philpotts, Brustman, Shi, Carlson and Denison1999).
1.6.6 Computer-Aided Construction of Three-Dimensional Images
Serial two-dimensional optical or X-ray tomographic images can be scanned and imported into suitable computer graphics programs to provide three-dimensional constructions (Johnson & Moore, Reference Johnson and Moore1993, Reference Johnson and Moore1996; Carlson et al., Reference Carlson, Denison and Ketcham1995, Reference Carlson, Denison and Ketcham1999; Pugliese & Petford, Reference Pugliese and Petford2001). Readily available computer software can also be used to animate the images, producing a more complete visualization of features, such as grain shapes, grain distributions and vein networks (e.g., Johnson & Moore, Reference Johnson and Moore1996; Carlson et al., Reference Carlson, Denison and Ketcham1999; Pugliese & Petford, Reference Pugliese and Petford2001).
1.6.7 X-ray Compositional Mapping
Maps of compositional zoning in crystals (Sections 3.12 and 4.12) are produced by multiple stage-scan chemical analyses made with wavelength-dispersive spectrometers on an electron microprobe, different colours being assigned to different concentrations of the analysed element. Examples are shown in Section 4.12. The technique can also be used for more clearly revealing mineral or compositional domains in fine-grained aggregates (e.g., Clarke et al., Reference Clarke, Daczko and Nockolds2001; Lang & Gilotti, Reference Lang and Gilotti2001; Williams et al., Reference Williams, Scheltema and Jercinovic2001; Daczko et al., Reference Daczko, Clarke and Klepeis2002a, Reference Daczko, Stevenson, Clarke and Klepeis2002b). Raw X-ray intensity maps can be converted to maps of oxide weight per cent by appropriate matrix corrections (Clarke et al., Reference Clarke, Daczko and Nockolds2001).
1.7 Quantitative Approaches
Though most work on rock microstructures is qualitative, involving description and interpretation, quantitative methods are also used. For example, grain measurement is important in the classification and interpretation of clastic sedimentary rocks in terms of transport and depositional environments (Section 2.2.2). Grain size is also used in the classification of igneous rocks, though less precisely, and crystal size distributions are being increasingly investigated in igneous and metamorphic rocks (Sections 3.5 and 4.3.1). Numerical modelling has been used to convert two-dimensional measurements of grain shapes and sizes in thin section to three-dimensional grain shapes and true crystal size distributions (Higgins, Reference Higgins1994; Peterson, Reference Peterson1996). Moreover, computer software is readily available to do this and to make animated images, as mentioned in Section 1.6. Interfacial angles have been measured in many metamorphic rocks, sulphide rocks and igneous cumulates, as indicators of mutual solid-state growth of minerals (Section 4.2). In addition, the orientations of inclusion trails in porphyroblasts have been used as indicators of tectonic processes (Section 5.10). Numerical simulation of the development of metamorphic and deformation microstructures is also well under way (Jessell, Reference Jessell1988a, Reference Jessell1988b; Jessell et al., Reference Jessell, Bons, Evans, Barr and Stüwe2001).
1.8 Some Terms
Though no hard and fast rule exists, it is probably best to use crystal for a volume of crystalline mineral with well-formed, planar faces (called crystal faces or facets), and grain for any other volume of crystalline mineral. For me, the shapes, arrangements and orientation of the minerals constitute a rock’s fabric. At the microscope scale, the fabric (microfabric) consists of the grain shapes and arrangement (the microstructure) and the spatial orientation of the minerals (the preferred orientation). However, many people use ‘fabric’ for ‘preferred orientation’, which is the usage recommended by the International Union of Geological Sciences (IUGS) Subcommission on the Systematics of Metamorphic Rocks (Brodie et al., Reference Brodie, Fettes, Harte and Schmid2002).
It would be good to get materials scientists more interested in rocks, as they are the great class of natural solid materials. Therefore, because ‘texture’ means ‘preferred orientation’ to most materials scientists and some structural geologists, it would be best not to use it instead of ‘microstructure’ as many petrologists do. However, though ‘microstructure’ is gaining in popular usage, ‘texture’ is common, and no ambiguity is caused among petrologists by using it. Actually, ‘microstructure’ appears to have priority, because the first publications on the microscopic examination of rocks referred to ‘microscopical structure’ or ‘microscopic structure’ (e.g., Sorby, Reference Sorby1851, Reference Sorby1858; Allport, Reference Allport1874). Moreover, the IUGS Subcommission on the Systematics of Metamorphic Rocks has recommended that the term ‘texture’ be replaced by ‘microstructure’, which is defined as ‘structure on the thin section or smaller scale’ (Brodie et al., Reference Brodie, Fettes, Harte and Schmid2002). The term ‘microtexture’, which unfortunately is starting to enter the literature, is unnecessary because ‘texture’ mainly refers to the microscopic scale.
Of course, every gradation in scale exists between the microscopic and mesoscopic (outcrop) scales, and so I have not been able to confine the discussion to the microscopic, although this is by far the main scale discussed.
1.9 Traditional Rock Groupings
Many rock-forming processes apply to more than one of the traditional igneous, sedimentary and metamorphic rock groups. For instance, similar basic principles governing the nucleation and growth of crystals apply to all rocks, and grain growth in the solid state occurs not only in metamorphic rocks (in which it is a universal process) but also in the late stages of formation of some rocks conventionally regarded as igneous. In addition, growth of new minerals in the solid state (neocrystallization) occurs not only in metamorphic rocks but also in the late-stage alteration (deuteric alteration) of igneous rocks, and in the low-temperature alteration (diagenesis or burial metamorphism) of rocks that many people would consider to be still sedimentary. Moreover, metamorphic rocks begin to melt at high temperatures, producing rocks with both igneous and metamorphic features. In addition, radiating crystal aggregates (‘spherulites’) commonly grow in glass, which, though technically solid, is liquid-like with regard to its atomic structure. Furthermore, exsolution, which is a solid-state process, occurs in both igneous and metamorphic minerals. As if that is not enough, fragmental material thrown out of explosive volcanoes produces rocks that are technically sedimentary, but consist entirely of igneous material, and may also show evidence of solid-state flow of glass. The result of this cross-linking of processes is that, though this book adheres roughly to the traditional sedimentary-igneous-metamorphic subdivision, processes discussed under one of these headings may also be relevant to another of these groups. These instances are cross-referenced.
1.10 Importance of Evidence
Science relies on evidence. An assertion made without evidence is not worth very much. Yet I often read statements such as: ‘the microstructural (textural) evidence indicates …’ This implies that the writers are asserting that their interpretations are so obviously right that they do not have to go to the bother of describing what they saw and evaluating the evidence.
Of course, recognizing evidence takes practice. As noted by A. F. Chalmers in What Is This Thing Called Science? ‘It is necessary to learn how to see expertly through a telescope or microscope, and the unstructured array of bright and dark patches that the beginner observes is different from the detailed specimen or scene that the skilled viewer can discern.’
Whenever you make interpretations based on microscopic examination of rocks, you should (1) describe clearly what you see and (2) evaluate the possible interpretations. If one or more interpretations are valid, you should not arbitrarily favour one of them, unless other evidence (e.g., field or chemical evidence) clearly points in that direction. This is the ‘method of multiple working hypotheses’ advocated by Chamberlain (Reference Chamberlain1890). In many instances, the microstructural evidence may not be at all clear, in which case, you should not use it to support a hypothesis. Maybe you will have to suggest equally valid alternative interpretations and leave it at that.
The paramount importance of evidence in making scientific inferences is emphasized in the following quotation.
On so important a question, the evidence must be airtight. The more we want it to be true, the more careful we have to be. No witness’s say-so is good enough. People make mistakes. People play practical jokes. People stretch the truth for money or attention or fame. People occasionally misunderstand what they’re seeing. People even sometimes see things that aren’t there.
Carl Sagan (The Demon-Haunted World) was referring to UFOs, but at least some of these statements could refer to petrologists interpreting rock microstructures. People do make mistakes and even see things that are not there, and though practical jokes may be uncommon in such a serious pursuit as petrology (!), people certainly do occasionally misunderstand what they’re seeing. We all do, in fact. Most important, we often want something to be true so much that we may be tempted to gloss over the evidence, whereas we should be doubly careful, in order to save ourselves falling into the trap of a woefully wrong interpretation, no matter how attractive it may seem.
It does not matter how many times an assertion is repeated or how loudly it is trumpeted in conversation, in the scientific literature, in textbooks or even on the Internet; it is only as good as the evidence for it. Another point to remember is that an interpretation presented by a great authority on the subject, though worthy of respect perhaps, is also only as good as the evidence for it. Such ‘arguments by authority’ can subdue interpretations based on careful accumulation of evidence (Vernon, Reference Vernon1996b).
Too often we see examples of interpretations based on inadequate evidence used to support a preferred model. Even some well-accepted interpretations may be wrong. A good example is the common belief that an ‘order of crystallization’ in igneous rocks can be inferred by looking at the microstructure. Generally this is impossible, as explained in Section 3.6. If the microstructure cannot give you the evidence, please do not try to extract it anyway!
In fact, the more I examine and read about rock microstructures, the more cautious I become about interpreting them, and that will be a constant theme in this book. As mentioned in Section 1.11, recent work on the direct microscopic observation of developing microstructures in organic compounds used as mineral analogues has revealed many unexpected processes, and has shown that similar microstructures may have very different histories. They remind us of the necessity for caution in the interpretation of natural rocks.
So I will try to give explanations that are sufficiently general to be regarded by most people as ‘reasonable’ on the available information and that students can infer largely from the optical microstructure. Where alternatives need to be discussed, the relevant publications will be mentioned. Moreover, where pitfalls exist, they will be pointed out, and it must be reemphasized that this book is only the most general of guides. It indicates what to look for and how to start (not stop) thinking about what is observed.
1.11 Kinds of Evidence Used
What evidence is useful in interpreting rock structures? Imagine you had never seen a rock section, either a thin section or a slab cut through a hand sample. How could you begin to interpret the crystal shapes and arrangements you see? You must have some guides. These are field relationships and experimental evidence on rocks and minerals, assisted by some general inferences from experiments on other materials, such as metals, ceramics, organic polymers and synthetic ice. For example, when Sorby first looked at thin sections of slates with the microscope, he would have already known that slates are formed by strong deformation and that the deformation is in some way responsible for their characteristic strong foliation (slaty cleavage). Furthermore, once Sorby had observed and described the features shown by the microstructure of the slate, others were in a position to recognize similar cleavages in thin sections of rocks from other areas. In this way, general guides to the interpretation of rock microstructures have been established.
If we can observe rocks forming, as with sedimentary and volcanic rocks, we are on strong ground for making inferences about how the microstructures were formed. We are on much shakier ground when it comes to intrusive igneous and metamorphic rocks. However, we can learn much from careful interpretation of field relationships, though strong differences of interpretation often occur. In addition, experiments on the cooling of melted rocks and the melting of solid rocks are valuable guides to the interpretation of rocks involving melts, and many recent experimental advances have been made in the interpretation of igneous microstructures (e.g., Lofgren, Reference Lofgren1971b, Reference Lofgren1973, Reference Lofgren, MacKenzie and Zussman1974, Reference Lofgren1976, Reference Lofgren and Hargraves1980; Lofgren et al., Reference Lofgren, Donaldson, Williams, Mullins and Usselman1974; Fenn, Reference Fenn, MacKenzie and Zussman1974, Reference Fenn1977, Reference Fenn1986; Donaldson, Reference Donaldson1976, Reference Donaldson1977, Reference Donaldson1979; Swanson, Reference Swanson1977; Swanson & Fenn, Reference Swanson and Fenn1986; London, Reference London, Martin and Cerny1992; Hammer & Rutherford, Reference Hammer and Rutherford2002; Hammer, Reference Hammer2004, Reference Hammer2009; Brugger & Hammer, Reference Brugger and Hammer2010a, Reference Brugger and Hammer2010b). However, we should keep in mind possible problems caused by the short duration of experiments.
Experimentally determined stability fields of mineral assemblages in different bulk chemical compositions reveal the conditions of pressure, temperature and fluid composition that occur during metamorphism. However, it is not as easy to conduct successful experiments on the development of microstructures in metamorphic rocks because of the high temperatures and pressures involved in the experiments and the generally small size of the samples used.
Many important experiments on mineral and rock deformation have been carried out (Chapter 5), but again we can only observe the finished product, not the stages along the way. Fortunately, experiments on ice deformation have helped our understanding of progressive microfabric development during deformation (e.g., Wilson, Reference Wilson1984, Reference Wilson, Heard and Hobbs1986; Wilson et al., Reference Wilson, Burg and Mitchell1986). Moreover, a new experimental technique using transparent and translucent organic compounds that behave somewhat similarly to minerals has been developed (Sections 3.3.5 and 5.2) and is being applied with great effect to the interpretation of microfabrics, especially deformation features (e.g., Means, Reference Means1977; Urai et al., Reference Urai, Humphreys and Burrows1980, Reference Urai, Means, Lister, Hobbs and Heard1986; Means, Reference Means1981; Urai & Humphreys, Reference Urai and Humphreys1981; Means, Reference Means1983; Urai, Reference Urai1983a, Reference Urai1983b, Reference Urai1987; Means & Jessell, Reference Means and Jessell1986; Means & Ree, Reference Means and Ree1988; Means, W. D. Reference Means1989; Ree, Reference Ree1991; Means & Park, Reference Means and Park1994; Park & Means, Reference Park and Means1996; Ree & Park, Reference Ree and Park1997). Because these compounds deform, melt and crystallize rapidly at room temperatures, the processes can be observed and photographed in progress in the microscope (‘see-through’ experiments). Of course, these materials are generally not minerals, but nevertheless, they have provided some startling and surprising insights into possible grain-scale processes that may occur in natural rocks.
Another technique developed recently is direct transmitted light observation of crystallization of minerals during cooling from realistically high temperatures, using the ‘moissanite cell’ (Hammer, Reference Hammer2009; Schiavi et al., Reference Schiavi, Walte and Keppler2009; Ni et al., 2014), as described in Section 3.3.5.
In the absence of reliable experimental evidence, it is necessary to fall back on ‘commonsense’ interpretations based on accumulated experience of the type outlined previously. This applies especially to metamorphic rocks. Unfortunately, commonsense isn’t so common, and what makes perfectly good sense to one person may make no sense at all to somebody else. The most important thing is to be as honest and logical as possible, and to evaluate (and if necessary retain as possibilities) every interpretation that can reasonably explain the observations. If the end result is the unsatisfying conclusion that you cannot make an unequivocal interpretation on the available evidence, leave it at that. No harm will be done. On the contrary, many a doubtful interpretation, presented as being reliable, has been accepted at face value and used in later work, thereby misleading subsequent researchers.
1.12 Complexity
A rock’s microstructure is the product of a complicated sequence of events and processes. So is a rock’s chemical analysis. Both may tell us something about the rock’s history, but neither can fully reveal all the historical complexities. This is a problem that petrologists have to accept. We do our best with the evidence available, without taking it too far, and we must acknowledge that our interpretations are often incomplete.
Another point to add to the complexity is that superficially similar microstructures may be formed in different ways, as with exsolution and epitaxial replacement producing similar intergrowths (e.g., Craig, Reference Craig, Jambor and Vaughan1990a, Reference Craig, Jambor and Vaughan1990b). For example, hematite lamellae in magnetite, usually inferred to be of replacement origin, can be due to exsolution in some rocks (Edwards, Reference Edwards1949). Another complication is the optical similarity between subgrains formed by recovery and similar features formed by fracture, as discussed in Section 5.4. Other complexities of rock microstructure will become apparent in the following chapters.
