Geology A Complete Introduction PDF

Geology
A Complete Introduction
601550_Geo_A_Comp_Intr_Book.indb 1 5/8/15 12:29 PM

This page intentionally left blank
Geology
A Complete Introduction
David Rothery

First published in Great Britain in 1997 by Hodder Education.
An Hachette UK company.
First published in US in 1997 by The McGraw-Hill Companies, Inc.
Previously published as Geology: The Key Ideas
This revised edition published 2015
Copyright © David Rothery 1997, 2003, 2008, 2010, 2015
The right of David Rothery to be identified as the Author of the Work has been
asserted by him in accordance with the Copyright, Designs and Patents Act 1988.
Database right Hodder & Stoughton (makers)
The Teach Yourself name is a registered trademark of Hachette UK.
All rights reserved. No part of this publication may be reproduced, stored in a
retrieval system or transmitted in any form or by any means, electronic, mechanical,
photocopying, recording or otherwise, without the prior written permission of
the publisher, or as expressly permitted by law, or under terms agreed with the
appropriate reprographic rights organization. Enquiries concerning reproduction
outside the scope of the above should be sent to the Rights Department, John
Murray Learning, at the address below.
You must not circulate this book in any other binding or cover and you must
impose this same condition on any acquirer.
British Library Cataloguing in Publication Data: a catalogue record for this title is
available from the British Library.
Library of Congress Catalog Card Number: on file.
Paperback ISBN 978 1 473 60155 0
eBook ISBNs 978 1 473 60157 4, 978 1 473 60156 7
1
The publisher has used its best endeavours to ensure that any website addresses
referred to in this book are correct and active at the time of going to press.
However, the publisher and the author have no responsibility for the websites and
can make no guarantee that a site will remain live or that the content will remain
relevant, decent or appropriate.
The publisher has made every effort to mark as such all words which it believes to
be trademarks. The publisher should also like to make it clear that the presence of
a word in the book, whether marked or unmarked, in no way affects its legal status
as a trademark.
Every reasonable effort has been made by the publisher to trace the copyright
holders of material in this book. Any errors or omissions should be notified in
writing to the publisher, who will endeavour to rectify the situation for any reprints
and future editions.
Typeset by Cenveo® Publisher Services.
Printed and bound in Great Britain by CPI Group (UK) Ltd., Croydon, CR0 4YY.
John Murray Learning policy is to use papers that are natural, renewable and
recyclable products and made from wood grown in sustainable forests. The logging
and manufacturing processes are expected to conform to the environmental
regulations of the country of origin.
John Murray Learning
Carmelite House
50 Victoria Embankment
London EC4Y 0DZ
www.hodder.co.uk

Contents
Preface ix
Acknowledgements x
1 Introduction: a mighty matter of legend 1
Rock types and the rock cycle
The structure of this book
2 Inside and outside: the solid Earth
and the atmosphere 9
The atmosphere and its development
The Earth’s interior
Seismic waves
The Earth’s core
Evidence from magnetism
The Earth’s crust and mantle – compositional layers
Summary of compositional layers
Another kind of layering – strong and weak
The Earth’s heat
Radiogenic heating
Radiometric dating and the age of the Earth
The origin of the Earth and the Solar System
A warning
3 Earthquakes 35
How earthquakes happen
Measuring earthquakes
Protecting structures from earthquake damage
Earthquake preparedness
Prediction and prevention
Tsunamis
4 Plate tectonics 51
Moving plates
Plates colliding
Plates moving apart
The break-up of a continent
The life-cycle of an ocean
Plates sliding past each other
The global picture
What makes the plates move?

5 Volcanoes 71
Magma origin and eruption
Magmatic gases and explosive eruptions
Volcanic hazards
Ignimbrite eruptions
Volcanoes associated with continental rifting
Volcanism at ocean ridges
Hawaiian volcanism
Lava hazards
Predicting eruptions
Widespread effects of eruptions
6 Igneous intrusions 105
Minerals
Granite
Basaltic intrusions in the oceanic crust
Dykes and sills
Grain sizes in igneous rocks
7 Metamorphism 123
Thermal metamorphism
Regional metamorphism
Metamorphic facies
Textures of metamorphic rocks
8 Weathering, erosion and transport 137
Chemical breakdown of minerals during weathering
Physical weathering
Sand-sized grains in wind and water
Moving ice
Glacial landscapes
Erosion by rivers and streams
Drainage patterns
Marine erosion
9 Deposition of sediment and the
origin of sedimentary rocks 155
Ripples and dunes
Turbidity currents
Current directions
Waves
Finer and coarser sediments
Other means of transport
Using sediments to interpret ancient environments
Depositional environments associated with rivers
vi

Flash floods
Deltas and estuaries
Logging a sedimentary succession
Beaches and other coastal environments
Evaporite deposits
Limestone
Turning sediment into rock
10 Deformation of rocks 189
Different types of steep fault
Low-angle faults
Thrust belts and collision zones
Brittle faults versus shear zones
Folds
Shapes and styles of folding
Unconformities
11 Physical resources 211
Reserves and resources
Origins of oil and gas
Oil and gas reservoirs
Exploration for oil and gas
The origin of coal
Mining techniques
Ores and their origins
Ore exploration
The mining of ores
Diagenetic and sedimentary ore bodies
Non-metallic raw materials
Making good the damage
Water resources
12 Past life and fossils 239
The origin of life
The first microbes
More complex cells
Multicellular organisms
The major animal phyla appear
Vascular plants and life on land
The mammals take over
Mass extinctions and adaptive radiations
How fossils form
Exceptional preservation
Plant fossils
The use of fossils in stratigraphy
Contents vii

13 A short history of the Earth 267
The Earth’s first crust
Archean plate tectonics?
The growth of cratons
The Proterozoic Eon
Phanerozoic plate configurations
Climate change
Human-induced climate change
The future history of the Earth
14 Solar System geology 293
Impact craters
The Moon
Mercury
Venus
Mars
The asteroids
Io
Europa
Large icy bodies
Geology among the stars
15 Going into the field 327
Things to do in the field
Field equipment
Field safety
Geological maps and guidebooks
Suggestions for field projects
Appendix 1: Minerals 343
Appendix 2: Rock names 349
Igneous rocks
Metamorphic rocks
Sedimentary rocks
Appendix 3: Geological time 355
Glossary 357
Answers 371
Index 379
viii

Preface
I am grateful to all who helped with or commented upon the
previous editions of this book. Those would probably never
have been written without the stimulus of teaching geology to
enthusiastic students out in the field, which is by far the best
environment in which to inspire them about the record of our
planet’s past preserved in the rocks. Let those short-sighted
bean-counters who would reduce fieldwork teaching beware!
For this fifth edition, I have taken the opportunity to clarify
the text in many small instances. I have also interpolated ‘key
ideas’, ‘case studies’ and a few other devices into the text that
I hope will be helpful as aids to learning. Each chapter, except
the first, now concludes with some simple multiple-choice
questions to help readers test and consolidate what they have
learned. Answers will be found near the back of the book.
I have increased the number of references (to online and
print resources) and distributed them at the end of the most
relevant chapter, under the heading ‘Digging deeper’.
I have updated various items relating to rapidly changing
topics (such as volcanic eruptions, earthquakes, climate change,
physical resources and Solar System geology), expanded the
glossary and the index, and rectified various minor errors or
omissions.
Preface ix

Acknowledgements
The diagrams in this book have been specially prepared, and
most of the photographs are my own. With those exceptions,
the illustrations are from the following sources:
Figures (illustrations in the main text): Figure 3.3 United
States Geological Survey; Figure 5.6 Stuart Hall; Figure 5.12
Peter Francis; Figure 9.10 United States Geological Survey/
NASA; Figure 12.5 Phil Allen (Production Geoscience Ltd)
and Simon Stewart (BP); Figures 13.3–13.6 are modified
from various sources, but draw heavily on the work of C.R.
Scotese and co-workers; Figures 14.2, 14.4, 14.8, 14.10–12,
14.14, 14.16, 14.17 NASA; Figure 14.3 NASA/Johns Hopkins
University APL/Carnegie Institution of Washington; Figure 14.5
Don P. Mitchell and Yuri Getkin; Figure 14.7 ESA/DLR/FU
Berlin (G. Neukum), Figure 14.9 NASA/JPL-Caltech/UCLA/
MPS/DLR/IDA.
Plates (colour pictures in the middle of the book): Plate 1
JAMSTEC/Earth Simulator Center; Plate 13 John Watson
(copyright Open University); Plate 14 NASA/JPL/Cornell;
Plate 15 NASA.

Introduction:
1
a mighty matter
of legend
In this chapter you will learn:
33about the scope of this book, and how
most chapters are intended to help you
understand particular processes
33the three main types of rock, and how
they are related by the ‘rock cycle’.

‘The green earth, say you? That is a mighty matter of legend,
though you tread it under the light of day!’* Although taken from
a work of fiction, about a fictional situation, those words may be
taken as the theme for this book. The Earth is a mighty matter
of legend. Its origin and the events that have taken place in its
history do indeed form a grand and epic tale. Moreover, various
parts of this story have been interpreted in many ways; some of
these are contradictory, and others are now thought to be untrue.
I should also point out that some of what geoscientists now
believe fairly firmly will undoubtedly turn out to be wrong, or
only partly correct. However, I think it is fair to claim that we
understand very well (for the purpose of an introductory book
such as this) most of the processes that shape the materials at,
and immediately beneath, the Earth’s surface.
Probably, what we lack most is a fully developed understanding
of the interdependence of all these processes. For example,
over geological timescales, measured in tens of thousands or
millions of years, the Earth’s climate has changed many times.
It is changing today, and the consensus among scientists is that
humans are to blame for the current rapid rate of global warming
because of the gases we have released into the atmosphere,
notably carbon dioxide through burning coal and oil and various
industrial processes. However, other factors drive climate change,
too. These include changes in the tilt of the Earth’s axis and
slight variations in the shape of its orbit, minor fluctuations in
the energy output of the Sun, the slow drift of the continents
that forces readjustment of the ocean circulation pattern, the rise
and fall of mountain belts, catastrophes such as giant volcanic
eruptions and impacts by asteroids and comets, and even the
emergence and evolution of life itself. All these factors have
played a role in driving the temperature and climate of the Earth
in one direction or another. Just how significant human-induced
global warming will be in the grand scheme of things remains
to be seen, though I think it is likely to impact severely on
ecosystems and human cultures in many parts of the globe.
Climate, in turn, influences how fast and in what manner rock
that is exposed at the surface becomes worn away, and where
* Aragorn to Eomer, The Two Towers, J.R.R. Tolkien, 1954

it is redeposited as sediment. Thus, all the factors listed above
must leave some kind of imprint on the history of our planet,
as revealed by a study of the rocks. The complexities are
boundless, but nevertheless we now have a reasonably complete
understanding of how the Earth as a whole functions.
People may become interested in geology, which is the scientific
study of the Earth, for a variety of reasons. Many are intrigued
by fossils, or pebbles that they pick up on the beach, and
wonder what they mean. Others want to know more about
earthquakes and volcanoes reported in the news, and whether
such events are likely in their own neighbourhood. Some may
be curious about where the oil comes from that powers their
car, or what their house bricks are made of. And some are just
overawed by the grandeur of the mountains or impressed by the
pictures of the Earth and of other Earth-like bodies sent back by
space probes, and decide that they want to find out more.
Whatever your own reason for opening this book, I hope you will
find what you want. I have tried to address all these concerns, and
others. Most chapters aim to help you understand a particular
type of geological process. You don’t have to read these chapters
in the sequence in which they are printed, but it will be easier if
you do. This is because I have attempted to build up the use of
terminology (geology is riddled with strange-sounding words!)
in a coherent fashion. Notable exceptions to the process-oriented
approach are the next chapter (which is a descriptive account of
the composition of the Earth) and the final chapter (which deals
with fieldwork and the simple equipment you will find useful if
you decide to go and look at rocks and fossils for yourself). At
the end of the book is a glossary, where the most important terms
(introduced in bold in the text) are defined.
Rock types and the rock cycle

It is helpful at this stage to define terms that describe each of the
three main kinds of rock:
1 Igneous rock is material that was once molten; it usually
contains crystals that grew within this molten material as it
cooled.

2 Metamorphic rock is made when heat or pressure (or both)
causes a pre-existing rock to recrystallize, but without melting.
3 Sedimentary rock forms at the surface, by deposition of
detrital grains or by precipitation from solution in water.
A simplified view of the origins of these rock types is shown in
Figure 1.1, which illustrates what is known as the rock cycle.
This includes the erosion, transport, deposition, burial, heating,
deformation, melting, cooling and exhumation of rock or rock
fragments. Rock on high ground is worn to fragments by frost,
rain and wind, and transported downhill to a place where these
fragments are deposited and buried. This is the erosional and
depositional part of the rock cycle. If these processes were the
sole agent of change, the ultimate fate of the Earth’s surface
would be for hills and mountains to be worn away, until
eventually everywhere was flat and buried by mud. However,
the Earth is a dynamic planet, so forces of deformation
and uplift are at work that continually bring igneous and
metamorphic rock, and hardened sediment, to the surface. This
provides an unending supply of material to be worked on by the
forces of erosion, and so the rock cycle continues.
volcanic eruption
erosion
transport
deposition
uplift
deformation sea
burial
igneous metamorphism sedimentary
rocks metamorphic rocks
melting rocks
Figure 1.1 Cross-section through the Earth’s surface to show the rock
cycle.
In nature, the rock cycle is not as simple as it may seem in

Figure 1.1. If you were to attempt to trace the history of a
particular volume of rock over hundreds of millions of years,

you might find it tracing a loop to and fro within just part
of the cycle, perhaps being alternately metamorphosed and
melted at depth and never being brought to the surface. On the
other hand, a grain of sediment may be deposited in a whole
succession of sedimentary environments, being plucked out by
erosion and transported to each new environment without ever
becoming deeply buried.
The rock cycle is the most important unifying idea in geology,
and it forms the main theme of this book. However, geoscientists
recognize many other interrelated cycles, too. For example, the
water that does much of the erosion, that transports and shapes
sedimentary grains, and that carries in solution the chemicals
that help to turn soft sediment into hard sedimentary rock is part
of a hydrological cycle involving evaporation and precipitation,
interaction with the ocean floor, and even passage into the
Earth’s interior and escape again during volcanic eruptions.
Similarly, there is a carbon cycle, involving an interplay between
atmospheric carbon dioxide and carbon – stored short-term in
living organisms and long-term in certain kinds of sedimentary
rock (carbonates such as limestone and in coal and oil). This
is an important influence on climate, because the more carbon
dioxide there is in the atmosphere, the more solar heat is trapped
and the warmer the planet tends to be.
The structure of this book

There is no obvious point at which to begin describing the rock
cycle. In this book I begin with its igneous parts, first near the
surface (Chapter 5) and then at depth (Chapter 6). Next, I deal
with metamorphism (Chapter 7); followed by the erosional and
depositional parts of the cycle (Chapters 8 and 9); and then the
remaining aspect of the rock cycle, related to the deformation of
rocks (Chapter 10). Subsequent chapters consider the physical
resources that are available to us as a result of the rock cycle
(Chapter 11); the history of life on Earth and of the Earth
itself (Chapters 12 and 13); and evidence that the processes
I have described occur on other planets, too (Chapter 14). The
final chapter provides an introduction to geological fieldwork
(Chapter 15).

First, though, I describe the Earth’s internal properties and
its atmosphere (Chapter 2) to give you the information you
need to set its geology in context. Then I discuss earthquakes
(Chapter 3) – which are one of the most obvious manifestations
of the unquiet Earth; and plate tectonics (Chapter 4) – which
is a description of how wandering continents and widening
or vanishing oceans are associated with the creation and
destruction of rock on a huge scale.

Dig deeper
The rock cycle can be explored at http://www.geolsoc.org.uk/
ks3/gsl/education/resources/rockcycle.html
and via an animation at http://www.pbslearningmedia.org/resource/
ess05.sci.ess.earthsys.rockcycle/rock-cycle-animation/
A site at the Smithsonian Museum of Natural History (Washington,
DC) has lots of good things to explore if you select the topic ‘Earth
and Planetary Sciences’ http://www.si.edu/Encyclopedia/Search/
Science and Technology
http://www.usgs.gov/ The home page of the United States
Geological Survey; complicated but you can find information on
most geological topics here.
http://www.bgs.ac.uk/ The home page of the British Geological
Survey.
http://www.gsi.ie/ The home page of the Geological Survey of
Ireland.
http://www.ga.gov.au/ The home page of Geoscience Australia.
http://www.gns.cri.nz/ The home page of New Zealand’s Institute
of Geological and Nuclear Sciences.
http://www.geolsoc.org.uk/ The home page of the Geological
Society of London, the organization to which most professional
geologists in the UK belong.
http://www.geologists.org.uk/ The home page of the Geologists
Association, representing amateur geologists in the UK.
http://ougs.org The home page of the Open University Geological
Society, a Europe-wide self-help group for serious or casual
students of geology. News of local meetings and field trips.
http://geosociety.org The home page of the Geological Society of
America, a largely professional society.
http://geology.about.com/ A free commercial site (with advertising)
offering up-to-date news and web links across the full range of
geology, including pictures of rocks, minerals and fossils.

Inside and
2
outside: the solid
Earth and the
atmosphere
33about the Earth in cross-section, from
the atmosphere to the core, and how
some of this information has been
obtained
33about the Earth’s age and origin.

Before exploring the geological processes that affect the Earth, I
had better establish just what we are dealing with. As Table 2.1
shows, the Earth is the largest of several rocky planets in our
Solar System. It is nearly spherical, measuring 12 714 km from
pole to pole and 12 756 km round the Equator, and orbits the
Sun at an average distance of 150 million km.
Table 2.1 The Earth and similar planetary bodies. Although the Moon orbits
the Earth rather than the Sun (and so is not a planet in the astronomical
sense), it is included here because geologically it belongs to the same class
of bodies.
Distance Density in
from Sun Equatorial Mass, relative tonnes per Surface
in millions radius in to Earth cubic metre atmospheric
Name of km km (6.0 × 1024 kg) (103 kg m–3) pressure (bars)
Mercury 58 2439 0.055 5.43 negligible
Venus 108 6052 0.815 5.25 92
Earth 150 6378 1 5.52 1
Moon 150 1738 0.0123 3.34 negligible
Mars 228 3394 0.107 3.90 0.0016
Spotlight: What is a planet?

In 2006 the International Astronomical Union formally defined a
planet as a body that orbits the Sun, that is large enough to take
on a nearly spherical shape, and that has been able to ‘clear its
orbit’ of other bodies of comparable size. Pluto crosses the orbit
of the much more massive planet Neptune, and so Pluto is no
longer officially a planet. It was put into a new category called
‘dwarf planet’ to denote bodies orbiting the Sun that are big
enough to be spherical in shape but which have not ‘cleared
their orbits’.
I will describe some of our Solar System’s other geologically

interesting bodies in Chapter 14, but until then my focus is very
much on the Earth.
10

The atmosphere and its development
Compared with the size of the planet, Earth’s atmosphere is
no more than a very thin layer. It is so sparse as to be virtually
unbreathable at the tops of the tallest mountains (8–9 km above
sea level), clouds rarely occur higher than at about 12 km, and
at 200 km it is so tenuous that it offers little or no resistance to
artificial satellites as they orbit.
However, so far as living things are concerned, the importance
of the atmosphere is out of all proportion to its size. This is
not just because it contains the oxygen necessary for breathing,
but also because the atmosphere moderates the temperature of
the globe and shields the surface from radiation that would be
harmful to life.
Most familiar living things would have a hard time surviving
on the Earth’s surface if the atmosphere were to become
radically changed. This is not to say that the present
atmosphere has been around since the Earth was formed. As
you will see towards the end of this chapter, the Earth’s origin
is reliably dated at just over 4.5 billion years ago. At that time
the Earth’s primordial atmosphere (most of which is thought
to have escaped from the interior by means of volcanoes) was
probably largely water vapour, carbon dioxide, sulfur dioxide
and nitrogen.
The appearance and subsequent evolution of exceedingly
primitive living organisms (bacteria-like microbes and
simple single-celled plants) began to change the atmosphere,
liberating oxygen and breaking down carbon dioxide and
sulfur dioxide. This made it possible for higher organisms to
develop.
When the earliest known plant cells with nuclei evolved about
2 billion years ago, the atmosphere seems to have had only
about 1 per cent of its present content of oxygen. With the
emergence of the first land plants, about 500 million years ago,
oxygen reached about one-third of its present concentration. It
had risen to almost its present level by about 370 million years
ago, when animals first spread on to land.
2 Inside and outside: the solid Earth and the atmosphere 11

Today’s atmosphere is thus not just a requirement to sustain life
as we know it – it is also a consequence of life.
Key idea: The atmosphere

Without the Earth’s atmosphere, conditions at its surface would be
radically different. Atmospheric composition has evolved over time,
notably because of interactions with living organisms.
The Earth’s interior

This book is concerned mostly with things that occur upon,
or at shallow depth below, the Earth’s surface. I will come
back to the atmosphere from time to time, but first we need to
investigate the deep interior. People living in regions where the
soil is thin are well aware that the outside of the solid Earth is
made of rock. The rest of us are reminded of this when we see a
quarry or a cliff. But could the Earth be made of rocky material
like this all the way to its centre?
A simple way to determine whether that is possible is to
consider the Earth’s density. The kind of rock you usually find at
the surface has a density of about 2.7 tonnes per cubic metre. If
the Earth were made of rock, it should have an overall density
only a little greater than this value, to allow for compression
(and hence greater density) at depth.
You can work out the Earth’s actual density as follows. The
Earth’s mass is known to a high degree of precision, because
it can be determined from how long it takes the Moon or
an artificial satellite to complete an orbit. It was shown in
a Table 2.1 column heading as 6.0 × 1024 kg. If you are not
familiar with scientific notation, this translates as 6 million
million million million kilogrammes, or 6 million billion billion
kilogrammes, or 6 thousand billion billion tonnes. The Earth’s
density can be found by dividing its mass by its volume. The
result is 5.52 tonnes per cubic metre, which is about twice the
density of surface rock. Clearly, therefore, the Earth is not just
rock. Since the outer part is self-evidently rock, you can deduce
that there must be something denser than the average value
hidden away at depth. In fact, the centre of the Earth is believed
12

to be composed mostly of iron, with a density of almost
13 tonnes per cubic metre.
The picture geologists have formed of the Earth’s interior is not
based merely on density arguments. Density tells us only that
there must be something much denser than rock somewhere
within the Earth, but not how this denser matter is arranged,
let alone what it is made of. For example, Figure 2.1 shows two
entirely different types of internal structure for the Earth, both
of which are consistent with the Earth’s density, but only one of
which makes sense when other evidence is taken into account.
Figure 2.1 Two possible arrangements of dense matter within the Earth.
Either of these would account for the observed average density, but only
the model on the right fits the observations discussed later in the text.
Seismic waves
The most useful information for determining the internal
structure of the Earth comes from the study of how vibrations
travel through it. As you can imagine, it needs pretty powerful
vibrations to be detected through the full thickness of the globe.
Fortunately, the vibrations from large earthquakes, known as
seismic waves, are sufficiently strong for this purpose (though
this is a mixed blessing for people living in the most earthquake-
prone areas, as you will see in the next chapter).
The reason why seismic waves are so useful for probing the
Earth’s interior is that the speed at which vibrations travel
through a substance depends on its physical properties. As
density increases, so the speed of seismic waves tends to

decrease. On the other hand, vibrations travel faster through a
more rigid substance than through a less rigid substance.
Let’s see how this can help us determine the structure of the
Earth, by supposing for now that the outer part of the Earth has
a uniform (rocky) composition. How would we expect the speed
of seismic waves to vary with depth? Rigidity should increase
with depth, as a result of the increasing pressure exerted by the
overlying material. The effect of this should be for the speed of
transmission of seismic waves to increase as depth increases. The
contrary tendency for speed to decrease because of increasing
density with depth is a much slighter effect, since rock is not
compressed very much however high the pressure becomes. We
therefore expect the speed of seismic waves to increase with depth
within the rocky part of the Earth, and this is indeed what we find.
You can demonstrate this when an earthquake happens at a
known location, by timing the arrivals of the first vibrations
received by sensitive vibration detectors, known as seismometers,
at different distances. Figure 2.2 shows how this works. Note
that the vibrations follow curved paths as a consequence of the
speed of the transmission increasing with depth.
Source
Figure 2.2 Because the speed of transmission of seismic waves increases

with depth, the first vibrations detected by distant seismometers are not
those that have travelled in a straight line, but those that have followed a
curved track, shown here, taking them on a deeper but faster route. The
first vibrations reach seismometer A about 27 minutes after the event, and
seismometer C only about 35 minutes after the event, even though C is
about one and a half times further away.
14

Studies of this sort, based on hundreds of thousands of
earthquakes detected at thousands of seismometers across the
globe, have enabled the speed of seismic waves at different
depths to be determined in considerable detail. The structure
of the Earth’s outer few tens of kilometres varies considerably
from place to place, and I will consider this shortly. The deeper
interior is much more uniform, seismic speeds changing with
depth but not varying much with location.
What helps us to pin down the density distribution within
the Earth is that at a depth of 2,900 km the speed of seismic
waves drops sharply. We can tell this because, instead of
continuing to follow an upward-curving track, the path of a
vibration reaching this depth is deflected downward, as shown
in Figure 2.3. At this depth, the composition evidently changes
Source
core mantle
shadow
zone
Figure 2.3 Paths of seismic waves deep within the Earth. The density of
the core is about twice that of the much more rigid overlying mantle, so
seismic speed is halved. The direction of wave travel is deflected sharply
downward at the core–mantle boundary, and there is a ‘shadow zone’ at
a specific range of distances from the source within which seismometers
cannot pick up directly transmitted waves. In addition to earthquakes,
underground nuclear explosions cause seismic waves powerful enough to
probe the deep structure in this way.

from rocky to something different, with about twice the density,
causing a dramatic decrease in seismic wave speed. This marks
the outer limits of the Earth’s dense core. The rocky material
surrounding the core is known as the mantle.
Key idea: Seismic waves and the Earth’s interior

The way that seismic waves (vibrations triggered by earthquakes)
travel through the Earth enables its internal structure to be
mapped out.
The Earth’s core

A remarkable thing about the outer core at this depth is that,
despite the immense pressure, it is not solid. This has been
demonstrated by studying the transmission of different kinds
of seismic waves, which so far I have glossed over. There are
two sorts of seismic wave that concern us when considering
the Earth’s interior: compressional waves and shearing waves.
A compressional wave (or P-wave) is just like a sound wave in
air; it consists of alternate pulses of compression and dilation.
A shearing wave (or S-wave) can be observed by shaking a
jelly; it is an alternate side-to-side wobble travelling through
the body of the material. An important distinction between
these two types of seismic waves is that compressional waves
can travel through anything, but shearing waves cannot pass
through a liquid. This is because a liquid offers no resistance
to shearing motion. Seismometers can distinguish between
the two types of waves, and from this we can tell that only
P-waves are transmitted through the outer core, which must
therefore be liquid.
That is not quite the end of the story, because an inner core
can be identified whose limit is marked by a rapid increase in
the speed of P-waves and within which we can tell (by indirect
means) that S-waves are once again transmitted. Figure 2.4 is a
plot showing the variation with depth of the speeds of P-waves
and S-waves (which can be calculated) and density (which can
then be deduced).
16

MANTLE CORE
P-waves density
speed of seismic waves (km per second)
density (tonnes per cubic metre)

10 10
P-waves
S-waves
5 density 5
S-waves
2000 4000 6000

depth (kilometres)
Figure 2.4 Speeds of compressional waves (P-waves) and shearing waves
(S-waves) within the Earth, and inferred density
Now you have seen the evidence needed to decide between the
alternative models of density distribution given in Figure 2.1, and
I hope you will agree that the right-hand model (symmetrical
distribution), rather than the left-hand model (irregular
distribution), has to be correct. In fact, we can go further than
this, because we have enough information to propose what the
dense stuff actually is. The solid inner core has the properties
we would expect of solid iron mixed with a small percentage
of nickel. Its properties match that of other metals too, such as
cobalt or titanium, but the iron core model is the only reasonable
one because it fits with the Earth being rich in the same metallic
elements as we find in the Sun and in meteorites. The liquid outer
core has a density too low to be pure metal. About 10 per cent of
its mass must be composed of one or more relatively light elements.
What these are cannot be proved, but oxygen, sulfur, carbon,
hydrogen and potassium (a light metal) are the most likely.
Key idea: The Earth’s core

The Earth has a dense core made largely of iron. The inner core is
solid, but the outer core is liquid. The core makes up 27.5 per cent
of the Earth’s mass, but only 9.25 per cent of the Earth’s volume.

Evidence from magnetism
There is one more factor bearing on the nature of the core that
I should touch on here, and this is magnetism. The Earth has a
magnetic field resembling one produced by a giant bar magnet
(Figure 2.5). Both common sense and the seismic evidence you
have just considered enable you to rule out a real bar magnet as
the origin of this magnetic field. Instead, the field is regarded as
the product of electrical currents generated by motion (a sort of
self-sustaining dynamo) within the fluid outer core.
Figure 2.5 Cross-section through the Earth from pole to pole, showing the
lines of magnetic field and an imaginary bar magnet at the Earth’s centre.
Key idea: Magnetism

The Earth’s magnetic field is generated electrically by dynamo
motion in the liquid outer core.
The Earth’s crust and mantle –

compositional layers
Now I will turn to the outermost few tens of kilometres of the
Earth, where the variation in seismic speeds is most complex.
18

Junctions between one rock body and another are sometimes
manifested by sharp changes in seismic speed, but on a global
scale the effect is one of gradually increasing speed with depth.
This is because of the dominating effect of increasing rigidity
when depth, and hence pressure, increases. P-wave speed
increases gradually from about 2 km per second just below the
surface to 6 or 7 km per second. Below this, a sharp jump to a
P-wave speed of 8 km per second is recognized throughout the
entire globe, at an average of 30 km below the continents but
usually 10 km or less below the ocean floor. This sharp change
to denser rocks is known as the Mohorovicic discontinuity (after
Andrija Mohorovicic, 1857–1936, the Croatian seismologist
who first recognized it). It is usually called the Moho for short.
Above the Moho are the rocks that belong to the Earth’s crust,
and below it is the mantle, which extends all the way to the core.
The crust is composed of slightly less dense rock than the mantle,
and has a greater proportion of silicon, aluminium, calcium,
sodium and potassium, but less magnesium. It is easy to explain
why the crust sits above the mantle; essentially it is just the light
stuff that has worked its way to the top. For reasons that you
will learn later, there are actually two distinct types of crust:
continental crust (underlying virtually all the land surface and
the shallow seas) and oceanic crust (forming the floor of the
deep oceans). The compositions of these two types of crust are
compared alongside the mantle composition in Table 2.2.
The thickness of the continental crust varies from about 25 km
in thin, stretched regions, to as much as 90 km below the
highest mountain ranges. Oceanic crust is much thinner, ranging
from about 6 to 11 km thick.
The wide variation in the thickness of the continental crust is
manifested not so much by mountains tens of kilometres high
(the highest mountain, Everest, reaches only 8848 m above
sea level), as by giant downward protuberances of the base of
the crust, which can be detected as variations in the depth of
the Moho. This shows that mountains are not held up by the
strength of the material on which they rest. Instead, crustal
regions of different thickness may be thought of as floating on
the denser mantle. The situation is illustrated in Figure 2.6.

Table 2.2 Average compositions of continental crust, oceanic crust and
the mantle. The names of the rock types most closely matching these
compositions are shown. Elements are named in the first column, but
compositions are expressed as the oxides of these elements (per cent
by weight) whose chemical formula is shown in the second column; this
is purely a convention and should not be thought of as indicating the
chemical species actually present. Iron is listed twice, as Fe2O3 and FeO.
SiO2 is commonly referred to as silica.
Continental
crust (granite– Oceanic crust Mantle
Element Oxide andesite) (basalt) (peridotite)
silicon SiO2 62 49 45
titanium TiO2 0.8 1.4 0.2
aluminium Al2O3 16 16 3.3
iron Fe2O3 2.6 2.2 1.2
iron FeO 3.9 7.2 6.7
magnesium MgO 3.1 8.5 38.1
calcium CaO 5.7 11.1 3.1
sodium Na2O 3.1 2.7 0.4
potassium K2O 2.9 0.26 0.03
sea level water

oceanic
continental crust
crust
57 km
mantle
Figure 2.6 Cross-sections showing crust of different thicknesses resting

in equilibrium upon the mantle. In the situation shown, the weights of
all five columns (crust plus mantle, and including water in the right-
hand column) from surface to base are equal. This demonstrates how
mountains of modest height (6 km in this example) are supported by
buoyant roots (extending 57 km below sea level in this example). Average
densities of water, continental crust, oceanic crust and mantle are 1.0, 2.7,
3.0 and 3.3 tonnes per cubic metre, respectively.
20

Remember this!
Remember that the mantle cannot really be a liquid, because
it transmits S-waves. Much of it behaves like a liquid when
considered over geological time, but it is not molten rock.
The concept of blocks of crust floating in equilibrium upon

the mantle is known as isostasy, and has been found to apply
almost everywhere. There are few examples known where
high or low regions are held up or pushed down by any
force other than simple buoyancy (which is really what is
meant by isostasy). Buoyancy shows us why oceanic crust is
virtually never found on dry land. It is thinner and denser than
continental crust, and so it floats lower.
Key idea: The mantle

The mantle makes up 68.4 per cent of the Earth’s mass and 49.5
per cent of the Earth’s volume.
Summary of compositional layers

Figure 2.7 summarizes what I have said so far about the Earth’s
compositional layers. The compositional difference between
crust and mantle is relatively slight, but sufficient to account
for the jump in the speed of seismic waves across the Moho.
The most abundant elements in both are silicon and oxygen.
Any compound made of a chemical combination of just these
two elements is known as silica. The rocky materials dominated
by silicon and oxygen are therefore commonly referred to as
silicates. Some depth-related changes in seismic speeds within
the mantle have been recognized (see Figure 2.4), but these are
thought to represent pressures at which atoms within crystals
become packed into denser, more rigid structures, rather than
being changes in chemical composition. Throughout, the mantle
is thought to have a chemical composition similar to that of the
rock type known as peridotite. The compositional distinction
between the core and the mantle is much more fundamental

Crust
Upper mantle
Transition zone
Lower mantle
Outer core
Inner core
Iron
with
small
Iron (and percentage
sulfur?) of
Silicate
(liquid) nickel
(peridotite)
(solid)
670 km 2900 km 5155 km 6370 km
400 km
6–11 km (oceans)
25–90 km (continents)
Figure 2.7 The compositional layers within the Earth. The lower mantle
has a denser structure than the upper mantle, but is not thought to differ
significantly in composition.
than that between the crust and the mantle. The core does not
even consist of silicates, but is dominated by iron.
Key idea: Crust and mantle

The Moho is the interface between the denser, silica-poor rocks
of the mantle and the less dense, relatively silica-rich rocks of
the overlying crust. Oceanic crust is thinner and denser than
continental crust, and the base of the continental crust is deepest
beneath mountain belts.
Another kind of layering – strong

and weak
When you consider how the Earth behaves, it is not always
chemical composition that is the most important factor. You
also need to bear in mind how strong the material is. You have
already met this concept when we considered the crust as blocks
‘floating’ in the mantle. Mountain belts need deep buoyant
roots only because the crust and underlying mantle do not have
22

the strength to support them. Actually, the weakest zone in the
mantle is not immediately below the crust, as you might think
from a simple interpretation of Figure 2.6. Instead, it begins
at a depth of about a hundred kilometres. At this depth there
is a slight drop in seismic speeds (discernible in Figure 2.4),
which has been interpreted as due to the presence of a small
percentage of molten material. However, the material as a whole
must still be well and truly solid because S-waves can travel
through it. This is known as the ‘low-speed layer’ (sometimes
the ‘low-velocity layer’) and it is weak enough to deform (over
geological time periods) under the weight of mountain belts or
ice sheets. For example, Scandinavia is still rising upwards at
a rate of a few millimetres per year in response to unloading
caused by the disappearance of a thick sheet of ice that covered
the region until about 10 000 years ago. It is flow in the weak
zone in the mantle that allows this slow rebound to happen.
You will discover in Chapter 4 that the outer shell of the Earth
is broken into large slabs or ‘plates’ that are gliding sideways.
This, too, would not be possible but for this weak zone, which
lubricates the bases of the plates.
In terms of how the outer part of the Earth responds to forces,
the crust and the uppermost part of the mantle constitute a
single rigid unit known as the lithosphere (from lithos, the
Greek word for ‘rock’). The lithosphere is rocky not just in its
composition but also in terms of its mechanical properties. In
contrast, below the lithosphere the mantle is relatively weak,
although its chemical composition is the same. This weak zone
has been named the asthenosphere (using the Greek word for
weak). The relationship between crust, mantle, lithosphere and
asthenosphere is summarized in Figure 2.8.
The weakest part of the mantle lies in the few tens of kilometres
immediately below the base of the lithosphere, and the
term asthenosphere is often restricted to just this thin layer
(coinciding with the low-speed layer). However, once you get
below the lithosphere there has been an important change in
the properties of the rock that persists all the way to the core.
Although solid in respect of the transmission of seismic waves
(which travel at many kilometres per second), the rock below

Oceanic crust
Moho Continental crust
Upper mantle
upper mantle L I T H O S P H E R E
Moho
Upper mantle
A S T H E N O S P H E R E
100 km approx
Figure 2.8 The mechanically distinct outer layers of the Earth: the rigid
lithosphere and the relatively weak asthenosphere.
the lithosphere is not at rest. It is circulating slowly, at speeds of

a few centimetres a year. This is usually described as ‘solid-state
convection’. Convection is what makes warm air rise and cold
air sink, or causes water to circulate in a saucepan (even before
it boils). It is a way of transporting heat outward. Put simply,
hot mantle rises upwards and transfers its heat to the base of
the lithosphere. Circulating mantle that has lost heat in this
way becomes slightly denser, and sinks downwards to complete
the loop (Figure 2.9 and Plate 1). Most of the heat deposited
at the base of the lithosphere works through to the surface by
conduction, but some is carried higher by molten bodies of rock
that can intrude high into the crust, or even reach the surface at
volcanoes.
core
Figure 2.9 Possible arrangement of convection within the sub-
lithospheric mantle. This model shows two layers of convection cells,
meeting at the seismically defined transition zone (see Figure 2.7).
24

Key idea: Strong versus weak
The Earth’s rocky part is layered by strength (strong lithosphere
and weaker asthenosphere) and by chemistry (core and mantle).
Even the asthenosphere is solid, not molten. However, it is mobile –
flowing by solid-state convection at a few centimetres per year.
The Earth’s heat

Considering the nature of the mantle has brought us to the
question of the Earth’s heat. It is common knowledge that the
interior of the Earth is hot, but most people probably think this
means that below a certain depth everything is molten. You
have already seen that such a notion is misconceived, because
seismic data prove conclusively that the mantle – even the bulk
of it that constitutes the asthenosphere – is virtually all solid.
The mantle is hot, but not as hot as you would think by
extrapolating from surface measurements. Heat is escaping to
the surface at a rate of about eight-hundredths of a watt per
square metre (0.08 W m−2), so you would need to collect all the
heat escaping from an area greater than 1000 m−2 to power a
100-watt light bulb. Near the surface of continents, the average
rate of temperature rise with depth (the geothermal gradient)
is about 30°C per kilometre. From this you could infer, by
extrapolation, a temperature of about 900°C at the base of
30 km thick crust and about 2700°C at the base of 90 km thick
crust, but you would be wrong to do so. Temperatures over
1000°C are not reached in the crust because crustal rocks would
melt at about this temperature.
The solution to this conundrum is that only about half the heat
escaping to the surface of the continents comes from below,
whereas the rest is actually generated within the crust. This means
that the geothermal gradient becomes less as you go deeper.
In fact, below the lithosphere, temperature rises at a rate of less
than half a degree per kilometre. One factor contributing to
this low value is that, as you have already seen, this part of the
mantle is convecting. Convection is much more effective than
conduction at transporting heat.

Spotlight
If convection could (magically) be stopped, then heat would build
up within the mantle and its temperature would rise by a degree
or so every million years until it melted. The action of solid-state
convection within the mantle is actually what stops it becoming
hot enough to melt.
The very effective convection in the liquid outer core keeps the
temperature gradient even lower there, so that the temperature
of the inner core is estimated at a surprisingly low value of
approximately 4700°C.
So where does the Earth’s heat come from? An unknown
proportion, perhaps around 30 per cent, is heat that was
trapped within the Earth as it formed. This primordial heat
continues to leak out, but is not being generated today. The rest
of the heat comes from processes that are continuing today. By
far the most important of these is decay of radioactive elements,
causing so-called radiogenic heating.
Radiogenic heating
Many elements have radioactive isotopes, but only three of
these are important heat sources in the Earth today. These
are potassium, uranium and thorium. Most potassium atoms
are stable and non-radioactive. Each of these contains a total
of 39 heavy particles (protons and neutrons) in its nucleus,
and is described as potassium-39, or 39K (using the chemical
symbol for potassium, K). However, about one potassium
atom in every 10 000 contains an extra neutron, making a
total of 40 heavy particles. These atoms of potassium-40 (40K)
are unstable, and undergo radioactive decay. All atoms of
uranium and thorium are unstable; thorium occurs as 232Th,
and uranium has two radioactive isotopes, 235U and 238U.
Although it is impossible to predict when any individual
radioactive atom will decay, there is a measurable probability
of it happening during a given time. When dealing with large
26

numbers of nuclei, probability takes on a deadly certainty. It
is therefore possible to determine the time it takes for half the
nuclei of a particular isotope to decay, which is known as the
half-life. Because the number of radioactive nuclei halves over
a fixed time, it never actually reaches zero (although it can
get very close to it for isotopes with much shorter half-lives
than those discussed here). The half-life of 238U is 4.5 billion
years. This, coincidentally, is roughly the age of the Earth, so
it means that the abundance of 238U in the Earth must now
be very close to half what it was when the Earth formed.
The half-life of 232Th is over three times the age of the Earth,
so about 80 per cent of the 232Th that the Earth started with
still remains.
Of the four heat-producing isotopes, 40K (half-life 1.3 billion
years) probably produces the greatest amount of heat today and
235U the least. However, 235U has the shortest half-life of the four,
about 710 million years, which is about one-sixth the Earth’s

age. To determine the rate of heat production by 235U when the
Earth was very young you have to multiply the present rate
by 26 or 64 times, from which it can be deduced that 235U was
the second most important heat-producing isotope during the
Earth’s first half-billion years.
All three of these elements tend to be concentrated into the
relatively light rocks of the continental crust, rather than the
mantle, which explains why as much as half of the Earth’s
radiogenic (radioactively produced) heat is generated in the
crust, although this makes up less than 0.1 per cent of the
Earth’s total volume. However, one of the great unknowns is
whether potassium is abundant in the outer core, where it is a
contender for the unknown light element. What is not in any
doubt at all is that the Earth’s radioactive power source must be
running down, though at a rate that is too slow to concern us
for most purposes.
Key idea: Radiogenic heating

Heat is generated inside the Earth by radioactive decay of uranium,
thorium and potassium.

Radiometric dating and the age of
the Earth
Apart from heat production, radioactive decay is important
to the geologist because it is a physical process that occurs at
known rates. The time since a rock or crystal was formed can
be determined by measuring the relative proportions of the
surviving parent isotope and its decay products trapped within
the same rock or crystal. This is called radiometric dating.
For example, 40K (the parent isotope) decays to 40Ar (its decay
product). Argon is not commonly trapped within minerals when
they are formed, so it is usually assumed that any argon found
within a mineral is the product of decay of 40K. Unfortunately,
argon can escape from within a crystal, so potassium-argon
dates are sometimes unreliable.
The decay of uranium (to lead) offers an important dating
technique, but 232Th is not used much because of its extremely
long half-life. There are other radioactive isotopes that are
insignificant producers of heat but whose decay products
enable radiometric dating to be performed. The most important
of these is the decay of rubidium-87 (87Rb) to strontium-87
(87Sr). Radiometric dating techniques can be used to determine
the time since a crystal grew from molten rock or by
recrystallization within rock without melting (a process known
as metamorphism). Extensions of these techniques have enabled
the age of the oldest components of meteorites to be determined
as 4.567 billion years, with the Earth being fully formed by
about 4.5 billion years.
Case study
No reputable scientist today doubts the age of the Earth, as
deduced from radiometric dating. Eighteenth-century geologists
mixed careful observation of the record of past events evidenced
in successive deposits of sedimentary rock and deduced that the
Earth must be immensely old. Notable among these was Scottish
geologist James Hutton, who wrote in 1788 that ‘we find no
vestige of a beginning, – no prospect of an end’. However, it took
a physicist, the Belfast-born William Thomson (later Lord Kelvin),
28

working at the University of Glasgow, who placed a credible number
on the age. In 1862 he published calculations that showed that it
would have taken somewhere between 20 and 400 million years for
the Earth to have cooled to its present temperature from an initially
molten state. He was right, but he didn’t know about radiometric
heating and based his sums on primordial heat only. In the 1920s
and 1930s radiometric dating became available and showed that
the Earth must in fact be a few billion years old, and an age of 4.55
billion years was accepted, based on meteorites, in 1956.
The origin of the Earth and

the Solar System
Now that you have completed this brief survey of the Earth’s
interior and gross composition, it is time to look at how
scientists think the Earth came into being. This takes us back
to the birth of the Sun, which grew by the collapse of a slowly
rotating interstellar cloud of gas (mostly hydrogen) and dust.
This cloud is referred to as the solar nebula. As it contracted,
the solar nebula spun faster and faster. The material that was
not drawn into the central point, where the Sun was forming,
became concentrated in a disc around the Sun. It was within this
disc that the planets grew. At first the cloud would have been
very hot, because of the gravitational energy converted to heat
by the contraction. As it cooled, things began to condense out of
the gas as tiny solid grains. The first grains to condense would
have been made of substances able to form at temperatures
above 1000°C. These would have included nickel-iron metal
and some of the silicate minerals. As the temperature declined
to a couple of hundred degrees centigrade, the minerals growing
could begin to trap water within their structure, but ice crystals
would not begin to grow until the temperature reached about
–90°C.
The tiny grains that formed within the solar nebula would
tend to stick together whenever they happened to come into
contact, and so progressively larger chunks would collect. Once
the process had begun, it may have taken as little as a few
thousand years to form centimetre-sized pieces. Throughout

the sequence of events the timescales are poorly understood,
but the following is a reasonable summary. After about
100 000 years, the biggest blocks had probably grown to about
10 km across and are dignified with the name ‘planetesimals’.
These were now big enough to make their gravitational
influence felt, and their growth would now proceed at an
accelerating rate until after about a further 50 000 years,
most of the planetesimals had collided and accreted into a few
dozen bodies a few thousand kilometres across, now known as
‘planetary embryos’.
Collisions (sometimes referred to as ‘giant impacts’) between
these surviving bodies would have been violent, sometimes
shattering both bodies, but more often pasting the debris
of the smaller body across the face of the larger one and
liberating sufficient heat to cause melting to a very great
depth. After about 100 million years all the planetary
embryos that were going to collide had probably done so,
leaving the Solar System with four inner rocky planets (see
Table 2.1). There were also four giant planets (Jupiter, Saturn,
Uranus and Neptune) further out from the Sun where it had
become cold enough for large quantities of water, methane,
ammonia and similar volatile substances to condense. In the
outer fringes, where the density of material was too sparse
to collect into large bodies, there are only relatively tiny icy
objects, of which Pluto is the most famous – but only the
second most massive – example.
The Earth probably acquired the Moon rather late. It probably
formed from the debris of the last planetary embryo collision to
affect the Earth. In this case, some fragments of the impacting
body bounced back from the Earth and collected together in
orbit around the Earth.
The irregularly shaped rocky and iron-rich bodies known as the
asteroids, which are up to a few hundred kilometres across and
found mostly between the orbits of Mars and Jupiter, were once
thought to be the remains of a planet that was broken up by a
giant collision. However, it now seems that they are surviving
planetesimals that never stuck together when they hit each
other, because their orbits were stirred up by their proximity to
30

Jupiter so that they hit each other too hard to allow accretion.
Most meteorites (which are chunks of rock or nickel-iron that
fall from the sky) are thought to be made of the same material
as the asteroids.
A warning
The origin story as outlined above may sound rather glib.
It is important to realize that you stand on the threshold of
the realm of legend here. Most scientists would accept this
story in outline, but there are many important details still in
dispute. There is good evidence that the majority of Sun-like
stars have their own planetary systems. Curiously, many of the
documented ‘exoplanets’ are in orbits very close to their stars,
which has earned them the apt name of ‘hot Jupiters’. Such
‘solar systems’ are clearly unlike our own, and it is believed that
giant exoplanets such as those must have formed farther away
from their star and then migrated inwards, a process that would
almost certainly destroy any intervening Earth-like exoplanet.
However, these examples are probably unrepresentative, because
giant exoplanets close to their stars are the easiest ones to find.
The first Earth-sized exoplanet with a surface cool enough for
liquid water was discovered as recently as 2007, and there are
hopes that within a further decade or two we may be able to
determine at least the atmospheric composition of such worlds,
even though details of their geology must remain a mystery.
Closer to home, the temperature history of the solar nebula
is uncertain. It could have remained too hot for much rocky
material to have condensed until substantial-sized bodies
of nickel-iron had formed. In this case, the later-formed
rock would have collected around these pre-grown cores,
and planetary bodies would develop with an inbuilt layered
structure. On the other hand, the temperature may have fallen
swiftly enough for chunks of a wide range of compositions to
have been around at the same time. Planetary bodies would then
have grown with a thoroughly mixed composition, and internal
layering would then have to be generated by some later process.
It is not too difficult to envisage how compositional layering
could develop from initially well-mixed material, especially

if a planet like the Earth grew by ‘giant impact’ collisions
between planetary embryos of similar size. As noted above,
these impacts would have been violent enough to cause global
melting, and this would have provided the opportunity for
dense components, such as iron, to sink and form the core.
At the same time, lighter materials, such as the silica-rich
minerals characteristic of the crust and mantle, would have
risen. However, a variant of the above story omits the planetary
embryo stage, and has each planet developing from the
largest planetesimal in its vicinity, growing by accreting other
planetesimals that were always much smaller than itself. In this
scenario, the heat released by impact energy may not have been
enough to cause global melting, in which case differentiation
into a compositionally layered structure would have to depend
on internal sources of heat, such as radioactive decay.
In the next chapter I will put speculation aside, and turn to
much firmer ground (metaphorically speaking!) and look at
what happens in an earthquake.
Dig deeper
Material about the Earth’s interior can be found at
http://www.bbc.co.uk/science/earth/surface_and_interior/
inside_the_earth
If you want to explore the difference between P-waves and
S-waves, try http://www.bgs.ac.uk/discoveringGeology/hazards/
earthquakes/seismicWaves.html
There is some complex information about the Earth’s magnetic
field at http://www.geomag.bgs.ac.uk/education/earthmag.html
32

Fact-check
1 Which of these is liquid?
a The mantle
b The asthenosphere
c The inner core
d The outer core
2 Which of these is richest in silica?

a The continental crust
b The oceanic crust
c The mantle
d The outer core
3 Which of these is the least dense?

a The continental crust
b The oceanic crust
c The mantle
d The outer core
4 What is the main role of convection in the mantle?

a It generates heat
b It drives the internal rotation of the core
c It transports heat outwards
d It carries silica to the crust
5 What was the main reason why oxygen appeared in the initially
oxygen-free atmosphere of the Earth?
a Sunlight
b Microbial activity
c Volcanic eruptions
d Temperature decrease
6 Which layer of the Earth does not allow S-waves to propagate

through it?
a The crust
b The mantle
c The outer core
d The inner core

7 Which of these produces most radiogenic heat in the Earth
today?
a 40K
b 232Th
c 235U
d 238U
8 What is isostasy?
a Different numbers of neutrons in nuclei of the same element
b A way of looking at relative buoyancy
c The way that the core generates the Earth’s magnetic field
d The way that planets grew from the solar nebula
34

3
Earthquakes
33about earthquakes and how they are
measured and predicted
33about the dangers posed by
earthquakes, and how risks can be
minimized.
35

You have already seen that earthquakes are useful for probing
the interior of the Earth. This is because the seismic waves
generated by earthquakes can travel right through the planet.
However, earthquakes themselves do not happen deep down,
and are pretty much confined to the lithosphere. It is their very
shallowness that makes the larger ones so devastating. In this
chapter I will concentrate on describing earthquakes. Their
fundamental causes will become apparent in the next chapter.
How earthquakes happen

There are many fault lines in the Earth’s crust. The San Andreas
fault in California is probably the most famous. A fault is a fracture
between tracts of crust that are moving relative to one another.
The typical average rate is around a millimetre per year. If this
movement were to happen gradually it would pose few problems
for people living nearby. Unfortunately, faults do not behave that
way. Instead, they tend to stick. Strain builds up for decades or
centuries until it reaches a critical level, and then everything gives
at once. The principle is shown in Figure 3.1. An imperfect analogy
to what goes on at the break-point of an earthquake happens if
you stack a house brick on top of another. If you hold the lower
one while gradually tilting it, you will find that the upper one does
not move at first. As the angle of tilt increases, the force tending to
make the top brick slide gradually increases until it is great enough
to overcome friction, and suddenly it will slide off.
(a) (b) (c) (d)

Figure 3.1 Plan view showing a road built across a fault (dashed line).
Strain builds up gradually (a)–(c), as the edges of the rock masses on
either side of the fault are bent elastically. Eventually (d), the strain
becomes great enough to overcome the frictional resistance along the
fault and the fault gives way. Slip on the fault is rapid and releases a lot of
pent-up energy as seismic waves.
Key idea: What causes an earthquake

An earthquake occurs when there is sudden movement across a fault.
36

Remember this!
Because the rock on opposite sides of a fault sticks while strain
builds up and then gives way without sudden slippage, the process
is referred to as ‘stick-slip’.
As soon as a fault has given way at one point, adjacent lengths

of the fault will slip also, but movement is unlikely to occur
along its whole length, which may be hundreds or thousands
of kilometres. Other than in exceptionally large earthquakes,
movement is usually restricted to a much shorter portion of
the whole fault. The strongest seismic waves are generated
at the initial break-point. The closer to the break-point, the
greater the energy and the greater the potential for destruction,
especially if the break-point is near the surface. Technically, the
break-point is called the earthquake focus, and the point on the
surface directly above it is referred to by that much-abused term
epicentre. Slip farther along the fault and readjustments close
to the focus usually cause a series of smaller aftershocks, which
continue for days (even years in extreme cases) after the initial
earthquake.
Little, if any, of the damage caused by an earthquake is done
by the P-waves and S-waves that you met in the previous
chapter. Those waves travel through the body of the Earth, and
their energy is spread over a rapidly increasing volume as they
propagate. It is other waves of the sort that travel along the
surface of the ground that do most of the harm. These include
rolling waves (like waves on the ocean, called Rayleigh waves)
and strong side-to-side shaking waves (called Love waves). The
P- and S-body waves travel faster than the surface waves, and,
if felt or detected by a seismometer, can give a few seconds’ (or
minutes’, if farther away) warning of the arrival of the more
damaging surface waves.
Key idea: Earthquake damage

Most earthquake damage is done by seismic waves that travel
along the surface, not by those that travel inside the Earth.
3 Earthquakes 37

Measuring earthquakes
When an earthquake makes the news, its size is usually
described by a number on the Richter scale. This scale (of
which there are several variants) is an attempt to specify how
strong the seismic waves must have been at the focus. This can
be estimated using the signal detected by a single seismometer
at a known distance from the focus or, better, by comparing the
signals detected at many seismometers at a variety of distances.
Spotlight: Charles Richter

The original Richter scale was devised in 1935 by the American
Charles Richter (1900–1985) and his colleague Beno Gutenberg (1889–
1960), whose name has rather unfairly been ignored by posterity.
The Richter scale is numerical, enabling comparative data to

be compiled on earthquakes of all sizes from around the globe.
For each one magnitude increase on the scale, the motion (or
amplitude) of the seismic waves at the focus increases tenfold
and the energy released increases roughly thirty-fold. The
most devastating earthquakes tend to be of magnitude 6.0 and
above. Fortunately, there is on average only one so-called ‘great
earthquake’ (an earthquake of magnitude 8.0 or stronger) each
year. Table 3.1 summarizes the occurrence of earthquakes on the
Richter scale.
Case study: Beware the Richter scale on its own

A disadvantage of the Richter scale is that it does not show how
much damage a particular earthquake is capable of. For example,
deep-focus earthquakes are inefficient at generating surface
waves, so they rarely cause much damage. Furthermore, what
effect an earthquake has depends on when and where it occurs.
A shallow magnitude-8.0 earthquake in the middle of nowhere, will
probably cause much less damage than a shallow magnitude-6.0
earthquake whose epicentre is in a major city, especially if it
happens during the rush hour. For example, on 24 May 2013 a
magnitude-8.3 earthquake at a depth of 600 km below the sea
west of Kamchatka (eastern Russia) did no harm whatsoever.
38

The effect of an earthquake also depends on the nature of the rock
near the surface. In the Mexico City magnitude-7.5 earthquake of
1985 (see Table 3.1), the soft sandy subsoil quivered so vigorously
that some buildings sank deep into it, causing many deaths. A
similar earthquake below a city built on hard rock would probably
have caused much less damage.
Table 3.1 Numbers of earthquakes per year in each magnitude range of the
Richter scale. Some noteworthy earthquakes of each magnitude are listed.
Magnitude Number
range per year Example (number of deaths in brackets)
8.0 and 1 San Francisco 1906 M8.3 (500); Tokyo 1923 M8.2
above (143 000); Kuril Islands 1994 M8.2 (12); Sunda Trench,
Sumatra 2004 M9.1 (283 000); Sunda Trench, Sumatra
2005 M8.7 (1,313); Coast of Peru 2007 M8.0 (514); Samoa
2009 M8.0 (192); Tohoku, Japan 2011 M9.0 (18 000)
7.0–7.9 18 Tangshan, China 1976 M7.6 (250 000); Mexico City 1985
M7.9 (30 000); Loma Prieta (San Francisco) 1989 M7.1 (61);
Kobe, Japan 1995 M7.2 (5000); Izmit, Turkey, 1999 M7.6
(17 118); Bhuj, India 2001 M7.7 (20 000); Muzzafarabad,
Kashmir 2005 M7.6 (80 000); Chengdu, Sichuan, China
2008 M7.9 (86 633); Sunda Trench, Sumatra 2008 M7.6
(1,117); Port au Prince, Haiti 2010 M7.0 (230 000);
Lamjung, Nepal, 2015 M7.8 (>7365)
6.0–6.9 134 Avellino, Naples, Italy 1980 M6.8 (8,000); Spitak, Armenia
1988 M6.8 (25 000); Northridge (Los Angeles) 1989 M6.7
(61); Bam, Iran 2003 M6.6 (26 200); L’Aquila, Italy 2009
M6.3 (283); Christchurch, New Zealand 2011 M6.3 (185)
5.0–5.9 1300 Plattsburgh, New York 2002 M5.1 (none); San Giuliano di
Puglia, Italy 2002 M5.3 (29); Market Rasen, England 2008
M5.2 (none)
4.0–4.9 13 200 Dudley, England 2002 M4.8 (none); Kent, England 2007
M4.3 (none); Running Springs, California 2014 M4.6 (none)
3.0–3.9 130 000 Fort William, Scotland 2005 M3.0 (none); Oakham,
England 2015 M3.8 (none)
2.0–2.9 1 300 000 Mansfield, England 2014 M2.6 (none)
< 2.0 millions
An alternative to the Richter scale is the Modified Mercalli scale.

This ranks earthquakes according to the intensity of their effects,
and can be used to map how concentric zones around the epicentre
are affected by a single earthquake. A version of it is shown in
Table 3.2. An alternative known as the European Macroseismic
scale is very similar, running from 1 to 12 instead of I to XII.
3 Earthquakes 39

Table 3.2 The Modified Mercalli scale of perceived earthquake intensity.
The more intense earthquakes are categorized by structural damage, less
intense earthquakes by effects perceived by people on the spot.
Mercalli intensity Description
XII Damage total. Lines of sight and level distorted. Objects

thrown in the air.
XI Few if any masonry structures remain standing. Bridges
destroyed. Rails bent greatly.
X Some well-built wooden structures destroyed; most masonry
and frame structures destroyed with foundations. Rails bent.
IX Damage considerable in specially designed structures; well-
designed frame structures thrown out of plumb. Damage great
in substantial buildings, with partial collapse. Buildings shifted
off foundations.
VIII Damage slight in specially designed structures; considerable
damage in ordinary substantial buildings, with partial collapse.
Damage great in poorly built structures. Fall of chimneys,
factory stacks, columns, monuments, walls. Heavy furniture
overturned.
VII Damage negligible in buildings of good design and
construction; slight to moderate damage in well-built ordinary
structures; considerable damage in poorly built or badly
designed structures; some chimneys broken. Difficult to stand
upright. Noticed by vehicle drivers.
VI Felt by all, many frightened. Some heavy furniture moved; a
few instances of fallen plaster. Damage slight.
V Felt by everyone; many awakened. Some dishes and windows
broken. Unstable objects overturned. Pendulum clocks may
stop.
IV Felt by many indoors, outdoors by few. At night, some
awakened. Dishes, windows, doors disturbed; walls make
cracking sound. Sensation like heavy truck striking building.
Standing cars rock.
III Felt quite noticeably by people indoors, especially on upper
floors. Many people do not recognize it as an earthquake.
Hanging objects swing. Feels like passing traffic.
II Felt only by a few people at rest, especially on upper floors.
I Generally detected by instrument only.
Some of the descriptions on the Mercalli scale are universally

applicable, such as whether the earthquake can be felt by people
who are walking about, or only by those lying down. However,
the amount of damage to buildings depends very much on the
nature of their construction. A well-designed structure can
40

withstand even a very major earthquake without collapsing,
even though it may be so badly damaged that the only safe
thing to do afterwards is to demolish it. Most earthquake-prone
regions have building codes designed to ensure that structures
are relatively collapse-proof, but failure to observe and enforce
these codes has led to many unnecessary deaths.
Remember this!
Schools in particular tend to collapse too easily, because building
regulations are ignored in order to save on construction costs.
Key idea: Richter versus Mercalli

The Richter scale measures the energy of an earthquake at its
source, whereas the Mercalli scale measures its felt intensity (and
hence likely damage) at a particular place.
Protecting structures from

earthquake damage
There are many ways to make buildings, elevated highways
and other structures resilient to earthquakes. It is not simply a
matter of making everything stronger. It is more important to
ensure that if a structure is rocked by an earthquake, the whole
of it sways together in the same direction (Figure 3.2).
As well as tying walls together so that they sway in unison, it
is important to ensure that a building can sustain any twisting
motion engendered by an earthquake. Digging extra-deep
foundations can also help, so that the structure rests on solid
bedrock where the amplitude of the vibrations is likely to be
least. If this cannot be achieved, then a building should be
situated where its foundations can rest on the same material all
round, so any subsidence happens evenly on all sides.
Another factor is the natural frequency at which a building
tends to shake. Taller buildings sway more slowly than low-
rise buildings, and there is a danger that high-rise buildings
will be ‘in tune’ with the shaking of the ground. This had an
3 Earthquakes 41

(a) (b) (c) (d)
(e) (f) (g) (h)
Figure 3.2 Two series of cross-sections to show the possible response

of a simple structure (two walls and a roof) to the ground-shaking
caused by an earthquake. In (a)–(d) both walls sway in the same direction
simultaneously, and the structure survives with relatively minor damage.
In (e)–(h) the walls sway in opposite directions, with catastrophic results.
Here the roof collapses, even though the walls manage to stay upright. In
inadequately designed multi-storey buildings, several floors have been
known to collapse on top of one another, crushing anyone unfortunate
enough to have been inside.
unfortunate consequence in the major earthquake that struck

Mexico City in 1985. In the most heavily damaged area of the
city, about 70 per cent of buildings between 10 and 20 storeys
high were damaged, because the low-frequency shaking of
the ground became amplified within the building by resonant
vibrations of their structure. In contrast, fewer than 20 per cent
of buildings less than five or more than 20 storeys high were
damaged.
Searching for survivors after an earthquake is a risky business.
Buildings that are apparently unscathed may in fact be
dangerously near to collapse, which could be triggered by an
aftershock at any time. There may also be a risk of explosions
from ruptured gas pipes. Fire is actually one of the biggest
hazards in the aftermath of a major earthquake.
Spotlight
Most of the death and destruction associated with the San
Francisco earthquake of 1906 was because of fires, caused by
toppled lamps and stoves. The majority of the buildings survived
the quake, but were then burned to the ground.
42

Many lessons have been learned about how to build earthquake-
resistant structures. However, even in the wealthier at-risk areas
large numbers of old and less resistant structures remain, as
demonstrated by some of the damage caused by late twentieth-
century earthquakes in California (Figure 3.3) and Japan.
Figure 3.3 This building in San Francisco was wrecked when its lower
floors gave way during the 1989 Loma Prieta earthquake.
Earthquake preparedness
In earthquake-prone regions, most people are well aware of the
risks they face. Instructions on how to be prepared, and what
to do in the event of an earthquake, are widely available and
are commonly rehearsed in schools. The following is an extract
from the advice pages of a Los Angeles telephone directory.
During an earthquake
1 If you are indoors, DUCK or drop down to the floor. Take COVER
under a sturdy desk, table or other furniture. HOLD on to it and
be prepared to move with it. Hold the position until the ground
stops shaking and it is safe to move. Stay clear of windows,
fireplaces, and heavy furniture or appliances. Don’t rush outside.
You may be injured by falling glass or building parts.
DO NOT try using the stairs or elevators while the building is
shaking or while there is danger of being hit by falling glass or
debris.
3 Earthquakes 43

2 If you are outside, get into the open, away from buildings and
power lines.
3 If you are driving – STOP if it is safe – but stay inside. DO NOT
stop on or under a bridge, overpass or tunnel. Move your car as
far out of the normal traffic pattern as possible. DO NOT stop
under trees, light posts, electrical power lines or signs.
4 If you are in a mountainous area, be alert for falling rock and
other debris that could be loosened by the quake.
5 In a crowded public place, do NOT rush for the exits. Stay calm
and encourage others to be so.
Even when there is no structural damage, the shaking that

occurs during an earthquake can cause breakage and injury.
Simple precautions include not hanging pictures, especially
glass-fronted ones, over the head of one’s bed, and securing
china ornaments in place with ‘earthquake wax’.
Key idea: Keep safe

If you are caught indoors by an earthquake, shelter under a table
until it is over.
Prediction and prevention

The business of predicting earthquakes is not simple. Usually,
the first rupture is the biggest, and the aftershocks in the
following weeks are of decreasing magnitude. If, instead,
earthquakes habitually built up towards a crescendo, it would
be more obvious when a big earthquake was due! One way
to estimate where an earthquake is likely to occur is to map
out areas along a fault where earthquakes have not been
happening lately (Figure 3.4). If a length of fault can be
identified with no recent break-points, but the lengths of fault
to either side have many recent epicentres, then it is probable
that a large amount of strain has built up in the seismically
quiet bit in the middle. This strain will have to be released
sooner or later, perhaps by a major rupture, like the snapping
44

of an elastic band stretched to breaking point. The trouble
is that the method of pinning down exactly when this will
happen has so far proved elusive.
city
fault
epicentres
magnitude <3
magnitude 3–4
magnitude 4–5
Figure 3.4 Map of a fault and the surrounding area, showing earthquake
epicentres over a 30-year period. The central part of the fault has not
ruptured recently, so the prospects of a major earthquake epicentre in or
near to the city in the next few years look pretty grim.
Not much can be done to prevent earthquakes happening.

One suggestion has been to pump water into fault planes
in areas where strain build-up is indicated by lack of recent
seismic activity. The hope is that the water will act as a
lubricant, allowing the crustal blocks to begin to slip past
each other gently instead of violently. However, such a
course of action is risky. Imagine the chances of re-election
for a politician on whose orders this was done, if (in order
to forestall a magnitude-7 earthquake that was likely some
time in the coming decade) the lubricated fault moved in such
a fashion as to cause a series of magnitude-5 earthquakes.
Even if only minor damage occurred, the body responsible
for the fault lubrication programme would get the blame,
and the lawyers would have a field day trying to establish
financial liability!
Tsunamis
One other aspect of earthquakes deserves to be mentioned, and
this is a potentially tragic consequence of earthquakes under the
3 Earthquakes 45

ocean. If the focus is at a shallow depth below the sea
floor, the sudden displacement of rock can create a series of
waves travelling through the ocean at a speed of several hundred
kilometres per hour. In deep water, such waves may be scarcely
noticeable, being less than a metre high and very gently sloped.
However, upon reaching shallow, coastal water, the waves slow
down and become both higher and steeper, with the result that
they may spill over the shore with tremendous force, sweeping
buildings (and people) away even tens of metres above sea
level. Popularly (but incorrectly) called a ‘tidal wave’, this
phenomenon is more correctly referred to by its Japanese name
of tsunami. A tsunami can also be triggered by landslides or
undersea volcanic eruptions.
Spotlight
Even though shaking caused by the 2011 magnitude-9.0 Tohoku
earthquake (Japan) reached intensity IX in some cities, virtually all
the fatalities resulted from a tsunami because buildings had been
designed to be resilient to earthquake damage.
Surprisingly, the first manifestation of a tsunami at the shore is

sometimes a fall in sea level lasting for several minutes, which
unfortunately can lure the curious on to the beach where they
are especially vulnerable to the dramatic rise in water level that
follows. The reasons for this are complex; on coasts close to the
epicentre, the initial fall in local sea level may be because the
land has been displaced upwards by the fault motion, but more
generally it happens when the first part of a wave to arrive is a
trough rather than a crest.
Key idea: Crest of a wave?

This shouldn’t surprise you. Generally speaking, it’s a 50:50 chance
whether a wave arrives crest-first or trough-first, but many people
visualize waves just as crests and forget about the troughs.
46

The wide-ranging effect of tsunamis is demonstrated
by the case of the Hawaiian Islands, whose north shores
face the earthquake-prone belt of Alaska and the Aleutian
Islands, over 3 000 km away. The largest historic tsunami
to hit Hawaii happened in 1946, and was triggered by an
earthquake in the Aleutians. This wave broke on the shore
and reached up to 17 m above normal sea level, sweeping
away entire villages and taking the lives of 159 people.
Today, yellow tsunami-warning speakers mounted on
telephone poles are a familiar sight in parts of the Hawaiian
Islands. It takes a tsunami several hours to travel from the
Aleutians to Hawaii, but there is much less warning time in
the case of tsunamis generated locally.
On 26 December 2004 a magnitude-9.1 earthquake, with its
focus at a depth of 30 km in the Sunda Trench, offshore of
Sumatra, happened when the rock on either side of a 400-km
length of fault jerked by about 10 m, rupturing an area of
fault surface almost the size of California. This triggered a
tsunami that devastated many shorelines around the Indian
Ocean, and caused almost all of the 283 000 deaths attributed
to this earthquake in Table 3.1. Tragically, there was no
tsunami early-warning system in the Indian Ocean, though
one has now been installed. This could have saved tens of
thousands of lives in Thailand, Sri Lanka and elsewhere.
However, Sumatra and the nearby small islands, where more
than 200 000 people were killed by inrushing water up to
30 m above the normal high-water mark, were so close to the
source that even the best early-warning system would have
saved very few.
The nearly 200 deaths caused by the 2009 magnitude-8.0
Samoa earthquake listed in Table 3.1 (whose undersea focus
was only 18 km deep) were a result of tsunami waves striking
Samoa, American Samoa and Tonga within a few hundred
kilometres of the epicentre.
3 Earthquakes 47

Key idea: Earthquakes and tsunamis
An earthquake whose focus is at a shallow depth below the sea
can cause a tsunami.
Dig deeper
More background to earthquakes and types of seismic waves
can be found at www.bgs.ac.uk/discoveringGeology/hazards/
earthquakes/
For advice on how to prepare for and survive an earthquake, see
http://www.earthquakecountry.info/dropcoverholdon/
The best website for up-to-date information on earthquakes
anywhere in the world is http://earthquake.usgs.gov/
48

Fact-check
1 Which of these is correct?
a The Richter scale and the Modified Mercalli scale are
different ways of measuring the same thing.
b The Richter scale measures the intensity of shaking caused
by an earthquake.
c The Modified Mercalli scale and the European Macroseismic
scale are approximately equivalent.
d The Modified Mercalli scale measures the strength of seismic
waves at an earthquake’s focus.
2 Which term is used to describe the process by which faults

move intermittently?
a Start–stop
b Slip-slop
c Stick-slip
d Creep-crack
3 If you are indoors and feel a powerful earthquake, what should

you do immediately?
a Shelter below a table.
b Run upstairs to fill the bath with water.
c Hurry outside.
d Check online to see if it is real.
4 Which of these is incorrect?

a When a tsunami arrives at a coastline, it can do so either
crest-first or trough-first.
b A tsunami can travel thousands of km across an ocean.
c Deep foundations make a building more resilient to an
earthquake.
d In an earthquake, tall buildings tend to sway faster than low-
rise buildings.
3 Earthquakes 49

Plate tectonics
4
33how the Earth’s outer shell is broken
into plates that glide slowly across the
weaker interior
33how this enables creation of new
lithosphere and destruction of old
lithosphere at plate boundaries marked
by earthquakes and volcanoes.
51

I have so far discussed earthquakes without paying much
attention to where in the world they tend to occur. Certain parts
of the globe suffer more earthquakes than others, and you will
have experienced at least a few minor tremors if you live in a
high-risk area.
Spotlight
Living in central England, not a region noted for its earthquakes,
I was woken up by Mercalli IV shaking from a magnitude-4.2
earthquake 50 km away in September 2000. However, I slept
through Mercalli IV shaking from a magnitude-5.2 earthquake
140 km away in February 2008, which was embarrassing as I was
asked about it on local radio a few hours later!
A glance at a map of earthquake distribution, such as

Figure 4.1, shows that earthquakes tend to occur in belts. The
most famous of these stretches all the way up the west coast
of the Americas, through the Aleutian and Kuril Islands and
southwards round the western side of the Pacific Ocean into
New Zealand, so that the belt almost encircles the Pacific
basin. This belt is also where many of the most active on-land
volcanoes occur, and has been called the Pacific Ring of Fire.
Figure 4.1 Map showing the distribution of earthquake epicentres over a

typical six-year period.
52

The coincidence of earthquakes and volcanic zones is no
accident, and the relationship between these two quite
different phenomena will become clear in this chapter and the
next. Earthquakes also occur beneath Indonesia on the north-
eastern fringe of the Indian Ocean, in a broad belt running
through the Mediterranean and eastwards in central Asia
(where there are very few volcanoes), in eastern Africa, and
in some much narrower zones, snaking their way through the
oceans, notably a zone running from north to south through
the Atlantic Ocean.
Moving plates
Earthquake zones are a manifestation of a phenomenon known
as plate tectonics, but the earthquake zones mentioned above fit
into the picture in various ways. Plate tectonics is the term used
to describe the manner in which the Earth’s lithosphere moves
around. As established in Chapter 2, the relative weakness
below the uppermost mantle allows the lithosphere to slide
about. However, the lithosphere is not a single unit. It is broken
into a series of rigid plates, usually described as seven major
plates and six minor plates (these are named on Figure 4.9).
Most plates carry areas of both continental and oceanic crust.
The Pacific Plate is the largest but is a notable exception in
having no substantial areas of continent.
Each plate is in contact with its neighbours on all sides, but
the plates are moving relative to one another. It is important to
realize that there are no gaps between these plates, so there are
no chasms open clear down to the asthenosphere. To see how
plates are able to move around without any gaps appearing,
I will describe processes at the margins of plates.
Key idea: Plates and plate tectonics

The lithosphere is broken into plates, which move across the
underlying asthenosphere. This motion is plate tectonics.

Plates colliding
First, what happens when plates are moving towards each
other? This situation is typified by Japan, shown in cross-section
in Figure 4.2. Japan is a piece of continental crust on the eastern
edge of a major plate. The floor of the Pacific Ocean belongs
to a different plate, which is moving towards Japan at a rate of
about 10 cm per year. Where the two plates meet, one is being
thrust down below the other, a process described as subduction.
Because oceanic crust is denser than continental crust, when two
plates meet it is almost invariably the oceanic plate that goes
under, as in this example. Under Japan, the Pacific Plate forms a
slab descending at an angle of about 45 degrees, but examples
of much steeper and much shallower subduction are found in
comparable situations elsewhere.
W volcanoes E E
sea level KEY
lithosphere
{ *** *
** *
continental crust
oceanic crust
* ** lithospheric
* **** mantle
asthenosphere * *
* ** asthenosphere * earthquake
* focus
*
100 km
Figure 4.2 Cross-section through Japan (see text for explanation).
As one plate slides over another, the front of the overriding

plate is compressed, and the rocks there may become buckled.
At the plane of movement between the two plates, motion is
far from uniform, progressing in the stick–slip fashion that
you encountered in the previous chapter. This gives rise to
earthquakes, and it was by plotting the depths of earthquake
foci that subduction zones were first recognized for what they
are. As you can see in Figure 4.2, earthquakes become deeper
from east to west under Japan. A similar situation, though an
east–west mirror image, occurs below Sumatra. Here the floor
of the Indian Ocean is being subducted below Indonesia, and
it was fault motion in that subduction zone (marked on the sea
54

floor by the Sunda Trench) that led to the tragic earthquake and
tsunami of 26 December 2004.
Inclined zones of earthquakes can be traced down to about
700 km, but no deeper. This is because, as a plate descends,
it is warmed by heat from the surrounding asthenosphere, so
it eventually ceases to be recognizable. The first part of the
subducting slab to lose its identity is its crust, because crustal
rocks melt at a lower temperature than the temperature of
the sub-lithospheric mantle that they encounter. The melted
material rises upwards to feed volcanoes, which tend to occur
in a belt about 70 km above the subduction zone. I will describe
volcanoes in more detail in the next chapter.
The top end of a subduction zone is marked by a trench on the
sea floor. Off northern Japan, the trench is a fairly typical 8 km
deep, but further south, the trench reaches over 11 km, which is
the greatest known depth in the world’s oceans.
When two plates meet at a subduction zone, one of them is
destroyed, so this setting is described as a destructive plate
boundary. However, it is only the oceanic part of a plate that
can be destroyed in this way. Let’s consider what happens
over time when both plates contain continents (Figure 4.3). At
first, subduction proceeds as normal, while the oceanic part of
one plate descends below the other. However, eventually the
continental part of the subducting plate reaches the subduction
zone. Continental crust is thicker and more buoyant than
oceanic crust, and this prevents the plate from continuing to
subduct. The edges of both continents become buckled, but one
will eventually be thrust over the other. Soon afterwards, the
subduction zone jams. The oceanic part of the descending slab
breaks free, leaving the two plates joined together above. Near
this suture, the crust may be double its normal thickness, and
this is where the highest mountains are found.
Remember this!
Where one plate passes below another, the process is called
subduction.

volcanoes
KEY
continental crust
oceanic crust
lithospheric
mantle
(a) asthenosphere
100 km
(b) (c)
Figure 4.3 Time series of cross-sections to show a continent–continent

collision (see text for explanation).
A recent example of such a continent–continent collision began

about 30 million years ago when India collided with Asia. You
can see the aftermath of this event in the high elevation of the
Tibetan Plateau, fringed to the south by the Himalayas. Another
consequence of this collision, and others that accompanied it,
is the diffuse belt of earthquakes running eastwards from the
Mediterranean through central Asia. As for the vanished ocean,
the only remains are a few slivers of oceanic crust and upper
mantle caught up in the suture zone, known as ophiolites.
Remember this!
An ophiolite is a slice of oceanic lithosphere that has escaped
subduction and been preserved on land.
Plates moving apart

So much for what happens where two plates are moving
towards each other. This cannot be the only sort of interaction
between plates, because if it were, the Earth’s surface would
56

have to be shrinking all the time. In fact, it is now known
that new ocean floor is being created at a rate sufficient to
compensate for the loss of oceanic parts of plates at destructive
plate boundaries. These sites are known as constructive plate
boundaries.
What goes on there is summarized in Figure 4.4. As two
oceanic plates are drawn apart, in a process referred to as
‘sea-floor spreading’, the underlying asthenosphere wells
upwards to avoid any gaps appearing. As the upwelling
asthenospheric mantle approaches the surface, it cools and
becomes part of the lithosphere belonging to the plates on
either side of the boundary. This new lithosphere is still
relatively warm. This makes it slightly less dense and more
buoyant than the older, colder lithosphere farther from
the boundary – so these boundaries are marked by ridges.
Typically, the crest of such a ridge lies at a depth of 2–3 km
below sea level, whereas the expanse of ocean floor to either
side is at an average depth of 4–5 km. There are earthquakes
associated with the rifting and subsidence of oceanic
lithosphere near constructive plate boundaries, but these are
all at shallow depths, giving a seismic belt entirely different in
character from a subduction zone.
sea level KEY

lithosphere
{ oceanic crust
lithospheric
mantle
asthenospheric mantle
100 km
Figure 4.4 Cross-section through a constructive plate boundary,
where upwelling asthenosphere accretes to the diverging edges of two
lithospheric plates. The oceanic crust is a result of partial melting of the
upwelling asthenospheric mantle (see text for explanation).
In addition, a small percentage of the upwelling mantle

melts. This is not because there is a heat source here; it
is a consequence of the drop in pressure, and is known
as decompression melting. When mantle of peridotite

composition (about 45 per cent silica, SiO2) begins to melt, the
molten rock (or magma) produced has a slightly higher silica
content. This is what gives rise to oceanic crust (Table 2.2)
where the composition is, on average, about 49 per cent silica.
This composition (described as ‘basaltic’, ‘basic’ or ‘mafic’)
results in the rock type known as basalt. The residual mantle
left behind has a reduced silica content (to compensate for the
enrichment of silica in the crust). However, because
the volume of melt produced is very much smaller than the
volume of mantle contributing to the melt, the chemical
change in the mantle is slight.
Key idea: Decompression melting

Decompression melting is partial melting that occurs as a result
of decreasing pressure, when mantle rock is carried upwards by
solid-state convection.
The ability of rock to melt in this way, so that a small volume

of melt enriched in silica (and other components) is produced,
is a very important process in geology, and is known as partial
melting. The whole of the oceanic crust has been produced
by partial melting of the mantle. You will look in more detail
at how the oceanic crust is constructed, and in particular
at the sub-sea volcanism associated with constructive plate
boundaries, in the next chapter.
Key idea: Partial melting

When rock melts, it does so by partial melting. This produces a
melt richer in silica than the bulk composition, and leaves a solid
residue poorer in silica than the bulk composition.
Decompression melting does not need a heat source. In fact,

partial melting to produce magma inside the Earth rarely
involves the addition of heat. There is more on this in the next
chapter.
58

Case study: Magnetic stripes on the ocean floor
One of the most compelling lines of evidence in favour of sea-
floor spreading is provided by magnetism. Rocks that have cooled
down from a melt inherit their magnetization from the Earth’s
magnetic field as it was when they cooled. Such rocks forming
today have their magnetization aligned with the present-day field.
However, the direction of this field has flipped many times in the
past, and rocks that formed when the field was in the opposite
direction are magnetized in the reverse direction too. It is not
known why the Earth’s magnetic field reverses, but it provides
another argument, if one were needed, for discounting the idea of
a giant internal bar magnet.
In the 1950s studies of the magnetization of the sea floor (made

by towing sensitive instruments behind research vessels) began
to reveal a curious stripy pattern. Half the stripes are magnetized
in the direction of the present field, but half are magnetized in
the reverse direction. This was elegantly explained in 1963 by
two British geoscientists, Fred Vine and Drummond Matthews,
who suggested the explanation of a mid-ocean ‘tape recorder’
(Figure 4.5). As new ocean floor is added at a constructive plate
boundary, it is magnetized according to the direction of the
Earth’s magnetic field. It retains this as it moves away. Reverse
magnetized stripes represent ocean floor formed at times when
the Earth’s field was also in the reverse direction, resulting in a
‘bar code’ pattern.
age (millions of years)
3
(a) (b )
Figure 4.5 (a) The mid-ocean ‘bar code’. Normal (black) and reverse
(white) magnetic stripes on the ocean floor show the polarity of the earth’s
magnetic field at the time when each piece of ocean floor was added to the
edges of the plates at a constructive plate boundary. The pattern is a mirror
image on opposite sides of the ridge. (b) The magnetic polarity timescale.

The dates of reversals in the Earth’s magnetic field, extending
back for a couple of hundred million years, have been determined
by studying the magnetization of volcanic rocks on land, whose
age has been established radiometrically. As a result, the age of
the ocean floor at any point can be found just by counting back the
number of magnetic stripes from the constructive plate boundary,
or by recognizing a distinctive ‘bar code’ pattern of reversals of
different durations, without the need to collect a sample.
The break-up of a continent

If a linear belt of upwelling in the asthenosphere establishes
itself below a continent instead of an ocean, the continental
crust first becomes stretched thin, then it ruptures. Eventually
a constructive plate margin may develop in between, with
formation of new oceanic crust. This is illustrated in Figure 4.6.
asthenospheric mantle
KEY
continental crust
oceanic crust
lithospheric
mantle
100 km
Figure 4.6 Time series of cross-sections to show the splitting of a continent

to form a new ocean. Once sea-floor spreading has begun, it continues at
a rate typically between 1 cm and 20 cm per year. Subsidence along some
of the faults that were initiated during the initial rifting event may continue
sporadically at the rifted edges of the continents for tens of millions of
years. This is the biggest cause of earthquakes not at plate boundaries.
60

Something of the sort seems to have been responsible for
the origin of the Atlantic Ocean. The apparent fit of South
America and Africa, if you imagine the Atlantic closed (and
not just the fit of their coastlines, but of more fundamental
geological structures too) has long been recognized. It gave rise
to theories of moving continents, described as continental drift.
The discovery of magnetic stripes on the ocean floor, and the
identification of subduction zones and other plate boundaries in
the 1960s, provided the modern explanation in the form of the
theory of plate tectonics.
Spotlight: Continental drift

Although others had previously suggested elements of the idea,
the hypothesis of continental drift is usually credited to the
German Alfred Wegener (1880–1930) in a book dated 1912. Few
geologists gave much credence to Wegener’s ideas during his
lifetime: partly because he was primarily a meteorologist, partly
because at the time geologists could conceive of no mechanism to
allow continents to move, and partly because the special nature of
ocean crust had not yet been discovered.
The life-cycle of an ocean

The rates of ocean floor creation, though varying from place
to place, are usually approximately equal on either side of
a constructive plate boundary. In consequence, the ridge
marking this boundary in the Atlantic Ocean is midway
between the two shores. Constructive plate boundaries in
general are therefore often called ‘mid-ocean ridges’. This is a
misnomer, because eventually the oceanic part of one or other
of the plates becomes detached from the continental part and
begins to subduct beneath it (Figure 4.7), as has happened
in the Pacific and Indian Oceans. Whether or not the ocean
continues to become wider depends on whether the rate of
subduction can outpace the rate of sea-floor spreading – but
subduction ensures that the constructive plate boundary does
not remain central in the ocean.

(a)
100 km
(b)
(c)
(d)
Figure 4.7 Time series of cross-sections to show the evolution of an ocean,

formed in the way shown in Figure 4.6. The scale is reduced here, and the
crust has not been distinguished. The stage reached by the Atlantic Ocean
is shown by (a), though it is wider than shown here. In stage (b) the left-
hand plate has split, and the oceanic part is being subducted beneath the
mainly continental part. The Pacific Ocean today is at stage (c).
Figure 4.7 shows what will almost certainly be the eventual

fate of the Atlantic Ocean. If you imagine this as a west–east
cross-section, then it implies subduction (with its attendant
earthquakes and volcanoes) beneath New York. However,
it could equally well happen on the European side instead.
Eventually, the constructive plate boundary could be drawn
into the subduction zone and, once it has been removed in this
way, the ocean must grow narrower more rapidly until, as in
Figure 4.3, there is a continent–continent collision.
Remember this!
It is thought that the cycle of continental splitting, ocean opening,
ocean closing and continental collision takes on average 400–500
million years.
62

You may have noticed when reading the above that, whereas
ocean floor is continually created and destroyed, there has been
no mention of formation of continental crust. Continental crust
is split and reassembled in ever-changing configurations by
the action of plate tectonics. Although it can be worn away by
the erosional part of the rock cycle (Figure 1.1), this leads to
redistribution rather than destruction of material, so continental
crust is never destroyed. Nor does it grow much these days,
except by the addition of small volumes resulting from volcanic
activity above subduction zones. I will postpone consideration
of the origin of continental crust until Chapter 13, except to
say that most is thought to go back over at least 2 billion years
(although much of it has been deformed many times since
then), and that some traces date back nearly 4 billion years. In
contrast, because it is recycled, the oldest known oceanic crust is
a mere 190 million years old.
Plates sliding past each other

Not all boundaries between plates are sites of convergence or
separation. Some mark places where plates are sliding past
each other. These are known as conservative plate boundaries.
Figure 4.8 shows a typical example within an ocean, where a
constructive plate boundary is offset by a fault, known as a
transform fault.
constructive
plate boundary
transform fault
fracture zone
10–100 km
(approx)
Figure 4.8 Map view showing how a transform fault offsets a constructive
plate boundary.

A transform fault may offset the ridge by anything from ten to
a couple of hundred kilometres. It is marked by a narrow zone
of shallow earthquakes. The transform fault is the bit between
the two offset ridge segments; its apparent continuation beyond
this lies within a single plate, and is not a plate boundary. This is
known as a ‘fracture zone’ and it is often manifested as a sharp
change in depth because of the different ages of ocean floor on
either side of it, but it has far fewer earthquakes because there is
no sideways slip between the rock on either side.
Conservative plate boundaries can run through continents, too.
The most celebrated of these, of which the San Andreas fault is
a part, runs through California. Here, the part of California on
the south-west side of the fault is attached to the Pacific Plate,
and is moving north-westwards at an average rate of 1 cm per
year relative to the rest of North America. Stick–slip motion
along this fault system is what causes most of the earthquakes
in the San Francisco–Los Angeles region.
Key idea: Plate boundaries

Plate boundaries can be destructive (where subduction occurs),
constructive (where plates draw apart and new ocean floor is
formed), or conservative (where plates slide past each other).
The global picture

You are now in a position to view the big picture of how the
plates and the boundaries between them fit together on a global
scale. This is shown in Figure 4.9. There are essentially two
settings in which destructive plate boundaries can occur. The
first is where one oceanic plate descends below another. When
this happens, volcanism on the overriding plate leads to the
construction of a series of volcanoes, which is the origin of the
volcanic island arcs in the western Pacific, Indonesia and the
Caribbean. On the other hand, when subduction occurs beneath
a continent, as where the Nazca Plate (south-east Pacific)
subducts beneath South America, the volcanoes grow on
64

pre-existing continental crust, giving rise to an Andean-type
volcanic mountain range. The case of Japan, with which we
began, is something in between. Part of Japan is old continental
material rifted away from the eastern edge of the Asian
continent, but much of it is young island arc material.
A potentially confusing feature of any map such as Figure 4.9
is that it attempts to portray on a flat sheet of paper something
that is really happening on the surface of a sphere. Plate
tectonics occurs on a globe, which means that rates of relative
plate motion must vary along a plate boundary. For example,
the constructive plate boundary between the Pacific Plate and
the Antarctic Plate is spreading at a rate of less than 6 cm a year
between Antarctica and New Zealand, but the rate gradually
doubles as the boundary is traced over the next 1000 km
north-eastwards. The relationships between individual plates
are continually adjusting to prevent any gaps opening between
them. This effect is particularly noticeable in trying to trace
through time the positions of triple junctions where three plate
boundaries meet.
North American Plate 21

Eurasian Plate
10
10
29
106 79 Caribbean
11 Plate 6
26 Philippine
Arabian 50 36 Plate Cocos
Plate Plate
Pacific 57
Plate
African South
46 American
Plate Indian–Australian
Plate Nazca Plate
Plate
102 8
84 32
8 73
10 Antarctic 12
Plate 12
Key: Plate Margins destructive conservative constructive

10 direction of motion (speed in mm per year)
Figure 4.9 Global map (on the same base as Figure 4.1) showing plate
boundaries and the rates of plate motion relative to Africa (which is
virtually stationary). Recent continental collision zones are shown here as
destructive plate boundaries.

Spotlight: Fudging the lines
Distortions in the map projection in the far north led into a bit of a
‘fudge’ in portraying the plate boundaries in Figure 4.9, which shows
no boundary between the Eurasian and North American Plates. These
are distinct plates, but in fact eastern Siberia (and, according to some
experts, northern Japan, too) is part of the North American Plate.
The missing plate boundary is an ill-defined and diffuse divergent
feature running north from somewhere near Japan. Beyond the map,
it passes close to the North Pole and eventually becomes the Mid-
Atlantic ridge, reappearing near the map’s north-eastern corner.
So far I have kept the discussion simple, by assuming that the

relative motions between plates are parallel. Although this is
commonly a good approximation, especially for constructive
plate boundaries, it does not always hold. There are plenty of
examples where collision between continents occurs obliquely.
Indeed, much of the western coast of North America from
California to Alaska is thought to consist of pieces of continent
described as ‘exotic terranes’, which slid in obliquely from the
south at various times during the past 200 million years. To
take another example, conservative plate boundaries within the
continents often have kinks in them. In the example shown in
Figure 4.10, the relative motion (as seen looking from one side
of the fault to the other) is from left to right. Overall, the fault
runs parallel with the direction of relative motion, but where
the fault line bends to the left there is a zone of compression
(manifested by buckling of rocks and thrusting of one unit
over another) and where it bends to the right there is a zone of
compression
extension
10 km (approx) fault
Figure 4.10 Map view to show zones of local compression or extension

where a conservative plate boundary bends one way or another.
66

extension (where the crust is stretched thin). The San Andreas
Fault system is like this, with both sorts of bend.
What makes the plates move?

In all this discussion of plate tectonics, I have avoided mention
of what drives it. This is still controversial, but most geologists
would agree on two things. The first is that, seductive as the idea
might seem, plate motions are not the direct surface expressions
of mantle convection (of the kind illustrated in Figure 2.9). The
plates seem to be moving around more or less independently of
this deeper process. The second is that, whatever the immediate
driving mechanism, the creation of new, hot oceanic lithosphere at
constructive plate boundaries and the subduction and recycling of
old, cold oceanic lithosphere at destructive plate boundaries is the
principal means by which the heat generated by radioactive decay
in the mantle escapes to the surface. Probably about twice as
much heat gets out this way as trickles out through the lithosphere
by conduction, and each of these considerably outweighs the
outward heat transfer associated with volcanic eruptions.
Spotlight: Sister planet?

One of the most perplexing contrasts between the Earth and
Venus, which is almost the same size, mass and density as the
Earth, is that Venus seems to lack plate tectonics. The answer may
be as simple as a deficiency of water inside Venus, making the top
of its asthenosphere too stiff to allow plate motion.
There are many forces that could be causing plates to move. As

already noted, simple conveyor-belt-type drag by flow in the
underlying asthenosphere is not considered likely. Nor does it
seem that plates are pushed apart by forcible injection of new
material along constructive plate boundaries. It is probably
closer to the truth to regard the upwelling here as a consequence
rather than a cause of plate divergence. Perhaps the most likely
driving mechanism is that the old, cold edge of a subducting
slab, sinking because it is negatively buoyant, drags the rest of
the plate with it.

Key idea: Heat transfer and plate tectonics
The mechanism that drives plate tectonics is unknown, but plate
recycling is the principal way in which heat is transferred from the
mantle to the surface.
I have mentioned volcanoes several times so far. In the next

chapter, I will consider them more fully.
Dig deeper
There is a lot of background information about plate tectonics at
http://www.geolsoc.org.uk/Plate-Tectonics
The Smithsonian Institution’s ‘This Dynamic Planet’ map relates
to volcanoes, earthquakes and plate tectonics, and allows you
to explore plate boundaries in detail http://www.volcano.si.edu/
tdpmap/
The Open University has some free plate tectonics material here:
http://www.open.edu/openlearn/science-maths-technology/
science/geology/plate-tectonics/content-section-0
and a plate tectonics game here: http://www.open.edu/
openlearn/science-maths-technology/slip-slide-collide
For an amusing but informative animation about subduction
and volcanoes with good accompanying text, see http://www.
oxfordsparks.net/video/underwater-volcano-disaster
There are many other plate tectonic resources at http://serc.
carleton.edu/NAGTWorkshops/geophysics/visualizations/
PTMovements.html
A landmark paper that established the ‘rules’ for plate tectonics
on a sphere is D.P. McKenzie and R.L. Parker, ‘The North Pacific:
An Example of Tectonics on a Sphere’, Nature, 216:5122 (1967):
1276–80.
68

Fact-check
1 Which is the Earth’s largest tectonic plate?
a Antarctic Plate
b Eurasian Plate
c North American Plate
d Pacific Plate
2 Why is it not generally possible for continental crust to be

subducted?
a Because it is too young
b Because it is too buoyant
c Because it is too rugged
d Because plate motion never takes continental crust to a
subduction zone
3 Which statement about plate tectonics is incorrect?
a It can occur on the Earth because the top part of the
asthenosphere is so weak
b Plate tectonics is the means by which continents plough
through the mantle
c It is largely independent of deep mantle convection
d Where plates are drawn apart in the ocean, there is usually a
ridge
4 How fast is the Pacific Plate moving relative to the African

Plate?
a It is moving east at about 10 cm per year
b It is moving northwest at about 10 cm per year
c It is moving south at about 5 cm per year
d It is moving north at about 3 cm per year
5 Which of these is most likely to be the main driving mechanism

for plate tectonics?
a Convection currents in the sub-lithospheric mantle
b Pushing apart by injection of magma at constructive plate
boundaries
c Dragging apart of plates because of subduction
d Magnetic pull caused by the outer core

6 How deep can the inclined pattern of foci defining a subduction
zone be traced?
a About 70 km
b About 250 km
c About 700 km
d About 1000 km
70

5
Volcanoes
33about the different kinds of volcanoes
characteristic of various geological
settings
33what makes volcanoes erupt, and what
hazards are posed by different kinds of
eruption.
71

You saw in the previous chapter that the world’s most obvious
volcanoes are above subduction zones in Andean-type mountain
chains or island arcs. The global distribution of volcanoes above
sea level (Figure 5.1) correlates well with destructive plate
boundaries (Figure 4.9). There is rather more volcanic activity
(though of a different kind and harder to detect because it is
usually several kilometres underwater) occurring continually
along constructive plate boundaries. In contrast, conservative
plate boundaries rarely show associated volcanism. In this
chapter I will consider the types of volcanic activity that happen
at plate boundary settings, and also look at volcanism that
occurs well away from the edges of plates.
Figure 5.1 Global map of volcanoes above sea level, showing volcanoes
known to have erupted in the past 10 000 years.
Mention the word ‘volcano’ and most people probably think of

something resembling the volcano in Figure 5.2; a fairly steep
conical mountain like Mt Fuji in Japan. This sort of volcano,
known as a composite cone volcano, is typical of settings above
subduction zones, though they are not all so symmetrical.
Spotlight: Volcano terminology

No one has been able to impose a set of standard terms for types
of volcano. What I refer to here as ‘a composite cone volcano’
would elsewhere be described as ‘a stratocone volcano’ or just ‘a
stratocone’.
72

Figure 5.2 A composite cone volcano, Acamarachi in northern Chile. The
date of its most recent eruption is not known, but is believed to have been
within the past 10 000 years.
Magma origin and eruption

Because mechanisms to generate magma occur there, volcanic
eruptions are common at constructive plate boundaries and
near destructive plate boundaries (above subduction zones).
Conversely, volcanoes are rare at conservative plate boundaries.
In subduction zone volcanism, much of the material that is being
melted at depth belongs to the crust of the descending slab. With
the exception of any sediments that have accumulated on top of
it, this will be basaltic in composition, implying a silica content
of about 49 per cent (Table 2.2). The melt derived from this must
be richer in silica (typically about 55–60 per cent) as a result
of partial melting (a process described in the previous chapter).
Melts of this intermediate silica content produced by partial
melting of basalt are called andesites (named after the Andes
mountains, where they are common).
As it begins to rise, this intermediate magma will often mingle
with batches of basaltic magma generated by partial melting
5 Volcanoes 73

in the wedge-shaped zone of mantle between the top of the
subducting plate and the base of the overriding plate. Melting is
stimulated there as a result of water having been squeezed out
of the wet crust of the subducting plate. This water rises into
the overlying mantle wedge, where it has the effect of lowering
its melting temperature by about a hundred degrees. This is
described as hydration melting.
Remember this!
Like decompression melting, hydration melting is a way of creating
magma without actually heating the source material. It is partial
melting that occurs as a result of adding water, typically above a
subduction zone.
Magma is usually slightly less dense than solid rock, so, once
formed, it tends to rise towards the surface. Basaltic magma will
be at about 1150 °C, whereas the temperature can be 400 degrees
lower for the most silica-rich magmas. At first, magma percolates
along the interfaces between crystals, but eventually collects into
larger blobs that can force their way upwards. Some magma
bodies cool so much that they solidify before reaching the surface,
forming intrusions (described in the next chapter). For now, we
are concerned only with what happens if the magma reaches the
surface (Plate 2). When it does, one of two things may happen. It
may ooze out and begin to flow downhill. This is described as a
‘lava flow’ (lava being the term used to describe magma once it has
reached the surface). Alternatively, it may break apart, sometimes
explosively, to produce pyroclastic (fire-broken) rock.
Key idea: Lava

Lava is just magma at the surface.
Magmatic gases and explosive

eruptions
The idea of magma exploding may seem surprising. It is,
however, a very common occurrence (Figure 5.3, Plate 3). What
74

makes magma explode is the sudden escape of gases that had
been dissolved in it. If these gases escape rapidly (like a can
of fizzy drink opened after it has been shaken), bubbles form
that expand the magma into a froth and may break it into
fragments. Lumps of solidified bubbly rock are referred to as
pumice. The small fragments are referred to as volcanic ash.
Figure 5.3 An eruption from the south-east crater of Mt Etna, Sicily. Gases
are escaping from magma within the vent, with force sufficient to drive
this mild pyroclastic eruption. Fine ash fragments are thrown to a height
of about 50 m above the rim of the crater, and then blown downwind in a
billowing cloud, sometimes called an ‘eruption plume’.
Spotlight: A burning issue

Beware! The term ‘ash’, although correctly describing the fine-
grained nature and typical greyish colour of volcanic ash, is
potentially misleading. Unlike the ashes of a fire, volcanic ash is
not a product of combustion. Terms such as ‘fire fountain’ (a kind
of eruption), ‘ring of fire’ (a colloquial description of the volcanoes
that surround the Pacific ocean), ‘smoke’ (to refer to ash or clouds
emitted by a volcano) and even the ‘pyro’ in ‘pyroclastic’ are nice
graphic descriptions but are flawed as metaphors.
5 Volcanoes 75

Magma rising above subduction zones typically contains a small
percentage of gas dissolved within it. Typical gas composition is
about 56 per cent H2O (water vapour), 28 per cent CO2 (carbon
dioxide), 14 per cent SO2 (sulfur dioxide) and 1 per cent or less
of each of H2S (hydrogen sulfide), HCl (hydrogen chloride), H2
(hydrogen) and CO (carbon monoxide). Above a subduction
zone, most of the dissolved water vapour can be explained as
seawater escaping, after having been transported down with the
descending slab (in fissures and hydrated minerals in the oceanic
crust or within wet sediments), and some of the carbon dioxide
probably comes from breakdown of subducted carbonates
(which are common in certain kinds of sedimentary rock, most
notably in limestone). However, magmas in other settings
also contain dissolved gas (though usually less) and it seems
that some of the gas is escaping from within the mantle. This
represents the tail end of the degassing process that is thought to
have been the source of most of the atmosphere (see Chapter 2).
Whether a volcanic eruption is explosive depends not on whether
the magma is shaken like beer in a can, but on how quickly the
magma rises and how easily the gases can escape. The great
pressure at depth keeps gases dissolved in the magma when it
is deep down, and usually bubbles begin to form only within a
couple of kilometres of the surface. Events during the final ascent
from this depth are crucial. If the magma rises slowly, then there
will usually be time for it to degas quietly, and on reaching the
surface the magma will form a lava flow. If it rises quickly, or
if the walls of the conduit are impermeable to gas, the gas will
remain trapped within the magma and it will erupt explosively.
Key idea: Explosive eruptions

Violent gas escape from magma when it rises to shallow depths is
what drives explosive eruptions.
In many volcanoes, gas escapes quietly over periods of tens or

hundreds of years without any associated eruption. The fissures
where the gas reaches the surface are known as fumaroles,
and may be at temperatures of up to several hundred degrees
centigrade (Figure 5.4).
76

Figure 5.4 Volcanologists studying fumaroles on the summit of Vulcano,
a volcanic island to the north of Sicily that gave its name to volcanoes
in general. On mixing with air, the volcanic water vapour forms a visible
steam plume. The volcanologists are wearing gas masks to protect against
additional acid gases, notably sulfur dioxide, in case the wind shifts.
Eruption conditions in andesite composite cone volcanoes tend

to fluctuate between quietly effusive and explosive. It is the
resulting alternation of lava flows and more widely dispersed
ash layers that is chiefly responsible for the relatively steep
conical shape of this kind of volcano (Figure 5.5).
crater vent
lava
ash conduit
Figure 5.5 A cross-section through a composite cone volcano, showing

the alternation of ash and lava responsible for both its name and its shape.
There is only a single vent shown here, but in many cases the conduit
branches at depth, feeding secondary (‘parasitic’) vents on the flanks of
the volcano.
5 Volcanoes 77

Volcanic hazards
Lava flows of andesite composition are fairly viscous. They tend
to have very blocky surfaces, and move more like a train of
rubble than a stream of liquid. As a result, they are rarely longer
than about 10 km. Although they may damage property, people
can usually escape, because such flows normally advance at only
a few metres per hour.
When andesitic volcanoes erupt explosively, the effects are
much more widespread. One of the biggest eruptions of the last
century occurred in June 1991 at Mt Pinatubo in the Philippines.
This volcano had not erupted for at least 500 years. As is often
the case, when it awoke, it erupted much more dramatically than
a similar volcano in the habit of erupting every few years, and a
column of ash reached a height of about 60 km over the volcano.
This type of eruption is described as plinian (Figure 5.6).
Figure 5.6 A plinian eruption column rising about 20 km above Lascar

volcano, in the Andes.
The term ‘plinian’ comes from Pliny the Younger, a Roman

magistrate and author, who as a youth wrote the earliest known
stage-by-stage account of a volcanic eruption. He witnessed
an eruption of Vesuvius in 79 ad, which culminated in the
78

destruction of the towns of Pompeii and Herculaneum. His
uncle, Pliny the Elder, was in charge of the local fleet, and was
overcome by volcanic gas when he went ashore.
During a plinian eruption, the ash is driven upwards from
the vent by the force of the explosive escape of gas. However,
columns of ash would not attain the great heights that they
do without the aid of buoyancy. What happens is that air
is drawn into the column, where it is heated by the hot
ash particles, and by mixing with the hot volcanic gases.
The heated air expands and, despite the weight of the ash
particles, the column as a whole becomes buoyant and rises,
until it reaches a height where its density matches that of the
surrounding air. The eruption cloud is then blown downwind,
and the ash begins to settle out to form ash ‘fallout’ deposit.
The larger particles fall faster, so fallout deposits tend to be
thicker and made of larger particles closer to the volcano
(Figure 5.7). The largest bits, ranging up to metres in size, are
known as volcanic bombs.
Figure 5.7 This volcanic ash (seen in cross-section on the island of

Vulcano, Italy) represents several eruptive pulses. In each case the larger
particles fell first, so that each layer becomes finer upwards.
5 Volcanoes 79

Spotlight: Volcanic bombs
If, like me, you ever fly home with volcanic bombs in your hand-
baggage, be very careful not to use the b-word to describe them if
security staff ask what you are carrying. They are ‘rock samples’.
By mapping out the thickness of ancient fallout deposits, it

is possible to determine the wind direction at the time of the
eruption, and also to estimate the intensity and volume of the
eruption. Where it accumulates thickly, the weight of fallout
ash can cause roofs to collapse and the fine particles can cause
choking, but it is rarely a major cause of death.
Sometimes, an eruption column becomes unstable and all or part
of it collapses on to the side of the volcano, and then sweeps
downslope as a cloud of searing-hot ash and gas travelling at
over 100 km per hour. This sort of pyroclastic flow is known
as a nuée ardente (French for ‘glowing cloud’) and it can also
be triggered by the collapse of a steep dome of extruded lava.
A nuée ardente is justifiably the most feared consequence
of a plinian eruption. One reason for this is the destruction
of the town of St-Pierre, the capital of the Caribbean island
of Martinique, and the death of all but two of its 29 000
inhabitants, in the 1902 eruption of Mt Pelée. One survivor was
a prisoner in a windowless jail cell, partly below ground level.
He was dug out badly burned two days later, and pardoned of
his crimes.
Fears of a similar eruption on the nearby island of Montserrat
led to the temporary evacuation of its most vulnerable areas in
1995, followed by their permanent evacuation in 1996 when
the island’s capital was abandoned (Figure 5.8). In the event, the
Montserrat dome collapse flows were smaller but more persistent
than on Martinique. Many people left the island, and with the
eruption still continuing, by 2006 the island’s population was less
than half what it had been ten years previously. Since then, more
people have drifted away for economic reasons.
Unlike fallout deposits, which blanket the topography
irrespective of slope, pyroclastic flows tend to be confined to
80

Figure 5.8 Progressive burial of Plymouth, the capital of Montserrat,
mainly by pyroclastic flows. Top: April 1997. Lower left: July 1998. Lower
right: May 1999.
valleys, a characteristic that enables the two types of deposit

to be distinguished. However, not all fallout material remains
in place. In particular, it is easily washed away by rainfall. In
tropical areas rain is liable to be torrential, so the watercourses
fill with a slurry of ash and water that, being denser than
water, moves with a great force. These volcanic mudflows
(often called lahars, an Indonesian word) caused most of the
damage resulting from the 1991 eruption of Mt Pinatubo.
Lahars can also be initiated if a volcanic eruption melts a glacier
5 Volcanoes 81

or snowcap. This is what happened during an eruption of
Ruiz, Colombia, in 1985, when approximately 23 000 people
(including most of the population of the town of Armero)
lost their lives. Readers who are old enough may remember
harrowing television pictures of 13-year-old Omayra Sanchez
(her lower body trapped by debris in the mudflow, and with
only her head showing above water) who succumbed after three
days’ attempts to dig her out.
Most volcanologists who grew up in the 1980s will tell you that
televised scenes from Armero were instrumental in setting them
on their career path.
The principal hazards associated with explosive eruptions of
andesitic volcanoes are summarized in Figure 5.9. We may also
add the risk of volcanic collapse, a process that first became
widely recognized after the May 1980 eruption of Mt St Helens
in Washington State, USA.
eruption plume
wind
eruption
column
ash fallout from plume
mud flows
caused by ash flows
heavy rain
on loose ash
earthquakes
Figure 5.9 Some of the hazards associated with explosive eruptions. The
most dangerous ash flows are nuées ardentes.
82

Case study: Mt St Helens, 1980
Here, two months’ warning of a major eruption was given by small
but progressively shallower earthquakes, minor ash eruptions
and a slight bulging of the northern flank of the volcano. However,
rather than culminating in an anticipated plinian eruption from the
summit crater, the northern flank gave way, forming a giant debris
avalanche. As soon as this collapse began, the confining pressure
on the magma within the volcano was released and it degassed
explosively, directing most of its force sideways. This directed
blast was supersonic, and 60 sq. km of what had been forest was
devastated (Figure 5.10).
Figure 5.10 Mt St Helens, three years after its 1981 eruption. The whole
summit was undermined and destroyed by the collapse and directed
blast. The steam within the collapse amphitheatre comes from a slow-
growing lava dome that may eventually restore the volcano to its previous
symmetrical cone shape.
The sideways nature of this directed blast caught even the

professionals by surprise. A volcanologist working for the US
Geological Survey, Dave Johnson, was stationed in what should have
been a safe location, on a high ridge several kilometres north of Mt
St Helens, having been assigned the duty of recording the eruption.
Tragically, his campsite caught the full force of the directed blast, and
5 Volcanoes 83

his body was never found. His young field assistant, Harry Glicken,
had been sent back to headquarters, and survived to become a well-
known volcanologist in his own right. Sadly Harry died 11 years later,
when he was caught by a nuée ardente generated by a collapsing
lava dome on the Japanese volcano Unzen.
The Mt St Helens collapse was triggered by the injection of

magma into the volcano, but it is possible that other volcanoes
collapse simply under their own weight. When this happens,
debris avalanches can travel for tens of kilometres, to judge
from mapping of ancient deposits. No major volcanic debris
avalanche has happened in historic times.
Other causes of death associated with volcanoes include the
release of suffocating gases, which may flow downhill, and
tsunamis generated not by earthquakes but by underwater volcanic
explosions, or by a volcanic landslide entering the sea or a lake.
Table 5.1 lists the fatalities from some notable volcanic events.
Not all volcanoes above destructive plate boundaries are
andesitic composite cones. In some places, basaltic melts
approach the surface, generated either by more complete
melting of the oceanic crust of the subducting slab, or because
of partial melting within the wedge-shaped volume of mantle
between the two plates. This basalt may flow out as quiet lava
flows, its lower viscosity allowing it to spread more thinly than
andesite over many square kilometres or (if the eruption is
driven by the force of escaping gases) it may result in a ‘scoria
cone’, formed of lumps of slightly frothy basalt (Figure 5.11).
On the other hand, very large volumes of more silica-rich
magmas are sometimes generated above subduction zones. This
is particularly likely where large proportions of sediment have
been dragged down with the subducting slab, or when there is
a lot of melting of the lower crust above the subduction zone.
These magmas typically have around 70 per cent silica, and are
described as acidic or felsic or granitic. These usually crystallize
at depth, forming the well-known rock type called granite
(discussed in the next chapter), but sometimes they do approach
the surface.
84

Table 5.1 Some notable volcanic events.
Volcano Year Fatalities Main cause(s) of death
Vesuvius, Italy 79 CE >3500 pyroclastic flows
Kelut, Indonesia 1586 10 000? unknown
Asama, Japan 1598 800 religious pilgrims killed at summit
Vesuvius, Italy 1631 >4000 pyroclastic flows
Merapi, Indonesia 1672 3000? pyroclastic flows
Laki, Iceland 1783 9350 starvation
Asama, Japan 1783 1500 pyroclastic flows, lahars
Unzen, Japan 1792 14 300 tsunami
Tambora, Indonesia 1815 92 000 starvation
Krakatau, Indonesia 1883 36 400 tsunami
Mt Pelée, Martinique 1902 29 000 pyroclastic flows
Taal, Philippines 1911 >1335 pyroclastic flows
Merapi, Indonesia 1930 1369 pyroclastic flows
Ruapehu, New Zealand 1953 151 lahar
Iliwerung, Indonesia 1979 539? tsunami
Mt St Helens, USA 1980 57 directed blast, lahars
Mayon, Philippines 1981 >200 lahar
El Chichon, Mexico 1982 1900 pyroclastic flows
Ruiz, Colombia 1985 23 000 lahar
Lake Nyos, Cameroon 1986 >1700 asphyxiation by gases
Pinatubo, Philippines 1991 800 fallout, lahars, disease
Soufrière Hills, 1997 19 pyroclastic flows
Montserrat
Casita, Nicaragua 1998 1600 lahar
Nyiragongo, Congo 2002 147 lava flows, asphyxiation
Merapi, Indonesia 2010 353 pyroclastic flows
Ontakesan, Japan 2014 56 airfall and asphyxiation
Key idea: Deadly hazards

The most deadly volcanic hazards are pyroclastic flows and lahars.
Lava flows and fallout are usually less deadly. Volcanic eruptions
can also cause tsunamis.
It is rare for granite magmas to ooze out quietly. This is because

granitic lava is even more viscous than andesitic lava, making it
5 Volcanoes 85

Figure 5.11 A basaltic scoria cone, in the Andes of northern Chile.
difficult for gases to escape quietly. Sometimes it does happen,

though, and a very thick flow or a steep-sided lava dome is
formed (Figure 5.12).
Ignimbrite eruptions
When granitic magma approaches within a few hundred
metres of the surface, the gas pressure usually fractures the
roof over the magma body, thereby releasing pressure and
encouraging whatever gas remains in solution to come out
all at once. This drives the most energetic variety of explosive
eruption known, which occurs at volcanoes colloquially
known as ‘supervolcanoes’. Such eruptions are known from
the study of ancient deposits rather than from observation,
because (fortunately for us) they are infrequent and no
example has occurred in recent times. To judge from the
fallout deposits that you can find preserved, a high plinian
eruption column develops, sometimes feeding a continent-wide
ash cloud.
The most characteristic deposit is probably a result of column
collapse. This takes the form of a very extensive sheet of acidic
86

Figure 5.12 An image of a 40-km-wide area in the Andes, recorded by
satellite. On the left is a 24-cubic-km acidic flow, whose front is 300 m
high. Its thickness and the large wrinkles on its surface indicate that it was
an extremely slow-moving viscous flow. On the right is an acidic dome,
only one-tenth the volume, which appears to have oozed out radially from
a hidden vent below its centre.
ash, in which the particles are often welded together because

they were still very hot when emplaced. This sort of deposit is
known as an ignimbrite. There are good recent examples in the
south-west USA (Figure 5.13), New Zealand and the Andes, and
400-million-year-old examples in Snowdonia (Wales) and the
Lake District of England.
Large ignimbrites are the result of the eruption of up to
around 100 cubic km of magma. This tends to leave a large
hole in the ground, rather than an obvious volcanic mountain.
Usually, what had been the roof over the magma body
collapses to form the floor of a large crater, termed a caldera
(Figure 5.14).
5 Volcanoes 87

Figure 5.13 The high cliffs in the background of this view expose a thick
section through an ignimbrite erupted from the Valles caldera, New
Mexico, USA. The lower cliff in the foreground is a plinian fallout deposit.
Note the person for scale in the centre.
Remember this!
Technically, a caldera is any crater bigger than 1 km in diameter.
I have described volcanism near destructive plate boundaries in

broad outline. There is not space to explore the many details of
eruptive processes in a book of this length. Now for volcanism
at constructive plate boundaries.
caldera
magma
body 10 km
approx
(a) (b)
Figure 5.14 Cross-section to show caldera formation. Between these two
stages, a large-volume ignimbrite-forming eruption occurs.
Volcanoes associated with

continental rifting
First, what happens when a continental plate is in the process
of splitting to initiate a new constructive plate boundary?
88

Figure 4.6 showed this happening above a zone of upwelling
in the asthenosphere. That is not always the case, but when it
does happen that way, the faults that allow the crust to stretch
are actually rather good pathways along which magma can
rise. Continental rifting associated with mantle upwelling is
often preceded by an approximately 10-million-year period of
eruption of mainly basaltic lavas (produced by partial melting
of the upwelling mantle asthenosphere), which reach the surface
even before much stretching of the crust has happened. Major
stretching and associated volcanism is going on today across
the African Rift, a large fault-bounded structure that runs from
Eritrea and Ethiopia southwards through Kenya and Tanzania
to Mozambique.
Remember this!
Africa appears to be splitting apart by east–west extension across
the African Rift (hence most of the African earthquakes shown in
Figure 4.1), although there is no guarantee that this will progress
far enough to generate a new ocean.
Crustal extension across the African Rift provides the upward

pathways for the magma feeding the large African volcanoes
like Kilimanjaro. Volcanism is at its most intense near the
northern end of the African Rift, corresponding to a hot spot
generally attributed to pipe-like upwelling (called a ‘plume’)
rising within the mantle asthenosphere, below the point where
the rift joins the Red Sea and the Gulf of Aden. Here the
eruptions have taken the form of vast sheets of basaltic lava
flows, known as continental flood basalts. Whether hot spots in
general are supplied by deep-seated mantle plumes (originating
as deep as the core-mantle boundary) became a matter of
controversy a few years ago. Probably some (maybe most) are,
but some (possibly only a few) are not.
There are many ancient examples of flood basalts that appear to
represent hot spots prior to continental rifting. One of the largest
and most famous is the 65-million-year-old Deccan Traps, in
India. These are a kilometre thick. Although reduced by erosion,
5 Volcanoes 89

they still cover an area of half a million square kilometres, and
are a result of hot spot volcanism fed by a mantle plume that
helped India to rift away from Africa and Madagascar. (Although
much less active today, the hot spot can still be recognized, being
marked by the site of the volcanically active island of Reunion.)
Other continental flood basalt provinces reflect the rifting of
the Atlantic Ocean, and can be found split into unequal parts
on either side. These include: the 125-million-year-old Parana
basalts of southern Brazil that match with their less extensive
counterpart (the Etendeka traps) in Nambia; and the smaller
but more famous 60-million-year-old Brito-Arctic province,
most of which is in east Greenland, but which includes the thick
lava flows of the Giant’s Causeway, in Northern Ireland, and the
Inner Hebrides islands, Scotland (Figures 5.15 and 5.16).
Figure 5.15 The island of Staffa, Scotland, formed of a 60-m-thick

basalt flow, the middle part of which (from present sea level up to the
roof of the cave) fractured into columns as a result of contraction during
comparatively slow and orderly cooling. The upper part is more rubbly. The
cave is Fingal’s cave, made famous by Mendelssohn in his overture The
Hebrides.
The 15- to 17-million-year-old Columbia River Basalts of

Washington, Oregon and Idaho constitute a continental
flood basalt province above a hot spot that has not been
90

Figure 5.16 Top view of columnar joints formed by slow cooling within the
flood basalt flow on Staffa, Scotland. Most columns are hexagonal, though
five-side examples can also be seen.
followed by a major rifting event. One flow field amounting

to 1300 cubic km of basalt travelled 300 km while confined to
the Columbia River Gorge, before filling it and spreading out
laterally to a width of over 100 km. The time it took to achieve
this is estimated as between 5 and 15 years.
Volcanism at ocean ridges

Once true ocean crust has begun to form, by sea-floor
spreading, the character of the volcanism changes. The basaltic
lava that erupts at ocean ridges to form the top layer of the
oceanic crust can take on a form completely unlike that of lava
on land. When lava is extruded under water, its surface chills
very rapidly, forming a flexible rind around the still-molten
interior. The lava typically resembles a pile of water-filled plastic
bin-liners, a morphology known as pillow lava (Figure 5.17,
Plate 4).
5 Volcanoes 91

Figure 5.17 Pillow lava, erupted on the ocean floor 95 million years ago at
a constructive plate boundary. This is now exposed on land, as part of an
ophiolite in the Oman mountains, Arabia. This is a rare example of a slice
of oceanic lithosphere that, instead of being subducted, was thrust on top
of continental crust during a collision.
Remember this!
Pillow lava does not have to form in the ocean. Basalt can also
chill to form pillow shapes when erupted into a lake or underneath
an ice cap.
Where lava is erupted fast enough under water at a constructive

plate margin, it may form sheet flows. Alternatively, it may
fragment to form glassy debris known as hyaloclastite.
However, explosions, as opposed to fragmentation, are virtually
92

unknown because the pressure exerted by the overlying depth of
water inhibits the violent escape of gases.
One special characteristic of volcanism at oceanic constructive
plate margins is the widespread occurrence of ‘hydrothermal’
(hot water) vents where superheated water emerges. This
happens because water that has percolated down through
the crust over a wide area is warmed and then expelled, by
convection, through vents along the axis of the ridge. On the
way, it dissolves some elements from the rock, and certain other
elements already dissolved in seawater come out of solution by
reacting with the rock.
This exchange of chemical elements between water and rock
at hydrothermal vents is an important part of the rock cycle.
It also regulates the salinity (saltiness) of seawater, which has
remained broadly constant for several hundred million years.
Hot water emerges from ocean floor vents as an acidic solution,
but as soon as this mixes with seawater some of its dissolved
constituents become insoluble, and form a smoke-like cloud of
metal sulfide particles. The most famous examples are vents at
about 350°C where the particles are black, hence their popular
name of ‘black smokers’. These hot vents are surrounded by
organisms like clams, crabs and ‘tube-worms’ that ultimately
depend not on plants that derive their energy from sunlight,
but on bacteria-like microbes that, in turn, feed by oxidizing
the sulfide particles. It has been suggested that life on Earth
(and elsewhere) could have originated in such a setting, and
only later spread to sunlit regions where microbial forebears of
plants gained the ability to photosynthesize (see Chapter 12).
Key idea: Hydrothermal vents

Hydrothermal vents at oceanic constructive plate boundaries
are important sites for exchange of chemicals between rock and
seawater, and help to stabilize the seas’ salinity.
I noted in the previous chapter that the partial melting at a

constructive plate margin is caused by the decompression
associated with the upwelling of asthenosphere at a rate
5 Volcanoes 93

necessary to replace the lithosphere drawn away by subduction.
However, when a portion of constructive plate boundary is
situated at a hot spot (which may or may not represent a plume
coming up from deep within the mantle), the volume of melt
produced is much greater than usual. As a result, the oceanic
crust may be two or three times thicker than normal and its
surface will be above sea level. The only really clear example of
this today is Iceland.
Hawaiian volcanism
I turn now to volcanoes that are unrelated to plate boundaries,
and consider the example of Hawaii. A plume from deep
within the mantle is involved here, but the plume hits the base
of the lithosphere more than 1000 km from the edge of the
Pacific Plate. This plate is moving north-westwards over the
plume, and as a result there is a chain of increasingly older,
extinct volcanoes stretching north-westwards from the Big
Island of Hawaii where the current activity is centred. With
time, the weight of these older volcanoes causes them to
subside, and they also suffer collapses and are worn down by
erosion, so only the youngest few remain above sea level. The
rest of the chain survives only as undersea peaks known as
seamounts.
During the most productive part of its lifetime, a Hawaiian
volcano erupts nothing but basalt. Basalt is the least viscous
common type of lava. There are two important consequences of
this for a volcano above sea level. First, any gas within the lava
can escape relatively easily, so eruptions are not usually highly
explosive. Second, the lava flows, being so runny, are able to
spread out thinly. They can also flow a long way before cooling
down, especially if the top of the flow freezes and then acts as
an insulating lid, allowing the molten lava to continue to flow
in a ‘lava tube’ within. In consequence, Hawaiian volcanoes are
characterized by gentle slopes of 2–3 degrees. Such an edifice
is known as a shield volcano, because in profile it resembles a
shield laid on the ground (Figure 5.18). When lava flows reach
the coast, their fronts become chilled by the seawater, so they
tend to spread sideways along the shore. This steepens the
94

shoreline to the extent that eventually it collapses, and much
of the submarine portion of the cone is made of collapsed lava
rubble rather than intact lava flows.
50 km
50 km
Figure 5.18 The profile of a Hawaiian-type shield volcano. Mauna loa

volcano on the Big Island of Hawaii counts as the highest volcano on Earth,
because it rises more than 9 km from its base on the floor of the Pacific
Ocean.
Key idea: Shield slopes

Basalt is the least viscous common silicate lava, so basaltic shield
volcanoes have gentle slopes.
Basalt lavas (like those erupted on Hawaii) take essentially two

forms. When moving relatively fast (because of a fast eruption rate
or a locally steep slope), the surface breaks into clinkery blocks.
This type is known by the Hawaiian term ‘a’a’ (pronounced
‘ah-ah’, and said by some to represent the cries of pain made by
someone trying to walk barefoot over such a jagged surface).
When moving more slowly, the surface forms a smooth, chilled
but pliable skin over the molten interior of the flow, and the
lava spreads as a series of smooth lobes described by another
Hawaiian word, pahoehoe (Plate 5). This does not fragment like
a’a, although drag by flowing lava beneath the skin can pleat the
surface over upon itself to form a pattern reminiscent of coiled
rope. Both a’a and pahoehoe are visible in Figure 5.19.
Spotlight: Pahoehoe
I was told by a native speaker of the Hawaiian language that the way
to pronounce pahoehoe is PA-hoy-hoy, although many geologists
say pa-HOEY-HOEY. Apparently, ‘hoe’ is a term describing the
pattern made by a stroke of a canoe paddle upon the water, and
pahoehoe translates as something like ‘flat swirly swirly’.
5 Volcanoes 95

Figure 5.19 Two kinds of basalt lava on Hawaii: clinkery a’a (left) and
ropey pahoehoe (right).
Key idea: Hot spot volcanism

Hot spot volcanism is unrelated to plate margins. The most
popular explanation is decompression melting at the top of a rising
‘mantle plume’, but this is a matter of active debate.
Lava hazards
Basaltic lava flows, even though they are the least viscous and
therefore fastest flowing of common lava types, usually advance
at less than 5 m per second. They rarely kill people because
there is usually plenty of time to get out of the way. Tragic
exceptions have occurred when pent-up basalt lava has escaped
rapidly from breeches in the flank of Nyiragongo volcano in
the Congo, taking an estimated 60–300 lives in 1977 and 147
in 2002. Normally, though, it is property and livelihoods rather
than lives that are threatened by basaltic volcanism. Lava flows
are virtually unstoppable, and barriers erected to try to pond
or divert them are easily overcome. One of the more notable
successes in lava flow diversion was in 1992 when the front
of a major lava flow on Etna, Sicily, had reached within a few
96

hundred metres of the town of Zafferana. The side of the flow
was deliberately breached by explosives partway along its
length. This caused the flow to spread out sideways high up on
the mountain and halted its downstream advance.
Predicting eruptions
About 60 volcanoes erupt each year. A tenth of these eruptions
finish within a day and some go on for years, but the average
duration is about seven weeks. How can you tell when an
eruption is due, and how can you predict what its effects will be?
The best way to anticipate an eruption’s effects is to study
previous eruptions. This is easy in the case of volcanoes that
erupt every few years, because there are likely to be eyewitness
reports, video recordings and so on that give you some idea
of what might happen again. However, it is the volcanoes
that erupt least frequently whose eruptions are usually the
most dangerous, because a greater volume of ash and/or lava
tends to come out in one go. In such cases, there may be no
historical records and the volcanologist must depend largely
on the traditional skills of the field geologist – specifically,
interpretation of past events based on detailed mapping and
logging of ancient deposits. In addition, computers can predict
the paths of ash, lava and mud flows on digital terrain maps.
Hazard zone maps based on these approaches are available for
many volcanoes.
Case study: Eruption precursors – reading

the signs
There are many possible precursor signs of an impending
eruption. If recognized, these can give days’ or even months’
warning. This makes volcanoes easier to predict than tectonic
earthquakes, in which, as you have seen, the first shock is usually
the biggest. Probably the single most useful eruption precursors
are special kinds of small earth tremors, readily distinguished
from tectonic earthquakes, whose depth typically decreases from
several kilometres to a few hundred metres during the build-up
to an eruption. These are associated with injection of magma
5 Volcanoes 97

into the volcano, or with migration of gas bubbles. In order to
pinpoint them, it is necessary to have an array of at least three
seismometers on or close to the volcano.
Seismometers are expensive instruments, and few volcanoes

have permanent arrays. At the time of the first historic eruption
of Pinatubo, Philippines, in 1991 the volcano had been little
studied and was not under surveillance until a small ash eruption
near the summit in March showed that it was waking up. An
array of portable seismometers was rushed into place, enabling
the locations of sub-volcanic tremors to be tracked as they
approached the surface. This gave enough notice for the most
vulnerable areas to be evacuated before the catastrophic eruption
in June. Most of the ensuing deaths were attributable to the
failure to sweep fallout ash off flat roofs (which collapsed under
the weight) and subsequent mudflows (lahars), rather than to lack
of eruption warning.
When magma finds its way into a volcano, the volcano swells in
size. In extreme cases, its contours may change by many metres.
This deformation is another useful precursor sign, and its pattern
may serve to indicate whether the eruption is likely to occur at the
summit or from a vent or fissure on one side. It can be studied by
ground-based surveying or by using radar data from satellites.
Other techniques include measuring tiny changes in gravity

that reflect injection or withdrawal of magma from a volcano,
measuring changes in the rate of gas emission or of gas
composition, and monitoring changes in surface temperature
using ground-based instruments or satellite-borne infrared
detectors.
Sadly, precursor signs are not always enough to save lives in

the case of small eruptions. On 28 September 2014, nearly 300
Japanese hikers were out and about on the flanks of Ontakesan
volcano, a popular weekend venue. Seismicity increased only
11 minutes before a gas-driven explosion that generated a
3-km-long pyroclastic flow that claimed 56 lives (Table 5.1). No
warning was given, and even if it had been there would not have
been time to evacuate the hikers.
98

Key idea: Eruption precursors
Volcanic eruptions are usually heralded by precursor signs, such
as an increase in shallow seismic activity, or ground deformation.
Widespread effects of eruptions

Volcanic eruptions don’t just affect people living locally.
Several intercontinental airliners have run into trouble by
unwittingly flying into volcanic ash clouds, whereupon their
engines have stalled. Luckily, on every occasion so far the
pilots have been able to restart the engines, though sometimes
only after a terrifying gliding descent most of the way to the
ground. Eruption cloud warning systems, based on images
from weather satellites, and specially designed aircraft-
mounted radar are now used to protect the most vulnerable
air traffic routes, notably the transpolar route between Europe
and Japan that passes over Alaska and the Kurils, and routes
to Australia via Indonesian airspace. In 2010 airspace over
much of Europe was closed for six days in April, and again
briefly in May, because of concerns about a high-altitude
volcanic ash cloud blown south-east from the Icelandic
volcano Eyjafjallajökull.
The biggest volcanic eruptions can have continental or even
global effects. The 1991 eruption of Pinatubo injected so much
fine ash and sulfur dioxide into the stratosphere that global
surface temperatures are estimated to have been reduced by
about half a degree centigrade throughout the following year.
This is because sunlight that would normally have warmed
the surface was blocked in the upper atmosphere. The largest
eruption in modern history was of Tambora, Indonesia, in 1815.
At least 50 cubic km of ash was erupted, which was responsible
for glorious red sunsets round the globe and a much greater
drop in temperature than after the Pinatubo eruption. The
following year, 1816, has been described as ‘the year without
a summer’ by northern hemisphere commentators. There were
widespread crop failures and probably more deaths than from
any other historic eruption.
5 Volcanoes 99

When you consider the inevitability of a similar eruption
somewhere in the world in the near future, not to mention an
even larger eruption of a supervolcano, you have to accept that
volcanoes could play a big role in short-term climate change.
The possibility of worldwide famine cannot be discounted.
Volcanic activity also has more benign effects. Magma at
shallow depths can heat the groundwater and cause it to gush
episodically through a vent in the form of a geyser, like the
archetypal Geysir in Iceland or Old Faithful in Yellowstone
Park, Wyoming. Volcanically heated water can also be piped
to provide free heating to nearby communities, as is done in
Iceland and parts of Japan. Heated water may also seep out
gently to form hot mud pools.
Spotlight: Volcanic mud pools

Bathing in hot mud pools is said to be therapeutic, though
personally I have found the clinging smell of rotten eggs (caused
by hydrogen sulfide) to be an annoyance that seems virtually
impossible to get rid of, no matter how long I shower for afterwards!
Hot springs are as much a manifestation of magma intruded at

depth as of magma erupting at the surface, and these intrusions
are described in the next chapter.
100

Dig deeper
There is a companion book in this series (also written by me),
which has much more information on the topics covered here in
Chapters 3 and 5: Volcanoes, Earthquakes and Tsunamis (2015).
Another that I can recommend is Eruptions That Shook the World
by Clive Oppenheimer (Cambridge University Press, 2011).
The best place for reports on volcanic activity is the website of
the Global Volcanism Program, http://www.volcano.si.edu/ This
has an excellent database and learning resources.
Online volcano resources aimed more at schoolchildren can be
found at Volcano World, http://volcano.oregonstate.edu/
Discussion of a 2014 eruption in Iceland and whether it could
lead to a disruptive ash cloud can be found here: http://www.
open.edu/openlearn/science-maths-technology/all-you-need-
know-about-icelands-volcanic-eruption
The Open University has some volcanic hazards material here:
science/geology/volcanic-hazards/content-section-0
5 Volcanoes 101

Fact-check
1 Which magma composition tends to produce the steepest
volcanoes?
a Basaltic
b Intermediate (andesitic)
c Felsic
d Acidic
2 Which of these is not a gas typically erupted at volcanoes?

a Carbon dioxide
b Oxygen
c Sulfur dioxide
d Water vapour
3 Which statement about magma is incorrect?

a Partial melting yields a magma that is richer in silica than
the starting material
b Partial melting to produce magma can be triggered by the
addition of water to a volume of rock
c Partial melting to produce magma can be triggered by
increasing the pressure on a volume of rock
d Magma is usually less dense than solid rock
4 Which types of plate boundary are typically associated with

volcanoes?
a Constructive only
b Destructive only
c Constructive and destructive
d Conservative, constructive and destructive
5 Which of these is incorrect?

a Hawaii’s volcanoes are associated with a constructive plate
boundary
b Iceland’s volcanoes are associated with a constructive plate
boundary and a hot spot
c Indonesia’s volcanoes are associated with a destructive plate
boundary
d Japan’s volcanoes are associated with a destructive plate
boundary
102

6 What do granitic calderas tend to erupt?
a Ignimbrite
b Lahars
c Pahoehoe
d Pillow lavas
7 What is a mantle wedge?

a A place where one tectonic plate is forcing its way between
two others
b A wedge-shaped zone of mantle between the top of the
subducting plate and the base of the overriding plate
c A kind of shield volcano
d A tool used by a geologist to collect a sample
5 Volcanoes 103

Igneous intrusions
6
33some basic information about the
mineral composition of rocks
33about crystallization of minerals as
magma cools slowly underground
33how this affects the properties of the
remaining magma and the kinds of rock
that this produces.
105

Rocks formed by cooling magma are called igneous rocks, from
ignis, the Latin word for fire. We were dealing with igneous
rocks of the volcanic kind throughout the previous chapter.
However, not all magma reaches the surface. Often it solidifies
at depth, to form an intrusive igneous rock.
Spotlight: Fire away

Here we go again. In reality, of course, there is no ‘fire’ involved
in the formation of these rocks, though they do require high
temperatures, close to 1000 °C.
Volcanic rocks cool so quickly that often the crystals within

them are too small for the unaided eye to see. However, an
intrusive rock cools more slowly, so there is time for the
crystals to grow to lengths of millimetres or centimetres before
the rock is completely solidified. Crystals as big as that are
easy to see when the rock becomes exposed on the surface as a
result of erosion.
In Chapter 1 I introduced the concept of crystals growing
during the cooling of molten rock. Now it is time to consider
the crystalline nature of igneous rocks more fully, before
looking at the sorts of intrusive bodies that igneous processes
can form.
Minerals
Because the crust and mantle contain so much silicon and
oxygen (Table 2.2), the common sorts of crystals that grow
in all except the most exotic kinds of magma are rich in those
two elements. The crystalline structure of rock-forming igneous
minerals is based on two- or three-dimensional arrays of silicon
and oxygen held together by molecular bonds. These minerals,
as well as the rocks they form, are commonly referred to as
silicates.
106

Remember this!
The terms mineral and crystal are sometimes used
interchangeably, in a casual fashion. Strictly, any substance with a
regularly repeating atomic structure can be described as a crystal.
A crystalline mineral has a specific atomic structure, which can in
some cases accommodate different mixtures of elements. Some,
non-igneous, minerals have a fixed recipe but lack any crystalline
structure.
The most familiar of the silicate minerals is pure silica. This

contains just silicon and oxygen, and has the formula SiO2.
In crystalline form this is usually the mineral quartz, though
at extremely high pressures the atoms are squeezed into more
tightly packed crystalline structures which are given different
mineral names. All other silicate minerals contain various
metallic elements in addition to silicon and oxygen. Quartz is
a robust, hard-wearing mineral, on whose significance I will
elaborate in Chapter 8. It is able to grow in magmas whose SiO2
content exceeds about 52 per cent. Quartz is abundant in rocks
of granitic composition and is usually present, though scarce, in
rocks of andesitic composition.
The term ‘silica’ is used with two different meanings in geology.
It can mean a mineral or substance described by the formula
SiO2, or else it can denote the SiO2 content of a rock or magma.
This is potentially confusing, so read on!
The silica (or SiO2) content quoted for a rock is based on
an overall chemical analysis of the elements making up the
rock. This should not be taken to imply that all or even any
of the silicon and oxygen occurs in the form of quartz or any
other pure silica mineral. On the other hand, when SiO2 is
given as the chemical formula of a mineral, it means that this
mineral contains Si and O only, in the ratio 1:2. This specific
ratio is dictated by the number of oxygen atoms with which
each silicon atom is able to form a bond, if no other element
is involved. The three-dimensional structure of such bonds
controls the properties of the crystal.

In magmas whose composition is basic (45–52 per cent silica)
or ultrabasic (< 45 per cent silica), metallic elements are so
abundant that the only silicate minerals contain a relatively
high proportion of metals, and no quartz crystals can grow. The
most extreme example is olivine, formula Mg2SiO4, in which
there are twice as many metal atoms (in this case, magnesium)
as silicon atoms. Actually, iron (Fe) atoms can fit in the same
place as magnesium atoms, so olivine can have any composition
ranging between Mg2SiO4 and Fe2SiO4, with Mg and Fe in any
proportion. The permitted variation in olivine’s formula is often
expressed by quoting it as (Fe,Mg)2SiO4. This degree of variability
in composition, which is achieved without changing the internal
structure of the crystal, is one of the attributes that distinguishes
most minerals from simple chemical compounds, where different
elements cannot substitute for each other. Usually in igneous
rocks, olivine is of a magnesium-rich variety, say Fe0.2Mg1.8SiO4.
Remember this!
What makes the so-called ‘solid-solution’ possible between Fe-
rich and Mg-rich end-members is that the Fe2+ and Mg2+ ions have
the same charge and are of similar size, and so can be substituted
for each other without changing the structure of the crystal.
Olivine most often occurs in association with pyroxene, another

metal-rich silicate, with the formula (Ca,Mg,Fe)2Si2O6. A rock
composed essentially of just olivine and pyroxene is described as
a peridotite, and these are the dominant minerals in mantle rocks.
In magmas of higher SiO2 content, the proportion of magnesium
tends to be lower, whereas aluminium (Al), calcium (Ca),
sodium (Na) and potassium (K) become more abundant.
One result is the appearance of minerals known as feldspars.
Plagioclase feldspar, the most abundant mineral in rocks of
andesitic composition (52–66 per cent SiO2), has a formula
ranging between NaAlSi3O8 and CaAl2Si2O8 (a continuous
range in composition is made possible by ‘double substitution’,
whereby Na and Si are replaced by Ca and Al in equal
proportions). Potassium feldspar (KAlSi3O8) occurs chiefly in
108

granitic rocks (>66 per cent SiO2), in association with quartz
and plagioclase feldspar.
The most important other silicate minerals are mica and
amphibole. These occur in granitic and andesitic rocks and can
form only if there was water present in the magma. They have
complicated formulae: muscovite mica is KAl2(AlSi3O10)(OH)2,
biotite mica is K(Mg,Fe)3AlSi3O10(OH)2, and amphibole is
(Na,K)Ca2(Mg,Fe,Al)5(Al,Si)8O22(OH)2.
Each of these types of mineral, and the varieties within each
type, can be distinguished by features such as colour, hardness,
cleavage (regular planes of weakness in the crystal resulting
from its atomic structure) and optical properties.
A few characteristics of the main minerals are summarized in
Appendix 1. With practice, it is possible to recognize many of
these when you find a coarse-grained igneous rock exposed
in the field, without the aid of analytical instruments. Having
identified each of the minerals making up a rock, it is then
possible to give the rock itself a name, based on the criteria
summarized in Appendix 2, and thence to deduce the conditions
under which the rock is likely to have formed.
Key idea: Minerals and silica content

The minerals that crystallize in a cooling magma depend on the
magma’s silica content. The less silica overall, the greater the
ratio of metallic elements (especially magnesium) to silicon in the
minerals.
Spotlight: Grain size

Definitions of coarse, medium and fine grain size are given in
Appendix 2. When ‘grain size’ is used to describe a sedimentary
rock, it refers to the size of particles that were individual grains of
sediment and is an entirely appropriate term. However, in igneous
and metamorphic rocks it refers to the most abundant size of
crystal, so really it should be called ‘crystal size’, as these crystals
were never really ‘grains’ as such.

Case study: Crystallization of minerals in
cooling magma
As a magma rises it gets cooler, because heat escapes into the
surrounding rocks. Eventually the temperature and pressure
will be appropriate for minerals to start crystallizing. The nature
of these minerals depends on the overall composition of the
magma. The growth of mineral crystals in a magma is not like the
freezing of water, where the whole liquid solidifies at a specific
temperature. Instead, each mineral species begins to crystallize
at a different temperature. Generally speaking, the minerals
richest in iron or magnesium crystallize at higher temperatures,
accompanied by a calcium-rich variety of plagioclase feldspar,
whereas quartz and the sodium- and potassium-rich feldspars
crystallize at lower temperatures. This process, which is
effectively partial melting in reverse, is known as fractional
crystallization.
If the crystals are carried along with the remaining magma until
everything has crystallized, then the overall average composition
of the eventual rock is the same as that of the initial magma,
even though the rock consists of several minerals each having
a different composition. However, if the first crystals become
(a) (b) (c)
andesitic magma
basaltic magma
olivine pyroxene
crystal crystal
Figure 6.1 Schematic cross-section (not to scale) showing how a body of
basaltic magma (a) can begin to grow crystals of olivine and pyroxene (b).
If these crystals begin to settle out, as in (c), they form a rock of ultrabasic
composition and the remaining magma (which may crystallize later) is
andesitic (‘intermediate’) in composition.
110

separated from the magma (perhaps because they settle to the
bottom, or because the magma is squeezed out), the remaining
magma will have a different composition, richer in silica than that
it started with. In this way, andesitic magmas can evolve from
originally basaltic magmas (Figure 6.1), and granitic magmas
can evolve from andesitic magmas. When these new magmas
crystallize, they will form rocks richer in silica than the initial
magma, whereas the rock formed by the early-formed crystals
that became concentrated elsewhere will be poorer in silica than
the initial magma.
Key idea: Changing composition

Separating crystals from melt during fractional crystallization
is a way to generate rock that differs in composition from the
original magma.
At the surface, a basaltic magma begins to crystallize at

about 1200 °C and is entirely solidified by about 1080 °C.
The corresponding temperatures for a granitic magma are
1100 °C and 960 °C, and those for an andesitic magma lie
somewhere in between. At depth, the pressure acts to increase
these temperatures by about 2 °C for every 10 km of depth.
However, this applies only for water-free magmas. Above
destructive plate boundaries, escaping subducted water means
that magmas are usually rich in water. The effect of this is
to reduce the temperature at which crystallization begins, so
that in a ‘wet’ granitic magma at 10–30 km depth, crystals
do not begin to form until the temperature has dropped to
about 720 °C. On approaching the surface, the water vapour
is liable to escape, as we saw in the previous chapter. If the
magma had already cooled to close to its ‘wet’ crystallization
temperature, then this loss of water will cause the magma
to crystallize rapidly because it will be well below its ‘dry’
solidification temperature, and the expansion and escape
of the water vapour and other gases is likely to shatter the
solidifying magma explosively.

Key idea: Wet and dry magma
Escape of water from magma can cause the magma to solidify.
Granite
Crystallization promoted by volatile loss is one reason why
granitic magmas, in particular, are more widely known for the
large intrusions that they form than for volcanic equivalents.
Another reason is that they are about a thousand times more
viscous than basalts, and so are less easily extruded on to the
surface. It is not uncommon to find coarse-grained solidified
intrusions of granitic composition that are tens of kilometres
across. These constitute the well-known rock type granite
(Plate 6).
Spotlight: One man’s granite …

The term ‘granite’ is so well known that stonemasons and the like
tend to use it for any crystalline rock, irrespective of whether its
mineralogical composition fits the geological definition of granite.
Authors of a poetic bent use the term even more loosely, to refer
to any rock that they consider to be hard and enduring, although in
fact granite is not particularly immune to the effects of wind and
rain (see Chapter 8).
How, then, does a granitic magma originating, say, 30 km deep

near the base of the crust, rise upwards? As magma forms it
seeps along grain boundaries, but it cannot flow freely enough
to escape until about 5 per cent of the source rock has melted,
which may take 10 000 years or more. If there is a nearby
plane of weakness, such as a fault or fracture, the magma may
escape up it in a matter of a thousand years or so, and spread
out higher in the crust to form a large granite mass. However, if
there is no easy pathway available, magma will remain trapped
at depth until it has grown into a body that is large enough to
force its way up to a shallower level by buoyancy alone. This is
thought usually to require melting of about a third of the source
112

rock. At first, the surrounding rocks will be very hot, and soft
enough to allow the granite magma to push them aside, rising
at a rate of a few centimetres or metres per year. As the granite
reaches progressively shallower depths, the surrounding rocks
will be colder and therefore more brittle, so that eventually the
granite can rise no further.
However, there are other processes that may aid magma ascent
beyond that stage. First, if the overlying rocks have a melting
point similar to or lower than that of the magma, they may
become assimilated into the magma. Essentially, the magma melts
its way through, continuing to rise until it has run out of heat.
Remember this!
This assimilation process is another way in which magma
composition can evolve, by the mixing in of other rock.
Secondly, the granite may pluck away blocks of rocks from

above. These sink down through the granite, allowing the granite
to pass upwards. This process is called stoping, and is shown
in Figures 6.2 and 6.3. The stoped blocks, which can be of any
size, may either sink to the bottom of the magma unscathed, or
may be melted by the heat of the magma and become assimilated
within it. Thirdly, granites can be squeezed upwards along gently
inclined faults, as happens in the Himalayas along faults related
to the collision between India and Asia.
Figure 6.2 Sketch cross-section to illustrate stoping at successive

intervals. A rising body of magma plucks blocks off the roof to make way
for itself. The scale could be anything from a metre to several hundred
metres across.

Figure 6.3 A pale granitic rock that became solidified in the act of stoping
its way upwards into a darker rock type. The geological hammer gives the
scale.
Most of the granites familiar to geologists ceased to rise at

depths of a few kilometres within the crust. The reason why
they can be seen today at the surface is that millions of years of
erosion have worn away the original overburden of other rock.
Exhumation (as it is called) is encouraged by the fact that, even
when solid, granite is less dense than most other rock types,
and it tends to buoy up the terrain isostatically, increasing its
vulnerability to erosion.
Granites can form wherever the composition of originally
basaltic or andesitic magmas has become sufficiently rich in
silica, and so they can be found in any setting. However, they
are particularly common above destructive plate margins
(caused in particular by the melting of the lower crust above
a subduction zone) and in the mountain belts at sites of
continental collision. Here, the thickness of the continental crust
is usually doubled (Figure 4.3); and because uranium, thorium
and potassium are concentrated in the crust, the rate of local
radiogenic heat production is doubled too. This encourages
melting within the thickened crust even after relative motion
between the two collided plates has ceased.
114

Figure 6.4 Half Dome, a famous granite pluton exposed by erosion in
Yosemite National Park, California. The sheer face, beloved of climbers,
is a result of erosion taking advantage of joints that were initiated as the
pluton was cooling.
Large granite massifs (such as that shown in Figure 6.4) are

termed plutons, after Pluto, the Roman god of the underworld.
Sometimes, geophysical evidence suggests that several plutons
are joined together at depth, in which case the larger unit is
referred to as a batholith. The relationship between plutons and
a batholith is shown in Figure 6.5.
Not all plutons are granitic in composition. For example, the
60-million-year-old continental flood basalts described in the
previous chapter (reflecting the rifting of the North Atlantic
Ocean between Britain and Greenland) are associated with
plutons of basaltic composition. Being course-grained, their
rock type is called gabbro rather than basalt (see Appendix 2).
Key idea: Plutons

A pluton is a coarse-grained body of igneous rock that crystallized
at depth.

10 km batholith plutons
Figure 6.5 Cross-section into the crust, showing several plutons as

upward projections from a more extensive batholith at depth.
Basaltic intrusions in the

oceanic crust
Gabbro also crystallizes to form the lower oceanic crust
at constructive plate boundaries. This gabbro layer can be
recognized seismically in the oceans and can be studied on
land in ophiolites, which (see Chapter 4) are slices of oceanic
crust and upper mantle that have been thrust over the edge of a
continent during a collision. It was once thought that the magma
chambers, within which these oceanic gabbros crystallize, are
ten or more kilometres wide. However, studies in the ocean
(attempting to map seismic reflections from the roof of these
chambers, or to locate the molten volumes where S-waves cannot
penetrate) have demonstrated that these magma chambers tend
to be narrow, possibly ephemeral, features. Much of the gabbro
appears to arrive in place as a crystal-rich mush.
Between the gabbro layer and the typical basaltic pillow lavas
erupted on to the sea floor, there is a layer that is neither
plutonic nor volcanic. This marks the depth at which pulses of
magma are injected upwards along a vertical fissure to feed the
lavas. When magma freezes in such a fissure, it forms a vertical
curtain of intrusive igneous rock called a dyke. In sea-floor
spreading, successive dykes intrude either beside or inside each
other, and so a layer is formed consisting of nothing but dykes
(some of them whole, others split), which is called a sheeted
dyke complex. The relationship between these layers in the
oceanic crust is shown in Figure 6.6.
116

pillow lavas
sheeted
dykes
gabbro
1 km
mantle
magma chamber
Figure 6.6 The constitution of the oceanic crust, showing a cross-section
through a spreading axis. Crust on either side is moving apart. Basaltic
magma cools slowly at depth to form the coarse-grained basaltic rock
type known as gabbro. The curved lines represent successive walls of
the magma chamber, which can maintain a constant volume only if the
rate of magma supply from below keeps pace with the rate of sea-floor
spreading. Some of the magma is injected up fissures to the sea floor,
where it is erupted to form basalt lava (often taking the form of pillow
lava), having the same composition as the gabbro but a much finer grain
size because it cools more rapidly. Magma that solidifies within the
fissures forms vertical curtains of medium-grained rock about a metre
wide, called dykes.
Spotlight: Dyke or dike?

‘Dyke’ is spelled ‘dike’ in America, but to a geologist it means the
same thing.
Dykes and sills

Dykes cool faster than plutons because they are thin curtains,
which allows them to lose their heat to either side more easily.
As a result, the crystals do not grow so big as in a plutonic rock
but grow coarser than in a volcanic rock. Dykes intruding other
rock types are common in volcanic areas of the continents,

but sheeted dykes, where dykes intrude nothing but dykes, can
form only where there has been continual spreading, and are
known only in ophiolites and below the ocean floor.
Because they are injected up cracks or fissures, dykes record
tension in the crust at the time of their formation. In many
volcanic provinces, dykes occur roughly parallel to one another
(at right angles to the direction of crustal extension), forming
a dyke swarm. Near a main volcano it is common for dykes to
radiate away from the volcano. Volcanic eruptions that occur
on the flank (side) of a volcano (rather than at the summit vent)
are usually where the top of a radial dyke reaches the surface.
Another form of dyke occurs where magma is injected up the
fracture that has allowed the floor of a caldera (Figure 5.14) to
subside. This is known as a ring dyke, because it has a circular
plan view.
Not all shallow intrusions cut up vertically like dykes. Sometimes
magma is injected in a horizontal sheet, especially where the
surrounding rock is composed of flat-lying beds and the magma
can squeeze along a bedding plane. The generally horizontal
igneous body that results is described as a sill. The relationship
between dykes and sills is shown in Figure 6.7. Sheets intruded
along inclined fractures are rarer. These have no special name,
but would be described as ‘inclined sheets’.
Key fact: Dykes and sills

Dykes are steep-to-vertical curtains of igneous rock, and sills are
near-horizontal sheets of igneous rock. Both are usually shallow
and tend to have medium grain sizes, except at their chilled
margins (see below).
Grain sizes in igneous rocks

I have made a generalization in this chapter that plutonic rocks
are coarse grained, shallow intrusions are medium grained, and
volcanic rocks are fine grained. Although invariably true for
plutonic rocks, there are circumstances where these rules are
broken for shallow intrusions and volcanic rocks. For example,
118

lava flow
dyke
sill
Figure 6.7 Schematic cross-section showing the difference between a

dyke and a sill. Both are sheet-like, tabular bodies, but a dyke is vertical
and tends to cut across any bedding in the rocks it intrudes (described as
a discordant intrusion). In contrast, a sill is intruded parallel to bedding
(concordant intrusion), though it may have the occasional discordant step,
as in this example.
if magma begins to crystallize by slow cooling at depth but is

then erupted, it may carry with it some sizeable crystals that had
already begun to grow. These will become surrounded by much
finer crystals that did not begin to grow until the magma cooled
rapidly at or near the surface. Recognizing isolated large crystals
(known as phenocrysts) in a lava or shallow intrusive rock
reveals that the magma may have had a complicated history,
cooling in at least two stages.
Another exception involves the very last fraction of a per cent of
magma to crystallize in a magma body. This is likely to be very
rich in water and volatile gases (such as fluorine), because these
tend to be excluded from crystals and so become progressively
concentrated in whatever melt remains as crystals grow. Strong
concentrations of these make it difficult for new crystals to begin
to form, but it also makes it easier for any crystals that do get
started to grow quickly. As a result, crystals grow very large, in
extreme cases a metre or more long, but more usually centimetres
in size. Such very coarse-grained igneous rocks are described as
pegmatites, and can be found in veins (filling irregular fractures)
within igneous intrusions or the surrounding rocks.

Key idea: Crystal size in igneous rocks
In an igneous rock, larger crystals tend to indicate slower
cooling. However, this does not apply to pegmatites, where
special conditions force crystals to grow large irrespective of
cooling rate.
Conversely, not all volcanic rocks are crystalline at all. Magma

erupted at the surface may chill so rapidly that there is no time
for crystals to begin to form. The rock is therefore a frozen
mixture of the elements that were present in the magma. It
lacks crystalline structure, and is described as glass. Pyroclastic
igneous rocks may consist largely of glassy fragments, and
many lavas consist of fine-grained crystals floating within a
glassy matrix.
Shallow intrusions (especially dykes, although generally of
medium grain size – with or without a few larger phenocrysts)
often have very fine crystals, or even a glassy texture, at their
edges. Such a chilled margin forms because the surrounding
rock was cold, causing the edge of the igneous body to freeze
quickly against it. Chilled margins are rare at the edges of
plutonic intrusions, and there are two factors contributing
to this. The first is that, being deeper down, things are warmer
in general anyway, so the surrounding rocks are not usually
cold. The second is that plutons are so big and contain so
much heat that they heat up the surrounding rock before
being cooled much themselves. This heat can cause the
surrounding rock to undergo dramatic changes through a
process known as metamorphism, which is the subject of the
next chapter.
120

Dig deeper
If you want to read more deeply about igneous rocks, I
recommend Igneous Rocks and Processes: A Practical Handbook
by Robin Gill (Wiley-Blackwell, 2010) and The Field Description
of Igneous Rocks by Dougal Jerram and Nick Petford (Wiley-
Blackwell, 2011).
The collection and study of minerals is an enthralling pastime
in its own right, and there are many books and websites dealing
with the subject. Starting points are:
http://webmineral.com/ This is a database of minerals, including
photographs and analyses.
http://www.minersoc.org/ The website of the Mineralogical
Society (UK) includes a picture gallery of spectacular mineral
specimens and links to other useful sites.
http://www.minsocam.org/msa/collectors_corner/index.htm The
‘collector’s corner’ at the website of the Mineralogical Society of
America contains pictures of minerals and guidance on rock and
mineral collecting.

Fact-check
1 Which of these terms denotes an intrusion of igneous rock
along a narrow near-vertical fissure?
a Dyke (or dike)
b Gryke (or grike)
c Pluton
d Sill
2 Which of these statements is correct?
a An igneous rock containing about 50 per cent silica contains
a lot of quartz
b If an igneous rock is low in silica, it probably contains a lot of
magnesium
c All silicate minerals contain silicon, oxygen and metals only
d Olivine is composed either of magnesium bonded with silica
or iron bonded with silica
3 Which of these is a correct example of ‘fractional crystallization’?
a A lava flow consisting of small crystals encased in volcanic
glass
b The metal/silicon ratio in a silicate rock
c The Fe/Mg ratio in an olivine crystal
d Separation of magma from the first crystals to grow, so that
the rock that forms from the separated magma is richer in
silica
4 Which of these is not true of granite?
a It is less dense than most other rock types in the crust
b It has similar grain size to gabbro
c It is often intruded into the lower oceanic crust at
constructive plate boundaries
d It can make space for itself by stoping
5 Which of these correct?
a A pluton is bigger than a batholith
b In an igneous rock, a phenocryst is larger than the crystals
that surround it
c A sill is steeper than a dyke
d Granite magmas are less viscous than basaltic magmas
122

Metamorphism
7
33how heat and pressure can cause
dramatic changes to rock
33how these changes can be recognized,
and what they can tell us about the
processes that caused them.
123

All rocks are composed of minerals. These may be crystals that
grew while a magma was cooling (as in the igneous rocks you
learned about in the previous chapters) or grains laid down to
form a sedimentary rock (by processes that you will encounter
in the next two chapters). Each mineral is stable only over
a particular range of temperature and pressure. Beyond that
range, it will tend to break down or combine with neighbouring
minerals to form new minerals.
Most igneous minerals are unstable at the low temperatures
prevailing at the surface, and they tend to rot. This chemical
weathering process is slow (like most chemical changes) because
of the low temperatures. However, when a rock is held at a
high temperature or pressure, its constituent minerals can
change dramatically and thus completely alter the character of
the rock. Recrystallization of this sort under the influence of
temperature and/or pressure, but without melting, is known
as metamorphism, and is responsible for the creation of
metamorphic rocks.
Key idea: Recrystallization without melting

Metamorphism involves recrystallization of a pre-existing rock,
without melting. It can be caused by pressure, or temperature,
or both.
Thermal metamorphism
The previous chapter closed with the example of a large hot
igneous body intruding into colder rocks. When the rocks
surrounding such an intrusion are examined in the field, they
may show the effects of having been heated across a zone
extending a few hundred metres from the contact. This is
described as contact metamorphism or thermal metamorphism.
Key idea: Thermal (or contact) metamorphism

Thermal (or contact) metamorphism is a result of heating adjacent
to an igneous intrusion.
124

Let’s take the example of a granite intruding a fine-
grained muddy sedimentary rock. Right next to the granite
this mudrock will have become very hard and splintery,
containing clusters of new minerals that grew within it as
a result of contact metamorphism. A splintery, spotty rock
such as this is described as a hornfels. Prominent among
the metamorphic minerals is likely to be sillimanite, having
the formula Al2SiO5. This never crystallizes from a magma;
it grows only by metamorphism in rocks that contain
aluminium and silica (a condition generally met in rocks
formed from mud). If you walk away from the edge of such
an intrusion, you should eventually notice that the hornfelsed
mudrock becomes less splintery, and the spotty pattern is
lost. Soon the sillimanite disappears, and instead there is a
different metamorphic mineral occurring as isolated stubby
crystals up to a centimetre long. These crystals are of the
mineral andalusite, whose formula is also Al2SiO5, but which
grows at lower temperatures than sillimanite and has a
different crystal structure. As you continue away from the
intrusion, the andalusite crystals become fewer and smaller,
until you find yourself in unmetamorphosed mudrock. You
have now walked right through the metamorphic aureole of
the granite.
Remember this!
Most minerals that grow during metamorphism are varieties that
never crystallize from a magma.
Such a metamorphic aureole is sketched in Figure 7.1. Note,

however, that sillimanite and andalusite grow only in rocks
that have the right chemical make-up. In rocks of other
compositions, the metamorphic minerals will be different. For
example, when a granite intrudes an impure limestone, the
most prominent metamorphic mineral close to the contact
is likely to be the iron-rich variety of olivine Fe2SiO4 (maybe
Fe1.8Mg0.2SiO4), which is virtually unknown in igneous rocks.
Going back to metamorphism of rocks that were originally
muddy, there is a third aluminium silicate mineral, called
7 Metamorphism 125

unmetamorphosed 1 km
andalusite
zone hornfels
texture
sillimanite
zone
granite
Figure 7.1 Sketch map of the metamorphic aureole of a granite that was
intruded into mudrocks. In this example, the outer metamorphic zone is
characterized by the metamorphic mineral andalusite, whereas the next
zone inwards has sillimanite instead. Adjacent to the granite, the surrounding
rocks have become brittle and developed a spotty hornfels texture.
kyanite, that can form only if the pressure is high enough.

The minerals sillimanite, andalusite and kyanite are described
as polymorphs of the compound Al2SiO5. Their crystalline
structures differ by having different arrangements of atoms, and
so the shapes in which each can grow are different. Figure 7.2
shows the conditions of pressure and temperature under which
each will form. From this we can see that the absence of kyanite
in the metamorphic aureole of the granite in Figure 7.1 shows
that the pressure at the time of metamorphism must have been
less than about 3 kilobars, equivalent to a depth of 10 km or
less. Conversely, the aureole of a granite that was intruded at
a depth of between about 10 and 15 km would contain an
inner kyanite zone, then an andalusite zone surrounded by a
sillimanite zone; whereas an intrusion deeper than about 15
km would lack the andalusite zone, and its aureole would pass
directly from kyanite to sillimanite. In addition, there might be
high-pressure minerals of other compositions such as garnet,
formula (Fe,Mg)3Al2(SiO2)3, which is sometimes of sufficient size
and quality to be regarded as a semi-precious gemstone.
You may wonder how it is that we can still find sillimanite
surviving in the aureole of a granite exposed at the surface today,
where clearly neither the pressure nor temperature are high
enough for conditions to lie within sillimanite’s stability field.
126

0 0
andalusite
3 10
pressure (kbar)
depth (km)
sillimanite
6 20
kyanite
9 30
400 500 600 700 800

temperature (°C)
Figure 7.2 The stability fields of the metamorphic minerals kyanite,
sillimanite and andalusite. These all have the formula Al2SiO5, but are
polymorphs stable over different ranges of temperature and pressure. The
pressure scale is labelled in kilobars (1 kbar is equal to a thousand times
atmospheric pressure), and the scale on the right gives the approximate
conversion to depth in the crust.
The answer is that rocks tend to retain the minerals that grew
within them, at or near the conditions of highest temperature
and pressure encountered during metamorphism. During cooling
(or exhumation) of a metamorphosed rock there is usually
insufficient impetus to drive the retrograde reaction from a high-
temperature (or high-pressure) mineral to its low-temperature
(or low-pressure) equivalent. This is fortunate, otherwise the
only minerals we would ever see at the surface would be those
stable at pressure of 1 atmosphere and about 0–30°C!
Remember this!
Minerals that survive outside their stability field are able to do so
because there is not sufficient impetus to initiate the change. In
such a situation they are described as being ‘metastable’.
Regional metamorphism
Very often, pressure, rather than temperature, is the main cause
of metamorphism. Usually the two go together, because as depth
(and hence pressure) increases, so does temperature. However, it
is useful to distinguish thermal metamorphism (which is caused
7 Metamorphism 127

by proximity to a hot intrusion) from regional metamorphism,
which is caused by the pressures and temperatures prevailing
regionally. Any rock that is formed near the surface, whether
volcanically or by deposition as a sediment, is liable to be
buried by subsequent deposits. Eventually, it may find itself
at a depth where pressure and temperature are sufficient for
metamorphism to begin. The whole of the lower crust consists
of regionally metamorphosed rocks, except where it is made of
recently intruded igneous intrusions.
Important departures from average conditions occur in collision
zones (Figure 7.3). Where sediments have been dragged down
into a subduction zone they will experience particularly high
pressures, but having been recently at the surface they will not
be so hot as most rocks at the same depth. They will therefore
experience high-pressure, low-temperature metamorphism, and
will be characterized by appropriate assemblages of minerals.
On the other hand, where the crust is heated on a regional
scale by large numbers of igneous intrusions, such as below the
volcanic region near a subduction zone, crustal temperatures
will be higher than is normal for such relatively shallow depth.
The rocks here experience low-pressure, high-temperature
metamorphism.
Low pressure, High pressure,

high temperature low temperature
continental crust oceanic crust
mantle lithosphere mantle lithosphere
50 km
Figure 7.3 Cross-section showing zones of low-pressure, high-

temperature metamorphism, and of high-pressure, low-temperature
metamorphism associated with a destructive plate boundary.
128

Key idea: Regional metamorphism
Regional metamorphism is a result of burial to depths where
pressure (and possibly temperature also) is high.
A zone of high-pressure, low-temperature metamorphic rocks

(consistent with formation in a subduction zone) found next to
a zone of low-pressure, high-temperature metamorphic rocks
(consistent with formation in heated crust above a subduction
zone) is called a ‘paired metamorphic belt’.
Remember this!
When found in ancient rocks, a ‘paired metamorphic belt’ is good
evidence that there was once a destructive plate boundary there.
Metamorphic facies
Just as in thermal metamorphism, the metamorphic minerals
that grow during regional metamorphism depend on the
original composition of the rock. For example, a metamorphic
mineral that requires magnesium cannot develop in a rock
that lacks this element. Because of this, rocks that have been
metamorphosed under identical conditions can contain
entirely different minerals. Geologists therefore classify
metamorphic rocks by groupings called facies. Each facies
reflects a particular range of temperatures and pressures, but
the minerals that develop in each facies differ according to the
original rock type.
The pressure and temperature conditions of the main
metamorphic facies are shown in Figure 7.4. Metamorphism
in the conditions near the top left of this diagram is described
as low grade and metamorphic grade is said to increase
with distance from this point. Table 7.1 contrasts the most
abundant minerals found in metamorphosed mudrocks with
metamorphosed basaltic rocks in some of the facies named in
Figure 7.4.
7 Metamorphism 129

Spotlight: Facies
Facies is a term that seems strange to most people when they
start learning geology. Pronounced ‘fasheez’, it is a nineteenth-
century coining from Latin, meaning ‘form, appearance or face’.
Apart from its use in metamorphism, ‘facies’ is also used to refer
to a sedimentary rock unit that formed under a certain range
of conditions, or (similarly) to a suite of conditions responsible
for deposition of a particular kind of sediment. For example, a
geologist might speak of a deltaic facies of sandstone, meaning
sandstone possessing various characteristics that indicate that it
was deposited in a delta.
Table 7.1 The mineral assemblages characterizing some metamorphic

facies in two different pre-metamorphic rock types.
Facies Originally mudrock Originally basaltic rock
greenschist muscovite, chlorite, quartz, albite, epidote, chlorite
sodium-rich plagioclase
amphibolite muscovite, biotite, garnet, quartz, amphibole, plagioclase
plagioclase
granulite garnet, sillimanite, plagioclase, calcium-rich pyroxene, calcium-
quartz rich plagioclase
eclogite garnet, sodium-rich pyroxene, sodium-rich pyroxene, garnet
quartz
One thing that Figure 7.4 does not show is that, in the presence
of water, granite crust will begin to melt at temperatures lower
than those required for granulite facies metamorphism. When
granulites are found, they are a sure sign of dry conditions in
the lower crust during metamorphism.
Key idea: Metamorphic facies

A metamorphic facies corresponds to a particular range of
temperatures and pressures at which metamorphism occurred,
but the minerals that develop in each facies differ according to the
original rock type.
130

temperature (°C)
0 200 400 600 800
0
Hornfels
2 Zeolite
10
4
Amphibolite
Greenschist
pressure (kbar)
Granulite
depth (km)
6 20
Blues
8
chist
30
10
12 40
Eclogite
conditions not found in nature

Figure 7.4 The pressures and temperatures required to produce
metamorphic rocks of each facies. These facies grade into one another,
and the boundaries between them are ill defined, but they have been
separated here for clarity.
Textures of metamorphic rocks

There is an alternative way to describe metamorphic rocks,
which, for the geologist in the field, is simpler and more
useful than facies. This is based simply on the texture of the
rock – texture being a term that encompasses the size of the
crystals, their shapes and their relative orientations. You have
already met hornfels, which is a textural description for a
fine-grained metamorphic rock in which spotty agglomerations
of metamorphic minerals have grown during contact
metamorphism.
Case study: Metamorphic foliation

In regional metamorphism, the rocks are affected by pressure as
well as temperature. When this happens, crystals tend to line up so
they lie at right angles to the direction of maximum compression.
This may be brought about both by rotation of existing crystals and
by growth of new crystals in this pressure-controlled orientation.
7 Metamorphism 131

The effect is most noticeable for those minerals that tend to form
flat or elongated crystals, and less apparent in minerals whose
crystals are more equidimensional. The alignment of minerals
into layers is described as foliation, and such rocks are said to
be foliated. If you can spot foliation in the field, it tells you about
the grade of metamorphism and the orientation of the deforming
stress. Sometimes, metamorphism causes elongated minerals
such as amphiboles to become aligned so that they all point in the
same direction. This is described as lineation.
The lowest grade of regionally metamorphosed rock is slate, which

is formed by zeolite facies metamorphism of mudrocks or fine
silts. Slate is well known for its ability to be split thinly along flat
foliation surfaces, hence its traditional use as a roofing material.
These planes of weakness, or cleavages, running through slate
are parallel to the metamorphic foliation. They have no relationship
with any sedimentary layering that may have been present in
the rock before it was metamorphosed. Slate is produced only
from rocks that were originally fine grained and sedimentary in
nature; if a basalt or a sandstone were to be metamorphosed
in the same facies as a slate it would show few obvious signs of
metamorphism, except maybe under microscopic examination. In
particular, a coarse-grained rock cannot develop a closely spaced
slaty cleavage.
Warning! Do not confuse cleavage in rock with cleavage in

minerals. Cleavage in a mineral is defined by planes of weakness
in its crystalline structure. Cleavage in rock is because virtually all
the fine-grained platy minerals have become oriented parallel to
one another.
If a slate is subjected to slightly higher grades of metamorphism

(still mostly in the zeolite facies, but at higher temperature or
pressure), fine flakes of metamorphic mica and chlorite minerals
grow on the cleavage planes. These give the rock a greenish
sheen. Often the cleavages become wrinkled, and the rock would
then be described as a phyllite.
At still-higher grades of metamorphism, the metamorphic

minerals grow bigger and can be identified with the naked eye.
Note the different significance of crystal size in igneous and
132

metamorphic rocks. In igneous rocks larger crystals usually
reflect slower cooling, whereas in metamorphic rocks larger
crystals are a result of higher temperatures and/or pressures of
metamorphism.
The next grade of foliated metamorphic rock after phyllite is called

schist. In a schist the metamorphic minerals are large enough
to be readily visible to the naked eye. The foliation is usually wavy
rather than flat, and is often made conspicuous by concentrations
of mica (which grows as shiny platy crystals). Schists resulting
from metamorphism under normal continental conditions (where
temperatures tend to rise by about 30°C for each kilometre
increase in depth) are formed under conditions of greenschist
facies metamorphism, so named because of colour imparted
by the green metamorphic minerals chlorite and (for rocks of
basaltic origin) epidote. However, under the high-pressure, low-
temperature conditions that define the blueschist facies, a blue
amphibole mineral (glaucophane) may develop. A schist texture
can develop irrespective of whether the rock was originally fine
grained or coarse grained.
Rocks metamorphosed under eclogite, amphibolite or granulite

facies conditions are too far gone to contain abundant mica, and
they rarely contain any other strongly planar minerals. In these the
foliation is picked out by dark and pale minerals segregating into
layers a few millimetres thick. This type of metamorphic rock is
described by the German term gneiss (pronounced ‘nice’).
The most extreme metamorphic rocks are gneisses that reached

the verge of melting. In these, veins and blobs of material
crystallized from a melt are interspersed through the rock, usually
strongly deformed and flattened into the plane of foliation. Such a
rock is referred to as a migmatite.
The metamorphic textures described above can develop only

in rocks that originally consisted of appropriate mixtures of
minerals. There are two common rock types that are each made
of essentially a single mineral and so cannot develop these
textures, irrespective of the facies conditions under which they
are metamorphosed. One of these is sandstone consisting of
7 Metamorphism 133

quartz grains. When a pure quartz sandstone is metamorphosed,
the grains tend to fuse together producing a harder rock known
as quartzite, but no new minerals can be formed. The other is
limestone, which is a non-silicate rock consisting of calcium
carbonate. When metamorphosed, limestone turns into marble,
which can be a beautiful decorative stone. Recrystallization of
calcium carbonate usually gives it a bright white colour, and any
impurities tend to concentrate into colourful veins.
Spotlight: Marble
Marble is another of those rock names (like granite) that is used in
a broader sense by non-geologists. Masons describe any durable
stone that can be polished to reveal a pretty surface as ‘marble’,
irrespective of whether or not it consists of calcium carbonate.
When we look at deformation of rocks, in Chapter 10, you will

see that shearing of rock in fault zones can lead to localized
metamorphism of another kind.
Metamorphic rocks and most igneous rocks are products of
processes within the crust. These are the hard rocks that often
form the most prominent landscape features. Now we will
begin to consider what happens to these rocks when they are
exposed at the surface, and see how detritus derived from them
accumulates to form the third main class of rocks, the sediments.
Dig deeper
Metamorphism is a difficult and complex subject. One standard
text is Petrogenesis of Metamorphic Rocks by Kurt Bucher and
Rodney Grapes (Springer, 2011).
A book relevant to both this chapter and the previous one is
Essentials of Igneous and Metamorphic Petrology by B. Ronald
Frost and Carol D. Frost (Cambridge University Press, 2013).
A useful starting point to learn about metamorphism on the
Internet is http://www.geolsoc.org.uk/ks3/gsl/education/
resources/rockcycle/page3576.html
134

Fact-check
1 What is a paired metamorphic belt?
a A zone of andalusite next to a zone of sillimanite
b A belt of blueschist next to a belt of greenschist
c A belt of high-pressure, low-temperature metamorphic
rocks next to a belt of low-pressure, high-temperature
metamorphic rocks
d Any two metamorphic facies occurring next to each other
2 What is a polymorph?
a A crystal that can grow in a variety of shapes
b A mineral with the same chemical formula as another but a
different internal crystal structure
c A metamorphic mineral that can grow in more than one
metamorphic facies
d A rock that has slatey cleavages in many different
orientations
3 Which of these is not correctly described as cleavage?

a The way two crystals stick together in a rock sample
b A flat plane of weakness in a slate
c A wrinkled foliation surface in a phyllite
d A flat plane of weakness within a crystal
4 Which of these metamorphic facies can represent simultaneous
high pressure and high temperature?
a Blueschist
b Greenschist
c Eclogite
d Hornfels
5 Which of these metamorphic facies can represent simultaneous

low pressure and low temperature?
a Amphibolite
b Granulite
c Hornfels
d Zeolite
7 Metamorphism 135

6 Which of these is a low-grade regionally metamorphosed rock
type?
a Hornfels
b Granulite
c Migmatite
d Slate
136

8
Weathering,
erosion and
transport
33how rocks are worn away by chemical
and mechanical processes
33about the landscapes produced by
erosion
33how rock fragments can be transported
by wind, water and ice.
137

You saw in Chapter 7 (Metamorphism) that each mineral is
chemically stable over only a restricted range of temperature
and pressure. Most minerals in igneous and metamorphic rocks
form at temperatures and pressures considerably higher than at
the surface, so they are no longer stable when the rock in which
they are embedded becomes exposed. However, these minerals
do not vanish in a flash (or otherwise we would not be able to
study them!).
Conversion to new minerals that are stable at low temperatures
and pressures happens very slowly indeed. In dry conditions
a mineral may survive in a metastable state for hundreds of
millions of years at the surface without appreciable change.
The presence of water allows many minerals to ‘weather’ away
rather more quickly, although the process may be virtually
imperceptible on human timescales.
Key idea: Metastable survival

Most igneous and metamorphic minerals are, strictly speaking,
unstable under the conditions prevailing at the Earth’s surface.
However, change is usually imperceptibly slow.
Chemical breakdown of minerals

during weathering
Water controls the types of minerals that develop during chemical
weathering. Large crystals of igneous or metamorphic silicate
minerals tend to rot away to tiny flakes of clay minerals. The clay
family of minerals is very large, but all clays have characteristics
in common. They are all hydrated, having the group OH (derived
from water) in their formula, and often water itself, H2O. Their
crystal structure consists of sheets of silicate held together only
loosely, which is what gives clay particles their microscopic flaky
nature and makes a handful of damp clay so pliable.
The rotting of feldspar to produce a clay mineral may be
described by the following chemical reaction between feldspar
and water (that is slightly acidic because of atmospheric carbon
dioxide dissolved in it).
138

2KAlSi3O8 + 2CO2 + 3H2O = Al2Si2O5(OH)4 + 4SiO2 + 2K+ + 2HCO3-
feldspar carbon water kaolinite (a dissolved dissolved dissolved
dioxide (in clay mineral) silica potassium bicarbonate
solution) ion
This sort of reaction is described as hydrolysis (which means

‘breaking apart by water’). In this example, the original mineral
(feldspar) is broken apart by the chemical action of water,
leaving particles of a clay mineral (in this case kaolinite) as a
solid residue. The clay mineral contains all the aluminium and
some of the silicon that was in the feldspar, but the potassium
and some of the silicon (as dissolved SiO2) has been washed
away in solution.
Case study: Hydrolysis

The effects of hydrolysis are far reaching, and are worth looking at
in a little more detail. Hydrolysis can take place both at the surface
and at shallow depths where water is circulating through the rock.
When carried to its extreme (and with similar reactions affecting
other minerals), this can turn a granite into a deposit of kaolinite,
which can be quarried as a source of china clay.
Hydrolysis contributes to the breakdown of micas, amphiboles
and many metamorphic minerals. Potassium, sodium, calcium
and magnesium tend to be removed in solution, but some can
be retained in many clay minerals, like illite KAl3Si3O10(OH)2 or
montmorillonite, which has the frighteningly complicated formula:
(Na,Ca,Al,Fe,Mg)8(Si3O10)3(OH)10.12H2O.
When pyroxenes and olivines react with water, their magnesium

and silica are carried away in solution, but their iron remains
as a solid residue in the form of iron oxides such as the mineral
hematite (Fe2O3).
The only common material that is more or less immune to

chemical attack is quartz, which is pure SiO2. Although SiO2 is
soluble in acidic groundwater, this is significant only for SiO2
released by chemical breakdown of complex silicate materials
during hydrolysis, such as the dissolved silica derived from

feldspar in the chemical reaction shown above. Quartz itself is
virtually insoluble under surface conditions.
Dissolved calcium and the bicarbonate ion resulting from

hydrolysis may react and combine in a new environment to produce
calcium carbonate, from which limestone is made. (You will look
at this in the next chapter). Other dissolved elements, notably
magnesium, find their way into seawater and become incorporated
into minerals in the oceanic crust by means of chemical exchanges
that occur below hot vents, like the black smokers mentioned
in Chapter 5. There is thus a chemical strand to the rock cycle,
involving solution and re-precipitation of elements.
Key idea: Chemical weathering

Chemical weathering is the destruction of minerals (and hence
rock) by chemical processes, usually involving water.
Physical weathering
Rocks at the surface suffer not only the indignity of chemical
attack; they also are exposed to the abrasive or grinding action
of wind, water and ice. These are very effective agents of
physical weathering. For example, if water repeatedly freezes
and thaws in cracks and cavities, the expansion that happens
when water turns to ice can prise away fragments of rock.
Even in dry conditions, a rock face may flake away because
of cracking caused by expansion and contraction of the rock
itself, as its surface heats by day and cools by night. It may also
be plucked away by the roots of plants. Furthermore, fragments
of rock embedded in a moving ice sheet (a glacier), or carried
along by flowing water or wind, will abrade any surface
that they come into contact with and, in turn, be abraded
themselves.
Where there is a steep rock face, such as a cliff undermined
by pounding waves, lumps of rock may fall and smash under
gravity. Rocks that occur in weakly bonded layers, like
140

slate that breaks easily along its metamorphic cleavage, or
sedimentary rocks that were deposited in successive beds, are
particularly prone to this – and so are rocks in which closely
spaced joint patterns have developed as a result of cooling
(e.g. Figures 5.15 and 5.16) or release of pressure during
exhumation. However, the large blocks that accumulate in this
way may be too big to be moved, even by strong waves or a
fast-flowing river.
For transport to occur, rocks must usually first be broken into
fragments less than a few centimetres across. Glaciers and
landslides are notable exceptions, being capable of moving
much larger pieces. On the other hand, the wind rarely blows
hard enough to move anything coarser than a large sand grain.
When a fine-grained rock has been broken into transportable
fragments, the particles may consist of several minerals still
stuck together. Effectively these are just small pieces of rock.
However, a transportable fragment of a coarse-grained rock is
likely to be comparable in size with the crystals within it, so
in what way does a coarse-grained rock break into fragments?
You might think that each constituent crystal would tend to
become separated from its neighbours. This can happen, but
a characteristic of igneous and metamorphic rocks is that the
minerals are usually intergrown and interlocking, like the
pieces of a three-dimensional jigsaw, so they are really rather
strongly held together. Such rocks usually break across crystals,
rather than between them. What tends to happen in physical
and chemical weathering is that pre-existing fractures within
each crystal are attacked and opened. These tiny fractures take
two forms.
1 Fractures may be caused by deformation of the rock, or
open up as the confining pressure grows less when the rock
is brought closer to the surface. This sort of fracture may be
found in any mineral.
2 Fractures also occur in minerals whose crystals contain
inherent planes of weakness, called cleavage planes. As you
learned in Chapter 6, mineral cleavages pick out planes of
inherently weak bonding in a crystal’s structure. Common

minerals with well-developed cleavages include the micas,
feldspars, amphiboles and pyroxenes. In a mica crystal,
all the cleavages are parallel, so it will tend to break into
thin flakes bounded by cleavage surfaces. The others have
cleavages in two or three orientations, predisposing them to
break into chunkier fragments.
Although larger rock fragments are bounded by fractures that
cut across crystals, most sand-sized particles produced by the
breakdown of a coarse-grained rock inevitably tend to consist
of a piece of a single mineral. The smaller the particle, the
greater its surface area compared with its volume, and so the
more vulnerable it is to chemical weathering and the shorter
its survival in the transport regime before it has rotted away.
I mentioned earlier that quartz is effectively insoluble and
immune to chemical weathering. Another peculiarity of quartz
is that it is the only abundant mineral lacking a well-developed
cleavage. Quartz is therefore physically very robust and it
survives transport very well. It should be no surprise to you
that sand grains on a river bed usually turn out to be mostly
quartz, unless collected close to their source, in which case
there may be a fair proportion of feldspar and fragments of
fine-grained rock too. The other solid residue (mostly clay
particles) is unlikely to be found in the same place as this
material because, being much finer grained, it is washed away
and will settle out eventually in much quieter conditions (as
you will see in the next chapter).
Key idea: Physical weathering

Physical weathering is the abrasion or fragmentation of rock.
Small particles produced thereby may be more vulnerable to
chemical weathering.
Incidentally, elsewhere you may meet the terms ‘erosion’ and
‘weathering’ apparently used interchangeably. This isn’t really
correct. Erosion is a physical process, so it is almost the same
as ‘physical weathering’ although erosion implies removal of the
weathered material, irrespective of whether it was produced by
physical or chemical means.
142

Remember this!
Quartz is by far the most abundant mineral that is resistant to both
physical and chemical weathering. This is why most sand grains in
a river or a sandy desert are made of quartz.
Sand-sized grains in wind and water

Let’s now look at what happens to a sand-sized grain during
transport (Figure 8.1). If broken off an igneous or metamorphic
rock, a grain is likely to be quite jagged with sharp corners. This
shape is described as angular. Each time a quartz grain bounces
into a stationary rock, or collides with another grain, there is an
opportunity for bits of the grain to be knocked off. The corners
are the most vulnerable parts, so these get worn away first,
and the grain takes on a progressively more rounded shape.
In contrast, a feldspar grain is more likely to break into small
fragments controlled by its cleavage planes, and hydrolysis will
cause these to rot away to clay particles.
Currents of water and wind transport grains in similar ways.
However, water is denser and more viscous than air, so it can
move a particular-sized grain in a current flowing much more
angular shape rounded shape
quartz
feldspar
clay particles
Figure 8.1 A grain of quartz starts off angular in shape, but becomes
progressively rounded because of collisions during prolonged transport,
which abrade its corners. On the other hand, a grain of feldspar is liable
to break apart along cleavage planes, and the resulting small pieces will
weather chemically, leaving clay particles as the only solid residue.

slowly. For example, water flowing at only 0.2 m per second
(less than 1 km per hour) is capable of picking up 1-mm-
diameter sand grains. Wind needs to be blowing at 10 m per
second to do the same.
Spotlight: Windy
A wind speed of 10 m per second is 36 km per hour, and corresponds
to wind force 5–6 (a fresh to strong breeze) on the Beaufort scale. You
wouldn’t want it any stronger if you hoped to fly a kite.
Because of the difference between wind- and water-transport

speeds, the collisions suffered by grains transported by wind are
more violent than for water-transported grains. Sand grains blown
around in a desert therefore tend to become very well rounded.
When there are no more corners to knock off, the effect of
subsequent collisions is to make a microscopic pit on the surface
at the point of impact, whose combined effect is to give the grain a
frosted appearance. Seen in close-up under a microscope, it looks
just like etched glass. Another result of the high wind speed needed
to transport sand grains is that rocks exposed to this ‘sandblasting’
may be sculpted into strange shapes.
The relatively gentle transport of grains by water means that
physical weathering proceeds more slowly there. Grains seldom
become very well rounded, their surfaces do not become frosted,
and obstacles do not suffer from the sandblasting effect.
What wind and water have in common is the way in which they
transport grains.
Case study: Transport of grains in a fluid flow

Grains can be transported by a fluid in either of two ways.
1 Grains that are only just small enough to be moved will tend to
stay close to the bed of the stream (in water) or to the ground (in
wind). Some simply get dragged or rolled along, but most of them
bounce, suffering a collision each time they land. The bouncing
mode of transport is called ‘saltation’, which literally means
144

‘jumping’. Grains transported by dragging, rolling or saltation are
described as belonging to the bed-load of the current.
2 Smaller grains, once picked up by the flow, can remain
suspended for long periods, suffering few collisions, and are
said to belong to the suspension-load of the current.
An example might be a fast river in which sand grains are bouncing
along its bed, constituting the bed-load, whereas fine clay particles
are carried along in suspension, discolouring the water and
forming the suspension-load.
Remember this!
The larger sedimentary particles transported by wind or water
belong to the bed-load, and the finer particles to the suspension-
load.
Moving ice
Ice is a rather different transport medium. A glacier is ice
flowing downhill at typical rates of between 20 and 200 m per
year, though occasionally a glacier can surge forward at over
300 m a day. It is a powerful agent of erosion, plucking rock
from the sides and floor of the confining valley. Glaciers also
carry debris that has fallen from above.
Once encapsulated in the ice, these fragments cannot bounce
along. Ice-transported fragments thus tend to remain angular.
The only way in which they can suffer abrasion is if they are
carried at the base of the flow, in which case they will be scraped
against the underlying rock. Debris transported by ice can
often be recognized by scratches (or ‘striations’) on its surface.
Corresponding glacial striations can be found on exposures of
bedrock that lay below a glacier millions of years ago.
You have now seen the effects of erosion and transport on
material that is being transported. I will complete a review of
this part of the rock cycle by describing briefly its effects on the
landscape.

Glacial landscapes
Staying with glaciers, you have already met their ability to pluck
away at the sides and floor of the valley along which they flow.
A valley that has been sculpted by a glacier can be recognized
by its U-shaped cross-sectional profile, very different from the
more V-shaped valleys carved by rivers. Another peculiarity of
glaciers is that they can scour deep hollows on the valley floor,
in a way quite impossible for liquid water to achieve. After
a glacier has melted, these hollows become filled with water,
making lakes. Figure 8.2 illustrates the effects of glaciation on
the landscape.
Glaciers, and more extensive ice sheets, cover most of
Antarctica and Greenland today, with just an occasional
ice-free peak protruding above the level of the ice. Valley
glaciers can be found in most of the world’s high mountain
ranges, and their rate of ‘retreat’ (because melting at the front
Figure 8.2 A landscape that has been sculpted by the action of glaciers.
The large U-shaped valley was carved by a major glacier, which filled the
valley to a depth of some hundreds of metres. Tributary glaciers fed into it
from the ‘hanging valleys ’ on the side, whose floors are far above the floor
of the main valley. The hanging valley on the left contains a mountain lake
or tarn, which has filled a glacially scoured hollow. The ridges between
adjacent hanging valleys are often quite sharp.
146

outpaces resupply by ice flow from the source) is one of the
key indicators of global warming. However, there are large
currently ice-free regions of the globe where the landscape
bears the unmistakable signs of glacial erosion. The areas
affected include North America (down as far as slightly south
of the Great Lakes), and Europe (as far south as southern
Britain). This landscape was developed during the past three
million years, when the polar ice sheets repeatedly advanced
Equator-ward and then retreated.
Periods of extensive ice cover are referred to as glaciations, and
the warmer intervals between (during which ice cover is much
less extensive) are called interglacials. Repeating glaciations
and interglacials make up an ice age, which is a climate
phenomenon usually lasting about 10 million years that has
happened roughly every 200 million years. There is no reason
to think that the current ice age is over. We are experiencing
an interglacial, though the accelerated rate of global warming
in recent years probably has more to do with human release of
greenhouse gases to the atmosphere than to natural cycles (see
Chapter 13).
Because so much water was locked up as land ice during the
last glaciation, global sea level was then more than 100 m lower
than now. Many glacial valleys were cut well below the present
sea level, and are now flooded to form long, steep-sided inlets of
the sea, known by their Norwegian name of fjords.
Remember this!
Sea-level rise today has more to do with the thermal expansion
of seawater (as the global temperature creeps up) than with the
melting of polar ice. Melting of floating ice (as in the Arctic ocean)
has no effect on sea level.
Among the more picturesque landscapes inherited from

recent glaciations are the U-shaped lake-filled valleys of the
English Lake District and the Scottish Highlands, and the fjord
coastlines of Scandinavia, British Columbia and New Zealand’s
South Island. Among the bleaker heritage of those times are the

relatively flat, glacially scoured plains of central Canada and
the ‘channelled scablands’ of Washington and Idaho (north-
west USA), carved by repeated catastrophic floods when the
alternately retreating and advancing ice sheet released the
pent-up waters of glacially dammed lakes such as Montana’s
7000-square-km Lake Missoula.
Spotlight: Frosty language

The terms used by geologists to describe features of a mountain
landscape produced by glacial action tend to come either from
the French Alps where mountain glaciers still survive, such as
arête for a sharp ridge between adjacent hanging valleys, or from
the mountains of Great Britain (where the last major glaciers
disappeared about 12,000 years ago) such as cwm (Welsh) or corrie
(anglicized from Scots Gaelic) for the semi-circular head of a
hanging valley, also called a cirque (French).
Erosion by rivers and streams

Beyond the sculpting of individual rocks and boulders that
I have already referred to, wind is not often a sufficiently
powerful erosive agent to have much effect on the landscape.
In ice-free regions, it is liquid water that is the key player. Even
regions with exceptionally arid climates suffer the occasional
downpour. When this happens, valleys that may have been dry
for a century become occupied by a torrential stream that may
last only for a day or so.
Unlike glaciers, rivers are nowhere near deep enough to fill their
valleys. Therefore they erode more at the base than the sides,
which explains the typical V-shaped cross-sectional profile of
the upper reaches of a river valley (Figure 8.3). Rivers tend to
flow less straight than glaciers, so when a valley is carved by a
river it is likely to be sinuous, with spurs of the hills on either
side overlapping. This means you cannot usually stand on the
floor of a river valley and see straight along it as you can in a
glacial valley. If a glacier forms in a valley previously cut by
a river, it will tend to straighten the valley by truncating the
overlapping spurs.
148

Figure 8.3 A landscape sculpted by the action of rivers and streams. If
the main valley were to become deepened and widened by a glacier, the
landscape could end up like that shown in Figure 8.2.
Key idea: Rivers versus glaciers

Glaciers sculpt the landscape in ways very different from rivers.
Drainage patterns
In detail, how a landscape is eroded depends as much on
the nature of the bedrock as on the agent of erosion. This is
particularly seen in the erosive effects of water. If the rock is
limestone, which is soluble, rainwater may drain through a
system of fissures and caves, and there may be no streams at
the surface for most of the year. If the rock was laid down in
a succession of sedimentary beds, the softer units will be worn
away most rapidly. If there is a strong pattern of faulting or
jointing running through the rock, rivers and streams will tend
to exploit these lines of weakness. The nature of the bedrock,
and its geological structure, can therefore often be inferred
from a map showing the drainage pattern (Figures 8.4 and 8.5).
Catastrophic floods released by retreating glaciers create a valley
pattern at odds with ‘steady-state’ conditions; the channelled

(a) (b)
(c) (d)
Figure 8.4 Varieties of drainage pattern: (a) dendritic, which develops
where the bedrock is fairly homogeneous; (b) a dendritic pattern offset
by recent fault motion – in this case movement on an east–west fault has
offset the drainage in the northern region to the left. If the fault remains
locked for long enough, the drainage pattern will readjust; (c) radial
drainage, developed over a central high point, such as a domal uplift or a
volcano; (d) rectangular drainage, which exploits planes of weakness such
as joints or softer beds of rock.
Figure 8.5 Block diagram showing how the rectangular drainage pattern
in Figure 8.4(d) can be related to the underlying geology. The structure is
simple, with alternating harder (shaded) and softer beds dipping to the
right. Streams develop along valleys where the softer rock has been worn
away at the surface, and occasionally cut gorges at right angles to these,
through the intervening ridges of harder rock.
150

scablands referred to above are crossed by flat-bottomed,
near-vertically sided canyons up to 200 m deep, carved into the
Columbia River flood basalts.
Key idea: Drainage patterns

Drainage patterns can give important clues to the nature of the
bedrock geology.
Marine erosion
On coastlines exposed to storms, waves can be a very powerful
erosive agent, and can cut the coastline back into a cliff. Erosion
is strongest near the level where waves are breaking, so it
does not cut down much below the intertidal level. A wave-
cut platform of planed-off rock is often exposed at low tide,
extending seawards for 100 m or more from the foot of the cliff.
There is no comparable limit to the height above sea level of the
top of a cliff, this being controlled solely by the height of the
land surface.
Spotlight: Exploring a wave-cut platform

If you visit a tidal coast, a wave-cut platform at low tide can make an
enthralling setting to explore the well-exposed rock sequences and
to search for any fossils within – but keep safe by staying away from
the cliff and being careful not to be cut off by the incoming tide.
Where the rock is a sediment composed of gently dipping

layers, a cliff is liable to be undermined, particularly if the
beds at the foot of the cliff are relatively soft rock. In such
circumstances it is common to find caves at the cliff foot. Faults
or joints can be exploited, too, and these often control the
locations of coves and bays.
As the coast retreats because of this sort of erosion, a headland
that once separated two bays may become a steep-sided island
called a sea stack. Some such islands inherit a cave running
through them and are known as sea arches.

Figure 8.6 illustrates successive stages of coastal erosion.
Coastal retreat on this scale takes thousands to hundreds of
thousands of years, depending on the strength of the rock and
the violence of the storm waves. In Figure 8.6 the coastline is
likely to be marked by cliffs, except within the bays, which are
sheltered from the strongest wave action and where sand is
likely to accumulate.
sea
weak zones
1 km
(a) (b)
(c) (d)
Figure 8.6 Maps of successive stages (a–d) of coastal erosion. In (a)
the coastline has weak zones running into it. These may be faults,
closely spaced joints or beds of softer than usual rock. In (b) and (c)
marine erosion has eaten away at the weak zones fastest, resulting in
a coastline of headlands and bays. In (d) the headlands have become
small islands.
You will consider the circumstances under which sediment can

be deposited, and the sort of rocks which these can eventually
form, in the next chapter.
Key idea: Marine erosion

Coastal erosion by storm waves is more effective on weaker rocks,
resulting in bays and headlands.
152

Dig deeper
Erosion by the river Conwy is documented in this two-minute
video from the BBC: www.bbc.co.uk/programmes/p00xq904
A place to starting learning about glaciers is http://nsidc.org/
cryosphere/glaciers
If you want to learn more about drainage patterns, this is
a good starting point: http://www.physicalgeography.net/
fundamentals/10aa.html
Erosion and transport of grains is explored in a series of short
videos here: http://www.open.edu/openlearn/science-maths-
technology/science/across-the-sciences/exploring-sedimentary-
processes

Fact-check
1 What does ‘hydrolysis’ mean?
a Rainfall
b Breaking apart by water
c Transport by liquid water
d Transport by liquid water or ice
2 What tends to happen to a quartz grain during transport by

a river?
a It dissolves
b It breaks down to clay particles
c It shatters along cleavage planes
d Its corners become worn away
3 A grain of sediment that is ‘saltating’ is…?

a Being transported in the suspension-load
b Being rolled along the bed of a flow
c Being bounced along by a flow
d Becoming coated in salt
4 Which of these is an incorrect description of a typical valley that

was carved by a river compared to one that was carved by a
glacier?
a Straighter
b More V-shaped in cross-section
c Fewer lakes
d Lacks ‘hanging valleys’ to the sides
5 Which of these statements is incorrect?

a Marine erosion on a coastline is strongest where waves are
breaking
b The drainage pattern developed over homogeneous bedrock
tends to be dendritic
c A lump of rock transported inside a glacier will soon be
ground down to powder
d Dissolved silica is an end-product of hydrolysis of feldspar
154

Deposition of
9
sediment and
the origin of
sedimentary rocks
33how beds of sediment often contain
information about the currents that
deposited the grains
33about the types of sediment that
accumulate in different environments
33how sediment can be turned into rock.
155

The best farmland in most countries is typically in regions where
rock is being (or has been) deposited in the form of sediments.
These tend to be low-lying areas of land prone to flooding by
rivers. Much more sediment is deposited, unseen, below the sea.
Sediments tend to be deposited in layers, known as beds.
After sediments have hardened into sedimentary rock, this
layering makes them among the most easily worked building
stones. They are sources of certain vital resources, such as coal,
and may act as underground reservoirs for water, oil or gas.
Sediments can accumulate to great thicknesses; for example,
the Grand Canyon of the Colorado River in Arizona has cut
through 2 km of sedimentary rocks, which were deposited in
several episodes between about 550 and 250 million years ago.
Spotlight: Sediment versus sedimentary rock

In casual use these terms are sometimes used interchangeably,
but strictly speaking there is a difference between ‘sedimentary
rock’ and ‘sediment’. A deposit of sediment at or near the surface
is typically young and unconsolidated. The chances are that it will
be removed by erosion and therefore will not survive to become
part of the geological record. A sedimentary rock is typically
older, and is likely to have been buried at some time. It may have
become hardened as a result of processes described at the end of
this chapter.
You have already considered the processes involved in erosion

and transport of detrital grains. This chapter begins by looking
at what conditions are necessary for grains to be deposited.
Many of the most familiar sedimentary rocks are formed from
accumulations of detrital material. These are known as clastic
rocks (clast being a term used to describe any fragment or
grain). If the clasts are predominantly of silicate minerals, as is
usually the case, the rock can be described as ‘siliciclastic’.
Clastic sedimentary rocks are classified by the diameter of their
most abundant clasts according to the scheme in Table Y of
Appendix 2. You should refer to this if you are concerned to
know the exact size distinctions that are drawn between clay
particles, silt, sand, gravel and so on.
156

Remember this!
Non-clastic sedimentary rocks include evaporites and some
varieties of limestone. (These are discussed later in the chapter.)
Most deposits of sediment contain only a limited range of clast

sizes. The reason for this is that a current of wind or water that
is just weak enough to deposit bed-load clasts of one particular
size will not be flowing strongly enough to bring in any clasts
that are much coarser than this. However, it will be flowing
with sufficient speed and turbulence to keep all the significantly
finer clasts in suspension, so those will be carried away in the
suspension-load and will not be deposited in the same place.
Ripples and dunes

Except in very quiet conditions, when fine silt and mud can
settle out from the suspension-load, clastic sediments tend to
form by deposition from the bed-load. The rolling and bouncing
(saltation) modes of transport in the bed-load give rise to some
distinctive features in the bed of the flow, which can very often
be recognized in ancient deposits. Everyone is aware that wind
tends to pile sand into large dunes. Water flow produces similar
(but usually smaller) sedimentary structures, of which ripples
are the best known. Ripples, dunes and the like are referred to
collectively as bedforms.
Many ripples and dunes are characterized by asymmetric cross-
sections. Each has a gentle slope on its upstream side and a
steeper slope on its downstream (or lee) side. Grains bounce or
roll up the gentle upstream slope, which is exposed to the full
force of the current. Some grains may be picked up (eroded)
from this upstream slope, to join those arriving from farther
upstream in the bed-load. As grains topple over the crest they
find themselves in a more sheltered region, where the current
is too weak to transport them, so they simply drop on to the
upper part of the lee side. This face is continually being over-
steepened by the arrival of sediment at its top. Every time it
gets too steep to be stable, an avalanche of grains slips down
the face. Each avalanche forms a layer resting on this slip-face
9 Deposition of sediment and the origin of sedimentary rocks 157

at the maximum angle that can be supported, which is around
30° to 40°. Grains are thus removed from the up-current side
of the bedform and deposited on its lee side, and so the ripple
or dune migrates slowly downstream. Bedforms stand a chance
of eventually being buried, and therefore being preserved in the
geological record, only if the net rate of erosion of grains from
upstream slopes is less than the net rate of deposition of grains
on downstream slopes.
Key idea: Preservation of sediments

A deposit of sediment is not necessarily destined to become
sedimentary rock, because it may be eroded away before this
can happen.
Figure 9.1 shows how bedforms like ripples and dunes migrate.
If there is no net deposition of sediment, all that happens is that
bedforms migrate downstream as grains are eroded from the
upstream face of a bedform and deposited on its downstream
slip-face. However, if more sediment arrives in the area than
can be transported away, then there is net deposition. Successive
bedforms migrate up the only partially eroded upstream faces of
previous ones, and a deposit of sediment accumulates.
The successive slip-faces within each bedform can often be
recognized if the deposit is subsequently excavated and seen
in cross-section, even after hundreds of millions of years
(Figure 9.2). Slip-faces dip more steeply than the deposit as a
whole, and the contrast in dip amounts gives rise to a feature
commonly called cross-bedding (Plate 7).
If you notice cross-bedding (sometimes called cross-stratification)
when you are looking at rocks in the field, this is a strong
indication that you are dealing with a clastic sediment. However,
cross-bedding can occur in pyroclastic flows, too.
The shapes of bedforms seen from above depend on the
conditions under which they form. A relatively slow flow in a
constant direction produces crests aligned at 90° to the current,
resulting in transverse ripples or transverse dunes. As the flow
158

1 flow 2 3
(a)
1 flow 2 3
4 5 6
(b)
Figure 9.1 Migration of ripples (centimetres in size) or dunes (tens or
hundreds of metres in size) seen in cross-section. In (a) a single ripple
or dune is shown, in a situation where there is no net accumulation of
sediment. The three views are at successive times, with the instantaneous
surface shown in the black solid line, and previous surfaces in the faint
grey lines. In (b) there is net accumulation, so that each bedform climbs
up the back of the previous one as it migrates. 1 to 5 are successive time
steps, and 6 is much later.
Figure 9.2 The face of this sand quarry in County Durham, England, shows
cross-bedding formed in desert sand dunes about 250 million years ago.

(a)
(b)
Figure 9.3 Two sorts of cross-bedding, both produced under conditions
of net accumulation of sediment. Each diagram is a block view, cut on
flat horizontal and vertical faces to show how the cross-bedding appears
on different planes. The actual surface, showing the three-dimensional
shape of the bedforms, is the left-hand part of the top of each view:
(a) transverse ripples or dunes; (b) crescent-shaped ripples or dunes. The
cross-bedding is different in each case: (a) has tabular cross-bedding;
(b) has trough cross-bedding. It would be difficult to distinguish between
these if the only cross-section you saw was looking at right angles to the
current direction (the left-hand face), but the distinction is clear when
looking along the current direction (the right-hand face).
160

becomes stronger, crests become wavier, and may separate into
individual crescent shapes. These contrasting bedforms are
associated with different patterns of cross-bedding that allow
them to be distinguished when seen in cross-section (as shown
in Figure 9.3). Observation of the nature and orientation of
cross-bedding can therefore provide very important clues about
the environment where an ancient sediment was deposited.
There are many variations on the basic ripple and dune
theme. Crescent-shaped wind-blown sand dunes are usually
isolated from one another, rather than touching at their tips
as suggested in Figure 9.3(b). These are called barchans. In
some areas dunes develop with their crests aligned parallel
to the dominant wind direction; these are known as seifs or
longitudinal dunes. They have slip-faces on both sides, usually
attributed to corkscrew-like vortices in the wind as it is
funnelled between the dunes. The largest dunes are hundreds
of metres high and many kilometres in length. Clearly, these
are major constructions and they are not going to re-orient
themselves every time the wind changes direction!
Most wind-blown dunes migrate by about a metre per year. The
shape of a dune, and the internal cross-bedding it inherits, are
influenced not only by the annual prevailing wind but also by
seasonal winds blowing from other directions. The surfaces of
sand dunes are often covered in transverse ripples, and these are
small enough to adjust in response to hourly variations in wind
strength or direction.
Key idea: Ripples and dunes

Ripples and dunes are bedforms resulting from deposition from the
bed-load during water or wind transport or from pyroclastic flows.
In a pyroclastic flow, despite the high speed of the flow, the high
ratio of the pyroclastic fragments to the volume of air entrained
within the flow may be so great that clasts are plastered on to
the ground, even when the speed of the flow would otherwise be
great enough to keep them moving in the bed-load. Pyroclastic
flow deposits, and also fallout ash, can be considered as clastic
sediments even though they have a volcanic origin.

Turbidity currents
Subaqueous analogues to dense pyroclastic flows, known as
turbidity currents, are particularly important for transporting
sediments off the edge of the continental shelf and on to the
ocean floor, notably into trenches. These can be triggered when
a submarine landslide stirs a lot of sediment into suspension.
This mixture is considerably denser than the surrounding
seawater, and it flows downslope as a turbulent mass of
sediment and entrained water.
Turbidity currents can be very powerful, and have been known
to snap submarine telephone cables. Initially, a turbidity current
is usually strong enough to scour up extra material at its base,
but as the slope eases it slows down and begins to deposit
its load. At first, deposition is by plastering down of clasts,
as can happen in pyroclastic flows, but as the speed wanes
and the proportion of sediment remaining in the turbidity
current decreases, ordinary climbing ripples form. These can
be distinguished from ripples formed in more familiar water
currents only by the context in which they are found. After a
turbidity current has passed by, the fine-grained material that
had been stirred up into suspension settles out, so the sandy
rippled part of the deposit is overlain by finely laminated silts
and muds. A deposit laid down by a turbidity current is called
a turbidite.
Key idea: Turbidites

Turbidites, deposits laid down by a turbidity current, are underwater
analogues to dense pyroclastic flows.
Current directions
In rivers, the direction of the current is rarely as fickle as the
winds. It is essentially uniform, even though the strength of
the current may vary seasonally. The consistency of current
direction controls the orientation of the cross-bedding, so that
unidirectional current flow can be recognized in ancient river
deposits.
162

alternating current 5 cm
directions
(a) (b)
Figure 9.4 Cross-sections showing ripple cross-bedding produced by
alternating tidal currents. In (a) the top of each layer of ripples has been
planed off (eroded) by the current that brought in the sediment that
formed the next layer of ripples. In (b) there is a quiet interval between
each reversal of the tide, allowing mud or silt (black) to accumulate in the
hollows between ripples. This is called ‘flaser bedding’.
In the sea, tidal currents change in both strength and direction,

usually on a twice-daily basis. In sediments deposited in shallow
marine conditions, where tidal currents are most obvious, this is
usually reflected by alternating directions of cross-bedding, and
sometimes by thin layers or lenses of fine-grained suspension-
load that were able to settle during the period of slack water
while the tide was turning (Figure 9.4).
Key idea: Current directions

Cross-bedded sediments laid down in a river show evidence of
unidirectional flow, whereas shallow marine deposits may show
evidence of alternating tidal currents.
Waves
At sea or in large lakes, wave motion can disturb the sediment if
water depth is less than about ten times the height of the waves.
Bedforms may grow in the shape of symmetric ripples whose
opposite faces have equal steepness (Figure 9.5), in contrast to
the asymmetric ripples produced by current flow. If a current
is flowing in a totally different direction from the alignment of
wind-generated waves, simultaneous current ripples and wave

2 cm
Figure 9.5 Cross-sections through a ripple produced in shallow water as a

result of wave motion.
ripples with competing orientations may result in a criss-cross

pattern of interference ripples. These can often be seen exposed
on modern beaches at low tide. During storms, wave action
may be effective at many tens of metres depth. Particularly
violent storms are thought to be the cause of what is known
as hummocky cross-stratification, when hummocks form that
are a few metres across but only a few centimetres high, with
correspondingly low-angle cross-bedding within them. When
waves break on a beach, they tend to plaster sand down in
successive layers, with little or no cross-bedding.
Remember this!
Wave motion can disturb sediment on the seabed at a depth up to
about ten times greater than the height of the waves themselves.
Finer and coarser sediments

Ripples and dunes are usually found only in sand or the coarsest
silty sediments. Fine silt or mud will not settle out from water
unless it is flowing at less than a few centimetres per second,
which is too slow to sculpt sediment into bedforms. Silt and
mud thus tend to accumulate in fine horizontal layers, referred
to as laminations.
Conversely, when gravel and pebbles are deposited by flowing
water, it is usually because the current has dropped suddenly,
which leaves too little time for bedforms to develop. However,
pebbles do tend to be oriented by the current and come to rest
with their tips pointing downward into the current. Each pebble
164

thus overlaps its neighbour, in a fashion described as imbrication
(Figure 9.6). This can be a useful indicator of current direction.
current
current 5-50 cm
5–50 cm
Figure 9.6 Cross-section showing imbricated pebbles deposited by fast-

flowing water. The tip of each pebble points downward and upstream.
Remember this!
It is virtually unknown to find an ancient deposit containing nothing
but pebble-sized clasts, because finer material (sand or mud) is
likely to be washed in later, filtering its way down through the gaps.
Other means of transport

You have seen previously that material may also be transported
by mudflows (lahars), by landslides and by moving ice. After
this material has been deposited, it is generally distinguishable
from a water-lain deposit because of its wider range of clast
sizes, the angular, poorly rounded shapes of the clasts, and the
lack of bedforms. Another distinguishing feature is that pebble-
sized clasts may be supported within a matrix of mud or sand,
whereas in a water-lain deposit pebbles almost always touch.
Deposition from mudflows and landslides occurs simply when
the flow stops moving. Deposition from a glacier happens as
the ice melts, usually at its downhill end, or ‘snout’. Often,
glacial detritus is reworked by streams of meltwater, in which
case it loses all or most of the characteristics listed above.
Debris surviving more or less where it was dropped by a glacier,
without being reworked, is referred to as moraine. When the
melting snout of a glacier has retreated up a valley in stages,
each successive position of the snout is marked by an accurate
ridge of moraine.

Key fact: Landslides etc.
Landslide deposits, mudflows and glacial deposits lack cross-
bedding (unless reworked by water or wind) and have wider ranges
of grain size than most other clastic sediments.
Using sediments to interpret

ancient environments
Most sedimentary rock was deposited under water. You
have already learned of the separate processes involved and
the bedforms that can be created. However, to interpret the
environment in which an ancient deposit was laid down, it is
rarely sufficient to rely on a narrowly focused approach. Usually,
it is better to weigh up many attributes of the deposit, and thereby
to characterize its facies as fully as possible, in order to build up a
picture using as many lines of evidence as you can gather.
Spotlight: In-depth interpretation

It would not be very informative if all you could say by way of
interpretation of a sedimentary rock was that it had been formed
under water. As an analogy, imagine you are a CSI (crime scene
investigation) forensic scientist, called to the scene where a dead
body has been found. The police would be unimpressed if you
merely reported that the victim had been shot. They would expect
you to work out what sort of gun had been used, and from what
distance and direction it had been fired. They would also want to
know how long the body had lain there before discovery, whether
it was found in the place of death or had been dumped there later,
whether or not there were signs of a struggle, and so on. The aims
of a geologist interpreting sedimentary rock are much the same.
To see how the evidence preserved in the sedimentary record can

be pieced together to interpret past environments, I will describe
a few key modern depositional environments. By establishing the
ranges of bedform, grain size, composition and texture, and their
lateral and vertical variations and extent in modern examples,
you can understand what to look for in ancient deposits.
166

Case study: The key to the past?
Learning from modern-day environments in order to understand
the geological record is a principle well known among geologists,
which can be summed up as ‘the present is the key to the past’, an
expression attributed to the Scotsman Sir Charles Lyell (1797–
1875), who was one of the founders of modern geology.
Lyell’s maxim was a central theme of his hugely influential book

Principles of Geology (1830), which championed the ‘uniformitarian’
view of Earth history. This holds that the processes sculpting the
planet today have operated continuously throughout geologic time,
in contrast to the view of the opposing school of ‘catastrophists’.
Lyell was essentially correct, although the present is not an infallible

guide to everything in the past. For example, nothing like channelled
scablands (Chapter 8) is forming today. The idea these had been
formed by a catastrophic process was proposed in the 1920s by local
geologists J. Harlen Bretz (1882–1981) and Joseph Pardee (1871–
1960), but it took the geological community 50 years to overcome its
uniformitarian mind set enough to accept this. Bretz was honoured
for his insight by the Geological Society of America’s highest award,
the Penrose Medal, in 1979 at the age of 96. He remarked to his son,
‘All my enemies are dead, so I have no one to gloat over.’
Key idea
The present is the key to the past’ is a good, but not infallible,
guide to interpreting a succession of sedimentary rocks.
Depositional environments
associated with rivers
Let’s begin by looking at rivers in their sluggish downstream
reaches, where they are depositing rather than eroding
material. Unless it has been straightened artificially, a lowland
river usually winds in a lazy way across its floodplain. This
floodplain consists mostly of muds and silts deposited as
flood waters recede, after an event when the rate of water

supply from upstream was so great that the flow could not
be contained within the channel. While a flood is abating, the
water covering the floodplain is sufficiently stagnant for much
of the suspension-load of silt and clay particles to settle out,
although normally (when the river is confined to its channel) the
suspension-load would be flushed right down the river until it
reached the sea or a lake. A floodplain is thus built of successive
layers of fine- or very fine-grained flood deposits, which provide
some of the globe’s richest agricultural land. On the other hand,
it is a foolish place to build houses because after exceptional
rainfall it is vulnerable to flooding, unless there are expensive
and well-maintained ‘flood defences’.
Slightly more sediment is deposited on the floodplain right next
to the river, and in this way riverbanks are built up, in some cases
to heights of several metres. These raised banks, known as levées,
are a characteristic feature of rivers such as the Mississippi.
Key idea: Floodplains

When it floods, a meandering river deposits fine sediment on its
floodplain.
The river itself usually flows too fast for its suspension-load
to be deposited, and in some places is likely to be transporting
sand in the bed-load. You may therefore find ripples when
you look on the bed of a river, but these can be preserved in
the geological record only if they are buried by later deposits.
Burial of ripples in a winding lowland river is possible without
clogging up the river, because the river does not maintain its
course. The reason for this is that the current has to flow faster
round the outside of a bend than on the inside. The riverbank
on the outside of a bend is therefore subject to undercutting
and erosion. In contrast, the current is slowest near the inside
of bends, so bed-load tends to be deposited against the bank
there, producing a sand and gravel deposit known as a point-
bar. Consequently, bends (or meanders) become progressively
exaggerated, as shown in Figure 9.7. A meander loop may
eventually become cut off, allowing the main course of the river
to straighten again, but another meander soon develops. The net
168

effect over tens of thousands of years is for the river to migrate
from side to side across its flood plain.
(a) (b) (c)
Figure 9.7 Map view at successive times showing how a meandering river
migrates across its flood plain, by erosion on the outside of bends and
deposition on the inside. In (c) a loop has been almost cut off, forming an
oxbow lake that will eventually silt up. X–Y indicates the cross-section
shown in Figure 9.8.
The kind of thing a geologist would hope to see in a cliff

cut into an ancient floodplain deposit can be pictured by
constructing a cross-section across a modern environment, such
as Figure 9.8. Most of the deposit is muds and fine silts, laid
down on the surface of the floodplain. Abandoned channels
become filled with laminated silts and muds, looking rather
similar to floodplain deposits, except that they will have steep
edges marking the sides of the channel where it was cut into the
abandonedchannel
abandonned channel active channel
levée levée
X Y
Key
flood plain deposits (mostly mud)
abandoned channel fill (laminated mud and silt)
point-bar deposits (cross-bedded sand)
Figure 9.8 Cross-section along the line X–Y in Figure 9.7, showing
lateral and vertical variations in the nature of the deposits. The channels
at ground level correspond to those seen in the map view. The buried
channels mark the course of the river at earlier times. These were
abandoned and then buried by floodplain deposits.

floodplain. Successive positions of active channels are marked
by point-bar accumulations of sand and gravel deposited on
the inside of bends. The extent of each such sand/gravel sheet
depends on how far the bend migrated before the channel
was abandoned. Within each point-bar deposit, the grain size
becomes finer upwards: with coarse sand or gravel at deeper
levels representing a faster current; and finer sand towards the
top where the current was slowest. The detailed structure of
ripple cross-bedding indicates the direction of the current.
Sediments laid down by a meandering river thus consist
predominantly of very fine-grained material (mud and silt), most
of which was laid down on the floodplain. However, the sheet-
like or lens-like bodies of gravel and sand within it provide the
evidence for migrating channels.
Not all rivers have this sort of floodplain. Downstream from the
ends of glaciers, or in regions where rainfall is strongly seasonal,
rivers have to carry a particularly high sediment-load at certain
times of the year. Typically, such a river occupies several
interconnecting channels (rather than being confined within a
single channel), separated by bars of gravel that are submerged
only when the river is in flood. This is known as a braided river
(Figure 9.9). Seen in cross-section, the deposits from a braided
river consist almost entirely of lenses of sand and gravel (Plate
8). Any mud is restricted to the occasional abandoned channel,
so it is relatively easy to distinguish between deposits from
meandering and braided rivers.
Figure 9.9 Map view of a braided river, which may be 100 m to about 10
km in width. Braided rivers are common in Texas (the Brazos is a classic
example) and other areas where discharge is strongly seasonal.
170

Flash floods
The mountainous fringes of desert areas experience river flow
that is even more strongly episodic. Here, channels may be dry
for almost the entire year, or even for decades. However, when
it does rain it may do so very heavily, and gullies that are almost
always dry may suddenly experience a flash flood, during which
they rapidly become brim-full with water, flowing fast enough
to roll even quite large boulders along as bed-load.
Spotlight: Flash floods

A flash flood is one of the chief hazards to beware of if you do
fieldwork in mountainous terrain. When I was a student doing the
mapping for my PhD in Arabia I once had to relocate my poorly
chosen campsite in a hurry after dark when heavy rain set in.
Where a gulley debouches on to the plains, the torrent is no

longer confined and so it disperses. This causes the speed of flow
to drop dramatically, so the larger clasts in the bed-load come
to rest. The deposit spreads out radially from the mouth of the
gully, to form an alluvial fan.
Seen in cross-section, an alluvial fan consists of successive
layers of imbricated boulders and pebbles, with sand washed in
between later (Plate 9). Channels within it may not be obvious
but, with care, erosion surfaces can usually be distinguished at
the base of each fresh influx of sediment.
Deltas and estuaries

In places where rivers reach the sea, you often find a delta
made of sediment deposited by the river. There are two factors
contributing to this.
33 The first is that, just as in an alluvial fan, the current slows
down where the flow spreads out, so any bed-load that
the river has managed to bring all the way to the sea will
become deposited.
33 The second is a result of something that happens to clay
particles as soon as the water becomes salty. Clay particles

are tiny and remain in suspension provided they stay as
individual particles. However, once they get into even slightly
salty water they start to stick together, or flocculate. The
clusters of clay particles created in this way are big enough
to sink, and so the suspension-load of the river is dumped as
soon as the river water mixes with the seawater.
Our planet’s two most famous deltas show some important
differences in morphology, resulting from differing local conditions
(Figure 9.10). The archetypal delta is that of the Nile. The likeness
between the shape of this flat, many-channelled tract of land to
the Greek capital letter delta, Δ, led the historian Herodotus to
coin the term delta to describe it in 490 bce. The Nile breaks into
a plethora of distributory channels, radiating outward from the
vicinity of Cairo, more than 100 km from the sea. The seaward
fringe of the delta is a gentle arc, rather than the straight edge
implied by the Greek character. In contrast, the Mississippi delta
has been likened to a bird’s foot in appearance, because the
levée-confined channel that feeds it has built out beyond the main
shoreline and it and the few distributory channels coming from it
stick out into the sea like the toes on a foot.
Figure 9.10 Views from space of two famous deltas. On the left is a
200-km-wide view of the classic delta of the river Nile, which flows from
south to north. Sea is dark, desert is pale, and the well-vegetated land on
the delta is mid-grey. On the right is an 80-km-wide view of the ‘bird’s foot’
delta of the Mississippi, which flows from north-west to south-east. Deep
water is dark, shallow water is light, and the complexly shaped land is in
mid-tones.
The main reasons for the different shapes of these two deltas are
that the Mississippi discharges much more sediment than the
172

Nile, whereas the Nile delta is exposed to stronger waves, which
tend to mould a delta front into an arc.
Neither the Nile delta nor the Mississippi delta is subject to
strong tides. In situations where there is a large tidal range
(for example at the mouth of the Ganges-Brahmaputra or the
Mekong), tidal currents scouring up and down the distributary
channels flush a lot of the sediment seawards and keep the
channels wide. Other river mouths lack any kind of delta, but
are funnel-shaped inlets of the sea known as estuaries. Most
of the river mouths around the British Isles and Australia, and
along the east coast of North America from Chesapeake Bay
northward are of this kind. Factors encouraging the presence
of an estuary rather than a delta include relatively low amounts
of sediment discharge and high tidal ranges. Some estuaries are
valleys that have been partly submerged by the rise in sea level
that happened as a result of the melting of continental ice sheets
at the end of the most recent glaciation.
In many ways, the tops of deltas resemble floodplains of
meandering rivers. The deposits found there consist of ripple
cross-bedded sand that accumulated on point-bars, silts and muds
laid down in abandoned channels, and mud deposited over the
delta plain during floods. Delta plains are often densely colonized
by vegetation; and roots, fallen leaves and branches can often
be found as fossils. Decaying vegetation may accumulate thickly
enough to form coal deposits (as you will see in Chapter 11).
Moving seawards from a modern delta, the sandy sediment
deposited near the shoreline may be reworked by wave action
(which is reflected in the style of cross-bedding). This area
is referred to as the delta bar. Channels are largely absent
here, and so is mud because such shallow water is usually
too agitated for clay to settle, even when it has flocculated.
Another diagnostic feature is broken or complete shells of
marine organisms, which are, of course, absent in any sediment
deposited higher up the river. Going slightly further offshore, to
the delta front, you reach the zone where most of the flocculated
clay is able to settle. This region is characterized by unbroken
shells of marine organisms, and burrows in which these and
soft-bodied organisms lived.

Key idea: Deltas and estuaries
Where a river enters the sea, the shape of any delta or estuary is a
result of interplay between the rate of sediment discharge by the
river and waves and tides in the sea.
Case study: Prograding shorelines and

Walther’s Law
Now I can introduce you to a very important concept. By their
very nature, because sediment is continually being brought there,
deltas tend to build out seawards over time. A shoreline advancing
seawards like this is said to be prograding. As a result, if you follow
events at a specific point (originally a few kilometres offshore) over
maybe 10 000 years you might record the following succession.
1 Purely marine deposits are encroached upon by delta-front
sediments. At first these will be the very finest clay material,
which has been washed farthest from the delta, but as the delta
draws nearer coarser clay and silt are deposited here.
2 Throughout this time the seabed is building up higher, because
of the accumulation of sediment.
3 Eventually you will find yourself on the shoreline, where the
sands of the delta bar are being deposited.
4 The shoreline continues to prograde seawards, until you are in
a distributary channel of the delta, whose base may erode away
some of the underlying delta bar sands.
5 After a while the channel migrates away, and you find yourself on
the delta plain, though there is always a chance that a channel
may migrate your way again if you wait long enough.
If you lack the leisure to wait 10 000 years, you could see the same
variety of environments by moving sideways a distance of ten or so
kilometres from offshore to delta top. You could also see a similar
succession of environmental changes by drilling a 100-m-long
vertical borehole through the top of a modern delta. This would
reveal delta plain sediments at the top, delta bar sediments
below, and progressively finer delta-front sediments below these
(Figure 9.11). In other words, the lateral juxtaposition of sedimentary
174

environments at an instant in time is reflected by environmental
changes seen in a vertical succession (representing the same
location over a period of time).
This relationship between lateral environmental zonation at an

instant in time and environmental changes over time at a single
location is known to geologists as Walther’s Law, after the German
sedimentologist Johannes Walther (1860–1937) who first stated it
in 1894.
Figure 9.11 shows the delta plain sediments overlain by a new set
of offshore marine sediments. This might surprise you because,
continuing our imaginary sojourn on a delta, you might expect that
eventually you would find yourself well and truly on dry land as
the shoreline continues to prograde. This may be the long-term
tendency, but it is quite likely that you will see several returns to
Offshore marine
Coal
Delta plain laminated silts and muds
Distributary channels, point-bar sands
Delta bar, sands reworked by waves
Delta front, laminated sand, silt

and mud. Shells and burrows of marine
organisms becoming common
towards the base
Offshore marine
Figure 9.11 Changing grain size and bedforms in a vertical succession
through a prograding delta. The width of the column is drawn to
correspond with changes in grain size. The basic succession, fine grained
(marine) at the bottom, becoming gradually coarser upwards until topped
by non-marine muds, is characteristic of deltas. The relative contributions
of waves, tides, river flow and sediment supply influence details in this
general picture.

marine conditions in the meantime. This is because, if your river
bursts through one of its levées near the upstream part of the delta,
the main discharge may be diverted into a new channel, in which
case a new area of delta will build out into a different area of the
sea. This will starve the sediment supply to your observation point,
which will slowly sink below sea level as a result of compaction of
the underlying sediments, and downward isostatic adjustment of
the crust as a whole because of the weight of sediment.
To take the example of the intensively studied Mississippi delta,

studies have shown that the present-day main channel (seen
in Figure 9.10), flowing south-eastwards to the sea from New
Orleans, came into existence only about 1000 years ago. For most
of the previous 5000 years, the main discharge went southwards
from an area near Baton Rouge to produce a (today mostly sunken)
delta to the west of the present one. The same period also saw
an interlude when the main discharge was eastwards from New
Orleans, producing a delta to the east of the present one (part of
which is still above sea level).
Logging a sedimentary succession

Geologists rarely have the luxury of cliffs that are high enough
and continuous enough, and running in the correct orientation,
to record the whole lateral extent of an ancient delta. However,
by interpreting the environmental changes seen in a single
quarry face, or in a core recovered from a borehole, or in a
series of isolated exposures, they can construct a log such as
that in Figure 9.11 and then try to use Walther’s Law to deduce
what the whole area would have looked like at the time when
the sediments were being deposited.
Beware of the technical use of ‘sequence’ and ‘succession’. An
expert geologist will distinguish ‘succession’, denoting rock
units that succeed each other in chronological order, versus
‘sequence’, to denote a succession that is uninterrupted by
significant time-gaps or erosion surfaces. Walther’s Law works
best within an individual sequence.
176

Beaches and other coastal
environments
The concepts underpinning Walther’s Law are applicable not
only to deltas; they work whenever there have been gradual
changes in sedimentary environment. Let’s take a look at
depositional coastlines.
Case study: On the beach

A sedimentary succession deposited over time at a place where
the shoreline was prograding (because of accumulation of sand
washed on to a beach) would begin at its base with marine
deposits characteristic of deepish water, below the level exposed
at low tide and too deep for storm waves to disturb. These
sediments, deposited in conditions of unvarying low energy, would
be very fine grained, probably either mud or limestone (discussed
shortly). The slow accumulation of this fine-grained deposit would
eventually bring the sea floor close enough to the sea surface for
storm waves to disturb it, and you might expect to see erosion
surfaces and sandy layers, possibly containing hummocky cross-
stratification representing big storms.
As the sea floor became higher and the general environmental

conditions became more energetic (with stronger tidal currents
and disturbance by waves at every low tide), mud or limestone
would be replaced by silts and fine sands. Such conditions are
favoured habitats for many burrowing organisms that feed off the
continual supply of nutrients brought in by the currents. Cross-
bedding, such as wave-generated ripples, may be disturbed or
obliterated by burrows. You might also find complete fossilized
shells or organisms preserved within their burrows.
Continuing up the succession, you would reach sand deposited on
the beach, between the low- and high-tide marks. This is generally
laid down in low-angle sheets with occasional ripple cross-bedding.
There would probably still be plenty of shelly material, but now
it would be mostly broken fragments. Sand on a modern beach
is usually unlike river sand because, although quartz may be
the dominant mineral, a large proportion of the sand grains are

actually fragments of broken shells and so are made of calcium
carbonate. If your beach is the sort with sand dunes behind it, the
top of the succession would contain the large-scale cross-bedding
characteristic of wind-blown dunes. Because the sand had not been
blown very far, the grains would be unlikely to have become very
well rounded and frosted, so would not resemble the grains typical
of desert sand dunes. Eventually, beach dunes become colonized by
plants, and so you might find a few rootlets in your deposit.
The succession of sediments just described would be most likely

to be preserved in the geological record if it sank below sea level
and became buried below another cycle of prograding shoreline
sediments. Otherwise, it would be vulnerable to erosion by river
action, or by a change in marine conditions. For example, if the
currents started to take more sand away than they brought in, the
beach and eventually the dunes behind it would be washed away.
There are many variations on the theme of coastal sediment

deposition. If the tidal range is great and the gradient gentle,
waves break a long way offshore, causing them to lose most
of their energy before they lap up to the high-tide mark.
Particularly on sheltered coastlines, on the landward side of
a sand beach exposed at mid-tide there is often an extensive
mudflat consisting of very fine-grained material that has settled
out in quiet, wave-free conditions while the tide is turning.
Today in temperate regions, these tidal flats become colonized
by coarse grasses to form salt marshes. The grass roots bind
the sediment together, preventing it from washing away and
encouraging it to build up above the normal high-tide level. In
humid tropical regions, the same function is served by mangrove
trees, whose roots spread out above ground level and act as a
trap for sediment.
Evaporite deposits
On low-lying arid coasts there may be too little rainfall to
support salt marsh or mangrove swamp vegetation. Instead,
above the normal high-tide mark, mud is matted down by layers
of salt-tolerant fibrous algae. Such a salt flat is known by its
178

Arabic name of sabkha. The mud here is dominated not by clay
minerals but by mud-sized grains of carbonates, chiefly calcite,
CaCO3, and dolomite, CaMg(CO3)2.
There are biological ways of growing these carbonates,
discussed shortly, but on a sabkha they form inorganically.
The sun beating down on the surface of the sabkha causes
evaporation, and seawater is continually drawn in through
the mud to replace what is lost. As the water evaporates, its
dissolved constituents come out of solution, often beginning
with carbonate minerals. Minerals (and rocks) formed in this
way are described as evaporites. If the rate of evaporation is
sufficient, other evaporite minerals begin to grow, notably
gypsum (CaSO4.2H2O), anhydrite (CaSO4), halite (NaCl, which
is table salt), and ultimately salts of potassium.
Spotlight: Handy evaporites

Gypsum deposits make alabaster, much used by sculptors
because of the ease of carving and from which plaster of Paris is
made. Table salt, of course, has many uses and what you sprinkle
on your dinner may have come from an ocean that vanished long
before your fish was caught.
Evaporite minerals do not form only in sabkhas. They can also

form at sea, provided evaporation of seawater is proceeding
faster than the rate of replenishment by fresh water. These
conditions are met in enclosed marine basins in subtropical
latitudes where there is little rainfall. Here, the seabed becomes
covered by crystals of evaporite minerals that have grown
near the surface (where the effects of evaporation are most
extreme) and sunk to the bottom. There are no extensive marine
evaporite deposits forming today, but drilling has revealed up to
2 km of evaporites on the floor of the Mediterranean, deposited
between 5.4 and 4.8 million years ago when the Mediterranean
was periodically cut off from the Atlantic by the emergence of
the Straits of Gibraltar above sea level (Figure 9.12).
To produce such a thickness of evaporites, the Mediterranean
would need to have been completely evaporated about 40 times.

Figure 9.12 These large ‘swallow tail’ crystals of the evaporite mineral
gypsum, today exposed on the uplifted island of Cyprus, grew on the floor
of the desiccating Mediterranean Sea about 5 million years ago.
Even today, the rate of surface evaporation from the surface

of the Mediterranean Sea exceeds the rainfall and river influx.
Calculation shows that, if the Straits of Gibraltar became closed
(preventing replenishment by Atlantic water), the Mediterranean
would take only about 1000 years to dry out.
Economically important evaporite minerals, which formed in an
enclosed marine basin known as the Zechstein Sea about
220 million years ago, are found in a deposit stretching from
north-east England to northern Germany. These include
potassium salts and are the foundation of the British chemical
industries of Teesside.
Limestone
Evaporites form only under special circumstances, but calcium
carbonate is the main rock-forming mineral in many sedimentary
environments. The main reason for its ubiquity is biological,
rather than physical or chemical, processes. A rock made mostly
of calcium carbonate is called limestone, which is much the
most abundant non-silicate rock type. Deposition of calcium
carbonate is an important part of the rock cycle – calcium that
180

was dissolved during weathering and the bicarbonate by-product
of hydrolysis come out of solution in tandem, maybe thousands
of kilometres from where the calcium was dissolved. Calcium
carbonate is most commonly found as the mineral known as
calcite, but there is also a polymorph called aragonite.
Seashells found on the beach are made of calcium carbonate,
and some limestones are made of accumulations of such shells,
or of shell fragments. More important globally are microscopic
(generally single-celled) plants and animals that live in the sea.
Most drift around near the sea surface where there is plenty
of sunlight. These are referred to as plankton. Less abundant
forms live on the sea floor. The most important animal forms
are foraminiferans, which are simple organisms similar to the
amoeba except for having multi-chambered calcareous shells,
commonly up to about a millimetre across. The most important
plant variety is the coccolithophores. These are smaller and,
rather than having a simple shell, they are covered in an array of
disc-like plates called coccoliths, typically only 20 micrometres
in diameter, that separate after death.
When microscopic plankton like these die, their shelly remains
sink slowly towards the sea floor. If conditions are gentle
enough, they can settle to form extensive deposits. England’s
famous White Cliffs of Dover are cut into a 70-million-year-old
accumulation of coccoliths that was deposited on the floor of a
calm, shallow sea that had no major rivers discharging muddy
sediments into it (otherwise the cliffs would not be so white).
This rather special kind of limestone is called chalk.
Deposits of fine-grained calcium carbonate derived from the
shells of marine plankton cover much of the ocean floor in
places far from land where there is little other sediment supply.
It is thought that the removal of carbon dioxide into limestone
through the action of marine plankton is a natural mechanism
that maintains the concentration of carbon dioxide in the
atmosphere at a stable level. This may eventually help to soak up
much of the carbon dioxide being released into the atmosphere
through burning fossil fuels, and so act as a natural brake on
mankind’s additions to the greenhouse effect in which carbon
dioxide reduces the Earth’s ability to radiate heat in to space.

Tiny shells gradually dissolve as they sink, especially in cold
waters. In the deepest and coldest regions, calcite shells never
make it to the ocean floor and deep ocean sediments tend to
be dominated instead by microscopic shells made of silica that
housed planktonic plants called radiolaria. When these accumulate
en masse they can eventually form a rock known as chert.
In some environments, limestone can form as a ready-made hard
rock rather than an unconsolidated deposit. This occurs where
the sea floor is encrusted by sedentary organisms that secrete
calcium carbonate. Today, the most important such organisms
are corals. These are animals, but most have photosynthetic
algae living symbiotically within them so they require clear
shallow water with sufficient sunlight to keep the algae happy.
Delicate branching or fan-like varieties grow in quiet conditions,
and more robust forms colonize rougher settings.
Coral reefs are familiar to many people through marine wildlife
television documentaries. If the substrate they have colonized
is subsiding, they can build upwards so as to keep pace and
thereby maintain the living part of the reef close to sea level.
Charles Darwin realized this in his explanation for coral atolls,
which are rings of reef made of robust, wave-resistant coral
sheltering a shallow lagoon that hosts more delicate corals
(Figure 9.13). Drilling has shown that the volcanic basement
to some atolls has sunk to more than 1000 m below sea level.
Mururoa atoll in the Pacific (which achieved notoriety in the
mid-1990s as the site of French underground nuclear testing) is
of this type.
(a) (b)
sea level
Figure 9.13 Cross-section showing two stages in the formation of a coral

atoll by subsidence. (a) An extinct volcano is fringed by a reef. (b) The
island sinks isostatically (because of the load exerted on the lithosphere
by the weight of the island), but the reef builds up to keep pace with the
subsidence.
182

Apart from forming atolls, coral reefs shelter many tropical
shorelines in places devoid of muddy river sediment. The biggest
and most famous of these is the 2000-km-long Great Barrier
Reef off the coast of north-east Australia.
Along some carbonate-rich coastlines, agitation by waves
causes a rolling motion that can cause near spherical millimetre-
sized grains of calcium carbonate to form, by precipitation in
concentric layers around a tiny fragment or silt particle. These
grains are called ooids, and they can accumulate into deposits
forming rocks called oolites, or oolitic limestone.
Key idea: Clastic and non-clastic

sedimentary rocks
Clastic sedimentary rocks are made of particles (clasts) that
have been somehow transported (in the bed-load or suspension-
load). Non-clastic sedimentary rocks form in situ by precipitation
(inorganic in the case of evaporites, organic in the case of
limestone reefs).
Turning sediment into rock

It has been possible to give only a general account of the more
distinctive sedimentary environments. I will finish by considering
what it takes to turn a deposit of sediment into a rock.
The main requirement is that it must remain buried rather than
being eroded away. It is easy to forget this obvious point when
looking at a sedimentary rock in the field. A 3-cm-thick layer
of ripple cross-bedded sand may have taken less than a day to
accumulate, but it may be the sole survivor of a million such layers
that were successively deposited and then stripped away, perhaps
with every turn of the tide. Such is the fate of most sediment
deposited near shorelines, in rivers and in desert dunes, and it is
only in deep marine conditions that we normally find continuous
(but very slow) deposition without any erosional breaks.
Once buried, processes begin that can turn soft sediment into
hard sedimentary rock. Reef limestones and some evaporites
are already hard when they form, but other sediments begin

as loose, unconsolidated deposits. However, with burial, and
over time, various chemical and physical changes take place,
collectively described as diagenesis.
The dominant physical process is compaction, caused by the
weight of later sediment deposited on top. Clay-rich muds
may begin with solids making up only about 20 per cent of the
volume, the remaining 80 per cent being spaces between clay
flakes filled by water. The deposit is said to have a porosity
of 80 per cent. By the time such a deposit is buried beneath a
kilometre or so of sediment, most of this pore water has been
squeezed out, and the porosity is reduced to less than 10 per
cent (Figure 9.14). Consequently, a 1-m layer of compacted
mudrock may represent several metres of uncompacted mud.
(a) 10 µm
(b)
Figure 9.14 Magnified cross-sections showing compaction of mudrock:

(a) loosely packed clay flakes oriented randomly soon after deposition; (b)
clay flakes rotated during compaction so they lie horizontally. In this state
the rock can be split readily along closely spaced horizontal planes and is
called a shale.
Sands and silts are made of clasts that are more equidimensional
than clay, so the potential for compaction and loss of porosity
is less (Figure 9.15). In these sediments, compaction is achieved
by a combination of rotation, bending and fracturing of grains.
A sandy sediment may begin with up to 40 per cent porosity
and end up with only 10 per cent porosity at a depth of about
a kilometre. At greater depths, the points of grains resting on
others may dissolve away through a process called pressure
solution, reducing porosity still further. Similar compaction
processes affect carbonate sediments.
Most of the chemical changes during diagenesis are controlled
by the nature of the pore waters. Silica that has been dissolved
by pressure solution may come out of solution to fill the
remaining pore spaces with silica cement. This can turn a loose
sandy deposit into a hard rock (sandstone) in which the grains
184

are just as thoroughly stuck together as crystals in an igneous
or metamorphic rock. Pressure solution in limestones results in
pore waters that are saturated in calcium carbonate, so calcite
cements are almost universal within limestones. Calcite cements
may also form in the pore spaces of other rock types, either
from solution of minor amounts of calcium carbonate originally
in the deposit, or from calcium carbonate that migrating pore
waters have carried in from elsewhere.
(a) 1 mm (b)
Figure 9.15 Magnified cross-sections showing compaction of sandstone.

(a) Grains soon after deposition. Each grain touches its neighbours,
but only at one point, which rarely lies exactly in the plane of the cross-
section. (b) The same grains after compaction. There has been some
rotation of grains, but most of the compaction has been achieved by
pressure solution, in which material has been dissolved at pressure points
where grains touch. Dissolved material has been precipitated nearby, so
that most of the remaining pore space is filled by cement.
There are some mineralogical changes that may take place, too.
One of the earliest of these affects limestones. In many shells,
the calcium carbonate takes the form of aragonite, a polymorph
of calcite. Fragmental debris created from the abrasion of shells
is therefore mostly aragonite, but this converts to calcite during
diagenesis, so hard rocks made of aragonite are unknown.
Sometimes the diagenesis of limestones from originally lime-rich
muds involves so much recrystallization that the resulting rock
consists largely of interlocking crystals of calcite. This aspect of
its texture resembles an igneous rock rather than a fragmental
sediment, although its composition and the common presence of
fossils makes misidentification unlikely.
Other diagenetic changes include the driving off, as a result of
pressure and mild heat, of water that was originally bound up
in clay minerals. New clay minerals, with less water in their
formulae, take their place. This sort of change is on the verge

of metamorphism, and it is really just a matter of choice what
to describe as extreme diagenesis and what to think of as low-
grade metamorphism.
Key idea: Diagenesis

Turning clastic sediment into rock (diagenesis) involves
compaction (due to burial), cementation and sometimes
mineralogic changes and loss of water, too.
You have now seen how sedimentary rocks are created from
components derived by weathering and erosion. Sedimentary
rocks are not a dead end in the rock cycle. Most ocean floor
sediments eventually get dragged into trenches at destructive
plate margins, where they are metamorphosed or even melted.
Sediments in shallow seas or deposited on land may be regionally
metamorphosed after burial to sufficient depth, or if they are caught
up in a continent–continent collision. Alternatively, they may be
raised up to be exposed to the forces of weathering and erosion.
The remaining links in the rock cycle are covered in the next
chapter, where you will see how rocks behave when subjected to
stresses that deform them.
Dig deeper
A vast book relevant to this chapter is Earth Surface Processes,
Landforms and Sediment Deposits by John Bridge and Robert
Demicco (Cambridge University Press, 2008).
The classic text about sedimentary rock types is Sedimentary
Petrology by Maurice Tucker, 3rd edn (Wiley-Blackwell, 2001).
The environments of sedimentary deposition and the attributes
of the sediments deposited there are nicely explored in a
20-minute slide-show video here:
https://www.youtube.com/watch?v=jW6w7Dfq9qo
There is some free online learning about sedimentary rocks at
science/geology/practising-science-reading-the-rocks-and-
ecology/content-section-0
186

Fact-check
a All sedimentary rocks are clastic
b All sedimentary rocks are silicilastic
c Mud formed of clay is siliciclastic
d All clasts are produced by erosion
2 Which of these is least likely to contain cross-bedding?

a Sand formed into ripples by wave action
b Sand deposited by the wind
c A mudflow
d A pyroclastic flow
3 Which of these bedforms is usually symmetric in cross-section?

a Wave-formed ripples
b Ripples made by a tidal current
c Ripples on the bed of a river
d Desert sand dunes
4 Which of these is a valid expression of Walther’s Law?

a The present is the key to the past
b In a sedimentary succession, grain-size tends to become
finer as you go upwards
c A thirsty sedimentologist always heads for the delta bar
d The lateral juxtaposition of sedimentary environments at an
instant in time is reflected by environmental changes seen in
a vertical succession, which represent the same location over
a period of time.
5 Which of these minerals would you not expect to find in an

evaporate deposit?
a Calcite
b Chlorite
c Gypsum
d Halite

6 Which of these minerals have identical chemical compositions?
a Anhydrite and gypsum
b Aragonite and calcite
c Dolomite and calcite
d Gypsum and halite
7 Which of these is not a characteristic attribute of a braided

river?
a High sediment load and seasonally varying flow
b Deposition mostly of sand and gravel
c Several interconnecting channels
d Migrating meanders
8 Which of these can happen during diagenesis?

a Cementation
b Compaction
c Mineralogic changes
d All three
188

10
Deformation
of rocks
33about the ways in which rock can
become deformed, and the processes
responsible
33how solid rock can fracture when it is
displaced by fault movement, and how in
other circumstances it can be bent into
structures known as folds
33how a deformed succession of rocks can
become buried by younger, undeformed
sediments.
189

Parts of the Earth’s crust are from time to time subjected to
stresses that tend to extend or compress them. These stresses
may be regional in extent (such as those associated with plate
tectonics), or may operate more locally (perhaps caused by the
extra load on the crust produced by a thick accumulation of
sediment or the growth of a volcano).
Geologists use the term stress to refer to the forces acting to
deform a volume of rock, and strain to describe the resulting
deformation. When rock is relatively warm, it is able to deform
in a ductile fashion, that is to say strain can be accomplished
by squeezing or stretching without fracturing. Ductile strain
occurs by rotation, deformation and recrystallization of grains
or crystals within the rock, and happens most readily if the
stress builds up slowly. Ductile strain is thus characteristic
of deformation at deep crustal levels, where the temperature
is higher. At shallower levels, rocks tend to fracture when
subjected to stress, especially if the stress builds up quickly.
This is how faults and joints form, and is described as brittle
deformation.
Key idea: Brittle and ductile deformation

During ductile deformation rock is squeezed, stretched or bent in
a plastic fashion, without fracturing, whereas brittle deformation
happens primarily by means of fracturing.
The effects of deformation are most readily seen in sedimentary

rocks, because their inherently bedded nature provides ready-
made markers. For example, an individual bed of limestone,
sandstone or mudrock may be traceable over tens or even
hundreds of kilometres. If its base and top were virtually
horizontal when deposited and if the bed is no longer horizontal
when we see it, then it must have been tilted. Similarly, if the
bed is no longer flat but curved, then we can deduce that it
has been folded. Moreover, some beds within a sedimentary
succession are usually distinctive, and enable us to match
displacements across faults.
190

Different types of steep fault
Faulting is displacement of rock across brittle fractures. There
are essentially three types of fault movement, all illustrated in
Figure 10.1. The rock all around has been made invisible so that
you can see what happens to this specific block of terrain. The
sequence here consists of an undeformed succession of horizontal
(a) (b)
(c1) (d1)
(c2) (d2)
Figure 10.1 Cross-sectional block diagrams illustrating three types

of fault, each starting from the situation in (a): (b) shows the area after
transcurrent movement across the fault plane; (c) shows what would
happen if, instead, the block on the right went down relative to the block
on the left (c1 without erosion, c2 after erosion); (d1) and (d2) show what
would happen if the block on the right went up instead of down. Note that
the fault plane is not quite vertical. We can therefore distinguish case (c) as
a normal fault and (d) as a reverse fault. (See text for further discussion.)

sedimentary beds cut by a dyke of igneous rock, which will act
as a convenient vertical marker. The sedimentary beds could
instead equally well be lava flows, layers of pyroclastic rocks or
a series of horizontal sills; the main thing is that there is a set of
individually recognizable horizontal layers, or strata.
Spotlight: Strata
Strata (singular: stratum) is a useful term that encompasses all
kinds of layers, not just sedimentary beds. However, the term
‘stratigraphy’ is used to denote the study of the distribution,
deposition and ages of sedimentary rocks in particular. William
Smith (1769–1839), a surveyor who made the first geological map
of most of the island of Great Britain, was known to some of his
contemporaries as ‘Strata Smith’.
In Figure 10.1(a), the area is shown before any fault motion has
taken place, though the position of the future fault is marked. In
this example, it is a plane that is not quite vertical but is dipping
steeply down to the right.
One thing that might happen is that the blocks of terrain on
either side of the fault could slip sideways past each other. This
is called strike-slip or transcurrent fault motion, and is what is
happening along most of the San Andreas Fault in California.
As Figure 10.1(b) shows, if the direction of motion is exactly
parallel to the bedding (or stratification), there is no resulting
mismatch of displaced strata either side of the fault. However,
the vertical marker, in this case a dyke, has been offset, and
the amount of offset can be measured to determine how much
movement there has been across the fault plane. If you were
to stand on the dyke on one side of the fault, you would see
that the other part of the dyke had been offset to the left. You
would see a leftwards offset whichever side of the fault you
were looking from. The fault illustrated is thus distinguished
as a left-lateral (or sinistral) transcurrent fault. If the offset
had been in the opposite sense, so that the marker had been
displaced to the right, it would be called a right-lateral (or
dextral) transcurrent fault.
192

Key idea: Transcurrent faults
Transcurrent (or strike-slip) faults are sites of sideways displacement.
Alternatively, the fault plane could separate regions whose

relative movement is vertical. When this happens, the offset
can be measured by the fact that strata that were originally
at different depths have become juxtaposed (it is immaterial
whether we regard one side as going up or the other side as going
down). However, there will be no evidence of offset displayed
by a vertical feature cut by the fault. Figures 10.1(c) and (d)
show alternatives, each starting from the situation shown in
Figure 10.1(a). In Figure 10.1(c) the terrain on the right has gone
down relative to that on the left. Usually, the side that is being
relatively uplifted experiences much more erosion, and the result of
planing off to a new horizontal surface is shown in Figure 10.1(c2).
As a result, an older bed on the left of the fault is exposed at the
surface next to a younger bed on the right of the fault. In Figure
10.1(d) the fault motion was in the opposite direction, and the
younger bed survives at the surface on the left of the fault instead.
Figures 10.1(c) and (d) are not quite mirror images of each other
because the fault plane is dipping steeply downwards to the
right instead of being vertical. In Figure 10.1(c) the downthrown
terrain is on the down-dip side of the fault; this is the commonest
kind of fault movement, and is known as a normal fault.
Key idea: Normal faults

Faults associated with local extension are normal faults.
Spotlight: Normal or abnormal?

‘Normal fault’ is a term inherited from eighteenth- or nineteenth-
century miners, who gave this name to the most common type
of fault that they encountered in coal mines, so to them it meant
‘normal’ in the sense of ‘ordinary’. It is merely coincidental that
the direction of displacement is at right angles to (i.e. normal to)
the line of the fault. In a transcurrent fault, the displacement is
parallel to the line of the fault.

In Figure 10.1(d) the downthrown terrain is on the up-dip side
of the fault. This is described as a reverse fault. Its motion is the
exact opposite of a normal fault.
In active fault belts, erosion may keep pace with the rate of
uplift, so that faults are rarely represented by cliffs. There is
certainly never an overhang at the surface of the kind shown in
Figure 10.1(d1), because this would crumble away as it formed.
However, sometimes erosion lags behind to the extent that the
position of an active normal or reverse fault is indicated by a
straight edge to a belt of hills.
If you compare stage (c) in Figure 10.1 with stage (a), you will see
that because the fault plane is not quite vertical, the distance from
left to right has increased slightly. Normal faulting is therefore
common in regions of crustal extension. If, however, the fault acts
as a reverse fault, ending up at stage (d), the left-to-right distance
has been slightly reduced. Reverse faults are thus characteristic
of regions of crustal compression. However, these are gross
generalizations. In nature, most faults are parts of linked
systems and it is unwise to read significance into a single fault.
Key idea: Linked faults

Faults of different types are often linked together.
Case study: Large faults

Although the San Andreas Fault is a plate boundary, this is not true
of all transcurrent faults. Some areas of continental crust have
many faults (of all kinds) that are not, and never have been, at the
boundary between two plates. Also, most faults that you will see
if you go on a field trip will probably be old, not having moved for
millions or hundreds of millions of years, and no longer pose a
significant earthquake hazard.
A good example is the Great Glen Fault in Scotland. This runs in
a southwest to northeast straight line across Scotland from Loch
Linnie via Loch Ness to Inverness, and can be traced a long way
offshore at either end. Its last movement was 100 km of right-
lateral displacement between about 90 and 50 million years ago.
194

A series of consecutively developed normal faults is shown in
Figure 4.6, which illustrates continental rifting. Figure 10.2
shows a similar situation at an enlarged scale, and defines some
terms that you may find used elsewhere to describe faulted
areas. Note that the faults in this figure are shown as curved
surfaces, becoming less steep at depth, which allows rotational
movement between adjacent blocks.
rotated
block
horst
graben step faulting
Figure 10.2 Cross-section showing some features associated with faults.

A graben is a block that has dropped down between two normal faults.
A horst is a block that has been left standing high as a result of blocks
on either side having dropped down. In nature, the scale of this diagram
would typically be hundreds of metres to tens of kilometres across.
Spotlight: Grabens or graben?

The German word Graben (meaning ‘ditch’ or ‘trench’) is used by
geologists to denote a down-dropped block between two roughly
parallel faults (see Figure 10.2). In German, the singular and plural
of Graben are identical. However, now that the word graben has
been adopted into English, without the initial capital letter as in
German, all sensible geologists put an –s on the end to make a
plural. In writing this, I have probably offended about 50 per cent of
my colleagues. I think it looks pretentious to write such phrases as
‘The East African Rift Valley is formed by a series of linked graben’
or ‘typically, horsts have graben on either side’, but take your pick.
It’s up to you; after all, we still say ‘strata’ and not ‘stratums’.
The previous discussion was based on using the displacement of

strata to determine fault movement. Sometimes there are other,
smaller-scale clues. The best known is aligned scratches or gouge

marks made on the fault plane itself when the rock masses
ground past each other. These are known as slickensides (not
slickenslides), originating from an English dialect term. However,
they tend to record only the last episode of fault motion, and this
may not be representative of the overall displacement.
Low-angle faults
The fault plane drawn in Figure 10.1 dips steeply and cuts
the strata at a steep angle. Sometimes fault planes lie at a low
angle to bedding, and may even become parallel to it. In this
case, where the motion is compressional, the fault is described
as a thrust fault, or more commonly just as a thrust. Where
the motion is extensional, the fault is described as a low-
angle normal fault or low-angle extensional fault. Faults that
develop parallel to bedding occur that way because some layers
are easier to fracture than others, and thrusts and low-angle
normal faults usually lie between beds or within weak beds
such as shales, avoiding stronger beds (such as well-cemented
sandstones and limestones). However, they must approach the
surface somehow, and this is usually achieved by the fault plane
cutting steeply through each strong bed, yet dipping gently
everywhere else. In this situation, it is more appropriate to write
of a fault surface rather than a fault plane, because it is certainly
not a smooth plane.
Key idea: Thrusts and reverse faults

Faults associated with local compression are reverse faults and
thrust faults (which are just low-angle reverse faults).
Figure 10.3 is a rather complicated block diagram, but it serves

to illustrate many of the most important features of low-angle
faults. A thrust is illustrated, but the same geometrical concepts
apply to low-angle normal faults. Stage (a) shows the situation
before displacement, with the position of the future fault surface
marked. The rock below the fault is described as being in the
footwall, whereas the rock above belongs to the hanging wall.
Thrust movement is going to occur from left to right parallel
196

thrust transcurrent fault
(a)
(b)
Figure 10.3 Cross-sectional block diagrams illustrating the motion of a

thrust sheet. In nature, the scale of this could be anything from a metre to
tens of kilometres across (see text for discussion).
to the side of the block facing you. Looking at this face first,
you can see that the fault surface runs between beds for most
of its length (perhaps there is a thin, weak layer of mudrock to
lubricate movement between each of the major beds drawn),
but cuts steeply upwards at two places. The steep bits are called
footwall ramps, and the flat bits are called footwall flats. The
fault does not reach the ground surface on this face, but you
would see it do so if the diagram were extended to the right.
The right-hand face of the block reveals the fault surface at the
base of the topmost bed close to the near corner, but the fault
surface turns upwards to reach the ground about a third of the
way along the face. If we trace the fault on to the top (ground)
surface, we see that it becomes a fault line running parallel to the
direction of thrust movement. This sort of boundary is referred
to as a sidewall ramp, and along this part of its length the fault
will behave as a transcurrent fault. The fault at the surface

makes a right-angle turn where it becomes a thrust again, then
turns again to become another length of transcurrent fault.
To put it more concisely, the sheet of rock that is about to
be transported by a thrust (referred to as a thrust sheet) is
bounded on its lower surface by a thrust fault and on its sides
by transcurrent faults.
The situation after thrust movement is shown in stage (b).
Looking at the near face of the block, we can see that there are
two places where a bed has been repeated by being thrust over
itself. Also, because the fault surface is not a plane but has steps
in it, the thrust sheet has had to deform; in fact, it has folded in
a ductile fashion. This folding occurs only in the thrust sheet,
and it does not affect the units in the footwall. In nature, the
up-folded areas would probably become eroded away, but
the changing dips of the beds exposed at the surface would
demonstrate that folding had taken place, and would enable us
to deduce the locations of underlying footwall ramps.
Key idea: Folds caused by footwall ramps

Folds may occur as a geometric consequence of compressional or
extensional low-angle fault movement over ramps in the footwall,
or as a result of crustal shortening independent of faulting.
Thrust belts and collision zones

You will meet folding in more detail shortly, but first there is
another important aspect of thrusts to consider. You saw in
Figure 10.3 that the footwall rocks below the thrust surface do
not fold like those above. However, very often, as deformation
proceeds, a new thrust develops below the first. What happens is
that the active thrust surface jumps forward by slicing ramps off
from the footwall, as illustrated in Figure 10.4.
By propagating forward in this way, a thrust develops into
a series of blocks (or ‘horses’) bounded above and below by
thrusts, called a duplex complex. Rocks in the hanging wall that
were folded during motion of the original thrust sheet are liable
to be folded again each time the geometry of the active thrust
198

surface changes. Figure 10.5 shows what the final result might
look like in cross-section, at the site of a single footwall ramp.
(a)
(b)
(c)
Figure 10.4 Progressive development of a thrust system, seen in cross-

section. Undeformed rock below the lowest thrust is shaded. Bedding is
not shown, but can be assumed to be horizontal, except where folded as
a result of thrusting. In each case, the thrust that is just about to become
active is shown by a heavy line. Inactive lengths of thrust are shown by
finer lines. Note that these become folded when they are transported over
footwall ramps.
Spotlight: Horse and horst

Do not confuse ‘horse’, an English word used to describe an
imbricated block in a thrust stack (bounded by thrusts above and
below) and looking like a horse trying to mount whatever is in front
of it (often another horse), with ‘horst’ (Figure 10.2), originally a
German word for ‘heap’, which denotes a block bounded on either
side by a normal fault.

roof thrust horse
sole thrust
Figure 10.5 Cross-section through a thrust system. This shows the sort
of structural complexity that results from progressive forward propagation
of the active thrust, as in Figure 10.4. A series of imbricated thrust slices,
called ‘horses’, is produced, each horse being bounded by thrusts above
and below. In this example, thrust motion is from left to right, so the first
horse to form was the leftmost one.
Figure 10.5 could represent a 100-km-long cross-section from

west to east under the eastern Canadian Rockies. These are a
collisional mountain belt with thrust displacement from west
to east. Alternatively, it could represent a 10-km-long cross-
section from east to west in the Moine thrust zone in north-west
Scotland, where regionally metamorphosed rocks have been
thrust westwards over a series of 500-million-year-old sediments.
Case study: The Tibetan plateau

The world’s biggest currently active collision zone provides an
excellent example of how different kinds of fault are related. The
Tibetan plateau has an average altitude of 5 km above sea level,
and became uplifted as a result of the collision of India into central
Asia, which began about 35–50 million years ago. The plateau is
bounded to the south by the Himalayas, which is a mountain belt
that has developed by means of a series of southwards-directed
thrusts. The plateau itself is crossed by many grabens delineated
by normal faults running north to south, which at first sight
seems odd in a region that ought to be experiencing north–south
compression rather than east-west extension.
The explanation becomes clear when it is recognized that the

plateau is bounded by and crossed by numerous transcurrent
200

faults, few of which had been recognized by geologists before they
were revealed on images from satellites in the 1970s. Some run
from southwest to northeast and are left-lateral faults; others run
from northwest to southeast and are right-lateral.
Tibet is buttressed by mountains on all sides except the east, and

the Tibetan lithosphere is being squeezed out eastwards by motion
across these faults. The north–south normal faults are just a
manifestation of Tibet’s east–west extension that accompanies its
north–south shortening (Figure 10.6).
Altyn
AltynTagh
Taghfault
fault
Karakoram
Karakoram
fault
fault TIBET
TIBET
Himalayan
Himalayanmain boundarythrust
main boundary thrust
INDIA
1000 km
Figure 10.6 Sketch map showing how the Tibetan plateau is bounded and
crossed by transcurrent faults, and has north–south normal fault systems
within it. Tibet is being squeezed out towards the east. The fault systems in
Tibet are so large that they were not appreciated until seen from space.
So, to really understand the tectonic history of an area, the

whole fault system needs to be mapped out. Sometimes it can
be deduced that fault motion has reversed one or more times
during a prolonged period of deformation, or that faults that
seem related were in fact active at quite different times.

Remember this!
The biggest faults have displacements of hundreds, or even
thousands, of kilometres. This is accomplished by a vast number
of slip events over millions of years in the stick-slip manner
described in Chapter 3. Slip is usually a few centimetres in one go,
and exceeds 10 m only in the biggest earthquakes.
Brittle faults versus shear zones

The best-known ancient faults are those that were active at
relatively shallow depths in the crust, say a few kilometres
or so, because these are the ones that are most likely to end
up exposed at the surface by erosion. The rock within a few
centimetres or metres of such a fault is often intensely fractured,
and is described as a fault breccia. Faults therefore tend to
be zones of weakness that erode away readily. Because of the
fracturing associated with them, faults may act as pathways for
groundwater, and can be important sources of underground
water in arid regions, especially where they occur in hard
igneous or metamorphic rock that is otherwise impermeable.
Low-angle faults can be mapped below ground because
they will reflect back seismic waves generated by explosions
deliberately set off for the purpose. It is believed that, below
a depth of 10–20 km, many faults become ductile features, in
which displacement is accommodated across a shear zone that is
tens or hundreds of metres wide for moderate-sized faults.
The recrystallization that occurs in a shear zone can be thought
of as a localized form of metamorphism, referred to as dynamic
metamorphism. Rock from a ductile shear zone with a streaked-
out texture indicating the shearing is referred to as a mylonite.
Folds
Folding is another kind of ductile deformation, and it can
occur at shallow depths. You have already seen that folds
are an inevitable side effect of thrusting. The Appalachian
Mountains are part of a folded and thrust mountain belt that
stretches from Alabama to Newfoundland and is the result of
202

a continent–continent collision between North America and
Africa that took place about 300 million years ago.
Traditionally, folds have been regarded as features produced
when the crust experiences compression. However, they can also
form in hanging walls in extensional settings, as a consequence
of footwall ramps on low-angle extensional faults.
Figure 10.7, showing a cross-section through folded rocks,
introduces some terms commonly used to describe folds.
An anticline is an arch-like structure, whereas a syncline is
U-shaped in cross-section. By definition, the axis or hinge of a
fold is where the curvature is greatest, whereas on the limbs to
either side of the axis curvature is less.
axial
anticline plane hinge line
fold limb
fold limb
syncline axial
plane
Figure 10.7 A perspective cross-section through some folded strata. This
could be on any scale from centimetres to kilometres across.
Key idea: Anticlines and synclines

Anticlines are folds resembling arches, whereas synclines are
U-shaped in cross-section.

Shapes and styles of folding
Folds can take a whole variety of forms, depending on the rate
of deformation, the depth, and the mechanical properties of
each stratum. The strata in Figure 10.7 are folded in continuous
curves, in contrast to the folds portrayed in Figure 10.3 where
the changes in dip are abrupt. In Figure 10.7 there is a tendency
for the strata to get slightly thicker near the fold axis: this is
accomplished by ductile deformation within each stratum,
manifested by distortion of features within it. In extreme
cases, the degree of distortion and accompanying rotation
and recrystallization can lead to the development of cleavage,
especially in mudrocks, with cleavage planes being roughly
parallel to the axis of the fold.
To find cleavage in folded rocks should come as no surprise
to you. You met cleavage as an essential element of low-grade
regional metamorphism in Chapter 7 and, by their very nature,
most regionally metamorphosed rocks have been deformed.
The axial planes of the folds in Figure 10.7 are very steep and
the hinge lines are horizontal. It does not have to be so: axial
planes can sometimes have very gentle dips, hinge lines can
plunge downwards at any angle, and opposite limbs of a fold
can diverge at a larger or smaller angle. Examples are shown
in Figure 10.8. In highly deformed regions, fold limbs are often
parallel to one another, and in these cases the folds are described
as isoclinal. If the axial plane dips at less than ten degrees,
the fold is described as recumbent (Figure 10.9). In multiply
deformed areas, younger folds may distort older folds, resulting
in a complex pattern (Plate 10). Folds may even be turned upside
down, in which case the anticlines look like synclines, and vice
versa. In such a situation, the right way up has to be determined
by inspection of small-scale structures such as cross-bedding.
Folds manifest themselves upon the landscape in a variety of
ways. It is rare for the crests of hills or mountain ranges to
correspond with the axes of anticlines, because erosion gets to
work and it is the more resistant strata that tend to form the
high ground, as shown in Figure 10.10.
204

(a) (b)
(c)
Figure 10.8 Styles of folding: (a) upright folds, having steep axial planes;
(b) recumbent folds, having near-horizontal axial planes; (c) isoclinal
recumbent folds, with fold limbs almost parallel.
Figure 10.9 Recumbent folds in thinly bedded siltstones on the north

Devon coast, England.

Figure 10.10 Cross-section through an eroded anticline, looking along
its hinge line, showing how strata that are more resistant to erosion form
ridges. Gentle slopes following the surface of a stratum are described as
‘dip slopes’. Opposite-facing steeper slopes (where a stratum has been cut
through by erosion) are called ‘scarp slopes’.
Unconformities
Sometimes tilted or folded strata that have experienced erosion
are later covered by younger deposits. These will share neither
the dip nor the previous deformation history of the strata they
bury. The junction between the two series of strata, which may
represent a gap of millions or even hundreds of millions of
years, is called an unconformity.
An angular unconformity, as in Figure 10.11, is usually easy
to spot because of a difference in the amount or direction
of dip between the strata below and above. However, it can
be very difficult to recognize unconformities where there is
no difference in dip. These occur where the upper strata of
an unfolded, untilted succession have been eroded away in a
planar fashion and then covered by younger strata, or simply
where deposition was interrupted by a pause during which no
sediment accumulated.
Key idea: Unconformities

An unconformity is where younger strata sit on top of older strata
that have usually been tilted or deformed and eroded before the
younger strata were laid down. It represents a time gap that may
be hundreds of millions of years.
206

(a)
(b)
Figure 10.11 Cross-sections showing younger strata deposited

unconformably on the eroded anticline (seen in Figure 10.10): (a) no further
erosion before deposition, so that old topography is preserved in buried form
below the younger strata; (b) the older rocks planed flat before deposition
recommenced. In both cases the junction between the older (folded) and
younger successions is described as an angular unconformity, because of
the difference in dip between the strata above and below.
Case study: Deep time

Although, technically, any subsequently buried erosional surface
can be called an unconformity, the term is not usually applied if
the presumed time gap is geologically short, the amount of erosion
slight, and the change in environment trivial. For example, the
erosional bases of the river and delta channels in Figures 9.8
and 9.11 would not normally be classed as unconformities.
However, in the case of angular unconformities, common sense

and logical thinking can tell you that a very considerable time must
have elapsed between the deposition of even the youngest rocks
below an angular unconformity and the oldest rocks above it.

The time unrepresented within the strata themselves is the time
needed to bury, then deform and then erode the succession below
the uncomformity, before deposition recommenced.
This realization, first articulated by the Scotsman James Hutton

(1726–97), led geologists in the late eighteenth century to challenge
the widely accepted young age of the Earth (a few thousand
years) implied by a literal interpretation of the Bible and to accept
intellectually a concept known as ‘deep time’. On the basis of his
observations, in 1795 Hutton memorably wrote of the age of the Earth
that he could ‘find no vestige of a beginning – no prospect of an end’.
In particular, Hutton documented an angular uncomformity

between steeply dipping metamorphosed shales and overlying
gently dipping sandstones at Siccar Point on the east coast of
Scotland. Radiometric dating has now shown that there is a
60-million-year gap between 425-million-year-old rocks below and
345-million-year-old rocks above Hutton’s unconformity.
One way to decide the duration of the time interval represented

by an unconformity is to determine the difference between
the ages of the beds immediately above and below. In many
situations, the most precise dating comes not from radiometric
techniques, such as those described in Chapter 2, but by looking
for differences in any fossils. You will examine fossils and their
value in stratigraphic dating in Chapter 12, but first I invite you
to consider the Earth’s physical resources.
Dig deeper
A recent, well-illustrated, book about the folding and faulting
of rock is Structural Geology by Haakon Fossen (Cambridge
University Press, 2010).
A university-level text on deformation fabrics in metamorphic
rocks is Structural Geology: The Mechanics of Deforming
Metamorphic Rocks by Bruce E. Hobbs (Elsevier Science
Publishing, 2014).
208

Fact-check
1 Which of these is not an example of ductile deformation?
a A fold
b Mylonite
c Cleavage in a mudrock
d A fault
2 What is a recumbent fold?

a One in which the hinge line is bent
b One in which the hinge line dips at less than 10 degrees
c One in which the axial plane dips at less than 10 degrees
d One in which the limbs are almost parallel
3 What is deep time?

a The age of the stratum in the lowest part of a mine
b The necessarily very long timescale represented by the
sequential deposition of rock, followed by its burial,
deformation, erosion and then the deposition of a new
succession of rock above the unconformity
c The depth below the ground surface to a buried unconformity
d The average rate at which a fault moves

a Normal faults have footwalls, whereas reverse faults have
hanging walls
b If a fault is curved, this can allow the displacement between
adjacent blocks of crust to be rotational
c A horse is an upthrown block between two normal faults
d A reverse fault is one whose direction of movement has
changed over time
5 Which of these is a left-lateral transcurrent fault?

a The Altyn Tagh Fault
b The Great Glen Fault
c The Karakoram Fault
d The San Andreas Fault

Physical resources
11
33about fuels and other useful materials
that can be extracted from the ground
33how some of them originate, and how
they can be found
33about the implications of extracting
physical resources much faster than
they are renewed by natural geological
processes.
211

The human race relies on geology (broadly defined) for all its
raw materials (except things that grow), and for almost all
its energy (except solar power). Winds, waves and tides can
be used to generate power, but the vast majority of power to
run our industrial society comes from burning fossil fuels,
specifically coal, oil and natural gas – all of which release
carbon dioxide into the atmosphere when burned, and so
contribute to global warming via the greenhouse effect.
Because humans are using up these fuels at rates vastly in
excess of the rates at which geological processes can form
them, they are classed as non-renewable resources, along with
the radioactive fuels used for power generated by nuclear
fission. Wind, wave and tidal power, and (to an extent)
geothermal power generated from the heat escaping from
shallow magma bodies, do not outstrip the rate of supply, and
are referred to as renewable resources.
In addition to fossil fuels, non-renewable resources include such
commodities as the ores from which metals are extracted, the
clay used to make bricks and porcelain, the limestone used to
make cement, and rock in general used to surface roads.
If we continue to use up raw materials at the present rate, some
things will run out sooner than others. For example, there is
no shortage of limestone that could be used to make cement.
(We might not be willing to accept the environmental impact of
quarrying away beautiful limestone hills, but that is a separate
issue.) On the other hand, in 2008 silver was being used up
at a rate sufficient to exhaust the world’s supply of that metal
within 13 years.
The rates at which we are using up many raw materials, and the
inadequacy of recycling, are causes for concern, but there is no
need to panic. The world will not actually run out of silver
13 years after 2008. More silver deposits are being discovered,
and silver is still being extracted from deposits that were
previously uneconomic to exploit. This pushes the price of silver
up, and encourages both recycling and the use of substitutes (see
the case study below).
212

Reserves and resources
The example of silver above highlights an important distinction
between resources and reserves, two terms that are often
misused. Their relationship is summarized in Figure 11.1. Strictly
speaking, ‘reserves’ means only those raw materials that can be
extracted profitably and legally under existing conditions. On
the other hand, ‘resources’ should be used only in a wider sense,
to describe the total amount of the raw material that could be
extracted if the price were high enough, or legislation allowed, or
technological advances made it feasible. Because exploration is
incomplete, estimates of resources are likely to be too small.
The earlier statement that the supply of silver was due to run out
in 13 years is thus rather inadequate. It should be rephrased to
state that the current reserves of silver were due to be exhausted
in 13 years. In practice, as reserves are used up by extraction
they are added to by decisions to extract the commodity from
deposits that were formerly regarded as not worth exploiting, or
by the discovery of new economic deposits. Thirteen years is thus
a fairly comfortable cushion, so long as new discoveries continue
at a rate sufficient to keep pace with extraction. Even so, every
non-renewable resource is liable to be exhausted eventually,
so recycling and the use of substitutes will inevitably become
increasingly important. Thus, we should extract non-renewable
resources with care and sensitivity, and consume them wisely.
increasing certainty of existence

discovered not yet discovered
already
economic now
used up
increasing economic feasibility
RESERVES
not yet economic
RESOURCES
Figure 11.1 The relationship between physical resources and reserves.

Global reserves of uranium (used as a fuel in nuclear reactors)
are about 70 years’ supply, and the untapped resources are
vast. Nuclear power is attractive because of the almost infinite
nature of its fuel supply and because it does not emit carbon
dioxide and thereby contribute to global warming. However,
a significant downside is that spent nuclear fuel is highly
radioactive and needs to be stored or buried securely.
In this chapter you will first look at fossil fuels, and then at
other non-renewable resources such as ore minerals and bulk
materials. You will see how their origins can be explained by the
processes encountered earlier in this book.
Key idea: Reserves and resources

‘Reserves’ denotes raw materials that have already been
discovered and can be extracted profitably and legally, whereas
‘resources’ means the amount potentially available in total.
Case study: Not running out (yet)

I left the two paragraphs referring to the state of global silver
reserves in 2008 unchanged from the previous edition of this
book, as it allows me to make a comparison. In 2013 the global
consumption of silver was nearly 31 thousand tonnes per year,
but by then the global silver reserves had risen to 520 thousand
tonnes, equivalent to 17 years supply at the 2013 rate of demand,
so the date at which silver was due to run out had receded further
into the future. This boost in silver reserves happened because
it had become economical to produce more silver from mines in
China, Khazakhstan, Mexico, Indonesia and Peru (partly through
more efficient recovery of metal from the ore), and because of
better prospects for extracting silver as a by-product in lead-zinc
mines in Australia and India.
Origins of oil and gas

Of all the fossil fuels, oil is the one that tends to make the
headlines. It has become so important to the global economy
that wars have been fought over it. Oil and natural gas have the
214

same origin, and are often found in association. Moreover, they
are both hydrocarbons (made of carbon and hydrogen), so I will
consider them together. They are also the fuels whose burning,
along with coal, is responsible for most of the human-induced
increase in atmospheric carbon dioxide, leading to global
warming by means of the greenhouse effect.
The raw material for hydrocarbons is dead organic matter,
chiefly the remains of micro-organisms and land plants.
For sufficient of this material to collect, the depositional
environment must have a large supply of dead organic matter
coupled with a lack of oxygen to stop it decaying to water and
carbon dioxide. These conditions are met in areas of the sea
floor where the surface waters have a high rate of production
(and mortality) of planktonic plants, which upon death sink to
the bottom to be buried in fine-grained, shaley sediment. More
rarely, the floors of lakes provide a suitable environment.
Once an organic-rich shale has accumulated, it is a potential
source rock for hydrocarbons. For this potential to be realized,
burial must occur, to allow heat, pressure and chemical
conditions to cause the necessary diagenetic changes in the
organic matter. The conversion from dead organic matter to a
hydrocarbon is described as maturation.
By the time burial has reached a depth of 1 km, the temperature
is likely to be about 50°C. Under these conditions, the organic
matter will be cooked into complex molecules called kerogens,
which consist of carbon, hydrogen and oxygen. With deeper
burial to about 6 km and temperatures up to about 180°C,
the kerogens break down to hydrocarbons, which are chains of
carbon molecules bonded with hydrogen, known as petroleum
(a term derived from the Latin words for ‘rock’ and ‘oil’).
The general formula is CnH2n+2, where C is carbon and H is
hydrogen. The chains get shorter according to the temperature
of maturation. Where n is 16 or more the product is a waxy
solid, but shorter chains are liquid. Oil in the ground is usually
a mixture of many different hydrocarbons, and is referred to
as crude oil, to distinguish it from refined oil that has been
industrially separated into its components.

At slightly higher temperatures, the chains become really short,
producing C4H10 (butane), C3H8 (propane), C2H6 (ethane) and
CH4 (methane), which collectively are known as natural gas. If
the temperature rises much above 200 °C, even methane breaks
down, driving off the hydrogen and leaving the carbon behind
as graphite. Thus hydrocarbons can form and survive only if
conditions stay within a fairly narrow band. There is a time
element, too, because hydrocarbons can form and break down
at lower temperatures, given sufficient time (Figure 11.2).
Maturation time (millions of years)

10 100
onset of oil generation

peak of oil generation
all oil breaks down to gas
Temperature (°C)
100
200
peak of gas generation
all gas breaks down to graphite
Figure 11.2 Production of hydrocarbons by maturation within a suitable

source rock, showing how this depends on temperature and time. Note
that the scale of the time axis is logarithmic, not linear.
Oil and gas reservoirs

It is one thing for hydrocarbons to form and to survive within
a rock. It is quite another matter to be able to extract them.
If the source rock becomes exposed at the surface, then it can
simply be dug up to get at the oil within it. This is the only easy
way of extracting non-liquid hydrocarbons (kerogens), but the
environmental costs of this sort of quarrying are so high that
most oil shales, as they are known, have not yet been exploited.
Liquid hydrocarbons are almost always extracted by drilling.
You might think that oil and gas wells are made by drilling
into source rocks. This is not the case, because although shales
216

are porous and can contain 10 per cent or so by volume of
hydrocarbons, they are not permeable. That is to say, the pore-
spaces are not sufficiently well connected to allow fluids to flow
through the rock at the rate required to feed a well.
With the notable exception of fracking (see the case study
below), most hydrocarbons are extracted not from their original
source rock but from a reservoir rock into which they have
migrated over geological time. A reservoir rock must be both
porous and permeable. Sandstones usually fit the bill, unless
most of the pore-space has been filled by cement, and these are
the most common reservoir rocks for hydrocarbons. A rubbly
limestone, formed in a reef, may make a good reservoir rock,
too, especially if groundwater has previously dissolved some of
the rock, thereby increasing its porosity and permeability.
Remember this!
Porosity is a measure of the proportion of a rock’s volume that is
unoccupied by rock, and so potentially available to be filled by a fluid.
Permeability is a measure of how freely this fluid can flow, which
depends on the degree of connectivity between pore-spaces.
In order for hydrocarbons to be held within a reservoir

rock, there must be a nearby source rock from which the
hydrocarbons have migrated, and the reservoir rock must have
remained below the temperature at which the hydrocarbons
break down. Once they have found their way into a potential
reservoir rock, oil and gas tend to be displaced upwards by pore
water, which is denser than oil. A final condition, therefore, is
that the hydrocarbons must somehow be trapped within the
reservoir rock, otherwise they would have long since seeped out
to the surface.
There are essentially two kinds of hydrocarbon trap. Structural
traps are when the reservoir rock is folded or lies next to a fault
in such a way that the hydrocarbon-bearing part is bounded
by impermeable rocks, so the hydrocarbons cannot escape
(Figure 11.3(a) and (b)). Stratigraphic traps occur below an
unconformity (Figure 11.3(c)).

impermeable caprock
water-saturated
gas water-saturated impermeable caprock
oil reservoir rock
reservoir rock
gas
oil
(a) (b)
impermeable caprock gas

oil
water-saturated
reservoir rock
(c)
Figure 11.3 Cross-sections to show oil traps: (a) a structural trap in

the form of an anticline or a dome; (b) a structural trap against a fault;
(c) a stratigraphic trap below an angular unconformity. In all three cases,
impermeable rock prevents hydrocarbons from escaping through the sides
and top of the trap. The areas labelled ‘oil’ and ‘gas’ show where
the reservoir rock is saturated with these.
Once drilling has reached the appropriate part of a hydrocarbon

reservoir, oil or gas may be driven to the surface merely by
pressure. Usually about 30–40 per cent of the oil may be
recovered this way. Most of the remainder can be extracted
by displacing it by pumping water or gas down a suitably
positioned borehole.
Key idea: Exploitable hydrocarbons

Hydrocarbons (petroleum oil and natural gas) are produced by
burial of organic matter, under the right circumstances. In order
to be exploitable by traditional means, they must usually have
migrated into some kind of trap within a reservoir rock.
218

The pressure inside an oil reservoir was demonstrated with
lamentable consequences in 2010 when an explosion destroyed
the Deepwater Horizon drilling rig in the Gulf of Mexico.
Pressurized oil gushed out for many weeks at a rate of tens of
thousands of barrels per day, defying attempts to cap the well
for several months and spawning vast claims for compensation
and damages amounting to tens of billions of dollars.
Spotlight: Barrels of oil

Oil, of course, is no longer shipped in barrels, but the ‘barrel’
remains the conventional unit for quoting oil volumes and prices.
A barrel of oil is equivalent to 42 US gallons (35 imperial gallons),
which is 159 litres. It probably derives from wine, whiskey and even
fish barrels of similar size used in the Pennsylvania oil fields in the
1860s, though the size was not standardized to 42 gallons until 1872.
Case study: Fracking

Hydraulic fracturing, usually abbreviated to fracking, has been
used commercially to increase the recovery of hydrocarbons from
conventional reservoirs since 1949. As the name suggests, it is
based on pumping pressurized water down a well to fracture
the rock at depth (thereby improving permeability). Sand grains
are added to the injected water to keep the induced fractures
open. Nearly 50 years later, fracking was adapted to extract
hydrocarbons (principally gas) still trapped in fine-grained source
rock in Texas, and is now applied in this manner on a vast scale in
the USA, Canada and China.
The use of fracking for extracting ‘shale gas’ in Europe is

controversial. It was banned in France in 2011. The UK government
supports fracking, but the ‘green lobby’ is opposed, even though
gas burns more cleanly than coal and well-heads are less visually
intrusive than wind turbines. Objections to fracking include fear
of pollution of groundwater (there were some cases in the USA,
associated with poorly regulated shallow fracking), concerns over
the amount of water that has to be pumped below ground, and fear
that fracking will cause earthquakes (it doesn’t, if you do it carefully).

The UK’s resources of shale gas are estimated at 500 years’
worth of supply. If only 10 per cent turns out to be economically
extractable, that’s still 50 years’ worth of gas. This would free the
country from reliance on supplies from Russia, which some would
argue to be a prudent step.
Exploration for oil and gas

Oil wells are usually 1–5 km deep. It is very expensive to
drill that deeply, so drilling is not the major technique used
in exploring for hydrocarbons. Sometimes, interpretation of
the surface geology may suggest that an oil trap exists at an
appropriate depth, but often these structures are not apparent
at the surface. In such a situation, initial exploration relies on
geophysical techniques such as mapping anomalies in the local
magnetic and gravity fields.
When an area appears promising, it is worth making a seismic
study. This involves using explosions to generate seismic waves,
which are picked up by an array of sensors. Seismic waves
generated this way are not as powerful as those from a major
earthquake, but they have the advantage of providing a sharp
signal, generated from a chosen point at a known time. By
interpreting the reflected seismic signals from successive rock
layers, a picture of the sub-surface can be built up.
Surprisingly, seismic surveys are most easily done under water.
A ship trails a kilometre-long array of sensors to record
pressure variations in the water, and periodically generates an
underwater explosion using a compressed air gun. Part of the
energy is reflected back by the sea floor (this is the principle of
echo-sounding to determine water depth), but the rest continues
into the sea-floor sediment, and a portion of it is reflected back
at each interface between layers.
Only when a seismic survey has revealed an anticlinal or domal
structure, or some of other kind of potential hydrocarbon
trap, might it worth drilling an exploration hole. In addition to
attempting to reach the hydrocarbons, much can be learned by
studying rock chippings deliberately flushed out of the borehole.
220

Various electronic, magnetic and radioactive sensors can be
deployed down the hole to record ‘wireline logs’, giving estimates
of the density, porosity, composition and even the dip of bedding
in each stratum that is passed through. Only when a successful
exploration well has been drilled, and the extent and capacity
of the reservoir have been ascertained, is it worth sinking
production wells. These may be drilled at any angle. Especially
for offshore production, a production platform may extract
hydrocarbons from a three-dimensional fan of production wells
that penetrate the reservoir over an extensive area.
The origin of coal

The other great fossil fuel is coal. Like the hydrocarbons, coal
derives from dead organic matter, but in this case the source
is mainly land plants. Several criteria must be met for coal to
form.
33 First, there must be abundant land plants living (and
dying) nearby. Coal is unknown in rocks older than about
425 million years (which is when the first land plants
appeared). It is particularly abundant in successions formed
at certain times between about 350 and 50 million years
ago – when climate, the arrangement of continents, and the
evolutionary state of land plants combined to favour its
formation (Plate 11).
33 The second requirement (as for hydrocarbons) is that the
dead plant material should accumulate in an environment
lacking oxygen (‘anoxic’) to prevent oxidation. Swamps meet
these first two criteria very well, having abundant vegetation
and a stagnant water table coinciding with the sediment
surface. The organic-rich soil here is described as peat.
33 The next criterion for an economic coal deposit is that there
should be rapid subsidence to allow a sufficient thickness to
accumulate. Ten metres of peat compacts down to only one
metre of coal.
33 The final requirement is that the deposit should be buried to
hundreds of metres to enable the peat to be turned into coal.

All four requirements are met in delta environments, where (as you
saw in Chapter 9) a swampy delta top (where peat can accumulate)
is continuously subsiding under the weight of sediment. If the main
distributary channel of the delta switches, the old delta top may
sink and become buried by marine or delta-front sediments
(Figure 9.11). The major coalfields of Europe, Asia and North
America formed in this way about 300 million years ago, in the
latter half of the period of geologic time called the Carboniferous
(called ‘Pennsylvanian’ by North American geologists).
The quality or ‘rank’ of coal depends mainly on how deeply it
was buried during diagenesis. Under anoxic conditions, dead
plant material is broken down by bacterial action to a woody
residue called lignin. With progressive burial, volatiles such
as water and methane are driven off, the deposit becomes
compacted, and the amount of energy that can be released by
burning a given weight of material increases. Peat has less than
50 per cent carbon, but the highest rank of coal, anthracite, is
over 90 per cent carbon, and (weight for weight) contains about
three times as much energy. Apart from the rank of coal, other
factors determining its usefulness include the impurities within
it. Clay minerals leave a solid residue (ash) after combustion,
iron sulfide produces the acid gas sulfur dioxide during
combustion, and sodium chloride corrodes boilers.
Remember this!
Most coal formed from dead plant material that accumulated
thickly in subsiding swamps.
Mining techniques
Coal can be extracted either by deep mining or by strip mining.
The latter (Figure 11.4) is economic for shallow deposits, but
becomes prohibitively expensive (and more environmentally
damaging) as the thickness of overburden that must be removed
to get at the coal increases. Strip mining is the favoured
technique in the great US coalfields of Pennsylvania.
In deep mining, a shaft is sunk to the level of a coal seam that is
thick enough (typically 0.5–3 m) to be worth exploiting. Coal
222

overburden restored
(yet to be replaced overburden
topsoil topsoil
stripped) (awaiting landscaping)
unmined seam
Figure 11.4 Area strip mining of a shallow coal seam, showing how the
overburden that has to be removed to work the active face can be dumped
behind. This obviates the problem of how to dispose of it and (provided
the topsoil has been kept separately) the landscape can be restored to
something like its original form. The same technique was historically used
to mine shallow ironstone seams in England.
was formerly removed by men with pickaxes, but nowadays

mechanical cutters are used. These make it cheaper to exploit
thick seams, but have made it uneconomic to work thin coals
underground. There are essentially two approaches to mining
coal or any other sheet-like deposit (such as evaporite minerals).
Each is dictated by the need to prevent the roof collapsing on to
the miners and equipment.
33 The simpler and more traditional method is pillar-and-stall
working, in which only about 50–85 per cent of the deposit
is removed, leaving the rest as pillars to support the roof
(Figure 11.5a). It is a matter of fine judgement how thick and
how closely spaced the pillars need to be if they are to bear
the weight of the roof; the thicker and more numerous the
pillars, the smaller the fraction of the deposit you can extract.
33 The alternative method is longwall mining, in which an
underground face is created that may be several hundreds of
metres long. A cutter operates along the seam beneath a roof
supported by hydraulic jacks, and these are slid forward as the
cutter passes, allowing the roof to collapse behind it (Figure
11.5b). Access is maintained by a tunnel joining each end of
the face to the mine haulage roadways. As the face advances,
the ground surface subsides, the effects of subsidence being
more extreme if the mine is shallow. However, this method
allows the whole of a seam to be extracted.

Longwall mining is the main method of extracting coal in
Britain’s few remaining coal mines, and in India, the world’s
third-largest producer of coal. In North America, where the USA
is the world’s second-largest producer of coal, pillar-and-stall
working is favoured (there called ‘pillar-and-room’), although
pillars are usually removed when the rest of the seam has
become exhausted, allowing the roof to collapse in a controlled
manner (as in longwall mining).
In 2013 China accounted for 45 per cent of the world’s coal
production, but such is its rate of industrial growth (demanding
coal-fired electric power stations) that since 2007 it has been a
net importer of coal. Most Chinese coal is obtained by deep-
mining using the longwall method. Risks are clearly taken,
because deaths per tonne of coal are dozens of times greater
than in the USA. China’s coal mining deaths peaked at nearly
7000 in 2002, but had fallen to just over 1000 in 2013.
(a) pillar
ground subsidence
(b)
collapsed roof cutter coal seam
Figure 11.5 Techniques for mining a seam of coal, or similar valuable

stratum. (a) Pillar-and-stall working. This works only if the roof is a strong
rock type such as sandstone or limestone. If the roof is weak shale it is
prone to collapse, unless the pillars are very closely spaced. (b) Longwall
mining, seen looking along the face that is being worked. The cutter is
moving into the page (i.e. away from us) and the roof is allowed to collapse
behind it. When it reaches the end of the face, the cutter will be moved to
the right and then begin to work its way back towards us.
224

Key fact: Coal mining deaths
In 2008 in China, one miner died for every 0.7 million tonnes of coal
extracted, whereas in the USA one miner died for every 25 million
tonnes of coal.
Ores and their origins

Metals are extracted from ores. Ore is a term referring to any
rock containing a sufficient concentration of a metal for it to
be extracted economically. The concentration of metal (the
ore’s ‘grade’) necessary to make it economic depends on the
value of the metal and the cost of extracting it. Iron is abundant
and cheap, and at present it is not worth extracting from ores
containing less than 50 per cent iron. It uses a lot of energy to
isolate aluminium from its ores, so it is not usually economic to
work ores of less than 30 per cent aluminium. Lead is worked
from ores whose grade is as low as 5 per cent; copper from ores
around 1 per cent grade or less; silver from ores of 0.01 per cent
(100 parts per million) grade; gold from ores of one to ten parts
per million grade; and platinum from ores as low as 0.1 parts per
million grade. When several metals occur in the same ore, it can
be profitable to exploit it for the combination of metals, although
it would have been uneconomic to extract just one of them.
Key idea: Ore

Ore is a general term for any rock containing a sufficient
concentration of a metal for it to be extracted economically.
Most ores result from igneous activity. Sometimes economic

concentrations are formed within a magma body when heavy
crystals such as chromite (an iron-magnesium-chromium oxide
that is the main ore mineral of chromium, and an important
source of platinum and related metals that it can contain in trace
quantities) or nickel, iron and copper sulfides sink to the bottom.
A prime example of this type of differentiated igneous ore body is
the 200-km-wide Bushveld Complex to the north of Johannesburg,
South Africa, which contains enormous reserves of chromium, iron,

platinum and gold. One of the most famous, and largest, nickel
ore bodies is the 50-km-long Sudbury igneous complex in Ontario,
Canada. Interestingly, this is thought to have originated not as the
ordinary sort of crustal melt, but as a melt formed when a 10-km-
diameter asteroid hit the Earth nearly 2 billion years ago.
Other ores form not within an igneous body itself but in the
surrounding rocks. These are a result of hot watery fluids
circulating between a cooling intrusion and the surrounding
rocks. The process is one of convection, driven by the heat
of the igneous body, and is described as hydrothermal (‘hot
water’) circulation. Essentially, what happens is that metals are
dissolved out of the intrusion by the hot fluids and deposited
nearby, usually as sulfides and sometimes as oxides. The major
‘porphyry copper’ ore bodies in the Andes are of this type, as
are the tin ores in Cornwall, England (where parts of the granite
have been altered to kaolinite or china clay by the hydrolysis
reactions described in Chapter 8).
A few metals, notably gold and silver, are chemically so inert
that they rarely form compounds but occur instead as pure
(‘native’) metal, usually as tiny flecks rather than in the chunky
nuggets of popular myth. If the circulating fluid passes along
fractures, precipitation may occur there, so that the ores
form within veins rather than being dispersed throughout the
neighbourhood. Such fractures are usually filled mainly by
undesirable minerals like quartz, forming a vein that is mostly
useless material but will be economic to mine for gold, provided
it contains more than about ten parts per million.
Spotlight: Nuggets
The largest gold nugget ever found, known as ‘Welcome Stranger’,
was found near Dunolly, Victoria, Australia in 1869. It weighed 72 kg
and soon ended up as ingots in the Bank of England. The largest
surviving nugget was found by a metal detectorist, also in Victoria,
in 1980, and is now on display in the Golden Nugget Casino in Las
Vegas. Any large nugget is worth more than its weight in gold; in
2011 a 2.8-kg nugget from California was sold at auction for three
times what it would have fetched had it been melted down.
226

The origin of hydrothermal ore deposits associated with
igneous intrusions is summarized diagrammatically in
Figure 11.6. Other hydrothermal ores occur at ocean-
spreading axes at vents, such as black smokers (Chapter 5).
At present it is far too expensive to mine these on the
sea floor, but they have been mined in ophiolites, notably
on Cyprus, which was the main source of copper for the
classical world (Plate 12).
surface
1 km approx. disseminated ore
vein ore
igneous hydrothermal circulation

intrusion
Figure 11.6 Cross-section showing the precipitation of ore bodies from

circulating hydrothermal fluids. The convection system is driven by the
heat of the igneous intrusion.
Smelting copper ore in Cyprus in ancient times has left behind

large heaps of slag (the worthless residue from which the
extractable metal has been removed). Measuring the volume
of the slag heaps allows us to work out how much wood must
have been burned, and it turns out that the island’s forests
would have had to have been felled several times over to feed
the furnaces. Cyprus was able to become rich on copper only
because it had a climate suitable for the rapid regrowth of
timber.

Some hydrothermal ore deposits are nothing to do with
igneous intrusions at all. For example, the lead and zinc ore
bodies in the Mississippi River basin, in Ireland and in the
Pennines of England occur within limestones, and appear to
have precipitated from warm mineralized fluids expelled from
deeply buried sandstones or shales. Warm groundwaters are
also suggested to explain the concentration of uranium into
ore bodies in the Athabasca Basin of northern Saskatchewan,
Canada, and in the Northern Territory of Australia.
Ore exploration
Exploration for ore bodies can be done in many ways. If the ore
body extends to the surface, it can be discovered by an old-style
prospector armed with a pickaxe and good eyesight, although
this is now rare. Even if the economic part of an ore body is
not exposed, hydrothermal circulation may have bleached
or otherwise discoloured the overlying rocks at the surface.
Surface effects may also be discovered by analysis of visible
and infrared images recorded by satellites or airborne surveys.
Other exploration techniques include mapping the magnetic
field, often by means of an airborne survey, and looking for
anomalous magnetic features that may be caused by metallic
ores. When a potential ore body has been located, it must be
drilled to assess the grade of the ore and confirm its extent.
Only about one in a thousand promising discoveries has all the
attributes required to be turned into a successful mine.
Key idea: A bit expensive?

Drilling is very expensive and is usually the last stage of
exploration for any buried resource.
The mining of ores

Deep mining is the standard method for most ore bodies
(Figure 11.7) but hydrothermal copper deposits tend to be so
extensive that they are usually exploited by surface mining,
also known as quarrying or ‘opencast’ working (Plate 12).
228

waste fill lifting gear
disused adit
vertical
shaft
unbroken ore
collapsed area
broken ore
haulage
level
Figure 11.7 Cross-section through a deep mine, exploiting an inclined ore

body. An adit is a horizontal or gently inclined tunnel, used in this example
to gain access to the upper part of the ore body, which has now been
exhausted and back-filled by waste. Ore is now being extracted through
deeper tunnels and raised mechanically up a vertical shaft.
The deeper the hole grows, the more expensive this technique
becomes, because the sides of the quarry must be expanded
to prevent the walls from collapsing. The only effective means
to ensure this is to cut the walls back into a series of steps.
Thus a progressively greater ratio of unwanted ‘overburden’
must be removed every time the floor of the quarry is lowered
(Figure 11.8).
Key idea: Mining costs

Deep mining is much more expensive than quarrying (surface
mining) or strip mining.
Diagenetic and sedimentary

ore bodies
Not all ores are formed as a result of hot fluids. Some form
during low-temperature diagenesis of sedimentary rock, if the
chemical conditions in the permeating water are correct. A good
example is provided by the ironstones of Northamptonshire

waste tip
(a) (b)
ore
bench
(c) (d)
Figure 11.8 Progressive stages in ‘surface mining’ or quarrying of an ore

body seen in cross-section, showing how the sides of the quarry are cut
into a series of steps or ‘benches’ for stability. Each time the quarry is
deepened, a greater amount of overburden must be removed. Eventually it
becomes uneconomic to continue. In this example, before the quarry can
be deepened beyond stage (d) the waste tips will have to be moved.
and Cleveland in England. These are shallow marine sandstones

of Jurassic age that are cemented by iron carbonate (siderite
FeCO3), and were sufficiently easy to access and of sufficient
grade to have been an important ore until the latter part of the
twentieth century.
Today, most of the global production of iron comes from
another type of sedimentary deposit, consisting of alternating
layers of silica and iron oxides and known as banded iron
formations, or BIFs. Virtually all BIFs are between 2.5 and
1.8 billion years old, and are believed to have formed on the
floors of shallow seas when (as you saw in Chapter 2) there
was much less free oxygen around, but sufficient to allow iron
to bond with dissolved oxygen to form insoluble oxides. BIF
ores have been mined in the Lake Superior district of the United
States since the nineteenth century. More than 15 per cent of
the global production of BIF iron is from Australia, notably the
Hamersley Basin in the north of Western Australia where it is
mined in huge open pits.
230

Other ores are an end product of chemical weathering. The
outstanding example of this is bauxite, the only globally
significant ore of aluminium. This is produced in humid tropical
conditions when hydrolysis gets to work on clay minerals. For
example, kaolinite (produced by the chemical reaction shown in
Chapter 8) breaks down to bauxite as follows:
Al2Si2O5(OH)4 + 5H2O = 2Al(OH)3 + 2H4SiO4

kaolinite (a water bauxite ore a form of dissolved
clay mineral) silica
Australia, China, Brazil, Guinea, India and Jamaica were

the world’s six major producers of bauxite ore in 2014. The
only other important ores produced in this way are of nickel,
found as insoluble silicate residue in soils formed by humid
tropical weathering of olivine-rich igneous rocks. About a
third of the global production of nickel comes from such
sources (notably from New Caledonia, Indonesia, Cuba,
Brazil and the Dominican Republic), but the majority is from
sulfide ores associated with ultrabasic lavas and the Sudbury
asteroid impact.
The remaining class of ores is placer deposits. These occur when
heavy, metal-rich grains become concentrated by sedimentary
processes. Gold is the most obvious example. All too often, in
‘them there hills’ where the gold is supposed to be, it occurs
as tiny flecks in veins far too diluted by barren quartz to
be worth mining. However, streams draining this area (and
therefore transporting locally derived sedimentary grains) can
act to concentrate the gold. This is because grains of gold are
considerably denser than silicate sand grains and will tend to
lag behind in the bed-load. When panning for gold, a prospector
tries to concentrate the gold still further, by swirling a few
handfuls of streambed sand or gravel in a basin, hoping to wash
away the silicate particles and leave the gold behind.
Most commercial gold production comes not from present-day
streams but by mining ancient placer deposits. The world’s most
prolific gold-producing area is the Witwatersrand Basin in South
Africa, where gold is found in pebbly and gravelly sediments

that appear to have been deposited by braided rivers about
2.5 billion years ago. The Witwatersrand conglomerates are
notable also for containing placer grains of uraninite, an oxide
of uranium (UO2).
Remember this!
All banded iron formations (BIFs) and placer deposits of uraninite
are billions of years old. They could not form today because there
is now too much free oxygen available in the environment.
Non-metallic raw materials

Some non-metallic raw materials are sufficiently valuable to be
mined underground like ores. These include fluorspar (CaF2),
which is the main source of fluorine for the chemical industry
and is also needed for aluminium smelting and as a flux in
steel making; and various evaporite minerals such as gypsum
(CaSO4.2H2O), which is used in plaster and as a filler for textiles;
and potash, a general name for potassium salts (e.g. KCl and
K2SO4.2MgSO4), which is used in fertilizers and other chemicals.
Spotlight: Salt and sweat

Half the UK’s annual production of potash is mined by the pillar-
and-stall method at depths of 1.2–1.5 km in Europe’s second-
deepest mine at Boulby in Yorkshire. At that depth the temperature
exceeds 40 °C, and the miners are reputed to wear little other than
a hard hat, boots and boxer shorts.
Modern industrial society requires other materials in bulk too.

The most obvious are: clay and shale for bricks; limestone for
cement, concrete and roadstone; sand and gravel for concrete,
roadstone and various construction purposes; and crushed hard
rock in general for roadstone. These are extracted by quarrying.
Sometimes a happy occurrence of mudrocks and coal seams gives
two economic products from a single operation (Figure 11.9).
232

Figure 11.9 An opencast clay pit in County Durham, England. Deltaic
mudrocks and silts are extracted and mixed in the right proportions to
make bricks. Coal is a useful subsidiary product from this site. The digger
is standing on the top of a coal seam, and another seam can be seen about
4 m up in the far wall. The right foreground has been back-filled.
Making good the damage

In Britain, sand and gravel are obtained most readily from
shallow quarries on the flood plains of lowland rivers. Large
quantities of these materials were deposited at the end of the
last glaciation, when rivers were transporting a much greater
bed-load than at present. After use, these gravel pits usually
become flooded, and may be used for watersports or turned into
nature reserves.
Deeper disused quarries and mines pose more problems. Many
quarries are used as ‘landfill’ sites, for disposing of domestic and
industrial refuse. With sensitive landscaping, an infilled quarry
may blend perfectly into the surrounding countryside, though
it annoys geologists if no rockfaces are preserved. Controls are
required to limit what types of waste may be buried in specific
localities. For example, if the quarry is in a permeable limestone,
nothing should be buried there that might contaminate the
groundwater.

Old mines are rarely backfilled, so the most lasting legacy of
mining on landscape may be subsidence. Above shallow mine
workings, buildings and roads on the surface may become
damaged centuries after a mine was abandoned. Mines are
less of an eyesore than quarries, especially if the heaps of mine
waste or ‘spoil’ are sculpted to blend into landscape. They can
be dangerous if left in place anyway, as was tragically shown
in the Welsh village of Aberfan in 1966 when a part of a badly
situated and water-saturated colliery spoil tip collapsed and
killed 142 people, most of them children in a nearby school.
Water resources
One other great natural resource that is taken from the ground
in many places is water. Water may fill the pore-spaces of
permeable rocks, and can be pumped from wells drilled into
these natural underground reservoirs. If the permeable layer is
capped by an impermeable layer, the groundwater may be under
sufficient pressure to rise to the surface without pumping, in a
so-called artesian well (Figure 11.10). The fountains in London’s
Trafalgar Square were originally supplied from an artesian well,
rainfall
impermeable layer
permeable layer artesian well
(aquifer)
Figure 11.10 Cross-section showing how a permeable layer (known as an

aquifer) soaks up water where it is exposed at the surface. This water is
prevented from escaping by impermeable rock above, but in this case the
hydrostatic head of pressure will drive water to the surface if a suitably
sited well penetrates into the aquifer.
234

and many of the oases in the Sahara desert are where artesian
water in a sandstone layer reaches the surface.
Groundwater is essentially a renewable resource because the
groundwater reservoirs are recharged by rainfall. However,
in dry regions such as Arabia and the southern United States,
groundwater is being extracted much faster than its recharge rate.
Perhaps the most priceless natural resource of all that the
Earth has to offer is living organisms. Traces of life can be
found in many ancient rocks. You have already seen how coal
and hydrocarbons are made from organic remains, and how
limestone is usually made from the shells of living organisms. In
the next chapter you will look at some of the more interesting
or important organisms that can be found fossilized in rocks,
and also look at the history of life on Earth.
Dig deeper
Annual statistics (and commentary) on global production of
metals and other commodities extracted by mining can be
found at
http://minerals.usgs.gov/minerals/
Similar information about coal, oil and gas production can be
found at http://energy.usgs.gov
A study published in 1972 called ‘The Limits to Growth’ that
predicted catastrophic economic collapse in the short term,
partly because it misunderstood the relationship between
reserves and resources, can be downloaded from
http://www.donellameadows.org/the-limits-to-growth-now-
available-to-read-online/ It was wrong in the short term, but in
the longer term it may be uncomfortably close to the truth.
To help you make up your own mind about fracking (and I think
you should), some pros and cons are discussed at
http://www.natureworldnews.com/articles/9011/20140912/the-
pros-and-cons-of-fracking.htm

A well-balanced article from the Guardian newspaper is here:
http://www.theguardian.com/commentisfree/2014/jul/29/anti-
fracking-battle-wind-emissions-nimbys-green-groups
The British Geological Survey’s discussion of shale gas
reserves is at http://www.bgs.ac.uk/research/energy/shaleGas/
home.html
… and the Geological Society of America’s analysis is at
http://geosociety.org/criticalissues/hydraulicFracturing/index.asp
236

Fact-check
1 Which is the correct definition of ‘reserves’ in the context of a
country’s natural resources?
a Areas where both exploration and extraction are prohibited
b Areas where extraction but not exploration is prohibited
c The amount that could be profitably and legally extracted
d The total amount that could be extracted regardless of price
2 How is porosity defined?
a The average size of the pores in a rock
b The fraction of a rock’s volume that is occupied by air or
other fluids
c The fraction of pores in a sedimentary rock that have
become filled by cement
d The rate at which fluids can seep through a rock
3 Which of these are hydrocarbons?

a Butane
b Coal
c Crude oil
d Methane
4 If a hydrocarbon source rock is buried at a temperature of

about 170°C for ten million years, what hydrocarbons would it
be likely to contain?
a Mostly oil
b A mixture of oil and gas
c Gas only
d None: all the hydrogen would have been driven off, leaving
only graphite behind
5 What would be a reasonable depth for an oil well?

a 300 m
b 3 km
c 10 km
d 30 km

6 What is the origin of most coal?
a Kerogens buried at a sufficiently high temperature for long
enough to drive off all the oxygen and hydrogen
b Dead plant material accumulated on a swampy delta top
c Thick accumulations of marine plankton that were buried
by an advancing delta
d A prairie fire
7 What is strip mining?

a An underground mining technique to extract a whole seam
of coal
b Removing a shallow seam of coal after stripping away the
overburden
c Tunnelling along an ore-bearing vein
d What miners have to do when it is very hot underground
8 What is the origin of hydrothermal ore deposits?

a Rain falling into a volcanic vent
b Rain falling on to an active lava flow
c Groundwater penetrating into a magma-filled chamber
d Groundwater circulating in and out of a cooling igneous
intrusion
9 Why is no iron produced from banded iron formations less than
a billion years old?
a The maturation time would be insufficient
b They have not been buried deeply enough
c They are protected by international treaty
d The composition of the atmosphere became unconducive
to their formation
10 What is an artesian well?

a One designed by a master craftsman
b One in which the pressure is sufficient to drive the water
to the surface
c One producing a brine which a potassium salt called
artesia is dissolved
d One that is dry for part of the year
238

12
Past life and
fossils
33about the history of life on Earth,
including mass extinction events
33how the remains of dead plants and
animals can become preserved as fossils
33about the kinds of information fossils
can give to the geologist.
239

For many people, fossils and geology are inseparable. There is
no denying the thrill of discovering the remains of an organism
that lived hundreds of millions of years ago, and that you are
the first person to see. The study of fossils has taught us most
of what we know about the history of life, and provided much
of the original evidence in support of the theory of evolution.
Abundant additional evidence for evolution comes from
molecular phylogeny, which looks at the degree of similarity
between genetic material in different kinds of organism.
Another reason why fossils are important is that signs of life
in sedimentary deposits, which can be either the remains of the
organisms themselves or traces such as burrows and the like,
can be at least as valuable as bedforms, grain size, and so on,
in helping you determine the nature of a past environment.
Moreover, whereas most characteristics of deposits in any
particular environment are the same regardless of age,
identification of varieties of fossil organisms provides the single
most useful way of dating a bed of rock. This is because many
species have survived for only limited periods, no more than a
few million years in the most useful cases.
Key idea: Biostratigraphic dating

Fossils can be used to determine the age of sedimentary rocks in
which they are found.
The origin of life

The presence of life on Earth used to be regarded as a great
marvel. However, some scientists now maintain that it would
be a surprise to encounter an Earth-like planet where life had
not developed. The chemical elements required for life can be
found in abundance in the oceans, and even more abundantly
in hot fluids escaping from hydrothermal vents such as black
smokers (see Chapter 5). Experiments in the 1950s showed
that amino acids (from which proteins can be built) can be
made by passing electric sparks through a mixture of ammonia,
methane and water. Lightning discharges into the Earth’s early
oceans could therefore have begun the sequence of chemical
240

processes that ended up with life. Amino acids have also been
found in carbonaceous meteorites and comets, so an alternative
possibility is that these building blocks of life were delivered
by rocky or icy debris falling on to the still-cooling and sterile
surface of the young Earth.
There are many steps on the road from simple amino acids
to complex molecules capable of self-replication (such as
DNA) and thence to cellular life. However, geological time
is so long and the number of molecules in the oceans so vast
that there is no need to call on divine intervention to account
for how life began, and since it began on Earth it is equally
likely to have begun in hordes of suitable places around other
stars. This is all speculation until life, probably some sort
of microbe, is found on (say) Mars or a moon of Jupiter or
Saturn. Even then, we would need to establish that this life had
an independent origin, rather than hitchhiking from world to
world on debris ejected by collisions.
The first microbes

It seems that life on Earth began maybe as long as 4 billion
years ago, most likely in the form of microscopic single-celled
organisms (microbes), like bacteria or a less familiar group
called archaea, on the fringes of underwater hydrothermal vents
where they would have been nourished by the abundant supply
of chemical energy. Chemical reactions occurring across the
skin of tiny bubbles could have led to the origin of the first cell
membranes. There are varieties of microbes still around today
that obtain their energy chemically and can survive only in
oxygen-free conditions. These are probably little changed from
the primitive cells that developed beside those ancient hot vents.
Spotlight: Oxygen intolerance

Today, oxygen-intolerant microbes are most noteworthy for their
function in breaking down raw sewage, a rather unprestigious role
for the closest relatives of Earth’s first living organisms.

After bacteria-like organisms had arisen, eventually new
varieties evolved and migrated to less chemically rich
regions where energy had to be extracted from sunlight (by
photosynthesis). The simplest photosynthetic organisms today
are the cyanobacteria, formerly called blue-green algae. As you
saw in Chapter 2, when these organisms spread and multiplied
they also began slowly to change the chemical balance of the
atmosphere. In particular they liberated oxygen, which was at
only 1 per cent of its present level about 2 billion years ago.
The earliest visible traces of life that can be found in rock
are cells and filaments of cells, found as fossils in chemically
precipitated chert and carbonate deposits. Radiometric dating of
associated volcanic rocks shows these to be 3.5 billion years old.
These fossils are only about a hundredth of a millimetre across,
and so can be discovered only by painstaking microscope
surveys on thin, translucent slices of rock. Some scientists
contend that these structures were produced chemically rather
than biologically, but all agree that if they do represent real cells
then they can only have been very simple, like cyanobacteria,
lacking a nucleus and other internal structures.
Less contentious fossil structures large enough to see with the
unaided eye date back almost as far. These are metre-scale
mat-like or dome-like laminated features within calcareous
sediments in the Pilbara region of Western Australia, thought to
be between 3.4 and 3.5 billion years old. They resemble modern-
day stromatolites, sometimes known as algal domes, which are
held together by strands of cyanobacteria and survive only where
extreme conditions, such as high salinity, protect them from
animal grazing. Fossil stromatolites became abundant about 2.3
billion years ago, and remained so for nearly 2 billion years until
their decline, presumably as a result of grazing by animals and
competition with more advanced plants.
Remember this!
The oldest known fossils are about 3.5 billion years old, though life
may have begun as long as 4 billion years ago.
242

More complex cells
The next step up the evolutionary ladder from microbes and
algal mats came with the appearance of more advanced cells
of the kind known to biologists as eukaryotes, to which plants,
animals and fungi belong. In contrast to archaea (from which
they are probably descended) and bacteria, which are both
described as prokaryotes, eukaryotes have thick cell walls, a
nucleus, chromosomes, and various other internal structures.
Bizarrely, it seems likely that the internal complexity of
eukaryote cell structure derives from smaller microbes having
taken up residence inside larger ones. The earliest known
eukaryote fossil cells are about 2 billion years old, and are
probably single-celled planktonic plants that lived in the sunlit
upper waters of the ocean.
Multicellular organisms
Fleshy multicellular marine algae (what you might call
‘seaweed’) appeared about 1.2 billion years ago, but the first
multicellular animals do not appear in the fossil record until
approximately 900 million years ago. These had soft, worm-
like bodies and so are imperfectly preserved as fossils. It is
not until 600 million years ago that a wide variety of forms
emerged, in both shallow and deep-water environments. These
were still soft-bodied organisms, but an important change
happened 540 million years ago, when animals developed
hard parts (such as shells) that do not decay upon death and
which therefore fossilize easily. From then onwards the fossil
record is rich, diverse and dominated by remains of animals
with hard parts.
Key idea: Milestones

Significant steps in the evolution of life are: the first eukaryotic
cells (about 2 billion years ago); the first multicellular algae (1.2
billion years ago); and the first multicellular animals (0.9 billion
years ago). These were all marine.

The interval of time since 540 million years ago has long
been known to geologists as the Phanerozoic Eon. If you are
unfamiliar with the divisions of geological time, you may find it
helpful to refer to Appendix 3. Phanerozoic means ‘visible life’,
and was so named because it is not until the beginning of the
Phanerozoic, during the period known as the Cambrian, that
sediments were deposited in which fossils are abundant and
easily recognized.
Remember this!
The three successive eras into which the Phanerozoic Eon is
divided are the Paleozoic (meaning ‘ancient life’), Mesozoic
(meaning ‘middle life’) and Cenozoic (meaning ‘recent life’). The
Cambrian is the first of six subdivisions (periods) into which the
Paleozoic is divided.
The time prior to the Cambrian is often informally described as

Precambrian. More formally, it is divided into three eons: the
Hadean Eon from the Earth’s origin until 3.9 billion years ago;
then the Archean Eon until 2.5 billion years ago; and finally
the Proterozoic Eon extending from the end of the Archean to
the start of the Phanerozoic. ‘Proterozoic’ means ‘earlier life’,
though as you have seen, the earliest life has now been traced
way back into the Archean.
The major animal phyla appear

Most of the major divisions (phyla) of the animal kingdom
arose during or soon after the Cambrian. These include:
1 two phyla with shells made of calcium carbonate (Figure 12.1),
namely molluscs (some with two shells and some with one
shell) and the less familiar brachiopods (two-shelled organisms,
but with body plans entirely different from molluscs)
2 a phylum known as the arthropods, with jointed external
skeletons (exoskeletons) made of a resilient organic material
called chitin, most famously represented in the fossil record
by trilobites (Figure 12.2) whose present-day distant relatives
include crustaceans (e.g. crabs and lobsters), spiders and insects
244

(a) (b)
1 cm
(c) (d)
Figure 12.1 Bivalve (a) and ammonoid (b), two distinct classes of the
mollusc phylum. Two brachiopods (c) and (d), which, although at first sight
similar to bivalve molluscs, belong to a different phylum.
1 cm
Figure 12.2 Two trilobites, representatives of a kind of arthropod that first

appeared in the Cambrian and became extinct during the Permian. These
two examples show the complete articulated exoskeleton, except that the
fragile antennae and legs are missing. Compound (insect-like) eyes can be
seen on either side of their heads.

3 a phylum known as the echinoderms, characterized by a
multi-element skeleton of calcium carbonate, often with five-
fold radial symmetry, that includes sea urchins, starfish and
crinoids
4 the hemichordate phylum, obscure and unimportant today,
but represented in the Paleozoic by a group of floating
colonial organisms known as graptolites (Figure 12.3),
whose rapid evolution and widespread distribution in marine
sediments make them exceptionally useful for dating rocks,
especially during the Ordovician and Silurian periods.
1 cm
Figure 12.3 Two graptolites: these one- and two-branched forms are
characteristic of the graptolite species abundant in the Ordovician and
Silurian, respectively. Only the hard parts are shown here, consisting of a
series of cups arranged along one or more branches. Before fossilization,
these were made of a protein called collagen. In life, a separate animal lived
in each cup, catching food particles as the colony drifted through the water.
Remember this!
The apparently sudden appearance of all these groups has been
dubbed ‘the Cambrian explosion’. Just how rapidly they all evolved,
and how far back into the Proterozoic non-fossilized ancestors
might have originated, is a matter of controversy.
One important phylum, the cnidarians (or coelenterata), which

includes the jellyfish, has been shown to have first appeared in
the late Proterozoic. This phylum is very rare in the fossil record
until about 460 million years ago, when corals first appeared.
The reason why corals are well represented in the fossil record
is that the soft coral polyp encloses itself in a hard calcium
246

carbonate cup. This is readily preserved after death, especially
in the case of massive colonial corals in which adjacent cups
reinforce each other.
The chordate phylum, which includes the vertebrates (and
therefore humans) was around from the early Cambrian,
although initially it is insignificant in the fossil record. However,
a flourishing fauna of jawless fish developed about 500 million
years ago, and bony fish with jaws followed about 80 million
years later.
Key idea: Hard parts appear

Marine animals with ‘hard parts’ that could easily become
fossilized appeared 540 million years ago (in the so-called
‘Cambrian explosion’), marking the start of the Phanerozoic Eon.
Vascular plants and life on land

About 420 million years ago vascular plants (those with internal
tubes to transport water and nutrients) began to colonize the
land, and were shortly followed by arthropods. The oldest
known land animal is a 428-million-year-old centimetre-long
moss-eating millipede. The insects are an essentially purely
terrestrial class of arthropod that is first known from fossils in
396-million-year-old cherty sediments.
Spotlight: Pneumodesmus newmanii

The oldest land millipede was discovered in 2004 by Mike Newman,
an Aberdeen bus driver who is also a keen amateur fossil hunter.
He found it fossilized in sandstone on a beach in Scotland and
passed his discovery on to experts at the National Museum
of Scotland, in Edinburgh. They made the formal description
necessary to document a new species, which they named
Pneumodesmus newmanii. The first word of the name means ‘air
breathing’ and the second honours its discoverer.

The chordates – represented by the vertebrates – is the other
animal phylum to have been conspicuously successful on land.
Some fish developed the capability to breathe out of water
about 400 million years ago. About 360 million years ago some
of their descendants evolved into the first primitive amphibians
(living on land as adults but breeding in fresh water), followed
within 20 or 30 million years by the earliest reptiles. Other
phyla that expanded on to land, such as segmented worms (the
annelids), have a poor fossil record because they lack hard parts.
Large animal life on land was dominated for a long time by
reptiles, which, unlike amphibians, could breed on land as well
as live there. The most famous are the dinosaurs that flourished
throughout the Mesozoic Era (encompassing the Triassic, Jurassic
and Cretaceous Periods: 250–65 million years ago). At this
time, the swimming reptiles such as ichthyosaurs (Figure 12.4)
and plesiosaurs were among the largest marine creatures. Other
relatives of the dinosaurs (the pterosaurs) took to the air.
1 metre
Figure 12.4 An ichthyosaur, a marine reptile from the Mesozoic Era.

Usually only the bones and teeth survive, but in this example an impression
of the skin is preserved, which enables details such as the dorsal fin and
the upper part of the tail to be shown in the outline around this skeleton.
Mammals were present on land for most of the Mesozoic, though

most were small and insignificant creatures, known mainly from
their teeth, which are the only parts of such small-boned creatures
to fossilize well. The earliest known bird, Archaeopteryx, is of
late Jurassic age (about 170 million years old). This magpie-sized
248

creature, known from fossils found in the fine-grained Solnhöffen
Limestone of Germany, had teeth and dinosaur-like feet and
would probably have been dismissed as a winged reptile but for
the traces of its plumage imprinted within the rock.
The first flowering plants appeared as recently as 100 million
years ago, midway through the Cretaceous, and did not surpass
the conifers in importance for a further 30 million years or so.
Back in Carboniferous times (360–300 million years ago), the
trees responsible for forming most of the world’s coal deposits
belonged to neither of these groups; instead, vegetation was
dominated by Lycopods, giant relatives of today’s small and
inconspicuous club mosses.
Spotlight: Mary Anning’s ichthyosaur

The first recognized ichthyosaur was found in the cliffs near Lyme
Regis in Dorset, England, by local girl Mary Anning (1799–1847)
when she was only 12 years old. In later life, she became well
known as a finder of fossils, and her opinions were sought by
eminent geologists (including Sir Charles Lyell; see Chapter 9).
The world is fortunate that, at the age of 15 months, the infant
Mary survived a lightning strike that killed the neighbour who was
holding her, and two other women.
The mammals take over

It was not until the dinosaurs disappeared at the end of the
Cretaceous Period (65 million years ago), at the so-called K-T
boundary, that mammals had their chance to diversify, which
led to them becoming the dominant group of land animals.
They also spread into the sea, where the ancestors of modern
whales first appeared about 50 million years ago.
The fossil record of human ancestors is not good. Complete
skeletons are unknown, and most of the story has to be pieced
together from skulls (usually incomplete) and other bone
fragments. The first human-like apes (‘hominids’) appeared
about 6–7 million years ago. The first bones recognized to be

from species belonging to our own genus, Homo, are about
2.2 million years old and our species, Homo sapiens, does not
appear until about a quarter of a million years ago.
This timescale should make it clear to you that movies and
cartoons showing humans hunting dinosaurs (or vice versa)
are very wide of the mark indeed, because they overlook the
matter of a 60-million-year gap! On the other hand, that other
well-known extinct animal, the mammoth (essentially a hairy
elephant), was very much a human contemporary. Mammoths
feature prominently in cave paintings, and the last ones died out
as recently as 12 000 years ago. It is debatable to what extent
mammoths’ final demise was due to hunting or to the loss of
their habitat because of climate change
Spotlight: Mammoth proportions

The word ‘mammoth’ was adopted from Russian into English and
other western European languages in the early 1700s, in the wake
of discoveries of mammoth bones in Siberia. A hundred years later
‘mammoth’ had become an adjective to denote enormous size,
such as a cheese weighing 1234 pounds that was presented to US
president Thomas Jefferson in 1802. In fact, mammoths weren’t all
that gigantic, standing about the same height as an Asian elephant
and less tall than an African elephant.
Mass extinctions and adaptive

radiations
Why did the dinosaurs become extinct at the K-T boundary,
after having been so successful for so long? This is a question
that has taxed geologists and palaeontologists (those who
study fossils) for many years. Various explanations have been
proposed, including rapid climate change (perhaps caused by
catastrophic methane escape from sea-floor sediments or by
sulfur dioxide aerosols from flood basalt eruptions) and genetic
damage by cosmic rays from a nearby supernova. Probably
several factors were at play, and the dinosaurs appear to have
been in decline for at least a few hundred thousand years before
250

the probable knockout blow was delivered, when a 10-km-
diameter asteroid or comet collided with the Earth. The impact
site of the main fragment has been identified as the 200-km-
diameter Chicxulub crater on the Yucatán coast of Mexico.
This vast structure is buried beneath more recent sediments and
was discovered by geophysical techniques in the 1980s. A few
other buried craters of corresponding age (as close as can be
told within the uncertainties of the measurements) have since
been found, notably the 24-km-diameter Boltysh crater in the
Ukraine and the smaller Silverpit crater beneath the North Sea
about 100 km off the Yorkshire coast of England (Figure 12.5),
suggesting that the incoming object broke into several fragments
before it struck.
The Silverpit structure was first suggested as a crater in 2002.
Its origin and even its age remain unproven and an alternative
explanation is that the strata have been undermined by the
dissolution of some of the underlying evaporite salts (discussed
in Chapter 9). However, the impact origin of the Chicxulub and
Boltysh craters is uncontested.
Figure 12.5 A 20-km-wide view of the Silverpit structure, revealed by

seismic surveying. This is buried below more than a kilometre of Cenozoic
sediments, and may be an impact crater of the same age as the K-T
boundary. The black areas are unsurveyed.

The Chicxulub impact would have injected a vast amount
of dust into the atmosphere. Supplemented by smoke from
widespread burning, this would have reduced the sunlight
reaching the surface to such an extent that photosynthesis
became impossible. Plants would have died, and the whole
food web depending on them would have fallen apart. Thanks
to surviving seeds and spores, recovery of the vegetation on
land and of marine planktonic plants would have begun within
a few years, but many groups of animals and some plants
became extinct. Whatever the cause, the effects were much
wider than merely the demise of the dinosaurs alone. Many
important marine groups vanished, notably the ammonoids
(Figure 12.1(b)), reef-building molluscs called rudists, and the
large marine reptiles. Most species of marine planktonic animals
of the kind known as foraminiferans disappeared too, although
the group as a whole recovered.
Because it was global in extent, what happened at the K-T
boundary is described as a mass extinction event, and has
effects that can be seen among organisms living in virtually all
environments. The fossil record shows several mass extinctions
during the Cambrian Period (there may have been one
immediately before the Cambrian explosion too) and a more
dramatic one at the close of the Ordovician Period. Because
there was no life on land yet, these mass extinctions are purely
marine. However, a mass extinction event at the end of the
Permian Period (marking the end of the Palaeozoic Era) wiped
out whole groups of land and marine animals and, biologically
speaking, was an even more severe event than the K-T mass
extinction. There was another large one at the end of the
Triassic Period.
Remember this!
Estimates vary, but the end-Permian mass extinction killed off
over 80 per cent of marine genera, over 90 per cent of marine
species and 70 per cent of land species; whereas the K-T mass
extinction saw the end of only 50 per cent of genera and 75 per
cent of species.
252

The causes of these earlier mass extinctions are uncertain, and
it is unlikely that all were caused by impacts. Flood basalt
eruptions or catastrophic changes in seawater chemistry or
oxygenation are possible factors. What is certain is that, after
each mass extinction, some varieties of organisms that had
survived were able to take advantage of the demise of their
competitors. They adapted, through evolutionary change, to
new lifestyles in a process described as adaptive radiation.
Case study: Adaptive radiation

The sudden blossoming of the mammals at the start of the
Cenozoic Era is a good example of adaptive radiation. The
disappearance of the dinosaurs at the K-T boundary meant
that there were no longer any large land-dwelling herbivores or
predators. The sudden lack of competition made it possible for the
hitherto small and unobtrusive mammals to diversify and, within a
few thousand generations, some groups that had moved into new
environments were well on the way to evolving into different species
better suited to take advantage of the new opportunities. This is a
classic case of the meek inheriting the Earth, if ever there was one!
The dinosaurs themselves had been the beneficiaries of an earlier

burst of adaptive radiation, after the end-Permian mass extinction
removed the previously dominant lineages of land reptiles.
Key idea: The mammals take over

Mammals were able to diversify (undergo ‘adaptive radiation’)
after the K-T boundary mass extinction event at the end of
the Cretaceous, 65 million years ago, wiped out most of their
competitors.
How fossils form

A fossil is any sign of a past organism entombed in rock.
Preserved footprints and burrows count as fossils, and these are
termed trace fossils. On the other hand, an impression of the
body where it came to rest after death, or some part of the body

that has actually been preserved, is described as a body fossil. It
is rare for fleshy material to survive – it is usually broken down
by bacterial decay or eaten by scavengers before there is time
for it to be buried, which is why the vast majority of fossils are
the bony, shelly or other hard parts of an organism.
To follow the stages by which an organism can be turned into
a body fossil, I will take as an example a bivalve mollusc, like
a cockle or some other two-shelled shellfish (probably familiar
to you from visits to the seaside). Mollusc shells are known
technically as valves, hence the term bivalve denoting a mollusc
with two shells.
To become a fossil is a long and involved process, and is an
unlikely outcome for the vast majority of individuals. The most
obvious requirement for an organism to end up as a fossil is that
it must become entombed within rock, usually by burial. There
are many chances that can prevent this happening. If our mollusc
has the misfortune to be killed by a predator, then the soft parts
will be eaten. The predator will probably have to break at least
one of the shells to gain access to the flesh, so it is unlikely that
both shells will survive intact. Even if they do, the two shells
may become separated. On the other hand, if the organism dies
a natural death, although the soft parts of the body will almost
certainly rot away, neither shell is likely to be damaged. The
shells are held together at the hinge by fibrous tissues, which are
among the last to rot, so there is a reasonable chance that the
two shells will remain held together until they become buried.
A lot depends on where the creature is when it dies. If it is the
sort of bivalve that lives in rivers, then after death the empty
shells are likely to be rolled downstream and will pretty soon
be broken up. If it is a marine bivalve that lived and died on
a rocky shore, the empty shells will probably be smashed
in the next major storm. If it dies on a beach, it would get
rolled around by the waves; the two shells would probably
separate and become broken and abraded, ending up as
unidentifiable fragments of shell of the sort that you find in
beach sand, or that can become concentrated to make some
kinds of limestone. Only if the organism dies in a low-energy
environment, somewhere quiet enough for muddy sediment to
254

be accumulating, does it stand a reasonable chance of escaping
mechanical breakdown of the sort described above long enough
to become buried by sediment.
However, there is a very important exception to the previous
argument, which is that many species of bivalve (including
cockles) spend most of their time burrowing within the sediment.
If a burrowing bivalve dies in its burrow, then it is already
buried, and the two shells will be held together (thanks to the
surrounding sediment) even after all the connecting tissues have
rotted away. An instant fossil, you might think, but this is far
from the end of the process (Figure 12.6). Very few fossils, and
only relatively recent ones, occur as unchanged shell material.
external mould
shell
sediment
1 2
burial
shell is
dissolved cast
external
mould
internal
3 mould
4
shell is
replaced
Figure 12.6 Stages in the fossilization of the hard parts of an organism. A

bivalve mollusc is shown here (in cross-section), but the same processes
can affect any kind of shell, tooth or bone.

The first thing to happen to a buried bivalve, minus all its soft
parts, is usually that some sediment will find its way inside
the shells. Once the sediment has hardened (by the diagenetic
processes described in Chapter 9), the rock contains an
impression of the outside and the inside surfaces of the shells,
known as an external mould and an internal mould respectively.
Even if, as often happens, the shells are subsequently dissolved
away, these moulds remain, and many fossils are of this type
(Plate 13). If the shell survives, usually it recrystallizes in some
way, so that its original microscopic crystalline structure is
lost. It often happens that the whole shell is dissolved by
groundwater, but if conditions revert, the shell-shaped void
in the rock can be refilled by precipitation, to produce a cast
whose mimicry of the surface texture of the vanished shell
is limited only by the quality of the moulds that it left in the
sediment. It can be very difficult to distinguish between a cast
and a recrystallized shell if these are of the same composition
(calcium carbonate, in the case of molluscs).
The steps described above apply to most types of fossil, with
various changes of emphasis. An ammonoid has only a single
shell, so the complete hard parts stand more chance of being
preserved intact than in the case of a bivalve mollusc, but on
the other hand ammonoids did not live in burrows, so in that
respect their chances of becoming fossilized are reduced. Sea
urchins and other echinoderms have a skeleton consisting of
a large number of fragile plates, which are easily separated
after death, so low-energy conditions and rapid burial are
more important than for bivalves (fortunately many kinds of
sea urchin are burrowers, so these can be found beautifully
preserved). Trilobites and other arthropods have segmented
skeletons, so rapid burial is especially important for these, too,
if the hard parts are to remain intact; moreover, the chitin of
which the skeleton is made does not survive over geological
time, and must be replaced by some form of mineral if anything
other than a mould is to be preserved. In vertebrates, the teeth
(made of a dense crystalline form of calcium phosphate) are the
most robust items, so these are the bits that are most commonly
found as fossils. Bones (in life, a porous growth of calcium
256

phosphate) are less commonly preserved. When they are, their
porous structure has usually been invaded and replaced by new
minerals precipitated from groundwater.
Key idea: Preservation potential

Only a tiny fraction of dead organisms become fossilized, and most
of the fossil record is biased in favour of organisms with hard, or
resistant, parts.
Exceptional preservation
The foregoing does not mean that flesh cannot be preserved,
though it is rare. Examples include complete 20-thousand-year-
old Siberian mammoths frozen in ice, and 100-million-year-
old fish preserved in limestone of the Santana Formation in
northern Brazil. These were killed by highly saline waters that
caused phosphate minerals to be precipitated inside their bodies
within a few hours of death, so that their fine scale cell structure
and even their stomach contents are sufficiently well preserved
to warrant study using high-powered electron microscopes.
Perhaps the best-known circumstance for a whole organism to
be preserved is when it has the misfortune to become trapped in
tree resin, which then hardens to form amber. Delicate insects
and even small frogs can be found within amber.
Spotlight: Cloning from preserved blood cells

It is even true that the blood cells from a blood-sucking insect’s
last meal can end up being preserved in this way, although cloning
a dinosaur from the fragmentary DNA within these cells (as in the
film Jurassic Park) remains a very distant and unlikely prospect.
Exceptional preservation may also occur when environmental

conditions change so suddenly that a representative cross-
section of a living community is wiped out. A classic example is
described in the next case study.

Case study: Exceptional preservation in the
Burgess Shale
In 1909 the American palaeontologist Charles Walcott (1850–1927)
found some strange fossils in the form of black carbon films in
the middle-Cambrian Burgess Shale of British Columbia, Canada.
Most specimens were of unknown types, but others were trilobites
and so on showing their soft parts. Walcott returned every year
until 1924, by which time he had collected 65 000 specimens. The
significance of the site was recognized internationally, and has
been even more thoroughly studied since Walcott’s time.
What seems to have occurred is that animals living on a muddy

sea floor were now and then transported by submarine landslides
into deeper water, lacking in oxygen, where they died and were
immediately buried. The lack of oxygen kept scavengers away and
inhibited decay, with the result that entirely soft-bodied animals
are preserved, as flattened organic residues. This sort of ‘mass
mortality’ deposit is important because it gives us an unbiased
snapshot of a living community. In the Burgess Shale, only about
one fossil in ten is a hard-shelled organism, and some still defy
classification into recognizable phyla. This serves as a warning of
how heavily most of the fossil record is skewed in favour of robust
organisms with easily preserved hard parts.
Spotlight: Hallucigenia
One of the most outlandish Burgess Shale fossils is the aptly
named Hallucigenia, a sort of worm with legs, which was originally
reconstructed upside down. It had a tubular body a couple of
centimetres long and two rows of spines. Those were originally
interpreted as stiff legs, but are now accepted as protective spines
along the creature’s back, whereas there are in fact seven or eight
pairs of legs on the other side, originally misinterpreted as a single
line of tentacles.
258

Plant fossils
Plants have no shells or bones, so you might expect them to
make rather poor fossils. However, the cell walls of land plants
are made of cellulose, which makes them more rigid than
animal cell walls. As a result, woody tissues and even leaves
tend to retain their form after death and burial long enough
for the spaces inside the cells to become filled by minerals
that precipitate from groundwater, such as chert. This is how
petrified wood is formed.
An alternative mode of preservation is when plant material is
buried in waterlogged soil or mud with a low oxygen content.
In this case, the leafy or woody material decays to carbon or a
carbon-rich compound.
A major problem in understanding fossil plants is that it is
very rare to find a whole plant preserved. Roots (as you might
expect) are relatively common, but these are rarely found still
attached to leaves or cones. This has led to different parts of the
same plant being given different fossil names!
On the microscopic scale, pollen and spores are often well
preserved, because their coats are made of sporopollenin, a
substance that is remarkably resistant to decay. In sediments up
to a few million years old, when many plant species were the
same as those today (or else their close relatives), analysis of
pollen is a very important way of establishing the environmental
conditions. For example, changing tree communities in
northern Europe towards the end of the last glaciation are well
documented by the pollen record, which indicates climatic
fluctuations between warm and cold superimposed on the
general warming trend.
The use of fossils in stratigraphy

It is unusual for sediments to contain minerals that
have grown within them whose age can be dated by the
radiometric dating techniques described towards the end of
Chapter 2. If a sediment is interbedded with a series of lava
flows or layers of volcanic ash, then radiometric dating of

these volcanic rocks can be used to bracket the age of the
intervening sediment. However, convenient layers of igneous
rock are often absent from sedimentary successions, and
anyway radiometric dating requires sophisticated laboratory
facilities that a geologist is unable to call on while in the
field. This is where being able to identify fossils pays off.
If you already know the time span over which a particular
kind of fossil organism was alive (because somebody else has
been able to determine this in a location where fossil-bearing
sediments are interbedded with radiometrically datable
volcanic rocks), then you can tell the age of a rock merely by
recognizing the fossils within it.
To take a simple example, trilobites are pretty distinctive
(Figure 12.2), and there is not much risk of mistaking a trilobite
for anything else. You have already seen that trilobites first
appeared during the Cambrian and became extinct during the
Permian. This means that if you find a trilobite in a rock then
its age must lie within these limits. This is a period of about 300
million years, which is of course rather a long time. However,
during this period many different species of trilobite evolved
and then died out, with an average species lifetime of around
10 million years. The two trilobites shown in Figure 12.2 are
clearly different from each other and, although they are both
trilobites, they are not very closely related. The differences
between closely related species are usually more subtle, and it
requires well-preserved specimens and an expert’s eye to make
an unequivocal diagnosis enabling the age of the sediment to be
pinned down to within 10 million years.
It is possible to improve on this if you are fortunate to find
fossils of two species that have overlapping age ranges,
preserved in the same layer of sediment. This can be seen with
reference to Figure 12.7. If you found only species C in a bed
of rock it would not pin the age down very closely, because
this species was alive from 500 to 467 million years ago.
Species B on its own would be more diagnostic, because it
was around for a shorter period (493–478 million years ago),
and species D on its own would give an even more precise age
260

species C
460
species B
470
age (millions of years)
species D
480
species A
490
500
Figure 12.7 The age ranges of four species of fossils (heavy lines) (see
text for discussion).
(470–460 million years). However, if you found species B in

the same bed as species A, then the age of this bed could be
pinned down to the 2-million-year period (480–478 million
years ago) where the age ranges of these two species overlap.
An almost equally precise date could be inferred if species C
and D were found in the same bed.
The above shows how you can infer quite precise ages
of sediments, especially if you can use several fossils in
combination. These principles can be extended to match up the
ages of sediments deposited in areas that are remote from each
other. For example, if species A and B were found together in
North America and in China we would know that these two
unrelated deposits were of almost exactly the same age (this
would be true even if we did not know the absolute ages from
radiometric studies). Correlating between rock successions in
this way is called biostratigraphy.
The rule that overlying (younger) strata have fossils different to
those in underlying (older) strata, and that the progression in
fossil types is always the same, is called the Principle of Faunal
Succession.

Case study: Superposition and the Principle of
Faunal Succession
The ‘law of superposition’ was formulated by Niels Stensen
(1638–86), a Dane whose name is often rendered as Nicolas
Steno. This states that sedimentary layers are deposited in a time
sequence, with oldest at the bottom and youngest on the top.
It seems nothing more than common sense today, but was an
important conceptual breakthrough in its time.
The Principle of Faunal Succession builds on this, and was
established as a tool for relative dating of distant rock units in the
early 1800s by pioneering geologists such as Englishman William
Smith (1769–1839) and Frenchman Georges Cuvier (1769–1832),
decades before Charles Darwin offered an explanation of why
evolution might occur.
Some fossils are better than others for biostratigraphic

correlation. The best ones are called zone fossils (or index
fossils). A good zone fossil:
33 has easily recognizable features that enable it to be
distinguished from other species
33 underwent rapid and irreversible evolutionary change
33 is numerically abundant
33 lived in a wide range of environments
33 lived across a wide geographical area.
Trilobites satisfy the first three points, but usually fail on the last
two. Most trilobite species had fairly specialized lifestyles, and
not being good swimmers their geographic ranges were limited
(for example, entirely different species occurred at the same
time on opposite sides of major oceans). The best zone fossils
tend to be organisms that swam or floated passively in the
oceans and whose hard parts, upon death, sank to the bottom
and can therefore be found in both deep- and shallow-water
marine sediments. Prime examples are graptolites (especially
useful in the Ordovician and Silurian), ammonoids (especially
262

in the Jurassic and Cretaceous), and microscopic planktonic
foraminiferans (notably during the Cenozoic).
A potential pitfall to beware of when using fossils for dating
is that fossils can be eroded from an old sediment, transported
somewhere else and then reburied. ‘Derived fossils’ such as these
will suggest erroneously old ages. However, the derived nature of
a fossil can usually be recognized by signs of abrasion and other
damage during transport. One way to be sure that a body fossil
belongs in the rock where it is found is if it is in ‘life position’, as
is the case, for example, with a bivalve mollusc that died in its
burrow or a coral that is still attached to the rest of the reef.
Now that you have seen how fossils can be used to correlate
ages of rocks between different parts of the globe, I will turn
your attention, in the next chapter, to global Earth history.
Dig deeper
The website of the Natural History Museum (London, UK)
provides access to a wide variety of information about fossils,
including some interactive ‘virtual reality’ three-dimensional
models: http://www.nhm.ac.uk/
A site at the Smithsonian Museum of Natural History
(Washington, DC) has lots of good things to explore if you
select the topic ‘Fossils and Dinosaurs’: http://www.si.edu/
Encyclopedia/Search/Science and Technology
A good account of the history of life in Earth is Life: an
Unauthorised Biography by Richard Fortey (HarperCollins/
Flamingo, 1998).
The book that brought the significance of the spectacular fauna
of the Burgess Shale to popular attention is Wonderful Life by
Stephen J. Gould (W.W. Norton & Co., 1989).
An equally good read by the man who eventually turned
Hallucigenia the right way up, and which takes issue with Gould
over the inevitability of evolutionary outcomes, is The Crucible
of Creation: The Burgess Shale and the Rise of Animals by Simon
Conway Morris (Oxford University Press, 1999).

Fact-check
1 What is or was a lycopod?
a A marine reptile
b A dinosaur that walked on two legs
c A kind of millipede that lived more than 200 million
years ago
d The kind of plants making up the trees from which most
of the Carboniferous coal was formed
2 Based on the example in Figure 12.7, if you found fossils of

species A and species C preserved in the same bed of rock,
what would be the most appropriate estimate of the age
of that deposit?
a 459–500 million years
b 466–480 million years
c 478–480 million years
d 466–493 million years
3 What is a trace fossil?

a One that is so badly preserved that very few signs of it
remain in the rock
b One that is too small to see without optical aid
c Marks preserved in a sedimentary rock that record the
passage of a living organism
d One that is diagnostic of a specific age range
4 When do the first multicellular animals appear on the fossil

record?
a 3.5 billion years ago
b 1.2 billion years ago
c 0.9 billion years ago
d 0.6 billion years ago
5 What is a stromatolite?
a A mound of sediment bound by strands of algae, that
flourished more than 2 billon years ago
b A species of marine plankton whose microscopic shells
make useful zone fossils
c A trace fossil made by scratch marks from a trilobite’s feet
d The internal mould of a bivalve mollusc
264

6 Which of these statements is or are correct?
a The K-T mass extinction affected the dinosaurs but not
marine fauna, because only creatures on land were badly
affected by the asteroid impact.
b Adaptive radiation is when cosmic rays cause mutations,
allowing evolution to accelerate
c Vertebrate teeth are made of calcium carbonate
d Trilobites are common Palaeozoic fossils thanks to their
calcium carbonate exoskeletons
7 Which of these is not a well-known phylum in the fossil record?

a Brachiopods
b Echinoderms
c Molluscs
d Trilobites
8 Which of these phyla was the earliest to appear?

a Arthropods
b Chordates
c Cnidarians
d Hemichordates
9 Which of these is not a characteristic of a species that makes a

good zone fossil?
a Much slower evolutionary change than other species
b Easily recognized and distinct features
c Lived across a wide geographical area
d Lived in a wide range of environments
10 Where did the main fragment of the ‘dinosaur killer’ asteroid

strike the Earth at the end of the Cretaceous?
a Mexico
b North Sea
c Siberia
d Ukraine

13
A short history
of the Earth
33about the Earth’s evolution from a young
semi-molten ball to a globe with a
familiar pattern of drifting continents
33how the global climate has changed on
a variety of timescales, and how the
pace of change is being forced by human
activities.
267

At the end of Chapter 2, you followed the story of the Earth’s
origin as far as its birth as a hot, young planet. You also saw
how the atmosphere has changed over time. Now I will trace
the other main events in the history of the Earth, and risk a
few predictions about the future. Note that the names given to
divisions of geological time and the current best estimates for
their ages in millions of years are gathered together for ease of
reference in Appendix 3.
Imagine the young, hot Earth about 4.5 billion years ago. Most
of the outer part is likely to be molten, because of the heat
liberated by giant impact events. The surface skin has cooled
and solidified, but every so often it is ruptured by upheavals
from below, as pockets of primordial gases burst out from
the interior, and punctured from above by impacting debris.
The magma ocean (as it is called) below the skin is convecting
vigorously, and from time to time the convection currents rip
the solid skin asunder, whereupon a vast tract of cold, and
therefore dense, skin founders as magma spills out on top of it.
A terrible place, but eventually it will turn into your home.
The Earth’s first crust

The fragile skin of the infant Earth is not what most geologists
would recognize as a planetary crust, because it is was merely
the chilled equivalent of the magma beneath it, and had the
same composition. However, during the next 100 million years
or so, the magma ocean below the skin would cool down
sufficiently to allow crystals to grow within it. This would occur
by fractional crystallization (see Chapter 6), and the nature of
the crystallizing minerals would depend on how pressure and
temperature varied with depth. Calculations suggest that in
an Earth-sized body a magma ocean would solidify from the
bottom upwards, and that the last layer to solidify, just below
whatever remained of the chilled skin, would consist mostly of
olivine and pyroxene. This layer would be denser than the layers
below, and so the primitive mantle would be gravitationally
unstable. It would therefore be liable to overturn, whereupon
less dense, deeper mantle would undergo partial melting as a
result of decompression while being displaced upwards. This
268

would feed vast volcanic eruptions to produce the Earth’s
first true crust. Deeper still, the process of core formation (see
Chapter 2) was probably essentially complete by now.
Key idea: The first true crust

The Earth’s first true crust probably formed after the initial magma
ocean had solidified and then overturned.
No crust of such antiquity survives on the Earth today, but

‘primary crust’ (of a different composition) can still be seen in
the pale highland regions of the Moon, which we will consider
in the next chapter.
It may seem counter-intuitive that Earth’s magma ocean should
solidify from the bottom up. The reasons are to do with the
way that crystallization is controlled by pressure as well as
temperature, and also by the traces of water evidently contained
in the Earth’s magma ocean.
Throughout the crust-forming process, the surface would have
suffered from a continuing rain of impacts by Solar System
debris. Although there were no planetary embryos left, it is clear
from looking at craters on the lunar highlands that there were
still plenty of small- to medium-sized planetesimals, capable
of forming craters up to hundreds of kilometres across. Most
would have been rocky (like asteroids) but some would have
been icy comets, and opinion is divided as to what proportion
of the Earth’s water was delivered by cometary impacts at this
late stage. There were even a few larger planetesimals whose
impacts created 1000-km-wide basins on the Moon as recently
as 3.9 billion years ago. The size of these largest impactors can
be deduced by bearing in mind that collisions within the Solar
System usually take place at speeds of a few tens of kilometres
per second (not per hour!), and a crater produced by such an
impact is about 30 times the diameter of the impacting body.
Dating of lunar craters also shows us that the rate of impact
events fell sharply after this time, and had declined to something
like its present level by about 3.8 billion years ago. This marks the
depletion of the widely dispersed debris left over from the birth of

the planets. Subsequent impacts have been mostly by stray comets
or by rocky or iron-rich chunks scattered out of the asteroid belt
by chance events. It is obvious that many of these must have hit
the Earth as well as the Moon. However, there are only about 200
confirmed impact craters on the Earth, because the vast majority
of the Earth’s impact craters have fallen victim to our more active
rock cycle; mostly they have been buried by sedimentation, erased
by erosion, or destroyed by tectonic activity.
To return to the Earth well before 4 billion years ago, with its
newly formed crust and still in the throes of heavy bombardment,
it is likely that the mantle was convecting vigorously, even though
by now it was solidified. Above sites where convection cells
diverged, the crust would be rifted apart and partial melting
of the upwelling mantle would yield ‘basaltic’ crust similar to
today’s oceanic crust. As you saw in Chapter 4, oceanic crust is
recycled at subduction zones. No oceanic crust of Archean age
has survived, except for some debatable and deformed fragments
caught up in collision zones. However, its presence can be inferred
because partial melting of oceanic crust at subduction zones
appears to have led to the growth of continental crust. Andesitic
and other magmas above sites of subduction impregnated and
overprinted the earliest crust, to the extent that no recognizable
remnants of that have survived (Figure 13.1).
The oldest known continental crust consists of metamorphic
rocks about 3.8 billion years old, with the most famous
examples being near Isua in south-west Greenland. The oldest
unit here is the Amitsoq gneiss (apparently a metamorphosed
granite), which is overlain by slightly younger metamorphosed
sediments (containing tiny flecks of carbon that appear to be
biological in origin) and volcanic rocks. Examples of similar
age occur in north-west Canada (the Acasta gneiss), in South
Africa (the Sand River gneiss) and Western Australia (the
Narryer gneiss). The oldest rocks in Europe are possibly gneisses
at Gruinard Bay in north-west Scotland, with a (disputed)
age of 3.3 billion years. Metamorphic rocks must have had
precursors, so it is certain that the earliest continental crust
must have formed even earlier. Recent advances in analytical
techniques have enabled individual mineral grains to be dated
radiometrically.
270

earliest crust
oceanic crust
(a)
andesitic
10 km magma
base of lithosphere
(b)
continental crust
(c)
Figure 13.1 Hypothetical cross-sections through the lithosphere about

4.2 billion years ago, showing how continental crust may have grown. In
(a), andesitic magmas produced by partial melting of subducting oceanic
crust are beginning to intrude and replace the earliest crust, and by stage
(c) the process is complete.
Remember this!
The oldest surviving crust is continental and is about 3.8 billion
years old.

The age record for mineral grains is held by 4.4-million-year-
old grains of zircon (a very hard and resistant mineral made of
zirconium silicate, ZrSiO4) collected from within early Archean
metamorphosed sediments (in fact a quartzite) from Jack
Hills within the Narryer gneiss terrane in Western Australia.
Although the sediments in which these zircon grains ended up
are ‘only’ 3.6 to 3.8 billion years old, the zircons themselves
evidently began life by crystallizing in an igneous rock at
least 600 million years previously. Whether this rock was true
continental crust or just the first volcanic crust will probably
never be known.
With the survival of the oldest rocks, we do not need to rely
so much on speculation. Subsequent developments in Earth’s
history can be read directly, although the record is patchy at
first. During the early Archean, the picture is one of relatively
small ‘continental’ masses, only about 50 km across, being
carried around by the rapid opening and closing of intervening
areas of oceanic crust. This oceanic crust differed from today’s
because the mantle from which it was extracted was about
200°C hotter than now. As a result, rather then being basaltic
in composition, it was poorer in silica and richer in magnesium,
and was of a rock type known as komatiite.
The edges of the small continental masses were marked by
volcanism, presumably because of subduction-related melt
generation where they overrode the oceanic crust. Erosion in
the interiors shed sediment into the oceans, and each time two
continental masses collided, the sediments and volcanic rocks
caught between were deformed, and deeper parts of the crust
were metamorphosed.
Archean plate tectonics?

It would not be wise to regard the Archean opening and closing
of oceans and the associated movement of continental masses as
plate tectonics in the form that occurs today. Both continental
and oceanic crust, and indeed the lithosphere as a whole, were
probably thinner than now, and we do not know whether these
components behaved like rigid plates as they do today.
272

The growth of cratons
Throughout the Archean Eon the volume of continental crust
grew by igneous activity wherever oceanic crust was consumed.
Between about 3.0 and 2.5 billion years ago, many of the
hitherto individual tracts of continental crust became welded
together by collisions. Many of these amalgamations of crustal
units developed into regions that have remained stable ever
since. These are described as cratons or shields. Each craton has
two elements: the original tracts of continental rocks, which are
loosely described as ‘granite’, and the deformed belts of volcanic
and sedimentary rocks, collectively termed ‘greenstone belts’,
that were caught up in the collision zone between each ‘granite’
block. Such an assemblage of crustal material is described as a
granite-greenstone terrain.
The largest granite-greenstone craton is the Canadian or
Laurentian Shield, which makes up most of the eastern half of
Canada from the Arctic Ocean to the Great Lakes (extending
south of the border into the Minnesota–Michigan region). Most
of the rest of North America is younger rocks that became
accreted around the shield as a result of subsequent deposition
of sediments, volcanic activity and plate collisions.
Each of today’s major continents has a cratonic element within
it. Australia has the Pilbara craton (Figure 13.2) and the Yilgarn
craton to its south. Africa has the Kaapvaal and Zimbabwe
cratons in the south and several others further north and west.
Europe has the Baltic shield (Finland and adjacent parts of
Russia), while Asia’s largest is the Aldan shield in the far south-
east of Russia. South American cratons occur in Brazil and
Venezuela.
Key idea: Archean processes

In the Archean Eon, continental crust began as areas (not really
‘plates’) only about 50 km across, that grew (by collision and
volcanism) into terranes (cratons) hundreds of kilometres across.

10 km
Figure 13.2 Sketch map of part of a granite-greenstone terrain in Western

Australia, near Marble Bar. This is part of the Pilbara block, a 500-km-wide
craton extending inland from Dampier and Port Hedland. ‘Granites’ are
indicated by the cross ornament, and the trend of deformation structures
in the intervening greenstone belts is shown by fine lines. Two dykes of
proterozoic age, postdating the assembly of this craton, are shown by
thick lines.
The Proterozoic Eon

Our story has now reached approximately 2.5 billion years
ago, which marks the start of the Proterozoic Eon. Various
authorities estimate that by this stage the volume of the
continental crust had reached somewhere between 40 and 80
per cent of its present value. However, Archean cratons do not
occupy anything like this proportion of today’s continental
surface. This is because the cratons represent only those regions
of Archean crust that have remained stable ever since, and
which happen not to be buried by younger sediments.
I will not attempt to recount Proterozoic history in detail. It is
complex and poorly understood, but becomes easier to unravel
as we approach the Phanerozoic (beginning 540 million years
ago). Essentially, Proterozoic history is one of rifting apart and
drifting together of cratonic blocks into new configurations,
and gradual increase in the total volume of continental crust by
means of volcanic and intrusive activity.
So far as can be told from the few deformed slivers that
remain, the oceanic crust had now become basaltic rather
274

than komatiite in composition, as a consequence of the mantle
having cooled.
Key idea: Proterozoic processes

During the Proterozoic Eon, continental cratons repeatedly rifted
and collided. Slivers of basaltic oceanic crust survive, caught up in
the collision zones.
Some rifts never quite managed to split apart, and became sites
of deposition for large volumes of sediment. Most of these
sediments are types familiar today, except for the lack (or
scarcity) of fossils. However, because the atmosphere was still
poor in oxygen, some sediments were kinds that do not form
today. Notable among these are the banded iron formations
(BIFs) referred to in Chapter 9. The commercially important
Hamersley Basin overlies the Pilbara craton south of the area
shown in Figure 13.2.
Phanerozoic plate configurations

I will now wind the clock on to Cambrian times, beginning
540 million years ago and marking the start of the Phanerozoic
Eon. Experts agree that well over 90 per cent of the Earth’s
continental crust had been created by this time. Henceforth the
record of the rocks offers enough clues for us to fit together a
fairly complete story of events, although there is not space to
trace it fully here.
Key idea: Plates in the Phanerozoic

Plate tectonics was already well established by the start of the
Phanerozoic Eon, and nearly all the current volume of continental
crust had been generated.
It is clear that in Cambrian times the distribution of continental

crust was radically different from today. Evidence for this
includes study of palaeomagnetism, which is magnetism locked
within rocks (especially volcanic rocks) since the time of their

formation. When igneous rock cools, it picks up an internal
magnetization lying parallel to the Earth’s magnetic field at the
time. This enables scientists to determine both its orientation
relative to the pole and its latitude at the time of its formation.
It does not distinguish whether the rock was in the northern
or southern hemisphere, because (as you saw in Chapter 4) the
polarity of the Earth’s magnetic field has flipped many times in
the past. However, this ambiguity can usually be resolved by
backtracking from the present location, provided that the same
region includes suitably magnetized rocks from several intervals
of geological time.
Spotlight: Palaeomagnetists
It isn’t really all that complicated, but it is a ‘dark art’ to some
geologists, who refer to their colleagues who specialize in
palaeomagnetism as palaeomagicians. The prefix ‘palaeo’ is from
the Greek ‘palaios’ meaning ‘ancient’. It is sometimes simplified to
‘paleo’, especially in American English. For no good reason, I have
used the spelling palaeomagnetism but Paleozoic in this book. You
can take your pick, and likewise with the dipthong in the mineral
name haematite/hematite.
Other evidence comes from the similarity of rock types in

continents that are now widely separated. Sediments (and any
fossils within them) are particularly useful in this respect, but
intrusive rocks and deformed belts can also be informative.
Palaeomagnetism may tell us that two continents were at
the same latitude at a certain time, but if they do not share
any geological features of this age, we may conclude that
they were not joined. Palaeomagnetism does not constrain
east–west separation, so it does not tell us how far apart the
two continents were. However, in the Phanerozoic there are
abundant fossils, so if our two continents have similar fossils,
then we can conclude that they were fairly close together;
whereas if the fossils in the two regions are unrelated species, it
is likely that they were separated by a wide expanse of ocean.
276

Key idea: Reconstructing the situation
in the Phanerozoic
Palaeomagnetism, coupled with study of any matching rock
units and the degree of similarity between fossil assemblages,
enables reconstruction of the distribution of continents during the
Phanerozoic.
In Cambrian times, most of the present continent of North

America had been assembled. The bits still missing were various
exotic terranes forming the present western coast (see Chapter 4),
most of the present east coast from Newfoundland to the
Carolinas, and also Florida. On the other hand, the continent
was augmented by the presence of Greenland (firmly attached to
what is now northern Canada), and by what is now Scotland and
the north-western half of Ireland (which were somewhere close
to Greenland). Clearly this continent was not identical to North
America, and so it is given a different name: Laurentia, referring
to the Laurentian shield at its heart.
Palaeomagnetic studies show that Laurentia lay astride the
Equator during the Cambrian, and that the part that would
later become North America was rotated about 90° clockwise
relative to its present orientation (Figure 13.3).
The nearest other continent at this time consisted of what is
today Scandinavia and most of Russia west of the Urals. This
continent was built around the Baltic Shield, and is referred to
as Baltica. Palaeomagnetism shows that Baltica was roughly
halfway between the Equator and one of the poles, and it is
most likely that it lay in the southern hemisphere.
To complete the Cambrian global picture, Africa, Florida,
South America, Antarctica, Australia, India, Tibet, Arabia,
‘Iran’ and southern Europe were joined in a ‘supercontinent’
known as Gondwana. This straddled the Equator but appears
to have been well out of the way of Laurentia. There were
three smaller isolated continents consisting of what are now
Siberia, the Kazakhstan region of central Asia, and China with
south-east Asia. Between Gondwana and Laurentia lay several

NZ
China
Aus Ant
Avalonia
SAm
SAm LAURENTIA Equator
SIBERIA Ind
Iapetus Tibet
Fl Kaz GONDWANA
England Ocean
Af Ar Africa
Scotland Iran
BALTICA
S Europe
Figure 13.3 Sketch map of the whole globe in the middle Cambrian,
about 520 million years ago, showing the distributions of the continents.
Present-day coastlines are drawn to aid identification, but do not
correspond to shorelines at the time. Abbreviations: Af, Africa; Ant,
Antarctica; Ar, Arabia; Aus, Australia; Fl, Florida; Ind, India; Kaz,
Kazakhstania; NZ, New Zealand; SAm, South America. Many tracts of
continental crust consisted of more than is implied by the names used
here; for example, ‘China’ included much of south-east Asia, and ‘Iran’
included Turkey. The named regions within Gondwana as yet had no
separate identities, and are outlined merely to aid recognition. England
was part of Avalonia, which consisted of several microcontinents, most
of which are now exotic terranes on the east coast of North America.
(Modified from various sources.)
microcontinental fragments, collectively known as Avalonia,

which would later become the Carolinas, Newfoundland, Nova
Scotia, the British Isles (except north-west Ireland and Scotland),
and parts of northern France and Belgium.
Throughout the Cambrian, the geologies of Laurentia and
Baltica have little in common. Fossils in marine sediments
on the facing margins of the two continents are generally
unrelated, whereas the unity of Scotland and North America is
confirmed by the presence of identical species of trilobites and
other animals. During the succeeding geological periods (the
Ordovician and Silurian), the assemblages of fossils in Laurentia
and Baltica gradually become more similar. Clearly, the ocean
between the two continents was becoming narrower, and the two
continents had collided by the early Devonian (Figure 13.4). This
now-vanished ocean is given the name Iapetus.
278

SIBERIA
China
Kaz
Equator LAURENTIA BALTICA
Avalonia
Hercynian Ocean Aus
Iran Tibet
S Europe
Ar NZ
Fl India
Ant
GONDWANA
Africa
S America
Figure 13.4 Map in the style of Figure 13.3, showing the continents in the
early Devonian, about 400 million years ago. See caption to Figure 13.3 for
explanation. (Modified from various sources.)
Remember this!
The Iapetus Ocean was a precursor to the Atlantic Ocean. The
name comes from Greek mythology, in which the Titan Iapetus was
the father of Atlas, who held up the heavens.
Various bits of Avalonia accreted on to Laurentia from the

middle of the Ordovician (about 470 million years ago)
onwards. English fossils became more similar to Scottish ones
during the Ordovician and Silurian, showing that the English
terrane was sliding into the Laurentia–Baltica collision zone. Its
speed looks prodigious if you compare Figures 13.3 and 13.4,
but it was only about 10 cm per year, which is comparable with
present-day speeds of relative plate movements.
The geology of the Avalonian terranes contains many indicators
of their histories. For example, many of them host andesitic
volcanic rocks of the kind generated above subduction zones,
showing their essentially island arc character, prior to being
accreted on to Laurentia. The Ordovician volcanic rocks of
the English Lake District are a good example of this, and the
Southern Uplands of Scotland consist of a deformed stack
of sedimentary rocks that accumulated in a subduction zone
trench. The edge of Laurentia that bore the brunt of the

collisions was deformed by folding and thrusting, and a number
of slivers of the Iapetus Ocean floor survive as ophiolites thrust
on to the continent, notably several in Newfoundland.
By the end of the Devonian, Laurentia, Baltica and all the bits
of Avalonia had become united into a single continent. The
collision zone was marked by a high mountain belt, whose
eroded remnant can be seen today in Norway, the Scottish
Highlands and the northern Appalachians. Their origin is
attested by deformation, metamorphism, and intrusion of
granites produced by melting due to crustal thickening (see
Chapter 4). This mountain building event (or orogeny) is
widely known as the Caledonian Orogeny, although in the
Appalachians it is usually called the Acadian Orogeny. Erosion
in the Caledonian mountains shed sediment into rivers that
deposited large amounts of sand in freshwater basins in
parts of the British Isles and eastern North America. These
Devonian Period sandstones have long been known as the Old
Red Sandstone, and the united Laurentia–Baltica continent is
sometimes referred to as the Old Red Sandstone continent.
Hutton’s famous unconformity at Siccar Point, discussed in the
case study at the end of Chapter 10, occurs between marine
sediments that had been strongly deformed and then eroded
during the Caledonian Orogeny, and overlying undeformed beds
of Old Red Sandstone.
Battered though the Old Red Sandstone continent was along
the multiple collision zone, its troubles were not yet over. In
Carboniferous times the enormous continent of Gondwana
shunted into it from the south, throwing up the fold mountain
belt of the Ouachitas, reactivating and enhancing many of
the Appalachian structures, and deforming much of western
Europe and the north-west fringe of Africa. This orogeny is
referred to as the Alleghenian Orogeny in America, and as the
Hercynian or Variscan Orogeny in Europe and Africa. North of
the new mountain belt, the deltas of rivers from the (by then)
partly denuded Caledonian mountains provided the setting for
widespread coal formation. To the south, Africa, South America,
India, Australia and Antarctica share glacial deposits of late
Carboniferous and Permian age, which attest both to their high
280

latitude and to the fact that these components of Gondwana
were still united at that time.
In the Permian, Siberia collided with Baltica along the line of the
present-day Ural mountains, which are a result of the associated
mountain-building event. With the arrival of the rest of non-
Gondwanan Asia shortly afterwards, virtually all the world’s
continental crust was united into a single ‘supercontinent’ known
as Pangea (Figure 13.5). Europe (including the British Isles) and
eastern North America now found themselves in low northerly
latitudes. On land, desert conditions prevailed, as they do today
at similar latitudes. Where the continental crust was flooded by
shallow seas, these were warm and poorly connected to the world
ocean, so marine deposition was mostly in the form of evaporite
deposits. Meanwhile, volcanic activity above subduction zones
and collisions with island arcs continued to extend the continent
in the present-day south-west of North America.
Asia
Tibet
S Europe
N America
Equator
Iran Tethys
Ocean
S America
Africa
India
Aus
Antarctica
Figure 13.5 The world in Triassic times (about 220 million years ago), when
the continents were united into the single supercontinent of Pangea. Tethys
is the name given to the wedge-shaped ocean between the northern and
southern parts of eastern Pangea. (Modified from various sources.)
Case study: Supercontinents and their names

The Tethys Ocean (Figures 13.5 and 13.6) is named after a sea
goddess of ancient Greek mythology, but Pangea is a modern
word constructed from Greek, meaning ‘entire Earth’. The word

was first used by Alfred Wegener in 1920 and took on its modern
meaning at a symposium held in 1927 to discuss Wegener’s theory
of continental drift.
The Triassic Pangea is unlikely to represent the only occasion

when all the continental crust was gathered together. It has
been suggested that the situation illustrated in Figure 13.3 is a
result of the break-up of a previous supercontinent that had been
centred on Laurentia in the late Proterozoic (1250–750 million
years ago). Geologists refer to this as Rodinia, from the Russian
for ‘Motherland’. Having split apart, the fragments of Rodinia may
have been reunited again about 600 million years ago in a short-
lived supercontinent known as Pannotia (meaning ‘all southern’)
before breaking apart into the configuration shown in Figure 13.3.
The hypothesized supercontinent prior to Rodinia (2.5–1.6 billion

years ago) is variously known as Columbia, Nuna or Hudsonland.
The break-up of the supercontinent of Pangea by continental

rifting was as long and complicated as its assembly had been.
Much of the ocean floor that was created by these events
still remains (especially in the Atlantic, Indian and Antarctic
Oceans), so we can use the pattern of sea-floor magnetic
stripes (see Chapter 4) to track the relative movements of the
continents. Break-up was initiated late in the Triassic, when
rifting began between North America and Africa, but this did
not progress far enough to generate oceanic crust until the
Jurassic. Shortly afterwards, India broke free from between
Africa and Australia and began to be carried northwards
because of subduction beneath central Asia and sea-floor
spreading to its south. Next, South America split from Africa,
giving birth to the south Atlantic Ocean (Figure 13.6), but the
north Atlantic Ocean between Greenland and Britain did not
begin to open until about 65 million years ago, after a false start
had attempted to split Greenland away from North America.
Australia broke away from Antarctica at about the same time.
The lines along which Pangea fragmented appear to have been
controlled at least partly by underlying hot spots, where mantle
282

N America
Asia
Atlantic
Ocean
Equator
Africa Tethys
S America Ocean
India
NZ
Aus
Antarctica
Figure 13.6 The world in late Cretaceous times (about 80 million years
ago). The Atlantic had been opening between North America and Africa
for about 100 million years already, and between South America and
Africa for about 30 million years, but had not yet begun to open between
North America and Europe. India was beginning its rapid movement away
from Africa and Antarctica that would lead to its collision with Asia about
40 million years later. (Modified from various sources.)
plumes caused heating, doming and eventually rifting of the

crust. Prior to rifting, extensive fields of basaltic lava flows were
erupted in some places, in the form of continental flood basalts
(see Chapter 5).
The westward movement of the Americas as Pangea broke
apart caused parts of oceanic plates on the floor of the Pacific
to be subducted beneath them. This is the origin of the present
Andes-to-Alaska volcanic belt, and volcanism helped to build
the Central America land bridge that currently links the two
continents. As already noted, the history of western North
America, in particular, is especially complicated because of the
large number of exotic terranes that were accreted on to it.
Key idea: Pangea to the modern world

During the Triassic Period, most of the continental crust was
united in a single supercontinent called Pangea. The break-up of
Pangea led to the present-day continental distribution.

Climate change
You have already come across several examples of climate
change in this book. When trying to use geology to interpret
the past, it is important to realize that some variations in
climate are local (related perhaps to the changing latitude of a
drifting continent, or proximity to an ocean or sea), whereas
others affect the whole globe. Some of the latter are relatively
brief catastrophes of the kinds leading to mass extinctions, as
described in Chapter 12.
Here I will summarize the main long-term global climate changes
that can be deduced from the geological record. This is essentially
a history of alternation between ice ages and warm periods. The
globe appears to have been particularly cold during two lengthy
episodes in the late Proterozoic, with glacial features found even
at low latitudes (a situation aptly described by the expression
‘Snowball Earth’) at about 720–660 and 650–635 million years
ago. These two episodes are informally known as the Surtian
and Marinoan Snowball Earths, respectively. A briefer interlude,
known as the Kaigas, is suspected about 750 million years ago,
and a more ancient Snowball Earth (known as the Makganyene)
may have prevailed at about 2200 million years ago.
The global climate was warmer for the remainder of the
Proterozoic and remained so throughout the Cambrian and most
of the Ordovician. Although full Snowball Earth conditions never
returned (see the case study below), a 20-million-year ice age has
been documented in the late Ordovician and early Silurian, when
glacial deposits are found in northern Africa (which was then at
the south pole) and adjacent parts of Gondwana.
Warm conditions then returned for nearly 100 million
years until an 80-million-year ice age lasting from the early
Carboniferous to the late Permian. Alternate advances and
retreats of the ice (glaciations and interglacials) and associated
falls and rises in sea level are probably responsible for the
cyclical nature of sedimentation related to the formation of
coal deposits in late Carboniferous times, though undoubtedly
subsidence of deltas and switching of delta channels played a
role, too (see Chapter 9).
284

The world was mostly warm again from the late Permian until
about the Eocene, when the climate began to cool. Antarctica
was ice-covered for most of this time, but ice sheets did not
spread from Greenland across North America and Europe
until about 2 million years ago. As noted in Chapter 8, from
the geological perspective there is no reason to suspect that
our present mild climate represents anything other than a brief
interglacial interlude between successive advances of the ice.
Case study: Snowball Earth versus ice ages and

their causes
The causes of ice ages are poorly understood, and are sure to be
complex. Subtle changes in the shape of the Earth’s orbit and the
tilt of its axis influence many sorts of climate variation on a variety
of timescales, but cannot be the complete explanation.
The greater the extent of ice cover, the more sunlight is reflected
by the Earth, and hence the less the solar heating of the globe.
This would make a Snowball Earth situation naturally self-
perpetuating. The Phanerozoic ice ages were all considerably less
severe than a full Snowball Earth situation, having high-latitude
(but not equatorial) ice sheets during their glaciations.
Variations in the amount of greenhouse gas of various kinds, such

as carbon dioxide, sulfur dioxide and methane (influenced by the
weathering and volcanic parts of the rock cycle and by biological
activity), and in the yearly average amount of cloud cover (which
depends on ocean circulation and the distribution of continents)
are probably additional key factors. In the early Proterozoic, the
Sun was probably shining less strongly than today (perhaps about
70 per cent), and were it not for the strong greenhouse effect
from the atmospheric carbon dioxide that had not yet been largely
broken down to liberate oxygen (see Chapter 2), the Snowball Earth
situation would have been the norm rather than the exception.
Sea level has fluctuated in various ways over geological time.

Naturally, it has tended to be low at times when large amounts
of water were held on land in the form of major ice sheets, and
highest when there have been no major continental ice sheets.
However, an even more important factor seems to be the vigour

of sea-floor spreading. When sea-floor spreading is particularly
active, the total volume of constructive plate boundaries in the
form of mid-ocean ridges is abnormally great. This displaces
seawater unusually far over the edges of the continents, giving
rise to atypically extensive shallow seas.
Sea level appears to have been as much as 600 m higher than
at present during the Ordovician, when sea-floor spreading
following the break-up of the late Proterozoic supercontinent
reached a crescendo. It had fallen to approximately its present
level by the Triassic, when Pangea was in existence, but rose to a
new high during the Cretaceous in response to the many young,
hot, mid-ocean ridges initiated by Pangea’s break-up.
Another factor that surely influences the climate on a global scale
is the distribution of the continents. For example, in the Triassic
(Figure 13.5) there was a single continental mass extending from
pole to pole, which would have prevented the establishment of
globe-encircling ocean currents. This is very different from the
Devonian (Figure 13.4), when there was probably a current
girdling the globe just south of the Equator, or to today, when a
circumpolar current is able to circle Antarctica between 50 and
60°S. Another characteristic of today’s climate is the monsoon,
which brings much-needed seasonal rain to India and east
Africa. According to computer-assisted climate models, without
the barrier to atmospheric circulation posed by the 5-km-high
plateau of Tibet (which is a result of the India–Asia collision
50–25 million years ago), there would be no monsoon.
Therefore, if there were no Tibet, the climate of the Indian Ocean
region, if not the globe, would be radically different.
Key idea: Ice ages and Snowball Earths

Ice ages (and extreme versions called Snowball Earths) happen at
irregular intervals, for unknown – but undoubtedly complex – reasons.
Human-induced climate change

The greenhouse effect is perfectly natural (and without it the
world would be about 30°C colder), but fears are growing that
human release of additional carbon dioxide into the atmosphere
286

by burning fossil fuels is responsible for slight, but perceptible,
global warming (and resulting rise in sea level) since about
1950. The amount of carbon dioxide in the atmosphere rose
at an accelerating rate from 280 parts per million in the year
1800, to 399 parts per million in 2014. After carbon dioxide,
methane is the second most important greenhouse gas released
by human activities (the main sources being cattle farming and
decomposition in landfill sites).
We may thus be forcing the climate to change at faster rates
than most geological processes. Justifiable concern over
the rate at which fossil fuels are being burned, led to the
coining of the term ‘carbon footprint’ to reflect the amount
of carbon dioxide emitted by a particular activity or process.
Some Western countries (such as the United Kingdom and
Germany) have managed to reduce their carbon dioxide
emissions since 1990. However, the government of the USA,
the country that emits by far the greatest per capita amount
of carbon dioxide, long refused to be bound by international
reduction targets. In 2007 China, the planet’s most populous
country and opening new coal-fired power stations at a rate
of two per week, overtook the USA as the globe’s greatest
source of carbon dioxide.
Before his election as US president in 2008, Barack Obama
pledged to reduce US carbon dioxide emissions to 1990 levels
by 2020, and to cut them to 80 per cent by 2050. The same
target for 2050 was endorsed by leaders of the ‘G8’ industrial
nations in 2009. However, although the need for emission
reductions to combat global warming had become part of
the political rhetoric of both the USA and China, the much-
heralded 2009 UN climate change conference in Copenhagen
failed to result in legally binding agreements. In 2014 a bilateral
agreement between President Obama and President Xi Jinxing
of China pledged that China’s carbon dioxide emission would
peak around 2030, whereas emission by the US would be at
least 26 per cent below their 2005 levels by 2025.
It remains to be seen whether good intentions prove sufficient.
India, the world’s third-largest emitter of carbon dioxide, has
so far refused to commit to any cuts. I am tempted to write

‘Don’t hold your breath’, but at least if you do that you will be
reducing your own carbon dioxide emissions!
The future history of the Earth

Leaving aside what may happen to climate, it is possible to
extrapolate the major aspects of present-day plate motions with
reasonable confidence for the next 50 or 100 million years. The
Atlantic must continue to widen, while sea-floor spreading on
the Mid-Atlantic Ridge carries the Americas away from Europe
and Africa. Eventually, a major subduction zone will form in the
Atlantic, but we cannot predict on which side this will happen.
Much of California seems set to continue its north-westwards
displacement by sideways slip along the San Andreas fault
system, and is due to pass British Columbia in about 50 million
years’ time.
The collision of India with Asia is largely a spent force.
Eventually, erosion will wear down the Himalayas, but they
will be surpassed in grandeur by mountains thrown up by the
forthcoming collision between Australia and south-east Asia. Its
first stages are manifested by thrust movements (and associated
earthquakes) in Papua New Guinea, and about 30 million years
from now the whole of the Indonesia region is likely to be
compressed into a continental collision zone. At the same time,
slow northward convergence of Africa towards Europe is likely
to unite the two continents, replacing the Mediterranean Sea
with a mountain belt higher and longer than the Alps.
Sea-floor spreading in the Red Sea will widen the gap between
Africa and Arabia, and may propagate along a transcurrent
fault line that runs along the present Jordan valley. Similarly,
crustal stretching across the African Rift Valley may develop
into sea-floor spreading, in which case a new ocean will separate
the main part of Africa from a strip extending between Somalia
and northern Mozambique.
Bounded on east, north and west by subduction zones, the
Pacific Ocean will continue to be ringed by Andes-type volcanic
belts and island arcs. If the pace of sea-floor spreading is unable
to keep up with the rates of subduction at the destructive
288

plate boundaries, the Pacific Ocean must contract. That being
so, North America will eventually swing into mainland Asia,
sweeping Japan to destruction in the process, and South
America will collide with New Zealand and Australia.
If this happens, then, with the possible exception of Antarctica,
the world in maybe 200 million years will once again bear a
single supercontinent, though with its components differently
disposed compared to how they had been arranged in Pangea.
Some tens of millions of years later, this, too, will break apart
and the cycle will begin again.
Looking even further ahead, the Earth’s radiogenic heat engine
is running down so slowly that there is no reason to expect that
plate tectonics will ever cease, unless perhaps the climate goes so
far off course that Earth loses all its water and its asthenosphere
becomes too strong to allow plates to slide. More likely, the
lithosphere may become a bit thicker, and sea-floor spreading a
bit slower, but the large-scale processes seem likely to continue
much as they do today, until (according to well-established
astronomical models) the Sun swells up and engulfs the Earth
some 5 billion years into the future.
Key idea: The future

Plate tectonics will probably continue to destroy and create oceans,
while reuniting and separating continents.

Dig deeper
Maps and animations of plate tectonic movements and the
changing pattern of global climate can be found at
http://www.scotese.com/
The Open University’s plate tectonics game, already suggested
in Chapter 4, includes a section on past and future plate
reconstructions: http://www.open.edu/openlearn/science-
maths-technology/slip-slide-collide
A book that describes how the Earth’s history has been
deciphered (relevant to earlier chapters as well as this one) is
Earth’s Deep History: How it Was Discovered and Why it Matters by
Martin J.S. Rudwick (University of Chicago Press, 2014).
A readable academic text that covers what its title says is How to
Build a Habitable Planet: The Story of Earth from the Big Bang to
Humankind by Charles H. Langmuir and Wally Broeker (Princeton
University Press, 2012).
A good starting point for climate change data is
http://climate.nasa.gov/
290

Fact-check
1 What is/was Rhodinia?
a A ‘Snowball Earth’ episode during the Proterozoic
b A Proterozoic supercontinent
c An ocean separating Gondwana from Laurentia during the
Paleozoic
d An ocean forming a vast embayment into eastern Pangea
in the Triassic
2 How is the Earth’s first differentiated planetary crust believed

to have formed?
a By accretion of meteorites on to the surface of the young Earth
b By chilling of the molten magma ocean
c By fractional crystallization of the magma ocean
d By partial melting of the mantle
3 Why was the Archean oceanic crust komatiite rather than

basalt in composition?
a Because the Archean mantle was significantly poorer in
silica than today’s mantle
b Because the Archean mantle was significantly richer in
silica than today’s mantle
c Because the Archean mantle was significantly hotter than
today’s mantle
d Because plate movement in the Archean was much faster
than today
4 What is a craton?
a A stable tract of continental crust, surviving from the
Archean
b A supercontinent
c The scar left by an asteroid collision
d A region where ‘banded iron formations’ accumulated
5 What fraction of the present-day volume of the continental

crust already existed at the end of the Proteozoic?
a 50 per cent
b 75 per cent
c 90 per cent
d 98 per cent

6 What is the Amitsoq gneiss?
a A craton
b An uplifted tract exposing the mantle below Greenland
c A representative of the first crust to form on the Earth
d A granite metamorphosed about 3.8 billion years ago
7 Which of these was not part of Gondwana in the Devonian?

a Africa
b Antarctica
c Arabia
d Avalonia
8 The closure of which ocean is associated with the collision of

India into Asia to form the Himalayas?
a Hercynian
b Iapetus
c Indian
d Tethys
9 Which two atmospheric gases contribute most to the

greenhouse effect?
a Carbon dioxide and ammonia
b Carbon dioxide and methane
c Methane and water vapour
d Nitrogen and water vapour
10 Why is it thought that there were no low-latitude globe-

encircling ocean currents in the Triassic?
a There was no polar ice-melting to drive ocean currents
b The supercontinent Pangea stretched from pole to pole, so
there was no available route for such a current
c Sea-floor spreading would have disrupted it
d The location of the Earth’s poles was different at that time
292

14
Solar System
geology
33about the geological processes that
have left their marks on the surfaces of
the other planetary bodies in our Solar
System.
293

One of the benefits of studying geology is that when you
have begun to understand the Earth, you can use your skills
and knowledge to deduce what has shaped the surfaces of
other planets. In return, study of other planets can help you
understand the history and development of our own planet, or
see it from a new perspective. I exclude the giant planets Jupiter,
Saturn, Uranus and Neptune because these are fluid all the way
down to their tremendously deep cores. However, the other
‘terrestrial planets’, Mercury, Venus and Mars, are rocky bodies
with (we presume) iron-rich cores that are directly comparable
with the Earth. So is the Moon, which counts as a terrestrial
planet so far as geologists are concerned, even though it orbits
the Earth rather than going directly around the Sun. The sizes,
masses and densities of these bodies were compared with the
Earth in Table 2.1.
Key idea: Planetary geology

Study of the Earth helps us understand other planets better, and
study of other planets helps us understand the Earth better.
There are many bodies in our Solar System other than planets
whose size and composition make them suitable for geological
study. These include the larger asteroids (which are generally
rocky), Io (a silicate-dominated satellite of Jupiter with active
volcanoes), and at least the 15 largest icy satellites of the giant
planets. These have an icy mantle overlying a silicate core,
but this ice behaves in all important respects like the silicate
material that constitutes the Earth’s outer layers. There are
also numerous icy dwarf planets in the ‘Kuiper Belt’ beyond
Neptune, some exceeding 2000 km in diameter. Pluto is one of
these, but it is no longer the largest known example.
Spotlight: Pluto not a planet

A Kuiper Belt object that was awarded the name Eris in 2006
is possibly bigger and certainly more massive than Pluto.
Recognizing this, in the same year the International Astronomical
Union controversially stripped Pluto of its status as a planet that
294

it had held since its discovery in 1930. In my opinion, this was a
logical and sensible decision. Others disagree, notably many of
those involved in NASA’s 2015 Pluto fly-by mission, New Horizons.
They seem peeved not to be the first to visit the last unexplored
planet, rather than excited to be the first to visit a Kuiper Belt
object. Notwithstanding its classification, Pluto and its largest
moon, Charon, are fascinating places.
Here I will try to give a taste of the geology of our nearer

neighbours. However, first it is worth saying something about
impact cratering. This is the clearest example of a process whose
effects are studied more easily on other bodies, and whose study
has led to better insights into the history of the Earth.
Impact craters
Any solid surface in the Solar System that is not wiped clean
by erosion, deformed out of recognition by tectonic processes,
or buried (by sediments, lava or volcanic ash) becomes
pockmarked with circular craters within a few hundred million
years. This is because of infrequent impacts by small chunks of
rocky or icy debris (asteroids and comets) that are remnants
of much denser swarms of debris left over from the birth of
the Solar System. As you read in the previous chapter, the rate
of bombardment had fallen dramatically by about 3.8 billion
years ago, but it still goes on. The Earth’s active geological
processes are fast enough to outpace the rate of crater
formation, and until the 1960s most geologists did not even
accept that any craters on the Earth are formed by impacts.
However (as you read in Chapter 13), nearly 200 impact
craters have now been identified on Earth. In contrast, impact
craters (many of them hundreds of kilometres in size) survive
everywhere on the geologically less active Moon, whose largest
craters can be seen through simple binoculars. Counting the
density of craters on planetary surfaces is the best way we have
of estimating surface ages, because (with the exception of some
areas of the Moon) we lack samples from known locations for
radiometric dating.

Comparative studies of craters on different bodies, and
simulations of crater formation in the laboratory and on
computers, have given us a pretty thorough idea of how craters
form. Figure 14.1 shows the accepted model. A projectile
(such as a fragment of an asteroid) hits the surface at about
10 to 40 km per second. At such a high velocity (described as
‘hypervelocity’), the crater-forming process is driven by shock
waves emanating from the point of impact. The shock waves
(a)
(b)
(c)
Figure 14.1 Cross-sections showing how an impact crater is formed.

(a) A hypervelocity projectile strikes the surface. A shock wave begins to
excavate a crater, and also destroys most of the projectile. (b) Excavation
by the shock wave spreads radially out from the point of impact. As drawn,
at this stage the crater has reached virtually its final diameter, and the
latest ejecta are barely able to flop over the rim. (c) A central peak has
risen by rebound, and the walls of the crater have slumped, giving the
crater its final form. Where there is no atmosphere to shield the surface,
impact cratering occurs at all scales, from microscopic to 1000 km. A
single central peak, as shown here, occurs only for a limited range
(15–140 km on the Moon, 7–70 km on the Earth). It takes about
100 seconds to excavate a 100-km crater.
296

excavate a hole that ends up about 30 times the diameter of the
projectile. As they do so, they fling out an expanding ‘curtain’ of
fragmented ejecta that becomes dispersed over the surrounding
terrain. The resulting crater is circular for all but the most
grazing of impact angles. Impact craters have a distinctive
and virtually unmistakable morphology, which includes a
floor lower than the surrounding terrain, a circular outline, a
surrounding blanket of ejecta and (in craters above a minimum
diameter that depends on the planet’s gravity) a central peak, or
ring or cluster of peaks.
Remember this!
Impact craters result from collisions at typical speeds of tens of
kilometres per second.
Studies of the rock below the floor of impact craters on the

Earth confirm the high pressures associated with the shock
of impact. The first such crater to be thoroughly investigated
was the Barringer crater (otherwise known as Meteor crater)
in Arizona. This is a small crater (barely a kilometre across)
that was formed about 20 to 30 000 years ago, but it has a
characteristic ejecta blanket around it, and drill holes show that
the rock below the crater floor is severely shocked.
Tremendously high impact pressures are also demonstrated
at the 15-million-year-old 24-km Ries crater near Nördlingen
in Bavaria, Germany. This is strewn with diamonds thought
to have been produced by shock compression of atmospheric
carbon dioxide.
Spotlight: Diamonds from the sky

Unfortunately for the good people of Nördlingen, the diamonds
in their crater are microscopic in size and of no conceivable use.
However (as you saw in Chapter 9), other meteorite impacts have
sometimes provided genuine bounty from the skies, as exemplified
by the Sudbury nickel ore body in Ontario.

The dangers posed by impacts are most clearly demonstrated
by the Chicxulub crater in Mexico, whose formation led
to the end-Cretaceous (K-T boundary) mass extinction (see
Chapter 12). Even an impact of the relatively small size that
produced the Barringer crater would destroy a nearby city, and
if it hit the ocean it could displace sufficient water to cause a
devastating tsunami. Telescopic sky surveys to quantify the
threat posed by ‘near-Earth asteroids’ suggest that these are
sufficiently numerous, that we should expect a 100-m-diameter
object to hit the Earth on average about every thousand years,
a 200-m object every 5000 years, and a 1-km impactor every
100 000 years.
The Moon
The Moon’s diameter is 25 per cent, its mass 1.2 per cent and
its density 60 per cent of those of the Earth. It is too small to
retain an atmosphere. Because it is so close to us, the Moon has
been explored more fully than any other body beyond the Earth.
A calibrated cratering timescale has been established, using
the relationship between radiometric ages of samples of lunar
material and the local density of impact craters. By extrapolating
this, we can make a reasonable estimate of the absolute ages
of cratered regions on other bodies in the inner Solar System.
This does not apply beyond the orbit of Mars, where the impact
history and the source of the impactors have both been different.
The Moon appears to have cooled down much more quickly
than the Earth. This is to be expected, because a planetary
body generates heat from radioactive decay in proportion to
its volume, but loses heat in proportion to its surface area.
The smaller the body, the more surface area it has relative to
its volume, so the faster it will cool. In the case of the Moon,
this cooling has resulted in a present-day lithosphere about
1000 km thick, which explains why the lunar surface shows no
substantial signs of current or recent geological activity.
The minerals in rock samples from the Moon are all nearly
anhydrous; that is to say, they contain very little H2O or OH.
This means no amphiboles, micas, clay minerals or hydrated
298

iron oxides, and only tiny traces of water trapped inside other
minerals. This discovery was one line of evidence that convinced
the scientific community that the Moon accreted from debris
ejected by a ‘giant impact’ collision, which would have deprived
the Moon of most of its volatiles.
Case study: Dry Moon?

In the decades after the Apollo Moon landings (1969–72), the
Moon was generally regarded as bone-dry. However, in 2008,
new analytical techniques began to reveal traces of OH and other
volatiles inside samples of sub-millimetre-size glass beads
erupted from lunar volcanic vents. Soon, water began to be
reported also inside some minor lunar minerals within lunar rock
samples. This evidence does not undermine the ‘giant impact’
theory for the Moon’s origin, because the Moon is still very dry (and
depleted in volatile elements) compared to the Earth.
In addition, it has now been shown that there is ice inside

permanently shadowed polar craters. Unlike the traces of water in
volcanic beads and within minerals, this has almost certainly been
delivered to the Moon by impacting comets, and represents the tiny
fraction of the cometary ice molecules that became caught in ‘cold
traps’ rather than escaping back to space.
The Moon’s geological processes waned so quickly that large

regions of its earliest crust have survived in the lunar highlands,
which are the regions that look brightest when you look at the
Moon. Samples from the Moon show that the rock type here
is anorthosite, which is composed of a calcium-rich variety of
plagioclase called anorthite. In contrast, the dark areas on the
Moon are covered by basaltic lavas that flooded most of the largest
impact basins. This happened mostly between about 3.9 and 3.1
billion years ago (dated radiometrically by samples collected during
the Apollo landings), though the youngest flows may be as ‘young’
as only 2 billion years. These dark areas were once (incorrectly)
thought to be lunar seas, and they are still known by the Latin word
for sea, ‘mare’ (pronounced MAH-ray; the plural form is ‘maria’
pronounced MAH-ria). A rapid decline in the rate of bombardment
in the early Solar System is demonstrated by the fact that the lunar

highlands typically have far more craters per unit area than the
maria, despite being only about 25 per cent older. Highlands partly
flooded by mare basalts can be seen in Figure 14.2.
Figure 14.2 A 200-km-wide view of the Moon, photographed from orbit

by Apollo 15. Solar illumination is coming from the right. Rugged highland
crust (in this example uplifted as part of the rim of an enormous impact
basin) occupies most of the view, but is flooded in the north-west by
smoother mare basalts. The sinuous trench is a collapsed lava tube, wider
than those on Earth because of the Moon’s lower gravity. There are more
craters in the highlands than on the mare surface, but none in this view is
large enough to have developed a central peak.
Because it is still preserved and has been sampled, we

understand the origin and nature of the Moon’s first crust
better than the Earth’s. As noted in Chapter 2, the Moon
almost certainly accreted from debris flung out by the last ‘giant
impact’ collision between the Earth and a planetary embryo.
Lunar accretion appears to have been energetic enough to cause
melting, so the young Moon had a magma ocean. As the lunar
magma ocean cooled and experienced fractional crystallization,
300

it entered a stage where anorthite crystals began to grow in the
upper part of the magma ocean. These were less dense than the
average magma and were able to rise to the surface, to form a
‘flotation crust’ that corresponds to what we now find in the
lunar highlands.
Remember this!
There are two reasons why a lunar-type flotation crust is unlikely
to have formed on the Earth. First, given Earth’s stronger gravity,
plagioclase could crystallize only in the very uppermost fraction of
any magma ocean where the pressure was low enough. Second,
unlike the Moon’s magma ocean, the Earth’s magma ocean
contained sufficient water to force crystallization to begin at its
base rather than near the surface.
The craters in the lunar highlands show virtually no signs of

deformation by tectonic processes, so the Moon’s lithosphere
must have become thicker (and/or more rigid) than the Earth’s
at an early stage. After the most recent episode of mare lava
eruption, the lithosphere evidently became too thick for
more magma to escape easily from below. Except for impact
cratering, the Moon has been geologically very quiet ever since.
Seismometers left in place by the Apollo programme detected
weak ‘moonquakes’ (some of them near the base of the Moon’s
thick lithosphere). There are occasional controversial reports of
gases leaking from isolated patches of the surface.
Key idea: The Moon

The Moon probably accreted from debris flung off by the last giant
impact collision to affect the Earth. The degree of dryness of the
lunar interior has recently been called into question.
Mercury
Mercury is the closest planet to the Sun. Though larger
than the Moon, it is still too small to retain an appreciable
atmosphere. Its density is 98 per cent that of the Earth. That

may seem unremarkable to you, but it is actually anomalous;
Mercury has only a twentieth of the Earth’s mass, so its gravity
is too weak to lead to the same amount of ‘self-compression’
in its interior. Mercury’s high density in comparison to its
mass can be reasonably explained only if it has a much larger
iron core (relative to its total size) than the Earth, or indeed
than Venus or Mars. Mercury is the only terrestrial planet
other than the Earth to have a strong magnetic field, and there
is good evidence that this is generated by a self-sustaining
dynamo operating in a liquid shell constituting Mercury’s
outer core.
Its large core means that Mercury must contain a greater
proportion of iron and a correspondingly smaller proportion
of silicates than any other terrestrial planet. How it came to
be made from such a different recipe is such an intriguing
mystery that both NASA and the European Space Agency
have put aside part of their budgets to send probes there this
century, the first since NASA’s Mariner 10 sent back images in
1974–5.
Mercury is a heavily cratered world, but is only superficially
reminiscent of the Moon. Volcanic plains cover a large
proportion of its surface, and were revealed by NASA’s
MESSENGER mission, in orbit about Mercury 2011–15, to
be iron-poor, magnesium-rich basalt containing up to 5 per
cent sulfur. Some giant thrust faults can be seen disrupting
the surface, and there are several volcanic vents that seem to
have been sites of explosive eruptions (Figure 14.3). Most
activity happened more than 3 billion years ago, but a few
thrusts and volcanic vents were probably active within the
past billion years. One process that definitely continues today
is the stripping away of patches of surface to a depth of ten
or so metres to form irregular-shaped flat-bottomed ‘hollows’.
With no atmosphere, and therefore no wind, the most likely
explanation is the loss of volatile material to space. No one
knows what this volatile substance is, nor the nature of the
magmatic gas that drove the explosive eruptions.
302

Figure 14.3 A 200-km-wide view of Mercury, obtained by MESSENGER.
Solar illumination is coming from the left. Most of the area is lava plains,
and the superimposed impact craters demonstrate a surface age of
about 3 billion years. In the centre lies a 30-km-long irregular depression
excavated by a series of volcanic explosions that occurred while the active
vent migrated to and fro.
Key idea: Mercury

Mercury is the only other terrestrial planet that generates a
magnetic field. It has a very large iron core, but very little iron in
the lavas that cover most of its surface.
Spotlight: How to name a space mission

Space agencies are renowned for the contrived acronyms used to
name their missions. MESSENGER is one of the best. It derives
from MErcury Surface, Space ENvironment, GEochemistry, and
Ranging, and was devised because in classical mythology Mercury
was the messenger of the gods. The name of the European Space
Agency’s Mercury orbiter (due to arrive in 2024, and in which I am
personally involved) is BepiColombo, named after Giuseppe (Bepi)
Colombo, an astronautical engineer.

Venus
In terms of basic properties, Venus ought to be our twin
planet. Its diameter is 95 per cent, its mass 82 per cent and its
density 95 per cent of the Earth’s. However, the two planets are
surprisingly dissimilar. Venus is shrouded beneath an opaque
atmosphere, but orbiting space probes have been able to image
the surface using radar. Venus’s atmosphere is denser than ours,
with a surface pressure over 90 times greater, and is 96 per cent
carbon dioxide. Its average surface temperature of 450°C is
much hotter than expected, for a world whose distance from the
Sun is only a third less than the Earth’s.
How Venus got this way is a tale of extreme climate change,
whose reach was long enough to affect even the properties of
its mantle. Venus probably once had liquid water in similar
abundance to the Earth, but the higher temperature closer to the
Sun meant that a greater proportion of it was evaporated into
the atmosphere. Water vapour acts as a greenhouse gas, so the
atmosphere became even warmer, causing even more water to
evaporate, leading to a ‘runaway greenhouse effect’ and ending
up with the hot dry world we see now. Venus’s atmosphere today
contains only a trace of its previous water vapour, because solar
ultraviolet radiation has long since split most of it into hydrogen
(which has escaped into space) and oxygen (which has reacted
with the rock). Venus’s present greenhouse effect is maintained
by the large amount of carbon dioxide in its atmosphere. Earth
probably once had a similar amount of gaseous carbon dioxide,
but most of this is now locked up in limestone, a rock type that
can form only in water and so is not expected on Venus.
The surface of Venus is better documented than much of Earth’s
ocean floor, because virtually all of it has been mapped by radar
images in sufficient detail to show features as small as about
100 m across. However, our understanding of Venus does not
compare with that of the Earth, because we lack seismic data
and other information that can be gathered only on the ground.
Perhaps the best way to sum up Venus’s geology is to remark
that it does not appear to be experiencing Earth-style plate
tectonics. There are widespread volcanoes (Figure 14.4)
304

Figure 14.4 A perspective view across part of Venus, constructed from
radar images obtained by the Magellan probe in orbit above the planet.
The scale varies from about 100 km across the foreground to nearly
1000 km across the background. The nearby fractured plains (dark) are
overlain by lava flows (pale) emanating from the shield volcano Sapas
Mons in the middle distance. Other volcanoes dominate the horizon.
associated with lava flows that appear to be basalt, and a few

highly deformed regions. However, nobody has come up with
a convincing model to link the various elements together on a
global scale in the way that plate tectonics does for the Earth.
A likely explanation is that the dryness of Venus’s interior
means that the top of its asthenosphere is not weak enough to
allow plates to glide freely across it, in contrast to the Earth
where ocean water is continually fed into the asthenosphere via
subduction zones.
The age of Venus’s surface is particularly perplexing. Crater
counting suggests an average age of about 500–800 million
years. It was once claimed that there is no significant age
variation from region to region, which forced the conclusion
that the whole globe had been resurfaced in a 10-million-year
orgy of flood basalt volcanism about 500–800 million years
ago. Now it seems more likely that the rate of resurfacing has

Figure 14.5 The surface of Venus, seen from the Soviet lander Venera
13. Crude chemical analysis of the slabby rocks beside the foot of the
spacecraft show a basaltic composition.
been on the wane ever since about 800 million years ago, and
that there are probably small tracts of terrain as young as a few
million years old. Certainly some of Venus’s volcanoes look
young, though there is no proof of present-day eruptions. The
few (possibly unrepresentative) images we have from the surface
show it to be lava strewn (Figure 14.5).
Key idea: Venus

Not just the atmosphere, but the whole geology of Venus, has been
profoundly influenced by a runaway greenhouse effect that caused
the planet to lose virtually all its water and prevented carbon
dioxide from becoming incorporated into its rock cycle.
Mars
Mars is smaller than Venus, but larger than Mercury. Its diameter
is 53 per cent, its mass 11 per cent and its density 72 per cent of
those of the Earth. Like Venus, its atmosphere is mostly carbon
306

dioxide. However, the surface pressure is less than a hundredth of
the Earth’s, because most of the original atmosphere has been lost
as a result of a combination of Mars’s weak gravity and lack of a
magnetic field to shield the upper atmosphere from erosion by the
solar wind.
Mars appears to have the cool, thick lithosphere that we would
expect for a planet of such small size. It has several impressive
volcanoes, including the largest in the Solar System. This is
Olympus Mons, whose summit is some 24 km above its base.
It would dwarf the Big Island of Hawaii, which is the Earth’s
largest volcanic edifice (Figure 14.6). The vast bulk of Olympus
Mons can be supported only because the lithosphere is so thick
and strong. Moreover, the concentration of volcanic activity at
this point over many hundreds of millions of years suggests that
Mars’s lithosphere is stationary relative to the upwelling plume
within the mantle that presumably feeds the volcano.
Olympus Mons
100 km
Mauna Kea Mauna Loa

sea level
Figure 14.6 Cross-sections to show the comparative sizes of Olympus

Mons on Mars and the Big Island of Hawaii (its two largest volcanoes are
named). The horizontal scale is as shown, but the vertical scale of both
cross-sections is exaggerated fivefold.
Large though it is, Olympus Mons covers too small an area

for impact crater statistics to give a reliable estimate of its age.
The best guess is that it last erupted about 30 million years ago.
There are several smaller (but still very large) volcanoes nearby
and others elsewhere on the globe.
For a smallish planet, Mars has a surprising number of
exceptional features in addition to its volcanoes. The most
fundamental is a twofold division into an ancient, highly
cratered highland terrain occupying most of the southern
hemisphere and a younger, low-lying terrain in the north. The
highland terrain has many large channels upon it, some of them

in dendritic networks that can hardly have been cut by anything
other than flowing water and would appear to have been fed
by rainfall. The younger surface age of the northern lowlands
is probably due to deposition of sediment transported from the
highlands. It has even been suggested that part of the northern
hemisphere was once occupied by a shallow sea. Whether or
not that is correct, Mars had so much water in the past that it
would be rather surprising if primitive life had not developed,
and searching for the conditions necessary to sustain life and
ultimately to find life itself are prime goals of missions to Mars.
Liquid water would not be stable at the surface under the low
atmospheric pressure that prevails on Mars today, so clearly
its climate must have been very different when water was
flowing in the large channel networks. Mars has lost much of its
atmosphere and cooled down over time, in contrast to Venus,
which has warmed up and developed a dense atmosphere.
Based on ages inferred from crater counting, the ‘wet’ period of
Mars’s history had ended by about 2 billion years ago. Images
and chemical analyses made by robotic rovers on the surface have
found convincing proof of Mars’s wet past, including cross-bedding
produced by flowing water (Plate 14), and minerals that can only
form in hydrous (wet) conditions. Water still occurs in the form
of ice in the northern and southern polar caps. In winter these are
enlarged by a seasonal covering of frozen carbon dioxide. There
is at least one example of what may be a dust-covered frozen sea
near the equator (Figure 14.7) and a large amount of water may
also be stored as ice within the Martian subsoil.
Spotlight: Opposing camps

The frozen sea and other ‘glacial’ features on Mars are subjects
of vigorous and sometimes downright hostile debate. An opposing
school of thought dismisses the ‘frozen sea’ as platy lava flows,
and finds volcanic explanations for every other ‘glacial’ feature.
Robotic rovers in the right places might settle the issue, but my
guess is that it will rumble on until a few examples have been
explored in person by humans. If any young reader is looking for an
exciting career goal, look no further.
308

Figure 14.7 This 40-km-wide view of Mars, from the European Space
Agency’s Mars Express orbiter, may show the dust-covered remains of a
frozen sea.
High-resolution images from orbiting spacecraft have revealed

gullies in crater walls, some of which appear to be experiencing
brief episodes of flow even today (Figure 14.8). The mechanism
for melting and releasing this water from the subsoil is unknown.
Even the longest young gully is only a few kilometres in length.
They are on a far smaller scale than the ancient dendritic
networks, and most of the water flowing down them must
evaporate into Mars’s atmosphere before it reaches the bottom.
Figure 14.8 Views of part of the inner wall of the same Martian crater in
2001 (left) and 2005 (right). Several narrow gullies can be seen. A new pale
deposit in a gully in the lower left of the 2005 view suggests that water
flowed in this gully between the two dates.

With the exception of these local and ephemeral discharges of
water, sediment transport and erosion of surface features are
today restricted mostly to landslides and wind effects. Extensive
fields of sand dunes occur, and dust storms sometimes obscure
the surface for weeks on end.
Mars is also notable for Valles Marineris, a 4000-km-long canyon
system in the southern highlands that reaches 7 km in depth and
over 600 km in width. This would dwarf Earth’s Grand Canyon
even more than Olympus Mons would overshadow Hawaii.
It probably owes its origin to tectonic fracturing, but has been
extensively modified by flowing water and landslides.
Key idea: Mars

Mars has large extinct volcanoes and ancient major valley systems
carved by flowing water. Young gullies, a few proven to be still
active, indicate minor local events when water flows at the surface.
The relative importance of lava flows and ice in forming some
landscape features is contentious.
The asteroids
As noted in Chapter 2, the asteroids (mostly orbiting the Sun
between Mars and Jupiter) are probably the remnants of
planetesimals, prevented from accreting into a planet because
of gravitational perturbations caused by Jupiter. Most are rocky
bodies, though some appear to be metallic (like the Earth’s core).
Meteorites have been studied in laboratories for centuries. These
have taught us much about the age and conditions of formation
and subsequent evolution of the meteorite parent bodies, which
are clearly almost all asteroids. Common meteorites are merely
fragments of ejecta thrown off asteroids during collisions. A few
rare meteorites come from the Moon and Mars, from which they
have been flung as ejecta from large impacts.
Images and other data from Vesta (the third largest asteroid,
500 km in diameter), obtained by NASA’s Dawn mission, which
orbited it in 2011 and 2012, showed a heavily cratered basaltic
crust (Figure 14.9) and a bulk density that requires the presence
310

Figure 14.9 A 200-km-wide view of the asteroid Vesta, seen by the Dawn
space probe. The steep curved slope near the centre seems to have been
the source of a major landslide.
of an iron core about 110 km in radius. To enable such a core-

crust differentiation to develop, Vesta’s interior must have been
hot at one stage. If this is correct, then the explanation probably
lies with short-lived radioactive isotopes, such as aluminium-26
(26Al), which we have good reason to believe could have been
very significant heat sources during the Solar System’s first
hundred million years or so.
Remember this!
Dawn (so named because its purpose was to investigate the dawn
of the Solar System) was launched in September 2007. After its
orbital tour of Vesta it was directed onwards to Ceres, the largest
asteroid, which it began to orbit in March 2015.
Various smaller asteroids have also been visited by spacecraft

(Figure 14.10). These are confirmed to be irregular in shape,

Figure 14.10 An oblique view across a 1.4-km-wide part of the asteroid
Eros, seen from orbit at a height of 5 km in 2000 by NASA’S NEAR-
Shoemaker probe. Impact craters can be seen, and also boulders up to
60 m in diameter.
but there is considerable variation in the abundance of impact

craters. Some have curious dust-filled depressions and a
surprising number of boulders strewn across their surface.
Key idea: Asteroids

Asteroids are small bodies that have been lithospheric throughout
most of the age of the Solar System.
Io
Io is the innermost of Jupiter’s four large satellites. By analogy
with the Moon (which it resembles in size, density and inferred
composition – except for having a larger core), we would expect
Io to be a densely cratered body with few signs of geological
activity in the past 2 or 3 billion years.
312

Spotlight: Moon or satellite?
Any planetary body that orbits a more massive one is a satellite of
the larger body. You can also call it a moon of the larger body. The
two terms mean the same thing. The Earth has one moon, which is
called the Moon, with a capital M, because that’s its name.
The reality is very different, because Io is actually the most

volcanically active body in the Solar System. Up to about ten
volcanoes are erupting at any one time (Plate 15), and the
surface is being covered by volcanic material at a rate more
than adequate to bury impact craters as fast as they are formed.
There are both explosive eruptions, in the form of 300-km-
high eruption plumes, and lava flows. When first seen on
images from the Voyager space probes in 1979, the lava flows
were widely interpreted as sulfur because of the red or yellow
colour of Io’s surface. Subsequent thermal infrared studies
demonstrated that the temperatures in the volcanic vents are
too high for molten sulfur, so it seems that we are dealing with
Earth-style silicate lava (basalt or an ultrabasic lava). Io’s colour
is almost certainly caused by sulfur, but it is merely a surface
dusting, as is sometimes seen on lava flows on the Earth.
Radiogenic heating cannot account for Io’s prodigious volcanism.
The explanation appears to be tidal heating of Io’s interior,
caused by tidal deformation of Io’s shape as it orbits Jupiter.
Key idea: Io
As a result of tidal heating, Jupiter’s satellite Io is the most
volcanically active body in the Solar System.
Europa
Europa circles Jupiter outside the orbit of Io, and completes one
orbit in exactly the time taken for Io to complete two orbits.
This relationship is responsible for maintaining the rate of tidal
heating in the two bodies, which is weaker in Europa than in
Io. Europa is slightly smaller and has an icy surface. However,
its average density is not much less than Io’s, and so the ice can

be no more than about 100–150 km thick above its silicate
crust and mantle. Close-up images of Europa’s surface show
good evidence that only the outer few kilometres of ice are solid
(Figure 14.11), below which there is likely to be a global ocean,
overlying a less active, submarine, version of Io.
Figure 14.11 A 35-km-wide region of Europa. The icy surface, which

has a characteristic ridge-and-groove pattern, is relatively intact in the
north-west but has begun to fracture. The further south-east you look, the
more fractured the ice becomes, until all you see is isolated rafts of ice
dispersed across a now refrozen sea.
If there are zones of hot rock not too far below the ocean floor,
then surely there is hydrothermal circulation and chemical
reaction between rock and water, and thus a likelihood of
hydrothermal vents on the ocean floor (like the ‘black smokers’
described in Chapter 5). As you read in Chapter 13, this is
the setting where life probably began on Earth, so why not
on Europa as well? Missions to Jupiter are both difficult and
expensive, but such is the lure of finding life that space agencies
are planning experiments that can be done from orbit (and
ultimately on Europa’s surface), to prove the existence of the
ocean and to find ways to sample it – by landing beside tidal
cracks where slush from below is squeezed up, or even by drilling
through the ice to allow a robotic submarine to go exploring.
314

Key idea: Europa
Europa may harbour life, thanks to the likelihood of an ocean below
its icy surface and tidal heating to power hydrothermal vents.
Large icy bodies

There are several other icy satellites that should interest the
geologist. The next one out from Jupiter is Ganymede (Figure
14.12), which is bigger than the planet Mercury. However, its
density is much less because it consists of about three parts ice
and one part rock (Figure 14.13). This preponderance of ice
over rock is a common factor from now on. Jupiter has a fourth
large satellite, Callisto, beyond Ganymede. Saturn has one large
icy satellite, Titan (that rivals Ganymede in size), and six others
more than 400 km across; Uranus has five satellites larger than
this; Neptune has two, of which Triton (2700 km across) is
the largest and most interesting. The Solar System’s inventory
Figure 14.12 A 110-km-wide area on Ganymede. The large impact crater

in the south-west has been rifted apart by faults associated with north–
south extension. The strange elongated crater in the south-east was
probably formed by the impact of a comet that had broken into at least two
fragments before it struck the surface.

ice (mantle)
2000 km
rock (core)
Figure 14.13 Cross-section through a large icy body, showing
differentiation into a dense rocky core surrounded by less dense ice. The
scale bar is for Ganymede (the largest example of its class), but many
smaller icy satellites and bodies such as Pluto probably have a similar
structure. Ice is not necessarily just frozen water; it can include other
‘volatile’ substances such as methane, ammonia, carbon dioxide and
nitrogen, and there may be one or more water-rich liquid layers within it.
of large icy bodies is completed by the largest of the icy dwarf

planets in the Kuiper Belt beyond Neptune, such as Pluto
(2300 km diameter), its most substantial satellite, Charon
(1250 km diameter), and Eris (2400 km in diameter).
We can ignore the hundred or so other known satellites of the
outer planets that are less than about 400 km across, and a huge
number of Kuiper Belt objects below the same size threshold,
because below this size the body’s own gravity is too weak
to pull it into a spherical shape. Unlike some of the asteroids,
we have no reason to suspect internal differentiation or other
geological processes in such irregular bodies.
It may seem strange to deal with predominantly icy bodies in
a book about geology, but their relevance should become clear
to you as I review the properties of ice. At the low surface
temperatures prevailing in the outer Solar System (–150°C at
Jupiter and –240°C at Neptune), ice is extremely strong and
rigid, just like rock on the Earth. However, in the interior of
an icy body with a supply of heat, ice is capable of flowing by
convection without melting. This distinction between rigid and
convecting behaviour is exactly the distinction described in
Chapter 2 between the lithosphere and asthenosphere of the
Earth (or, indeed, any other terrestrial planet), except that we
are dealing with ice rather than silicates.
316

A further parallel with silicates arises because of how ice melts.
Unless it is pure, ice does not melt at a single temperature.
Salty ice (such as we might expect in Jupiter’s icy satellites
because of reactions between water and rock) begins to melt at
a few degrees below 0 °C, sweating out a brine instead of pure
water. Ice that consists of a mixture of frozen water and frozen
ammonia (as is likely at Uranus and beyond) begins to melt at
about –100 °C, yielding a liquid that is about two parts water
and one part ammonia. Both are examples of partial melting,
just like what happens in the mixtures of silicate minerals that
make up rock. ‘Ice’ becomes even more like mineral-rich silicate
rock at Neptune, where methane and carbon dioxide are added
into the mix.
Thus, with a lithosphere/asthenosphere distinction and the
potential to produce compositionally distinct ‘magma’ by
partial melting, and thus capable of contributing to a crust
that is chemically distinct from the mantle, icy bodies have
the essential attributes of a terrestrial planet. There is too little
rock within them for much radiogenic heating but, fortunately
for the geologist, most of them show evidence for current
or previous episodes of tidal heating. For example, parts of
Ganymede’s surface are intensely fractured, though the number
of superimposed impact craters shows that the fracture pattern
was created in the distant past.
There is not space here to describe each of these fascinating
bodies. Highlights include a 500-km-diameter satellite of Saturn
called Enceladus, whose surface is criss-crossed by fractures.
Images from the Cassini spacecraft orbiting Saturn have
captured jets of water-ice crystals tainted by traces of methane,
ammonia, carbon monoxide, carbon dioxide and various simple
organic molecules venting into space from a set of fractures near
the south pole. These plumes are presumably powered by tidal
heating (Figure 14.14) and are the icy equivalent of pyroclastic
volcanism.
Moving out to the satellites of Uranus, we find features such
as ‘lava’ flows (sketched in Figure 14.15), which are thought
to be composed of an ammonia–water mixture produced by
partial melting. It is known that such a mixture is vastly more

Figure 14.14 Main view: a 120-km-wide region of Enceladus, revealing the
intensely fractured nature of its crust in this region. The largest fractures
are warm, and are the source of ice particle eruptions. Inset: a reduced-
scale view of part of Enceladus, seen looking towards the Sun, so that only
a narrow crescent is illuminated. Jets of ice particles erupted into space
(from the region shown in the main view) are revealed by the way they
scatter the sunlight.
50 km (approx)
Figure 14.15 Sketch made from an oblique view of part of Miranda (a

470-km-diameter satellite of Uranus) that has a complex geological
history. Several impact craters are visible (roughly circular). One 10-km
crater in the lower left has been partly obliterated by a resurfacing event,
whereas another is superimposed over the boundary and so must postdate
the resurfacing. Two icy ‘lava’ flows are shown in bold outline.
318

viscous than pure water, and would behave very much like a
silicate lava flow on Earth. Neptune’s large satellite Triton has
a seasonal polar cap of frozen nitrogen, punctured by geysers
powered by pressurized escape of nitrogen gas, and a surface
geology dominated by other ices including methane, carbon
dioxide and ammonia (Figure 14.16), although the most
abundant ice in the mantle is probably water.
Figure 14.16 A 600-km-wide region of Triton. The ragged fringe of the

solar polar cap of nitrogen ice runs across the bottom. Beyond this, a
complex dimpled icy surface is cut by fissures and buried by patches of
smoother material that may be icy lava flows.
Key idea: Ice lithospheres and asthenospheres

In icy satellites where ice predominates over rock, the ice functions
mechanically as lithosphere and asthenosphere, and may be
compositionally differentiated into a crust and mantle. In these
bodies, the rock is a deeply buried core. The same may be true
inside the icy dwarf planets of the Kuiper Belt.

To conclude, I will backtrack to Saturn’s largest satellite, Titan.
Based on the diversity of geological processes that it exhibits, this
is the most Earth-like body in the outer Solar System. Titan has a
dense atmosphere in which the most abundant gas, as on Earth,
is nitrogen. The atmospheric pressure at the surface is 50 per cent
greater than Earth’s; but rather than oxygen, carbon dioxide and
water vapour (as on Earth), the rest of Titan’s atmosphere is
mostly methane and it is much colder (−180 °C at the surface).
It has a high-altitude smog layer (which makes the surface hard
to see) made of hydrocarbon molecules created by the linking
together of methane under the influence of solar ultraviolet
radiation. Neither active icy volcanism nor tectonics has been
proven on Titan, though they seem likely, and there are certainly
wind-blown dunes (made of ice grains rather than sand).
However, what makes Titan so special is that it has its own
equivalent of Earth’s hydrologic cycle. Instead of water, this
is based on methane. From time to time (possibly seasonally),
methane condenses in the atmosphere to make methane rain.
This feeds some impressive dendritic river networks. Low-lying
areas across most of the globe are currently ‘dry’ (and home to
vast fields of wind-blown dunes), but in the polar regions there
are lakes and seas (presumed to be liquid methane) that are up
to several hundred kilometres across (Figure 14.17). One day,
a geologist from Earth may stand on the shore of one of those
methane seas to study coastal erosion (such as you saw at the
end of Chapter 8), and sediment transport on beaches of icy
sand (analogous to the terrestrial beaches in Chapter 9).
Key idea: Titan

Titan has a ‘methanological’ cycle analogous to Earth’s
hydrologic cycle.
Remember this!
Titan has been explored in detail only by the NASA–ESA orbiter
Cassini, which began to orbit Saturn in 2004. The only probe to
land on a satellite of another planet, Huygens was detached from
Cassini and parachuted to Titan’s surface in January 2005.
320

Figure 14.17 A 250-km-wide region of Titan, seen by the imaging radar
on board the Cassini orbiter, showing a rugged coastline with numerous
offshore islands. This is a very Earth-like scene, except that the land is
made of ice, and the sea is made of liquid methane.
Geology among the stars

By early 2015, planets had been discovered at about 1200
other stars. These are referred to as ‘exoplanets’. At first,
most of the discoveries were Jupiter-sized or bigger, orbiting
rather close to their stars, because these are the easiest to
detect. However, rocky planets comparable in size to the
Earth and cool enough to have liquid water have begun to
show up as well. In our lifetimes, we should get spectroscopic
data on exoplanets’ atmospheric compositions, but we
cannot expect to see their surfaces in sufficient detail for
any kind of geological study. The best we might hope for
is to detect the occurrence of supervolcano eruptions by
means of their effect on atmospheric composition. However,
with many unsolved questions remaining about the Earth,
and the geological exploration of our Solar System barely
begun, there is plenty to keep the geologist busy, as well as
fascinated, nearer to home.

Dig deeper
If you want to know more about the geology of planets and
moons, then I can hardly do other than recommend two books
of my own: A Very Short Introduction to Planets (Oxford University
Press, 2010) and A Very Short Introduction to Moons (Oxford
University Press, 2015). There’s a related video of me enthusing
about planetary geology and why Mercury is so intriguing at
http://ukcatalogue.oup.com/product/9780199573509.do
A highly readable account of the exploration of Europa, arguably
the icy moon most likely to host life, and which lays bare some
quite bitter infighting between rival schools of thought, is
Unmasking Europa: The Search for Life on Jupiter’s Ocean Moon by
Richard Greenberg (Praxis Publishing, 2008).
If you are intrigued by the possibility of detecting volcanic
eruptions on exoplanets, see http://www.universetoday.
com/73103/extrasolar-volcanoes-may-soon-be-detectable/
The best website for finding images of planetary surfaces is
http://photojournal.jpl.nasa.gov/
There are 30 hours of free online learning about moons available
at https://www.futurelearn.com/courses/moons and http://www.
open.edu/openlearn/science-maths-technology/moons/content-
section-overview
322

Fact-check
1 Why are almost all impact craters nearly circular in outline?
a Because that is the shape of most impacting bodies
b Because most impacts occur vertically
c They are excavated by shock waves radiating from the
point of impact, so the direction of impact doesn’t matter
d They are blasted out from below by shock waves reflected
from the Moho
2 Why does the Moon show fewer signs of recent geological

activity than the Earth?
a It contains less heat-producing radioactive elements
per kilogram than the Earth
b It has no atmosphere
c Because most have been buried below the lava flows that
formed the lunar maria
d Because its greater surface-to-volume ratio has allowed it
to cool down more rapidly
3 What is anorthosite?
a A lunar rock-type that lacks the mineral ‘orthosite’.
b A lunar rock-type representing its flotation crust
c A mineral produced by shock-metamorphism when a
hypervelocity impact occurs in basalt
d A mineral produced by shock-metamorphism when a
hypervelocity impact occurs in siliciclastic sediments.
4 Which rocky planetary body other than the Earth generates its
own magnetic field?
a Mars
b Mercury
c The Moon
d Venus

5 What is the most likely explanation for the much greater total
amount of carbon dioxide in Venus’s atmosphere, compared to
the Earth’s?
a The planetary embryos from which Venus formed
contained a much greater proportion of carbonaceous
material
b Most of the Earth’s atmospheric carbon dioxide has
become locked up in carbonate rocks
c There has been much more volcanic outgassing of carbon
dioxide on Venus than on the Earth
d Earth’s atmospheric carbon dioxide has been largely
stripped away by the solar wind
6 Which of these does Mars have (or probably have)?

a A iron-rich core
b A dense atmosphere
c A magnetic field
d A thin lithosphere
7 Which of these statements is/are correct?

a Asteroids are the source of all known meteorites
b Asteroids and comets are responsible for the craters on
the Moon
c The asteroid Eros has no visible impact craters on its
surface
d Some asteroids are internally differentiated
8 Which moon has active volcanoes erupting silicates?

a Ariel
b Europa
c Io
d Titan
9 Which moon experiences rainfall?

a Enceladus
b Europa
c Titan
d Triton
324

10 Where would you find Valles Marineris?
a Mars
b Mercury
c The Moon
d Venus
11 Where is the largest volcano in the Solar System?

a Earth
b Io
c Mars
d Venus
12 Which moon is venting plumes of ice crystals to space?

a Ariel
b Enceladus
c Europa
d Titan

15
Going into
the field
33about the importance of making
intelligent observations in the field
33to develop a safe and responsible
attitude to fieldwork.
327

Using spacecraft to explore the surfaces of other planetary
bodies is both fun and informative. Detailed images of the
Earth itself from space can also be illuminating, but there can
be no denying that, for many geologists, whether professionals
or hobbyists, the main rewards from geology come from doing
it in the field. Fieldwork can be a great source of pleasure, but
it is also by far the best way to appreciate the sort of evidence
from which most of our understanding of the Earth’s history is
derived. If you can possibly get yourself into the field to look at
some real rock, I urge you to do so.
Spotlight: Fieldwork
When I was a student doing fieldwork in the Arabian desert, our
local contact in the relevant government office asked me, in some
confusion, ‘Where exactly are these fields that you work in?’
The term fieldwork is one of those geological jargon words that
sometimes confuses laypeople. However, in this case at least it is a
term that we have in common with botanists, zoologists, ecologists
and even sociologists. When geologists speak of ‘going into the
field’, they mean spending time outdoors in a place where the rock
succession or geological process to be investigated can be seen.
Thus, fieldwork can be conducted in places as diverse as a sandy
beach, a roadside cutting, a remote stretch of moorland or the
flanks of an active volcano.
If you belong to a geology club, or are following a course

of study in geology, you should have opportunities to go on
organized field trips. If you have never been in the field and
have decided, after reaching this stage in the book, that you
want to go deeper into geology, then I urge you to get out
there and begin looking at some rocks and landscapes. This
chapter suggests some ways to begin practical geology in the
field. It therefore describes things to look out for, discusses
the equipment you will find useful, and suggests a code of
practice to enable you to work safely and without annoying any
landowners.
328

Key idea: Fieldwork
Fieldwork is studying rocks or geological processes in situ. It can
give you insights that you will never learn online or from a book.
Things to do in the field

It is sensible to decide in advance what you want to achieve
during a field trip. For example, are you just going out for a
general look at the rocks, or do you want to search for the answer
to a particular riddle? If you are going in a group organized by
somebody else, you can leave the decisions about where to go and
what to look at to them, but to get the most out of a field trip it
pays to approach each locality in a methodical fashion.
When you arrive at a rock exposure (a place where bedrock is
visible), perhaps a cliff face, do not rush straight up to it. I will
have something to say about the safety of working under cliffs
shortly, but it makes scientific sense to stand back for a while
anyway.
You should spend time taking in the general features of your
site. These are best seen from a distance. The sort of thing to
look out for is whether or not there is any layering in the rock,
which might take the form of bedding, layers within an igneous
intrusion, metamorphic foliation, or a set of fractures (joints).
You may not be able to decide which of these four you are
seeing, but for now you should be content with spotting the
pattern; you will probably be able to distinguish between the
options when you get within touching distance.
While you are still ten or more metres away, you should look
for variations. For example, if you can see some layering, try
to decide whether these layers maintain a uniform orientation
throughout the exposure, or whether they are folded.
You should also try to work out whether layers are continuous,
by visually tracing some of them sideways. If they are not, try
to decide whether this is an inherent property of the layers
(perhaps each one represents a deposit laid down in a channel
that had a finite width, or maybe you are seeing large-scale
cross-bedding of the sort that occurs in wind-blown sand

dunes), or whether originally continuous layers have been
truncated by an igneous intrusion or by a fault (which will
almost certainly visibly offset the layers).
If the structure within an exposure is complicated, making a
simple sketch can often help you begin to sort things out. This
should not be a piece of representational art, showing every crack,
shadow and tonal variation. Rather, it should show only those
features that are of geological significance. For example, bedding
is significant and should be drawn, but if you try also to show all
the irregularities in a rock face left either by natural weathering
or quarrying then you risk obscuring the essential features. Always
include a scale in a field sketch. Most adults are 1.6–1.8 m tall,
so if you can persuade somebody to stand (safely) against your
exposure for a moment, you can very easily mark a scale-bar on
your sketch. A sketch is usually more useful than a photograph,
because it shows only what you need and is easy to annotate while
you are in the field. Whether you take a sketch or a photograph, it
is important to note what direction you were looking in, so record
the compass direction to the centre of the picture.
Before you finish your preliminary assessment, choose a suitable
place to begin your close-up examination. Avoid approaching
anywhere you are likely to be hit by material falling from above,
or where the ground beneath you could be unstable. It is best to
try to make sense of any strange areas after you have worked
out the main story, so choose a spot that looks representative of
the exposure in general.
Key idea: Assessing an exposure

Rock exposures are best assessed at a distance, before going
up close.
By the time you arrive at the rock face, your initial observations
will have suggested a number of checks needed to test any
hypotheses that have begun to form in your mind. If you have
decided that this exposure was probably sedimentary rock, you
will expect to be able to confirm this by close inspection, which
should reveal the individual grains making up the sediment
330

(unless these are very fine), small-scale bedforms and (if you
are lucky) trace fossils and/or body fossils. On the other hand,
a coarse-grained igneous rock will manifest itself with clearly
visible interlocking crystals, including minerals of the sort that do
not survive transport in the sedimentary regime (see Chapter 8).
In order to identify the rock, you will need to recognize and
identify its main minerals, with the aid of Appendices 1 and 2, or
a more detailed field guide to rocks and minerals.
You may see features close up that are worth sketching or
photographing. Examples would be small-scale cross-bedding,
fossils, or the shape of the contact at the edge of an igneous
intrusion. Just as with any more distant sketch (or photograph),
you should try always to include an object that will give a clue
as to scale.
Spotlight: Scales for photos

A coin, a lens cap, a hammer or a finger are probably the most
common scale objects in geological photos. The last has the
advantage that you are unlikely to accidentally leave it behind when
you leave.
If you are not visiting a quarry or a cliff, you are probably

dealing with comparatively small exposures, unless you are
in an arid environment where there is little or no soil cover.
However, the principles outlined above still apply. Stand back
and assess the visible area of rock as a whole, however small,
before rushing in and pressing your nose against it.
Key idea: The importance of a scale

Include items to give an indication of scale in sketches and photos.
Field equipment
It is a common misconception that geologists spend their time
digging up rocks and fossils. In fact, it is rare for a geologist
to go into the field armed with a spade. The trick is to go

somewhere where the things you want to see are already at
the surface. In many cases, the natural weathering of the rock
picks out features more clearly than they can be seen on freshly
exposed surfaces. Exposures of rock are most common in hilly
or mountainous regions, especially along water courses, but they
may be particularly good on the coast. Quite apart from the
damage and disturbance that digging would cause, the labour
involved in removing two or three feet of topsoil will rarely be
repaid by the small patch of dirty and possibly frost-shattered
bedrock that this would reveal.
The most important items of field equipment are a perceptive
eye and an enquiring mind. The latter might be stimulated
by appropriate reading, and the former can be augmented by
carrying a hand lens.
A typical hand lens is smaller than a Sherlock Holmes-style
magnifying glass, but provides a magnification of about
ten times. This can be extremely helpful for such things as
determining the shapes (angular or rounded) and textures
(frosted or glassy) of sedimentary grains, spotting cleavage
traces in minerals, and making out fine details in fossils. There
are Apps for smartphones that turn their camera and screen into
a magnifier, but beware of running your battery down.
You should also carry a notebook, for recording your
observations and making sketches. Pencil is best because it
can be rubbed out to correct mistakes, and does not smudge
as badly as ink when it rains. If you might want to go back
on another day to check or extend your observations, it is
important to record your exact location. Marking this on a map
and/or a sketch map of your own should do the trick. Noting
down co-ordinates read from any GPS (Global Positioning
System) device that you are carrying is second best.
Sometimes you may want to collect a small sample. If you are
just starting out as a geologist, it is a particularly good idea to
build up your own collection of the main types of rocks and
fossils (but see my later warning about selfish or indiscriminate
collecting). If you do find something that you want to take
away, then you need something to put it in, and some way of
332

labelling it. One specimen per small plastic bag (each containing
a sample number on a slip of paper) is a good way to proceed,
provided you record the location of each find in your notebook.
If not with spades, then it is with hammers that geologists are
usually associated. Unlike spades, hammers are a standard
item in many geologists’ field gear. In situations where the rock
surface is badly weathered or too covered by lichen for it to be
identifiable, a hammer can be invaluable for breaking off a piece
to expose a fresh surface. However, natural freshly broken pieces
can very often be collected near the foot of cliffs, in which case
there is no need to hammer the rock face at all. The same goes
for collecting fossils; you will usually find many more, and avoid
damage, by searching through rock falls or spoil heaps. If you
do use a hammer, then it must be a purpose-made geological
hammer. The head of a hammer made for knocking in nails
will not be made of adequately tempered steel, and is liable
to splinter if you hit a hard rock with it. A steel shard from a
shattered hammer can blind you if it catches you in the eye.
Key idea: Hammers

Hammer (if at all) with discretion and care.
The other standard item of field equipment is a device called a

compass clinometer, which is used to measure the orientation
of bedding, foliation, fault planes and so on. This is particularly
important if you are being ambitious enough to draw your own
geological map. It is valuable too when, for example, you are
visiting several isolated exposures of a sedimentary rock and
want to determine whether the bedding in each one has the
same orientation. If it does not, then the rock succession has
probably been folded.
A compass clinometer has a magnetic needle, functioning as in
any ordinary magnetic compass, and a second, non-magnetic
needle that is pivoted from one end so as to hang vertically
under its own weight. To appreciate how to use a compass
clinometer, it is necessary to understand how the orientation of
any plane surface can be specified in a unique and unambiguous

manner. This can be done with just two measurements. The
first is to imagine a horizontal plane intersecting the geological
surface (such as a bedding plane) whose orientation we want
to determine. The two planes must intersect at a horizontal
line and the angle between this line and north is defined as the
strike of our geological surface (see Figure 15.1). To measure
the strike, the base-plate of the compass clinometer is held
horizontally, with one edge pointing along the strike direction.
The angle that this edge makes with the free-floating magnetic
compass needle is the strike, and can be read on the dial. The
only other measurement we need to know is how steeply the
geological surface is dipping, relative to horizontal. This is
called the dip of the surface. To determine this, the compass
clinometer is held with one edge running directly down the
steepest gradient that can be found on the geological surface.
The dip can then be read using the free-hanging needle.
dip (degrees)
bed of rock
North
strike (degrees)
dip (degrees)
imaginary horizontal surface
Figure 15.1 Definitions of strike and dip. A bed of rock is used in the
example, but the definitions apply to any geological feature (such as a
dyke, a joint or a fault) whose local orientation can be expressed as a
plane. Strike is the angle, relative to north, made by any horizontal line
on this geological surface. Dip measures how steep the surface is, and is
defined as the angle of slope measured downwards from horizontal.
In the example shown in Figure 15.1, the strike is 150° and

the dip is 35°. This would be recorded in our field notebook
as 150/35. This is unambiguous if strike is always measured
clockwise from north (as in this example), but to play safe it is
common to note down the approximate direction in which the
surface is dipping. In this example it is dipping down towards
334

the north-east, so we could record 150/35NE. It is standard to
omit degree symbols when recording strike and dip, because in
handwritten field notes a degree symbol can later be mistaken
for a number zero. The exact direction down-dip (which you
do not need to record) must be 90 degrees different from the
strike; in this case towards a compass bearing of 150 − 090 =
060 degrees. Figure 15.2 shows a compass clinometer in use.
Figure 15.2 A compass clinometer being used to measure the strike of the
edge of a pale igneous dyke, which has been intruded into a darker rock.
The photograph was taken looking down on to a roughly horizontal surface.
In this example, the strike is approximately 160° relative to magnetic north.
The edge of the dyke dips at 80° towards the east (left in this view), so the
strike and dip of the dyke would be written down as 160/80E.
Key idea: Compass clinometers

Use a compass clinometer to record the orientation (strike and dip)
of bedding and other planar features.
If you want to record your visit photographically, it is best to

have a camera with adjustable focus, or even a choice of lenses.
Some landscapes are best recorded using a wide-angle lens, but
detailed close-up photographs require a macro lens. Digital
cameras allow you to check your photo as soon as you have

taken it. This is very useful, but beware that the resolution of the
camera display may be too poor to confirm that fine details are
in focus. These days, smartphones can take pretty good pictures,
and have a large enough screen for you to check the focus.
Other items to include in your field equipment depend on the
nature of the rocks that you want to examine. If it is in a poorly
cemented sandstone, you may find a trowel or similar object handy
for scraping away loose grains in order to reveal the original cross-
bedding. If it is a shale and you hope to find fossils, a chisel comes
in handy for splitting the rock along its bedding surfaces.
Finally, if there is any likelihood that your field trip will involve
working close to steep or vertical faces above 3 metres high, then
you should take a hard hat. Hard hats are cheap and readily
obtainable from builders’ merchants and do-it-yourself stores.
I usually wear a slightly more expensive climber’s helmet, which
is less bulky and comes with a chinstrap.
Field safety
In the field, you should take care to minimize the likelihood
of injury to yourself and to others. As already noted, the
most basic precaution is to assess the inherent risks before
approaching a rock face. You should wear your hard hat
whenever there is the slightest chance of anything gravel-sized
or bigger falling on to you. While it won’t help you survive a
major rock fall, a hard hat can save you from concussion or a
worse head injury caused by small lumps of debris.
Key idea: Hard hats

Always wear a hard hat if working close to steep faces.
If it is absolutely necessary to use a hammer, then you must

make sure that nobody is within about 3 metres of you. Rocks
such as flint and many igneous varieties are particularly hard
and brittle. Hammering these can send sharp splinters flying.
Most older geologists have met someone who has lost the
sight of an eye in this way, and it is foolish to hammer without
wearing goggles or safety glasses.
336

Be adequately clad. Walking boots with ankle protection are the
most sensible footwear in rough terrain, and you should carry
protection against sun, rain and cold as appropriate. When
visiting a coastal exposure, be aware of the state of the tide
and plan ahead to avoid being cut off. When visiting a remote
area, general backcountry rules apply: be wary of dangers such
as being caught in a gully or canyon in the event of a flash
flood; make sure you have enough food and drink to survive if
your return is unexpectedly delayed; and leave a copy of your
intended route with someone in case you have to be rescued.
Take a mobile phone or similar means of communication.
Key idea: Safety

Plan for you own safety, and that of any colleagues, before you go
into the field.
Fieldwork code of practice

The rules governing geological collecting and other activities vary
from place to place. However, there are various widely applicable
guidelines that make sense wherever you are.
• If you are entering a quarry or other private land, seek

permission. Keep to paths, close gates behind you, and do not
climb over walls or fences. Quarry owners will usually insist that
you wear a hard hat at all times, and that you stay well clear of
machinery.
• If you are working in a national or state park, nature reserve or
other conservation area, then find out the rules about collecting.
Some places will allow you to collect samples only from loose
debris, while others forbid collecting entirely.
• Even where there are no local rules, be aware of the damage you
might cause. Hammer, if at all, with discretion. Do not damage
rock faces, and do not try to extract fossils or mineral specimens
that are well embedded. You will usually end up wrecking them,
so it is much more sensible to take a photograph and leave the
item in place for others to admire.

Geological maps and guidebooks
If you want to know about the geology of a particular area,
then an obvious first step is to buy a geological map. A good
bookshop will be able to order these for you, and some can be
obtained from online book suppliers. A geological map usually
shows the outcrop of each rock unit, which represents the
distribution of rock types that would be seen at the surface if
the superficial deposits (such as soil) were removed. Sedimentary
rocks are usually colour-coded by age, whereas igneous and
metamorphic rocks are usually distinguished by type rather than
by age. Major faults are usually shown by bold lines.
Spotlight: Outcrop and exposure

I’ve lost count of the arguments I’ve had with colleagues over the
terms ‘outcrop’ and ‘exposure’. As I’ve defined them in this book,
they are different – and it is a useful distinction. Some geologists
in the field refer to a patch of bedrock that is sticking out of the
ground as ‘an outcrop’, and this blurs the distinction, because the
bedrock next to it hidden below, say, a metre of soil is ‘outcrop’
too – just not exposed outcrop.
Most geological maps are published by geological survey

organizations. For example, the geology of the United Kingdom
is conveniently covered in two sheets at a scale of ten miles to
the inch (published by the British Geological Survey). Most of
the country is available on larger-scale maps, 1:50 000 being the
commonest. These include additional information such as strike
and dip, mineral veins, and so on. The equivalent organization
in the USA is the United States Geological Survey, but many
state geological surveys publish maps, too.
Although a map will show you the distribution of rock types,
it will not usually tell you where these are well exposed or
conveniently (and legally) accessible. This is where geological
field guides come in handy. These books usually contain sketch
geological maps, too, so for beginners, at least, there is often
no need to invest in a large-scale geological map. Many field
guides are written by enthusiastic geologists who know and love
338

the area, others are more in the nature of official guides round
geological trails in national parks.
Key idea: Field guides

A geological guidebook to a particular area can be expected to
tell you where the best and/or most easily accessible exposures
can be found.
Suggestions for field projects

Here are a few suggestions for geology fieldwork projects that
you could tackle after reading this book. There will probably
be one or two that can conveniently be attempted in your home
area or on your next holiday.
33 If there are quarries in your area, find out whether these
occur in every local rock type. If they do not, then try to
work out why some rock types are (or were) quarried and
others not.
33 Visit a river or stream and decide whether it is mainly
eroding or depositing sediment. Does it do the same thing
all year round? Could it be described as a braided or a
meandering channel? Can you find any abandoned meanders
(like the oxbow lake in Figure 9.7)?
33 Take your hand lens to the seaside. If it’s a sandy beach, you
should be able to identify quartz grains (clear, glassy) and
shell fragments (milky, opaque) among the sand particles.
What else can you see? Where do you think it has come from?
If it’s a pebbly beach, how many different rock types are
represented among the pebbles? Are they all local? Do they
match the rock types exposed in nearby cliffs? If not, why?
33 Find an exposure with fossils in it, and try to identify them
and work out how they lived (swimming, fastened to the sea
floor, in burrows, etc.). Are they all of the same kind? Are
most of the fossils complete or fragmentary? What can you
deduce about the environment in which the fossils (i) lived
and (ii) died? Are (i) and (ii) the same?

33 Is there any history of mining in your area? What is or was
extracted, and how was it formed? Can you find any signs of
subsidence? What, if anything, has been done to landscape
the spoil-heaps?
33 What are the sources of the stone used (i) to build houses, (ii)
to build churches, colleges, and municipal buildings, (iii) to
decorate the façades of banks, department stores and so on,
(iv) for gravestones? Is it all local, or has some been brought
in from tens or hundreds of miles away? Gravestones usually
bear dates; do some varieties of stone weather differently
(and at different rates) from others, and if so, why?
33 Visit some really big hills or mountains. Why are they
there? How many rock types are present? Can you find any
evidence of deformation (such as folding or cleavage)? Has
the landscape been eroded into its present form mainly by
the action of flowing water or by ice?
The possibilities for field projects are endless, and you may
think of ideas that appeal to you more than those I have
suggested. Best of all, join a geological society, which will offer a
programme of talks and field trips.
Dig deeper
A good book to help you learn how to make geological
observations in the field beyond the level established here is
Geological Field Techniques by Angela L. Coe, Tom W. Argles,
David A. Rothery and Robert A. Spicer (Wiley-Blackwell, 2010).
There are many colour guides available to help you identify rocks
and minerals in the field, such as Rocks and Minerals by Chris
and Helen Pellant (Bloomsbury, 20014) or Minerals, Rocks and
Fossils by A.C. Bishop, A.R. Woolley and W.R. Hamilton (Philips,
1999).
If your fieldwork is in the British Isles and you expect to find
fossils, the standard books perfectly well illustrated by black-
and-white sketches are British Palaeozoic Fossils, British
Mesozoic Fossils and British Caenozoic Fossils (2012, 1983, 1979)
published by the Natural History Museum. These should also be
340

useful if you are working near the British Isles or in Paleozoic
terrain that was close to any part of the British Isles during the
Paleozoic.
To help you begin fieldwork, try joining your local geological
society (look on the Internet). If you are in or near the UK,
then the Open University Geological Society (open to anyone
interested in geology) has local branches that meet for lectures
and run domestic and overseas field trips http://ougs.org
The Geological Society has fieldwork resources accessible via
http://www.geolsoc.org.uk/education
Geologists’ Association field guides to many parts of the UK and
Europe can be obtained via http://www.geologistsassociation.
org.uk/guidesales.html
The group of enthusiasts responsible for conservation of
important geological sites in the UK has a website at
http://wiki.geoconservationuk.org.uk
The online archive of the Geological Society of America field
guides (mostly north America, but some Europe) is at
http://fieldguides.gsapubs.org/content/by/year
And, finally, there are many apps now available for smartphones
and tablets that can simulate a compass clinometer or even show
you a geological map of your current location. In the UK, try
http://www.bgs.ac.uk/opengeoscience/

Fact-check
1 Strictly speaking, what is an ‘outcrop’ of a geological unit?
a Any cliff or knoll more than 5 metres high
b Any place where it sticks out of the ground
c The area where it would be visible if all superficial material,
such as soil, were removed
d A mining spoil-heap where samples of it can be found
2 Which of these is not a usual item of geologist’s field
equipment?
a A compass clinometer
b A hand lens
c A hard hat
d A spade
3 Which of these is definitely not the first thing you should do

when arriving at a new field locality?
a Examine the rock with your hand lens
b Record the location and date in your notebook
c Stand back and make a sketch
d Work out which way you are facing
4 In which direction does a stratum dip if its strike and dip are
215/30NW?
a 125
b 215
c 260
d 305
5 The orientation of which of these could you not record as strike

and dip?
a Bedding
b The contact between a dyke and the rocks into which it had
been intruded
c A metamorphic cleavage
d Slickensides
342

Appendix 1: Minerals
Many thousands of minerals have been documented. Guides
to identify even the few that are relatively common use many
pages to show colour photographs and to list their properties.
This appendix is not intended as a substitute for such a guide;
rather it is a brief introduction that indicates some of the ways
in which the very commonest minerals may be identified.
It is not straightforward to define what is meant by the term
‘mineral’. A simple statement that ‘a mineral is a naturally
occurring crystalline substance with a well-defined chemical
composition’ is a good starting point. However, this is too
restrictive to be an adequate definition, because the exact
chemical composition of a mineral may vary, even though the
arrangement of atoms within its crystalline structure is fixed.
On the other hand, some minerals that are regarded as distinct
from each other share the same atomic arrangement, and are
distinguished by differences in their chemical composition
(which may radically affect their colour or other properties).
Some minerals have identical chemical composition but with
their atoms differently arranged, giving the crystals totally
different properties. These are described as ‘polymorphs’. In
this book you have met the three aluminosilicate polymorphs
sillimanite, andalusite and kyanite, and two forms of calcium
carbonate: calcite and aragonite. A few substances of non-
crystalline form, notably amorphous (or ‘cryptocrystalline’)
silica such as chalcedony, flint and opal, are usually regarded as
minerals, too.
A thorough analysis of a mineral would require chemical
analysis, study of its optical properties using a microscope, and
X-ray analysis of its atomic structure. However, any common
mineral occurring as a crystal bigger than a few millimetres
long usually shows enough clues for the moderately experienced
geologist to make a fairly confident identification. There are
several things to look out for. These include lustre (does it
look metallic or glassy?), colour (but impurities can make this

misleading), cleavage (how many sets of cleavage planes, if any,
and at what angle do they intersect?), general shape (related to
cleavage in some, but not all, minerals), hardness (how easily
can it be scratched?).
The hardness of minerals is judged on Mohs’ scale of hardness,
which runs from 1 (softest) to 10 (hardest). A harder mineral
is able to scratch a softer one. Common minerals are used to
define each point on the scale, as show in the table below:
Table a Mohs’ scale of hardness
1 talc can be scratched by fingernail

2 gypsum can be scratched by fingernail
3 calcite can be scratched by copper coin
4 fluorite can be scratched by steel
5 apatite can be scratched by steel
6 feldspar scratches glass with difficulty
7 quartz scratches glass with ease
8 topaz scratches glass with ease
9 corundum scratches glass with ease
10 diamond scratches glass with ease
The most common rock-forming minerals, notable ore minerals,

and various precious and semi-precious stones are listed in the
tables below:
Table b Common rock-forming minerals (all these have a glassy lustre)
Name Hardness Distinguishing properties
Amphibole (e.g. 5.5 Dark mineral, may be greenish, often
hornblende) elongated crystals, in basic igneous
(Na,K) Ca2(Mg,Fe,Al)5 rocks and some metamorphic rocks, two
(Al,Si)5O22(OH)2 cleavages at about 60°.
Andalusite 7.5 Red, often altering to silvery or whitish
Al2SiO5 mica. May form cross-shaped crystals.
Occurs only in metamorphic rocks.
Polymorphs of the same composition are
kyanite and sillimante.
Anhydrite 3–3.5 Forms pale crystals, often breaking into
CaSO4 rectangular fragments.
Apatite 5 The most common phosphate mineral.
Ca5(PO4)3(F,Cl,OH) Crystals can take a variety of forms.
344

Biotite (mica) 2 Virtually black, one excellent cleavage, so
K(Mg,Fe)3AlSi3O10(OH)2 that it occurs as fine flakes in igneous and
metamorphic rocks.
Calcite 3 Usually whitish, sometimes good crystals,
CaCO3 three directions of cleavage (not at 90°),
reacts with weak hydrochloric acid
(producing bubbles of carbon dioxide).
Chlorite 2.5 Usually too fine to see, but gives a green
(Mg,Al,Fe)6 (Si,Al)4O10(OH)8 sheen to fine-grained metamorphic rocks.
Epidote 7 Usually green elongated crystals, in
Ca2(Al,Fe)Al2(SiO4) metamorphic rocks.
(Si2O7) (O,OH)2
Feldspar 6 Usually white or pinkish crystals, two
KAlSi3O8 to NaAlSi3O8 poorly developed cleavages roughly at 90°.
(alkali feldspar) or
NaAlSi3O8 to CaAl2Si2O8
(plagioclase feldspar)
Fluorite (also known as 4 Colourless to pale purple or orange,
fluorspar) normally occurs as cubes in veins
CaF2 associated with ore minerals, four
excellent cleavage directions.
Garnet 7 Red, brown or green, no cleavage, most
(Ca,Fe,Mg)3Al2(SiO4)3 commonly as well-formed crystals in
metamorphic rock.
Gypsum 2 Usually whitish, can occur as good
CaSO4.2H2O crystals, or as finely crystalline alabaster.
Illite n/a Illite is the most common sort of clay in
KyAl4Si8–yAlyO20(OH)4 mudrocks. In the formula given, the value
of y is usually in the range 1 to 1.5.
Kaolinite n/a A clay mineral with a comparatively
Al2Si2O5(OH)4 simple chemical formula. This is the main
component of china clay.
Kyanite 4–7 Light blue, bladed crystals that may form
Al2SiO5 rosettes. Occurs only in metamorphic
rocks. Polymorphs of the same
composition are andalusite and sillimanite.
Montmorillonite n/a A clay mineral that can absorb several
(Na,Ca)0.3(Al,Mg,Fe)4Si4O10 times its own weight of water, as implied
(OH)2.nH2O by nH2O in its formula.
Muscovite (mica) 2 Usually colourless, one excellent cleavage,
KAl2(AlSi3O10)(OH)2 so that it occurs as fine flakes in igneous,
metamorphic and sedimentary rocks.
Olivine 6.5 Pale olive to very dark green, no cleavage.
(Fe,Mg)2SiO4
Pyroxene 6 Common dark mineral in basic igneous
(Ca,Mg,Fe)2Si2O6 rocks, two cleavages at about 90°.
(augite, diopside)
(Mg,Fe)2Si2O6 (enstatite)
(Continued)

Quartz 7 Colourless or white, but may be tinted by
SiO2 impurities, no cleavage but sometimes
shows nice crystal faces.
Serpentine 3–4 Has various polymorphs, including
Mg3Si2O5(OH)4 crysotile, which is fibrous and forms
asbestos. A product of hydrous alteration
of olivine, pyroxene and amphibole.
Sillimanite 7 Long needle-shaped brown or grey
Al2SiO5 crystals. Occurs only in metamorphic
rocks. Polymorphs of the same
composition are andalusite and kyanite.
Talc 1 When occurring in bulk, talc is described as
Mg3Si4O10(OH)2 ‘soapstone’. A product of hydrous alteration
of olivine, pyroxene and amphibole.
Zircon 7.5 An accessory (rare) mineral in igneous
ZrSiO4 rocks, which owing to its hardness is very
robust, survives transport well and can
be found in sedimentary and even some
metamorphic rocks.
Table c Common ore minerals (all these have a metallic lustre)

Cassiterite 6 Black tetragonal crystals, more often
SnO2 (tin oxide) found in water-washed sediments than
within hard rock.
Chalcopyrite 3.5 Yellow, less brassy than pyrite,
CuFeS2 tendency to tarnish.
(copper-iron sulfide)
Galena 2.5 Lead-grey in colour, three good cleavage
PbS (lead sulfide) directions, may occur as cubes.
Hematite 6 Black to dark red in colour, no cleavage,
Fe2O3 (iron oxide) may occur in kidney-like shapes.
Illmenite 5–5.5 Black, tabular or massive crystals.
FeTiO3 (iron-titanium oxide)
Magnetite 5.5–6 Greyish black. Common opaque mineral
Fe3O4 (iron oxide) in igneous rocks.
Molybdenite 1 Bluish-grey. Platy or massive crystals.
MoS2
(molybdenum sulfide)
Pyrite 6 Brassy yellow in colour, often occurs as
FeS2 (iron sulfide) cubes, no cleavage. (This is ‘fool’s gold’.)
Sphalerite 4 Brown, black or red, tetrahedral
(zinc blende) crystals.
ZnS
346

Table d Notable precious and semi-precious stones
Name Formula Notes
Amber – Arguably not a mineral. This is fossilized resin.
Amethyst SiO2 Violet variety of quartz, coloured by impurities,
notably iron.
Diamond C (pure carbon) A polymorph that forms under high pressure, this
is the hardest known mineral.
Emerald Be3Al2Si6O18 A form of beryl, coloured green by minute traces
of chromium.
Garnet (Fe,Mg)3Al2(SiO4)3 Magnesium-rich varieties are deep red and iron-
rich varieties are usually violet.
Jet – A lustrous variety of coalified fossilized wood
(lignite).
Lazurite (Na,Ca)4–8(AlSiO4)6 Lapis lazuli is a precious stone, noted for its
(SO4,S)1–2 attractive blue colour, consisting of a fine mixture
of lazurite with calcite, pyroxene and other
minerals.
Opal SiO2.nH2O Amorphous, non-crystalline, hydrated silica.
Ruby Al2O3 A form of corundum, coloured red by traces of
chromium.
Sapphire Al2O3 A form of corundum, coloured blue by traces of
titanium and iron.
Topaz Al2SiO4(F,OH)2 May be yellow, blue, or even red.

Appendix 2: Rock names
No two rock bodies (nor, for that matter, two specimens of
rock collected from the same body) have exactly the same
composition. Even closely similar specimens will differ minutely
in their chemical ingredients, or in the size or shape of the
crystals or grains that make up the rock. However, a basalt and
a granite are so clearly different in appearance and origin that it
is both useful and sensible to have distinct names for them. This
appendix lists the more common rock names and the criteria
used to distinguish them.
Igneous rocks
These are rocks that formed from a molten state. Most igneous
rocks are composed of an interlocking collection of crystals
that grew as the melt (magma) cooled. The crystals usually
have a random orientation and, generally speaking, the slower
the magma cooled, the larger the crystals. Consequently,
most igneous rocks can be identified as igneous simply on the
basis of their textures. Coarse-grained igneous rocks (average
crystal size greater than 2 mm across) can be classified with the
unaided eye; medium-grained igneous rocks (average crystal
size between 0.25 mm and 2 mm) usually need a hand lens; and
fine-grained igneous rocks (average crystal size less than
0.25 mm) require a microscope to be certain of their nature
and composition.
There are several notable exceptions to these general remarks:
33 Flow or settling during crystallization can cause crystals to
become oriented in layers.
33 Some volcanic igneous rocks may have cooled so rapidly that
no crystals had time to grow, with the result that the rock has
the appearance, and sub-microscopic structure of glass.

33 Volcanic rocks that were erupted explosively or generated by
collapse of a lava dome (pyroclastic rocks) are made of either
glassy or crystalline fragments.
33 The very coarse grain size of rocks described as pegmatites
(see Chapter 6) bears no relation to the rate of cooling.
Whatever their texture, igneous rocks are formally classified on
the basis of their chemical composition, notably by the proportion
of silica (SiO2) in their total chemical make-up. Chemical analyses
cannot usually be made in the field, but the overall chemistry of a
rock is reflected in the kinds of minerals it contains so, fortunately,
it is possible to identify rock types in this way. The more silica a
rock has, the more acidic it is said to be, and igneous rocks are
divided into acidic (otherwise known as ‘felsic’), intermediate,
basic (or ‘mafic’) and ultrabasic (or ‘ultramafic’) types.
Acidic (felsic) igneous rocks

These consist of more than 66 per cent silica (SiO2). This does
not mean that they are more than 66 per cent quartz (the mineral
with SiO2 as its formula), because all silicate minerals contain
some SiO2 in their formulae. The most abundant minerals in an
acidic igneous rock are alkali feldspar, quartz and plagioclase
feldspar; with lesser amounts of muscovite, biotite and sometimes
amphibole. If the rock is coarse-grained it is called granite, if
it is medium-grained it is called microgranite, and if it is fine-
grained it is called rhyolite. The term ‘granitic’ is used loosely to
describe acidic igneous rocks in general; for example, a rhyolite
is a fine-grained granitic igneous rock. A pyroclastic rock of this
composition could be referred to as a ‘rhyolitic tuff’.
Intermediate igneous rocks

These contain between 52 per cent and 66 per cent silica. The
most abundant mineral is plagioclase feldspar, with lesser
amounts of biotite and amphibole, and sometimes alkali
feldspar, quartz or pyroxene. A coarse-grained intermediate
igneous rock is called diorite, its medium-grained relative is
called microdiorite, and the fine-grained equivalent is called
andesite. The term ‘andesitic’ is used to describe these rocks
in general, so that a diorite can be said to be a coarse-grained
andesitic rock.
350

Basic (mafic) igneous rocks
These contain between 45 per cent and 52 per cent silica (but
the mineral quartz is absent). The most abundant minerals
are plagioclase feldspar and pyroxene, and sometimes olivine.
A coarse-grained basic igneous rock is called a gabbro, its
medium-grained equivalent is called a dolerite (British usage)
or diabase (American usage), and the fine-grained variety is
called basalt. The term ‘basaltic’ is used to describe these rocks
in general.
Ultrabasic (ultramafic) igneous rocks

These contain less than 45 per cent silica (usually not reaching
less than about 38 per cent). The most abundant mineral is
olivine, with pyroxene and a lesser amount of plagioclase
feldspar. Only the coarse-grained variety is common, and is
named peridotite. Lava of ultramafic composition was common
in the Archean (and occurs rarely in the Proterozoic and
Phanerozoic), and is known as komatiite.
Other distinctions
The names highlighted above are just the most common
names, and there is a bewildering variety of other names still in
use – recognizing that silica content is not the only significant
variable in rock composition and reflecting minor differences
in composition (such as iron to magnesium ratio). Some names
are inherited from the days when each rock type was named
after the place where it was found. For example, one variety of
basalt is still referred to as ‘hawaiite’, after Hawaii, and others
(benmoreite, mugearite) are named after type localities in the
west of Scotland. Komatiite is named after its type locality on
the Komati river in South Africa.
The relative abundance of the ‘alkali’ metals sodium and
potassium (which tend to go into alkali feldspar) compared
with calcium (which goes into plagioclase feldspar) is often used
to distinguish varieties of igneous rock that have similar silica
contents. Names such as syenite, monzonite and dacite reflect
a decreasing abundance of alkalis in alkali-rich rocks that are
broadly of granitic to andesitic composition.

Metamorphic rocks
These are rocks that form when heat or pressure, or both,
cause a pre-existing rock to recrystallize. Very often, the effect
of pressure is to cause crystals to grow in, or to re-orient
themselves into, layers. When this happens, the metamorphic
rock has a layered texture, which is described as ‘foliation’. In
fine-grained metamorphic rocks the crystals cannot be seen by
the unaided eye, but the foliation is manifested by the tendency
of the rock to fracture along flat or wavy ‘cleavage’ surfaces
parallel to the foliation.
Metamorphic rocks are most easily classified simply on textural
grounds, using the same fine-, medium- and coarse-grain size
division as in igneous rocks, as shown in the following table.
Table X Classification of metamorphic rocks by texture
Textural type Nature of foliation Grain size
Slate Very closely spaced, almost flat fine
Phyllite Wavy or crenulated, sheen caused by minute fine
flakes of mica or chlorite on foliation surfaces
Schist Undulating planes, characteristically with medium to
abundant mica coarse
Gneiss More widely spaced than in schist, usually coarse
alternating layers of dark and light minerals
Migmatite Poor foliation, shows signs of having begun to coarse
melt
It is often convenient to qualify the simple terms listed above in

recognition of prominent minerals within the rock. For example,
a schist with abundant mica and distinct crystals of garnet may
be described as a garnet mica schist.
Not all metamorphic rocks have a foliated texture. For example,
a rock metamorphosed by the heat of an igneous intrusion
(thermal or contact metamorphism) usually develops a spotty
texture and is described as a hornfels. Furthermore, rocks
that, prior to metamorphism, consisted of just one abundant
mineral are unlikely to develop a new mineral assemblage
during metamorphism. The original texture may be destroyed
by adjustment of crystal boundaries, but there will not be
a strong oriented fabric. A quartz-rich sandstone that has
352

been metamorphosed consists of interlocking quartz crystals
and is described as quartzite, and a limestone that has been
metamorphosed consists of interlocking calcite crystals and is
described as marble.
Chapter 7 introduced the concept of metamorphic facies,
corresponding to the temperature and pressure under which
metamorphism took place. The facies cannot usually be
determined by simple inspection; however, when the facies is
known it is normal to write of (say) an amphibolite facies gneiss
or a zeolite facies slate.
Sedimentary rocks
These form by deposition of detrital grains or by precipitation
from solution. Detrital (or ‘fragmental’ or ‘clastic’) sedimentary
rocks are most simply distinguished from igneous and
metamorphic rocks on textural grounds, because the grains of
which they are composed are non-interlocking fragments (clasts)
rather than interlocking crystals. Sedimentary rocks formed by
precipitation do have interlocking crystalline textures, but are
made of quite different minerals from igneous rocks, notably
calcite (in limestone) or evaporite minerals such as gypsum.
Clastic sediments are classified according to grain size in the
scheme shown in the following table, which shows names for
the rock type and the different sizes of clasts (note: ‘breccia’
comes from the Italian and is pronounced ‘brechia’).
Table Y Classification of sedimentary rocks by grain size

Grain size Term for clasts Rock type
> 256 mm boulders conglomerate (rounded clasts) or breccia
(angular clasts)
64–256 mm cobbles conglomerate (rounded clasts) or breccia
(angular clasts)
2–64 mm gravel (pebbles conglomerate (rounded clasts) or breccia
4–64 mm) (angular clasts)
0.0625–2 mm sand sandstone
0.002–0.0625 mm silt siltstone or mudrock (referred to as shale if
it splits easily into layers)
0.002 mm clay claystone or mudrock (referred to as shale
if it splits easily into layers)

Rocks of these compositions may sometimes also be described
according to their supposed mode of origin (which is a risky
business, because the geologist may get this wrong!). For
example, a breccia or conglomerate with a muddy matrix that
is thought to have been deposited by a glacier would be called
a tillite.
Sandstones are further classified according to the mineralogy of
their grains. A sandstone that is over 90 per cent quartz grains
is described as a quartz sandstone, whereas one containing
a significant amount of feldspar, clay minerals or calcite is
called an arkose, muddy sandstone or calcarous sandstone,
respectively. One that has a significant proportion of fragments
composed of fine-grained rocks is described as a greywacke
(pronounced as three syllables).
Just as a hardened rock made of sand is referred to as a
sandstone, so a rock made from silt is referred to as a siltstone
and one made from clay is a claystone. See Table Y for these and
other terms.
A rock that consists of over 90 per cent calcite is referred to as
a limestone. A limestone with an admixture of quartz grains or
clay minerals would be called a sandy limestone or a muddy
limestone. Limestones are further distinguished according
to whether the calcite occurs as fragments of shells (skeletal
limestones), as ooids (oolitic limestone) or as faecal pellets
(peloidal limestone), and whether the grains are cemented
together by crystalline calcite (sparite) or held together by fine-
grained calcite mud (micrite). If the calcite has been replaced
by the mineral dolomite, then strictly the rock should be called
a dolostone, but is more commonly referred to as dolomite (a
rare instance in which the same word is used for a rock as for
a mineral).
Another common sedimentary rock name is chert, which
describes a rock formed of microscopic silica crystals (which
can be either a chemical precipitate, as in the case of flint, or
shells of microscopic organisms such as radiolarians).
354

Appendix 3:
Geological time
Geological time is divided according to the scheme shown in
the following diagram. Note that this is not drawn to a uniform
scale. The numbers refer to the age in millions of years of the
boundary between each division, as determined radiometrically
(i.e. by measuring the products of radioactive decay). These
dates are still subject to revision.
The largest unit of geological time is an eon. The scheme shown
here names four eons. The oldest known rocks date back to the
beginning of the Archean, and the Hadean is the time before for
which we have only indirect evidence. Some authorities do not
regard the Hadean as a separate eon, and include it within the
Archean.
The Phanerozoic Eon is divided into three eras: Paleozoic,
Mesozoic and Cenozoic (which mean ‘old-life’, ‘middle-life’ and
‘modern-life’ respectively). In many kinds of sedimentary rocks
of Phanerozoic age, fossils are sufficiently abundant and show
sufficiently clear evolutionary change to enable geological time
to be split into a succession of progressively smaller divisions.
Each era is divided into a number of periods. Each period is
divided into a number of epochs (shown here only for the
Paleogene and Neogene Periods), and further subdivided into
ages (not shown here).
The Holocene Epoch covers the past 10 000 years, and is
sometimes called the Recent. The time span covered by the
Cenozoic was formerly split into two eras: the Tertiary Era
(from the start of the Paleocene to the end of the Pliocene)
and the Quaternary Era (Pleistocene and Holocene), and both
these terms are still in widespread use. The boundary between
the Cretaceous and the Tertiary, which is marked by the most
recent major episode of mass extinction, is often referred
to as the K/T boundary (K being the usual abbreviation for
Cretaceous).
Appendix 3: Geological time 355

A handy website where you can see an updated timescale in
detail is http://www.geosociety.org/science/timescale/
(Epochs)
Holocene 0.01
(Periods) Pleistocene
2.6
Pliocene
Cenozoic
Neogene 5
23
Era
Paleogene Miocene
65 23
Cretaceous Oligocene
Mesozoic Era
Phanerozoic Eon
145 34
Jurassic Eocene
199
Triassic 56
251 Paleocene
Permian 65
299
Carbon- Pennsylvanian
iferous 318
Mississippian
Paleozoic Era
359
Devonian
416
Silurian
443
Ordovician
488
Cambrian
541
Precambrian
Proterozoic Eon
2500
Archean Eon
4000
Hadean Eon
4560 million years
Geological timescales
356

Glossary
This glossary contains those terms that I think the newcomer to
geology will have most trouble with. I have indicated the chapter(s)
in which each term makes an important appearance. Names of
several minerals and rock types not listed here may be found in
Appendices 1 and 2. There is a stratigraphic timescale in Appendix 3.
acidic Descriptive term for igneous rocks containing more

than 66 per cent silica (SiO2). Alternatively referred to as
‘felsic’ (Ch 5).
andesite A fine-grained igneous rock type richer in silica than
basalt, commonly erupted at volcanoes above subduction
zones (Ch 5).
anticline An arch-like fold, in which beds on either side dip
away from each other (Ch 10).
ash Fine, fragmentary rock, produced by a pyroclastic
(explosive) eruption (Ch 5).
asthenosphere The weak zone within the mantle (but below its
top), which is overlain by the lithosphere (Ch 2).
banded iron formation (bif) A sedimentary deposit consisting
of alternate layers of iron oxide and silica, of the kind con-
stituting most of the global iron reserves. Most are between
3800 and 1800 million years old (Ch 11).
basalt An igneous rock type with the same composition as
oceanic crust. Strictly speaking, a basalt must be fine-grained;
coarser-grained rocks of the same composition are described
as basaltic or basic (Ch 2, 4).
basic Descriptive term for igneous rocks containing more
than 45–52 per cent silica (SiO2). Alternatively referred to as
‘mafic’ (Ch 4, 6).
bed The commonest and simplest term for a layer of
sedimentary rock. Also used to refer to the floor of the sea or
the surface over which wind or water is flowing (Ch 9).
Glossary 357

bedform A ripple, dune or other similar feature produced by
flowing water or wind (Ch 9).
bed-load Particles dragged, rolled or bounced along during
transport by wind or water (Ch 8).
biostratigraphy Determining the (relative) ages of
sedimentary rocks by means of their fossil content (Ch 12).
breccia A rock composed of coarse angular fragments. Can
be a sediment or a product of intense local fracturing during
faulting (Ch 10).
brittle Deformation of rock by fracturing. A characteristic of
rapid deformation at relatively low temperatures and
pressures (Ch 10). See ductile.
caldera A volcanic crater more than 1 km across, formed by
subsidence of its floor during an eruption (Ch 5).
carbon cycle A term describing the complex natural cycle in
which carbon is transferred between the atmosphere (as car-
bon dioxide), the oceans (in solution or in the shells of
living organisms) and the Earth’s interior (as carbonate
rocks, as hydrocarbons or as carbon dioxide emitted by
volcanoes) (Ch 1).
cement The precipitated mineral holding grains together in a
sedimentary rock, most commonly calcite or silica (Ch 9).
chilled margin The very fine-grained/glassy edge of a
shallow igneous intrusion, which cooled rapidly against the
cold rock into which it was injected (Ch 6).
clast A particle or lump (of any size) that is transported and
deposited in a sediment (Ch 9).
clastic rock A sedimentary rock composed of clasts (Ch 9).
cleavage 1) Regular planes of weakness in a crystal, which
are a consequence of its atomic structure (Chs 6, 8). 2) The
same term describes closely spaced planes of weakness in
a deformed or metamorphosed rock, which are caused by
alignments of platy minerals (Ch 7, 10).
358

conservative plate boundary A site where two tectonic plates
are sliding past each other. In the oceans, the fault across
which the motion occurs is called a transform
fault (Ch 4).
constructive plate boundary A site where two tectonic plates
are moving apart, and each is being added to by the forma-
tion of new oceanic lithosphere (Ch 4).
continental drift The relative movement of continents, which
is a consequence of plate tectonics (Ch 4).
core The Earth’s high-density iron-rich interior (3470 km in
radius). It consists of a liquid outer core and a solid inner
core (Ch 2).
craton A stable area of continental crust, also known as a
shield, that has remained undeformed since the Archean
Eon (Ch 13).
cross-bedding Sedimentary bedforms seen in cross-section,
where the dip of the incrementally added layers was
originally steeper than the dip of the bed as a whole (Ch 9).
crust The compositionally distinct, slightly lower-density rocky
layer overlying the Earth’s mantle (Ch 2).
crystal A solid substance formed by a repeating, three-dimen-
sional pattern of atoms of various elements. The symmetry
of this pattern is reflected in the characteristic shape of the
crystal. Most minerals occur in crystalline forms (Ch 1, 6).
decompression melting Partial melting stimulated in the
asthenosphere when mantle rises upwards (in a plume or
below a constructive plate boundary). Melting is caused by
the drop in pressure rather than the addition of heat (Ch 4).
destructive plate boundary A site where one tectonic plate is
destroyed by subduction below another (Ch 4).
diagenesis The chemical and physical processes that turn an
unconsolidated sediment into a hard sedimentary
rock (Ch 9).
Glossary 359

dip The angle between the horizontal plane and a geological
surface, such as a bedding plane (Ch 15).
ductile Deformation of rock by bending or squeezing, but
without fracturing. A characteristic of slow deformation at
relatively high temperatures and pressures (Ch 10).
See brittle.
dyke A curtain of igneous rock, intruded vertically (Ch 6).
earthquake Shaking of the ground, caused by sudden
movement within the Earth (Ch 2, 3).
effusive A term describing an eruption that occurs non-
explosively, such as flow of lava from a vent (Ch 3).
evaporite A mineral or deposit of minerals formed by
evaporation of seawater (Ch 9).
evolution A theory, famously propounded by Charles Darwin
in the nineteenth century and subsequently backed up by a
great weight of genetic and fossil evidence. It says that
pressure to survive in a competitive environment increases
the chances of individuals with successful adaptations
surviving to breed, and hence to pass on their adaptations to
their progeny. Life has diversified from the first primitive
microbes to its present state, by the slow accumulation
of random mutations that gave competitive advantage in
particular environments (Ch 12).
exposure A place where bedrock is actually visible at the
surface. Sometimes casually referred to as an outcrop,
although, strictly, that term has a different meaning (Ch 15).
facies The temperature and pressure conditions under which
metamorphic rocks were metamorphosed, or the suite of
conditions responsible for deposition of a particular kind of
sediment (Ch 7).
fault A fracture in the lithosphere across which motion takes
place (Ch 3).
fieldwork Studying rocks and geological processes ‘in the field’,
meaning ‘where they occur’ (Ch 15).
360

flood basalt Extensive, thick pile of flat-lying basalt lava flows
(typically more than a million cubic kilometres in
volume), thought to be generated above the site of a
particularly vigorous mantle plume (Ch 5).
floodplain Flat, low-lying ground beside a river, over which
it floods after very heavy rainfall. A river’s course tends to
migrate to and fro across its floodplain over tens of
thousands of years (Ch 9).
focus The break-point where earthquake motion begins and
where most of the energy is released (Ch 3).
fold A structure formed when layers of rock become bent
(Ch 10).
foliation An alignment of minerals in layers, most commonly
caused by regional metamorphism (Ch 7).
footwall The volume of rock below an inclined fault surface,
which is not deformed when the overlying rock is displaced
by fault motion (Ch 10).
fossil The remains of a past organism. A fossil is either a body
fossil (being a fossilized body part) or a trace fossil (such as a
footprint or burrow) (Ch 12).
fossil fuels Coal, oil and natural gas, which are being used up
at rates much faster than the rates at which they are being
created by geological processes (Ch 11).
fracking (hydraulic fracturing) A technique for boosting or
enabling the extraction of hydrocarbons by artificially in-
creasing the permeability of the host rock (Ch 11).
fractional crystallization The process whereby the first crystals
to form in a magma as it cools have a composition different
from that of the magma as a whole (Ch 6).
geothermal gradient The rate at which temperature increases
with depth inside the Earth (Ch 2).
glaciations Cold periods (tens or hundreds of thousands of
years) during an ice age when ice sheets are extensive, and
Glossary 361

punctuated by interglacial periods (Ch 8). The term is
sometimes wrongly equated with ‘ice age’.
glass Volcanic glass is magma that cooled so quickly that crys-
tals had no time to form (Ch 6).
gneiss High-grade metamorphic rock characterized by
alternating dark and pale layers (Ch 7).
granite A coarse-grained intrusive igneous rock, richer in silica
than andesite. Rocks of this composition are described as
granitic or acidic or felsic (Ch 5, 6).
greenhouse effect The warming of a planet’s atmosphere by
means of absorption of infrared radiation by certain
atmospheric gases, notably carbon dioxide (Ch 9, 11, 13).
hanging wall The volume of rock above an inclined fault
surface, which may be deformed when the rock is displaced
by fault motion (Ch 10).
hornfels A splintery, spotty rock produced by contact
metamorphism (Ch 7).
hot spot A place where a plume rising from deep within the
mantle hits the base of the lithosphere, giving rise to large
volumes of basaltic magma (Ch 5).
hydration melting Partial melting simulated by the
addition of water, notably where water escapes from a
subducting plate into the mantle wedge above (Ch 5).
hydrocarbons Oil and natural gas, consisting of chains of
carbon atoms bonded to hydrogen, or in the case of
methane single carbon atoms bonded to hydrogen (Ch 11).
hydrological cycle The complex natural cycle in which water is
continuously transferred between the atmosphere, the oceans
and the Earth’s interior (Ch 1).
hydrolysis Chemical breakdown of a mineral to clay
particles by the action of (usually slightly acidic) water
during weathering (Ch 8).
ice age A period (of the order of ten million years long) when
the climate alternates between colder and warmer, so that
362

there is a succession of glaciations (when polar ice sheets and
mountain glaciers are particularly extensive) and interglacials
(when, as at present, the ice has retreated) (Ch 8, 13).
igneous Referring to a rock or mineral formed by
solidification from a molten state (Ch 1).
ignimbrite An extensive pyroclastic deposit formed from
shards of acidic volcanic ash and pumice fragments (usually
welded together). Typically emplaced from a pyroclastic flow
as a result of the collapse of an eruption column during the
eruption of a supervolcano, or a smaller eruption of similar
type (Ch 5).
interglacials Interludes of relatively warm climate between
successive glaciations of an ice age (Ch 8).
intermediate Descriptive term for igneous rocks containing
52–66 per cent silica (SiO2), such as an andesite (Ch 5).
intrusive Referring to an igneous rock emplaced below the
surface (Ch 6).
island arc A usually arcuate line of islands built by v olcanism
above a subduction zone (Ch 4).
isostasy Another way of saying ‘buoyant equilibrium’.
It is normally found that mountain regions have deep
roots. In contrast, the Moho (the base of the crust) is
farther from the centre of the Earth below low-lying
regions. Seen from a geological perspective, the crust
thus ‘floats’ in equilibrium upon the mantle, although on
human timescales the mantle behaves as a solid rather
than a liquid (Ch 2).
joint A smooth fracture through a body of rock, caused by
release of pressure or cooling. In contrast to a fault, there is
no movement between opposite sides of a joint (Ch 8).
k-t boundary The boundary between the Cretaceous (K) and
the younger sediments of the Cenozoic or Tertiary (T) Age.
It coincides with an important mass extinction event, which
was probably caused by the impact of an asteroid or comet
(Ch 12).
Glossary 363

lava Molten rock at the surface; also used to describe
rock formed in this way, after it has solidified (Ch 5).
limestone Rock made mostly of calcium carbonate
(Ch 7, 8, 9).
lineation A linear fabric in a rock, such as when elongated
metamorphic minerals are aligned (Ch 7).
lithosphere The rigid outer shell of the Earth, consisting of the
crust and the uppermost mantle (Ch 2).
magma Molten rock; usually used to refer to molten rock
at depth, whereas molten rock at the surface is called lava.
Magma often includes dispersed crystals, so the term ‘melt’
is sometimes used when referring specifically to the liquid
component (Ch 4).
mantle All the silicate (rocky) material surrounding the core of
the Earth, with the exception of the thin silicate crust (Ch 2).
mass extinction A relatively brief interlude (tens or hundreds
of thousands of years) during which a high proportion of
species of many kinds become extinct. The K-T boundary
mass extinction was probably a result of an asteroid impact
(Ch 12).
metamorphic Referring to a rock or mineral formed by
recrystallization (without melting but in response to heat
and/or pressure) of a pre-existing rock or mineral. This
process is called metamorphism, and can usually be described
as either thermal or regional (Ch 1, 7).
metamorphism Recrystallization of the fabric of a rock without
melting, under conditions of high temperature and/or pressure.
Thermal or contact metamorphism is a result of proximity to
an igneous intrusion. Regional metamorphism is a result of
burial to great depth (Ch 7). Dynamic m
etamorphism can
occur during ductile fault movement (Ch 10).
mineral A naturally occurring crystalline substance with a
well-defined chemical composition. Most rock types consist
of several different minerals (Ch 6). (See Appendix 1 for a
fuller discussion.)
364

Moho The boundary between the crust and the mantle (Ch 2).
non-renewable resources Fossil fuels, ores and bulk materials,
which are being used up at rates much faster than the rates at
which they are being created by geological processes (Ch 11).
normal fault Steep fault with vertical displacement, where
the rock on the down-dip side of the fault has moved
downwards (Ch 10).
ophiolite A slice of oceanic crust and upper mantle that has
been thrust over the edge of a continent (Ch 4).
ore A metal-bearing mineral from which the metal can be
extracted at a profit (Ch 11).
orogeny A mountain-building episode resulting from a
collision between two continents, or between a continent
and an island arc (Ch 13).
outcrop On a geological map, the area where a particular unit
of rock would be visible if all superficial cover were removed
(Ch 15). See exposure.
partial melting This is what happens when any rock composed
of a variety of minerals begins to melt. The first melt to form
has a different chemical composition (richer in silica) from
the solid (Ch 4).
peridotite A silicate rock type poor in silica, having the same
composition as the Earth’s mantle (Ch 2, 4, 6).
permeability The ease with which liquid or gas can flow
through a porous rock, which is dependent on how well
connected the pore-spaces are (Ch 11). See porosity.
phenocryst A larger crystal surrounded by smaller crystals in
an igneous rock (Ch 6).
pillow lava The characteristic morphology shown by basaltic
lava when erupted underwater (Ch 5).
placer deposit An ore concentrated in a sediment because the
ore forming the densest grains becomes concentrated by
sedimentary processes (Ch 11).
Glossary 365

plate tectonics A description of how the rigid plates into
which the Earth’s lithosphere is divided move about
(Ch 4).
polymorphs Minerals that have identical chemical formulae
but different crystalline structures. Examples are kyanite,
sillimanite and andalusite (Al2SiO5) and calcite and aragonite
(CaCO3) (Ch 7).
porosity The proportion of a rock body that is occupied by
liquid- or gas-filled spaces (Ch 9, 11). See permeability.
prograding Description of a shoreline that is building out
seawards (Ch 9).
pyroclastic Term used to describe an eruption (or the resulting
rock) in which fragments are produced by explosive volcanic
activity or some other process (such as collapse of a lava
dome) that generates hot fragments (Ch 5).
pyroclastic flow A denser-than-air mixture of ash and larger
bouncing rock fragments mixed with hot air that flows
turbulently downhill (Ch 5).
radiogenic heating Heating of a planet’s interior through
the energy released by decay of the nuclei of unstable,
radioactive atoms (Ch 2).
radiometric dating A technique in which the relative
proportions of products of radioactive decay and their parent
isotopes are used to determine the age of a rock or mineral
(Ch 2).
regional metamorphism Metamorphism as a result of p ressure
and heat at depth within the crust, which can be seen in for-
merly deep rock that has been exposed at the
surface as a result of uplift and erosion (Ch 7).
reserves Those resources that can be extracted profitably and
legally under existing conditions (Ch 11). See resources.
reservoir rock A permeable body of rock into which
hydrocarbons or water have migrated, and from which they
can be extracted through a well (Ch 11). See source rock.
366

resources The estimated total amount of a commodity (fossil
fuel, ore or raw material) available in the world (Ch 11).
See reserves.
reverse fault Steep fault with vertical displacement, where the
rock on the down-dip side of the fault has moved upwards
(Ch 10).
rock cycle A complex network of processes including the erosion,
transport, deposition, burial, heating, deformation, melting,
cooling and exhumation of rock or rock fragments (Ch 1).
schist Metamorphic rock characterized by wavy layers of
minerals, whose crystals can be distinguished by eye (Ch 7).
sedimentary Referring to a rock formed from a deposit of
detrital clasts, or by precipitation from solution in water.
Sedimentary minerals are those that form in, or are found in,
sedimentary rocks (Ch 1).
seismic waves Vibrations, from earthquakes or artificial
explosions, that travel through the Earth. P-waves are
compression-dilation waves, and S-waves are shearing waves
(Ch 2).
seismometer Device for recording the vibrations caused by
earthquakes or movement of magma (Ch 2, 3, 5).
shale A mudrock or siltstone that breaks naturally into finely
spaced layers parallel to its bedding (Ch 9, 11, Appendix 2).
shield Another word for craton (Ch 13).
shield volcano A gently sloping basaltic volcano, so named
because of its cross-sectional profile (Ch 5).
silica Any compound with the formula SiO2. The term is also
used to refer to the percentage of a rock (by weight) that
can be expressed as SiO2 irrespective of whether this occurs
as pure silica or combined with other elements in silicate
minerals (Ch 2).
silicate Either a rock type or a mineral rich in silicon and oxy-
gen. The Earth’s mantle and most of its crust are formed of
silicates, as are virtually all igneous rocks (Ch 2, 6).
Glossary 367

sill A horizontal sheet-like igneous intrusion (Ch 6).
slate Low-grade, fine-grained, metamorphic rock, prone to
fracturing on planes parallel to its imposed foliation (Ch 7).
slickensides Scratches or grooves on a fault surface, formed as
the rock masses on either side ground past each other
(Ch 10).
source rock A body of rock within which hydrocarbons form
(Ch 11). See reservoir rock.
strain The amount of deformation that has occurred to a
volume of rock as a result of the stress to which it has been
subjected (Ch 10). Strain may be accomplished by deforma-
tion that is either brittle or ductile.
strata Layers of rock, irrespective of whether they are
sedimentary. The singular is ‘stratum’ (Ch 10).
stress The forces acting to deform a volume of rock (Ch 10).
strike The direction, conventionally measured clockwise from
north, of a horizontal line drawn on a geological surface,
such as a bedding plane. Strike and dip together define the
three-dimensional orientation of the surface (Ch 15).
subduction The descending of one tectonic plate at an an-
gle below another, which happens at a destructive plate
boundary (Ch 4).
suspension-load Particles suspended for long periods without
touching the bottom during transport by wind or water
(Ch 8).
syncline A U-shaped fold, in which beds on either side dip
towards each other (Ch 10).
terrane A fault-bounded region of the crust that has a different
geological history to that of the adjacent crust, because they
were formerly not neighbours. (The alternative spelling, ‘ter-
rain’, refers to landscape in general.) (Ch 13)
thermal metamorphism Metamorphism of pre-existing rock
near the margin of an igneous intrusion, and driven by
368

heat leaking out of it. It is also known as ‘contact
metamorphism’ (Ch 7).
thrust A low-angle fault which has pushed older rocks over
younger rocks (Ch 10).
transcurrent fault A steep fault across which relative
displacement is sideways. It is alternatively known as a
strike-slip fault (Ch 10).
transform fault A fault in oceanic lithosphere offsetting two
lengths of a constructive plate boundary, which has sideways
slip across it (Ch 4).
tsunami A series of water waves triggered by an earthquake,
landslide or volcanic eruption. The waves are not high while
they travel through deep water, but they become much higher
and extremely dangerous as they approach the shore
(Ch 3, 5).
ultrabasic A type of rock that has even less silica than a basalt
(<45 per cent), such as a peridotite. Alternatively referred to
as ultramafic (Ch 6).
unconformity A boundary between underlying older rocks
and an overlying set of younger strata, usually manifested by
a difference in dip and representing a considerable time gap
(Ch 10).
vein A fracture that has become filled by minerals precipitated
from solution, commonly quartz or calcite, and sometimes
including ore minerals (Ch 7, 11). Coarse veins described
as pegmatite may crystallize from the last fraction of melt
remaining as an igneous intrusion cools down (Ch 6).
volcano A landscape feature where magma has erupted
(Ch 2, 5).
weathering Slow decay of rock and its constituent minerals
upon exposure at the Earth’s surface, by either physical or
chemical attack (Ch 7, 8).
Glossary 369

Answers
Chapter 2
1 d
2 a
3 a
4 c
5 b
6 c – The outer core is liquid. A liquid lacks the rigidity
necessary for the propagation of S-waves, which are
shearing waves.
7 a – If you answered (b), you were probably thinking of
isotopes.
8 b
Chapter 3
1 c – (a) is wrong, because the Richter scale is a measure of an
earthquake’s strength at its source (at its focus) whereas the
Mercalli scale is a measure of the effects of an earthquake
(the intensity of shaking) at any point on the surface.
2 c
3 a
4 d – Taller buildings sway faster than low-rise buildings.
Chapter 4
1 d
2 b – Continental crust is less dense than ocean crust. This
makes it buoyant and hard to subduct, so that when it does
arrive at a subduction zone (d) the subduction zone jams.
3 a – If you thought (b) was correct, then you have really
not understood plate tectonics. The lack of a mechanism
Answers 371

for continents to plough through the mantle was a major
objection raised to Alfred Wegener’s theory of continental
drift. In fact, continents take their own mantle with them;
a plate consists of lithosphere, and the base of each plate is
lubricated by the weak asthenosphere.
4 b – Figure 4.9 shows speeds in mm per year relative to
Africa, including four speed arrows on the Pacific plate
ranging from 73 mm per year to 106 mm per year, all
pointing northwest. Rates differ from place to place because
the motion is occurring on the surface of a sphere. A rate
of 10 cm per year to the northwest is the only reasonable
choice out of the four.
5 (c) is suggested in the text as the most likely; (a) is
specifically ruled out in the text.
6 C
Chapter 5
1 b – Felsic (c) and acidic (d) are alternative terms for the
same, high-silica, composition. It is the mix of viscous
lava flows and pyroclastic deposits typically erupted from
andesite volcanoes that tends to make them steeper than
other types.
2 b – Free oxygen is not present in magma and does not
escape from volcanoes (this is why the Earth’s early
atmosphere had no oxygen; see Chapter 2).
3 c – a is a very important point that was introduced in
Chapter 4. (b) describes hydration melting. (c) is the wrong
way round: in fact, it is decreasing pressure that can lead to
partial melting. This is called decompression melting. (d) is
the reason why magma tends to rise towards the surface.
4 c – Volcanoes are rare at conservative plate boundaries,
because there is no particular way to generate magma in such
a setting.
5 a – Hawaii sits above a hot spot, remote from the nearest
plate boundary. If you weren’t sure about (c) and (d), you
should have compared Figures 5.1 and 4.9.
372

6 a – (b) is not an eruption product but a mudflow formed
from mixing ash with water. (c) and (d) can form only from
basaltic lava.
7 b
Chapter 6
1 a – A dyke is very steep, whereas a sill is its (near)
horizontal equivalent.
2 b – If you answered (a), you were probably confusing
‘silica’ with quartz. A basalt (Table 2.2) contains about
50 per cent silica, but this will all be bonded with metals
in silicate minerals such as pyroxene and feldspar. There
will be none left over to form quartz. (c) is wrong because
amphibole and the mica family also contain OH (derived
from water). (d) is wrong because magnesium and iron can
substitute for each other in any proportion in olivine.
3 d
4 c – The lower oceanic crust is typically gabbro, and never
dominated by granite. Granite is coarse-grained acidic
(felsic) igneous rock, whereas gabbro is coarse-grained basic
(mafic) igneous rock, so (b) is true.
5 b
Chapter 7
1 c
2 b – This chapter gave the example of kyanite, sillimanite
and andalusite that have the same chemical formula but
different arrangements of atoms.
3 a
4 c – This information can be read off Figure 7.4.
5 d – This information can be read off Figure 7.4.
6 d
Answers 373

Chapter 8
1 b
2 d – Quartz has no cleavage (c) and is chemically inert (a, b).
The wearing away of corners is more gradual in water-
transport than in wind-transport.
3 c – Saltation is bed-load transport by bouncing.
4 a
5 c
Chapter 9
1 c – Clay is a silicate mineral, and clay-rich mud forms by
the settling out of clay particles, so mud formed of clay is
indeed a siliciclastic sediment. (d) is wrong because ooids
are clasts that grow by precipitation, and shells (including
microscopic examples) that become sedimentary clasts
originally grew organically.
2 c
3 a
4 d – (a) is Lyell’s principle of uniformitarianism.
5 b – Chlorite is a mineral that grows during low-grade
metamorphism of mudrocks.
6 b – Aragonite and calcite are polymorphs. Anhydrite and
gypsum (a) are two forms of calcium sulphate, but gypsum
also has two water molecules in its formula.
7 d
8 d
Chapter 10
1 d – The cleavage example (c) counts as an essentially
ductile process because it occurs through rotation and
recrystallization.
2 c – A fold with near-parallel limbs is isoclinal. Figure 10.8c
shows a fold that is both recumbent and isoclinal, but if
374

you rotate the picture by a quarter turn it becomes upright
and isoclinal.
3 b
4 b
5 a – The arrows on Figure 10.6 show the directions of
displacement on the Altyn Tagh and Karakoram Faults.
The right-lateral displacement of the Great Glen Fault is
noted in a case study, and the right-lateral nature of the
San Andreas Fault can be deduced from Figure 4.9 and the
preceding text.
Chapter 11
1 c – Option (d) is a definition of resources.
2 b – Option (d) is a definition of permeability.
3 a, c and d – Crude oil (c) is a mixture of many kinds of
hydrocarbon; butane (a) and methane (d) are gaseous
hydrocarbons. Coal (b) is essentially carbon with
impurities, but the carbon atoms are not bonded to
hydrogen, which is how hydrocarbons are defined.
4 C – This can be answered by reading from Figure 11.2:
170 °C for 10 million years corresponds to the peak of
gas generation. At this temperature, all oil would break
down to gas in less than 3 million years (allowing for the
logarithmic timescale).
5 b – Oil wells are typically 1–5 km deep.
6 b – The anoxic conditions prevent oxidation, and
subsidence and sedimentation allow a thick deposit to
accumulate and eventually become buried. Maturation
of kerogens to the extent described in (c) would result
in carbon dispersed in the source rock as graphite, but it
could not then migrate and become concentrated.
7 b – Strip mining is illustrated in Figure 11.4.
8 d – They can also form from warm fluids expelled from
sediments. Molten rock (options (a), (b) and (c)) is not
involved.
Answers 375

9 d – BIFs are mostly 2.5–1.8 billion years old. Before then,
there was too little oxygen in the atmosphere, and after
then mostly too much.
10 b
Chapter 12
1 d
2 b
3 c – Dwelling burrows and grazing trails are common
examples of trace fossils.
4 c
5 a
6 None of them. (a) is incorrect because the K-T mass
extinction put paid to many kinds of marine life, too.
(b) is incorrect because adaptive radiation is when evolution
is encouraged by the availability of vacant habitats, such as
after a mass extinction. (c) is incorrect because vertebrate
teeth are made of calcium phosphate, not carbonate. (d) is
incorrect because (like all arthropods) the exoskeleton of a
trilobite was made of chitin.
7 d – They are all well known in the fossil record, but
trilobites belong to the arthropod phylum rather than
constituting a phylum on their own.
8 c – The cnidarians (or coelenterata) appeared in the late
Proterozoic, whereas the others appeared in the Cambrian.
9 a – In fact, rapid evolutionary change is required.
10 a – The crater is buried by later sediments, but has been
found straddling the present-day coast of Yucatán in
Mexico.
Chapter 13
1 b – The oceans (c) and (d) were the Hercynian Ocean and
Tethys, respectively.
2 d – (c) can probably explain the Moon’s highland crust.
(b) would produce a chilled crust of the same composition as
376

the magma ocean, and so would not be a true, differentiated,
crust.
3 c
4 a
5 c
6 d – It is an example of the oldest surviving crust, but cannot
be the first crust that formed.
7 d – Figure 13.4 shows that Avalonia was part of Laurentia
at this time.
8 d
9 b
10 b – See Figure 13.5 and the associated discussion.
Chapter 14
1 c
2 d – Most of the mare basalts (c) are more than 3.1 billion
years old.
3 b – Anorthosite is a rock made mostly of a calcium-rich
plagioclase called anorthite.
4 b
5 b
6 a
7 b and d – If you look at Figure 4.10 you can see several
craters on the surface of Eros (so (c) is not correct), though
many appear to be dust-mantled.
8 c
9 c – Titan’s rainfall is methane, not water
10 a
11 c – The largest volcano is Olympus Mons (Figure 14.6).
Answers 377

12 b – The Hubble Space Telescope captured images of
possible plumes on Europa (c) in 2012 (not mentioned in
the text above), but failed to relocate them in 2014 and
they remain unproven.
Chapter 15
1 c – The text in (b) describes an exposure rather than an
outcrop. Many geologists do not bother with this useful
distinction, so if you chose (b) you have plenty of company.
2 d
3 a – You may be putting yourself in danger if you get
close enough to use a hand lens before making an overall
assessment from a safe distance. The other options are all
parts of that assessment.
4 d – The direction in (a) would be looking up-dip, not
down-dip.
5 d – Strike and dip is a way of recording the attitude of a
plane in three dimensions, so it is applicable for (a), (b)
and (c). However, slickensides are linear features, so
the concept of strike and dip is inapplicable (you could,
however, measure the direction of the lines – their ‘trend’ –
and the angle they make to the horizontal – their ‘plunge’).
378

Index
a’a, 95–6 bacteria, 241–2
Acamarachi volcano, Chile, 73 Baltic shield, 273, 277
adaptive radiation, 253 Baltica, 277–81
African Rift, 89 banded iron formations (BIFs),
airfall deposits, 79 230, 232, 275
alabaster, 179 barchans, 161
algae, 178, 182, 242 barrels of oil, 219
alluvial fans, 171 basalt, 58, 84, 94–7, 274–5
aluminium, 20, 108, 225, 231 flood basalts, 89–91, 115
amber, 257, 347 metamorphosed, 129–30, 132
amethyst, 347 other planets, 299–300, 302,
amino acids, 240–1 305, 310
ammonoids, 245, 252, 256, basaltic magma, 73–4, 110–11,
262–3 116–17
amphibole, 109, 130–1, 133, batholiths, 115–16
139, 344 bauxite, 231
amphibolite facies, 131 beaches, 177–8
andalusite, 125–7, 344 bedforms, 157–61, 163
andesite, 73, 77, 78, 82, 350 bed-load, 157, 168
anhydrite (CaSO4), 179, 344 beds, 156
Anning, Mary, 249 biotite (mica), 109, 130, 345
anorthosite, 299 bivalves, 244–5, 254–7
anticline, 203, 204, 206–7, 218 black smokers, 93, 227, 240, 314
apatite, 344 blueschist facies, 131, 133
aragonite, 181, 185 brachiopods, 244
Archean Eon, 244, 270–4, 351, braided rivers, 170
355, 356 Bretz, J. Harlen, 167
argon, 28 bricks, 212, 232
arthropods, 245 brittle deformation, 190, 202
ash, 75, 78–80 buildings, earthquake damage,
clouds, 99 41–3
asteroids, 30–1, 226, 294–6, 298, Burgess Shale, Canada, 258
310–12 burrows, 177, 240, 255–6
asthenosphere, 23–5, 57, 93–4
atmosphere, 11–12, 242 calcite, 185, 344, 345, 353, 354
other planets, 301, 304, 306–7, calcium carbonate (limestone), 134,
308, 320 140, 180–3, 185, 246
atolls, 182 calderas, 87–8
Avalonia, 278–80 Callisto, satellite of Jupiter, 315
Index 379

Cambrian period, 244–7, 252, 275–8 concrete, 232
carbon cycle, 5 conglomerate, 353
carbon emissions, 287–8 conservative plate boundaries,
carbon footprint, 287 63–4, 66, 72–3
cassiterite, 346 constructive plate boundaries,
caves, 151 57–67, 72–3, 88–94, 116, 286
cement, 184–5 contact, intrusive 124, 331
manufactured, 212 contact (thermal) metamorphism,
chalcopyrite, 346 124–7
chalk, 181 continental collision, 55–6, 62
channelled scablands, 149–51, 167 continental crust, 19–21, 27, 55–6,
chert, 182, 354 63
Chicxulub crater, Mexico, 251–2, formation, 270–83
298 rifting, 57–61, 66, 88–94, 116
chilled margins, 120 continental drift, 61
China continental flood basalts, 89–91, 115
carbon emissions, 287 convection, 24–6
coal mining, 224–5 copper, 225, 227
chlorite, 130, 132, 345 coral, 182, 246–7
chromium, 225 reefs 182–3, 263
clastic sedimentary rocks, 156–61, core of the Earth, 15–18, 21–2, 24, 26
183, 353 craters (impact), 251–2, 295–8
clay minerals, 138–9, 179, 185, 222, lunar, 269–70, 295, 300–1
231, 354 on Mars, 309
clay particles, 142–3, 171–2, 184 craters (volcanic) 75, 77, 83, 87–8
claystone, 353 see also mudrock cratons/shields, 273–4
cleavage, 109, 132, 204 Cretaceous Period, 249
metamorphic, 141 cross-bedding, 158–61, 162–3
mineral, 132, 141–2 crust of the Earth, 19–22
cliffs, 151–2, 181, 331, 333 (see alsocontinental crust;
climate change, 2, 5, 147, 221, 250, oceanic crust)
284–9 fault lines, 36–7
Mars, 308 formation, 268–83
Venus, 304 heat, 25, 27
coal, 221–5 melting 30, 84, 114, 133
coastal erosion, 151–2 strength, 22–5
coastlines, deposition, 177–83 subduction see destructive plate
coccolithophores, 181 boundaries
compaction, 184–6 crystals, 3, 21, 28–9, 74, 106–12,
compass clinometers, 333–5 116–20, 124–6, 131–3, 138,
composite cone volcanoes, 72–3, 77 141–2, 179–80, 185, 190, 225,
compressional waves (P-waves), 268, 301, 317, 331, 343–7,
16–17, 19, 37 349–50, 352–4
380

currents, 162 interior, 12–18
Cyprus, copper ores, 227 magnetic field reversal, 59–60
mantle see mantle
Darwin, Charles, 182, 262 origin, 29–32
dating radiogenic heating, 26–7
biostratigraphic, 240, 261, 262 earthquakes, 13–16, 36–7
cratering timescale, 298, 305 aftershocks, 37, 42, 44
magnetic reversals, 60 damage caused by, 41–3
radiometric, 28–9, 208, 259–60, distribution, 52–3
270 measuring, 38–40
debris avalanches, 83–4 plate tectonics, 53–7, 65, 67–8
Deccan Traps, India, 89–90 prediction, 44–5
decompression melting, 57–8 preparation, 43–4
deformation, 190–208 scales, 38–41
brittle, 190 tsunamis, 45–7
ductile, 190, 204 volcanic, 97–9
deltas, 171–6, 222 echinoderms, 246, 256
density of the Earth, 12–16 eclogite, 130–1, 133
deposition, 206–7 emerald, 347
destructive plate boundaries, 54–5, Enceladus, satellite of Saturn,
57, 61–3, 64–5, 72–6 317–18
Devonian period, 278–80, 286 epidote, 130, 133, 345
diagenesis, 184–6, 215, 222, 2297 Eris, 294, 316
diamond, 297, 344, 347 Eros asteroid, 312
dinosaurs, 249, 250–3 erosion, 4–5, 142, 148–52, 194,
dip, 334–5 206–7
dolomite (mineral), 179, 354 glacial, 145–7
drainage patterns, 149–51 granite, 114–15
drilling, 216–17, 220–1 marine, 151–2
ductile deformation, 190, 202–6 rivers, 148–9, 168–9
dunes and ripples, 157–61, 164 wind, 140, 148
duplex complexes, 198 eruptions (see also volcanoes)
dykes, 116–19, 192 cloud/column, 86–7
directed blast, 82–3
Earth, 10 explosive (pyroclastic), 74–7, 80–1
age, 28–9 plinian, 78–82, 86
atmosphere, 11–12 prediction, 97–9
core, 16–18 estuaries, 173–4
crust see crust of the Earth eukaryotes, 243
density, 12–16 Europa, satellite of Jupiter, 313–15
future, 288–9 European Space Agency (ESA),
heat, 25–7 302–3, 320evaporites, 179–80,
history, 268–86 281
Index 381

evolution, 11, 241–53 synclines, 203, 204
exhumation, 114 foliation, 132–3, 352
exoplanets, 31, 321 foraminiferans, 181, 252, 263
exotic terranes, 66, 277, 283 fossil fuels, 212, 214–25, 287
exposures, rock, 329, 330 fossils, 240–53, 276–9
extinctions, mass, 250–3 body, 254, 263, 331
cells, 242
facies, metamorphic, 129–33, 353 collecting, 333–4, 336
facies, sedimentary 130, 166 formation, 253–9
fallout deposits, 79–82, 86, 88, 98, human, 249–50
161 marine, 177, 246–7, 278
faults, 36–7, 44, 66–7, 190–208 oldest, 242–3
compressional, 16–17, 19, 37 plant, 259
extensional, 196, 203 in sediment dating, 259–63
Great Glen, 194 trace, 331
low-angle, 196–8, 202 zone, 262
normal, 191, 193–6, 200–1 fracking (hydraulic fracturing), 217,
reverse, 191, 194, 196 219–20
San Andreas, 36, 64, 67, 192, fractional crystallization, 110
194, 288 fracture zones, 64
thrust, 196–200, 302 fumaroles, 76–7
transcurrent, 192–4, 197–8,
200–1, 288 gabbro, 116–17
transform, 63–4 galena, 346
feldspar, 108–9, 110, 344, 345, 350, Ganymede, satellite of Jupiter, 315,
354 317
rotting, 138–9, 143 garnet, 126, 130, 345, 347, 352
fieldwork, 328–31, 339–40 gas
equipment, 331–6 magmatic (volcanic), 74–7, 80,
safety, 336–7 86, 98, 119
fish, 256 natural, 212, 214–21
flaser bedding, 163 primordial, 268
flash floods, 171, 337 geological maps, 338
flesh, preservation, 257–8 geothermal gradient, 25
flint, 336, 354 geothermal power, 212
flocculation, 172 geysers, 100
flood basalts, 89–91, 115, 151, 250, giant impacts, 30, 32, 268, 299 this
253, 283, 305 far 30 Apr
floodplains, 167–70, 173, 233 glaciation, 146
fluorspar (CaF2), 232 glaciers, 140, 145–8, 170
folds, 202–6, 340 transport of sediment, 165–6
anticlines, 203, 204, 206–7, 218 glass, 120
isoclinal and recumbent, 204–5 Glicken, Harry, 84
382

global warming, 2, 147, 212, 214, horses, 198–200
286–8 horsts, 195, 199
gneiss, 133, 270, 270, 352 hot (hydrothermal) vents, 93, 314
gold, 225, 226, 231 hot spot volcanism, 89–90, 96
Gondwana, 277–9, 280–1, 284 human fossils, 249–50
grabens, 195, 200 hummocky cross-stratification,
grains 164, 177
shape and texture, 142–4, 166 Hutton, James, 28–9, 208, 280
size, 109, 118–20, 170, 175 hyaloclastite, 92
transport, 143–5, 157–66, 177–8 hydration melting, 74
granite, 84, 85–6, 112–16 hydraulic fracturing (fracking),
granite-greenstone terrain, 273 219–20
granitic magma, 86, 111–12 hydrocarbons, 214–21
granulite facies, 130–1, 133 hydrological cycle, 5, 320
graphite, 216 hydrolysis, 138–40
graptolites, 246, 262 hydrothermal circulation 93, 226–8,
gravel, 164, 231, 233, 353 314
Great Glen Fault, Scotland, 194 hydrothermal vents, 93, 240–1, 314
greenhouse effect, 181, 212, 215,
286–8 Iapetus Ocean, 278–80
on Venus, 304, 306 ice, 316–21
greenschist facies, 130–1, 133 glaciers, 145–8
groundwater, 234–5 transport of sediment, 165–6
Gutenberg, Beno, 38 ice ages, 147, 284–6
gypsum (CaSO4.2H2O), 179–80, ichthyosaurs, 248, 249
232, 345 igneous rock, 3, 106–20
acid (granitic), 84, 107, 350 see
Hadean Eon, 244, 355 also granite
Half Dome, California, USA, 115 basic, 108, 344, 350–1
halite (NaCl), 179 intermediate (andesitic), 107–9,
Hallucigenia, 258 279, 350
Hamersley Basin, 230, 275 intrusive, 106, 116, 119, 276
hand lenses, 332, 339 pyroclastic, 74, 120, 192, 350
Hawaii, USA ultrabasic, 108, 110, 350–1
tsunamis, 47 volcanic, 60, 106, 118–19, 279,
volcanoes, 94–6, 307 349
headlands, 151, 152 ignimbrite eruptions, 86–8
heat flow, 24 illite, 139, 345
heat generation, 25–7, 67, 114, 268, illmenite, 346
289, 298, 311, 313, 316–7 imbrication, 165
hematite (Fe2O3), 139, 276, 346 interglacials, 147
Himalayas, 56, 200–1, 288 Io, satellite of Jupiter, 294,
hornfels, 125–6, 131, 352 312–13
Index 383

iron, 20, 29–31, 108, 125, 139, 225, lead, 28, 225, 228
229–30, 270, 311 levées, 168–9
in the Earth’s core, 12, 17, 22, 32, life 11–12, 235
294, 302 extraterrestrial 308, 314
in Mercury, 302 origin, 2, 93, 240–1
ironstone, 223, 229–30 on land 247–50
island arcs, 64, 65, 72, 279, 281, 288 limestone, 5, 76, 125, 140, 149,
isostasy, 21 177, 180–3, 185, 196, 212, 217,
228, 232–3, 235, 249, 254, 257,
Japan, plate tectonics, 43, 46, 54–5, 304
65–6, 72, 98–100, 289 metamorphosed, 134, 353
jet, 347 oolitic, 183, 354
Johnson, Dave, 83–4 lineation, 132
joints, 141 lithosphere, 23–5, 53, 56, 67
Jupiter, satellites, 312–15, 317 other planets, 298, 301, 307, 319
longitudinal dunes, 161
kaolinite, 226, 231, 345 longwall mining, 223–4
Kelvin, Lord, 28–9 low-angle faults, 196–8, 202
komatiite, 272, 351 low-speed layer, 23
K-T boundary, 249, 250, 252–3 Lyell, Sir Charles, 167
Kuiper Belt, 294–5, 316
kyanite, 126–7, 345 magma, 58, 73–4, 89, 98, 100,
106–7, 112–3, 212, 270, 268–9,
lahars/mudflows, 81–2, 85, 165–6 317
landfill, 233 chambers 116–7
landslides crystallization, 110–12, 119–20,
marine, 258 124, 301, 349
on Mars, 310–11 dyke formation, 116–19
transport of sediment, 141, 165–6 explosions, 74–7
lapis lazuli, 347 frozen, 120
Lascar volcano, 78 granitic, 86, 111–12
Laurentia, 277–80, 282 ocean, 268–9, 300–1
Laurentian Shield, 273, 277 magnetic stripes, 59–61, 282
lava flows, 74, 76–8 magnetism
andesitic 85 Earth’s core, 18
basalt, 89–92, 94–7, 283 exploration tool 220–1, 228
domes, 80–1, 83–4, 87 Mercury, 302–3
granitic, 85–6 of the ocean floor, 59–61, 282
hazards, 96–7 palaeomagnetism, 275–7
icy, 319 magnetite, 346
other planets, 305, 308, 310, 313, mammals, evolution, 248–50, 253
319 mammoths, 250
underwater, 91–2 mantle, 19–22, 268, 270
384

convection 24–6, 67 minerals, 110–20
decompression melting, 57–8 evaporites, 179–80
heat, 25–6 limestone, 185
plumes, 89, 90, 94, 282–3, 307 metamorphism, 120, 124–34
strength, 22–5 mining, 232–4
maps, 39, 44, 52, 65–6, 80, 84, 97, oldest, 272
116, 126, 149, 192, 202, 220, silicates, 106–9
228, 332–3, 338 weathering, 138–43
marble, 134 mining
marine current directions, 163–4 coal, 193, 222–5
marine deposition, 171–83 minerals, 214, 232–4
marine erosion, 151–2 ores, 228–9
Mars, 10, 241, 294, 306–10 Miranda, satellite of Uranus, 318
Martinique, 80 Mississippi delta, 172–3, 176
mass extinctions, 250–3, 284, 298, Modified Mercalli scale, 39–40
355 Moho (Mohorovicic discontinuity),
mass mortality, 258 19, 21
Matthews, Drummond, 59 Mohs’ scale, 344
Mauna Loa volcano, 95, 307 molluscs, 244–5, 252, 254–5
melting, 55, 57, 73, 316 molybdenite, 346
decompression, 57–8 montmorillonite, 139, 345
hydration, 74 Montserrat, 80–1
Mercury, 10, 294, 301–3 Moon, 298–301
Mesozoic Era, 248 craters, 269–70, 295, 300–1
metal ores, 225–32 origin, 30
metamorphic aureole, 125–6 moraine, 165
metamorphic facies, 129–33, 353 mountains, 19, 22–3, 55, 200, 280,
metamorphic grade, 129, 131–3 340
metamorphic rock, 4, 131–4, 141, Mt St Helens, USA, 82–4
270, 352–3 mud, 162–5, 167, 169, 173, 178
metamorphism, 120, 124–34 porosity, 184
contact/thermal, 124–7 mudflows/lahars, 81–2, 85, 165–6
dynamic, 202 mudrock, 125–6
regional, 127–9, 131 compaction, 184
meteorites, 17, 29, 31, 241, 297, 310 muscovite, 109, 130, 345
methane, 30, 216, 222, 240, 285, mylonites, 202
287, 320–1
mica, 109, 132–3, 139, 142, 345 NASA, 302, 310, 320
microbes, early, 241 natural gas, 212, 214–21
microcontinents, 278 Neptune, satellites, 319
mid-ocean ridges, 61, 286 Newman, Mike, 247
migmatites, 133, 352 nickel, 17, 29, 31, 225–6, 231, 297
millipede fossils, 247 Nile delta, 172–3
Index 385

non-renewable resources, 212–34 phenocrysts, 119
normal faults, 191, 193–6, 200–1 phyla, 244–9
nuées ardentes, 80–2, 85 phyllite, 132
physical weathering, 140–2
Obama, President Barack, 287 pillar-and-stall mining, 223–4, 232
oceanic crust, 19–21, 55–6, 270 pillow lava, 91–2, 116
spreading, 57–60, 91–4, 116–17, Pinatubo, Philippines, 98, 99
286, 288 placer deposits, 231
oceans plagioclase feldspar, 108–9, 110
life-cycle, 61–3 planetary embryos, 30, 32, 269, 300
turbidity currents, 162–4 planetesimals, 30, 32, 269, 310
oil, 214–21 planets, 10, 29–32, 67, 240, 270, 294
Old Red Sandstone continent, 280 definition, 10, 294–5
olivine ((Fe,Mg)2SiO4), 108, 110, 125, plankton, 181
139, 268, 345 plants, 11, 93, 140, 178, 181–2, 215,
Olympus Mons, Mars, 307 221–2, 242, 243, 247, 249, 252
oolites, 183 fossils, 259
opal, 347 plate boundaries, 194
ophiolites, 56, 116, 227, 280 conservative, 63–4, 66
ores, 225–32 constructive, 57–61, 66, 88–94,
organisms, evolution, 240–53 116
orogeny, 280 destructive, 54–5, 57, 61–3, 64–5,
outcrops, 338 72–6
oxbow lakes, 169 plate tectonics, 53–7, 65, 190, 289,
oxygen levels, 11, 215, 221, 230, 304–5
232, 241–2, 253, 258, 259, 275, cause, 67–8
285 platinum, 225
plinian eruptions, 78–82, 86
P-waves, 16–17, 19, 37 Pluto, 10, 294–5
Pacific Plate, 54 plutons, 115–16, 117, 120
pahoehoe, 95–6 point-bar deposits, 170, 173
palaeomagnetism, 275–7 polymorphs, 126, 181, 185, 343
Pangea, 281–3, 286, 289 pore water, 184–5, 217
Pardee, Joseph, 167 porosity, 184, 217
partial melting, 57–8, 73–4, 84, 268, potash, 232
270–1, 317 potassium, 19–20, 26–7, 108, 139,
peat, 221 351
pebbles, deposition, 164–5 decay to argon, 28
pegmatites, 119–20, 350 salts, 180, 232
peridotite, 21, 108 potassium feldspar, 108–9
permeability, 217 Precambrian, 244, 356
Phanerozoic Eon, 244, 247, 274–83, preservation, exceptional, 257–8
285, 355 pressure solution, 184–5
386

Principle of Faunal Succession, 262 current directions, 162–3
progradation, 174–8 deposition by, 167–71
prokaryotes, 243 erosion, 148–51
Proterozoic Eon, 244, 246, 274–5, roadstone, 232
282, 284–6 rocks
pumice, 75 ancient, 272
pyrite, 346 cycle, 3–5, 63, 93, 140, 145,
pyroclastic eruptions, 74–7, 80–1 180–1, 186, 270, 285, 306
pyroclastic flows, 80–2, 85, 120, extraction of hydrocarbons,
161–2 216–21
pyroxene ((Ca,Mg,Fe)2Si2O6), 108, faults, 190–208
110, 139, 268, 345 fractures, 141–2
grain size, 109, 118–20
quarries, 12, 139, 176, 212, 216, names, 349–54
228–9, 230, 232–4, 330–1, 337, observations, 330–1
339 sedimentary see sedimentary
quartz, 107–8, 134, 139–40, 142–4, rocks
339, 344, 346 texture, 131–3
quartzite, 134, 272, 353 transport, 141, 143–5
types, 3–4
radiogenic (radioactive) heating, rubidium, 28
26–7, 32, 67, 114, 289, 298, ruby, 347
311, 313, 317 Ruiz, Colombia, 82
radiometric dating, 28–9, 60, 208,
242, 259–61, 270, 295, 298–9, S-waves, 16–17, 21, 23, 37
355 sabkhas/salt flats, 178–80
recrystallization, 28, 124, 134, 185, saltation, 144–5, 157
190, 256 Samoa, 47
reefs, 182 San Andreas Fault, California, 36,
regional metamorphism, 127–9, 64, 192, 194, 288
131 sand, 141–5, 152, 157–65, 168, 170,
reserves, 213–14 171, 173, 177–8, 183
reservoir rocks, 217 sandstone, 130, 184–5, 190, 196,
resources, 213–14 217, 228, 230, 235, 247, 280,
nonrenewable, 212–34 336, 352–3
renewable, 234–5 dunes 157, 159–61
reverse faults, 191, 194, 196 metamorphosed 132–4
Richter, Charles, 38 sapphire, 347
Richter scale, 38–9, 41 satellites, 312–16, 317–21
Ries crater, Germany, 297 Saturn, satellites, 315, 317–18,
ring dykes, 118 320–1
ripples and dunes, 157–61, 164 Scandinavia, 23
rivers schist, 133, 352
Index 387

scoria cones, 84, 86 silver, 212–13, 214, 225, 226
sea-floor spreading, 57–60, 91–4, slate, 132, 352
116–17, 286, 288 slickensides, 196
sea level, 147, 285–6 slip-faces, 158
seamounts, 94 Smith, William, 192
sedimentary rocks, 4, 156–7, 166–7, Snowball Earth, 284–6
183–6, 190–2, 229, 273, 279, Solar System, 269, 294–321
330, 333, 338, 353–4 origin, 29–32
dating, 240 rocky planets, 10
sediments source rock, 215
dating, 259–63 sphalerite, 346
becoming sedimentary rocks, Stensen, Niels, 262
183–6 stick–slip motion, 37, 54, 64
transport, 141, 156–83 stoping, 113–14
seifs, 161 storms, 151
seismic surveys, 44, 116, 202, 220, strain, 190
251 strata, 192, 203–4, 206–7
seismic waves, 13–16 stratigraphic traps, 217–18
measuring, 21, 38–9 stratigraphy 192, 240, 259–61
P-waves, 16–17, 19, 37 stress, 190
S-waves, 16–17, 21, 23, 37 strike–slip/transcurrent faults,
seismometers, 14–16, 38, 98, 301 192–3, 198, 200–1
serpentine, 346 stromatolites, 242
shale, 184, 216–17, 228, 232, 258, strontium, 28
336, 353 structural traps, 217–18
shear zones, 202 subduction, 54–5, 57, 61–3, 64–5,
sheeted dykes, 116–18 72–6
shells, 173, 177–8, 181–2, 185, 235, Sudbury igneous complex, Canada,
243–4, 254–6, 258, 339, 354 226, 231, 297
shield volcanoes, 94–5 Sumatra, tsunami, 47
shields/cratons, 273–4 Sun, origin, 29–32
siderite, 230 supercontinents, 281–2
silica (SiO2) superposition, 262
chemical, 21–2, 58, 73, 139–40 supervolcanoes, 86
minerals, 107, 109 suspension-load, 157, 168, 172
silicates, 21–2, 29, 106–9, 125, synclines, 203, 204
138–9, 156, 231, 272, 294, 302,
313, 317 talc, 344, 346
sillimanite, 125–7 terranes, 66, 277, 283
sills, 118–19 Tethys ocean, 281, 283
silt, 162, 164, 167, 170, 177, 184, texture of rocks, 131–3
353 thermal (contact) metamorphism,
siltstone, 353 124–7
388

Thomson, William (Lord Kelvin), clouds, 99
28–9 volcanic bombs, 79–80
thorium, 26–7, 114 volcanic mudflows, 81–2, 85,
thrust faults, 196–200, 302 165–6
Tibetan plateau, 200–1 volcanoes, 64–5, 72–97
tidal flats, 178 disruptive effects, 99–100
tidal heating, 313–14, 317 dykes, 118
tidal waves see tsunamis on Mars, 307
tides, 150, 163, 173, 178, 212 predicting, 97–9
time, geological, 28–9, 208, 241,
244, 355–6 Walcott, Charles, 258
tin, 226 Walther, Johannes (Walther’s Law),
Titan, satellite of Saturn, 315, 175–7
320–1 water
topaz, 344, 347 chemical breakdown of minerals,
transcurrent faults, 192–4, 197–8, 138–40
200–1, 288 erosion, 148–52
transform faults, 63–4 on Mars, 308–9
transport, 141, 143–5, 157–83 on the Moon, 299–300
trenches, ocean floor, 162–4 resources, 234–5
Triassic Period, 283, 286 transport of grains, 143–5
trilobites, 260, 262 transport of sediment, 157–65,
Triton, satellite of Neptune, 319 167–83
tsunamis, 45–7, 84 waves, 163–4, 173, 177–8, 183
turbidity currents, 162–4 erosion, 151–2
weathering, 138–43, 186
unconformities, 206–8, 280 chemical, 124, 138–40, 231
uranium, 26–7, 214, 228, 232 physical, 140–2
decay to lead, 28 Wegener, Alfred, 282
Uranus, satellites, 317–19 wells, 219, 220, 234–5
USA, carbon emissions, 287 wind, transport of grains, 144, 157,
61
Valles Marineris, Mars, 310 Witwatersrand Basin, South Africa,
valleys, 148–51 231–2
vascular plants, 247
veins, 119, 133–4, 226–7, 231, 338 Xi Jinxing, President of China, 287
Venus, 67, 294, 302, 304–6, 308
vertebrates, 256 Zechstein Sea, 180
Vesta asteroid, 310–11 zeolite facies, 131–2, 353
Vine, Fred, 59 zinc, 228
volcanic ash, 75, 78–80 zircon grains, 272, 346
Index 389

Geology A Complete Introduction PDF

Uploaded by

Copyright:

Available Formats

Geology A Complete Introduction PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Geology A Complete Introduction PDF

Uploaded by

Copyright:

Available Formats

Geology

601550_Geo_A_Comp_Intr_Book.indb 1 5/8/15 12:29 PM

601550_Geo_A_Comp_Intr_Book.indb 3 5/8/15 12:29 PM

601550_Geo_A_Comp_Intr_Book.indb 4 5/8/15 12:29 PM

601550_Geo_A_Comp_Intr_Book.indb 5 5/8/15 12:29 PM

601550_Geo_A_Comp_Intr_Book.indb 6 5/8/15 12:29 PM

601550_Geo_A_Comp_Intr_Book.indb 7 5/8/15 12:29 PM

601550_Geo_A_Comp_Intr_Book.indb 8 5/8/15 12:29 PM

601550_Geo_A_Comp_Intr_Book.indb 9 5/8/15 12:29 PM

601550_Geo_A_Comp_Intr_Book.indb 10 5/8/15 12:29 PM

601550_Geo_A_Comp_Intr_Book.indb 1 5/8/15 12:29 PM

601550_Geo_A_Comp_Intr_Book.indb 2 5/8/15 12:29 PM

Rock types and the rock cycle

1 Introduction: a mighty matter of legend 3

601550_Geo_A_Comp_Intr_Book.indb 3 5/8/15 12:29 PM

In nature, the rock cycle is not as simple as it may seem in

601550_Geo_A_Comp_Intr_Book.indb 4 5/8/15 12:29 PM

The structure of this book

1 Introduction: a mighty matter of legend 5

601550_Geo_A_Comp_Intr_Book.indb 5 5/8/15 12:29 PM

601550_Geo_A_Comp_Intr_Book.indb 6 5/8/15 12:29 PM

1 Introduction: a mighty matter of legend 7

601550_Geo_A_Comp_Intr_Book.indb 7 5/8/15 12:29 PM

601550_Geo_A_Comp_Intr_Book.indb 9 5/8/15 12:29 PM

Spotlight: What is a planet?

I will describe some of our Solar System’s other geologically

601550_Geo_A_Comp_Intr_Book.indb 10 5/8/15 12:29 PM

2 Inside and outside: the solid Earth and the atmosphere 11

601550_Geo_A_Comp_Intr_Book.indb 11 5/8/15 12:29 PM

Key idea: The atmosphere

The Earth’s interior

601550_Geo_A_Comp_Intr_Book.indb 12 5/8/15 12:29 PM

2 Inside and outside: the solid Earth and the atmosphere 13

601550_Geo_A_Comp_Intr_Book.indb 13 5/8/15 12:29 PM

Figure 2.2 Because the speed of transmission of seismic waves increases

601550_Geo_A_Comp_Intr_Book.indb 14 5/8/15 12:29 PM

2 Inside and outside: the solid Earth and the atmosphere 15

601550_Geo_A_Comp_Intr_Book.indb 15 5/8/15 12:29 PM

Key idea: Seismic waves and the Earth’s interior

The Earth’s core

601550_Geo_A_Comp_Intr_Book.indb 16 5/8/15 12:29 PM

density (tonnes per cubic metre)

2000 4000 6000

Key idea: The Earth’s core

2 Inside and outside: the solid Earth and the atmosphere 17

601550_Geo_A_Comp_Intr_Book.indb 17 5/8/15 12:29 PM

Key idea: Magnetism

The Earth’s crust and mantle –

601550_Geo_A_Comp_Intr_Book.indb 18 5/8/15 12:29 PM

2 Inside and outside: the solid Earth and the atmosphere 19

601550_Geo_A_Comp_Intr_Book.indb 19 5/8/15 12:29 PM

sea level water

Figure 2.6 Cross-sections showing crust of different thicknesses resting

601550_Geo_A_Comp_Intr_Book.indb 20 5/8/15 12:29 PM

The concept of blocks of crust floating in equilibrium upon

Key idea: The mantle

Summary of compositional layers

2 Inside and outside: the solid Earth and the atmosphere 21

601550_Geo_A_Comp_Intr_Book.indb 21 5/8/15 12:29 PM