Milestones in Geology (Geological Society Memoir No. 16)

Milestones in Geology Reviews to celebrate 150 volumes of the

Journal of the Geological Society

Geological Society Memoirs

Series Editor A. J. FLEET

, !j

Parliamentary-style meeting room of the Geological Society at Burlington House before 1975.

The meeting room after renovation.

Milestones in Geology Reviews to celebrate 150 volumes of the

Journal of the Geological Society E D I T E D BY

M. J. LE BAS University of Leicester, UK

Memoir No. 16 1995 Published by The Geological Society London

THE GEOLOGICAL SOCIETY The Society was founded in 1807 as the Geological Society of London and is the oldest geological society in the world. It received its Royal Charter in 1825 for the purpose of 'investigating the mineral structure of the Earth'. The Society is Britain's national society for geology with a membership of 7500 (1993). It has countrywide coverage and approximately 1000 members reside overseas. The Society is responsible for all aspects of the geological sciences including professional matters. The Society has its own publishing house, which produces the Society's international journals, books and maps, and which acts as the European distributor for publication of the American Association of Petroleum Geologists and the Geological Society of America. Fellowship is open to those holding a recognized honours degree in geology or cognate subject and who have at least two years' relevant postgraduate experience, or who have not less than six years' relevant experience in geology or a cognate subject. A Fellow who has not less than five years' relevant postgraduate experience in the practice of geology may apply for validation and, subject to approval, may be able to use the designatory letters C Geol (Chartered Geologist). Further information about the Society is available from the Membership Manager, The Geological Society, Burlington House, Piccadilly, London W1V 0JU, UK.

Published by the Geological Society from: The Geological Society Publishing House Unit 7 Brassmill Enterprise Centre Brassmill Lane Bath BA1 3JN UK (Orders: Tel 01225 445046 Fax 01225 442836) First published 1995 The publisher makes no representation, express or implied, with regard to the accuracy of the information contained in this book and cannot accept any legal responsibility for any errors or omission that may be made. © The Geological Society 1995. All rights reserved. No reproduction, copy or transmission of this publication may be made without written permission. No paragraph of this publication my be reproduced, copies or transmitted save with the provisions of the Copyright Licensing Agency, 90 Tottenham Court Road, London W1P 9HE. User registered with the Copyright Clearance Center, 27 Congress Street, Salem, MA 01970, USA: the item fee code for this publication is 0435-4052/95/$7.00.

British Library Cataloguing in Publication Data A catalogue record for this book is available for the British Library 1SBN 1-897799-24-1

Typeset and Printed by Universities Press (Belfast) Ltd, Northern Ireland

Distributors USA A A P G Bookstore PO Box 979 Tulsa OK 74101-0979 USA (Orders': Tel (918) 584-2555 Fax (918) 548-0469) Australia Australian Mineral Foundation 63 Conyngham Street Glenside South Australia 5075 Australia (Orders: Tel (08) 379-0444 Fax (08) 379-4634) India Affiliated East-West Press pvt Ltd G-1/16 Ansari Road New Delhi 110 002 India (Orders: Tel (11) 327-9113 Fax (11) 326-0538) Japan Kanda Book Trading Co. Tanikawa Building 3-2 Kanda Surugadai Chiyoda-Ku Tokyo 101 Japan (Orders: Tel (03) 3255-3497 Fax (03) 3255-3495)

Foreword

vii

LE BAS, M. J. Introduction

1

RUDWlCK, M. J. S. Historical origins of the Geological Society's Journal

5

WINDLEY, B. F. Uniformitarianism today: plate tectonics is the key to the past

11

HALL, R. P. & HUGHES, D. J. Early Precambrian crustal development: changing styles of mafic magmatism

25

ROGERS, G. & PANKHURST, R. J. Unravelling dates through the ages: geochronology of the Scotting metamorphic complexes

37

BLUCK, B . J . W . Q .

57

Kennedy, the Great Glen Fault and strike-slip motion

BROWN, M. P - T - t evolution of orogenic belts and the causes of regional metamorphism

67

MCKERROW, W. S. The development of Early Palaeozoic global stratigraphy

83

FORTEY, R. A. Charles Lapworth and the biostratigraphic paradigm

93

RILEY, N. J. Dinantian (Lower Carboniferous) biostratigraphy and chronostratigraphy in the British Isles

105

CALLOMON,J. H. Time from fossils: S. S. Buckman and Jurassic high-resolution geochronology

127

SAVAGE, R. J. G. Vertebrate fissure faunas with special reference to Bristol Channel Mesozoic faunas

153

COCKS, L. R. M. Triassic pebbles, derived fossils and the Ordovician to Devonian palaeogeography of Europe

165

ALLEN, J. R. L. Sedimentary structures: Sorby and the last decade

175

SELLWOOD,B. W. Structure and origin of limestone

185

WALKER, G. P. L. Flood basalts versus central volcanoes and the British Tertiary Volcanic Province

195

WILSON, M. Magmatic differentiation

205

ATHERTON, M. P. Granite magmatism

221

RANKIN, A. H. Hydrothermal orefields and ore fluids

237

BAILEY, D. K. Carbonate magmas

249

Index

265

Foreword The Geological Society, which is the senior Earth science society in the World, was founded in 1807 for the purpose 'of investigating the mineral structure of the Earth'. In keeping with the place of science in society at the time, it soon received its Royal Charter (1825). The Society's role today is not so different in essence: as a learned society it is primarily concerned with the furtherance of scientific knowledge. This is achieved through debate and, of particular relevance here, through the publication of the results of scientific investigation, analysis and discussion of findings. The Society's principal medium for publication is the Journal. It first appeared in 1845 and has continued, without break, since that time. Hence we arrive at volume 150, and this book celebrates that event. I am sure that readers of this book will not only learn much about how our science has progressed and where the frontiers lie, but will also find interesting the manner in which the Geological Society played the major role in this advance. The Society has grown over the years both in its membership (now over 7000) and in the range of its activities, publications and responsibilities. To its role as the leading UK Earth science society, has been added that of representing professional geologists in the UK and, through the European Federation of Geologists, throughout Europe. I am pleased of this opportunity to recommend this book, edited by the Journal's Chief Editor, Dr Mike Le Bas, to all Earth scientists. His introduction sets the scene. Charles Curtis President 1992-1994

Foreword

vii

LE BAS, M. J. Introduction

1

RUDWlCK, M. J. S. Historical origins of the Geological Society's Journal

5

WINDLEY, B. F. Uniformitarianism today: plate tectonics is the key to the past

11

HALL, R. P. & HUGHES, D. J. Early Precambrian crustal development: changing styles of mafic magmatism

25

ROGERS, G. & PANKHURST, R. J. Unravelling dates through the ages: geochronology of the Scotting metamorphic complexes

37

BLUCK, B . J . W . Q .

57

Kennedy, the Great Glen Fault and strike-slip motion

BROWN, M. P - T - t evolution of orogenic belts and the causes of regional metamorphism

67

MCKERROW, W. S. The development of Early Palaeozoic global stratigraphy

83

FORTEY, R. A. Charles Lapworth and the biostratigraphic paradigm

93

RILEY, N. J. Dinantian (Lower Carboniferous) biostratigraphy and chronostratigraphy in the British Isles

105

CALLOMON,J. H. Time from fossils: S. S. Buckman and Jurassic high-resolution geochronology

127

SAVAGE, R. J. G. Vertebrate fissure faunas with special reference to Bristol Channel Mesozoic faunas

153

COCKS, L. R. M. Triassic pebbles, derived fossils and the Ordovician to Devonian palaeogeography of Europe

165

ALLEN, J. R. L. Sedimentary structures: Sorby and the last decade

175

SELLWOOD,B. W. Structure and origin of limestone

185

WALKER, G. P. L. Flood basalts versus central volcanoes and the British Tertiary Volcanic Province

195

WILSON, M. Magmatic differentiation

205

ATHERTON, M. P. Granite magmatism

221

RANKIN, A. H. Hydrothermal orefields and ore fluids

237

BAILEY, D. K. Carbonate magmas

249

Index

265

From Le Bas, M. J. (ed.), 1995, Milestonesin Geology, Geological Society, London, Memoir No. 16, 1-4

Introduction M . J. L E

BAS

Department of Geology, University of Leicester, Leicester LE1 7RH, UK

Science advances by taking new and unexpected turnings, pioneers opening up pathways which later workers follow and explore. This book takes the reader along several such geological paths that have followed from observations and theories first printed in The Quarterly Journal of the Geological Society of London (since 1971, the Journal of the Geological Society). Following the paths, one sees the history of geological thought during the nineteenth and twentieth centuries. To achieve this desired structure for the book, leading geologists were invited to present their personal views on significant topics that had been brought to the fore in earlier contributions to the Journal, to evaluate the evidence presented and to give their view of how these seminal papers affected our present understanding of geological processes, and further to hazard where future paths of investigation may lie. Bringing these together under one cover serves two purposes: first to celebrate 150 years of continuous publication of papers by the Geological Society of London: second, it makes a British review of the current 'battle lines' across many fields of geological research. Not only will the serious research investigator discover the several turning points which have governed the paths of his study, but the general reader also will discover the delights, the fortunes and machinations taken by many leading British geologists. Others will use the chapters to answer the question: 'Does historical contingency govern the paths of geological exploration, as it has been said to govern the evolution of living creatures?' Most of the papers have already been published as Celebration Papers during 1993 in Volume 150 of the Journal of the Geological Society, hence the size and format of this book. To emphasize the historical context, all the chapters are preceded by the abstracts or prefaces of the seminal papers reproduced from the Journal. The exception is Rudwick's, which sets the scene on why the Geological Society of London rated so highly the publication of geological observation. He recounts the creation of the Journal, and how it survived growing pains to become a leading international ,journal. One of the several maxims followed by geologists is that of uniformitarianism. It was defined by Hutton over 200 years ago and popularized by Charles Lyell in his Principles of Geology and in his many papers read before and published by the Geological Society. Earlier this century, it was considered applicable only to the Phanerozoic. Nowadays with the great advances made in dating techniques, Windley is able to argue cogently that the plate tectonic paradigm can be successfully applied not only to the Proterozoic but even to the Archaean. He shows how the growth of the North American plate and the Kapvaal craton can be explained by the plate tectonic model, and how

secular changes in heat production changed the course of development of igneous and metamorphic processes at subducting plate boundaries. The application to the Archaean is still not universally accepted, but that working model helps interpret the rocks formed during the first half of the Earth's existence. The magmatism during those early times used to be thought to be no different from present-day ones, on the principle of uniformitarianism. In 1951 came the seminal work of Sutton and Watson on the Lewisian gneiss complexes and the mafic dyke swarms cutting them. This became the cornerstone on which gneiss and greenstone dyke complexes were interpreted, and later contributed to the concept of terranes. Pursuing this, Hall & Hughes narrate the magmatic differences that emerged, mainly as the result of the early high heat flow: the unstable komatiitic volcanic-dominated crust mainly in the Archaean; and then the onset of noritic magmatism and the concomitant crustal accretion super-event, as markers of the transition from the Archaean to the Proterozoic. Their contribution provides greater understanding of the evolution of Precambrian mafic magmatism and the formation of the Earth's early crust. The single great technique that enabled the mysteries of Precambrian metamorphic complexes to be unravelled, was the application in 1961 of R b - S r isotope systematics to age determination, by the Oxford school led by Moorbath. K - A r isotope studies had not been enough; too often they produced only thermally reset ages. Nowadays, any study of high-grade metamorphic rocks automatically includes isotope determinations, and the same now applies to igneous rocks. But the whole process has become very sophisticated, and Rogers & Pankhursl present a thorough analysis of the process as applied to Scotland. They show how the techniques were expanded to include the use of the isotopes of lead, uranium, samarium and neodymium, with great success but not without considerable controversy, much of it still running (e.g. Ben Vuirich). Another break-through in studies of the Earth's crust via the rocks of Scotland was the identification by W.Q. Kennedy in the 1940s of the extent of the Great Glen Fault. Here was a fault of apparent massive strike-slip displacement; hitherto faults had been mainly normal, reversed or thrust phenomena. Bluck analyses the extent of this and other major Scottish faults, and the implication is made that massive displacement along lines of fracture are possible through the Earth's crust. That such displacements could occur was an essential ingredient to the theory of sea-floor spreading and to the analysis of many tectonic basins and oil-bearing structures, even to creating space as sphenochasms for the permissive emplacement of granites. Another process which came to be understood through careful field work in Scotland was the identification by George Barrow 100 years ago of metamorphic zones in the

2

M . J . LE BAS

Dalradian schists. It marked the birth of meaningful metamorphic petrology. Barrow's 1893 paper quickly became a classic and was much quoted, even though he mis-identified the heat source for the metamorphism; demonstrating that observation is more important than interpretation. Brown shows how a knowledge of mineral chemistry, textural and field relations which identify stability relations, together with a knowledge of the time relations deduced from age determinations, can produce paths across petrogenetic grids which give information on how pressure and temperature varied with time in different tectonic environments. Even within a single tectonic environment, more than one P - T - t path can be recognized. The next four papers are on stratigraphical classification and correlation, particularly using fossils, one of the main bases of our science and one whose terminology eventually becomes part of everyday language. McKerrow discusses Palaeozoic stratigraphy in general, beginning with the works of Sedgwick and Murchison, and then sets out a logical sequence of steps that might, perhaps should, be followed in developing the stratigraphy of an area, once the structural relations of strata within an area are appreciated. Fortey focuses on the contribution of Lapworth's work on graptolites in the Southern Uplands of Scotland, Riley compares Vaughan's and later work on Early Carboniferous bio- and chronostratigraphy, and Callomon, in reviewing Buckman's pioneering work on the Jurassic of Dorset, goes on to discuss the current limits of resolution attainable in biochronology. It is interesting to reflect on how views on some of the problems addressed and concepts advanced by our predecessors have changed over the years. For instance, Fortey shows how durable Lapworth's biostratigraphy (as we should now call it) has been, while ideas on structure and palaeogeography that he derived concurrently with it have changed almost beyond recognition within the scientific lifetimes of many still active in the field. As Callomon describes, Buckman's practice in collecting and recording fossils in the field, and his ideas on the prevalence of gaps in the rock record, were far in advance of their time, entirely compelling, and relevant to many current concerns in sedimentology and sequence stratigraphy. Yet his contribution has been under-recognized for many years, partly because his views on evolutionary palaeontology were based on theories now superceded, and in his own later practice he departed from his earlier standards. Savage relates the remarkable story of the discovery by Charles Moore of Mesozoic terrestrial vertebrate remains within ,:fissured Carboniferous Limestone, and how the search widened as more species were discovered. This most fortunate means of preservation provided an abundance of material as well as species, and this occurrence and others subsequent have given vertebrate palaeontology a foundation upon which much of our present-day understanding of evolutionary history is based. Even more remarkable is the story unfolded by Cocks of four different faunal assemblages within one set of strata. When fossiliferous quartzite pebbles were found in the Budleigh Salterton Pebble Bed of Triassic age along the Devonshire coast, Salter realized in 1864 that they were different from anything known in Britain but could correlate some of them with Lower Palaeozoic strata in Europe. Then Davidson in 1870 showed there were Devonian fossils as well, and modern work reveals there are four faunas, two

Ordovician and two Devonian, with tectonic reconstruction now explaining all in terms of adjacent palaeocontinents. Present-day geologists sometimes forget the importance of accurate description, going straight to interpretation often based on generalizations of assumed facts. Sorby was a 'quantifier' and had a particularly keen eye, creating meticulous drawings to accompany his precise writings. His President's Address on the study of sedimentary structures, published in 1908 after his death, is taken by Allen as the starting point for marrying description to interpretation. After detailed descriptions, he examines the current understanding of aeolian bedforms, sand-wave bedding, tidal bedding, marine storm bedding, hummocky and swaley cross-stratification, soft sediment deformation and dewatering structures, particularly reviewing the past ten years research into these. Sorby appears again in the next contribution. His 1879 President's Address on the structure and origin of limestone can still be read with pleasure and profit by any modern student. He was far ahead of his time, and virtually invented geological microscopy. From that starting point, Sellwood develops current ideas on limestone classification, their environment of deposition and diagenesis, and their significance in sequence stratigraphy. In the last quarter of the eighteenth century, controversy arose between Geikie and Judd on the interpretation of the Hebridean volcanic complexes of Scotland. Until then, most igneous rocks were regarded in isolation, but Geikie and Judd both realised that there was an association of rock types waiting to be interpreted. This opened the 'school of Hebridean petrology' which was so strongly developed by Harker, Bailey and others, and has flourished ever since. Walker investigates the association of flood basalts with volcanic centres and shows that the former have much more to tell us than most have hitherto supposed: evidence of whether the basalt lavas and dykes were fissure-fed or emanated from point-source vents, and the crustal tension implied; tilting of volcanic fields perhaps related to inflation or deflation of volcanic edifices; and direction of flow of magma. As volcanic hazards become a matter of daily concern, more needs to be learnt about the magmatic plumbing of volcanoes, and the magma flow direction of dykes and sills, now laid bare by erosion within old volcanic structures, could supply the vital data. The study of the anisotropy of magnetic susceptibility would seem to be potentially significant in this respect. In contrast to the structural and mineralogical approach taken by Walker, Wilson takes the geochemical approach, which is equally applicable to volcanic and plutonic rocks. One hundred years ago, Harker began a school of thought on the controls of crystallization applied to the differentiation of basic magma as exemplified by variations seen within individual plutonic masses. Nowadays, finegrained rocks are considered to be better representatives of variation in magma chemistry, and Wilson reviews the many possible processes from fractional crystallization, assimilation and magma mixing to thermogravitational diffusion and liquid immiscibility that can be responsible for magmatic differentiation. These processes concern mainly basaltic magmas which, by and large, are partial melt products of the upper mantle of the Earth. By contrast, most granites have their origin in the partial melting of the continental crust of the Earth. They are the

INTRODUCTION layman's best known coarsely crystalline rock. Most granites are superficially similar, but all are individually distinct when studied geochemically. Fifty years ago, the 'granite controversy' raged, with the Read school upholding the metamorphic and migmatitic association, and the TilleyBowen school maintaining that granites were the product of crystal fractionation from basic magmas. Atherton reviews the pros and cons of granitization, the 'room problem' for plutons, the evidence of melting experiments, the use of rock and mineral geochemical analytical data in determining the production of acid magma by partial melting or by fractional crystallization, and the manner in which isotopes may identify the source rock and the region. He takes the example of Garabal Hill in Scotland, which is one of the few Caledonian complexes including ultrabasic to acid igneous rocks, where crustal contamination might be argued. Also considered are thoughts on the association of basins, granites and thermal highs with high-T/low-P metamorphism. In the penultimate chapter, Sorby's observations on fluid inclusions published almost 140 years ago are shown by Rankin to have led directly to the present state of knowledge about the pressures and temperatures of fluids in rocks, especially igneous-related ones. Fluid inclusions tell us much about mineralization processes, fluids being the carriers of the ore components. Rankin analyses the contribution of current fluid inclusion studies to understanding the many mineralization processes, taking several classical examples from the UK and abroad, some related to igneous bodies and others to tectonically driven crustal circulation of fluids. He also includes some original thermometric data on the main British ore fields. Only recently has undoubted carbonatite been discovered in Britain (and published in the Journal of the Geological Society, 1994, 151, 945). These exotic igneous rocks confounded geologists early this century, who could not believe in igneous 'limestones'--an apparent contradiction. The Journal of the Geological Society has a long history of publishing papers on African geology, a product of the past colonial era. In 1956 Campbell-Smith presented his review of African carbonatites, coinciding with a similar review by Pecora in the USA, and the two changed world opinion. Geologists flocked to the 1960 International Geological Congress in Norway and Sweden and saw the Fen and Alno carbonatite complexes. Calcite carbonatite became an acceptable igneous phenomenon. CampbellSmith's review showed the igneous nature of carbonatites: their occurrence as cross-cutting dykes with fine-grained margins (i.e. chilled) and as small plugs with thermal contacts marked by alkali metasomatic reaction aureoles (fenitization). Their origin remains controversial with three main current theories: they were produced by fractional crystallization of nephelinitic magmas; they were separated by liquid immiscibility from a nephelinitic (or melilititic) magma; they were produced by direct partial melting of the upper mantle. Bailey examines the last of these, on the premise that many carbonatites are found without associated nephelinite/phonolite, and on the experimental evidence that dolomite carbonatite melts can be produced in the mantle under CO2-saturated conditions. This view is contrary to the powerful consensus that now exists: that carbonatites are essentially infracrustal differentiates of alkali silicate melts, to the extent that most modern

3

discussions of petrogenesis begin with this assumption. Bailey's chapter re-dresses this imbalance in the literature. The relation of dolomite carbonatites to calcite carbonatite remains uncertain; perhaps like 'granites and granites' there are 'carbonatites and carbonatites'. Their importance in understanding Earth history is undoubted because, having minimal partial melt compositions, they are potentially the best natural products to give clues to the chemical and thermal evolution of the Earth's mantle. Many more seminal papers could have been selected from the pages of the Journal of the Geological Society for essaying in this book. One which changed the character of the British geologist, is that by Howell Williams on 'The geology of Snowdon (North Wales)' published in 1927 (83, 346). Beginning with clear field descriptions, he explicitly interprets glowing avalanches, mass-flow epiclastic deposits, and proximal and distal water-lain tufts from rocks considered by many to be among the most difficult to interpret. Until this exposition, pyroclastic rocks had been by-passed by geologists in Britain, but this paper fired imaginations. The area described became a training ground, and has spawned several generations of geologists renowned for their pyroclastic expertise. On more traditional grounds is the 1938 President's Address by O.T. Jones 'On the evolution of a geosyncline' (94, lx). For the next 30 years, students pondered on geosynclines which were understood to be crustal downwarped structures filled with sediment, and in so doing opened the subject of conditions of sedimentation and the sources of the sediments. Distinct sedimentary basins were recognized. When 'plate tectonics' burst on the scene, the data accumulated supplied the vital items allowing reinterpretation of geosynclines as oceanic trenches. A paper which was to turn geological interpretation upside-down was Bob Shackleton's 1957 paper in the Journal (113, 361) on 'Downward-facing structures of the Highland Border'. These Scottish schistose rocks had been observed to be largely flat-lying but synformal near Aberfoyle. Shackleton's revelation that they were all upside-down led to the re-interpretation that the 'flat belt' of Loch Tay was the lower half of a nappe with the synform being the inverted anticlinal nose of the nappe, this structure extending across the whole of Scotland. Whereas many geological advances are made on a broad front of carefully documented data, here the break-through depended on a few astute observations of way-up criteria on a bleak mountain side. Some regard geophysics as beyond the bounds of normal geology. It is not. In 1906, Oldham presented 'The constitution of the interior of the Earth, as revealed by earthquakes' (62, 456), which foretold how geophysics would contribute to the fundamentals of geology, i.e. the constitution of the Earth's core, mantle and crust. He pointed out that the seismograph 'enables us to see into the Earth' and that the three wave motions observed allowed interpretation of a shell-structure of the Earth. Having defined the depth to the core-mantle boundary and shown the seismic 'shadow zone', he goes on to discuss the possibility of other discontinuities (now mostly confirmed). This in turn has led on to an explanation of the Earth's geomagnetic field, and to the constitutions of the Moon and planets. This was a truly seminal paper. In bringing all these topics together, it is hoped that

4

M.J.

readers will find that the separate topics are not so unrelated; some topics need the others to sustain them, some merge into new topics, but all combined are essential to advancing the frontiers of science. The chapters also give hints on how this frontier may be further advanced. I thank most sincerely each of the authors of the chapters, for writing so assiduously to the briefs given them.

LE BAS I am grateful to the editors of the Journal of the Geological Society for guidance in the early stages of planning this volume and for editing the versions of the chapters published in volume 150 of the journal, and am particularly indebted to John Hudson who provided unstinting assistance at several critical stages.

From Le Bas, M. J. (ed.), 1995, Milestones in Geology, Geological Society, London, Memoir No. 16, 5-8 First published in Journal of the Geological Society, London, Vol. 150, 1993, pp. 3-6

Historical origins of the Geological Society's Journal MARTIN

J. S. R U D W I C K

Science Studies Program, University o f California San Diego, La Jolla, CA 92093-0104, USA Transactions, its earliest periodical (from 1811), published the full texts of a few selected papers, with fine illustrations, but generally long after they had been read at one of the meetings. Conversely, the Proceedings (from 1826) recorded all the papers soon after they had been delivered, but only in abstract and without illustrations. The launching of the Quarterly Journal (1845) was an attempt to combine the advantages and eliminate the disadvantages of the earlier periodicals. After a shaky start, it proved highly successful through the rest of the nineteenth century and much of the twentieth, and was the direct forerunner of today's Journal. Abstract: The Geological Society's

The Geological Society was a publishing body even before it was founded. That paradox is easily explained. One of the reasons for its foundation was the desire of a group of London 'men of science' (the later term 'scientist' would be anachronistic and highly misleading in this context) to give permanent form to meetings that had been concerned with the publication of a specific scientific work. Another reason was the frustration felt by others at the inability of the Royal Society to provide an adequate publication outlet for geological work, and particularly for work that was highly factual in character and localized in content. Those two reasons for the foundation of the Society (there were others too) epitomize the two distinct kinds of publishing activities that have characterized learned societies ever since they proliferated in the eighteenth century. On the one hand there was, and still is, the need to publish the completed results of scientific research and thereby place them permanently on record. On the other hand there was, and still is, the need to inform those with particular interests about the current work of others with the same interests, whether the reasons for seeking such information are those of competition or collaboration or a mixture of the two. Scientific societies have tried to meet both needs through their own publishing activities. By its very nature, the detailed results of scientific research generally appeal to only a relatively small and specialized public, and are therefore often unattractive to ordinary commercial publishers. One solution to this problem that was widely adopted before the twentieth century, and not only for scientific works, was to appeal for subscribers to a particular book before the printing process began; the subscribers' advance payments guaranteed the publisher against loss, and any further sales made after publication could go towards a profit. An alternative solution, however, was for all members of a scientific body, who by definition were a specialized public with common interests, to receive a continuing series of shorter publications in return for a continuing subscription. In effect, members with a great interest in one particular subset of papers received copies of those papers, which might not otherwise have been published at all, in return for subsidizing the papers that were of great interest to other members. This was, and of course remains, part of the rationale behind the publication of any specialized scientific periodical, and the Geological Society's Journal was and is no exception.

At the same time, however, the specialized common interests of the members of any scientific society provide the opportunity for the exchange of opinions and conclusions, and often of course for vehement controversy; indeed the desire for such exchanges has been one of the most common reasons for founding such societies. But unless all the members meet regularly face-to-face, and even more if they are spread widely and unable to meet in that way, they have often felt a need for some kind of newsletter to keep them informed of the current activities of others. Again, this was, and remains, part of the rationale behind the publications of scientific societies, and again the Geological Society was and is no exception. Some of those who founded the Geological Society in 1807 were already subscribers to an important but costly publication. This was a three-volume monograph (1808) on the mineralogy and crystallography of calcium carbonate, by Jacques Louis, Count de Bournon, a French aristocrat who had fled to England from the Revolution in France. The work dated from before the profusion of crystal forms was explained satisfactorily in terms of a small number of types of symmetry and sets of crystal faces. In de Bournon's view, and that of his subscribers, his work required many expensive engraved plates, in order to reproduce a large number of detailed drawings of specific crystal specimens: in terms of illustration, crystallography was in the state in which palaeontology necessarily still remains. So the work was expensive, and could best be published by subscription. The subscribers were of course united by their common interest in research such as de Bournon's, and it was natural for them to regard themselves as a potential core for a permanent society to foster that kind of scientific work. Most of them, however, were already Fellows of the longestablished Royal Society, which had its own Philosophical Transactions for the publication of high-level scientific research (though not of book-length works such as de Bournon's). When, after the Geological Society was founded, its leaders began to talk about starting a periodical of its own, some of the members who were also FRSs were highly critical of that proposal; a few, including the Royal Society's autocratic president, Sir Joseph Banks, even resigned from the Geological Society and for a time put its future in jeopardy. In fact, however, the proposal had been for a periodical that would supplement, and not necessarily compete with, the Philo-

sophical Transactions.

6

M A R T I N J. S. R U D W I C K

The concern of the leaders of the Geological Society was that the Royal would not, and perhaps could not be expected to, publish geological papers with highly detailed descriptions of mainly local interest. In other words they argued that there was a need for a more specialized periodical, to supplement the Royal Society's coverage of all the natural and mathematical sciences. But they also had in mind such Continental periodicals as the Parisian Journal des Mines (founded 1795), which published much practical material of interest to mining geologists as well as reports of more fundamental importance. The conflict within the early Geological Society reflected in part a difference of opinion as to whether it should model itself on a learned society such as the Royal Society, or make itself useful in a more practical way to the owners and managers of Britain's mineral resources. In the event, the former opinion triumphed, and the Society was not, after its earliest years, notably congenial to those whose interests were mainly practical or commercial, still less to those (such as the mineral surveyor William Smith) who did not share the wealth and social status of the Society's leaders. Even within the former model, however, there was in fact a clear precedent for the Geological Society's publication plans. This precedent, which had never aroused the hostility of Banks and the Royal Society, lay in the volumes of the Linnaean Society's Transactions (founded 1791), with their detailed and specialized papers on plant taxonomy. In any case, the Geological Society soon launched its own Transactions, modelled on those of both the Royal and the Linnaean. The quarto format was decidedly lavish, and clearly designed to match the gentlemanly tastes and pockets of the Society's members. The first volume (1811) was priced at £1 12s [£1.60], the second (1814) at £3; these were substantial sums. Added to the membership fee, which rapidly rose to £4 a year, an active gentlemanly interest in geology did not come cheaply. (As a rough-and-ready guide to real values, a n inflation factor of at least 50, and perhaps even 100, should be applied to these prices to make them comparable to modern prices.) Since the price of the Transactions was not included in the membership fee, only the more enthusiastic or more wealthy members bought the volumes, and the sales languished accordingly. As with de Bournon's book, a major expense was the illustrations. Engraving on copper was by far the best medium for the pictures, maps and other diagrams that geological papers required; but engraving was a highly skilled and timeconsuming craft, and correspondingly expensive. Furthermore, many geological illustrations required, or at least were greatly enhanced by, the use of colour. This could only be provided by applying watercolour washes by hand to every copy; and although this work was generally done by poorly paid female labour, it added further costs to the final plates. Still, the volumes were impressive, with handsome letterpress and fine illustrations. The Transactions helped to establish the scientific reputation of the Society, and of the selfconsciously new science of geology, both in Britain and abroad. But the periodical remained a medium of record rather than one for reporting work in progress. The intervals between successive volumes narrowed to about two years, as the Society became more established and the quantity of completed research increased; but there was still generally a long delay between the reading of a paper and its eventual publication. This was ill-suited to a science that was burgeoning rapidly into a major area of research internationally. Members of the Society could and did often seek alternative outlets for more rapid publication; but monthlies such as the Philosophical

Magazine, which at this period carried many geological papers, could not provide comparable illustrations, which were so important in geology. After the first decade, the Society took over the management of the Transactions from the commercial publishers who had handled it initially. A 'Second Series' was launched in 1822 to give the work a new look and to boost its sales. At the same time the opportunity was taken to adopt the new and cheaper technique of lithography in place of copper engraving. The price of the volumes was roughly halved, and authors could now be offered more space for their illustrations; an added bonus was that for most geological subjects (except perhaps maps) the more subtle tones of lithography were positively an advantage. Meanwhile, however, the Society had hardly taken any steps to improve the exchange of provisional ideas and ephemeral information, beyond the primary arena of its meeting room. An 'arena' is what its meetings had famously become: in contrast to the other learned societies in London, the Geological permitted discussion of the papers that had just been read. This was at first a cautious experiment, because there were those who feared it would lead to acrimonious argument; but it soon became an established and successful tradition of lively debate. Almost from its foundation, however, the Society had appealed for the collaboration of those living outside London. Its founders recognized that a geographical spread of the membership was even more valuable for geology than for many other sciences, since widely scattered members could report on local areas that they knew thoroughly. Such informants were enticed with the offer of free 'honorary' membership. But these provincial members could not get a first-hand impression of the current state of geological opinions in the Society, unless they were able to attend its meetings in person, on trips to London that for many of them were expensive, uncomfortable and therefore infrequent. The Society's very first publication, mooted almost immediately the Society was founded, and issued three years before the first volume of the Transactions, was in fact directed at these provincial members, and at those of the 'ordinary' or London members who found themselves travelling for any reason. The publication was a small booklet of 'Geological Inquiries' (1808), which listed the kinds of observations that could usefully be made, and the kinds of specimens collected, in more or less remote areas. It was probably inspired by, and partly based on, the famous 'Agenda' published in 1796 by the great Swiss naturalist Horace-B6n6dict de Saussure. Like that model, it was based on the belief that far more empirical information needed to be collected in the field, before it would be appropriate or profitable to indulge in high-level theoretical speculation about the structure or history of the earth. The Society's booklet certainly produced plenty of local information, most of it in the form of letters to the first President, George Bellas Greenough; in due course he incorporated much of it in his great geological map of England and Wales (1820). Together with the provincial members themselves, the 'Inquiries' gave the Society a network of local informants, so that its premises in London quickly became a centre of research material for the whole of Britain and beyond. However, this still did not give those informants much in return. In 1826, just 20 years after the foundation of the Society, a decision was taken to publish summaries of the papers that had been read, without waiting for their possible and eventual appearance in full in the Transactions. This marked the start of

H I S T O R I C A L O R I G I N S OF THE G E O L O G I C A L SOCIETY'S JOURNAL the Society's Proceedings, a publication that in effect complemented the older and grander periodical. The papers had been summarized in writing since soon after the Society was founded, but only in manuscript for its official minute books. From 1827 the summaries began to be printed and distributed to the Fellows (as they had been termed since the Society's formal incorporation in 1825). The Proceedings was published as a small octavo booklet about six times a year, during the Society's 'season' from November to June. Each issue contained summaries of the papers read at the most recent meetings, together with the names of new Fellows elected and other Society business. One issue each year was devoted to the business of the AGM, and also contained the president's 'Anniversary Address'. The latter had grown from a mere review of the Society's domestic affairs into a summary and assessment of all the papers read during the previous year. Some presidents expanded their survey beyond the Society, giving a major evaluation of the state of geological research nationally and even internationally, and often focusing on some particular aspect of the science. The Proceedings immediately became an important medium for the rapid exchange of news and views about geology in Britain. The periodical was not primarily designed to keep provincial Fellows informed, and indeed they were again at a disadvantage: in view of the high costs of postage, the newsletter (as it was in effect) was distributed only within London, and provincials did not receive it unless they could arrange for a friend in London, or their London club, to hold it or forward it for them. But in practice it was distributed and read widely beyond the capital. Furthermore, the summaries of papers could soon be read even by those who were not FGSs, because the general scientific monthlies took to reprinting them from the Proceedings. So any author who had his paper read at a meeting of the Society could be sure of having at least a summary in print, and widely read, within a month or two. By contrast, the authors of papers selected for publication in the Transactions (after a refereeing procedure much like that of the present) often had to wait a couple of years or more, before seeing their work fully in print and with its illustrations attached. As the volume of work presented at the Society's meetings grew, and its average quality improved, so the disadvantages of this two-track system of publication became more and more apparent. The Transactions languished again, as authors became impatient at the long publication delays; sales remained small, and the financial burden on the Society correspondingly great. Conversely, although the Proceedings provided rapid publication, it was at the cost of omitting the details, and particularly the illustrations, that would have given the papers most of their value and persuasive power. The effects of that dilemma can be seen in the successive issues of both periodicals. The number of papers published in the Transactions declined, in proportion to the number read, while the summaries published in the Proceedings became on average progressively longer. Even a few illustrations crept into the latter, as the Society began to adopt the technique of wood engraving. This was less effective for fine detail than copper engraving; but it was adequate for small maps and sections, it was much cheaper, and above all a wood engraving could be printed on the same page as the text to which it referred, rather than having to be bound separately at the end of the volume. In 1842, a substantial issue of the Transactions brought the problem to a head, because although it was a scientific success

7

it finally made the financial burden of the periodical almost intolerable. The following year the trend mentioned above was formally recognized, when the Society resolved to modify the format of the Proceedings to include much fuller summaries of the papers, with small illustrations on a regular basis. Even a few folding lithographed plates, of maps, sections and fossils, were included. But this palliative failed to yield the anticipated increase in sales. So in 1844 the Society tried another tack. The commercial publishers Longmans agreed to produce a new Quarterly Journal in octavo format, at their own risk and profit and for a trial period of one year. This was to incorporate the Proceedings, now extended to full texts of the papers, and fully illustrated with wood engravings and larger lithographed plates. A 'second, or miscellaneous part' would make the new periodical still more attractive, by reporting on recent geological books and other publications in Britain, and by printing abstracts or extracts, in translation, of significant work from abroad. The intention was that the Transactions would meanwhile continue 'when a paper could only be advantageously given in quarto'. The Quarterly Journal started to appear in 1845, but after the first year Longmans reported that they had made a loss on the venture and would not renew the agreement. In retrospect the reason for the failure is clear. The Society had allowed Fellows to continue to receive the Proceedings free, as they had always done, as an alternative to subscribing to the new quarterly (incorporating the Proceedings) at the commercial price. As the Society's centenary historian commented, many Fellows were evidently 'more concerned in appending F.G.S. to their names than in adding the Quarterly Journal to their bookshelves' (Woodward 1907, p. 157). However, the format of the new periodical was so attractive that its publication was continued at the Society's own expense and risk. Significantly, the Proceedings were no longer to be available separately; Fellows were now faced with an all-ornothing choice. Conversely, the Transactions virtually came to an end as soon as the Quarterly Journal began. Three small issues appeared in 1845~,6, printing papers that had been in the pipeline before the change was decided. By the time a final issue appeared a full decade later, the Transactions had clearly become redundant. The Society had thus decided, in effect, to adopt a compromise between the two earlier forms of periodical, between lavish but slow publication on the one hand, and quick but abbreviated publication on the other. As its name implied, the Quarterly Journal was published rather less frequently than the old Proceedings, but much more frequently than the Transactions. Like the former, it ensured reasonably quick publication; like the latter, what it published were the full texts of papers. Its octavo format made it look like the Proceedings; but it provided illustrations virtually as good as those in the Society's original periodical. They ranged from small wood engravings embedded in the pages of text, to substantial folding engraved plates of geological maps, some of them handcoloured, and lithographed plates of fossils and geological sections. Although initially regarded as an uneasy compromise, the Quarterly Journal proved to be a highly successful formula. It combined the advantages of both its predecessors, with just the right balance to satisfy most authors and most of their readers. In particular, it combined in an adequate manner the functions of both newsletter and medium of record. After the first few years its cost was absorbed into the Fellows' annual fee, so that its purchase became in effect a compulsory condition of

8

M A R T I N J. S. R U D W I C K

membership; that ensured a steady and predictable level of sales, which made it financially sustainable. The Quarterly Journal continued to serve as the Society's sole periodical throughout the rest of the nineteenth century and beyond the middle of the twentieth. The volumes became fatter, and the techniques of illustration were improved, or at least enlarged, by the adoption of photography for landscapes, rock exposures and fossils, and of chromolithography and other methods for coloured geological maps and sections. But the format remained almost unchanged until 1971, when the 'Quarterly' was dropped and the present Journal appeared in its place. Significantly, it has reverted to a larger format similar to the original Transactions, allowing for many larger illustrations to be included without the expense of fold-outs. Even before that change, the need for a separate newsletter had reemerged, for the quick publication of relatively ephemeral material; in that respect the modern Circular (Newsletter from 1972-1990), and its recent successor Geoscientist, represent a revival of one of the functions of the old original Proceedings. In conclusion, the Society's periodicals are now once more surprisingly similar, in form and function, to those of its earliest decades and first Golden Age.

Bibliographical note The system of references conventional in scientific papers is ill-suited to a historical article such as this. Readers who want to pursue this topic further will find that the following historical works ('secondary' sources, in historians' jargon) provide some starting points; they also give references to the contemporary ('primary') sources on which all historical research is properly, indeed necessarily, based. It should be noted that although the pace of research in the history of science is quite as intensive as in geology, historical books and articles generally enjoy a much longer useful life than those in the sciences. Woodward's centenary history (1907) of the Society is still a valuable source, since it prints much otherwise unpublished material from the Society's archives; but it is chaotically organized, and scarcely attempts any historical analysis or interpretation. My article on the foundation of the Society (Rudwick 1963) was based particularly on the manuscript papers of the Society's first president; a more recent analysis of

the Society's "prehistory' is by Weindling (1979). kaudan (1977) and Miller (1986) both analyse the micropolitics behind the Society's early emphasis on fact-gathering and its rejection of theorizing; Moore et al. (1991) describe its museum and early collecting activities. The present paper is, as far as I am aware, the only analysis, albeit a very brief one, of its early publications; my earlier review of the origins of what I termed the 'visual language' of geology (Rudwick 1976) discusses the importance of illustrations, and emphasizes the crucial role of the Society's publications in the establishment of a consensual practice that routinely combined maps, sections and other illustrations. Recent detailed analyses of two major geological controversies serve incidentally to demonstrate the role of the Society's publications in the concrete practice of geologists during the period covered in this paper: they are Secord's account (1986) of the famous arguments over the Cambrian and Silurian systems, and my account (Rudwick 1985) of the establishment of the Devonian.

References LAUDAN, R. 1977. Ideas and organizations in British geology: a case study in institutional history. Isis, 68, 527 538. MILLER, D. P. 1986. Method and the 'micropolitics' of science: the early years of the Geological and Astronomical Societies of London. In: SCHUSTER, J. A. & YEO, R. R. (eds) The politics and rhetoric of scientific method. Reidel. Dordrecht, 227-257. MOORE, D. T., THACKRAY, J. C. & MORGAN, D. L. 1991. A short history of the museum of the Geological Society of London, 1807-19l 1, with a catalogue of the British and Irish accessions, and notes on surviving collections. Bulletin of the British Museum (Natural History), Historical series, 19, 51-160. RUDWlCK, M. J. S. 1963. The foundation of the Geological Society of London: its scheme for cooperative research and its struggle for independence. British Journaljor the History of Science, 1, 325 355. -1976. The emergence of a visual language for geological science 1760-1840. Histoo, of science, 14, 149-195. - - - 1985. The great Devonian controvers:v. the shaping of scientific knowledge among gentlemanly specialists. University of Chicago Press, Chicago. SECORD, J. A. 1986. Controversy in Victorian geology: the Cambrian-Silurian dispute. Princeton University Press, Princeton. WHNDLING, P. J. 1979. Geological controversy and its historiography: the prehistory of the Geological Society of London. In: JORDANOVA, L. J. & PORTER, R. S. (eds) Images of the earth. British Society for the History of Science, Chalfont St Giles, 248-271. WOODWARD, HORACE B. 1907. The history of the Geological Society of London. Geological Society, London.

Received 18 August 1992; accepted 21 August 1992.

THE

QUARTERLY JOURNAL OF

THE

GEOLOGICAL SOCIETY OF LONDON,

EDITED

RY

THE VICE-SECRETARY OF THE GEOLOGICALSOCIETY.

VOLUME THE FIRST.

1845.

LONDON: LONGMAN, BROWN, GREEN, AND LONGMANS, PA.TERNOSTER-ROW.

MDCCCXLV.

From Le Bas, M. J. (ed.), 1995, Milestones in Geology, Geological Society, London, Memoir No. 16, 11-23 First published in Journal of the Geological Society, London, Vol. 150, 1993, pp. 7-19

Uniformitarianism today: plate tectonics is the key to the past BRIAN

F.

WINDLEY

Department o f Geology, The University, Leicester LE1 7RH, UK

Abstract: James Hutton published the first two volumes of The Theory of the Earth in 1795 and the third volume was published posthumously by the Geological Society in 1899. Charles Lyell in his four addresses (1836, 1837, 1850, 1851) to the Society put the uniformitarian paradigm of Hutton (the present is the key to the past) into the perspective of his era. Uniformitarianism today can be expressed in the view that plate tectonics is the key to the past. This paper summarizes key data and ideas which confirm that the plate tectonic paradigm can be applied convincingly back to the beginning of the geological record. In spite of the fact that heat production was greater in the early Precambrian than now, tectonophysical and geochemical processes that produced oceanic and continental rocks since the early Archaean have not been fundamentally different from those that operate today.

'The Present is the key to the Past' was the uniformitarian paradigm of James Hutton (1788). He published the first two volumes of his book Theory of the Earth in 1795. In the conclusion of the second volume he said 'In pursuing this object I am next to examine facts with regards to the mineralogical part of the theory etc', but he never published his intended third volume before his death in 1797. The manuscript was passed via Playfair and Webb Seymour to Leonard Horner who gave it to the Geological Society in 1856, where it was re-discovered in 1895 by F.D. Adams (1938, p. 242). In 1899 the Geological Society published volume 3 of Theory of the Earth as a book edited and indexed by Archibald Geikie. The uniformitarian paradigm was promoted and developed by Charles Lyell in his Principles of Geology (1830 and in 10 subsequent editions). Lyell also presented to the Society two anniversary addresses in 1836 and 1837 and two presidential addresses (1850 on tectonics; 1851 on palaeontology), in which he addressed the question of how far the leading contemporary discoveries had confirmed the uniformitarian argument, namely: 'that the ancient changes of the animate and inanimate world, of which we find memorials in the earth's crust, may be similar both in kind and degree to those which are now in progress'. Also having coined the term 'metamorphism' in the third edition of his book, he took the opportunity to follow the development of 'his metamorphic theory' in 1850. The uniformitarian idea of Hutton and Lyell was an important progenitor of the way of thinking of many generations of geologists. Lyell was not concerned with the building of the history of a continent, so he started with the Recent and worked backwards to 'conduct us gradually from the known to the unknown' (Bailey 1962). Adopting a similar time procedure, the aim of this paper is to summarize key ideas and data that suggest that the plate tectonic paradigm can be applied back to the beginnings of the geological record.

constrained, modern analogues of pre-Mesozoic orogens. One or two decades ago there was not much information about mid-early Precambrian island arcs, accretionary prisms, oceanic plateaus, foreland basins, indentation and escape tectonics, ophiolites with sheeted dykes, suture zones, basic dykes intruded in failed arms and passive margins during continental break-up, the seismic pattern of the crust, and terrane accretion in the lower and upper crust. Kerr (1985) and Kr6ner (1984) concluded that modern-style plate tectonics began at 2 Ga and Meissner (1983) at 1 Ga. However, more recent advances in all the above fields now enable us to postulate reasonably that plate tectonics goes back to 4 Ga.

The Phanerozoic Accretionary and collisional orogens can be considered to be two ends of a spectrum of orogens (Murphy & Nance 1991). The former developed largely by the amalgamation of numerous island arcs, accretionary prisms and ophiolites, and they represent almost total crustal growth of juvenile material; Phanerozoic examples include the Kun Lun orogen in Central Asia ($eng6r & Okurogullari 1991), and incomplete, ongoing, accretionary orogens include the Japanese islands and the Cordillera of western North America. Collisional orogens formed largely by the abutment of one continental block against another, and represent little or no crustal growth; modern examples include the Swiss Alps and the central-eastern Himalayas. The important developments in Phanerozoic geology that are relevant to the uniformitarian argument will be considered in the Precambrian sections below where Phanerozoic analogues can be discussed in their appropriate context.

The Late Proterozoic (1.0-0.6 Ga) Current evidence sugggests that in the last 400 Ma of Proterozoic time, widespread terrane accretion and plate collision led to the formation of a supercontinent, which rifted and broke-up into separate continental blocks before the inception

The plate tectonic uniformitarian model Ideas about the origin of orogens and the continental crust are evolving fast, and this information now provides us with better 11

12

B.F.

~,..

on

~

WINDLEY

$5rvMa;rm~:7~-~/Belomorian

~

S~ avecon, ~ ~orw.eg

SUPERIOR

Penokean Yavapai

:i:i:i:ili:i

~.~Mazat ...)i...~..z.al7.~

1.9-1.8GaJuvenilcrust e 1.8-1.7GaJuvenilcrust e 1.7-1.6GaJuvenilcrust e

of the Phanerozoic. Most prominent are the many orogens grouped within the terms Pan-African, Cadomian and Avalonian. The Pan-African includes the Arabian-Nubian Shield, the Mozambique belt and the Damaran orogen.

A ccretionary orogens The Arabian-Nubian Shield. This is an assemblage of accreted island arcs, ophiolitic belts, and probable microcontinents and oceanic plateaus, and thus provides good evidence of processes of lateral crustal growth and modern-type obduction-accretion tectonics (Kr6ner 1985; Stoeser & Camp 1985; Windley in press a). Disrupted ophiolites occur in linear belts up to 900 km long defining sutures between island arcs and microplates (Kr6ner 1985; Pallister et al. 1988). Some ophiolites contain a complete (Penrose definition) succession (Shanti & Roobol 1979). In Arabia in addition to the island arcs there are remnants of pre-Pan African (i.e. > 1.0 Ga) microcontinents and possibly oceanic plateaus, whereas in Egypt and Sudan the deformed passive continental margin of the Mozambique belt was partly transformed into an active margin along which there are ophiolites and inter-thrust arc volcanic rocks (Kr6ner 1985). In the Shield, there are three ages of island arcs that are very similar to modern arcs formed at sites of plate convergence (Stoeser & Camp 1985). (1) The earliest are chemically immature bimodal suites of low-K tholeiites and sodic dacites/rhyolites depleted in lithophile elements. After deformation, they were intruded by plutons of diorite and trondhjemite at 910 Ma. The lavas have chemical characteristics similar to immature island arcs such as the Tonga-Kermadec and Lesser Antilles arcs.

Fig. 1. Map of Laurentia showing the distribution of Early Proterozoic collisional orogens in the north of the Baltic Shield and three belts of Early Proterozoic accretionary orogens that extend from W. USA to Finland. Modified after Hoffman (1989). MK, Makkovik; KL, Killarney.

(2) Younger lavas are predominantly calc-alkaline and low-K arc tholeiites, andesites, dacites and tufts which were intruded by granitic batholiths dated at 816 Ma and 743 Ma. These are similar to more mature, partly emergent, intraoceanic island arcs in the western Pacific. (3) The youngest voluminous lavas have calc-alkaline or high-K, calc-alkaline compositions with moderately high lithophile element abundances; they are comparable to volcanic arcs as in Central America and Indonesia which are transitional between island arcs and continental margin volcanic arcs.

Collisional orogens The Mozambique belt. This complicated high-grade and highly deformed orogen in East Africa is still understood only in reconnaissance outline. Shackleton (1986) suggested that widespread thrusts, nappes and high-grade metamorphism imply crustal thickening as a result of continent-continent collision tectonics, and Burke & Seng6r (1986) proposed that the belt was the site of a Tibetan-style continental collision. Berhe (1990) described many ophiolitic remnants in deep crustal gneisses. The most detailed, recent work in the Mozambique belt was by Key et al. (1989) in Kenya who concluded from considerable field and geochronological results that the belt represents a deep crustal section through a Pan-African continent-continent collision zone. Orogens surrounding the West African craton. This Precambrian craton is surrounded by Pan-African sutures, arcs and collisional orogens. In Morocco there is a complete ophiolite at Bou Azzer dated at 788 Ma that is overlain by an island arc

U N I F O R M I T A R I A N I S M : PLATE TECTONICS consisting of calc-alkaline lavas and diorites (Bodinier et al. 1984). Many ophiolites, accretionary m61anges and fore-arcs occur as dismembered slivers on a suture between the craton and the island arc (Saquaque et al. 1989). In the Sahara on the east side of the craton in the central Hoggar, there is a collisional orogen that retains evidence of a complete Wilson Cycle spanning the period 900-550 Ma (Caby et al. 1981).

The Mid-Proterozoic (1.6-1.0 Ga) During the mid-Proterozoic a number of orogens formed, the best-known of which is the Grenville in North America (Davidson 1986) that was preceded by its genetically-related period of so-called anorogenic magmatism (Windley 1993). The Grenvillian Wilson Cycle started with prominent 1.481.43 Ga anorogenic magmatism in Canada, especially anorthosites, and in the central/southern USA, mostly rhyolitic ashfall tufts and peraluminous granites (Van Schmus et al. 1987). This magmatism most likely developed on the continental margin of the Grenvillian Ocean; modern analogues border the Atlantic Ocean (Kay et al. 1989; Windley 1993). Closure of the ocean by subduction is indicated by the 1.28-1.25 Ga island arc of the Central Metasedimentary Belt of Ontario, and by an island arc associated with an incomplete ophiolite in Texas which was thrust northwestwards onto a foreland and shelf (Garrison 1981). Collision of the Belt with adjacent continental blocks gave rise to the 1.25-1.22 Ga Elsevirian orogeny and the 1.12-1.03 Ga Ottawan orogeny. The result of these orogenies was the formation of the collisional Grenville orogen, which consists of several major inter-thrusted terranes (Rivers et al. 1989) bounded by sutures that can be recognized on COCORP deep seismic profiles (Culotta et al. 1990), and which shares some fundamental similarities with the Himalayas (Windley 1986). The northwestdirected deformation caused by the terminal Ottawan orogeny fractured the foreland giving rise to the 1.1 Ga Keweenawan rift, that in origin is comparable to the Rhine graben caused by the Tertiary Alpine deformation in Europe.

The Early Proterozoic (2.5-1.6 Ga) No significant orogens formed from 2.5 Ga to 2.1 Ga (a supercontinent?), but from 2.1 Ga to 1.6 Ga many orogens did form of both accretionary and collisional type. A ccretionary

orogens

Early Proterozoic 'growth' orogens, include: 1.7-1.6 Ga Mazatzal (North America) 1.8-1.7 Ga Killarney, Central Plains and Yavapai (all in North America) 1.9-1.8 Ga Svecofennian (Baltic Shield), Ketilidian (Greenland), Makkovik and Penokean (North America) 2.1 Ga Birimian (West Africa). Except for the Birimian, all the above orogens belong to a mega-orogen that extends across what is now North America and Europe and which youngs southwards (Fig. 1). Just key examples will be discussed. The Svecofennian. Extending from Central Sweden and Finland southwards to the Tornquist Line in Poland, this 1200 km wide orogen developed by the accretion of 1.9 Ga island arcs (Park 1991) and accretionary prisms, and by extensive crustal melting in the period 1.8-1.55 Ga. Extensive isotopic data in-

13

dicate that it contains no Archaean material (Huhma 1987; Patchett et al. 1987; Romer 1991). Extending along its northern margin with the KolaKarelian orogen to the north, the Lule&-Kuopio suture zone contains ophiolitic lenses. The 1.96 Ga (U-Pb) Jormua ophiolite with sheeted dykes was thrust about 30 km onto the northern continental margin (Kontinen 1987). Within the Svecofennian orogen there are several island arcs, whose lavas are chemically comparable with modern calc-alkaline arc lavas (Pharaoh & Brewer 1990). U-Pb zircon data indicate that many of the arc lavas were erupted in the short period of 1.92-1.87 Ga contemporaneously with the intrusion of innumerable 1.91-1.86 Ga, subduction-derived granitic plutons (Nurmi & Haapala 1986). Between many of the Svecofennian arcs there are biotitebearing granitic gneisses and schists which, because of chemical similarities, have been widely regarded as metagreywackes and metapelites, and which were most likely derived from accretionary prisms. Thrusting and folding was associated with high amphibolite facies metamorphism that locally reached granulite grade. Crustal thickening led to the formation of three types of crustal melt granites, the last of which were 1.7-1.55 Ga rapakivi granites and coeval gabbros, anorthosites and basic dykes (Haapala & R~im6 1990). These formed as a result of the internal slow heating of the thickened crust, its final extension and collapse, and thus to decompression melting of the mantle and melting of depleted granulitic lower crust (Windley in press a). The Ketilidian. This orogen in South Greenland (Allaart 1976) is an incomplete segment of an Early Proterozoic accretionary orogen which contains an Andean-type batholith (Fig. 2; Windley 1991 & references therein). A northern foreland of Archaean gneisses is overlain unconformably by a shelf-foredeep succession deposited by turbidity currents into basins on the deepening shelf, a 30 m thick sulphide-facies iron formation (chert-pyrite-shale) similar to that which commonly occurs on the outer ramp of Early Proterozoic foredeeps, and tholeiitic pillow lavas and basicfelsic pyroclastics, like those in the axial zones of other Early Proterozoic foredeeps (Hoffman 1987). The above succession has been thrust northwards over the foreland and back-thrust near the suture, where it and the basement thrusted gneisses are intruded by several 1.775-1.675 Ga granites that contain appreciable crustal-melt components (Kalsbeek & Taylor 1985). These relations are comparable to those that occur in the deformed foreland of modern collisional orogens such as the Himalayas. The Kobberminebugt suture is a 15 km wide vertical shear zone that contains relict greenschist-grade pillow lavas and gabbros, copper and gold mineralization, and late 100 m thick mylonite zones. The Julianehaab batholith is a 80-100 km wide Andean-type tonalitic-granodioritic batholith that contains relicts of pillow lavas, pyroclastic rocks and extensive noritic gabbros (Allaart 1976) that probably belong to an early island arc into which the major calc-alkaline batholith was intruded (Windley 1991). The arc rocks are similar to those in the Kohistan arc in the Himalayas of North Pakistan, the lower part (magma chamber) of which is occupied by the Chilas complex of noritic gabbros (Khan et al. 1989). The southernmost part of the Ketilidian orogen consists largely of metamorphosed, accretionary prism-type, supracrustal rocks that were deformed in three sub-horizontal thrust nappes and metamorphosed at 1.8 Ga. The thrust slab was

14

B.F.

KETILIDIAN OROGEN

~::i:':':':':':':':':"

! 62 °

WINDLEY

'

'

":i~iiiiiii i i i i i i i i i i:' ili!iii "

'; :':':'~

'

o~e~'~'~

+ ~!i!iiiiiiiiii~.~.-'.-.:-.--::. :, Shelf / ..:::.:::::....::::::::::::::::, Archaean ~ : : i - . " \ ..:.:-:- ' basement ~ ' . ' . ' . - : ' \

. ek . ~"

:.i~

~,+~_o,~ .,~

.i~!~

~

3~:.:i~.:.:.:.:.:.:.:.~/

~#

:':':'. ', ,~,

!~'~:!......?.~ "'":-! ".

#\s

~\

-...

, , f i e i~c" -, ----' ,

k)l~ s ~"

,," ,- ,- -"+ ++ + ,"+ -;-,-

5

AnOU . . . . . battloinH

" ' : t : ~ .:.:-:~.I,;~: :-~.'~:i~-*.':.:~ ~ ~ : ,.-..-..2200Ma) were attributed to the presence of excess At, whereas younger ages (with the exception of the sample cited above) were considered to be due to variable Ar loss during Laxfordian metamorphism. In contrast to Rb-Sr mineral and K-At studies, which may be reset by later thermal events, Rb-Sr whole-rock dating may in certain circumstances see through these later events to yield information regarding the igneous crystallization of protoliths. Using this philosophy, Chapman (1979) analysed three dolerite dykes--two from Scourie and one from Kylesku--using Rb-Sr whole-rock techniques. Owing to the low Rb/Sr ratios and to the lack of fractionation of Rb/Sr within each individual dyke the errors on the ages were high, although it was stated that the ages and the initial ~7Sr/~'Sr ratios for the individual dykes were not significantly different. However, each dyke had a distinct (if restricted) range of Rb/Sr, and so, by combining the data for the three dykes, a greater spread of Rb/Sr was achieved and hence a reduction in the calculated error. The combined age of 2390 + 20 Ma was thought to represent the emplacement of the Scourie dykes; this date also placed a minimum age on the Inverian metamorphism. No evidence was presented to substantiate the c. 1950 Ma age of Evans & Tarney (1964). In order to obtain more precise ages from individual dykes Cohen et al. (1988c) undertook a Sm-Nd mineral study of Scourie dykes in which an apparently primary igneous mineralogy was still preserved: the detailed results of this study were given in Waters et al. (1990). Olivine gabbro and quartz dolerite dykes gave Sm-Nd mineralwhole-rock ages of 2015 + 42 to 2102 + 77 Ma and 1758 + 7 and 1982 + 44 Ma respectively (Table 1). In the case of the Graveyard dyke a feldspar datum lay below the regression line and was therefore excluded from the calculation. A Rb-Sr mineral-whole-rock age of 1978 + 13 Ma was also obtained from the Rhegreanoch dyke, within error of the Sm-Nd age. Rb-Sr mineral-whole-rock ages for the other dykes generally gave younger ages than the Sm-Nd data. An exception was the Graveyard dyke which gave a Rb-Sr amphibole-feldspar age of 2027-1-11 Ma compared to the Sm-Nd age of 1758-t-7 Ma. This discrepancy was explained by invoking recrystallization and growth of garnet and ilmenite during Laxfordian metamorphism; amphibole and feldspar were deemed to have formed shortly after dyke emplacement, and to have remained closed systems throughout the later event. Clearly a primary igneous mineralogy had not remained undisturbed in this case. No evidence was found for dyke emplacement at 2400Ma, which led Cohen et al. (1988c) to state that the Rb-Sr age of Chapman (1979) was 'erroneous', and probably reflected contamination with continental crust. Interestingly, the Graveyard dyke at Scourie was also one of those used by Chapman in the Rb-Sr study. Whereas the studies considered so far have either been prone to post-emplacement thermal disturbance of isotope systematics, or have had to make assumptions about the geochronological significance of isotopic equilibration between minerals and/or whole-rocks, U-Pb geochronology on phases such as zircon, baddeleyite and titanite, potentially represent self-contained systems with high

42

G. R O G E R S & R. J. P A N K H U R S T Table 1. Geochronological data for Scourie dykes (Waters et al. 1990)

Dyke

Rock type

Phases

Badnaban

Olivine gabbro

Rhegreanoch

Olivine gabbro

Loch Torr an Lochain Graveyard

Olivine gabbro

Poll Eorna

Quartz dolerite

Quartz dolerite

Isotope System Age (Ma)

Fsp-cpx-amph-WR Bi-WR Fsp-cpx-amph-WR Fsp-cpx-amph-WR Fsp-WR

Sm-Nd Rb-Sr Sm-Nd Rb-Sr Sm-Nd

2031 + 62 1714 + 8 2015 + 42 1978 + 13 2102 + 77

Amph-gt-ilm-WR Amph-fsp IIm-WR Fsp-cpx-ilm-WR IIm-WR

Sm-Nd Rb-Sr Rb-Sr Sm-Nd Rb-Sr

1758 + 7 2027 + 11 1738 + 11 1982 + 44 1733 + 7

Fsp, feldspar; Cpx, clinopyroxene; Amph, amphibole; Bi, biotite; Gt, garnet; Ilm, ilmenite; WR, whole-rock

data of Waters et al. (1990) discussed above. The c. 2200 Ma K-Ar whole-rock ages of Evans & Tarney (1964) may therefore represent prolonged cooling following dyke intrusion at deep crustal levels, whereas the younger K-Ar whole-rock age of 1950 Ma may indicate more rapid cooling of later dykes intruded at higher crustal levels. In conclusion, if the Scourie dykes do, in fact, represent a unique structural time marker, then the time interval over which there has to be structural quiescence is about 400 Ma. The final recognition of such a long interval led Park (1991) to state that 'in view of the geochronology, it is likely that such tectonic activity did take place [during the 400 Ma] and may ultimately be proved by more adequate dating.' The studies of Cohen et al. (1988c), Waters et al. (1990) and Chapman (1979) indicate some of the problems of geochronological techniques which require an assumption of isotopic equilibration between coexisting phases or wholerocks, and highlight the great care required in determining which phases are primary and which may be either secondary or have recrystallized. If incorrect textural observations are made then isolated results may yield erroneous interpretations. It is clear, however, that if suitable material can be found and careful petrographic analysis undertaken, then both detailed Sm-Nd and U-Pb mineral studies may yield meaningful emplacement ages for mafic dyke suites. The significance of such ages in the structural evolution of the Lewisian complex requires a

closure temperatures to parent-daughter migration. Although material for analysis has to involve the selection of high-integrity grains in order to avoid the effects of low-temperature Pb loss, and there is always the possibility of analysing grains which have experienced multiple episodes of growth (e.g. Pidgeon & Aftalion 1978), the presence of two internal U-Pb radiometric clocks enables departure from simple closed-system behaviour to be generally readily identified. The development of improved techniques for the production of accurate and precise U-Pb data on small samples (e.g. Krogh 1982b; Parrish & Krogh 1987), and the discovery that mafic dykes may contain trace amounts of zircon and/or baddeleyite (Heaman et al. 1986; Krogh et al. 1987) enabled Heaman & Tarney (1989) to obtain ages on three individual Scourie dykes. A bronzite picrite from Beannach and an olivine gabbro from Strathan +3 gave baddeleyite ages of 2418_+TMa and 1992_~Ma respectively (Fig. 3). The latter was interpreted as the time of dyke emplacement whereas the former, owing to the slight discordancy of the data, was thought to be the minimum age of intrusion with the true age being a little older. Zircons from a norite from Badcall Bay yielded discordant data but their 2°7pb/2~pb ages of 2166-2179 Ma were considered to represent minimum estimates for the age of the dyke. The U-Pb data provide clear evidence for two phases of dyke emplacement, the first at c. 2418 Ma and the second at 1992 Ma, this latter date being consistent with the

a

0.37

2,420

b 2,000

0.45

j,

co

¢o o~

/

/

:~ 0 " 3 6 co

Fig. 3. U-Pb concordia diagrams for

tn IX

J~ 0 . 4 3 IX co 0

1,992+3/-2Myr I

0"35

-~1,071Myr 0"41 8"6

818

I 9-0

91"2

I 9"4

207pb[235 u

91"6

91-8

034 58

5.9

6.0

2o7pb/235

6 1

6.2'

two Scourie dykes, confirming the episodic emplacement of the suite. (a) Beannach dyke. (b) Strathan dyke. Error ellipses are drawn at the 2a level. (Figures reprinted with permission from Nature (Heaman & Tarney 1989): copyright (1989) Macmillan Magazines Limited, and from L.M. Heaman).

U N R A V E L L I N G DATES more reliable knowledge of the timing of the igneous and metamorphic events within the region, without which the structural debate may continue in sterile argument.

The Moinian Supergroup The age, metamorphic history and wider affinities of the Moinian Supergroup have long been contentious issues. Following the great debates of the last century (see Oldroyd 1990 for a review) two main hypotheses emerged regarding the evolution of the area: (1) that the Moinian and Torridonian sediments were of the same age and that orogenesis was entirely Caledonian (Peach in Peach & Horne 1930); (2) that the metamorphism was, at least in part, pre-Cambrian in age (Horne in Peach & H o r n e 1930). With the advent of radiometric dating techniques Giletti et al. (1961) set out to address this problem through Rb-Sr and K-Ar dating of micas from Moinian metasediments and pegmatites. Biotites from schists covering much of the strike length of the Moine gave an average age of 420 + 15 Ma which Giletti et al. interpreted as dating widespread metamorphism of the region. The age also placed a maximum age on the movement of the post-metamorphic Moine thrust. More interesting, however, w e r e Rb-Sr muscovite ages of 690-750Ma from pegmatites from Knoydart and Sgurr Breac. Giletti et al. put forward several hypotheses in which the dates might be a function of later isotopic disturbance, yet all reasonable solutions required the pegmatites to be Precambrian in age. This led Giletti et al. to conclude 'that the Moine sediments, at least in the Knoydart-Morar area are older than 740 m.y.' and that the pegmatites were formed 'possibly at the time of the first, or an early, metamorphism of the Moine sediments', thus supporting Horne's view of a Precambrian metamorphism. The Moinian Supergroup has been shown to have experienced polyphase deformation and two main metamorphic events (e.g. Ramsay 1963; Powell 1974). The timing of the metamorphism and associated deformational episodes are largely constrained by dating intrusive bodies which both pre- and post-date the events. The initial results of Giletti et al. stimulated further research into the geographical extent of the Precambrian event and the timing of the early and late metamorphisms. The results of this work, inextricably linked with the highly contentious issue of the structural interpretation of the Moine (see Harris & Johnson (1991) and references therein), have prompted much heated debate and several major issues are still unresolved. Various studies have shown that the late event in the Moine is dated by: (1) a bulk fraction U-Pb zircon age of 456 + 5 Ma for the Glen Dessarry syenite (van Breemen et al. 1979) which was deformed during this event (Roberts et al. 1984); (2) a concordant U-Pb monazite age of 450 + 10 Ma and Rb-Sr muscovite ages of 438-450 Ma for late pegmatites (van Breemen et al. 1974); (3) concordant U-Pb monazite ages of 455 + 3 Ma from the Glenfinnan area (Aftalion & van Breemen 1980); (4) Rb-Sr muscovite ages from semi-pelitic units of 427 + 8 to 462 + 10Ma (van Breemen et al. 1978). The high closure temperature of the U-Pb monazite system (c. 725 °C; Parrish 1990) suggests that the peak of Caledonian metamorphism was c. 455 Ma, with pegmatite injection and cooling below 500 °C following shortly thereafter. In a classic piece of work Long (1964) showed that the Carn Chuinneag granite, which was intruded between the

l

~k 'Sr

43

CARN CHUlNNEAG GRANITE

/~

°Sr

1 10 I . . . . . . . . . . . . .

/

o,t

0-8J ~

07

/

/ f \ ~ ~ M

/

1/

~ Mine~l,Isochron 403±5Ma

~Y /

WR~qbm="Isochron

( ./ "WR 10

20

30

750

87Rbl~ 86Sr I~ 1000

Fig. 4. Rb-Sr isochron diagram for whole-rocks and minerals from the Carn Chuinneag granite (Long 1964) illustrating isotopic homogenization of the minerals during regional metamorphism while the whole-rocks remained closed systems. two metamorphic episodes (Shepherd 1973; Wilson & Shepherd 1979), gave a Rb-Sr whole-rock age of 5 4 8 + 1 0 M a , but that the minerals from one of these whole-rocks yielded an age of 403 + 5 Ma (Fig. 4). These results were taken to indicate that the granite was intruded at 548 Ma, but that during the later metamorphism, whereas the whole-rocks remained closed systems, the minerals within the whole-rocks isotopically equilibrated to give the age of the reheating event. The whole-rock data clearly indicated that the earlier metamorphism had to be pre-550Ma. Pidgeon & Johnson (1974) performed bulk fraction U-Pb zircon analyses on three facies of the pluton. Data from the Inchbae and Lochan a' Chairn phases defined a reverse discordia with lower intercept ages of 563 + 10 and 502 Ma respectively, whereas the riebeckite gneiss gave a simple discordia line but also with an upper intercept of 522 -1-20 Ma. All these phases were considered to have been emplaced at c. 560 Ma, broadly consistent with the Rb-Sr date. A Rb-Sr whole-rock age of 416 + 15 Ma from the riebeckite gneiss showed that although the Inchbae and Lochan a' Chairn facies had remained closed to Rb-Sr migration on the scale of the whole-rocks during the Caledonian, the riebeckite gneiss had been open. Consequently, whereas the studies at Carn Chuinneag indicated the potential benefits of combined mineral and whole-rock studies, they also sounded a severe note of caution regarding whole-rock methods in polymetamorphic terrains. Further work on the older pegmatites of Giletti et al. has produced Rb-Sr muscovite ages from 647 + 20 to 776 + 15 Ma (Long & Lambert 1963; van Breemen et al. 1974, 1978; Powell et al. 1983; Piasecki & van Breemen 1983). In a detailed study of the C a m Gorm locality van Breemen et al. (1974) found that there was probably only minor disturbance of the Rb-Sr muscovite systematics during the Caledonian metamorphism. In contrast K-Ar muscovite dates from the Ardnish pegmatite (Powell et al. 1983) yield younger ages of 498-410 Ma, indicating partial or total Ar loss during the Caledonian, in accord with other KoAr determinations from the NW Highlands (Giletti et al. 1961; Miller & Brown 1965; Brown et al. 1965a, 1965b). In an attempt to address the possibility that the Rb-Sr muscovite ages might reflect partial Caledonian overprinting of Grenvillian pegmatites van Breemen et al. (1974, 1978)

44

G. R O G E R S & R. J. P A N K H U R S T analysed bulk fraction zircons from an MP1 pegmatite lit from the gneiss which gave a lower intercept age of 556 -t- 8 Ma. Yet again there was no hint of a Grenvillian age in the data (Fig. 5). Nonetheless, Aftalion & van Breemen, largely on the basis of the 1028Ma age, constructed elaborate models of multi-stage Pb loss to account for the observed zircon discordance in terms of a Grenvillian crystallization age for the gneiss. Sm-Nd mineral dating has also thrown a spanner into the works regarding the presence of a Grenvillian event in the Moine. Sanders et al. (1984) obtained Sm-Nd g a r n e t clinopyroxene-whole-rock ages of 1082-1-24 and 1010 + 13Ma for eclogites from the Glenelg inlier which are considered to date the eclogite facies metamorphism. As the Morar group sits unconformably on the Glenelg inlier (Clough in Peach et al. 1910; Ramsay 1958) and is only at low metamorphic grade (e.g. Fettes et al. 1985) it follows that the Morar group must have been deposited after c. 1000Ma. Given that there is stratigraphic continuity throughout the Moinian succession, and that the early Moinian metamorphism is considered to have occurred at pressures of about 6.5 kbar (Fettes et al. 1985), there is considerable difficulty in reconciling the Sm-Nd data with the Rb-Sr data from the Ardgour gneiss (1028 + 46 Ma). In summary, whereas the ages of the Caledonian metamorphism and the Knoydartian pegmatite emplacement are now fairly well constrained, the significance of the Knoydartian event and the timing of the Precambrian metamorphism are still unclear.

performed U-Pb analyses on monazite and zircon - which have higher closure temperatures - from two localities. Monazite analyses were concordant at 780 + 10 Ma whereas discordant zircon data gave ages of 7 4 0 + 3 0 and 815 + 30Ma. Given that these pegmatites were emplaced into rocks of low metamorphic grade, van Breemen et al. (1978) concluded that the ages must represent the time of pegmatite intrusion rather than reflecting slow cooling. Furthermore, the field evidence of pegmatite concordance with the main foliation in the host lithologies suggested that this was also the time of peak metamorphism (termed the 'Knoydartian' by Bowes (1968) and the 'Morarian' by Lambert (1969)). In this interpretation they agreed with the views of Giletti et al. (1961), Long & Lambert (1963), Bowes (1968) and Lambert et al. (1979) in ascribing the pegmatites to orogenic activity. Evidence against this interpretation was provided by Powell et al. (1983) who showed that the Ardnish pegmatite, which they had dated using Rb-Sr on muscovites at 776 + 15 Ma, post-dated the early folding and metamorphism of the Moinian metasediments and was deformed during a later event. The timing of the early metamorphism, however, remained unclear, although a Grenvillian age was plausible in the light of other data (see below). Perhaps the most controversial aspect of Moinian geochronology has centred on the age of the Ardgour gneiss. This was originally held to have been produced by in situ metasomatism during the peak of metamorphism (e.g. Dalziel 1966), but has since been shown to have been intruded during the early metamorphism (Barr et al. 1985). Brook et al. (1976), using large samples, produced a Rb-Sr whole-rock age of 1028 + 46 Ma which they interpreted as indicating that the early metamorphism was Grenvillian in age. Pidgeon & Aftalion (1978) investigated the gneiss using bulk fraction U-Pb zircon techniques. The data defined a reverse discordia giving a lower intercept age of 574 + 30 Ma and an upper intercept age of 1556 ' ~ Ma, which they had great difficulty in reconciling with the 1028 Ma Rb-Sr age of Brook et al. (1976). Aftalion & van Breemen (1980) also --

oso

I

I

Torridonian The dating of unfossiliferous sedimentary successions which do not contain volcanic horizons or igneous intrusions presents a considerable geochronological challenge as such combinations militate against dating strategies such as have been used in the Moine and Dalradian. The Torridonian sandstones are one such succession. Although samples from the Torridonian sandstones

~~

T

i

16

li re 206pb' 238 0

-~ Ojo~" /

o 20

1200Ma'x

P ~tlgl~ 1 lOOMax / ,~55.53uNM R __ _ _ >d" / ~ "106+841JNM GRANITI,,. b Sr AGE,o,~,..,oo l OOO~ ~ ~ / / / .•. .• . .' t~ . . . .

0 151

900Ma . ~ ~ ' ~

1200Ma/"~::~ ~: E M P L A C E M E N T T1 ~

! r8 L

[

......

Calculated

C | GRANITE-

o 2t

ooM'°°",~.o#%~ Re,o,

o ~1

61+450NM

~..

600M~'

o lo i

"84.53pM1

~'~ z....

1 0 0 0 M ~ " -"xZ ~-~RC70S ~1161JNM

r 2 ~ / - GLEN DESSARY SYENITE

k '4,°°7

a. MONAZlTE AGE Or

~/~ ....

R C 9 0 9 and RC 1 5 2 4

0

10

20 2Lo r5 = 2 ° 7 p b

Z i r c o n s i z e fractions: RC1524 Paragneiss Glenfinnan [J R C 7 0 5 G r a n i t i c G n e i s s :\ R C 9 0 9 Lit in Granitic G n e i s s

30

/

40 3'0

235 U

r5

50

6 ~ 40 - - - -

Fig. 5. U-Pb concordia diagram for samples from Glenfinnan. Inset shows modelled zircon evolution assuming Pb loss during Caledonian and Grenvillian times. See Aftalion & van Breemen (1980) for detailed discussion. (Figure reproduced from Aftalion & van Breemen 1980 with permission from Springer-Verlag and M. Aftalion).

UNRAVELLING DATES which unconformably overlie the Lewisian complex were not analysed by Giletti et al. (1961), the data they obtained regarding the age of the Laxfordian led them to conclude that the Torridonian must be younger than c. 1600 Ma, or possibly 1160 Ma based on an Rb-Sr biotite age from a gneiss from Loch Torridon whose interpretation was problematical. If the Moine and Torridonian sediments were lateral equivalents (e.g. Sutton 1963) then the latter must be older than the c. 740 Ma pegmatites in the Moine. However, given that the Torridonian is unconformably overlain by Cambrian sediments, Giletti et a l . ' s data permitted the Torridonian to be younger than the Moine and to have been derived from it. In order to assess the age and provenance of the Torridonian, Moorbath et al. (1967) analysed both detrital grains and individual pebbles. In the Applecross Formation muscovites from schistose pebbles gave K-Ar ages of 1659-1802Ma whereas Rb-Sr K-feldspar dates from microcline and quartz porphyry pebbles ranged from 1320-1637 Ma. Detrital muscovites from the Diabaig Formation gave a restricted range of Rb-Sr and K-Ar ages from l160-1190Ma. From these results Moorbath et al. concluded that the schistose pebbles were not derived from the Moine metasediments as these had been metamorphosed at least 740 Ma ago, and that some of the pebbles had a provenance in rocks which had been last metamorphosed c. 1700Ma, which they equated with the Laxfordian complex. The ages from the detrital micas indicated that the Torridonian must be younger than about 1190 Ma. In an attempt to define the age of the Torridonian more closely Moorbath (1969) used Rb-Sr whole-rock analyses of shales to construct isochrons for the Stoer and Applecross Formations, which gave ages of 968 + 24 and 788 + 17 Ma respectively (Fig. 6). Moorbath argued that these reflected isotopic homogenization during diagenesis which would have closely followed deposition. This interpretation was questioned by Smith et al. (1983) who suggested, on the basis of palaeomagnetic results which indicated that the Stoer and Applecross Formations were c. 1100 and 1040 Ma respectively, that diagenesis could have occurred significantly later than deposition. Allen et al. (1974) presented 4°Ar-39Ar data on exotic quartz tourmaline pebbles from the Applecross Formation

0-85

I

•

45

which gave ages from 802-2926 Ma. Taking the results at face value and using known palaeocurrent indicators they concluded that the data were consistent with a source in the Preketilidian and Ketilidian belts of Greenland and the Grenville of Labrador. However, it is now recognized that tourmalines are prone to excess Ar and give anomalously old ages (Damon & Kulp 1958); consequently no reliable information can be gleaned from the data. Rb-Sr dating of biotites has provided evidence of uplift and cooling of the Lewisian complex of northern Harris and Lewis at c. 1100 Ma (Cliff & Rex 1989). By comparing these data with Rb-Sr and K-Ar ages for detrital muscovites from the Torridonian and a K-Ar biotite age from a gneiss boulder in a Lower Torridonian tilloid (Moorbath et al. 1967), Cliff & Rex suggested that the northern part of the Outer Hebrides could have acted as a source for some of the Torridonian sediment. As zircon has a high closure temperature for U-Pb compared with either K-Ar or Rb-Sr in micas, U-Pb dating of individual zircon grains can potentially yield information regarding the ages of formation of the sources of sedimentary piles rather than later uplift and/or metamorphic ages. Moreover, because they are generally mechanically and chemically resistant they are most likely to survive sedimentary processes and so provide a full coverage of the provenance spectrum, provided, of course, that all the source rocks contain zircon. Rogers et al. (1990) analysed single detrital zircon grains from the Applecross Formation. The analyses were generally less than 1% discordant and fell into three groups: 1088-1193Ma; 1625-1662Ma; 26282857 Ma. These ages provide a maximum age for the the Applecross Formation of 1088 Ma consistent with both the palaeomagnetic and earlier geochronological results. Rogers et al. concluded that the data were more consistent with a provenance in Labrador than E Greenland. Given the lack of published evidence for zircons of c. 1150 and c. i650 Ma from the Outer Hebrides, a source in such an area also seems unlikely. Thirty years on, and despite the increasing sophistication of geochronological approaches, the problems of the age and affinities of the Torridonian are still unresolved. Stratigraphical correlation with the Moine is still equivocal and would appear to require more detailed knowledge about the age of the Torridonian sediments and the tectono-

a

26-5

Sr87 Sr 88

/

0.85

~ . , 26- 7 /

2~6_4 c 26-3

0-80

• Sr 87 _ Sr86

12a~/ /-i A

b /

/

0-80

26_//~8'26-6

Fig. 6.

Rb-Sr whole-rock isochrons for siltstones and shales from (a) the Stoer and (b) the Applecross Formations of the Torridonian sandstone. (Figure adapted from Moorbath 1969 with permission from Scottish Journal of Geology).

0"75

6_1o ~/~26_12 968-+24Ma ~ 26-11 26-9

0 -7(2 0

-~ 0 . 7 0 8 6 + 0 . 0 0 1 6

Rb87 I

12e~'~12d 12c~12h 0-75 - 12fQ, ,.J U'i29 12b /3A '~4A -,~0.7215+0,0014 0.7C I 0 5

788+_1 7Ma Rb87 Srr "~-6II~ I 10

46

G. R O G E R S & R. J. P A N K H U R S T

thermal history of the Moine. Nonetheless, significant progress has been made regarding the provenance of the sediments which points to areas of fruitful research in the future.

The Dalradian Supergroup Nowadays we look to geochronology to determine both the time of metamorphic recrystallization within a complex belt and its post-orogenic uplift history. These two aspects were already inherent in the treatment of the Dalradian metamorphic complex presented by Giletti et al.

following a brief early Ordovician climactic episode of metamorphism. They drew age contours ('chrontours' or 'thermochrons'), taken to represent lines of synchronous cooling through the respective closure temperatures, and proposed early uplift of the low-grade rocks against the bounding faults and later uplift of the hotter, high-grade rocks in the central areas (Fig. 7). Their model also related the cycle of deposition, deformation, metamorphism, uplift and erosion to the still new ideas for subduction and closure of the Iapetus Ocean to the south, and they further suggested that the post-orogenic Newer Granites (which had mostly given ages of 420-370Ma) were triggered by pressure-release during the uplift phase.

M i n e r a l ages: cooling histories Seven Rb-Sr mica model ages from the Grampian Highlands and three from the Connemara schists of western Ireland were reported by Giletti et al. By subsequent standards this is a small body of data, but they were enough to reveal an unexpected problem. Whereas their data for the Moinian Supergroup (see above) had suggested a Late Silurian/Early Devonian metamorphism at about 420 Ma, consistent with the stratigraphically-controlled age of Caledonian deformation in the Southern Uplands, England and Wales, most of the Dalradian and Connemara schist ages were significantly older at about 475Ma. Although this supported a correlation of the Dalradian and Connemara schists, it also indicated that both had experienced an Early-MidOrdovician metamorphism. Additionally, three of the Dalradian biotite ages were younger and within error of the Moine schist results, suggesting that the later Moinian event had also affected the Dalradian rocks; consequently, 475 Ma could be regarded only as a minimum age for the early metamorphism of the Dalradian. The idea of mineral phases having different 'blocking temperatures', with muscovite being more resistant to overprinting than biotite, was already emerging at the time of Giletti et al.'s paper, and in their Discussion reply to R. Rutland, the authors cautioned that the younger age of 420Ma might not represent the real time of prograde metamorphism but merely that of post-metamorphic cooling. In this they admitted a fundamental property of mineral geochronology, but also predicted how this might be turned to advantage in the interpretation of the thermal history of metamorphic belts. The following years saw the acquisition of a great deal more mineral data in an attempt to resolve this issue (K-Ar: Miller & Brown 1965; Brown et al. 1965a; Harper 1967: Rb-Sr: Bell 1968). These showed a wide spread of ages, from 500 to less than 400Ma in both the Moine and Dalradian. This was still generally taken to represent the effects of overprinting of a late metamorphism on an early one, although authors such as Brown et al. (1965a) carefully discussed the alternative possibility of slow cooling. Particularly influential was the idea of Harper (1967) that low-grade rocks that had only reached their maximum temperature during a short period prior to uplift, should give a better estimate for the metamorphic age than high-grade rocks; he obtained K-Ar whole-rock ages of c. 490-520 Ma from the southern margin of the belt, close to the Highland Boundary fault. The case for the cooling hypothesis was advanced by Dewey & Pankhurst (1970), who interpreted the entire body of data for both the Moine and Dalradian rocks as representing uplift and cooling

"~

IIW,'

• ) .'J / o / ~,~ f ~ ' // . ,'~'" / " \~_.~" ~// L//'/ ~ / / //-~--~z~

/

.} /;11 zx/ ,"),11/~1 /;1~

.)//,'I,'/J" Ii ~

I//

IIi

j / / t " ~ X-~',\ \ ~7. " ~-~'o.. ~. fl}" ~Oo~ - ~ . \ \ ,

/

>soo 490-499 480-489 470-479 460-469 450-459 440-449

° Q 0 • [3 .

4. 2o0-- .49 29 410-419 400-409 > 400

A • v v

%

2 ff

./J~

K-,,.o.m.,.

AA

J

I

40 miles

I

Fig. 7. K-Ar muscovite age-contours ('chrontours' or 'thermochrons') for the Scottish Highlands, as presented by Dewey & Pankhurst (1970). These were interpreted as representing lines of synchronous uplift and cooling through the blocking temperature (c. 350 °C) following a relatively brief climactic episode of deformation and metamorphism, 480-500 Ma ago. Early uplift occurred along the bounding faults in the marginal parts of the orogen (Highland Boundary fault and Moine thrust), whereas the central high-grade areas that had been most deeply buried did not finally cool to 350 °C until 80-100 Ma later. Subsequent work has suggested a more complex pattern of local uplift events, and attainment of peak metamorphism as late as 455 Ma in the NW Highlands (Figure adaptation reproduced by permission of the Royal Society of Edinburgh from Dewey & Pankhurst 1970).

UNRAVELLING DATES This was a very 'broad-brush' approach to the problem. Proper control of time-temperature trajectories for metamorphism requires a great deal of high-quality data. The principles usually applied have been developed from the pioneering work in the Swiss Alps (e.g. J/iger 1979). K-Ar hornblende and Rb-Sr muscovite are generally thought to have relatively high closure temperatures of about 550 and 500°C respectively, K-Ar muscovite about 350°C and biotite (Rb-Sr and K-Ar) about 300°C. The highest temperatures of metamorphism require independent control, possibly using Rb-Sr whole-rock or U-Pb zircon geochronology, whereas fission track data are necessary to date cooling down to 100°C. These methods must be applied to a volume of rock small enough to have had a uniform cooling path, and with sufficient precision to distinguish the separate stages: even then, no direct time-temperature information can be obtained for the prograde heating path prior to the maximum temperature. It is rare that the full set of such measurements is available; the only case in Scotland is that of the Glen Dessarry syenite intruded into the Moinian Supergroup (see above). The data for Glen Dessarry, collated and interpreted by Cliff (1985), are consistent with a fairly simple pattern, showing an initial cooling rate of 30 °C/Ma, falling to 10°C/Ma over the first 40 Ma, and followed by dramatically slower cooling from about 300 °C (Fig. 8). This last stage is, however, governed by an apatite fission track age which may reflect the effect of later re-heating rather than regional post-Caledonian cooling. In general, closure temperatures must depend on a variety of factors, such as mineral composition, grain-size, cooling rate and fluid interactions. Furthermore, Giletti (1991) has claimed that the variations in diffusive exchange rates for Sr are such as to cause major errors in the estimated closure temperature for Rb-Sr in biotite. Various closure temperatures have been proposed for Sm-Nd garnet systems (e.g.c. 500-700 °C: Humphries & Cliff 1982; 900 °C: Cohen et al. 1988b). Mezger et al. (1992)

I

I

I

Zircon

1000

0

0

800 Ik,.

=1

600 E

400

1--

l~

hene

hHornblende ~ Muscovite \ Muscovite Apatite Biotite - - ~

20O

Ap-f-t

(,)

i

~

400

I

I 300

I

Age (Ma) Fig. 8. Cooling pattern for the Glen Dessarry syenite showing the relationship between blocking temperature for each phase and the age determined for that phase. U-Pb ages shown by circles; Rb-Sr ages, filled stars; K-Ar ages, squares; Fission track age, open star. Fission track datum represents mean fission track age north of the Great Glen fault (Hurford 1977). All other data from van Breemen et al. (1979). (Figure reproduced from Cliff 1985 with permission from R.A. Cliff).

47

have recently argued for a closure temperature of c. 600 °C in metamorphic garnets 2716Ma and 2710MR and the other at 2480-2490 MR. Such a late high-grade event would help to explain

the Sm-Nd mineral ages of 2490Ma of Humphries & Cliff (1982) which would thus reflect more rapid post-metamorphic cooling rather than slow cooling over 150-200 MR. By using U-Pb techniques on other minerals Corfu et al. (in press) were able to more fully document the thermal history of the area. Titanites from a metasedimentary layer at Scourie More were shown to have grown during the early Inverian (2480Ma) and to have uffered Pb loss and probable regrowth during Laxfordian events (1750Ma). Titanites from other gneisses indicated Laxfordian growth, but also isotopic disturbance and new growth at c. 1670MR. This late event was also reflected in rutile growth (HeRman & Tarney 1989; Corfu et al. in press). Within the Laxford Front zone, but within gneisses of the Laxfordian complex, Corfu et al. (in press) found a completely different pattern of zircon discordance to that of the Scourian complex. One near-concordant zircon fragment gave a 2°7pbF°7pb age of 2882 MR, but despite detailed picking and air abrasion, the other data were all strongly discordant. The intense early Inverian event of the central region was not evident in this sample. These new detailed U-Pb studies have revealed the importance and varied nature of early Inverian events (c. 2480Ma), have highlighted the occurrence of a late event (c. 1670MR), and have indicated contrasting thermal histories between the Scourian and Laxfordiancomplexes. However, the complexity and intensity of processes occurring repeatedly in such a high-grade terrain have obscured the timing of the earlier events, such that definitive ages for protolith formation and early Badcallian metamorphism are still uncertain. Despite the application of increasingly sophisticated techniques many fundamental questions are still unanswered.

Additional references CORFU,F., HEAMAN,L.M. & ROGERS,G. 1994. Polymetamorphic evolution of the Lewisian complex, NW Scotland, as recorded by U-Pb isotopic compositions of zircon, titanite and rutile. Contributions to Mineralogy and Petrology, 117, 215-228. EVANS, C.R. & LAMBERT, R.ST.J. 1974. The Lcwisian of Lochinver, Sutherland; the type area for the lnverian metamorphism. Journal of the Geological Society, London, 130, 125-150.

Added November 1994.

From QJGS, 1 17, 233-234. A GEOCHRONOLOGICAL STUDY OF THE METAMORPHIC COMPLEXES OF THE SCOTTISH HIGHLANDS BY BRUNO J. GILETTI~ PH.D.~ STEPHEN MOORBATH~ M . A . D . F t t I L . F.G.S. AND RICHARD ST. JOHN LAM'BERT~ M.A. P H . D . F . G . S . S u b m i t t e d 19 October 1960 ; revised m a n u s c r i p t received 13 F e b r u a r y 1961 ; read 4 J a n u a r y 1961

[PLA~ I X ] Co~m~rs I. I n t r o d u c t i o n . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . I I . SnmrniEry of geological d a t a on t h e age of t h e m e t a m o r phic complexes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . (a) The Lewisian complex . . . . . . . . . . . . . . . . . . . . . . . ,. (b) The Moine a n d D a l r a d i a n Series . . . . . . . . . . . . . . . . III. Analytical methods ................................. IV. Geochronological d a t a a n d discussion . . . . . . . . . . . . . . . . . (a) T h e Scourie area . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . (b) L a x f o r d i a n m e t a m o r p h i s m . . . . . . . . . . . . . . . . . . . . . . (c) The Moine Series . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . (d) T h e D a l r a d i a n Series . . . . . . . . . . . . . . . . . . . . . . . . . . (e) R o c k s f r o m C o n n e m a r a , I r e l a n d . . . . . . . . . . . . . . . . . V. Correlations w i t h o t h e r areas . . . . . . . . . . . . . . . . . . . . . . . . . VI. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . V I I . A p p e n d i x . Localities a n d descriptions of a n a l y s e d samples V I I I . L i s t of references . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

PAGE 234 234 234 235 238 240 241 243 245 249 253 253 254 255 262

SUMM~RY R u b i d i u m - s t r o n t i u m a g e - d e t e r m i n a t i o n s are p r e s e n t e d for m i n e r a l s a n d whole rocks from the Lewisian, Moinian a n d D a l r a d i a n m e t a m o r p h i c complexes o f Scotland a n d from the C o n n e m a r a schists of western I r e l a n d . Ago d a t a from the Lcwisian eoml)lex confirm t h a t it was affected b y two m a j o r periods of m e t a m o r p h i s m . P e g m a t i t e s associated w i t h the Scourian p a r t of t h e Lowisian complex are s h o w n to be a t least 2460 m . y . old, whereas t h e L a x f o r d i a n m e t a m o r p h i s m occured a b o u t 1600 m.y. ago. T h e effect of t h e L a x f o r d i a n m e t a m o r p h i s m on t h e Scourian p e g m a t i t e s is to p r o d u c e a s c a t t e r of ages in which coexisting p o t a s s i u m feldspars a n d biotites show the p a t t e r n potassium-feldspar age > biotite age. Six biotites, a microcline a n d a m u s c o v i t e from t h e Moine Series h a v e ages in t h e range 435 to 405 m.y., showing t h a t a widespread Caledonian (sen~u atr/c~) metam o r p h i s m affected the Moine Series 420 4- 15 m . y . ago. Two p e g m a t i t e s f r o m t h e K n o y d a r t - M o r a r a r e a yielded m u s c o v i t e s w i t h ages of 740 m . y . a n d 665 m . y . ; a s u r v e y of the geochemical possibilities a n d consideration of the geological s e t t i n g of the p e g m a t i t e s suggest t h a t the Moine s e d i m e n t s in this area are older t h a n 740 m . y . a n d m a y h a v e u n d e r g o n e an e a r l y m e t a m o r p h i s m before this date. Specimens from t h e D a l r a d i a n Series of P e r t h s h i r e suggest a m a j o r m e t a m o r p h i s m a t 475 -t- 15 m . y . ago, i n t e r p r e t e d as L o w e r or Middle Ordovician in age. Two whole-rock a n d t h r e e mineral a n a l y s e s f r o m the p r o - m e t a m o r p h i c B e n Vuroch granite-gneiss suggest t h a t the intrusion was f o r m e d 600 4- 100 m . y . ago a n d t h a t a partial r e c o n s t i t u t i o n occurred 415 -[- 10 m . y . ago. T h e Ben V u r o c h g r a n i t e complex as a whole appears to h a v e b e h a v e d as a closed s y s t e m w i t h respect to r u b i d i u m a n d s t r o n t i u m d u r i n g later m e t a m o r p h i s m . Three specimens of m u s c o v i t e a n d biotite from the Cormomara schists of western I r e l a n d have a m e a n age of 475 m . y . ; this finding t e n d s to s u p p o r t t h e generally supposed e o n t e m p o r a n e i t y of t h e I ) a l r a d i a n a n d C o n n e m a r a m e t a m o r p h i s m s . B i o t i t e f r o m t h e G a l w a y g r a n i t e h a s a n age of 365 ± 10 m.y., which suggests t h a t this granite m a y be c o n t e m p o r a n e o u s with o t h e r d a t e d Caledonian granites of the British Isles. F o u r p o t a s s i u m - a r g o n ages s u p p o r t the conclusions on the age of t h e L a x f o r d i a n a n d Caledonian-Moinian m e t a m o r p h i s m s .

From Le

Bas, M. J. (ed.), 1995, Milestonesin Geology, Geological Society, London, Memoir No. 16, 57-65

W. Q. Kennedy, the Great Glen Fault and strike-slip motion B.

J.

BLUCK

Department of Geology and Applied Geology, University of Glasgow, Glasgow G12 8QQ, UK Abstract: At the time it was written, Kennedy's paper on the Great Glen Fault had clear evidence for

a known lateral displacement, and the evidence was so well presented that it convinced a sceptical geological world that such movements were possible. The acceptance of large scaled lateral movements led to the concept of great fundamental fractures and, with the advent of plate tectonics and a climate of mobilistic thinking, many of these great fractures were later recognized as plate or terrane boundaries. Along with this thinking, new criteria evolved for recognizing those fractures that had been involved in major displacements--in fact the concept of throw became replaced by the concept of role. Role was identified from the history of the blocks on either side of the fracture, and where that history was incompatible with them being together, then a large role was possible for the fault itself. Taking a new look at the Great Glen Fault in these terms, it becomes clear that there are insufficient data on the rocks on either side to allow any conclusions about the nature and timing of its role to be deduced. If the later, displacements, which are at present the main concern of researchers, are the sum of the movements on the fracture then it is ironically the least significant of the four NE-SW fractures in Scotland.

There can hardly be a student who graduated in the 1950s and 1960s who either did not read or was not aware of the paper by W. Q. Kennedy (1946) on the Great Glen Fault. Its impact lay in two areas: he raised the possibility of large scale lateral movements within continental crust, a concept not highly regarded in the fixist days before plate tectonics; the other in the masterly way in which evidence from a range of disciplines was mustered to focus on a single solution. In both these areas lay the future of a large segment of geological thinking. Lateral displacements along faults had been described in New Zealand (McKay 1890) and subsequently in Japan and California. But, as Sylvester (1988) pointed out, there was a reluctance on the part of geologists to apply the clear evidence for lateral movements in recent earthquake zones to the geological record. This reticence was partly lodged in the difficulty in understanding how the crust at either end of the fault accommodated the movement. Kennedy grew up geologically in an environment which was making much of the concepts of faulting. H e was born in Scotland, where, at that time, there was not only the greatest concentrations of faults on the ground, but where obvious large scaled structures, with clear geomorphological and structural signature cut the country from shore to shore (Fig. 1). It was the country where controversy over major structural features had been in existence for, and occasionally raged for, some 80 years. During his undergraduate days with Gregory at Glasgow he would have no doubt been introduced to these controversies and also to the concept of rifting (and the excitement aroused by Africa). Subsequently at the Geological Survey, he was introduced to the concept of wrench faulting by such fellow officers as Anderson who he readily acknowledges as a stimulus to his own paper of 1946. More apposite to the culmination of these ideas in the geographical region of the Great Glen, he would almost certainly have read Cunningham Craig (in H o m e & H i n x m a n 1914) who had already discussed the existence of strike slip movements in the vicinity of the Great Glen Fault.

The 1946 paper (originally given in 1939) was sufficiently rigorous and convincing to justify its publication in geological climate unsympathetic to strike-slip faulting on anything but the smallest scale. Although it may not have overcome the prejudice of the larger geological community against strike-slip movements of some magnitude, it released a number of important papers describing large scaled lateral movements on fault systems throughout the world. Kennedy's paper had, as daring papers so often do, crystallized a whole undertow of feeling that there was more to seen in the world of strike-slip faulting than the prevailing climate of prejudice would allow to be seen. In these ensuing papers, and to some extent in his own paper of 1946, there were faults described which had a long and variable geological history of movement: they had greatly different displacements on them at different times (e.g. the San Andreas Fault by Taliaferro 1941). The concept of the great fundamental fault, as De Sitter (1956) was later to call them, was born; and along with its birth came a change in geological thinking towards a more mobilistic view of geology. There was much discussion on the nature of fundamental fractures, and Bailey & McCallien (1953) pointed out that there were a group of major fractures which were associated with serpentinites and may, for that reason, be tapping magmas from great depths. As more of these large fracture zones were described throughout the world, so grew the numbers of papers which recorded or estimated that blocks of crust could move hundreds of kilometres instead of a few. The 100km lateral displacement estimated for the Great Glen and Dead Sea Rift faults was followed by 450 km for the Alpine fault in New Zealand (Wellman in Benson 1952); > 5 6 0 k m for the San Andreas (Hill & Dibblee 1953) amongst many others. Yet there still remained a whole gamut of faults being described having all the characteristics of fundamental fractures but for which no lateral movement could be determined. These faults were marked by a lack of correlation on either side of them: the Highland Bounday 57

58

B . J . BLUCK

~

I

ol)hioliteaml)hibolite sole 490Ma

thrusting & orl)hism c . 4 6 0 M a ges 4 6 0 - 3 9 0

ROCKS Cover

t Dalradian ] Grampian t Central Highland Granulites

t Crust

c.2.0 S. of

Moine Foreland

0

Km

100 J

Fig. 1. Outline of the major fractures in Scotland, the distribution of the basement units and some important features along the Great Glen Fault. MT, Moine Thrust; GGF, Great Glen Fault; HBF, Highland Boundary Fault; SUF, Southern Upland Fault; SL, Solway line; BV, Ben Vuirich; CC, Carn Chuinneag.

Fault, as described by Anderson (1946) was such an example.

A n e w role for f u n d a m e n t a l fractures At the same time as there was a change in attitude towards a more mobile earth, triggered by Kennedy's 1946 paper, so another Celt working in another Celtic realm had challenged another geological dogma. Although Dana (1873) and Bertrand (1897) had impicitly or explicitly invoked a tectonic control on sedimentation, it was O. T. Jones (1938) who revived and redefined the notion at a critical time, pointing out that in the Lower Palaeozoic rocks of North and Central Wales, tectonics had an influence on sedimentation. Accepting this was the first step towards rejecting a view of geological history which saw it as comprising long periods of quiescence when sedimentary rocks were laid down punctuated by short periods of world-wide change during which time these rocks were deformed and thrown into mountain chains. The mountains were then slowly eroded down to yield sediment for the next cycle. Even Krynine (1945) who was a vigorous proponent of the tectonic control on sedimentation still produced a cycle in which shelves were converted to orogens. The mobilistic view, as presented by Gilluly (1949), effectively brought to an end the concept of punctuated orogeny. Following on the recognition of the nature and diversity

of fundamental fractures and the close relationship between tectonics and sedimentation, came the recognition that faults had an essential role in controlling sedimentation and stratigraphy. Their movements were not accomplished in an instant nor always associated with a stratigraphic void, but it became clear that they exerted a strong control on the nature of the sedimentary record, they often defined source-basin margins and some of the history of their activity could be read from the rocks they helped to generate. It was in this atmosphere of tectonic control on sedimentation, and with the newly described models for the genesis of the Swiss molasse where there was a direct influence of faulting on sedimentation, that the role of great fundamental faults was extended in this country, again by Kennedy (1958) when he turned his attention to the big fractures to the south of the Great Glen. Seemingly detached from his view of the Great Glen Fault, Kennedy again entered the scene of fundamental fractures with postulating that the evolution of Midland Valley of Scotland was largely controlled by the Highland Boundary Fault to the north and Southern Uplands fault to the south. As with the Great Glen Fault, Kennedy had a dynamic view of Midland Scottish geology. Here he traced the long history of the Faults, particularly the Highland Boundary Fault, recognizing activity stretching back to the Arenig when he saw its earliest inception. During Early Devonian times he envisaged that a developing graben had been bounded on the north by the Highland Boundary Fault and to the south by the Southern Uplands Fault, and clearly saw the Highlands and Southern Uplands as the sources of the sediments which filled the rift. The mid-Devonian folding he read as the closing of the 'jaws' of the rift as the Southern Uplands and Highlands closed on each other and folded the rocks of the Midland Valley between. In this account of Kennedy, we see that him interpreting fundamental fractures, such as the Highland Boundary fault, as having differing roles at different times, but he never suggested that either the Southern Uplands or the Highland Boundary faults as having a history of substantial strike-slip. It is instructive to view now the repost by George (1960) who, in strong contrast to Kennedy came from a background of stratigraphy worked out on horizontal or gently folded rocks in South Wales and the Borders. It was the stratigraphy rooted in sequences of events through time rather than sequences through space and time. He saw the stratigraphy of the Midland Valley much more in a layer-cake form (although Kennedy himself was not free of such a view) with no evidence for the bounding faults having anything like the history of control on sedimentation as envisaged by Kennedy, but recognized and extended the concept of multiple movement along the fractures. In addition he pointed out that there were faults within and bounding the Midland Valley which had a control on sedimentation during the Carboniferous.

A n e w e r still role for f u n d a m e n t a l fractures The advances in understanding the nature of fundamental fractures in the interval from 1946 to 1966 were minimal compared with the advances which followed in the next 15 years. An acceptance of plate tectonics had the effect of releasing geologists from the constrains of limited movement of crustal blocks. Movements of continental masses were

W. Q. K E N N E D Y & THE G R E A T GLEN F A U L T proposed which were so large scaled as to make the propositions of Kennedy over 20 years earlier seem trivial. However, of more direct relevance to the interpretation to fundamentral fractures such as the Great Glen was the discovery in 1970s of exotic terranes which had accreted to the west coast of North America but which had a provenance in the east coast of Asia. These blocks had accreted during subduction at various points along the western edge of N America and had migrated northwards along the coast towards Alaska, their probable final home (Jones et al. 1972; Howells 1989). In order to achieve these large scaled movements they had to be bounded by large scaled faults, and one of the identified Earth's fundamental fractures of earlier times, the San Andreas Fault, was identified as such a terrane-bounding fracture. Workers in other regions were quick to see the importance of the terrane concept and, freed from the old constraints of limited lateral movement, began identifying many terranes and boundaries with varying degrees of lateral movement. But in any case the scales of movement now conceived were greater than the lateral persistence of a recognizably similar geology on either side of the fault. The old criteria used to identify major strike-slip displacement (i.e. the distance between correlatable elements across the fault) changed to one where there was no expected correlations across faults. A new set of criteria evolved by which large scaled faulting was determined and t h e y had very little to do with those used by Kennedy on the Great Glen Fault. After Dewey (1969) had demonstrated that the Lower Palaeozoic rocks of northern Britain had formed on a destructive plate margin, there was a significant change in the climate of thinking amongst those working in Palaeozoic geology. Discoveries in Cyprus and Newfoundland of oceanic crust sited on continental blocks prompted much discussion about the mechanics and scales of movement at destructive margins of this kind. When, later, the nature of the Ballantrae Ophiolite was clearly determined (Church & Gayer 1973; Dewey 1974), the problem came to Scotland and it became clear to those working in the Caledonides, as it had to those in the Mediteranean and Appalchians, that there was considerable tectonic significance in having oceanic crust lying on the continent: it implied that there was an enormous displacement of crust, and the faults which bounded the ophiolite were of unimaginable throws. Even the most conservative thinkers had eventually to concede that if the origin of the ophiolite was indeed in an oceanic setting, then from the growing background of what we understood from the oceans and their continental boundaries, its emplacement implied substantial structural activity. In this we see the development of a new argument: the geological nature of two blocks; their origin, associati~on and history of genesis have far more significance to nature of their boundaries than the visible structure of the boundary itself. And it left structural geologists who look closely at the fabrics of fault zones with a new challenge and wider boundaries within which to work. In the Ballantrae Complex this logical demand for substantial displacement was further compounded when, following Williams & Smyth (1973), Spray & Williams (1980) and Treloar et al. (1980) were able to determine that there was a depth of provenance for lithological elements of the metamorphic aureole which exceeded 12 kbar (36 km).. In this, as in many other examples along major tectonic

59

boundaries where many kilometres of displacement can be demonstated, the preservation of the history of movement is recorded only in fragmentary evidence along the lateral extent of the fault zone. Subsequent movement along this critical thrust at Ballantae has, in places, obscured these metamorphic rocks and placed unmetamorphosed spilite in contact with serpentinite with only a minor shear between. The lessons are clear: faults and the rocks in fault zones may only partly record the history of movement between blocks; subsequent movement may cut-out a critical earlier history so it is dangerous to read the history of a fracture from any one point or sector of its length. As already stated, the magnitude of movement must be seen in the context of the history of the blocks on either side of the fault as well as the fault zone itself. The acceptance of this type of evidence for large scale movement, paved the way for the more mobilistic views that followed in the immediately succeeding years. Barber (1985) and Bluck (1985) followed with the view that the major fractures of Scotland, including the Great Glen Fault, bound allochthonous blocks which were suspect terranes i.e. they were blocks which had no direct evidence to indicate that they were adjacent to each other for their entire history. The presence of large faults which bounded deformed blocks with apparently different histories was sufficient proof of suspect terranes, and the analogy with Mesozoic terranes of the American west set a new thinking going in the British Palaeozoic.

Blocks and boundaries: the way forward There are now clear indications in the Tertiary and Recent history of areas like SE Asia and western North America that continents are growing by the continual addition of discrete terranes with a history quite different from that of the block to which they are accreting. Taking this concept a stage further, Dalziel (1991, 1994) has attempted reconstructions of the 1.0 Ga continent and its subsequent break-up and re-assembly into Pangaea. In this reconstruction the re-arrangement of continents requires them to move great distances so that regions now quite remote from each other (western South America and UK for example) may have been juxtaposed in the Neoproterozoic. In the Wilsonian cycle of megacontinent growth and dispersal, there is therefore great potential for bringing together blocks with totally different histories and provenances to amalgamate onto continents with a history quite different from any of the blocks accreted to it. This dispersal of continents is achieved by ocean spreading, but it is the repeated change in the direction of spreading (as can be ciearly demonstrated on most destructive margins today), which makes it probable that continental blocks of all sizes are spread widely before they are re-assembled. But within this regime of widespread dispersal, there is a lower order of lateral movement which is of considerable significance. Tectonic elements, such as arcs, fore-arcs etc. on a single destructive margin, may be broken up and move laterally to re-assemble along the same continental edge as has been demonstrated for areas of the western Pacific such as the Phillipines (Karig 1983). The boundaries between these major tectonic elements on destructive marginsare often zones of weakness, so that fore-arcs can be detached from arcs and back-arcs and

60

B.J.

independenly move along continental margins far from their position of genesis. C a l e d o n i a n terranes a n d their b o u n d a r i e s

It follows from this discussion that a most important step to take in the Caledonides was to define the tectonic elements and terranes and locate their limits. By 1977, following closely on the work of Dewey (1974) and Church & Gayer (1973), McKerrow et al. (1977) made a major step forward in interpreting the Southern Uplands as an accretionary prism. England and Wales was now firmly regarded as a fragment alien to Laurentia, and in looking for the southern limit to the Southrn Uplands, there grew mounting evidence for the presence of another fracture in the Caledonides which had not been recognized by either Kennedy or George: the Solway line. The Solway line is a major fracture sitting alongside the others in Scotland. It owes its birth in the geological literature to the same sort of reasoning which was applied to terranes in the Mesozoic of the American west and to the Ballantrae Complex--it had to be there to satisfy the juxtaposition of an Ordovician arc in the Lake district with a fauna in its sediments quite different from the accretionary prism to the north. In terms of what can be seen in present-day plate regimes, the Lake District Borrowdale arc would have had a fore-arc in front of it and a southerly dipping plate beneath it: the Southern Uplands was a fore-arc, but its structure and stratigraphy supported the palaeontology in suggesting it to be a fore-arc above a plate that was subducting towards the north. There was a lot of ground clearly missing between the two tectonic elements. There was yet another turn in the thinking about major faults: throw was not a meaningful concept any more. An ocean plate of uncertain width had been consumed along the zone where now lies the Solway line. In addition, the presence of a fault that had yet to be seen, was accepted as a necessity because without it a section drawn from southern Scotland to northern England did not make sense in terms of the present-day distribution of tectonic elements. In this way a new form of critical thinking was used in evaluating major tectonic boundaries; a form of thinking used so effectively along the west coast of North America. So great was the role of this previously undiscoverd fault that the meaning of the known and fully exposed fractures had to be questioned in this new light. As with terrane workers elsewhere in the world, the stratigraphy, sedimentation, palaeontology, igneous and metamorphic history and tectonic history all had to be evaluated and rigorously examined for any mismatches or evidence of missing ground across the known major fractures. The Southern Uplands Fault. As an example of the new approach to the evaluation of blocks and boundaries, the case of the Southern Upland Fault is taken. The fault terminates the accretionary prism of the Southern Upland on its northern margin. The Southern Uplands, through the meticulous work of Walton and his many students over the years (Kelling 1962; Walton & Oliver 1991), was demonstrated to have a provenance in a metamorphic block which was associated with volcanic and plutonic rocks and an ophiolite. A fore-arc (accretionary prism) should have an arc to the north of it and it seemed safe to assume that the

BLUCK volcanic rock fragments in the greywackes of the Southern Uplands came from such a source. In the area around Girvan there is a proximal, fault-controlled sequence which overlaps in age with the finer grained turbidites of the Southern Uplands. Age determinations from boulders of granite in the conglomerates at Girvan showed them to have come from a contemporary igneous province which was clearly only a little distance to the north (Longman et al. 1979). This implied that the source of the Southern Uplands sediment was within the region of the Midland Valley or its lateral equivalent. It then became clear that the nature of that igneous source" was almost certainly a dissected arc and the Girvan sequence was its fore-arc. On the assumption that the Southern Uplands was a trench sequence, and there were many doubters (Murphy & Hutton 1986; Stone et al. 1987), there was now a problem that the gap beween the trench and the arc was only a few kilometres wide, so it was proposed that the Southern Uplands block was allochthonous, having been thrust over a continental basement and the gap thus reduced (Bluck 1985). Geophysical investigations have shown shallow continental basement beneath the Southern Uplands (Hall et al. 1983) but this can be regarded only as supportive of the view that they have been displaced northwards if the hypothesis of them being an accretionary prism is correct: a back or fore-arc for instance can be founded on continental crust. As with the Solway line, the history of the Southern Upland Fault is determined from close reasoning over the history of the blocks on either side: the fracture is only poorly exposed and very little of its history is likely to have been preserved in the fault itself. It is easy to imagine that after their initial suturing, movement continued between the receiving continent and the donated terrane and that later movement was likely to overprint or somehow obscure the earlier record of initial suturing. Interpretation of the geological history of terrane boundaries which has undergone this type of accretion is therefore likely to be thwart with potential problems of an incomplete structural record of the amalgamation. The Highland Boundary Fault zone. This boundary differs from the two previously discussed fractures in that there is comparatively good exposure of the margins of the blocks on either side of it. In addition, there is a range of rock types and ages (from Cambrian to Carboniferous) which are available to record the history of movement and for these reasons it is discussed in a little detail (Fig. 2). The Highland Boundary Fault has a sinuous trace across Midland Scotland, bifurcates at its southern end and variably dips to the northwest (Dentith et al. 1992), southeast, or is vertical. To the north lies the metamorphic basement of the Dalradian, and to the south it bounds rocks of Cambrian, Ordovician, Silurian, Devonian and Carboniferous age. The Dalradian block is a polyphase folded, late Proterozoic (Halliday et al. 1989) metamorphic sequence of passive margin style rocks which have been metamorphosed at least once. A n early phase of folding (Tanner & Leslie 1994) and possibly a phase of metamorphism is cut by the Ben Vuirich granite which is 590 ± 2 Ma (Rogers et al. 1989) or 597 = +11 Ma (Pidgeon & Compston 1992), and a later phase of folding, metamorphism and uplift occurred in the interval 515 to c. 430 Ma. (Dempster 1985).

W. Q. KENNEDY & THE G R E A T GLEN FAULT DALRADIAN

MIDLAND

TERRANE

Ma

Carboniferous

-~c-~- ~

I

STATE OF TERRANES

VALLEY

TERRANE

HBFZ

AMALGAMATED

Carboniferous overstep

\ c=,o,.

J

Peneplain

2

Devonian Igneous activity

Silurian

~+ * \

-

Peneplain 450

1

,° . , 0 , o o 0

]~,~... ~ ~ / ' ~ ~ ~ -~ ° ° ' ~ ~

Valley (UORS) Thrust convergence:

~

Strathmore syncline

AMALGAMATION

Strike-slip basins

Ordovi I cia~500 rapid uplift

61

\ \

\

APART

arc, back arc basin ~,~/~.-,~,.,~~/~ ~,~ • a.~, • -

The history of the south side is particularly revealing and is traced from Cambrian times. (1) C a m b r i a n is represented on the Island of Bute by a sliver of metamorphic rock associated with serpentinite which resembles the sole to an ophiolite (Henderson & Robertson 1982). A cooling age of 540 Ma was reported from the amphibolite suggesting an obduction at about that time (Dempster & Bluck 1991). Near Callander, a sliver of distal, shelf-type Lower Cambrian limestones, overlaps in age with this cooling age. The association and relationship to the fault history of both these slivers is uncertain but is likely to be related to the convergence of the Midland Valley and Dalradian terranes (see Fig. 2). (2) Ordovician. Undoubted rocks of Ordovician age have been described (Curry et al. 1982) and these are overlain by sedimentary rocks which are themselves overlain by Old Red Sandstone rocks. The latter, formerly regarded as approximately Devonian are now thought to be well into the Silurian (Thirlwall 1988; Richardson et al. 1984). This part of the Highland Border Complex is therefore likely to be essentially Ordovician in age and comprises limestones, black shales, metamorphic rocks and basic-ultrabasic igneous bodies. The juxtaposition of the Highland Border Complex with the Dalradian block is incongruous in that the Dalradian block which was being uplifted in Ordovician time should have been yielding sediment to a coeval basin now preserved in the Highland Border Complex (Bluck 1985). The fact that it did not, suggested that the two blocks were not near each other in Ordovician times. This in turn suggested that there was a substantial displacement on the Highland Boundary Fault and one or both were displaced terranes in the sense used by workers on the W USA. Tanner (1994) has recently suggested that the earlier work of Johnson & Harris (1967); Harris & Fettes (1972) and others in having the Highland Border Complex or part of it belonging to the Dalradian, may be correct. This would therefore remove the evidence for the Highland Boundary Fault being a terrane boundary of any significance. This would appear to be the classical case of weighing the evidence from the fault zone itself against the evidence from

Fig. 2. Diagram illustrating the history of the blocks on either side of the Highland Boundary Fault zone (HBFZ) and how this is recording the interaction of the blocks north and south of the fault. Peneplain 1, refers to the Late Silurian peneplain as recorded for example in the Lorne area where Late Silurian lavas rest on a flat Dalradian surface; Peneplain 2 is the Late Devonian-Carboniferous peneplanation.

the history of the blocks on either side. However, if it is agreed that there is a major break between the Dalradian and Highland Border Complex then it is instructive to see if younger rocks at the Highland Border have recorded the amalgamation of the two disparate blocks. (3) S i l u r i a n - D e v o n i a n rocks of this age occur in the asymmetrical Strathmore syncline which runs parallel with the outcrop of the Highland Boundary Fault. On the steep northern limb of the syncline, the Highland Border Complex provided the basin floor for the Old Red Sandstone south of the fault as the Dalradian does to a few outliers of Old Red Sandstone to the north. The Highland Border Complex was therefore flat lying or gently dipping at the begining of Old Red Sandstone time. The provenance of the sediment of this age in the basins north and south of the faults is very different despite being of roughly the same age: Dalradian clasts dominate the sequences to the north and polycyclic quartzites and volcanic rocks dominate the very much thicker sequences to the south. These basins south of the fault have no unequivocal Dalradian clasts in them, but the Dalradian block could not have been far away from the Strathmore basin on its southern side as an ignimbritic flow overlies both Dalradian and Strathmore basin sediment. Clearly this was a time of terrane amalgamation, when the Midland Valley and Dalradian blocks were joining together but not clearly exchanging sediment. The two basins north and south of the fault are clearly not displaced margins of one single basin, but are distinctive basins which have been brought together by subsequent fault movement and intervening ground has been lost in this convergence. One stage of that convergence history occurred before the deposition of the Upper Old Red Sandstone when the Lower ORS was folded into the asymmetrical Strathmore syncline. The axis of this syncline runs parallel to the Highland Boundary Fault and indicated a thrust convergence in this zone (Fig. 2) By Upper Old Red Sandstone times, in a coarse fanglomerate near the Highland Boundary Fault in the SW Midland Valley, the first clasts of unequivocal Dalradian provenance entered the Midland Valley sequence (Fig. 2). A

62

B.J. Ma

450

--

NW HIGHLANDS

500

-

DALRADIAN BLOCK Reburial

Cambro-Ordovician

Erosion of mountain b e l t

550

Torridonian

Collapse of Dalradian basin

?

600

650

Dalradian basin

700

-

750

--

800

L

Torridonian basin

Gondwanan b a s e m e n t

Fig. 3. Diagram showing the nature of the late Proterozoic history of the foreland (west of Moine thrust) and Dalradian blocks. When there was a passive margin on the west of the Moine Thrust, there was the large expanse of ductile folded and at least partly metamorphosed rock lying to the west. Bluck & Dempster (1991) point out this paradox. slight relative uplift of the Dalradian is probable at this time, possibly as a result of further convergence of the Midland Valley and Dalradian terranes. (4) Carboniferous. Carboniferous volcanic and sedimentary rocks transgressed the fault to rest on Dalradian, but subsequent movement in the fault zone alternately threw Carboniferous rocks down to the northwest and to the southeast. The Highland Boundary Fault has a fairly complete record of the movements along it and from reading these in terms of the history of the blocks on either side a convergence history can be reconstructed. But this has to be read from the a study along the fault: any one locality may record only its latest (post-Carboniferous) movement. In this respect it is similar to the many terrane boundary faults seen elsewhere in the world.

Back to the Great Glen Armed with the degrees of freedom which the concepts of plates and terranes have given and in the light of the more intensive examination of faults which these concepts have stimulated, the Great Glen Fault can be now approached with a very different mind from the one which took Kennedy there some 50 years ago. But a bigger framework of knowledge brings with it a wider range of prejudice. Examination of terrane boundaries, such as the Highland

BLUCK Boundary Fault, has taught the geological community to be suspicious of reading the history of fundamental fractures in terms only of the displacements which we can now observe: these may be young modifications produced on reactivation of the older fractures. The magnitude and role of the fault is often demonstated by critically examining the whole history of the rocks which lie on either side of it. Moreover some of that history is often written in the slivers which occur along its entire length. The Great Glen Fault is thought to extend into the Shetland Isles where it is identified as the Walls Boundary Fault (Flinn 1992). Its extension to the southwest is far less certain but it is clear that its great length through Scotland can be matched only by its poor exposure compared with the other large fractures. A critical line of evidence, the composition and history of the blocks within the fault zone as used in the Highland Boundary Fault, is not widely available on this fracture. However, there is ample exposure of the rocks on either side over a considerable lateral distance, and Kennedy made dramatic use of them in his original work. Since Kennedy's time there has been much work done on attempting to establish the nature, timing and magnitude of the movements on the Great Glen Fault. Palaeomagnetic measurements have led to a view of unusually large displacements (Van der Voo & Scotese 1981; Storetvedt 1987) which are usually unacceptable because of the constraints of the existing geology on either side of the fault and more rigorous palaeomagnetic studies (Torsvik 1984). But others have estimates which vary greatly amongst themselves in timing, amount and nature of displacement. Rogers & Dunning (1991) and Hutton & McErlean (1991) show clear evidence for sinistral shear in the region of the Great Glen at c. 425 Ma. Hutton & McErlean (1991) see further evidence for sinistral movements contemporaneously with the intrusion of dykes at 410-395 Ma. The main source of this evidence comes not from the fault itself but from shears thought to be related to it. Although movement along the fault itself is difficult to establish, Donovan & Mayerhoff (1982), Parnell (1982) and Rogers et al. (1989) have appealed to displacements of the Old Red Sandstone outcrops in the Moray Firth region. Rogers et al. (1989) suggest post-Frasnian to pre-Permain dextral movements of 2 5 - 1 2 0 k m movement, but Flinn (1992) deduces a pre-Carboniferous sinistral net movement of c. 100 km and a dextral 65 km movement in Jurassic times. Most of the discussion of the Great Glen Fault, both recently and at the time of Kennedy, has concentrated on establishing the magnitude of its throw. In the light of terrace accretion tectonics the whole emphasis with respect to major faults has changed: the throw of the fault is now subordinate to its role. From the more recent work on the three major Scottish fractures to the south, and particularly illustrated by the Highland Boundary Fault (Fig. 2), major fractures are evaluated on the following criteria. (1) Degree of separation of the two blocks on either side of the fracture: were they great distances apart? At this stage oceanic crust normally separates the blocks. (2) The history of amalgamation of the blocks. (3) Their post-amalgamation history. Magnitudes or throws of displacement have relevance only to the third and possibly to part of the second of these criteria. The work on the Great Glen Fault so far, including Kennedy's, has addressed only the third and possibly second

W. Q. K E N N E D Y & THE G R E A T GLEN F A U L T of these criteria: it is clear that the fault, in Silurian and post-Silurian, times can be regarded as undergoing relatively minor adjustments. In order to establish if it is a terrane-scaled fracture satisfying the first of the above criteria, it is necessary to examine the pre-Silurian history of the blocks on either side and evaluate the compatiblity of their histories.

The pre-Silurian Great Glen fault The pre-Silurian rocks on either side of the Great Glen fault are basements of various ages and associations (Fig. 1). The ways of reading basement history are very different from ways of reading the Palaeozoic history as it has been applied to the ground further south. Although the basement history may not be subject to the same degree of rigour, its evaluation should not be based on no rigour at all, and if one considers that the Scottish Highlands are amongst the most intensively worked basement areas in the world, there should emerge some substantial criteria by which to evaluate the role of the Great Glen fault in terms of the basements on either side of it. As with the analysis of the Scottish terranes to the south, it is important first to establish the blocks and then their boundaries. There are four recognized basement blocks to the south of the Great Glen Fault. (1) The furthest south is the Dalradian block which comprises a passive margin type sequence (Anderton 1985) which has been repeatedly folded up to four times. An early phase of folding (Tanner & Leslie 1994) and possibly a phase of metamorphism is cut by the Ben Vuirich granite which is 590 + 2 Ma (Rogers et al. (1989) or 597 = +11 Ma (Pidgeon & Compston 1992), and some of the subsequent fold phases are seen to post-date the intrusion. After a substantial gap of c. 75Ma, a major cooling event was recorded so the Dalradian block, if the data relating to Ben Vuirich are accepted, was probably subject to two periods of metamorphism. (2) To the northwest of the Dalradian lies the Grampian Group that comprises, as far as can be read, a shelf-like sequence which is apparently cut by the c. 750 Ma suite of pegmatite sheets. Many workers (e.g. Harris et al. 1978) include the Grampian Group within the Dalradian sequence and point out, along with Treagus (1987) and Winchester (1992), that there is a transition between them. These rocks are, in turn, at least partly in shear contact with rocks of the Central Highland Division (3) The Central Highland Division comprises migmatitic gneisses the age of which is uncertain but which have been correlated with Moine rocks on the north side of the Great Glen fault (Piasecki & Van Breemen 1982). Southwest, along the Great Glen, the Central Highland Division is cut out of the sequence, and rocks of the Grampian Group lie in contact with faults near to and parallel with the Great Glen Fault (Phillips et al. 1993). (4) Further to the southwest, on the islands of Colonsay and Islay rocks previously thought to be Lewisian have now been shown to be Ketilidian (Marcantonio et al. 1988). The Central Highland Division has many similarities with the Moine rocks to the north and most workers accept them to belong to the same group. This view of continuity across the Great Glen is supported by the presence of the c. 750 Ma group of pegmatites along some of the notable shears within both the Moine on the northern side and the

63

Central Highland Division to the south. There is therefore no evidence for post-750 Ma displacement of terrane scale along this zone. There are however a number of important points to bear in mind before assigning the Great Glen fault to the category of a minor fracture. (1) There is a considerable similarity in late Proterozoic sediments throughout areas of the world, and to some extent the distinctiveness of sedimentary sequences is often variably obscured when they are metamorphosed. In addition dykes and sheets can be intruded over a wide area in extensional regimes, although this applies to basic more than to acid ones. These considerations may still leave room for considerable displacement along the fault. (2) There is an ophiolitic assemblage in the Feltar mass in the Shetland Isles, and at the SW Scottish end of the fault, is Ketilidian crust which does not have clear associations with much else in Scotland. Terrane-scaled movement most commonly take place on oceanic crust which often gets preserved somewhere along the fracture. (3) The history of the rocks on the northern side of the Great Glen is also puzzling and not easily related to those on the south. They are thought to rest on gneisses considered to be Lewisian and are now recognized to be polymetamorphic. The earliest metamorphic episode was thought to be pre-750 Ma, the age of a suite of pegmatites which are found in some of the slides (Piasecki & van Breemen 1982). The later, c. 460 Ma metamorphic event was accompanied by at least three folded thrust sheets (Barr et al. 1986) which involved ductile translation towards the WNW at temperatures estimated to be c. 600 °C (Barr et al. 1986). Retrogression began at c. 440 Ma with cooling ages down to c. 400 Ma. There are three problems raised by these data. (1) The Moine at 460 Ma (Caradoc) and at its present level would need to have a cover of c. 20 km to maintain the temperatures prevailing at that time (Soper & Barber 1982), and even at 420 Ma (Ludlovian) the rocks were at c. 350 °C implying, at reasonable geothermal gradients, a cover of at least 10km. The nature, source and the means of post-Caradoc disposal of this cover are uncertain. (2) This type of deformation and metamorphism implies plate-scale interactions, but the Dalradian block lies beween the Moine and the paratectonic zone. Along with the cooling of the Moine the Dalradian also cooled rapidly at 460 Ma--as did many metamorphic blocks southwest along strike in the Appalachians. The provenance of the cover to the Moine can only be speculated, but there are a number of possibilities (see Bluck & Dempster 1991). It could either: (a) have been rooted in the area now the Great Glen fault; (b) have come over both the Dalradian and the Moine and be rooted to the south of the Dalradian outcrop; (c) the northern margin of the Dalradian (no longer visible) covered the Moine. Only in (1) above needs there have been any large scaled movement on the Great Glen Fault. (3) Whilst this ductile folding and metamorphic history was taking place east of the Moine thrust, there was a passive margin sequence being deposited to the west. Bluck & Dempster (1991) has pointed out the paradox of this situation (Fig. 3), and there may therefore be good reason to examine in detail the role of the Great Glen Fault and its subsurface relationship with the Moine Thrust.

B.J.

Conclusions There exists no compelling evidence for the the Great Glen Fault to have been a terrane boundary: the later history shows throws which are not in the mega-shear class. H o w e v e r there are anomalies in the history of the basements on either side of the fracture, but a great deal of work needs to be done on refining the significance of these, the boundaries to a n o m a l o u s ground as well as on the details of the timing of events before anything can be resolved from them. If ground so well k n o w n as this in Scotland can throw up anomalies of this kind which are not resolvable with the current state of information, then large b a s e m e n t areas the world over must have m a n y problems yet to be discovered. I wish to thank B. E. Leake for pointing out some details of the life of W. Q. Kennedy, and T. Dempster, G. Rogers and N. J. Super for valuable discussion.

References ANDERTON, R. 1985. Sedimentation and tectonics in the Scottish Dalradian. Scottish Journal of Geology, 21,389-406. ANDERSON, J.G.C. 1946. The geology of the Highland Border: Stonehaven to Arran. Transactions of the Royal Society of Edinburgh, 41, 479-515. BAILEY, E.B. & McCALLIEN, W.J. 1953. Serpentinite lavas, the Ankara melange, and the Analolian thrust. Transactions of the Royal Society of Edinburgh, 62, 402-442. BARBER, A.J. 1985. A new concept of mountain building. Geology Today, 1, 116-121. BARR, D., HOLDSWORTH, R.E. & ROBERTS, A.M. 1986. Caledonian ductle thrusting in a Precambrian metamorphic complex: the Moine of Northwestern Scotland. Geological Society of America Bulletin, 97, 754-764. BENSON, W.N. 1952. Meeting of the geologic division of the Pacific Science Congress in New Zealand, February 1949. International Proceedings Geological Society of America Bulletin, 63, 11 - 13. BERTRAND, M. 1897. Structure des Alpes Francaises et recurrence de certain facies sedimentaires. Comptes Rendus, Congress International 6th Session, 1894, 164-177. BLUCK, B.J. 1985. The Scottish paratectonic Caledonides. Scottish Journal of Geology, 21, 437-464. & DEMPSTER, T.J. 1991. Exotic metamorphic terranes in the Caledonides: tectonic history of the Dalradian block, Scotland. Geology, 19, 1133-1136. CHURCH, W.R. & GAYER, R.A. 1973. The Ballantrae ophiolite. Geological Magazine, 110, 497-510. CURRY, G.B., INGHAM, J.K., BI.UCK, B.J. & WILLIAMS, A. 1982. The significance of a reliable Ordovican age for some Highland Border rocks in Central Scotland. Journal of the Geological Society, London, 139, 451-454. DALZIEL, I.W.D. 1991. Pacific margins of Laurentia and East AntarcticaAustralia as a conjugate rift pair: Evidence and implications for an Eocambrian supercontinent. Geology, 19, 598-601. -1994. Precambrian Scotland as a Laurentia-Gondwana link: Origin and significance of craton promontories. Geology, 22, 589-592 DANA, J.D. 1873. On some results from the earth's contraction from cooling, including a discussion of the origin of mountains and the nature of the earth's interior. American Journal of Science, 5, 423-443; 6, 6-14; 1114-115; 161-171. DEMPSTER, T.J. 1985. Uplift patterns and orogenic evolution in the Scottsh Dalradian. Journal of the Geological Society, London, 142, 111-128. -& BLUCK, B.J. 1991. The age and tectonic significance of the Bute amphbolite, Highland Border Complex, Scotland. Geological Magazine, 128, 77-80. DENTITH, M.C., TRENCH, A., & BLUCK, B.J. 1992. Geophysical constraints on the nature of the Highland Boundary Fault zone in western Scotland. Geological Magazine, 129, 411-419 DE S~YrER, L.U. 1956. Structural Geology. Mc Graw-Hill. London. DEWEY, J. F. 1969. Evolution of the Caledonian/Appalachian orogen. Nature, 222, 124-129. 1974. Continental margins and ophiolite obductions: Appalachian/ Caledonian system. In: BURKE, C.A. 8¢~ DRAKE, C.L. (eds) Geology of continental margins. Springer-Verlag, New York, 933-950.

BLUCK

DONOVAN, R.N. & MEYERHOff, A. A. 1982. Comments on 'paleomagnetic evidence for a large (2000 km) sinistral offset along the Great Glen fault during Carboniferous time' Geology, 10, 604-605. FLINN, D. 1992. The history of the Walls Boundary fault, Shetland: the northward continuation of the Great Glen fault from Scotland. Journal of the Geological Society, London, 149, 721-726. GEORGE, T.N. 1960. The stratigraphical evolution of the Midland Valley. Transactions of the Geological Society of Glasgow, 24, 32-107. GILLULY, J. 1949. Distribution of mountain building in geologic time. Geological Society of America Bulletin, 60, 561-590. HALL,J., POWELL, D.W., WARNER, M.R., EL-ISA, Z.M.H., ADESANYA, O. BLUCK, B.J. 1983. Seismological evidence for shallow crystalline basement in the Southern Uplands of Scotland. Nature, 305, 418-420. HALLIDAY, A.N., GRAHAM, C. M., AFTALION, M. & DYMOKE, P. 1989. The depositional age of the Dalradian Supergroup: U-Pb and Sm-Nd isotopic studies of the Tayvallich volcanics, Scotland. Journal of the Geological Society, London, 146, 3-6 HARRIS, A.L. & FETrES, D.J. 1972. Stratigraphy and structure of the Upper Dalradian rocks at the Highland Border. Scottish Journal of Geology, 8, 253-264. --, BRADBURY, H. J., JOHNSON, H.D. &SMITH, R.A. 1978. Ensialic basin sedimentation: the Dalradian Supergroup. In: BowLs, D.R. & LEAKE, B.E. (eds) Crustal evolution in Northwestern Britain and adjacent regions. Seel House Press, Liverpool, 115-138 HENDERSON, W.G. & ROBERTSON, A.H.F. 1982. The Highland Border Rocks and their relation to marginal basin development in the Scottish Caledonides. Journal of the Geological Society, London, 139, 433-450. HILL, M.L. & DIBBLEE, T.W. JR. 1953. San Andreas, Garlock and Big Pine faults, California--a study the character, history and tectonic significance of thier displacements. Geological Society of America Bulletin, 64, 443-458. HORNE, J. 8¢ HINXMAN, L.W. 1914. The Geology of the country round Beauly and Inverness: Scotland. Geological Survey Memoir sheet 83. HOWELLS, D.G. 1989. Tectonics of suspect terranes. Chapman Hall, London. HUTI'ON, D.W.H. & McERLEAN, M. 1991. Silurian and Early Devonian sinistral deformation of the Ratagain granite, Scotland: constraints on the age of the Great Glen fault system. Journal of the Geological Society, London, 148, 1-4. JOHNSON, M.R.W. & HARRIS, A.L. 1967. Dalradian-?Arenig relations in parts of the Highland Border, Soctland and their significance to the chronology of the Caledonian Orogeny. Scottish Journal of Geology, 3, 1-6. JONES, D.L., SILBERLING, N.J. & NELSON, W.H. 1972. Southerastern Alaska--a displaced continental fragment? US Geological Survey professional Paper, 800B, B211-B217 JONES, O.T. 1938. On the evolution of a geosyncline. Quarterly Journal of the Geological Society of London, 94, lx-cx. KARIG, D.E. 1983. Accreted terranes in the northern part of the Philippine archipealigo. Tectonics, 2, 211-236 KEELING, G. 1962. The petrology and sedimentation of Upper Ordovician rocks in the Rhinns of Galloway, southwest Scotland. Transactions of the Royal Society of Edinburgh, 65, 107-137. KENNEDY. W.Q. 1946. The Great Glen Fault. Quarterly Journal of the Geological Society of London, 102, 41-76. 1958. The tectonic evolution of the Midland Valley of Scotland. Transactions of the Geological Society of Glasgow, 23, 107-133. KRYNINE, P.D. 1945. Sediments and the search for oil. Producers Monthly, 9, 17-22. LONGMAN, C.D. BLUCK, B.J. & VAN BREEMEN, O. 1979. Ordovician conglomerates and the evolution of the Midland Valley. Nature, 280, 578-581. MARCANTONIO, F., DICKIN, A.P., McNurr, R.H. & HEAMAN, L.M., 1988. A 1800 Ma Proterozoic gneiss terrane in Islay with implications for the crustal struture and evolution of Britain. Nature, 335, 62-64. MURPHY, F.C. & Hurroh, D.H.W. 1986. Is the Southern Uplands of Scotland really and accretionary prism? Geology, 14, 354-357. McKAY, A. 1890. On the earthquake of September 1888 in the Amuri and Marlborough Districts of the South Island. New Zealand Geological Survey Report of Geological Explorations 1888-1889, 20, 1-16. MCKERROW, W.S. LEGGETT, J. K. & EALES, U. H. 1977. Imbricate thrust model of the Southern Uplands of Scotland. Nature, 267, 237-239. PARNELL, J. T. 1982. Comment on 'paleomagnetic evidence for a large (2000 km) smistral offset along the Great Glen fault during Carboniferous time' Geology, 10, 605. PHILLIPS, E.R., CLARK, G.C. & SMITH, D.I. 1993. Mineralogy, petrology and microfabric analysis of the Eilrig Shear Zone, Fort Augustus. Scottish Journal of Geology, 29, 143-158. PIASECKI, M.A.J. & VAN BREEMEN, O. 1982. Field and isotopic evidence for a c. 750 Ma tectonothermal event in Moine rocks in the Central Highland

W.

Q.

KENNEDY

& THE

region of the Scottish Caledonides. Transactions of the Royal Society of Edinburgh, Earth Sciences, 73, 119-134. PIGEON, R.T. & COMPSTON, W. 1992. A SHRIMP ion microprobe study of inherited and magmatic zircons from four Scottish Caledonian granites. Transactions of the Royal Society of Edinburgh, Earth Sciences', 83, 473-483. RICHARDSON, J.B., FORD, J.H. & PARKER, F. 1984. Miospores, correlation and age of some Scottish Lower Old Red Sandstone sediments from the Strathmore region (Fife and Angus). Journal of Micropalaeontology, 3, 109-124. ROGERS, D.A. MARSHALL, J.E.A. & ASTIN, T.R. 1989. Devonian and later fault movements long the Great Glen fault system, Scotland. Journal of the Geological Society, London, 146, 369-372. ROGERS, G. & DUNNING, G.R. 1991. Geochronology of appinite and related granitic magmatism in the W Highland of Scotland: constraints on the timing of transcurrent fault movement. Journal of the Geological Society, London, 148, 17-27. - - , DEMPSTER, T.J., BLUCK, B.J. & TANNER, P.W.G. 1989, A high precision U-Pb age for the Ben Vuirich Granite: implications for the evolution of the Scottish Dalradian Group. Journal of the Geological Society, London, 146, 789-798. SOPER, N,J. & BABER, A.J., 1982. A model for the deep structure of the Moine thrust zone. Journal of the Geological Society, London, 139, 127-138. SPRAY, J.G. & WILLIAMS,G.D. 1980. The sub-ophiolite metamorphic rocks of the Ballantrae Igneous Complex. Journal of the Geological Society, London, 137, 359-368. STONE, P., FLOYD, J.D. BARNES, R.P. & L1NTERN, B.C. 1987. A sequental back-arc and foreland basin thrust duplex model for the Southern Uplands, Journal of the Geological Society, London, 144, 753-764. STORETVEDT, K.M. 1987. Major late Caledonian and Hercynian shear movements on the Great Glen fault. Tectonophysics, 143, 252-267. SYLVESTER, A.G. 1988. Strike-slip faults. Geological Society of America Bulletin, 100, 1666-1703. TALIAFERRO, N.L. 1941. Geological history and structure of the central Coast

GREAT

GLEN

FAULT

65

Ranges of Calilifornia. In: JENKINS, O.P. (ed.) Geologic" formation and economic development of the oil and gas fields of California, Part 2. Geology of California and the occurrence of gas. California Division of Mines Bulletin, 118, 119-163. TANNER, P.W.G. 1994. Caledonian Terrane relationships in Britain: Programme with Abstracts. B.G.S. Keyworth, 10. & LESLIE, A.G. 1994. A pre-D2 for the 590 Ma Ben Vuirich Granite in the Dalradian of Scotland. Journal of the Geological Society, London, 151, 209-212. THIRLWALL, M. 1988. Geochronology af Late Caledonian magmatism in Northern Britain. Journal of the Geological Society, London, 145, 951-968. TORSVIK, T. 1984. Palaeomagnetism of the Foyers and Strontian granites, Scotland. Physics of the earth and Planetary Interiors, 36, 163-177. TREAGUS, J. E. 1987. The structural evolution of the Dalradian of the Central Highlands of Scotland. Transactions of the Royal Society of Edinburgh, Earth Science, 78, 1-15. TRELOAR, P.J. BLUCK, B.J., BOWES, D.R. & DL'DEK, A. 1980. Hornblendegarnet metapyroxenite beneath serpentinite in the Ballantrae complex of SW Scotland and its bearing on the depth of provenance of obucted ocean lithosphere. Transactions of the Royal society of Edinburgh, Earth Science, 71, 201-212 VAN DE VOO, R. & S¢OTESE, C. 1981. Paleomagnetic evidence for a large (sinistral) offset along the great Glen fault during the Carboniferous time. Geology, 9, 583-589. WALTON, E.K. & OLIVER, G.J.H. 1991. Lower Palaeozoic-stratigraphy, In: CRAIG, G.Y. (ed.) Geology of Scotland 3rd Edition. Geological Society, London. 161-193. WILLIAMS, H. & SMYTH, W.R. 1973. Metamorphic aureoles beneath ophiolite suites and alpine peridotites: tectonic implications with West Newfoundland examples. American Journal of Science, 273, 594-621. WINCHESTER, J. A. 1992. Comment and Reply on 'Exotic metamorphic terranes in the Caledonides: Tectonic history of the Dalradian block, Scotland' Geology, 20, 764-765. -

-

From

QJGS, ] 0 2 , 41-42. THE GREAT GLEN FAULT BY WILY.TAM QUARRIER KENNEDY, D.SC. F.G.S.

Read 8 February, 1939 [PT,AWm I I I ] CONTENTS I. I n t r o d u c t i o n . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . I I . GeneraI features of t h e dislocation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . (a) S t r u c t u r a l c h a r a c t e r s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . (b) Topographic effects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . (c) Seismic a c t i v i t y . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . I I I . S u m m a r y of fault-line geology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . IV. The n a t u r e of t h e displacement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . (a) General discussion . . . . . . . . . . . . . . . . . . ~. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . (b) E v i d e n c e of lateral displacement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . V. I n t e r p r e t a t i o n of m o v e m e n t . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . (a) Age of t h e m a i n (lateral) displacement . . . . . . . . . . . . . . . . . . . . . . . . (b) D y n a m i c s of m o v e m e n t . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . (c) Tectonic significance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . VI. List of works to which reference is made . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Page 42 43 43 46 46 47 52 52 54 64 64 67 70 71

S~r~ARr The powerful dislocation which intersects Scotland along the line of the Great Glen has, in the past, been r e g a r d e d b y most geologists as a n o r m a l or dip-slip fault with a p r e d o m i n a n t vertical d o w n t h r o w to the south-east• A reconsideration of the ~atire problem n o w suggests t h a t this view is no longer tenable and t h a t the dislocation is, in reality, a lateral-slip or wrench fault with a horizontal displacement of approximately 65 miles. Such an interpretation is s u p p o r t e d b y several independent lines of evidence, as follows : - • (1) The dislocation possesses physical characters unlike those of most n o r m a l faults b u t similar to the g r e a t strike-slip shears of the California Coast Range. (2) I t belongs to t h e same s y s t e m as the S t r a t h c o n o n , E r i c h t - L a i d o n a n d Loch Tay faults, all of which h a v e p r o v e d lateral displacements of up to 5 miles. (3) I t displaces t h e g r e a t belt of regional injection which affects the Moine Schists of the n o r t h e r n a n d G r a m p i a n Highlands, the n a t u r e a n d a m o u n t of the displacement being consistent w i t h lateral shift b u t not with vertical d o w n t h r o w . (4) I t similarly displaces t h e m e t a m o r p h i c zones of t h e H i g h l a n d s in an equally significant m a n n e r . (5) I t t r u n c a t e s t h e S t r o n t i a n Granite, the southern p o r t i o n of which, according to the detailed s t r u c t u r a l evidence, is missing. The missing portion, moreover, can be identified in t h e F o y e r s mass which outcrops on the o t h e r side of the fault-line some 65 miles to t h e n o r t h - e a s t a n d is similarly t r u n c a t e d b y the fault. These two m a j o r Caledonian intrusions consist of identical rock t y p e s a n d are s t r u c t u r a l l y homologous. (6) Finally, t h e occurrence of Lewisian a n d Torridonian rocks in I s l a y and Colonsay and t h e presence of t h e Moine Thrust-plane in the former island are more readily explained on the a s s u m p t i o n of a lateral r a t h e r t h a n a vertical displacement along the fault. A l t h o u g h the dislocation is still active, the available evidence indicates t h a t the mare lateral m o v e m e n t was accomplished prior to t h e deposition of the Upper Carboniferous s e d i m e n t s of LochaHne a n d subsequent to t h e intrusion of the S t r o n t i a n and F o y e r s (Lower Old R e d Sandstone) granites. Middle Old R e d Sandstone s t r a t a along t h e G r e a t Glen h a v e , moreover, suffered intense crushing and d e f o r m a t i o n during t h e faulting, w h i c h m u s t , therefore, be referred p a r t l y if n o t wholly to a postMiddle Old R e d S a n d s t o n e epoch. The sinistral n a t u r e of t h e displacement, i.e. t o w a r d s the south-west on the northwest side of t h e f r a c t u r e a n d t o w a r d s the north-east on its south-east side, implies t h e o p e r a t i o n of a stress s y s t e m involving regional compression a c t i n g in a general n o r t h - a n d - s o u t h direction a c c o m p a n i e d by an east-and-west relief of pressure. This is r e g a r d e d as evidence of the fact t h a t the Herc).~ian forces, to which the f o r m a t i o n of t h e G r e a t Glen F a u l t is ascribed, were a l r e a d y m o p e r a t i o n during Upper Old R e d S a n d s t o n e or L o w e r Carboniferous times.

From Le Bas, M. J. (ed.), 1995, Milestones in Geology, Geological Society, London, Memoir No. 16, 67-81 First published in Journal of the Geological Society, London, Vol. 150, 1993, pp. 227-241

P-T-t

evolution of orogenic belts and the causes of regional metamorphism MICHAEL

BROWN

Department of Geology, University o f Maryland at College Park, College Park, M D 20742, USA Abstract: Barrow (1893) introduced three important ideas that furthered understanding of metamorphic processes: (i) the use of critical index minerals in argillaceous rocks to define metamorphic zones and elucidate spatial features of regional metamorphism; (ii) the concept of progressive metamorphism; and (iii) the concept of magmatic advection of heat as a possible cause of regional metamorphism. This article expands upon these themes by reviewing our understanding of the dynamic evolution of orogenic belts as interpreted from the P-T-t paths of metamorphic rocks, and by considering the likely causes of the different kinds of regional metamorphism that we observe within orogenic belts. Understanding metamorphic rocks allows the distinction of two fundamentally different types of orogenic belt defined by relative timing of maximum T and maximum P. Orogenic belts characterized by clockwise P - T paths achieved maximum P before maximum T, the metamorphic peak normally post-dated early deformation within the belt and additional heating above the 'normal' conductive flux has been related to the amount of overthickening. By contrast, orogenic belts characterized by counterclockwise P-T paths achieved maximum T before maximum P, the metamorphic peak normally pre-dated or was synchronous with early deformation within the belt and additional heating above the 'normal' conductive flux has been related to the emplacement of plutons. Techniques used to constrain portions of P-T-t paths include: the use of mineral inclusion suites in porphyroblasts and reaction textures; thermobarometry; the use of fluid inclusions; thermodynamic approaches such as the Gibbs method; radiogenic isotope dating; fission track studies; and numerical modelling. We can utilize specific mineral parageneses in suitable rocks to determine individual P-T-t paths, and a set of P-T-t paths from one orogenic belt allows us to interpret the spatial variation in dynamic evolution of the metamorphism. Recent advances are reviewed with reference to collision metamorphism, high-temperature-low-pressure metamorphism, granulite metamorphism, and subduction zone metamorphism, and some important directions for future work are indicated.

of zones to include, in order of decreasing metamorphic grade below the staurolite zone, a garnet zone, a biotite zone and a zone of digested clastic mica, bounded at lower grade by a local margin with clastic mica, and to formalize the sillimanite zone on the map. In the discussion of Barrow's 1893 paper, Teall commented that 'It appeared p r o b a b l e . . , that the line separating the sillimanite zone from the cyanite zone was an isothermal'. This point was picked up by Barrow in his 1912 paper which included a description of how a metamorphic zone is defined, as follows: 'Proceeding across Zone i we reach rocks in which biotite is developed; this marks the commencement of Zone ii. But no matter how far to the north-west we go, brown mica is usually present in many of the rocks; so that the distance to which it extends has no zonal value; it is the line marking the first oncoming or the 'outer limit' of the mineral that gives the zonal line. Farther north-west the common (non-calcareous) garnet is met with; this, too, extends into the highest zones, so that again the line showing its first oncoming or 'outer limit' is the zonal line . . . . It gradually becomes clear that these 'outer limit' lines correspond with isothermals; i.e., they indicate the point where the rocks have been raised to a sufficiently high temperature to develop the index-mineral of the zone'. In the Quarterly Journal of the Geological Society for 1924, Tilley proposed the replacement of Barrow's 'local margin with clastic mica' and 'zone of digested clastic mica' by the zone of chlorite. Tilley argued that chlorite is a typical mineral of the lowest zone of metamorphism and

This review takes as its starting point the classic article published in the Quarterly Journal of the Geological Society in 1893 by George Barrow 'On an I N T R U S I O N of MUSCOVITE-BIOTITE GNEISS in the S O U T H E A S T E R N H I G H L A N D S of S C O T L A N D , And Its A C C O M P A N Y I N G M E T A M O R P H I S M ' . It is instructive to note, in the centennial year of this paper, that one of the questions still debated in metamorphic geology was its underlying theme, namely what is the principal cause of regional metamorphism, as reflected in Barrow's introductory statement 'It is proposed to show in the present communication that this area contains several masses of intrusive rock which are probably connected underground, and that the highly crystalline character of the surrounding schists is mainly the result of thermometamorphism'. That paper represents the first attempt to bring precision to a study of regional metamorphism by publishing a 'map of a portion of North-East Forfarshire embracing the area of outcrop of the muscovite-biotite-gneiss, and showing the zones of occurrence of the silicates of alumina which are connected with the intrusion'. The map (Barrow 1893, plate XV) illustrated a staurolite zone, a cyanite [sic] zone and indicated an area where sillimanite occurs, referred to as a sillimanite zone in the text. In the explanation of plate X V Barrow wrote 'The zones of staurolite- and cyanite-gneiss or schist really represent the variation in height above the upper limit of the underlying gneiss'. Nineteen years later Barrow extended the area of his zonal metamorphic map to the Highland fault, at the same time increasing the number 67

68

M. BROWN

that it generally decreases in amount as successive higher-grade zones are entered, interpreted to reflect its consumption during synthesis of higher-grade index minerals. Furthermore, as Tilley pointed out, white mica persists throughout, and is often very abundant in the higher-grade zones. Finally, it was Tilley who explicitly made the point that rocks of restricted chemical composition must be utilized for such zonal mapping, as follows 'The widespread distribution of sediments of argillaceous t y p e . . . at once marks o u t . . , these rocks as a suitable group in which to study successive and progressive mineral changes. Such are the rocks which Barrow used in his zonal m a p p i n g . . . ' . Thus was established the classic sequence of index minerals that commonly develop in pelites during 'normal' regional metamorphism, widely referred to as 'Barrow's zones of progressive regional metamorphism'; this particular sequence of index minerals is diagnostic of medium-P regional metamorphism which traditionally has been referred to as 'Barrovian metamorphism'. In addition to producing a map of metamorphic zones characierized by specific index minerals, Barrow (1893) considered the reasons for the great extent of the area affected and the intensity of the alterations produced by what he called the 'thermometamorphism of the Southeastern Highlands.' Barrow wrote '... we are led to the conclusion that these gneisses occur in huge sills or laccolites [sic] having approximately horizontal upper s u r f a c e s . . , it follows that one of the chief factors in increasing or decreasing metamorphism of rocks affected must be the variation in depth of the sills below the surface.., the intensity of the metamorphism is doubtless largely due to the great depth below the surface of the rocks affected by the intrusion.' Finally, Barrow commented that ' . . . these special features m a y . . , be due to the depth in the earth's crust at which the metamorphism took place, rather than to any physical conditions peculiar to early geological t i m e . . . and strengthens Dr. Barrois's conclusion that 'regional metamorphism and contact-metamorphism are much the same thing". Barrow's depth of geological understanding is reflected in his interpretation, in his 1912 paper, of his regional metamorphic map (Barrow 1912, folding map), as follows 'A key to the form of the zones of this area of regional metamorphism is given by the 'outer limit' lines of the sillimanite zone (6) and of the biotite zone (2). • . . sillimanite-bearing rocks occur over an area of, roughly, 200 square m i l e s . . , roughly t r i a n g u l a r . . , but the line marking the outer limit of biotite is almost s t r a i g h t . . . (3) defining the limit of garnet, slowly diverges from the (2) just described. The divergence is more marked in the case of staurolite (4); its boundary begins to assume a rude parallelism with (6), defining the outer limit of sillimanite. The parallelism is still closer in the case of cyanite (5) . . . . It thus gradually becomes apparent that while the line marking the lower zones is almost straight, the intermediate zones surround lenticles of the most highly altered rocks containing sillimanite; further, the whole masses of crystalline rocks are somewhat lenticular in f o r m . . . It thus appears that the Highland a r e a . . , is essentially built up on the lines of an aureole of metamorphism around a granite intrusion; but instead of aureoles we have zones or belts which diverge more and more from a lenticular, highly crystalline nucleus, until the lines bordering the lower temperature zones are nearly straight.' Thus, Barrow was firmly convinced that this regional metamorphic terrane is

nothing more than a large-scale contact aureole around numerous small granitic intrusions; he saw no essential difference between regional and contact metamorphism• Although Barrow's work is widely known today, at least through the naming of medium-pressure regional metamorphism and associated metamorphic field gradients as 'Barrovian', it was not well known early during this century in Europe, possibly because of the rather non-informative title of his 1893 paper (above). Indeed, Goldschmidt (1916, summarized in Mason 1992), was able to map zones of progressive metamorphism in the Trondhjem district of the Norwegian Caledonides marked by the index minerals chlorite, biotite and garnet in argillaceous sediments, unaware of the earlier work of Barrow in the southeast Highland of Scotland. In a clear example of the parallel development of ideas in science, Goldschmidt had summarized the aim of his research in his inaugural lecture as Professor of Mineralogy at the University of Stockholm on 28 September 1914 (quoted in Mason 1992, from a translation by G. Kullerud) with the following 'It is . . . of great i n t e r e s t . . , to determine the physical conditions under which an individual mineral has been formed• It is of much greater i m p o r t a n c e . . , to study thoroughly a sizeable area in order to investigate the temperature-pressure distribution during a certain geological era. Such an investigation, no doubt the first of its kind, is being performed by myself in the Norwegian mountain areas, from Ryfylke (near Stavanger) to Trondhjemsfjorden, in order to determine the temperature and pressure conditions in this part of the earth's crust during the formation of the Norwegian Caledonides at the beginning of the Devonian . . . the sum of all observations gives us a picture of the temperaturepressure distribution during the formation of a mountain chain.' Indeed, Goldschmidt could not have described better the aim of modern metamorphic petrology. During the 100 years since publication of Barrow's seminal paper, Scotland has proven to be a fertile ground for the development of ideas in metamorphism. With respect to the zonal distribution of metamorphism and the quantitative estimation of P and T, the reader is referred to Chinner (1966), Atherton (1977) and Harte & Hudson (1979). Metamorphic reactions in the higher grade zones of the Barrovian type area have been investigated by McLellan (1985), who has emphasized also the importance of both sub-solidus and anatectic processes in the generation of migmatite leucosomes (McLellan 1983, 1989) interpreted by Barrow (1893) to be related to the granites by fractionation.

Orogeny and regional metamorphism Orogeny is characterized by a distinctive relationship between sedimentation, tectonic deformation, regional metamorphism and magmatism. It leads to structural inversion of sedimentary basins, mountain building and eventual exhumation of metamorphic belts, either during the same cycle or subsequently. Modern orogenic belts are located at convergent plate boundaries and along lines of continental and/or arc collisions. Orogenesis is responsible for crustal differentiation through anatexis and transfer of granitic magma from the lower crust to the upper crust, leaving behind a depleted residuum. Large tracts of the continents of the Earth are formed by rocks that have been metamorphosed on a regional scale, that is, their secondary mineral assemblages indicate that

P-T-t

E V O L U T I O N OF O R O G E N I C BELTS

these regions have been subjected to elevated temperatures and pressures at some time in their past. Generally accepted tectonic settings for regional metamorphism are continental margins associated with subduction, such as the west coast of South America; island arcs, such as those of Southeast Asia; and zones of continental collision, such as the Alps and Himalayas. Traditionally, such regional metamorphism has been separated from metamorphism that occurs in aureoles surrounding intrusives, so-called contact metamorphism, because of the scale of the area affected on the one hand and because of the spatial relationship to the intrusive heat source on the other hand. However, to maintain that metamorphism of regional extent shows no apparent relation to intrusive rocks as heat sources is to deny one of the main conclusions of Barrow's 1893 paper. Indeed, magmatic arcs at convergent plate boundaries are a locus for plutonic magmatism and, as a consequence, contact metamorphism, and such a relationship occurs on a regional scale (e.g. Barton & Hanson i989). The observation that belts of regional metamorphism typically contain abundant intrusive rocks leads to the postulate implied in Barrow's paper that intrusive rocks collectively increase the regional thermal gradient and might be a primary cause of some regional metamorphism, even though in the particular case discussed by Barrow the 'intrusive rocks' are sub-solidus and anatectic migmatites (McLellan 1983, 1989). On the other hand, Harker (1932) emphasized that regional metamorphism is characterized by both a wide areal extent and a regional-scale temperature distribution, as shown by mineral zones, independent of the distribution of individual plutonic masses. It is plausible that medium-P regional metamorphism may grade with decreasing crustal depth into regional-scale contact metamorphism; an example, likely representing such an oblique crustal section, may be the New England Appalachians from Connecticut to Maine (Tracy & Robinson 1980; Armstrong et al. 1992; De Yoreo et al. 1989a, b).

69

values of mantle heat flux, thermal conductivity and heat production.

Progressive regional m e t a m o r p h i s m

In detail, regional metamorphic terranes have been classified by their mineral assemblages into facies series types such as andalusite-sillimanite, kyanite-sillimanite and jadeite-glaucophane (Miyashiro 1961). It is implicit in this classification that an entire metamorphic terrane can be described by a single geothermal gradient and that higher-grade mineral assemblages develop from lower-grade ones similar to those now found along the present erosion surface. This view derives directly from Barrow (1893) who argued that the sequential development of a coarse-grained sillimanite-gneiss from a kyanite-gneiss, and the kyanitegneiss from a staurolite-schist within one continuous stratigraphic horizon, presented a conclusive proof of progressive metamorphism. However, metamorphism is not a static process but an evolutionary one. With our acceptance of a dynamic tectonic environment for regional metamorphism, we realize now that rocks follow more complex routes in P - T space, reflecting burial, heating and exhumation. This does not negate the concept of progressive metamorphism, and it is likely that, with the exception of circumstances in which heating rates are relatively rapid, the progressive model for medium- to high-grade segments of P - T paths is essentially correct, but the sequential change is not simply the one observed by following the sequence of assemblages along the metamorphic field gradient. The traditional view has been replaced by one in which individual rocks (e.g. Thompson & England 1984) and minerals (e.g. Spear & Selverstone 1983) can be used to derive paths in P - T space, that can be related to the tectonic setting (e.g. England & Thompson 1984). The derivation of such paths along the length and breadth of an orogenic belt enables us to unravel the three-dimensional reality of orogenic processes.

Paired m e t a m o r p h i c belts

Some metamorphic terranes have developed as paired belts, in which one member is characterized by high-temperature metamorphism and the development of migmatites and anatectic granites, and is interpreted as having been developed at a site of high heat flow such as that beneath an associated volcanic arc; the other member is characterized by blueschists and eclogites that indicate relatively low geothermal gradients and relatively high-pressure conditions, interpreted as having been developed at a site of low heat flow such as a subduction zone. Commonly, the first member is referred to as a low-pressure-high-temperature metamorphic belt, although many such belts exhibit a high-pressure history prior to the development of the final low-pressure mineral assemblage as in the Abukuma Plateau (Kano & Kuroda 1968; Hiroi & Kishi 1989) and in Southern Brittany (Cogn6 1960; Jones & Brown 1990). Intrusions advecting heat are commonly proposed as the cause of the thermal anomaly for these high-temperature metamorphic belts because of the abnormally high geothermal gradients implied (De Yoreo et al. 1991; Furlong et al. 1991). However, Treloar & Brown (1990) have shown that moderate overthickening of sedimentary basins during structural inversion may lead to high-temperature metamorphism at mid-to-lower crustal depths for reasonable

P - T - t paths of metamorphism Two fundamentally different types of orogenic belt are distinguished by relative timing of maximum T and maximum P, as revealed by metamorphic rocks within the belt. One type of orogenic belt is characterized by an evolutionary path in P - T space that is clockwise (Fig. 1; CW paths). Orogenic belts of this type are generated by basin inversion or crustal thickening followed by erosional exhumation and/or extensional thinning and/or lithospheric delamination and orogenic collapse (Oxburgh & Turcotte 1974; Bird et al. 1975; Houseman et al. 1981; Thompson 1981; Thompson & England 1984; Thompson & Ridley 1987). Orogenic belts characterized by clockwise P - T paths achieved maximum P before maximum T, and the metamorphic peak normally post-dated early deformation within the belt. Such an evolutionary path will lead to decompression dehydration-melting of common crustal rock types (Thompson 1982, 1990; Jones & Brown 1990), and may lead to granite magmatism which is a consequence of the regional metamorphism (e.g. Patin6-Douce et al. 1990). Experiments on natural rock compositions (Le Breton & Thompson 1988; Rushmer 1991) and the results of thermal models (De Yoreo et al. 1989a) indicate significant volumes of crustal melt can be generated through crustal thickening.

70

M. BROWN

500

700

900

1100 70

CW 17.5

15.0

CWa

50

12.5 N

E

.{3

10.0 ~[_

a

09

30

7.5

5.0

2.5

Ms(ss) Ab Q t z / \ a5 ~1 ~3 ~ CCW

for

500

/ BA

10

TC

AIs V

700

900

Indeed, significant amounts of melt are generated for overthickening as small as 10-15 km (see De Yoreo et al. 1989a) and De Yoreo (1988) has shown that partially molten (>0.3 melt fraction) sections of crust substantially thicker than 1 km may be generated in less than 40 Ma. Crustal anatexis may be a normal consequence of some types of high-temperature metamorphism, particular those which involve decompression at high temperature (e.g. Jones & Brown 1990). The relationship between anatectic migmatites and higher-level granites has been considered by D'Lemos et al. (1992), and the thermal consequences of crustal melting have been considered by Haugerud & Zen (1991) who make the point that the complement of a high-level zone of heating by magmatic injection is a zone of retarded heating in the middle-to-lower crust because melting buffers the geothermal gradient. The issue of magmatic advection driving high-temperature metamorphism v. crustal melting to derive granites subsequently emplaced into the high-temperature metamorphic belt needs to be resolved on an individual basis. -1"he other type 0i~ orogenic belt is characterized by an evolutionary path in P - T space which has a counterclockwise direction (Fig. 1; CCW path). In such an evolution, heating preceded crustal thickening or the two may have

1100

Fig. 1. Pressure-temperature diagram to show: (1) Some of the initial melting reactions in metapelites (from Thompson 1990, and primary references therein). (2) Initial melting reactions for amphibolites, including the H20saturated solidus for olivine tholeiite and an approximate H20-saturated solidus for quartz tholeiite based upon albite + quartz + H20 (after Thompson 1990 and primary references therein). (3) P - T - t paths (CW; 50 km and 70 km depth after thickening) for thickening of continental crust from 35 km to 70 km, followed by erosional thinning in 100 million years, after a post-thickening isobaric metamorphism of 20 million years for an initially 'hot' geotherm that certainly reaches granulite facies conditions (after Thompson 1990 and primary references therein). (4) P - T - t path (CCW) for heating followed by crustal thickening and near isobaric cooling (after Thompson 1990 and primary references therein). CW, clockwise path in P - T space; CCW, counterclockwise path in P-T space; GWS, granite wet solidus; BWS, basalt wet solidus; IAT, island arc tholeiite; BA, alkali basalt. Reactions and processes that occur as a consequence of evolution along these paths (a-j and p-y) are discussed in Brown (1993).

gone hand-in-hand. Models to generate such counterclockwise paths include intraplating of mantle-derived magmas (Bohlen 1987, 1991; Bohlen & Mezger 1989) and crustal thickening with concomitant mantle lithosphere thinning (Loosveld & Etheridge 1990; Sandiford & Powell 1991). Orogenic belts characterized by counterclockwise P - T paths achieved maximum T before maximum P, and the metamorphic peak normally pre-dated or was synchronous with early deformation within the belt. Once again, such a process will generate dehydration melting (Thompson 1990) and may lead to granite production as a consequence of regional metamorphism (Collins & Vernon 1991). Clockwise and counterclockwise paths may occur in adjacent parts of the same orogenic belt, as exemplified by the Acadian orogenic events in the northern Appalachians (Tracy & Robinson 1980; Schumacher et al. 1989; Armstrong et al. 1992), and illustrated in Fig. 2. In order to unravel the history of an orogenic belt we n e e d knowledge of the change of pressure and temperature with time, and on the relationship of these to deformation. Information that will enable us to address this issue potentially includes the following: data on the type and age of protolith lithologies and if possible the tectonic environment of their formation; data on the P - T evolution

P-T-t

E V O L U T I O N OF O R O G E N I C BELTS

71

14 12

Western Acadian

10

L_

n

0 0

200

400

600

800

T (°C) Fig. 2. P - T diagram to show postulated typical P - T trajectories for each of the metamorphic realms discussed by Armstrong et al. (1992) from central and western New England, USA. The patterned ovals indicate the approximate part of each trajectory at which the peak P - T conditions were recorded. Note: for broad metamorphic realms such as the Western Acadian and Taconian, what Armstrong et al. have shown is only one typical path from a nested family of similar paths.

of the metamorphic rocks; data on the relationship between metamorphic mineral growth and deformation; age data to constrain the timing of prograde and peak metamorphic conditions; age data from pre-, syn- and post-orogenic plutons; data on cooling and exhumation using various mineral geochronometers; and the unroofing history of the belt as reflected in the erosional debris deposited in its foreland. Although a complete understanding of the evolutionary history of an orogenic belt requires much or all of this information, studies to date are rarely so complete. Nevertheless, our understanding of the relationship between metamorphism and tectonics has increased dramatically during the past few years through the combination of several of these types of field, analytical, and numerical investigations of metamorphic P - T - t paths. M i c r o s t r u c t u r a l studies a n d the use o f textures

The identification of clockwise v. counterclockwise paths requires the relationship between mineral growth and deformation to be established from textural relationships; an example is given in Fig. 3. In effect, we need to recognize both a sequence of overprinting metamorphic events and the relationship between a particular mineral assemblage and the deformation phases that have affected the rock during thickening and exhumation. Textural analysis enables us to establish relative timing of metamorphic and deformational events (see Fig. 3), which may then be quantified using the increasingly sophisticated isotopic techniques that can be applied to individual minerals. Pioneering work on microstructural studies, in particular the relationship between porphyroblast growth and matrix development, was undertaken by Zwart in the Pyrenees (1962) and Johnson in the Scottish Highlands (1963). Careful petrographic analysis

Fig. 3. Relatively straight inclusion trails in the centre of kyanite (Ky) porphyroblast curve through the rim (lines emphasize trail orientation) and are consistent with early syn-foliation growth, a conclusion supported by the overall shape of the porphyroblast which exhibits small 'tails' that have grown into the foliation at the top and bottom. The dominant schistosity which encloses the kyanite porphyroblast is $2 in this particular rock, which is from the Port aux Basques Complex in Southwest Newfoundland, Canada. Long dimension of field of view is 4.5 mm, crossed polars.

is critical, yet many metamorphic textures are ambiguous, and interpretations consequently may be subjective; for example, we cannot yet agree on whether or not porphyroblasts rotate during deformation or, more likely, accept that in some cases porphyroblasts have rotated, but in other cases they have not (see the debate between Passchier et al. (1992) and Bell et al. (1992), and primary references therein). Suitable textures to elucidate P - T - t paths include relict and replacement features, porphyroblast-inclusion-matrix relationships and high strain zones cutting through metamorphic belts that may have developed during the exhumation part of the P - T - t path. Further, it is important to decide which minerals, if any, might represent equilibrium assemblages that can be used in quantitative calculation of P - T conditions at points on the P - T - t path. Minerals of particular growth stages can be used to elucidate t at points on the P - T - t path. Thus, a combination of microstructural, thermobarometric and geochronological studies will allow the identification of a well-constrained P - T - t - d e f o r m a t i o n path. Particularly important, but largely unknown, is information on the rates of processes such as heating, mineral reactions, partial melting and tectonic

72

M. BROWN

deformation; one example of such information is given by Mezger (1990).

Equilibrium v. disequilibrium The main development that has occurred in metamorphic petrology during the past twenty years is the realization that our previous obsession with 'equilibrium' ignores the evidence of a dynamic evolution represented by mineralogical and chemical 'disequilibrium'. Equilibrium is the basis of the metamorphic facies concept, proposed by Eskola in 1914 and developed by Goldschmidt and Eskola, in particular during a visit by Eskola to work with Goldschmidt in Oslo during 1919-1920 (Eskola 1920); and it is the sequence of metamorphic facies exposed along the erosion surface through a metamorphic belt that represents the metamorphic facies series of Miyashiro (1961). The thermodynamic basis for the metamorphic facies concept was provided by Thompson (1955) which set the ground for quantitative geothermobarometrical work that has proven so profitable in the quantification of metamorphic P and T over the last three decades. One basic aim of modern metamorphic petrology is to relate observed mineral assemblages to

P - T - t history and to utilize this information to distinguish between various possible tectonic processes that may operate in orogenic belts. To assess 'peak' P - T history, we have relied on thermodynamics and phase equilibria, and we have made t h e assumption that either the mineral assemblages, through use of a petrogenetic grid, or the mineral chemistries, utilizing thermobarometry, will reveal a P and T that are geologically significant. If equilibrium is achieved and preserved over significant portions of the mineral assemblage, as reflected by completely homogeneous mineral chemistry and straight boundaries between adjacent grains, then this procedure will be valid; but by the very nature of equilibrium, the effects of any previous or subsequent processes are lost completely. Thus, the dynamics or history of a rock are relegated to a secondary role. It follows from this discussion that a study of a whole series of samples from a single metamorphic belt that are thought to reflect equilibrium can only yield individual points on each of a set of P - T - t paths and can yield no information about the dynamic aspects of the tectonic processes involved. Disequilibrium features in rocks, however, reveal dynamic history because features such as inclusions in porphyroblasts, replacement textures and

!

0

12 Fig. 4. (A) Partially resorbed garnet (Grt) from granulite facies metapelite, Sharyzhalgay complex, Lake Baikal, Russia, exhibits a partial orthopyroxene necklace (Opx) that outlines the original garnet porphyroblast (dashed line). Inside the orthopyroxene necklace, a symplectite (Sym) cdmposed of cordierite, orthopyroxene and biotite has partially resorbed the garnet. This delicate texture often is interpreted to represent decompression, and indicates further that any deformation associated with decompression was concentrated in rocks other than this one since the texture would not have survived significant ductile strain. Long dimension of field of view is 13.5 mm, plane light. (B) Detail of rock shown in (A) to illustrate two reactions preserved by the textures. First, garnet (Grt) and quartz (Qtz) reacted to give granular orthopyroxene (Opx) and plagioclase (PI), orthopyroxene nucleated against quartz and plagioclase nucleated against garnet (now replaced by subsequent symplectite development). Second, during decompression garnet and quartz, probably in the presence of melt, have reacted to cordierite (Crd), orthopyroxene and biotite (Bt) intergrown as a symplectite. This reaction has not gone to completion which suggests that decompression was relatively rapid. Long dimension of field of view is 2.0 mm, plane light.

P-T-t

EVOLUTION OF O R O G E N I C BELTS

mineral zoning reflect a succession of changing conditions along a P - T - t path; an example is given in Fig. 4. These features have been utilized in an increasing number of studies during the past ten years; see, for example, information in Tracy (1987) and Peacock (1991b) and the many examples of P - T - t paths collected together in the book 'Evolution of metamorphic belts' (Daly et al. 1989). M o d e l l i n g studies

Possible causes of temperature change in the crust include variation in the conductive heat flux from the mantle, advection of heat through emplacement of mantle-derived magma into the crust, variation in the heat productivity from radioactive decay, physico-chemical processes that involve thermal energy such as metamorphic reactions and melting, including magmatic transfer of heat from the lower crust to the upper crust, depression of the crust through thickening and elevation of the crust through thinning. The dependence of the temperature field upon dynamic processes and the consequent pressure-temperature-time evolution of metamorphic rocks have been investigated through thermal modelling of orogenic belts (for a review of early work see Thompson 1981). This modelling has shown that the P - T - t path a rock follows is the result of a complex interaction between tectonic processes and heat flow/heat generation/heat transfer mechanisms. The regional scale complexity of metamorphic belts reflects a variety of factors that include lateral cooling at depth along major tectonic boundaries (Harte & Dempster 1987) and spatial variation in thermal conditions and structural history (e.g. Jaupart & Provost 1985; Allen & Chamberlain 1989; Sonder & Chamberlain 1992). These variations were recognized by Turner (1981) and Richardson (1970) and are acknowledged in most subsequent modelling studies, which by their nature are inherently simplified, either because of our incomplete knowledge of the physical properties of rocks, or by reduction of the number of dimensions from three to two, or even to one, or by regarding certain kinds of geological processes as instantaneous, such as thrusting during crustal thickening (England & Thompson 1984; Shi & Wang 1987; Haugerud 1989; Chamberlain & Sonder 1990; Peacock 1990, 1991a). Nonetheless, we are approaching a first-order understanding of those heat and mass transfer processes that drive metamorphism, exemplified by the saw-tooth geothermal profile that results from models of crustal thickening by instantaneous thrusting and its subsequent evolution with time to a new stable geothermal gradient (e.g. Thompson 1981; England & Thompson 1984). The tectonic history of a region, the thermal history of rocks within that region, and the P - T - t evolution of those rocks are related. In a one-dimensional analysis, by which we mean that the crust is regarded as composed of columns of rock with equal physical properties and between which there is no lateral heat transfer, this relationship can be represented as a surface in P - T - t space. The tectonometamorphic histories of rocks are represented by lines on this surface (Haugerud 1989, fig. 1). The projections of such rock histories onto the three two-dimensional planes of the P - T - t box represent the T-t paths described by the geochronologist, the P - T paths studied by the metamorphic petrologist and the P - t paths inferred by the tectonicist. Of course, the strict correlation of T and t refers to the cooling paths of a metamorphic belt; minerals that grow below their

73

blocking temperatures for which an estimate of pressure of formation can be made may be used to relate P and t. One approach to understanding the P - T - t evolution of rocks in particular tectonic settings and for particular heat-transfer mechanisms is that of forward modelling of the thermal response to tectonism, as demonstrated, for example, by England & Thompson (1984), Haugerud (1989) and Chamberlain & Sonder (1990). However, the metamorphic petrologist investigates the inverse problem, that is the determination of P, T and t from rocks, and from these data infers a tectonic history. A combination of the forward approach used by the modeller and the inverse approach used by the petrologist in understanding the P - T - t evolution of rocks will lead to a better understanding of the history of orogenic belts than either alone. Further, thermal modelling allows us to determine the potential for reaction by giving us rates and duration of overstep of particular equilibrium boundaries, it allows us to predict when reactions should be frozen-in and to determine the amount of bulk difusional resetting of phases such as garnet that are commonly compositionally zoned. Outcomes

A variety of different techniques has been used to constrain portions of P - T - t paths. These techniques include: the use of mineral inclusion suites and reaction textures, for example, in the Wopmay orogenic belt of northwest Canada (St-Onge 1987), in Southern Brittany (Jones & Brown 1990) and in British Columbia and New Hampshire (Selverstone & Chamberlain 1990); thermobarometry, for example, in Antarctica (Harley et al. 1990; Harley & Fitzsimons 1991) and in the Arunta Complex in central Australia (Goscombe 1992); thermodynamic approaches such as the Gibbs method, for example utilizing garnet (Spear et al. 1984; Spear 1988, 1989); radiogenic isotope dating, for example in the Himalayas (Zeitler 1989), the Pikwitonei granulite domain in the Canadian Shield (Mezger 1989) and in the New England Appalachians (Wintsch et al. 1992); the use of fluid inclusions to constrain physical conditions during exhumation (for example, Hollister et al. 1979); and numerical modelling (for example, Thompson & England 1984; Chamberlain & Sonder 1990; Peacock 1990). A number of recent reviews address methods of obtaining P - T - t path information from metamorphic terranes, with many examples, to which the reader is referred for more detailed information (.Ghent et al. 1988; Daly et al. 1989; Harley 1989; Spear & Peacock 1989; Haugerud & Zen 1991; Hodges 1991; Jamieson 1991). Our ability to separate partially overprinted 'equilibrium' mineral assemblages in 'disequilibrium' rocks has enabled the use of thermobarometry to determine different sets of P - T data for individual rocks, although one must proceed with due caution (Selverstone & Chamberlain 1990). The locus of peak or slightly post-peak P - T conditions as preserved by the mineral assemblages--the metamorphic facies series (Miyashiro 1961), metamorphic geotherm (England & Richardson 1977), piezothermic array (Richardson & England 1979), P - T array (Thompson & England 1984), metamorphic field gradient (Spear et al. 1984) or set of metamorphic zones (Harte & Dempster 1987)--results from the intersection of P - T paths for individual rocks with the erosion surface (England & Richardson 1977). Thus, each location along the metamorphic field gradient

74

M. B R O W N

represents a unique point of pressure, temperature and time. Since these points generally will not be contemporaneous, the metamorphic field gradient is necessarily the locus of diachronous P - T conditions. P - T - t paths deduced from petrographic, thermobarometric, fluid inclusion and geochronometric data can be the result of a single orogenic cycle or may be the cumulative effect of several orogenic cycles, in which case the apparent P - T - t path determined from evidence in the rocks may have no real meaning, being a composite of information partially preserved from more than one path. Additionally, single-cycle P - T - t paths are commonly represented as simple smooth curves, whereas in reality they are likely to have a more complex form because rates of burial and uplift, with or without magmatic heating, vary during the orogenic cycle or because of the effects of progressive deformation (e.g. Hames et al. 1990; Dempster 1985; Dempster & Harte 1986). An example of the complexity that occurs in nature and of our increasing ability to resolve such complexity is provided by Goscombe (1992) who has separated an earlier counterclockwise P - T - t evolution from a subsequent clockwise P - T - t evolution within the polymetamorphic Arunta Complex in central Australia. One rapidly advancing area is in the development and application of a wider range of mineral geochronometers; this has proven particularly important in the elucidation of prograde parts of P - T - t paths and exhumation processes. For example, Mezger et al. (1989) have utilized the U-Pb system on garnet to date points during the prograde evolution of parts of the Pikwitonei granulite domain in the Canadian Shield. With respect to the exhumation process, the Adirondack Mountains of New York cooled at time-integrated rates of c. 1.5 °C/Ma for at least 150 million years following the last phase of high-grade metamorphism (Fig. 5), suggesting only limited vertical tectonic displacement and approximate isostatic equilibrium (Mezger et al. 1991). By contrast, the Southern Brittany Migmatite Belt T (°C) 700

Central x \ \Highlands.,,.

G~u'net~,]

_ ~•

600

.

\

.

\x

Titanite

500

Homblende

Southern Highlands

400

•-- ..

Biotite

Rutile # Rutile

300 200

Cooling History Adirondack Highlands

I

1050

I

I

I

1000

950

900

I

850 800 time (Ma)

Fig. 5. A possible temperature-time cooling path for the Adirondack Highlands based on mineral ages of garnet, monazite, titanite, hornblende, rutile and biotite. The solid vertical bars and shaded areas indicate the mineral ages or range in mineral ages and the range in estimated closure temperatures. The closure temperatures used for the different minerals are as follows: garnet, U-Pb, >800 °C; monazite, U-Pb, 700-650 °C; titanite, U-Pb, 670-500 °C (depending on grain size); hornblende, 4°Ar-39Ar, 500-400°C; rutile, U-Pb, 430-380 °C; and, biotite, 4°Ar-39Ar, c. 300 °C. Data from Mezger et al. (1991) to which the reader is referred for further discussion.

900 800

I

I

I

700

' ~ 1

600

Cooling history Southern Brittany

500

Monazite

~-]

Hornblende

Muscovite

400 300

I

Zircon

--]

Toc

Biotite

200 100

time (Ma)

I

i Apatite

I I I I 400 350 300 250 Fig. 6. A possible temperature-time cooling path for the Southern Brittany Migmatite Belt based on mineral ages of garnet, monazite, hornblende, muscovite, biotite, and apatite. The boxes indicate the range of mineral ages and the range in estimated closure temperatures, based upon likely peak metamorphic temperatures for zircon and fast cooling for hornblende, muscovite, biotite and apatite. The closure temperatures used for the different minerals are as follows: zircon, U-Pb, c. 775 °C; monazite, U-Pb, 730-640 °C; hornblende, 4°Ar-a9Ar, c. 500 °C; muscovite, 4°Ar39Ar, c. 400 °C; biotite, Rb-Sr, c. 325 °C; and, apatite fission track, c. 125 °C. Data are from Peucat (1983) and Dallmeyer & Brown (1992). 0

(Jones & Brown 1990) exhibits extremely rapid cooling with time-integrated rates of c. 50°C/Ma. during 10 million years. (Fig. 6), suggesting tectonic exhumation (Dallmeyer & Brown 1992). Barrow's 1893 paper was concerned with spatial variations in metamorphic grade across an orogenic belt and the underlying causes of regional metamorphism. Much of modern work, however, deals with evolving P-T conditions within individual rock samples or within small parts of orogenic belts. This contrast reflects a change in emphasis, but we are beginning to return more to the spatial variation in P - T - t evolution as the methodology becomes more routine and its application more widespread across orogenic belts. As examples, regional variations across the Scottish Highlands have been summarized succinctly by Harte & Hudson (1979; see also Dempster 1985 and Dempster & Hafte 1986), and regional variations in depth of burial and the implications for denudation history of the southern New England Appalachians have been discussed by Zen (1991; see also Hames et al. 1990). A major feature of European orogenic belts is the great variety of facies types side-by-side or superimposed. For example, the Variscan belt contains very-high-pressure rocks over a wide area, from Poland to Portugal, some of which have been overprinted by a low-pressure-high-temperature metamorphism, reflected in orthopyroxene-bearing assemblages that overprint eclogitefacies mineralogies. The spatial relationships may reflect tectonic dismemberment of a paired metamorphic belt, and even though the structural history may be locally complex, the overall thermal pattern remains deeply ingrained.

R e c e n t a d v a n c e s in r e g i o n a l m e t a m o r p h i s m Any attempt to 'cubbyhole' regional metamorphism into types will produce some overlap between tectonic setting,

P-T-t

E V O L U T I O N OF O R O G E N I C BELTS

metamorphic processes and metamorphic rock-types. Nonetheless, it is convenient to highlight recent advances in our understanding of different aspects of regional metamorphism by considering four particular types of metamorphism, acknowledging that the distinction between them is not perfect and that my selection of both types and highlights is a personal one. Collision metamorphism

Metamorphism that is a result of substantial crustal thickening due to collisions between continental elements and/or arcs dominates the literature on regional metamorphism. P - T - t paths that result from collision metamorphism are clockwise in P - T - t space and their general characteristics are well understood, both from the standpoint of actual examples, such as the Appalachians (Armstrong et al. 1992, and references therein), the Caledonides (Anderson et al. 1992, and references therein) and the Himalayas (Searle et al. 1992, and references therein) and from a theoretical standpoint in terms of the heat transfer processes involved (England & Thompson 1984, and references therein). Recently, there has been a dramatic increase in the use of 4°Ar/39Ar methods to determine the polymetamorphic and exhumation histories of these metamorphic terranes (e.g. McDougall & Harrison 1988). Also, during the past decade the emphasis has shifted from the Appalachians and Caledonides to the Himalayas; although the former remain important both historically in the development of ideas and in terms of ongoing research, some of the most exciting advances have resulted from detailed work in the Himalayas. The Himalayan mountain chain developed as a result of the Eocene to Recent collision between India and Eurasia; the ongoing convergence has led to the exhumation and exposure of the high-grade metamorphic core of the Himalayan orogen. These high-grade metamorphic rocks of the Greater Himalaya lie above the north-dipping Main Central Thrust system, a major intracontinental thrust system that accommodated a significant proportion of the total shortening across the orogen (Le Fort 1975). The thrust system separates the high-grade metamorphic rocks from generally lower grade metasedimentary rocks of the Lesser Himalaya, although the metamorphic grade of this unit increases to the east (e.g. Swapp & Hollister 1991). Further, metamorphic studies along the Himalayan mountain chain show that high-grade rocks occur at structurally shallower levels than lower-grade rocks, a phenomenon referred to as 'inverted metamorphism' (Gansser 1964). One of the issues that remains unclear is the extent to which polyphase metamorphism is reflected in the observed mineral assemblages in the high-grade metamorphic core of the Himalayan orogen (Hodges et al. 1988; Inger & Harris 1992), and the degree to which the metamorphism is diachronous, propagating to the south (Searle et al. 1992). Syn-metamorphic displacements on fault systems result in thermal decoupling across such systems and the Main Central Thrust system probably was active during high-grade metamorphism and anatexis of the Greater Himalaya rocks (Hodges et al. 1988; Hubbard & Harrison 1989; Searle et al. 1992). This high-grade metamorphic core is .truncated by north-dipping, low-angle normal faults and shear zones of the South Tibetan detachment system (Burg et al. 1984; Burchfiel et al. 1992), structures that developed

75

under the influence of gravity to moderate the extreme topographic and crustal thickness gradients produced by displacement on contractional structures (Burchfiel & Royden 1985). Recent work suggests that displacement on the South Tibetan detachment system was penecontemporaneous with contractional deformation on the Main Central thrust zone (Burchfiel et al. 1992; Hodges et al. 1992; Searle et al. 1992); the thermal consequences of this complex tectonic activity are described by Inger & Harris (1992) and Hodges et al. (1993). Further, Hodges et al. (1993) emphasize that radically different P - T - t paths can be found at different structural levels beneath such extensional structures; this leads them to suggest that P - T - t paths in collision belts may be complex and thus not uniquely diagnostic of the unroofing mechanism, particularly in regions characterized by penecontemporaneous extension and shortening. To the north of the South Tibetan detachment system, in the Tibetan zone, Neogene metamorphic core complexes occur within the Tibetan sedimentary sequence. One example, the Kangmar dome, has been studied in some detail by Chen et al. (1990). On the basis of structural analysis, they infer that the domal structure formed as a consequence of extensional deformation; this draws inevitable comparisons with the Tertiary metamorphic core complexes of the northwestern American Cordillera. High-temperature-low-pressure

metamorphism

High-temperature-low-pressure metamorphic terranes have been explained in a number of different ways; however, a common cause for this type of regional metamorphism is not going to be established because fundamentally different tectonic processes result in similar P - T conditions. High-temperature-low-pressure metamorphism is perceived as a problem because of the extreme thermal anomaly implied by calculated geothermal gradients, commonly in the range 60-150 °C/km (De Yoreo et al. 1991), hence this type of metamorphism is a 'freak of nature'. As a direct consequence of this, the role of advective heat transfer in the formation of high-temperature-low-pressure metamorphic belts has been over emphasized and tectonic transport of heat to change the thermal gradient with time, quite plausible given deformation at reasonable rates and the poor thermal conductivity of rocks, has been underestimated. There are three endmember processes that result in high-temperature-low-pressure metamorphism: (i) contractional deformation and crustal thickening, for example structural inversion of a sedimentary basin developed on thin lithosphere, which results in a clockwise path in P - T space and subsequent tectonic exhumation that produces high-temperature-low-pressure metamorphism as a consequence of high-temperature decompression (e.g. Jones & Brown 1989; Jones & Brown 1990; Treloar & Brown 1990; Dallmeyer & Brown 1992; Thompson 1989); (ii) regionalscale contact metamorphism which results in isobaric heating and cooling (e.g. Lux et al. 1986; Barton & Hanson 1989; De Yoreo et al. 1989a, b); and, (iii) magmatic advection of heat during crustal thickening, which results in a counterclockwise P - T path and either isobaric cooling from the high-temperature-low-pressure peak or even increasing P during cooling (e.g. Wells 1980; Bohlen 1987; Vernon et al. 1990; Collins & Vernon 1991). Metamorphism associated with crustal extension has

76

M. BROWN

become widely recognized in the past decade, both in metamorphic core complexes and in collision belts, and the role of extension in the generation of high-temperaturelow-pressure metamorphism should not be underestimated; Peacock (1991b) gives a good summary. As we have seen in the section on collisional metamorphism, a hightemperature-low-pressure metamorphic event is commonly superimposed on an earlier higher pressure metamorphism as a consequence of extensional collapse of an overthickened orogenic belt (Dewey 1988; Inger & Harris 1992; Hodges et al. 1993). Variscan massifs of the Pyrenees are characterized by high-temperature-low-pressure metamorphism that provoked much discussion in the 1980s (e.g. Wickham & Oxburgh 1985; Lux et al. 1986; Wickham 1987; Wickham & Oxburgh 1987). Ironically, many of these massifs have been shown to have followed P - T - t paths that involved decompression under prograde and retrograde conditions, that is to say they are clockwise, such as the Bosost and Lys-Caillaouas massifs of the central Pyrenees (Pouget 1991; Kriegsman et al. 1989), the Canigou massif in the eastern Pyrenees (Gibson 1991), and the Trois Seigneurs and Saint Barth616my massifs of the North Pyrenean Zone (Kriegsman 1989; de Saint Blanquat et al. 1990). Further all of these authors have shown that the Variscan hightemperature-low-pressure metamorphism of the Pyrenees occurred in an extensional tectonic environment that produced the subhorizontal regional foliation; this followed only moderate overthickening evidenced by earlier tectonic structures. Granulite m e t a m o r p h i s m

This topic of regional metamorphism has proven intellectually productive during the past few years, and much information can be found in two recent books (Vielzeuf & Vidal 1990; Ashworth & Brown 1990). Some granulite facies terranes clearly are the result of collisional metamorphism, such as the Grenville Province in North America (Anovitz & Chase 1990); these terranes exhibit little variation in metamorphic conditions over large areas, and apparent disequilibrium textures, such as symplectites of orthopyroxene and plagioclase after garnet and quartz, may be preserved in rocks of the same bulk composition over thousands of square kilometres. Such metamorphic terranes followed clockwise paths in P - T space but have remained incubated as the post-orogenic lower crust to generate a long, nearly isobaric cooling history from high temperatures. Other granulite facies terranes appear to be related to extensional tectonics, and such an example has been described by Armstrong et al. (1992) from central Massachusetts, where a component of advective heat transfer from the mantle also is thought to be important. Finally, granulite facies metamorphism may be driven substantially by advective heat transfer, exemplified by the Proterozoic low-pressure granulites of southwest Finland (e.g. Schreurs & Westra 1986). The extreme conditions characteristic of granulite-facies terranes are of two kinds: high-temperature, such as found in the Enderby Land granulite terrane (e.g. Ellis 1980; Sheraton et al. 1987); and high-pressure, such as found in the European Variscides (e.g. Carswell & O'Brien 1992). With respect to the high-pressure granulites, such rocks may well represent the exposed roots of collisional mountain belts, metamorphism being the result of burial during crustal

overthickening. However, high-temperature granulites require a gross perturbation of the normal continental geothermal gradient, and many of these terranes preserve evidence of prolonged residence in the middle-to-lower crust after deformation and metamorphism. In spite of the debate in the literature during the 1980s concerning the origin of granulite-facies terranes (e.g. Bohlen 1987; Ellis 1987), the general cause of granulite-facies metamorphism in many high-temperature terranes may be external to the rocks that we observe, and possibly external to the crust (Vernon et al. 1990). Much of the argument during the 1980s about the origin of granulite-facies terranes stemmed from two particular kinds of incomplete P - T - t paths, derived largely from evidence preserved from the retrograde rather than the prograde metamorphic history, which has allowed the division of many high-grade metamorphic terranes into two types (Harley 1989): those which show near-isobaric cooling and those which show near-isothermal decompression. Isobaric cooling paths have been identified from many granulite terranes (see reviews by Bohlen 1987 and Harley 1989) but the tectonic setting in which they are generated, either crustal thickening or magmatic accretion (Ellis 1987; Bohlen 1987; Bohlen & Mezger 1989; Harley 1989; Bohlen 1991) or, possibly, lithopheric extension (Sandiford & Powell 1986), remains a matter of debate. Furthermore, since evidence for the prograde path is generally lacking, rocks which exhibit isobaric cooling paths can have followed either a clockwise or a counterclockwise path in P - T space. As an example, the granulite facies rocks of the Grenville province are best modelled by early high-pressure conditions followed by exhumation to lower-middle crustal levels and then slow cooling (Anovitz & Chase 1990). Isothermal decompression paths occur as part of clockwise P - T evolution but have steeper d P / d T s l o p e s than those generated by the erosion-controlled exhumation of the kind discussed by England & Richardson (1977) and England & Thompson (1984). The model of Albar~de (1976) corresponds closely with many of the petrologically determined P - T - t paths, as by Hollister (1982) in the Coastal Range of British Columbia, Canada, Brown & Earle (1983) in Timor, Droop & Bucher-Nurminen (1984) in the Alps, and Harris & Holland (1984) in the Limpopo mobile belt of Southern Africa. Such 'fast exposure paths' (Harley 1989) can be generated by rapid exhumation and probably reflect tectonic thinning by extension of previously thickened crust (e.g. Sonder et al. 1987; Ruppel et al. 1988). Subduction zone metamorphism

Advances in our understanding of subduction zone metamorphism have been achieved through field and petrologic investigations of ancient and modern subduction zones and through numerical modelling of heat and mass transfer at convergent plate margins. Again, this type of metamorphism grades into collisional metamorphism because arcs and continents will be transported to subduction zones where they will slow subduction and generate intracontinental deformation. Recently, an increasing number of localities with evidence of very high pressure, reflected in coesite or coesite pseudomorphs and other exotic minerals, has been identified; this demonstrates that crustal material can be subducted to, and exhumed from, depths greater than 100 km in collision zones, for example in the Alps (Chopin 1984, 1987; Schreyer 1988), in the

P-T-t

E V O L U T I O N OF O R O G E N I C BELTS

Caledonides of Norway (Smith 1984), and in the Dabie Mountains in central China (Wang et al. 1989). Further evidence of such extreme pressures is the occurrence of diamond-bearing metamorphic rocks derived from crustal protoliths (Sobolev & Shatsky 1990; Xu et al. 1992). The structural and metamorphic features of such very highpressure metamorphic rocks require rapid exhumation that is likely tectonic, effectively involving transport as fault-bounded tectonic slices. Without rapid tectonic exhumation the very high-pressure mineral assemblages are thermally consumed, hence slow erosion-controlled exhumation allows this type of metamorphism to be overprinted under lower-pressure/higher-temperature conditions because it is 'time, that devours all things!'. In terms of P - T - t paths, Ernst (1988) has divided subduction zones into two different types according to the retrograde legs of the P - T - t paths. One type comprises the collisional blueschist belts, such as the Western Alps, which undergo widespread retrogression in the greenschist and/or epidote amphibolite facies during near isothermal decompression associated with the collision of a continental fragment, oceanic plateau, or arc with the subduction complex. Other high-P belts, such as the Franciscan Complex of California in the USA, in which continued subduction results in exhumation P - T - t paths that approximately retrace prograde paths, exhibit minimal retrogression and the preservation of metamorphic aragonite. A correlation evidently exists between the nature of the retrograde metamorphic trajectory and continued underflow that inhibits back-reaction, in comparison with collision and abrupt deceleration of convergence that allows pervasive back-reaction. Thus, retrograde blueschist parageneses can help to constrain the tectonic history of the subduction zone.

Quo vadimus? Further resolution of problems in metamorphism requires advances in a number of different areas. These include, but are not limited to, the following. (1) The continued development of an accurate thermodynamic data base for minerals that occur in metamorphic equilibria through an improvement of our thermodynamic knowledge of individual minerals, and in particular the activity-composition relations in the P - T range of interest. The data on thermodynamic properties of minerals are becoming more accurate and precise with time and the importance of using internally consistent data sets has been realized; however, thermodynamically calibrated thermobarometers must continue to be evaluated against experimentally based equilibria and natural occurrences. More accurate and precise determination of metamorphic P - T conditions will place tighter constraints on modelling studies and advance our ability both to understand metamorphic processes and to identify different tectonic settings through their characteristic metamorphism. (2) The development of a better understanding of the kinetic response of minerals to changes in P and T, and in particular improvement of knowledge of diffusion rates and closure temperatures in minerals that are useful geochronometers. The past ten years have seen significant improvements in our understanding of processes such as intracrystalline diffusion that can modify significantly mineral compositions from their peak metamorphic values and thus obscure the peak conditions (e.g. Spear & Florence

77

1992, and references therein). However, the recent discovery of oxygen isotope zoning within garnet (Chamberlain & Conrad 1991) reminds us how poorly we understand the kinetics of diffusion, although it should be noted that this provides an additional record, or 'tape recording' of events in the evolution of the rock that are likely to have been completely obliterated in the matrix. (3) The further development of radiogenic isotope dating methods for a wider range of minerals to improve our ability to measure time at different points on the prograde and retrograde segments of P - T - t paths, and the increasingly widespread use of the laser 4°mr/39Ar method and the ion microprobe as geochronological tools. Uranium-lead mineral dating is the best source of high-precision ages. The relatively high U/Pb ratios of widely occurring accessory minerals such as zircon, titanite, monazite, rutile and ilmenite, the relatively rapid change in the 2°7pb*/2°6pb* with time, and the relatively short half-lives of 235U and 238U all contribute to the potential of small age uncertainties; uncertainties of less than + 0.1% relative are possible for concordant U-Pb ages. Increased use of high-precision U-Pb ages will improve our understanding of the time involved in the high-temperature parts of P - T - t paths. The usefulness of garnet to the future quantification of rates of tectonometamorphic processes in metamorphism is exemplified in a number of recent papers. Christensen et al. (1989) have utilized the Rb-Sr method in single garnet crystals from schists in southeast Vermont to study the rates of petrological processes, such as growth rate, estimated at 1.4 +0.92/-0.45mm per million years, and average time interval of growth, c. 10.5 + 4.2 Ma. Garnet and its mineral inclusions provide a sequential record of P - T change, strain and chemical reactions during metamorphism; therefore, the technique offers the potential for determination of the rates of those processes as well. In the Vermont example, the growth interval and the observed amount of rotation recorded by inclusion trails, assuming that the porphyroblasts have rotated with respect to the matrix, indicate that the average shear strain rate during garnet growth was 2.4 + 1 . 6 - 0 . 7 x 10 -14 per second. By contrast, Burton & O'Nions (1990) have used a combination of the Sm-Nd, U-Pb and Rb-Sr systems to give whole rock isochron and mineral isochron ages that have revealed in detail the chronology of processes in small-scale granulite formation from Kurunegala in Sri Lanka. Finally, the Sm-Nd system offers the possibility of dating different zones within minerals such as metamorphic garnet (Burton & O'Nions 1991).

Epilogue The past 100 years have produced significant and dramatic progress in our understanding of metamorphic processes, even though the relative contribution of some processes remains unresolved. Recently, we have begun to recognize and appreciate the importance of extension in collisional mountain belts, interpreted previously largely in terms of contraction. Most of us now accept that clockwise and counterclockwise P - T - t paths occur in nature and reflect substantially different tectonic and magmatic processes. It is apparent that high-temperature-low-pressure metamorphism can be generated by different tectonic processes. Granulites not only represent extreme conditions of both pressure and temperature but also are produced by tectonic

78

M. B R O W N

processes that result in both clockwise and counterclockwise P - T - t paths. The causes of regional metamorphism are multiple and may occur individually or in unison, in addition to the 'normal' conductive heat flux from the earth's interior; they include internal radiogenic heat production in thickened continental crust and structurally inverted sedimentary basins, magmatic advective heat transfer both from the mantle into t h e crust and from the lower crust into the upper crust, and lithospheric extension resulting in an enhanced conductive heat flux from the underlying asthenosphere. It is often alleged that serendipity plays a large part in research. Perhaps George Barrow was lucky to be born in the UK during the period when the Geological Survey was at its peak mapping the Scottish Highlands, but without his ability to make careful observations and his insight in their interpretation he could not have written the perspicacious paper that has provided much of the foundation for metamorphic geology today. It is an unfortunate measure of progress in science that some of what Barrow wrote about in detail in the Scottish Highlands has been reinterpreted. The pegmatites, critical to the model of thermal metamorphism preferred by Barrow, represents sub-solidus and anatectic migmatites and the tectonic setting of the metamorphism was one of plate collision. Nonetheless, his work will be remembered in perpetuity as the type example of medium-pressure or 'Barrovian' metamorphism.

States and thermal modelling. Geological Society of America, Bulletin, 101, 1051-1065. BELL, T. H., JOHNSON, S. E., DAVIS, B., FORDE, A., HAYWARD, N. and WILKINS, C. 1992. Porphyroblast inclusion-trail orientation data: eppure non son girate! Journal of Metamorphic Geology, 10, 295-307• BIRD, P., TOKSOZ, M. N. & SLEEP, N. H. 1975. Thermal and mechanical models of continent-continent convergence zones. Journal of Geophysical Research, 80, 4405-1406. BOHLEN, S. R. 1987. Pressure-temperature-time paths and a tectonic model for the evolution of granulites. Journal of Geology, 95, 617-632. --, 1991. On the formation of granulites. Journal of Metamorphic Geology, 9, 223-230. -& LINDSLEY, D. H. 1987. Thermometry and barometry of igneous and metamorphic rocks. Annual Reviews of Earth and Planetary Science, 15, 397-420. & MEZGER, K. 1989. Origin of granulite terranes and the formation of the lowermost continental crust. Science, 244, 326-329. BROWN, M. 1993. The generation, segregation, ascent and emplacement of granite magma: Insights from migmatites. Earth Science Reviews. & EARLE, M. M. 1983. Cordierite-bearing schists and gneisscs from Timor, eastern Indonesia: P - T conditions of metamorphism and tectonic implications. Journal of Metamorphic Geology, 1, 183-203. BURCHFIEL, B. C. & ROYDEN, L. H. 1985. North-south extension within the convergent Himalayan region. Geology, 13, 679-682. , CHEN, Z., HODGES, K. V., LIu, Y., ROYDEN, L. H., DENG, C. & Xu, J. 1992. The South Tibetan detachment system, Himalayan Orogen, -

-

-

extension contemporaneous with and parallel to shortening in a collisional mountain belt. Geological Society of America, Special Paper 269. BURG, J. P., BRUNEL, M., GAPAIS, D., CHEN, G. M. & LIU, G. H. 1984• Deformation of leucogranites of the crystalline Main Central Sheet in Southern Tibet (China). Journal of Structural Geology, 6, 535-542. BURTON, K. W. & O'NIONS, R. K. 1990. The time scale and mechanism of granulite formation at Kurunegala, Sri Lanka. Contributions to Mineralogy and Petrology, 106, 66-89. & --, 1991. High resolution garnet chronometry and the rates of metamorphic processes. Earth and Planetary Science Letters, 107, 649-671. CARSWELL, O. A. & O'BRIEN, P. J. 1992. Spatial and temporal relationships between HP and LP metamorphic assemblages in the Central European Variscides. 29th International Geological Congress, Kyoto, Japan, Abstracts, 2, 581. CHAMBERLAIN, C• P. & CONRAD, M. E. 1991. Oxygcn isotope zoning in garnet. Science, 254, 403-406. & SONDER, L. J. 1990. Heat-producing elements and the thermal and baric patterns of metamorphic belts. Science, 250, 763-769. CltEN, Z., LIU, Y., HODGES, K. V., BURCHFIEL, B. C., ROYDEN, L. H. & DENG, C. 1990. The Kangmar dome: A metamorphic core complex in southern Xizang (Tibet). Science, 250, 1552-1556. CHINNER, G. A. 1966. The distribution of pressure and temperature during Dalradian metamorphism. Quarterly Journal of the Geological Society of London, 122, 159-186. CtlOPIN, C. 1984. Coesite and pure pyrope in high-grade blueschists of the Western Alps: A first record and some consequences. Contributions to Mineralogy and Petrology, 86, 107-118. , 1987. Very high-pressure metamorphism in the Western Alps: Implications for subduction of continental crust. Philosophical Transactions of the Royal Society, London, A321, 183-197. CHRISTENSEN, J. N., ROSENEELD, J. L. & DE PAOLO, O. J. 1989. Rates of tectonometamorphic processes from rubidium and strontium isotopes in garnet. Science, 244, 1465-1469. COGN~, J. 1960. Schistes crystallins et granites en Bretagne mridionale. Le domaine de I'anticlinal de Cornouaille. M6moires pour servir l'explication de la Carte G6ologique d6taill6e de la France• COLLINS, W. J. & VERNON, R. H. 1991. Orogeny associated with anticlockwise P - T - t paths: Evidence from low-P, high-T metamorphic terranes in the Arunta inlier, central Australia• Geology, 19, 835-838. DALLMEYER, R. D. & BROWN, M. 1992. Rapid Variscan (c. 300 MR) exhumation of Eo-Variscan (c. 400 MR) metamorphic rocks from South 40 39 Brittany, France: New A r / Ar age data and tectonic implications•

I acknowledge the contribution made to my metamorphic education by all the participants in IGCP Project 235 (1984-1990). Rapid, critical and constructive reviews that substantially improved this article were provided by T.R. Armstrong, G.T.R. Droop, E.J. Krogstad, E.L. McLellan, K. Mezger, P.J. O'Brien, J.C. Schumacher, R.J. Tracy and two anonymous reviewers. I thank K. Mezger for the provision of Fig. 5 and R.J. Tracy for the provision of Fig. 2, and J. Martin for proficient word processing; however, I take responsibility for those misperceptions and infelicities that remain.

References ALBAREDE, F. 1976. Thermal models of post-tectonic decomprcssion as cxcmplified by the Haut-Allier granulites (Massif Ccntral, France)• Bulletin de la Soci~t( Gdologique de France, 18, 1023-1032. ALLEN, T. & CIIAMBERLAIN,C. P. 1989. Thermal consequences of mantled gneiss dome emplacement. Earth and Planetary Science Letters, 93, 392-404. ANDERSON, M. W., BARKER, A. J., BENNETT, D. G. & DALLMEYER, R. D. 1992. A tectonic model for Scandian terrane accretion in the northern Scandinavian Caledonides. Journal of the Geological Society, London, 149, 727-741. ANOWTZ, L. M. & CHASE, C. G. 1990. Implications of post-thrusting extension and underplating for P - T - t paths in granulite terrancs: A Grenvillc cxamplc. Geology, 18, 466-469. ARMSTRONG, T. R., TRACY, R. J. & HAMES, W. E• 1992. Contrasting styles of Taconian, Eastern Acadian and Western Acadian metamorphism, Central and Western New England. Journal of Metamorphic Geology, 10, 415-426. ASHWORTII, J. R. & BROWN, M. 1990. High-temperature metamorphism and crustal anatexis. Unwin Hyman, London, UK. ATHERTON, M. P. 1977. Carncgie Review Article: The metamorphism of the Dalradian rocks of Scotland. Scottish Journal of Geology, 13, 331-370. BARROW, G. 1893. On an intrusion of muscovite-biotite gneiss in the south-eastern Highlands of Scotland, and its accompanying metamorphism. Quarterly Journal of the Geological Society of London, 49, 330-358. --, 1912. On the geology of lower Dee-side and the Southern Highland Border. Proceedings of the Geologists' Association, 23, 274-290. BARTON, M. D. & HANSON, R. B. 1989. Magmatism and the development of low-pressure metamorphic belts: Implications from the western United

-

-

-

-

-

•

Geological Society of America Annual Meeting, Cincinnati, Ohio, Abstracts with Program, 24, A236. DALY, J. S., CLIFF, R. A. & YARDLEY, B• W. D. (eds) 1989. Evolution of Metamorphic Belts. Geological Society, London, Special Publication 43.

-

DEMPSTER, Z. J. 1985. Uplift patterns and orogenic evolution in the Scottish Dalradian. Journal of the Geological Society, London, 142, 111-128. & HARTE, B. 1986. Polymetamorphism in the Dalradian of the central Scottish Highlands. Geological Magazine, 123, 95-104. DEWEY, J. F. 1988. Extensional collapse of orogens. Tectonics, 7, 1123-1139. DE SAINT BLANQUAT, M., LARDEAUX, J. M. & BRUNEL, M. 1990. Petrological -

P-T-t

EVOLUTION

arguments for high-temperature extensional deformation in the Pyrenean Variscan crust (St. Barth616my massif, Arirge, France). Tectonophysics, 177, 245-262. DE YOREO, J. J. 1988. Thermal models for granite genesis following crustal thickening. Los, Transactions, American Geophysical Union, 69, 770. , Lux, D. R. & GolDOa'rl, C. V. 1989a. The role of crustal anatexis and magma migration in regions of thickened continental crust. In: DALY, J. G., CLIIW, R. A. & YARDLEY, B. W. D. (eds) Evolution of metamorphic belts. Geological Society, London, Special Publications, 43, 187-202. & , 1991. Thermal modelling in low-pressure/hightemperature metamorphic belts. Tectonophysics, 188, 209-238. --, ~ ~¢. , DECKER, E. R. & OSBERG, P. H. 1989b. The Acadian thermal history of western Maine. Journal of Metamorphic Geology, 7, 169-190. D'LEMOS, R. S., BROWN, M. d~ STRACHAN, R. A. 1992. Granite magma generation, ascent and emplacement within a transpressional orogen. Journal of the Geological Society, London, 149, 487-490. DROOP, G. Z. R. • BUCHER-NURMINEN, K. 1984. Reaction texturcs and metamorphic evolution of sapphirine-bearing granulites from the Gruf Complex, Italian central Alps. Journal of Petrology, 25, 766-803. ELLIS, D. J. 1980. Osumilite-sapphirine-quartz granulites from Enderby Land, Antarctica: P - T conditions of metamorphism, implications for garnet-cordierite equilibria and the evolution of the deep crust. Contributions to Mineralogy and Petrology, 74, 201-210. ,1987. Origin and evolution of granulites in normal and thickened crusts. Geology, 15, 167-170. ENGLAND, P. C. & RICHARDSON, S. W. 1977. The influence of erosion upon the mineral facies of rocks from different metamorphic environments. Journal of the Geological Society, London, 134, 201-213. & THOMPSON, A. B. 1984. Pressure-temperature-time paths of regional metamorphism, I. Heat transfer during the evolution of regions of thickened continental crust. Journal of Petrology, 25, 894-928. ERNST, W. G. 1988. Tectonic history of subduction zones inferred from retrograde bludeschist P - T paths. Geology, 16, 1081-1084. ESKOLA, P. 1920. The mineral facies of rocks. Norsk Geologisk Tidsskrift, 6, 143-194. FURLONG, K. P., HANSEN, R. B. & BOWERS, J. R. 1991. Modeling thermal regimes. In: KERRICK, D. M. (ed.) Contact metamorphism. Reviews in Mineralogy, 26, Mineralogical Society of America, 437-505. GANSSER, A. 1964. Geology of the Himalayas. Wiley-lnterscience, London, -

-

UK. GHENT, E. D., STOUT, M. Z. & PARRISH, R. R. 1988. Determination of metamorphic pressure-temperature-time ( P - T - t ) paths. In: NISBET, E. G. & FOWLER, C. M. R. (eds) Heat, metamorphism and tectonics. Mineralogical Association of Canada Short Course, 14, 155-188. GIBSON, R. L. 1991. Hercynian low-pressure-high-temperature regional metamorphism and subhorizontal foliation development in the Canigou Massif, Pyrenees, France-Evidence for crustal extension. Geology, 19, 380-383. GOSCOMBE, B. 1992. High-grade reworking of central Australian granulites: Metamorphic evolution of the Arunta Complex. Journal of Petrology, 33, 917-962. HAMES, W. E., TRACY, R. J. & BODNAR, R. J. 1990. Post-metamorphic unroofing history dcduced from petrology, fluid inclusions, thermochronometry, and thermal modeling: An example from southwestern New England. Geology, 17, 727-730. HARKER, A. 1932. Metamorphism. Methuen, London, UK. HARLEY, S. L. 1989. The origins of granulites: A metamorphic perspective. Geological Magazine, 126, 215-247. & Fn'ZSlMONS, I. C. W. 1991. Pressure-temperature evolution of metapelitic granulites in a polymetamorphic terrane: The Rauer Group, East Antarctica. Journal of Metamorphic Geology, 9, 231-243. , JENSEN, B. J. & SHERATON, J. W. 1990. Two-stage decompression in orthopyroxene-sillimanite granulites from Forefinger Point Enderby Land, Antarctica: Implications for the evolution of the Archean Napier Complex. Journal of Metamorphic Geology, 8, 591-613. HARRIS, N. B. W. & HOLLAND, T. J. B. 1984. The significance of cordierite-hypersthene assemblages from the Beitbridge region of the central Limpopo Belt: Evidence for rapid decompression in the Archean? American Mineralogist, 69, 1036-1049. HARTE, B. & DEMPSTER,T. J. 1987. Regional metamorphic zones: Tectonic controls. Philosophical Transactions of the Royal Society, London, A321, 105-127. & HUDSON, N. F. C. 1979. Pelite facies series and the temperatures and pressures of Dalradian metamorphism in E Scotland. In: HARRIS, A. L., HOLLAND, C. H. d(¢ LEAKE, B. E. (eds) The Caledonides of the British Isles--reviewed. Geological Society, London, Special Publications, 8, 323-344. HAUGERUD, R. A. 1989. On numerical modeling of one-dimensional -

-

-

-

OF

OROGENIC

BELTS

79

geothermal histories. Computers and Geosciences, 15, 825-836. & ZEN, E-AN 1991. An essay on metamorphic path studies or Cassandra in P - T - t space. In: PERCHUK, L. L. (ed.) Progress in metamorphic and magmatic petrology. Cambridge University Press, Cambridge, UK, 323-348. HIROl, Y. & KIsm, S. 1989. P - T evolution of Abukuma metamorphic rocks in north-east Japan: Metamorphic evidence for ocean crust obduction. In: DALY, J. S., CLIFF, R. A. and YARDLEY, B. W. D. (eds) Evolution of Metamorphic Belts. Geological Society, London, Special Publications, 43, 481-486. HODGES, K. V. 1991. Pressure-temperature-time paths. Annual Reviews of Earth and Planetary Science, 19, 207-236. --, PARRISH, R., HOUSH, T., Lux, D., BURCHFIEL, B. C., ROYDEN, L. & CHEN, Z., 1992. Simultaneous Miocene extension and shortening in the Himalayan orogen. Science, 2,58, 1466-1470. --, BURCHFIEL, B. C., ROYDEN, L., CHEN, Z. and LIu, Y., 1993. The metamorphic signature of contemporaneous extension and shortening in the Central Himalayan orogen: Data from the Nyalam transect, Southern Tibet. Journal of Metamorphic Geology, 11, In Press. , HUBBARD, M. S. & SILVERBERG,D. S. 1988. Metamorphic constraints on the thermal evolution of the central Himalayan orogen. Philosophical Transactions of the Royal Society, London, A326, 257-280. HOLLISTER,L. S. 1982. Metamorphic evidence for rapid (2 mm/yr) uplift of a portion of the Central Gneiss Complex, Coast Mountains, B.C. Canadian Mineralogist, 2,0, 319-332. , BURRUSS, R. C., HENRY, D. L. & HENDEL, E-M. 1979. Physical conditions during uplift of metamorphic terranes, as recorded by fluid inclusions.Bulletin Min~ralogique, 102, 555-561. HOUSEMAN, G. A., McKENZIE, D. P. & MOLNAR, P. 1981. Convective instability of a thickened boundary layer and its relevance for the thermal evolution of continental convergcnt belts. Journal of Geophysical Research, 86, 6115-6132. HUBBARD, M. S. & HARRISON, T. M. 1989. 4°Ar/39Ar age constraints on deformation and metamorphism in the Main Central Thrust Zone and Tibetan Slab, Eastern Nepal, Himalaya. Tectonics, 8, 865-880. INGER, S. & HARRIS, N. B. W. 1992. Tectonothermal evolution of the High Himalayan Crystalline Sequence, Langtang Valley, Northern Nepal. Journal of Metamorphic Geology, 10, 439-452. JAMIESON, R. A. 1991. P - T - t paths of collisional orogens. Geologische Rundschau, 80, 321-332. JAUPART, C. & PROVOST, A. 1985. Heat focussing, granite genesis and inverted metamorphic gradients in continental collision zones. Earth and Planetary Science Letters, 73, 385-397. JOHNSON, M. R. W. 1963. Some time relations of movement and metamorphism in the Scottish Highlands. Geologie en Mijnbouw, 41, 121-142. JONES, K. A. & BROWN, M. 1989. The metamorphic evolution of the Southern Brittany Migmatite Belt, France. In: DALY, J. S., CLIFF,R. A. d~. YARDLEY, B. W. O. (eds) Evolution of Metamorphic Belts. Geological Society, London, Special Publications, 43, 501-505. & --, 1990. High temperature 'clockwise' P - T paths and melting in the development of regional migmatites: An example from Southcrn Brittany, France. Journal of Metamorphic Geology, 8, 551-578. KANO, H. & KURODA, Y. 1968. On the occurrences of staurolitc and kyanite from the Abukuma Plateau, northeastern Japan. Proceedings of the Japanese Academy, 44, 77-82. KRIEGSMAN, L. M. 1989. Deformation and metamorphism in the Trois Seigneurs massif, Pyrenees--~vidence against a rift setting for its Variscan evolution. Geologie en Mijnbouw, 68, 335-344. --, AERDEN, D. J. A. N., BAKKER, R. J., DEN BROK, S. W. J. & SCHUTJENS, P. M. T. M. 1989. Variscan tectonometamorphic evolution of the eastern Lys-Caillaouas massif, central Pyrenees----evidencc for late orogenic extension prior to peak metamorphism. Geologie en Mijnbouw, 68, 323-333. LE BRETON, N. & THOMPSON,A. B. 1988. Fluid-absent (dehydration) melting of biotite in metapelites in the early stages of crustal anatexis. Contributions to Mineralogy and Petrology, 99, 226-237. LE FORT, P. 1975. Himalayas: The collided range. Present knowledge of the continental arc. American Journal of Science, 275-A, 1-44. LOOSVELD, R. J. H. & ETHERIDGE, M. A. 1990. A model for low-pressure facies metamorphism during crustal thickening. Journal of Metamorphic Geology, 8, 257-267. Lux, D. R., DE YOREO, J. J., GoiDo'rri, C. V. & DECKER, E. R. 1986. The role of plutonism in the formation of low pressure metamorphic belts. Nature, 32,3, 794-797. MASON, B. 1992. Victor Moritz Goldschmidt: Father of modern geochemistry. The Geochemical Society, Special Publication No. 4. Trinity University, San Antonio, Texas, USA, 184 pp. McDOUGALL, I. & HARRISON, T. M. 1988. Geochronology and thermo-

-

-

-

80

M.

BROWN

chronology by the 4°Ar/39Ar method. Oxford University Press, New York, USA, 212 pp. MCLELLAN, E. L. 1983. Contrasting textures in metamorphic and anatectic migmatites: An example from the Scottish Caledonides. Journal of Metamorphic Geology, 1, 241-262. ,1985. Metamorphic reactions in the kyanite and sillimanite zones of the Barrovian type area. Journal of Petrology, 26, 789-818. , 1989. Sequential formation of metamorphic and anatectic migmatites related to thermal evolution, eastern Scotland. Journal of Geology, 96, 165-182. MEZGER, K. 1990. Geochronology in granulites. In: V1ELZEUF, D. & VIDAL, P. H. (eds) Granulites and crustal evolution. Kluwer Academic Publishers, Dordrecht, The Netherlands, 451-470. --, HANSON, G. N. & BOHLEN, S. R. 1989. U-Pb systematics of garnet: dating the growth of garnet in the late Archean Pikwitonei granulite domain at Cauchon and Natawahunan Lakes, Manitoba, Canada. Contributions to Mineralogy and Petrology, 101, 136-148. --, RAWNSLEY,C. M., BOHLEN, S. R. & HANSON, G. N. 1991. U-Pb garnet, sphene, monazite and futile ages: Implications for the duration of high-grade metamorphism and cooling histories, Adirondack Mountains, New York. Journal of Geology, 99, 415-428. MIYASHIRO, A. 1961. Evolution of metamorphic belts. Journal of Petrology, 2, 277-311. OXBURGH, E. R. & TURCOa~rE, D. L. 1974. Thermal gradients and regional metamorphism in overthrust terranes with special reference to the eastern Alps. Schweizerische Mineralogische und Petrographische Mitteilungen, 54, 641-662. PASSCHIER, C. W., TROUW, R. A. J., ZWART, H. J. and VISSERS, R. L. M. 1992. Porphyroblast rotation: eppur si muove? Journal of Metamorphic Geology, 10, 283-294. PATIN6-DOUCE, A. E., HUMPHREVS,E. D. & JOHNSTON, A. D. 1990. Anatexis and metamorphism in tectonically thickened continental crust exemplified by the Sevier hinterland, western North America. Earth and Planetary Science Letters, 97, 290-315. PEACOCK, S. M. 1990. Numerical simulation of metamorphic pressuretemperature-time paths and fluid production in subducting slabs. Tectonics, 9, 1197-1211. , 1991a. Numerical simulation of subduction zone pressure-temperaturetime paths: Constraints on fluid production and arc magmatism. Philosophical Transactions of the Royal Society, London, A335, 341-353. --, 1991b. Metamorphic geology. Reviews of Geophysics, Supplement, 486-499. PEUCAT, J. J. 1983. Gochronologie des roches m(tamorphiques (Rb-Sr et

U-Pb). Exemples choisis au Groenland, en Laponie, dans le Mct~sif Armoricain et an Grande Kabylie. Memoire de la Soci6t6 G6ologique ct Min6ralogiquc de Bretagne, 2 8 . POUGET, P. 1991. Hercynian tectonometamorphic evolution of the Bosost dome (French-Spanish central Pyrenees). Journal of the Geological Society, London, 148, 299-313. RICHARDSON, S. W. 1970. The relation between a petrogenetic grid, facies series and the geothermal gradient in metamorphism. Fortschrifte Mineralogische, 47, 65-76. , & ENGLAND, P. C. 1979. Metamorphic consequences of crustal eclogite production in overthrust orogenic zones. Earth and Planetary Science Letters, 43, 183-190. RUPPEL, C., ROYDEN, L. & HODGES, K. V. 1988. Thermal modeling of extensional tectonics: Application to pressure-temperature-time histories of metamorphic rocks. Tectonics, 7, 947-957. RUSHMER, T. 1991. Partial melting of two amphibolites: Contrasting experimental results under fluid-absent conditions. Contributions to Mineralogy and Petrology, 107, 41-59. SANDIFORD, M. & POWELL, R. 1986. Deep crustal metamorphism during continental extension: Modern and ancient examples. Earth and Planetary Science Letters, 79, 151-158. & --, 1991. Some remarks on high-temperature-low-pressure metamorphism in convergent orogens. Journal of Metamorphic Geology, 9, 333-340. SCHREYER, W. 1988. Subduction of continental crust to mantle depths: Petrological evidence. Episodes, 11, 97-104. SCHUMACHER, J. C., SCHUMACHER, R. & ROBINSOn, P. 1989. Acadian metamorphism in central Massachusetts and south-western New Hampshire: Evidence for contrasting P - T trajectories. In: DALY, J. S., CLIFF, R. A. & YARDLEY, I . W. D. (eds) Evolution of Metamorphic Belts. Geological Society, London, Special Publications, 43, 453-460. SEARLE, M. P., WATERS, D. J., REX, D. C. & WILSON, R. N. 1992. Pressure-temperature and time constraints on Himalayan metamorphism from eastern Kashmir and western Zanskar. Journal of the Geological Society, London, 149, 753-773.

SCHREURS, J. & WESTRA, L. 1986. The thermotectonic evolution of a Proterozoic low-pressure granulite dome, west Uusimaa, SW Finland. Contributions to Mineralogy and Petrology, 93, 236-250. SELVERSTONE, J. & CHAMBERLAIN, C. P. 1990. Apparent isobaric cooling paths from granulites: Two counter examples from British Columbia and New Hampshire. Geology, 18, 307-310. SHERATON,J. W., TINGLY, R. J., BLACK, L. P., OFFE, L. A. & ELLIS, D. J. 1987. Geology of Enderby Land and Western Kemp Land, Antarctica. Australian Bureau of Mineral Resources Bulletin 223. SHI, Y. & WANG, C. 1987. Two-dimensional modeling of the P--T-t paths of regional metamorphism in simple overthrust terranes. Geology, 15, 1048-1051. SMITH, D. C. 1984. Coesite in clinopyroxene in the Caledonides and its implications for geodynamics. Nature, 310, 641-644. SOBOLEV, N. V. & SHATSKY,V. S. 1990. Diamond inclusions in garnets from metamorphic rocks: A new environment for diamond formation. Nature, 343, 742-746. SONOER, L. J. & CHAMBERLAIN,C. P. 1992. Tectonic controls of metamorphic field gradients. Earth and Planetary Science Letters, 111, 517-535. , ENGLAND, P. C., WERNICKE, B. P. and CHRlST1ANSEN, R. L. 1987. A physical model for Cenozoic extension of western North America. In: COWARD, M. P., DEWEY, J. F. & HANCOCK, P. L. (eds) Continental Extensional Tectonics. Geological Society, London, Special Publications, 28, 187-201. SPEAR, F. S. 1988. The Gibbs method and Duhem's theorem. The quantitative relationships among P, T, chemical potential, phase composition and reaction progress in igneous and metamorphic systems. Contributions to Mineralogy and Petrology, 99, 249-256. , 1989. Relative thermobarometry and metamorphic P - T paths. In: DALY, J. S., CLIFF, R. A. & YARDLEY, B. W. D. (eds) Evolution of Metamorphic Belts. Geological Society, London, Special Publications, 43, 63-81. - & FLORENCE, F. P. 1992. Thermobarometry in granulites: Pitfalls and new approaches. Precambrian Research, 55, 209-241. -& PEACOCK, S. M. 1989. Metamorphic pressure-temperature-time paths. Short Course in Geology, 7, American Geophysical Union, Washington, D.C. - & SELVERSTONE, J. 1983. Quantitative P - T paths from zoned minerals: Theory and tectonic applications. Contributions to Mineralogy and Petrology, 83, 348-357. --,--, HICKMOIT, O., CROWLEY, P. & HODGES, K. V. 1984. P - T paths from garnet zoning. A new technique for deciphering tectonic processes in crystalline terranes. Geology, 12, 87-90. Sr-ONGE, M. R. 1987. Zoned poikiioblastic garnets: P - T paths and syn-metamorphic uplift through 30 km of structural depth, Wopmay orogen, Canada. Journal of Petrology, 28, 1-21. SWAPP, S. M. & HOLL1STER, L. S. 1991. Inverted metamorphism within the Tibetan slab of Bhutan: Evidence for a tectonically transported heat-source. Canadian Mineralogist, 29, 1019-1041. THOMPSON, A. B. 1981. The pressure-temperature (P, T) plane viewed by geophysicists and petrologists. TERRA Cognita, 1, 11-20. --, 1982. Dehydration melting of pelitic rocks and the generation of H20-undersaturated granitic liquids. American Journal of Science, 282, 1567-1595. --, 1990. Heat, fluids, and melting in the granulite facies. In: VIELZEUF, D. & VIDAL, P. H. (eds) Granulites and crustal evolution. Kluwcr Academic Publishers, Dordrecht, The Netherlands, 35-57. --, & ENGLAND, P. C. 1984. Pressure-temperature-time paths of regional metamorphism II. Their inference and interpretation using mineral assemblages in metamorphic rocks. Journal of Petrology, 25, 929-955. -& RIDLEY, J. R. 1987. Pressure-temperature-time ( P - T - t ) histories of orogenic belts. Philosophical Transactions of the Royal Society, London, A321, 27'45. THOMPSON, J. B., JR. 1955. The thermodynamic basis for the mineral facies concept. American Journal of Science, 2,53, 65-103. THOMPSON, P. 1989. Moderate over-thickening of thinned sialic crust and the origin of granitic magmatism and regional metamorphism in low-P-highT terranes. Geology, 17, 520-523. TILLEY, C. E. 1924. A preliminary survey of metamorphic zones in the Southern Highlands of Scotland. Quarterly Journal of the Geological Society of London, 81, 100-112. TRACY, R. J. 1987. Metamorphic geology. Reviews of Geophysics, 25, 1115-1122. - & ROBINSON, P. 1980. Evolution of metamorphic belts: Information from detailed petrologic studies. In: WONES, D. R. (ed) The Caledonides in the USA. VPI and SU Memoir, 2, 189-196. TRELOAR, P. J. & BROWN, M. 1990. Numerical modelling of high-temperature 'clockwise' P - T paths and origin of low-P metamorphic belts. Los, Transactions, American Geophysical Union, 71, 1662.

P-T-t

-

EVOLUTION

TURNER, F. J. 1981. Metamorphic petrology: Mineralogical, field and tectonic aspects. Second edition, McGraw-Hill, New York. VERNON,R. H., CLARKE,G. L. & COLLINS,W. J. 1990. Local, mid-crustal granulite facies metamorphism and melting: An example in the Mt. Stafford area, central Australia. In: ASHWORTH, J. R. & BROWN, M. (eds) High-temperature metamorphism and crustal anatexis. Unwin Hyman, London, UK, 272-319. VIELZEUF, D. & VIDAL, P. H. 1990.Granulites and crustal evolution. Kluwer Academic Publishers, Dordrecht, The Netherlands, 585 pp. WANG, X., LIOU, J. G. & MAO, H. K. 1989. Coesite-bearing eclogite from the Dabie Mountains in central China. Geology, 17, 1085-1088. WELLS, P. R. A. 1980. Thermal models for the magmatic accretion and subsequent metamorphism of continental crust. Earth and Planetary Science Letters, 46, 253-265. WICKHAM, S. M. 1987. Crustal anatexis and granite petrogenesis during low-pressure regional metamorphism: The Trois Seigfieurs Massif, Pyrenees, France. Journal of Petrology, 28, 127-169. , & OXBURGH, E. R. 1985. Continental rifts as a setting for regional metamorphism. Nature, 318, 330-333. & , 1987. Low-pressure regional metamorphism in the Pyrenees and -

OF OROGENIC

BELTS

81

its implications for the thermal evolution of rifted continental crust. Philosophical Transactions of the Royal Society, London, A321, 219-242. WINTSCH,R. P., SUTFER,J. F., KUNK,M. J., ALEINIKOFF,J. N. & DORAIS, M. J. 1992. Contrasting P - T - t paths: Thermochronologic evidence for a late Paleozoic final assembly of the Avalon Composite Terrane in the New England Appalachians. Tectonics, 11, 672-689. Xu, S., OKAY,A. I., JI, S., SENG()R, A. M. C., Su, W., LIu, Y., & JIANG, L. 1992. Diamond from the Dabie Shan metamorphic rocks and its implication for tectonic setting. Science, 2,56, 80-82. ZEITLER, P. K. 1989. The geochronology of metamorphic processes. In: DALV J. S., CLIFF, R. A. & YARDLEY, B. W. D. (eds) Evolution of Metamorphic Belts. Geological Society, London, Special Publications, 43, 131-147. ZEN, E-AN 1991. Phanerozoic denudation history of southern New England Appalachians deduced from pressure data. American Journal of Science, 291, 401-424. ZWART, H. J. 1962. On the determination of polymetamorphic mineral associations and its application to the Bosost area (central Pyrenees). Geologische Rundschau, 52, 38-65.

Received 22 October 1992; revised typescript accepted 11 November 1992

Addendum Our understanding of high-temperature-low-pressure metamorphism has been advanced recently by new work from the Ryoke Belt in Japan and the Chugach Metamorphic Complex in Alaska, USA. In the case of the Chugach Metamorphic Complex, observation of the relative timing of deformation, metamorphism and plutonism leads to a model of ridge subduction followed by plate reorganization to account for the abnormally high geothermal gradients in the forearc at the subduction zone separating North America from the Pacific Ocean Basin during the Eocene (Sisson & Pavlis 1993). Perhaps more significant, is the recognition that the Ryoke Belt in Japan also might be a consequence of ridge subduction (Nakajima 1994). The metamorphism of ~he Ryoke Belt has been considered typical of Miyashiro's (1961) low-pressure facies series (andalusitesillimanite type). It is the type example of the high-temperature component of a paired metamorphic belt, with the Sanbagawa Belt to the oceanward side being the high-pressure m e m b e r of the pair. The main part of the Ryoke Belt extends for a length of c. 1000 km, but has a width of only 30-50 km. Metamorphic rocks occupy about one-third of the total area of the Ryoke Belt, because of the large amount of granitic rocks that also occur and which are characteristic of this belt. Higher grade metamorphic zones within the Ryoke Belt exhibit evidence for both fluid-conserving melt-producing reactions and fluid-absent-melting reactions, in particular reactions that involve biotite with aluminosilicate±quartz. A t the highest metamorphic grade exposed, biotite-K-feldspar-cordierite-garnetbearing assemblages are characteristic in rocks with a migmatitic layering. A n upper limit on temperature is provided by the absence of hypersthene-bearing assemblages, which indicates that the stability of biotite + quartz was not exceeded at the crustal level now exposed. Peak metamorphic conditions in the highest grade zones of the Ryoke Belt metamorphism likely correspond to c. 4 kbar and c. 750 °C (Brown & Nakajima 1994). The sequence of mineral assemblages developed in pelites that cover a range in Mg/(Mg + Fe) suggests that the prograde P - T path may be close to isobaric, at least in the higher grade zones, and that P may not vary significantly along the belt (Brown & Nakajima 1994). A t present, there are insufficient data to assess the exhumation P - T path, but the fine-grained nature of the rocks suggests rapid cooling, and

some replacement of garnet by biotite may indicate that the retrograde P - T path also may be close to isobaric. K-Ar and Rb-Sr ages of Ryoke granitoids and metamorphic rocks indicate that the metamorphism is older in the west and younger in the east, following the same systematic eastward younging identified in the San-yo granitoids immediately to the north of the Ryoke Belt (Nakajima et al. 1990). The San-yo and Ryoke granitoids and the Ryoke metamorphic rocks were formed during approximately the same interval of time, and the magmatism and metamorphism shifted eastward between 105 Ma to 65 Ma from Southwestern to Central Japan. Cooling rates are high at 40-80 °C Ma -1. The along-arc age variation is incompatible with a tectonic model based on steady-state subduction, and the metamorphic and granitic rocks are interpreted to have formed by subduction of a single ridge segment that migrated along the Eurasian trench margin with time. Observations consistent with this interpretation include: the narrowness of the belt and the diachronous nature of the metamorphism; the approximately isobaric prograde P - T path; the fine-grained nature of even the highest grade metamorphic rocks; high temperatures at middle crustal depths; and the rapid cooling rates. This model requires the juxtaposition of the Sanbagawa Metamorphic Belt against the Ryoke Metamorphic Belt to be a younger event as a consequence of sinistral strike-slip displacement on the Median Tectonic Line, which raises questions about the usefulness of the concept of 'paired' metamorphic belts (Brown & Nakajima 1994).

Additional references BROWN, M. & NAKAJIMA, T. 1994. High-T-low-P metamorphism in the Ryoke Belt of Japan: consequences of ridge subduction. Geological Society of America, 1994 Annual Meeting, Abstracts with Programs, 26, 7, A-214. NAKAJIMA,T. 1994. The Ryoke plutono-metamorphic belt: Crustal section of the Cretaceous Eurasian continental margin. Lithos, in press. - - , SHIRAHASE,T. • SHIBATA,K. 1990. Along-arc variation of Rb-Sr ages of Cretaceous granitic rocks in southwest Japan. Contributions to Mineralogy and Petrology, 104, 381-389. SISSON, V.B. & PAVL1S, T.U 1993. Geologic consequences of plate reorganization: An example from the Eocene southern Alaska fore arc. Geology, 21, 913-916.

Added November 1994.

From QJGS,47, 330, 343. 29. On an IN~RUSI01~ of MUSC0VITE-BIOTITE GlWglSS ~n the SOUTHEA.STERlffHIOHLA'NDSof SCOTLAND,and it$ ACCO~tPAI~IINO METAmORPHISm. By G~OSa~ BARROW,Esq., :F.G.S. (Communicated by permission of the Director-General of the Geological Survey. Read March 22nd, 1893.) [PLATES X~r. & XVI.] CONTENTS.

I. II. III. IV. V.

VI. VII. ¥III. IX.

Introduction ..................................................................... Distribution and Mode of Occurrence of the Igneous Rocks ......... Petrological Characters oF the Igneous Rocks ........................... Minerals of the Metamorphic Rocks ....................................... Rocks of the Metamorphic Area ............................................. (a) The Sillimanite-zone. (b) The Cyanite-zone. (c) The Staurolite-zonc. Sedimentary Origin of the Metamorphic Rocks ........................ Evidence of Progressive Metamorphism .................................... General Conclusions, and Summary of Results ........................... Analyses of the Rocks .. .......................................................

Page 330 330 339. 337 343

351 352 352 354

L INTRODUCTION.

Ta~ area to which attention is directed in the following pages lies in the north-eastern corner of Forfarshire, and forms part of the singularly flat table-land of the South-eastern Highlands. I t is essentially a moorland district, much covered with peat and heather, and is drained by two rivers, the North Esk and the South Esk. The rocks of which the area is composed consist principally of gneisses and schists ; these are clearly seen in the craggy sides of the valleys through which the two Esks and their tributaries flow. Boulders of these rocks may be noticed in the rough walls by the roadside as one drives up the gleus, and their intensely crystalline aspect is a most striking feature. A brief visit to the crags and the flat-topped moorland speedily convinces the observer t h a t this crystalline aspect is one of the chief characteristics of the district. I t is proposed to show in the present communication that this area contains several masses of intrusive rock which are probably conneeted underground, and that the highly crystalline character of the surrounding schists is mainly the result of thermometamorphism.

V. I~0CKS OF TIIE ~ETAMORPHIC 2~R:EAo

A brief description may ~now be given of the principal types of rock in which the minerals above described occur. They may be divided into four groups: firstly,, those of the silhmanite-zone ; secondly, those of the cyanite-zone ; thirdly, those of the staurolitezone : and lastly, those lying between the third zone and the Great Highland Fault, as seen on the banks of the North :Esk.

From Le Bas, M. J. (ed.), 1995, Milestones in Geology, Geological Society, London, Memoir No. 16, 83-90 First published in Journal of the Geological Society, London, Vol. 150, 1993, pp. 21-28

The development of Early Palaeozoic global stratigraphy W. S. M c K E R R O W Department o f Earth Sciences, Parks Road, Oxford OX1 3PR, UK Abstract: The major steps in the development of Early Palaeozoic stratigraphy are examined, with special emphasis on early Journal papers by Murchison and Sedgwick, and on their conception of systems and series, which permitted long-distance correlation. Unlike other periods, the Ordovician and Silurian were originally split into series; most stages have only been defined in the past 60 years. From 1880, Lapworth's graptolite zones have allowed much greater chronological precision. More recently, other methods have been developed for recognizing small time divisions, including studies in gradational evolution. A significant new advance is the correlation (by biostratigraphy) of short-lived physical events such as magnetic reversals and sea-level and climatic changes.

Sedgwick (1845), in the first paper in Volume 1 of the Journal, presented transverse sections showing the stratigraphic and structural relations of Lower Palaeozoic rocks in North Wales. He acknowledged, on page 6, the palaeontological help of J. C. Sowerby and his assistant, J. W. Salter, but, as we shall see, fossils were scarce in many of these formations, and Sedgwick was clearly more interested in unravelling the structure. In contrast, Murchison applied the methods of William Smith by describing (also with the expert help of Sowerby and Salter) the fossils present in each formation of his Silurian System. By 1845, Murchison had traced the Silurian across much of northern Europe, and was collaborating with several European geologists (e.g. Murchison & Verneuil 1845) to define younger parts of the Palaeozoic. While many of the objectives in modern stratigraphy are similar to those of Murchison and Sedgwick, we have now more techniques available. Just as important, we also have a different approach to biostratigraphy, mainly because we know more about how the Earth works and how animals live than did the geologists of 150 years ago. After a brief discussion of developments prior to 1845, this review gives an account of the evolution of Early Palaeozoic stratigraphy, followed by a more personal view of where biostratigraphy is heading. The principal steps in the development of stratigraphy are set out in a logical sequence which means that they are not all presented in chronological order.

and sections showing beds from the Coal Measures up to the Muschelkalk of Thuringia (Dunbar & Rodgers 1958, p. 290; Geikie 1897, p. 100). The much more celebrated A. G. Werner, working in the same region of Germany as Lehmann and Ffichsel, is famous for his all-embracing doctrine on the origin of rocks; he thought that they were mostly precipitates from sea water: the Neptunist theory. But Werner also proposed, in 1787, five broad stratigraphical units (based, in part, on the work of his associates, in the regions of Erzgebirge, Saxony and Bohemia) which he claimed were world-wide in scope (Geikie 1897, pp. 102-115; Rupke 1983, pp. 112-14). Werner's stratigraphical units were: 5. 4. 3. 2. 1.

Volcanics Unconsolidated deposits Floetz [now classed as Permian to Tertiary] Transitional strata, including greywacke Primitive strata, including granite, gneiss, etc.

Werner is more widely known than Lehmann or Ffichsel. The new theory (together with the ability to give a good lecture and have notable students) provided Werner with more publicity, even although he did not publish very much. Thanks to Hutton and others, Werner's ideas on Neptunism were later discounted, but he retains a position in the development of stratigraphy. In 1831, when Murchison and Sedgwick started field work in Wales and Shropshire, their aim was to examine the beds below the Old Red Sandstone, which were still known by Werner's term: 'Transition rocks or greywacke' (Fig. 1).

Step 1: defining a local succession of formations and mapping them in the field Stratigraphy did not begin with William Smith. In 1751, the French geologist, J. E. Guettard, appears to have created the first geological map (Geikie 1897, pp. 21-2), when he portrayed the geographical distribution of rocks (with fossil localities) in France and western Europe. But Guettard gave no indication that he had any ideas on the chronological or structural relations of the strata he mapped (Geikie 1897, p. 96). A few years later (in 1756), the German geologist J. G. Lehmann, working in the Harz and the Erzgebirge mountains, was the first to publish sections showing the sequence and structure of rocks (Geikie 1897, p. 96-7). Then, in 1762, G. C. Ffichsel, a contemporary of Lehmann, published both maps

Step 2: recognizing characteristic fossils in each formation The British must share the credit with other Europeans for recognizing the importance of fossils to stratigraphy. In the seventeenth century, the Danish physician and mineralogist, Nicholaus Steno recognized that different strata had different fossils. Subsequently, in 1703, the British scientist Robert Hooke clearly envisaged the potential of fossil shells to provide the 'criteria of chronology' in sedimentary rocks, but he does not appear to have put his suggestion into practice (Dott & Batten 1988, pp. 19-24). By 1799, William Smith had drawn up a table of strata in 83

84

W . S . MCKERROW

WERNER 1787

SEDGWICK 1855

MURCHISON 1859

LYELL 1865

LAPWORTH 1879

1993 Pridoli

Upper Upper Silurian

Ludlow

Silurian

Silurian

Wenlock Upper Llandovery

Middle

Transition

t"

Silurian

~

.....

Silurian

Lower Llandovery

Upper strata

Ashgill Cambrian Lower

including

Caradoc 0

Lower Silurian

greywacke

Silurian

Llandeilo Ordovician Llanvim

Middle Arenig Cambrian

O < 0" > Z

Tremadoc Upper Cambrian Cambrian

Cambrian

0

Middle Cambrian

~

Lower Cambrian

~

Lower Cambrian Cambrian Longmynd

Fig. I. Some classifications of the Lower Palaeozoic (modified from Secord 1986, fig. 9.4). England (from the Coal Measures to the Chalk) in which he listed some characteristic fossils in each formation, but this was not published until 1813 (Townsend 1813; Geikie 1897, pp. 230-1). At the same time, following the lead of GiraudSoulavie (Geikie 1897, pp. 204-8), Cuvier & Brongniart (1808) were compiling faunal successions in Tertiary beds in the Paris Basin, although they were at first more interested in the history of life than in the correlation of strata (Hancock 1977, p. 6). Subsequently, Smith (1816-1819) showed how the use of fossils permitted formations to be mapped across much of England. Smith may never have heard of Guettard, Lehmann or F(ichsel, and he had little comment to make on the theories of Werner and Hutton. In 1817 Smith stated: 'My observations ... are entirely original, and unencumbered by theories, for I have none to support.' (Hancock 1977, p. 4). The only theory essential to construct a geological map is the belief that formations are entities. Smith certainly had most influence on subsequent developments in biostratigraphy. Even as late as 1822, Cuvier and Brongniart were still attempting to fit the French sequences into Werner's scheme, but when confronted by [Cretaceous] beds in Poland, with different lithologies but the same fossils

as in France, Brongniart started to argue in favour of the prime position of fossils as a means of correlation (Hancock 1977, p. 7). These new observations also brought new theories. While Lyell preached uniformitarianism, Cuvier thought that the faunal changes which occurred in his successions were the results of a series of extinction events (Hallam 1989, pp. 3740). The catastrophic theories of Cuvier show some signs of revival in recent years but, as Smith observed, theories are not essential for the application of fossils to correlation.

Step 3: facies and type localities One potentially confusing issue in biostratigraphy is the presence of different facies (with different fossils) occurring in rocks of the same age. The problem of facies was recognized early on by Brongniart, Fitton and Phillips and discussed in detail by Gressly (1838), who coined the term. A year after Gressly's paper was published, De la Beche had no great difficulty in regarding the Old Red Sandstone as a local facies of the marine Devonian (Rudwick 1985, pp. 267-8). In 1842, after his visit to New York, Lyell recorded Old Red Sandstone fish and marine

GLOBAL S T R A T I G R A P H Y Devonian rocks sandwiched between the Silurian and Carboniferous (Rudwick 1985, p. 381). Because of facies changes, the beds represented in different regions had different aspects, although the stratigraphical sequences in which they occurred often contained enough fossils for their approximate age to be determined. This was usually done by reference to previously described sequences. The concept of type localities (invented by d'Orbigny) has thus proved a useful tool when correlating over large distances (Hancock 1977, p. 11). Step 4: systems, periods and eras In the 50 years after the Geological Society was founded (in 1807), several of its fellows played a crucial role in the development of global stratigraphy. The long-distance correlation of formations was closely linked with their reclassification into larger groups. At first, the word 'system' had a variety of meanings, but when Murchison (1835, 1839) defined the Silurian System both by its rock sequences and its fossils, it soon became recognizable across Europe and in America, and the term 'system' gradually assumed its modern meaning (Bassett 1991, pp. 16-20; Rudwick 1985, p. 446). The use of systems (with the present definition) was pioneered by the British, who based most of them on British strata. The Carboniferous System was established by Conybeare & Phillips (1822) to include the Old Red Sandstone; its limits only developed later, when the Devonian and Permian systems had been defined. Most systems (though often with rather imprecise boundaries) were established within the following 20 years (Rupke 1983, pp. 128-9); these included Lyell's (1833) subdivision of the Tertiary, and the establishment of the Cambrian by Sedgwick in 1835 (published a year later in: Sedgwick & Murchison 1836), of the Silurian and the Permian by Murchison (1835, 1841b), and of the Devonian by Sedgwick & Murchison (1839). When established, most of these systems were known to have characteristic fossil assemblages, but at first the Cambrian was hard to recognize outside Wales because it lacked any well-documented diagnostic fossils. In the first volume of the Journal, Murchison & Verneuil (1845) emphasized how the Silurian, Devonian and Carboniferous each had distinct organic remains in the same superposition across much of northern Europe. They also showed how the Permian System could be defined with reference to known sequences in Germany and Russia, a stratigraphical method which most of us would applaud. Murchison & Verneuil commented on the similarities between the fauna and flora of the Permian and the Carboniferous, even though a marked unconformity was present at this level in many areas of Europe (now termed the Hercynian Orogeny). They (Murchison & Verneuil 1845, p. 82) also noted that 'The Triassic system does not contain a single Palaeozoic form, whether animal or vegetable'; there was a very marked change in the fossils above the Permian, even though strong stratigraphic breaks at this level were not common. By 1845, Murchison & Verneuil had recognized that unconformities are much less useful than faunal changes for international correlation; this elementary principle of global stratigraphy has taken over a century to be widely applied. Much earlier, in 1838, Sedgwick chose the term 'Palaeozoic' to denote the Cambrian and Silurian jointly (Rudwick 1985, p. 242). Later, Phillips (1840, 1841) redefined the term Palaeozoic and suggested the corresponding terms: 'Mesozoic'

85

and 'Kainozoic' [subsequently 'Cainozoic' and now 'Cenozoic'] (Harland et al. 1989, pp. 30-1; Rudwick 1985, p. 363). Phillips' new eras had apparently made Murchison and Verneuil look more closely at the faunal changes across the Permian/Triassic boundary. In the first paper published in the Journal, Sedgwick (1845) gave an account of the rocks of North Wales; this was a continuation of an earlier paper published in the Proceedings of the Geological Society. At this time, Sedgwick had temporarily changed his mind about the use of the term 'Cambrian' and used 'Protozoic' instead; this may have been related to the discovery, by the Geological Survey, of Llandeilo and Caradoc fossils within the supposed 'Cambrian' of Wales (Hallam 1989, pp. 78-9). The boundary between the Cambrian and the Silurian was first agreed (in 1834) by Sedgwick and Murchison to run through unmapped territory in Wales (Secord 1986, fig. 3.9; Bassett 1991, pp. 8, 15-16), but by 1845, Murchison, Sedgwick and the Geological Survey all agreed that similar fossils were present on both sides of this boundary. By the 1850s, Murchison's solution (Fig. 1) was to extend the Silurian downwards to include all the fossiliferous rocks from the Lingula Flags (now classed as Upper Cambrian) to the Ludlow, whereas Sedgwick claimed everything up to and including the May Hill Sandstone (recognized today as Upper Llandovery) as Cambrian (Hallam 1989, p. 82). These arguments on the Cambrian-Silurian boundary are well known, but there was a more significant difference between Sedgwick and Murchison: while Sedgwick continued to stress the importance of physical stratigraphy and structure, Murchison realized (and put into practice) his belief that fossils were the best method of long-distance correlation (Secord 1986; Hallam 1989, p. 83-4). Sedgwick (1852), writing on the rocks of southwest England, was still maintaining 'that no good classification either of subdivisions or systems, or of subordinate formation, can ever be attempted without a previous determination of the physical groups'. This paper was written after the 'Great Devonian Controversy', recounted by Rudwick (1985) was over, but at a time when much stratigraphical uncertainty still prevailed about the ages of many rocks in Devon and Cornwall. Nevertheless, Sedgwick eventually put some of Murchison's precepts into practice by recording every fossil available to him; in addition, Sedgwick recruited M'Coy as his palaeontological assistant. While both Sedgwick and Murchison were presumably moderately competent palaeontologists, they both used specialists whenever possible. Rudwick (1985, p. 444) points out that the 'clinching evidence' for the resolution of the Devonian controversy was 'not only the result of Murchison's competent but quite conventional field work in Devon, the Rhineland and Russia, but also the product of fossil specialists such as the impoverished Lonsdale and the less than gentlemanly Sowerby and Phillips ... more or less professional palaeontologists'. Murchison's 1845 geological map of eastern Europe and Russia was based on long-distance correlation by fossils (Johnson 1982). Palaeontology was also important in Murchison's subsequent travels, for example, his visits to Scandinavia and Russia (Murchison 1847), and to Germany and Bohemia in 1853 (Bassett 1991, p. 38) while he was compiling the first edition of Siluria (Murchison 1854). By the 1850s, the Silurian System was becoming global in extent. In the United States, James Hall had compiled a list of Silurian fossils from New York (Murchison 1841a), and D. D.

86

W . S . MCKERROW

Owen (1846) had mapped and described the Silurian System of the mid-west (Johnson 1977); and shortly after, Sharpe (1848) had listed shelly fossils from each formation in New York and compared them with equivalents in Britain. Charles Darwin (1846) reported on fossils from the Falkland Islands resembling Silurian forms (they are actually Devonian), and Strachey (1851) recognized Silurian fossils in the Himalayas. Murchison has a claim to be the first global geologist; his map (Murchison 1854, p. 475, reproduced by Bassett 1991, p. 41) of the geographical distribution of 'Palaeozoic formations' around the world was the first of its kind. Murchison's renown, though deserving of high recognition, was gained through numerous publications which did not always give due credit to others (Torrens 1990); he appears to have been following a contemporary custom (Flinn 1992). Murchison originally grouped together what we now call Caradoc and Llandovery beds. In 1852, M'Coy recognized that Murchison's 'Caradoc beds' contained two distinct faunas, so that what we now term 'Ordovician' and 'Silurian' could be distinguished palaeontologically (Hallam 1989, pp. 80-1). M'Coy and Salter also described fossils from the Cambrian (in its modern sense) of Britain. At about the same time, Barrande (1852-1911) distinguished three successive Early Palaeozoic faunas in Bohemia. Thus, by the mid-1850s, although not designated by any formal nomenclature, the importance of a tripartite division of the Early Palaeozoic was becoming recognized internationally (Hallam 1989, p. 83) and the boundaries (Fig. 1) were incorporated in the literature by such authorities as Lyell (1865). The Ordovician System was originally defined by Lapworth (1879) as extending from the base of the Arenig Series to the base of the Llandovery Series. Although these series were quite well defined by their faunas, Lapworth appears to have been influenced by the presence of two unconformities in Wales and Shropshire: one below the Arenig and another below the Llandovery, both reflecting geographically restricted tectonic events (Woodcock 1990; Toghill 1992). Subsequent international decisions have now classified the Tremadoc Series with the Ordovician on the basis of internationally recognizable faunal changes at its base (Norford 1991 and references therein). The very fact that system boundaries are the subject of so much discussion, illustrates their subjective nature. Modern work on the definitions of systems is now concerned with obtaining international agreement on precise definitions of the base of each system, using type sections where good zonal fossils are present. McLaren (1977) showed, for the Devonian, just how it should be done; since then, the base of the Silurian (Cocks 1988) and the Carboniferous (Paproth et al. 1991) have been agreed; the base of the Ordovician is close to settlement (Norford 1991), while work is still in progress on defining the base of the Cambrian (Cowie & Brasier 1989). The International Geological Congress at Bologna in 1880 recognized the distinction between stratigraphical and chronological divisions: the duration of a system was recognized as a period (Hancock 1977, p. 15). This reflected some new thinking about different types of stratigraphical units, which is still not universally agreed (see also Dunbar & Rodgers 1957, pp. 290-2).

Caradoc sandstone' (p. 13), but in the same volume Murchison & Verneuil (1845, p. 81) refer to 'the whole Palaeozoic series'. In these earlier papers, the term 'series' was used primarily in a lithological sense. Two years later, Murchison (1847, p. 2) compared some Swedish beds and their faunas with the 'limestones of Wenlock' and others with the 'Ludlow formation', but his only time-stratigraphical terminology comprised 'Lower Silurian', 'Upper Silurian' and 'the Old red (Devonian) system'. Eventually, the major divisions of the 'Cambrian' and 'Silurian' systems came to be termed 'series', while the major divisions of the Devonian and younger systems (or more accurately: periods) were called 'stages'. It was not until 1859, seven years after Sedgwick and M'Coy had distinguished the Caradoc and May Hill fossils, that Murchison (by then Director of the Geological Survey) proposed the term Llandovery as a series (Hallam 1989, p. 81; Bassett 1991, p. 38), as opposed to 'Llandovery rocks', the phrase he had earlier employed in The Silurian System (Murchison 1839). This proposal was also prompted by the recognition of a postCaradoc unconformity in Shropshire by Aveline and Salter (Bassett 1991, p. 31). Local unconformities like this Shelveian event (Toghill 1992), rather than faunal changes, continued to play intrusive roles in the definition of many series. Apart from in Norway (see below), smaller divisions than series were not, however, discerned during the first three-quarters of the nineteenth century, so, until Lapworth (1879-80) published a list of graptolite zones, the series (which were based primarily on shelly fossils) formed the main pillar of international correlation in the Early Palaeozoic. Because many benthic animals could not cross wide oceans, independent Ordovician series have been defined in North America and elsewhere using native shelf faunas, and it is only in recent years that these have been correlated, with moderate precision, with the British series (Ross et al. 1982; Fortey et al. 1991, fig. 8). The term 'stage' was first employed in stratigraphy for ammonite-rich sequences in the Mesozoic by d'Orbigny (184251) who defined it 'solely according to the identity in the composition of the faunas' (see Hancock 1977, p. 9). Stages were thus originally, as subsequently they always have remained, chronological terms, while the Early Palaeozoic 'series' started life as formations, and then with the growth ofbiostratigraphy, subsequently developed into time-rock units. Many think they still have a role as such, but this is not universally accepted (see Harland et al. 1989, p. 21). The work of Norwegian geologists allowed the immediate recognition of Murchison's Silurian series in the Oslo area (Johnson 1982); of these, Kjerulf (1857) is the most remarkable: he proposed a sequence of stages for the local Silurian, which has only been superseded in the past 15 years. During the last 60 years, following the lead of Bancroft (1933, 1945), the original series (some would now call them epochs) of Murchison and Sedgwick are today divisible into numerous stages, both in Britain and abroad. These are being redefined by relation to type localities, so that, even in the Ordovician, when many indigenous faunas prevailed, international correlations are becoming ever more precise (Williams 1969; Barnes & Williams 1991; Holland & Bassett 1989).

Step 5: series and stages

Step 6: graptolite zones

In Early Palaeozoic stratigraphy, the use of the term 'series' began haphazardly. Sedgwick (1845), in the first volume of the Journal, refers to the 'Bala series' (p. 11) and the 'series of

Oppel (1856-8) showed that the Jurassic Period could be divided into 33 zones based on pelagic ammonites. He pointed out that the recognition of zones 'involves exploring the verti-

GLOBAL S T R A T I G R A P H Y cal range of each separate species in the most diverse localities, while ignoring the lithological development of the beds' (Hancock 1977, p. 12). The application of Oppel's conception of zones to the Early Palaeozoic had to wait until Hall and Lapworth examined the pelagic graptolites. Geologists working on the Palaeozoic owe at least as great a debt to Lapworth as they do to Murchison and Sedgwick. Lapworth (1873) produced a series of papers on the palaeontology of graptolites. He then showed that many of his newly defined species had very short time ranges (Lapworth 1879-80). Subsequently, and most significantly, Lapworth (1878, 1889) applied his new graptolite zones to unravelling the stratigraphy and the structure of southern Scotland. The introduction of Lapworth's graptolite zones raised several problems in Early Palaeozoic stratigraphy. At first the zones were based almost entirely on sequences in the Southern Uplands, where brachiopods and trilobites are absent or rare; there was therefore great uncertainty in correlating the zones with the previously established series based on shelly fossils. Later, Lapworth (1889) described graptolites occurring with trilobites and brachiopods in southwest Scotland, where the pre-Ashgill benthic faunas are largely different (because of the wide Iapetus Ocean) from those in England and Wales where the series were defined. New correlations are still being proposed between the graptolite zones and the series and stages (e.g. Fortey et al. 1991; Holland & Bassett 1989), and studies in graptolite evolution are now allowing even finer chronological divisions to be discerned (Rickards 1989).

87

gradational changes occur only in a minority (?< 10%) of benthic lineages. A sudden appearance, followed by stasis, is the norm for many genera of trilobites and brachiopods, so most benthic taxa can never be expected to be used as fine time indicators. Numerous alternatives to the original graptolite zonal scheme are now available (see Cowie & Brasier 1989; Barnes & Williams 1991; Webby & Laurie 1991; Holland & Bassett 1989; Bassett et al. 1991). In North America and Australia some endemic taxa are employed as indices for graptolite zones. The development of micropalaeontology permits correlation of many sections previously considered to be barren. Conodonts are now almost as useful as graptolites (more so in some facies) and there is lively discussion (e.g. Barnes 1988; Norford 1991) on the relative merits of the two groups in biostratigraphy. Acritarchs (e.g. Fig. 2), chitinozoans and ostracodes also provide useful stratigraphical indices. In the Early Cambrian, various small shelly fossils are proving accurate guides to correlation, while in the Silurian vertebrates and plant spores can also be employed. These different schemes result in the duplication of zonal indices for the same time interval, but this should not lead to arguments about usage or priority. All correlation schemes must (eventually) be related to internationally agreed series and stages. Most of the finer subdivisions should, perhaps, remain informal, to be emended as our knowledge develops (e.g. Fig. 2).

Step 8: event stratigraphy Step 7: other palaeontological criteria for zonation Biostratigraphy, in the long run, must rely on evolutionary changes in animals and plants, but these do not always occur progressively. Moreover, many changes in fossil sequences are not related to evolution: they can be due to local or regional changes in facies or, in some cases, to migrations. Ziegler (1965) showed that similar Silurian benthic communities occur in similar environments at different times, and thus that benthic assemblages cannot be used p e r se to indicate fine time intervals. Subsequently, it has been shown (e.g. Ziegler et al. 1968; Cocks 1989) that it is only in lineages with gradational evolutionary changes, and which have been studied by quantitative methods, that benthos can be used as precise indicators of small time divisions (Fig. 2). The total range of a species can then be determined by recording the ancestors in the beds below and the descendants in the beds above. As early as 1945, Bancroft showed that enough changes occurred in some Late Ordovician brachiopods and trilobites for the subdivision of the series into stages (see appraisal in Williams 1969, p. 245). Williams (1948) and Sheldon (1987) have also described gradational evolution in a few families of Ordovician trilobites. In the Silurian, gradational evolution has been described in S t r i c k l a n d i a which enabled successive taxa of these brachiopods to be employed as time indicators within the Llandovery Series (Williams 1951; Baarli 1986); gradational evolution is also known in E o c o e l i a (Ziegler 1966). At present, using macrofossils, it is difficult to give precise ages to Llandovery shelly facies without reference to members of one of these two brachiopod lineages (Fig. 2). We now realise that, at least with shelf benthos, a few wellstudied lineages showing gradational evolution are the best basis on which to create fine time divisions. But appreciable

Some Early Palaeozoic eustatic changes in sea level are related to the widespread development of black shales (Leggett et al. 1981) and also to more local occurrences of red shales (Ziegler & McKerrow 1975). More significantly, lowering of sea-level can be correlated with some extinctions (e.g. Leggett et al. 1981; Fortey 1984; Johnson & McKerrow 1991; Nielsen 1992), especially in the pelagic realm. During the Early Palaeozoic, the only large sea-level change related to major extinctions in the shelly faunas was in the late Ashgill (Brenchley & Newell 1984; Brenchley 1984; Owen et al. 1991). Several sea-level changes can also be correlated with increased intercontinental migrations of benthic faunas (Scotese & McKerrow 1990; Sheehan & Coorough 1990). Some environmental changes are indicated by changing ratios in stable isotopes in beds relatively unaffected by diagenesis or metamorphism (e.g. Corfield et al. 1992; Kirschvink et al. 1991). We are still at the stage of trying to distinguish which sealevel changes are globally synchronous (e.g. Johnson et al. 1991). The methods suggested by Vail et al. (1977) and Haq et al. (1988) may eventually allow recognition of Early Palaeozoic stratigraphic sequences bounded by synchronous stratigraphical breaks on cratons, but in orogenic regions many unconformities are only regional in extent and some are diachronous (McKerrow et al. 1991; Toghill 1992). For example, Woodcock (1990) has shown clearly that, in the Welsh Basin, the three big stratigraphical divisions recognized by Lapworth (Woodcock rightly calls them 'supergroups' rather than 'systems') are bounded by regional unconformities. By contrast, on the cratons of North America (Ross & Ross 1992), the Russian Platform (Kaljo & Nestor 1990), the Andean Platform (Baldis et al. 1992), the Australian Platform (Nicoll et al. 1992) and the Yangtze Platform (Johnson et al. 1985), many stratigraphic sequences have synchronous boundaries and are likely to be related to eustatic sea-level changes.

88

W. S. M C K E R R O W

Llandovery area litho stratigraphy

Eocoelia evolution

Stricklandiid evolution (Williams1951 emended)

Graptolite zones

(Ziegler 1966

centrifugus

N

0.5 km 3) than are normally erupted from polygenetic volcanoes, and the whole accumulation has the aspect of a low plateau or plain. Where lavas infill and flow down valleys, they present the aspect of flooding the topography. Eruptions tend to be either from fissures or from point-source vents. It should be borne in mind that fissure-vents evolve with time into single-point vents (where volcanic plugs may develop) as wall-erosion locally widens the fissure (Bruce & Huppert 1990). In accounts of flood-basalt fields, attention is usually directed at the giant fields exceeding 100 000 km 2 in area and 100 000 km 3 in volume that are distributed sparsely through the geological record. Many small to moderate-sized flood-basalt fields also occur and are better analogues to the British Tertiary basalts. Good examples of moderate-sized fields are Rahat and K h a y b a r / I t h n a y n / K u r a in Saudi Arabia, both 20 000 km 2 (Camp & Roobol 1989; Camp et al. 1991) and the McBride and Nulla fields in Queensland, 5800 and 6600km 2 respectively (Stephenson et al. 1980). The volcanism in each field has been spread over the past 5 to 10 Ma and each field has the potential to erupt again. Some of the lava flows particularly in the Queensland fields are very large (Stephenson & Griffin 1976).

central volcano on the rift zone (Saemundsson 1986). Active central volcanoes include Hekla, Askja, and Krafla. Several scores of extinct central volcanoes are known amongst the Tertiary lava piles of eastern, northern and western Iceland, each more or less enclosed by flood basalts (Fig. 2).

Studies o f lava -flow structures Lava flows constitute the bulk of the British Tertiary and North Atlantic Provinces. In a logical scheme of things they would attract the greatest attention and might reasonably be expected to help distinguish flood basalts from central volcanoes. Strangely however most of the structures described by Geikie (he commented that 'a more detailed description of them seems to be required') still have not been explained. A notable exception from this neglect is the columnar-jointing shown by a few lava flows. The occasional spectacularly columnar jointed lavas as at the Giant's Causeway and Staffa attracted much attention from early geologists, some of whose explanations for the columnar rock would be classified today as fanciful. Tomkeieff (1940) recognized the multi-tiered character of columnar-jointed flows and initiated the modern terminology of colonnade and entablature for the tiers. The regularity of the prisms in the colonnade is now attributable to solidification of lava under static conditions, as when it infills a depression, and the closer joint spacing in, and common greater thickness of, the entablature are due to water cooling (Saemundsson 1970). Recent morphometric research (yet unpublished) on lava flows in Hawaii has an important bearing on the environment of basalt effusion. The thickness of basaltic lava is quite sensitive to the ground-slope angle. On slopes exceeding about 4° the average flow unit is about 1 m thick whereas on slopes under 2° it is 5 m or more thick. The lavas of Antrim and north-western Skye have thicknesses generally indicative of slopes of ~. .-""r - , . ~

/

~ : ~ : : : ~ : ' : '

MASSIF

ANDALUSITE / IN"

IN / 2000 l

SEIGNEURS

i :

:

SILLIMANITE . i IN""

~ '

:

: : '

:

:

"

:

'

:

~ ~ 7

~

,

OUT ~ _

- _ .o

/ + +

+

+

+

+

+

~r~+~++++

ANDALUSlTE

,""

"

-""~'+ +" +' +"

/

/

~'-'-'

+

,i~+:+~-:+

,_,.:-:,,_,£',-,,.:,':7,\-t,--~--

+

+

+

+

+

+ +

:+ +

1000 Carbonate

. . . . . . . . .. . .. . . . . . . . . . v . v . v re- .5.2.0 6366 Cu 1.3 1544 Pb 0.35 109 Total value (millions) 8019 Pennine orefield (AIston, Askrigg, Derbyshire)t (Dunham 1983) Pb >6.0 1404 Zn 0.34 165 CaF 2 6.5 618 BaSO 4 1.5 75 Total value (millions) £2261

Production (millions of tonnes)

Value (£ millions

(Alderton 1993) 2.5 2.0 0.25

7958 2314 78 10350

(Ixer & Vaughan 1993) 7.5 1755 0.36 174 6.8 646 1.6 80 2655

* Values based on London Metal Exchange Prices December 1993. t Values based on 75% of London Metal Exchange prices for December 1993 for Pb and Zn and on prices quoted for metallurgical grade fluorite and drilling-mud grade barite in November 1993 issue of Industrial Minerals.

li,.,

I

1850

I

I

1900

1950

2000

15COPPER

10o3

o

x

5-

O~ ILl Z Z 0 I.-

llllHIIlllllllllll 0 1850

aqueous solutions (hydrothermal fluids) in the formation of mineral veins in the Erzgebirge. But it was the French scholar, Ellie de Beaumont who really promulgated the hydrothermal theory of ore genesis in a series of papers in the mid-nineteenth century, citing in his paper of 1847 the presence of large fluid inclusions in support of this theory. The presence of relatively large, but very rare fluid-filled cavities, clearly visible to the naked eye had been described and scientifically investigated earlier by several eminent British scientists of the time, including Sir H u m p h r e y Davey (1822) and Sir David Brewster (1823), but their geological significance was not appreciated at the time.

S o r b y ' s contribution o f 1858 [

1900

~. . . . . . . .

~IEIIII]]

1950

2000

1950

2000

LEAD

50 40 30 20 10 0 1850

1900 YEARS

Fig. 2. United Kingdom mine production of base metals: ten-year averages, 1850-1990. Modified from Highley et al. (1991). Note the rise in tin production from 1970 to 1990 following the opening of the Wheal Jane mine in Cornwall. Since its closure in 1991, South Crofty is the last remaining underground tin mine in Cornwall and the only significant producer of base metals in England.

The impetus for Sorby's seminal work on fluid inclusions published in 1858 was the acquisition of a new scientific instrument for studying very thin sections of rocks and tiny crystals grown in the laboratory: the optical microscope. The preconceived notion that fluid inclusions were rare, scientific curios was soon dispelled when Sorby discovered that under the microscope ' . . . it is easier to see that the proportion of many millions to the cubic inch is very common in some minerals.' Through a series of careful experiments on laboratorygrown, water-soluble crystals, including potassium chloride, sodium chloride and potassium bichromate, Sorby beautifully demonstrated the following features. (1) Tiny droplets of mother liquor may be trapped and preserved as fluid inclusions during the growth of crystals. In natural samples they provide a unique record of the nature and composition of ancient mineral-forming fluids. (2) When crystals are formed from aqueous solutions at elevated temperatures differential contraction of the contained fluid takes place in a manner similar to the development of the head-space in a mercury-in-glass

240

A. H. R A N K I N

thermometer. This results in the development of a 'vacuity' or contraction vapour bubble in the inclusion fluid on cooling. (3) The relative size of the vacuity varies depending on the temperature at which the crystal grew. The important conclusion was that, by determining the temperature at which the liquid and vapour components become homogeneous again (the homogenization temperature, T,), natural fluid inclusions could also be used as geothermometers for a variety of rocks and minerals. Sorby then carried out a further series of simple experiments on the thermal expansion of various salt-water solutions contained within glass tubes and established the following empirical relationship between vapour-liquid ratios in fluid inclusions and the temperatures and pressures at the time of trapping: v = (Bt + Ct2)(1 - 0.00000271p) - 0.00000271p.

where: v -- relative size of the vacuity t = temperature in degrees centigrade p -= pressure in atmospheres B and C = constants whose values depend on the nature and strength of the salt solution in the cavity. Sorby took care to warn his readers that these equations were accurate only for moderate values of temperature and pressure, and they were advised to adopt them provisionally. He also recognized some of the potential pitfalls of his new geothermometric method which could lead to erroneously high estimates of temperatures. These included: (i) heterogeneous trapping of discrete vapour bubbles (air in his experiments); (ii) leakage if the crystals are subsequently subjected to higher temperatures through, for example, the use of Canada balsam as the mounting medium. Sorby's observations (Fig. 3) were not restricted to crystals grown from aqueous solution. He recognized and described glass inclusions in crystals of iron silicates, and of Humboldtilite in slags formed from copper-nickel and iron smelting, and good examples of ~stone cavities' in pyroxene

from blast furnaces at Masborough in his native Sheffield. He inferred that both were products of trapped silicate melts. Thus, by simple microscopic examination of natural crystals, Sorby had devised a method of determining whether rocks and minerals had been formed from igneous fusion or the action of water, an issue that was still being hotly debated at the time. Sorby applied this new-found method to rocks and minerals from a number of geological environments including the granites, elvans (quartzporphyry dykes) and mineral veins of Cornwall. He recognized the existence of stone cavities in both the elvans and granites of St Austell and Land's End, but noted their absence in quartz from associated mineral veins where aqueous inclusions predominated. He estimated a temperature of around 200°C for the mineral veins from St Michael's Mount and the Camborne area. These were only slightly lower than his estimate for the granites themselves ( m e a n = 216°C). Whilst the inclusion evidence suggested hydrothermal processes were important in the formation of the mineral veins, the apparent co-existence of aqueous and stone inclusions in the granite led Sorby to postulate the development of a separate aqueous phase during cooling of the granite melt. In this respect Sorby appeared to be amongst the first committed magmatic hydrothermalist as far as mineral deposits were concerned, and a strong supporter of Ellie de Beaumont's views (1847) on the matter. S o r b y ' s o w n w o r k ' u n d e r the m i c r o s c o p e '

When Sorby delivered his classic paper to a meeting of the Geological Society in London on December 16 1857 it received considerable attention, not least from its chairman at that time, Leonard Horner. According to Sorby (reported in Judd 1908), Horner commented that he ' . . . had been a member of the Geological Society ever since its foundation, and during the whole of that time he did not remember any paper having been read which drew so largely on their credulity'. There was considerable scepticism and criticism of Sorby's work, as Sorby himself explained (see Judd 1908): 'In those early days people laughed at me. They quoted

2O

16 21

18

20,

STP~UCTUI'~.~E

ow

GR~fSTALS.

/'~ Fig. 3. Original sketches by Sorby (1858) of fluid inclusions in laboratorygrown crystals.

H Y D R O T H E R M A L O R E F I E L D S AND ORE FLUIDS Saussaure that it was not a proper thing to study mountains through a microscope'. The fate of further development of the subject, at least in the UK, was really determined through publication of another paper in the Quarterly Journal of the Geological Society by an equally eminent geologist of his time, J. A. Phillips, who by a quirk of fate happened to be one of Sorby's few supporters! Phillips carried out extremely detailed studies on Cornish veins and granites and obtained very wide temperatures using Sorby's fluid inclusion calculation method. Phillips even constructed a crude microscope heating stage but failed to homogenize many inclusions. He, therefore, concluded that because, gas-liquid inclusions showed such considerable variations in the relative size of their vacuities, Sorby's thermometric method was fallacious (Phillips 1875). What both authors failed to realize, and what left the subject in limbo for more than 100 years, was that fluid inclusions may be secondary, reflecting post-depositional recrystallization, as well as primary in origin.

241

outer barite zone (Fig. 4), provides a landmark in British ore geology. It initiated debate, speculation and scientific enquiry on the nature and origin of mineralizing fluids and the role of granites in ore genesis, which has extended over 50 years. Strongly influenced by previous work on zonation of the mineral lodes of Cornwall, and their association with granite (Davison 1927), Dunham postulated that: (i) falling temperature away from a focal point for mineralization was 'the prime cause of the zone of distribution' for the Pennine ores; (ii) mineralization was the result of juvenile fluids related to concealed granite at depth coincident with the fluorite zone outcropping at surface. Remarkably, gravity surveys (Bott & Masson-Smith 1957) and drilling (Dunham et al. 1965) confirmed the existence of a granite cupola directly under the orefield (Fig. 5), although it was soon apparent that the granite far from being co-eval with the overlying mineral deposits (post-Carboniferous) was very much older (Devonian) and could not possibly be the direct source of the mineralization

Development of ore genetic theory 1900-1965 The first half of the present century saw rapid advances in the application of fundamental principles of physics and chemistry to the study of ore deposition, and the growth in field observations arising out of an increasingly scientific, as opposed to serendipic, approach to mineral exploration. The most influential work in the first part of this period was that of W. Lindgren in the U S A who formulated a comprehensive and eloquent classification scheme (Lindgren 1933) based on field observations and the inferred physico-chemical conditions of formation. Lindgren recognized that hydrothermal deposits were the most important class of base and precious metal deposits, so much so that he further sub-divided them into hypothermal, mesothermal and epithermal classes (Table 2). Many modifications have subsequently been made to Lindgren's original proposal but many of the concepts and terms originally introduced still remain (Guilbert & Park 1986). Occasional attempts by the magmatists to challenge the hydrothermalist's view of ore deposition (e.g. Spurr 1923) were soon dispelled, although Spurr is credited by Guilbert and Park (1986) as being the first to provide a generalized statement of the theory of mineral zones, so well illustrated in Cornwall and the north Pennines. The first detailed accounts of primary zonation in these areas were provided by Davison (1927) and Dunham (1934) respectively.

Contribution o f Dunham's 1934 paper The recognition and delineation by Dunham (1934) of a distinctive pattern of mineral zonation in the north Pennine orefields in which a central fluorite zone is surrounded by an Table 2. Lindgren's (1933) classification of hydrothermal

ore deposits according to temperature and depth (pressure) or formation Sub-division

Temperature range

Hypothermal Mesothermal Epithermal

500-300 ° 300-200 °C 200-50 °C

Developments in fluid inclusion techniques and methodology Little interest was shown in the use of fluid inclusions as geological indicators during the early part of the twentieth century. However, during the period 1940-1965 a steady stream of pioneering papers in North America (notably by Roedder and Smith), the USSR (notably by Kalyuzhnyi, Lemmlein and Yermakov) and France (notably by Deicha), culminated in the reconciliation of many of the problems which had given cause for such doubt following Sorby's early work (for reviews see Deicha 1955; Smith 1953; Lemmlein 1956; Kalyuzhnyi 1960; Yermakov 1965; Roedder 1972). These included: (1) the recognition that several generations of inclusions, representing both primary and secondary crystallization processes, may be preserved in a single crystal; (2) the realization that both necking-down and leakage can alter the liquid-vapour ratios of fluid inclusions, and the recognition that such effects can give rise to erroneous temperature estimates; (3) the development of suitable microscope heating stages for accurately recording homogenization temperatures of fluid inclusions, and the realization (at least in North America and Europe) that the alternative decrepitation method of fluid inclusion geothermometry was unreliable. After some 100 years of inactivity and apparent disinterest in fluid inclusions within the UK following Sorby's 1858 paper, studies on British ore deposits suddenly erupted. The catalyst for this was K. C. Dunhan who invited a young researcher from the USA, F. J. Sawkins, to his department in Durham in the 1960s to apply the new-found methods of fluid inclusion studies to the Pennine and south west England mining districts (Sawkins 1966a, b).

Depth Great Intermediate Shallow

Characterization o f the Pennine orefluids: fluid inclusion and related studies Sawkin's pioneering fluid inclusion study of mineralization from the Alston block showed that, as Dunham (1934) had

242

A.H.

RANKIN

J / x //

B

.EXHA~

J \

\ \\ k

\

\

/

A

I

\

",,

,

t

,.1/

U

F

-/

/I \'~,x~

/

.~, " ' ' ~ /

I

c.*oeL

F

.,4

•

WOL$1 N G m A i,~

J, \ I

\

iN

\

t.*NC~ON

L

T

F

~k

• *PPLtB,

MILES

\-

o

~

2

DISTRIBUTION OF GANGUE MINER.&LS IN DEPOSITS OF THE PEN.'INE TYPE. O.P.F. = O u t e r P e n n i n e f a u l t . s y s t e m . S.F. = Stubliek fault-system, B.F.D. = Burtreeford monocline.

Yennine t y p e veins, t h i n b l a c k lines. G r e a t S u l p h u r Vein t~,'pe veins, d o t t e d lines. Z o n a l b o u n d a r i e s , h e a v y b l a c k lines.

Zone F = F l u o r s p a r . ,, A = F l u o r s p a r a n d b a r y t e s , or n e i t h e r m i n e r a l . ,, B = B a r i u m m i n e r a l s .

Fig. 4. Map reproduced fom Dunham's paper of 1934 showing the inner fluorite zone and outer barite zone of the North Pennine orefield (Aiston block)•

postulated, formation temperatures for the outer barite zone (less than 130°C) were significantly lower than those recorded for the central fluorite zone (up to about 177 °C for fluorite and from 181-216 °C in early quartz). Sawkins thus proposed a two-fluid model for ore genesis involving the mixing of granite-derived fluids and cooler barium-rich fluids derived from surrounding sediments. A two-fluid model was also proposed by Solomon et al. (1971) on the basis of isotopic evidence, although these authors postulated that mineralization in the fluorite zone was due to circulation of connate brines rather than juvenile fluids. Sawkins' preliminary results were soon substantiated by more detailed fluid inclusion studies on the Alston Block (Smith 1974; Smith & Phillips 1974) and extended over the next 15 years to include the mineralization of the Askrigg Block and Derbyshire Dome (Roedder 1967; Rogers 1977, 1978; Smith 1973, 1974; Greenwood & Smith 1977; Small 1978; Atkinson et al. 1982; Christoula 1992). Salinities from all three orefields typically cluster around 20-25 eq. wt% NaCI, but a

decrease has been noted from north to south in the homogenization temperatures of fluid inclusions in fluorite from these three areas (Rogers 1977; Atkinson et al. 1982; see also Table 3). Recent unpublished results based on more refined methods have essentially confirmed these trends (Christoula 1992) as shown in Fig. 6. However, a high temperature ( T h = l l 0 - 1 6 0 ° C ) , low salinity fluid (05 eq. wt% NaCI) is also discernable in apparently primary inclusions in fluorite from Derbyshire. This is in agreement with the preliminary results presented by Roedder (1967) and by Moser et al. (1992). Liquid-hydrocarbon-bearing fluid inclusions are also reported from the Derbyshire orefield. Detailed studies by these authors have revealed that the hydrocarbon inclusions have all the hallmarks of natural petroleum seepages in the area, but coexist only with the anomalously high temperature, low salinity fluids that have a restricted occurrence in the northeastern part of the orefield (Fig. 6). First melting temperatures between - 6 5 and - 5 0 ° C ,

HYDROTHERMAL

I I

Northumberland basin ' ....e',-......-.-'--~--

I

i - - _, ./ . - - / .

.

.

/ ~

~. ~" / • "~ NEWCASTLE

~ . A L S .T O N

~"

~ ~. ~ ~

\ ,.~

\

/

~

i

:

i

i

AN ,ill

I

/

I

:

i

~ ~

t

~" " I i iiii"

•

I I

North Sea

Stainmore trough

" - - -- ~

:)

J .,~

Fault zones

.-""

-'"

?

"'" Outer limit of Fluorite Zone

" :+:::::::::~ Outline of ,::+:+:+:+::-~' ::+: :.7.:.:.:.:.:-:.,, ~:~:~:;" Buried Granite

Table 3. Summary of published homogenization temperatures of fluid inclusions in fluorite from the Pennine orefield

Alston

110-177 °C

Askrigg

92-143 °C 92-164 °C

Derbyshire

70-140 °C 74-127 °C 92-127 °C 74-158 °C

Sawkins (1966a) Smith & Phillips (1974) Smith (1973) Rogers (1978) Small (1978) Roedder (1967) Smith (1973, 1974) Rogers (1977) Atkinson et al. (1982)

indicative of calcium (with or without magnesium)-bearing brines (Roedder 1984), are commonly reported from 'normal' high salinity, aqueous inclusions from the Pennine orefields. Recent estimates of the ratio of sodium to calcium based on careful micro-thermometric studies of these inclusions have revealed a remarkable uniformity throughout the ore field in terms of their calcium contents (Christoula 1992). The above fluid characteristics are typical of those reported from a number of Mississippi Valley Type deposits worldwide (see Table 4). This was one of the main reasons why consensus opinion gradually changed from an ore genetic model involving a juvenile, granite-derived source, to one in which basinal brines, analogous to modern day oil-field brines (Carpenter et al. 1974) was favoured; but see Russell & Skauli 1991). The other was that the underlying Weardale granite clearly postdated mineralizing events by at least 100 million years and could not possibly have been the direct source of the mineralizing fluids. D u n h a m was amongst the first to advance the basinal brine theory, as opposed to his original 'juvenile' source theory, for the Pennine mineralization. He also proposed that the ore field represents a fluoritic sub-type of the Mississippi Valley Type class of deposit (Dunham et al. 1983).

km

I I

I

DURHAM

',,"iiiiiiiiiiii ...--.___...--i ~" ~,

243

I

~

~ ~

Fluoritezone

~.

"\

OREFIELDS AND ORE FLUIDS

' e l

Cusp within Granite

Fig. 5. Location of the fluorite zone of the Alston block, North Pennine orefield, in relation to Carboniferous basins, basement structures and the underlying Weardale granite. Modified from Greenwood & Smith (1977) after Bott (1967) and Dunham (1934).

Apart from variations in homogenization temperatures from north to south, and also in the potassium content of the inclusion fluids, other mineralogical and geochemical differences are apparent between different areas of the orefields (Dunham 1983; D u n h a m & Wilson 1985; Brown et al. 1987; Colman et al. 1989a). The most significant of these are a granitic association of trace elements and minerals, observed within the Alston Block mineralization, but absent within that of the Derbyshire dome. The favoured explanation for such a difference is the variable influence of granites beneath different areas of the orefield (Brown et al. 1987; Rankin & G r a h a m 1988; Colman et al. 1989a). Current theories of ore genesis for the north Pennine orefield thus link directly back to Du n h a m ' s classic study in which he suggested that the epigenetic vein mineralization is associated with buried granites. The major difference between Dunham's original hypothesis and the one generally accepted today is based on the fact that the underlying Weardale granite is substantially older than the overlying mineralization. Rather than acting as direct sources of mineralizing fluids, it is now generally considered that the granites have simply acted in variable ways as heat engines and loci for channelling mineralizing fluids from a sedimentary source (e.g. Brown et al. 1987). Ore fluids o f Cornubia: f l u i d inclusion a n d related studies Although Dunham's (1934) recognition of mineral zonation in the North Pennine orefield post-dated its recognition in south west England (Davison 1927), the first major publication on fluid inclusions in minerals from the Cornubian orefield for almost a century also came from Sawkins (1966b) working at D u r h a m University, Dunham's stamping ground. Sawkins (1966b) published only a preliminary account of his findings based on a few samples but they clearly showed a temperature decrease from the early tin-tungsten mineralization (Th = 300-450 °C), to the

244

A. H. R A N K I N 240

'

'

,

I

'

Table 4. Typical fluid inclusion characteristics of Mississippi Valley

I

type deposits compared with those of the Pennine fluorite deposits Alston

200 CL

E

Block

160

o

tO t~

120

.N

tO O~ 0

80

E 0 e-

40

I

240

1'0

20

I

I

a.

E

I

'

160 120

0

c

80

E 0

t-

40

I

o 240

lO

'

2'0

I

30

40

I

South Pennines m a.

E O

200

160

cO

120 N

T: ®

8' E

Pennine Fluorites?

100-150 °C

70-180°C

>l.O

>1.0

15 to >20

15-27

Na, Ca, K C1 Liquid and gaseous hydrocarbons

Na, Ca, K C1 Oil present in inclusions from Derbyshire and light hydrocarbons from Askrigg and Alston$

* Based on summary by Roedder (1984). t Temperature, salinity, density and solute data from sources quoted in Table 2 and also in Christoula (1992). $ From Ferguson (1991) and Moser et al. (1992).

.N_ O~ 0

'

4O

Askrigg Block

200

0 ,,,,., co '.~ t~

3'0

Temperature (Th) Fluid density (g cm 3) Salinity (eq. wt% NaCI) Major cations Major anions Organic matter

Typical MVT Deposit*

8o

0

40 0

10

20

30

40

Salinity (Equiv. wt. % NaCI)

Fig. 6. Homogenization temperature versus salinity plots showing the 'fields' for fluid inclusion data from each of the three Pennine orefields. The main fields show constant salinities but decreasing temperatures of homogenization southward. Compiled from recent data from Christoula (1992). Note the presence of a separate, atypical grouping of low salinity inclusions restricted to the Castleton area of Derbyshire. Their origin and significance is not yet fully understood (see Moser et al. 1992). later copper-iron-arsenic-zinc-sulphide mineralization (Th--200-350°), to the final lead-zinc-fluorite mineralization (Th = 100-180°C). Further evidence that temperatures decreased away from centres of intense hydrothermal activity and mineralization within the granites, the so called 'emanative centres' of Dines (1956), was provided by a novel use of fluid inclusions as exploration guides for blind ore bodies in the region by Bradshaw & Stoyel (1968).

Major contributions to our understanding of metaUogeny and fluid processes associated with the Cornubian granites were made in the 1970s in three doctoral theses by Alderton (Kings College, London), Charoy (CRPG, Nancy, France) and Jackson (Kings College, London) in 1976, 1979 and 1977 respectively. A steady stream of papers on the fluid inclusion characteristics of specific mineralization and alteration styles followed (Halls et al. 1985 give the bibliographic details). By now, the necessity of distinguishing between primary and secondary inclusions, which had so hampered Sorby's early advances in the area, had been realized (e.g. Jackson et al. 1977). Sawkins' preliminary findings were substantiated. Based on studies at St Michael Mount (Jackson & Rankin 1976) and at Cligga Head (Jackson et al. 1977; Charoy 1979) in Cornwall, the fluid inclusion evidence for main stage tin-tungsten-sulphide mineralization pointed to high temperature, moderately saline fluids (Th = 200-450°C, salinity-- 5-20eq. wt.% NaC1). The fluids responsible for late-stage, cross-course veins, either barren of mineralization or carrying minor iron, lead and zinc minerals, were confirmed to be of much lower temperature and salinity, sometimes approaching that of pure water. In contrast, studies of fluorite from a number of cross-course lead-zinc deposits showed that the fluids responsible for this stage of low temperature mineralization (Th=100-150°C) were substantially more saline (2025eq. wt% NaC1) with low eutectics indicative of high calcium contents (Alderton 1978). This led Alderton to propose a separate basinal brine source (cf. Mississippi Valley Type-fluids in Table 3) for fluids responsible for this style of mineralization compared to the magmatic and/or meteoric source generally envisaged for the earlier mineralization. More comprehensive and systematic fluid inclusion studies, coupled with stable isotope studies of hydrogen and oxygen, were carried out in the 1980s on mineralization and alteration assemblages from other parts of the Cornubian orefield (notably by Alderton & Rankin 1983; Jackson et al. 1982; Shepherd et al. 1985). These results, recently

H Y D R O T H E R M A L O R E F I E L D S AND ORE FLUIDS

245

Conclusions and wider implications 400

.,,.. ~ . . ~? .

,.,.~

oc 3 0 0

g

.

7..

o

oE

.¢

200

-

loo

-

-?--9 ~

Kao 0

)

1I0

'

;0

'

3'0

'

4'0

Salinity (Equiv. wt. % NaCl)

Fig. 7. Homogenization temperature versus salinity plot showing 'fields' for fluid inclusions from the Cornubian orefield, based on compilations of data from various sources as summarized by Jackson et al. (1989) and Wilkinson (1990b). Note the large spread of data for 'main-stage' Sn-W-Cu mineralization and the similarity between data from the 'cross-course' lead-fluorite mineralization and the Pennine fluorites (see Fig. 6). The separate low salinity, low temperature field marked 'kao' is characteristic of late stage quartz in intensely kaolinized granite (Alderton & Rankin 1983). The field for main-stage mineralization may be extended to higher salinities when the examples from the Dartmoor area are considered.

supplemented and reviewed by Jackson et al. (1989) and Wilkinson (1990a), have not substantially altered earlier views of the temperature and salinity ranges of the fluids, except that the temperature range has now been extended down to less than 100°C for late-stage fluid processes associated with the formation economic china clay deposits (Fig. 7). The range of fluids previously envisaged as responsible for mineralization has also been extended to include a component from metamorphic sources (as confirmed by Wilkinson 1990b). In essence, the fluid inclusion, isotopic and geochronological evidence points to a multi-stage process involving fluids from a variety of sources and a protracted period of mineralization from the time of the emplacement of the granite (270-300 million years ago) extending to Mesozoic and even up until the present day. What about the granites themselves? After all, Sorby's original 1858 contribution to the development of fluid inclusion studies in the UK was directed mainly towards the granites from Aberdeenshire and Cornwall. His original observation that fluid inclusion abundances vary markedly from one area to the next provided the main impetus for systematic studies of fluid inclusion populations in granite quartz from Devon and Cornwall (Alderton & Rankin 1983; Rankin & Alderton 1983). These studies have revealed an assemblage of mainly secondary aqueous inclusions whose abundance and distribution reflects the extent to which different parts of the batholith have been affected by early and late-stage fluids of diverse origin; a feature which explains why Phillips' (1875) comments on Sorby's methods were so valid in the early days.

Historically, the role of granites was thought to be of paramount importance in providing a source of both fluids and metals for both the north Pennine (Dunham 1934; Sawkins 1966a) and Cornubian orefields (Dines 1956). Magmatic differentiation processes and the exsolution of metal-bearing, hydrothermal fluids from the cooling granites of Cornubia are still believed to play an important role in early tin-tungsten-copper mineralization in this area (Jackson et al. 1989). Geochemical evidence of such fluids may be found trapped and preserved in inclusions in tourmaline veins (Bottrell & Yardley 1988; Wilkinson et al. 1994), in breccia pipes (Halls et al. 1985) and in the granites themselves (Rankin & Alderton 1983; Alderton et al. 1992). However, convective fluid flow models involving the episodic circulation of fluids from different sources over long periods of time, are now generally accepted for at least some of the main-stage mineralization and most of the late-stage mineralization and alteration including kaolinization (Sams & Thomas-Betts 1988; Jackson et al. 1989). It appears that the high heat production capacity of the radioelement-rich granites provided the driving force for convection of late-stage meteorite waters and basinal brines in the region (Tammemagi & Smith 1975). In the north Pennine orefields the underlying Weardale granite is no longer considered to be the direct source of mineralizing fluids. High homogenization temperatures of the fluid inclusions and the presence of 'granitic' minor minerals in surface veins (Brown et al. 1987) has led to the suggestion that the underlying granite has acted beyond its previously conceived role as a broad structural control on the mineralization; it is now believed to have supplied at least some of the trace elements (Rankin & Graham 1988: Christoula 1992) and radio-thermal heat to the mineralizing fluids. Hydrothermal theories for ore genesis have come a long way since the early pioneering work of Sorby and Dunham. In the past 20 years fluid inclusion studies have become firmly established as the most important source of information on the physical and chemical properties of ancient mineral-forming fluids. A range of modern instrumental methods are now available for the geochemical analysis of the tiny droplets of fluids contained within the inclusions (Roedder 1990; Rankin et al. 1993), even to the extent that variations in fluid inclusion geochemistry may be used in mineral exploration (Alderton et al. 1992). Systematic and integrated geological, mineralogical, geochemical and fluid inclusion studies have now been carried out on a range of mineral deposit types worldwide. Based on these studies and consideration of experimental and thermodynamic data for mineral transport and deposition in high temperature fluids, it is now widely accepted that hydrothermal processes are the most important of all primary ore-forming processes in the Earth's crust (Roedder 1984; Guilbert & Park 1986). Our definition of a hydrothermal ore-forming fluid has changed substantially since they were first envisaged by the early ore geneticists as 'emanations form cooling granites'. Based on fluid inclusion and stable isotopic evidence we now recognize that such fluids are of diverse origins encompassing the whole temperature range between those responsible for diagenetic and igneous processes (Fig. 8). Most hydrothermal ore-fluids are demonstrably alkali-chloriderich brines, of variable salinities, and with ore metal

246

A.H.

NATURE

AND

FLUIDS

ORIGIN IN THE

OF HYDROTHERMAL EARTH'S

CRUST

"Any hot aqueous fluid that exists in the Earth's crust"

T°C = ~50 to >500°C

RANKIN

M. W e s t e r m a n , who are continuing the traditions set out by Sorby in investigating, in yet m o r e detail, the further mysteries and scientific importance of fluid inclusions in the granites of south west England. The helpful c o m m e n t s of an a n o n y m o u s referee and the Editor, M. J. Le Bas, are also gratefully acknowledged. I am also i n d e b t e d to M. J. Le Bas for first introducing me to fluid inclusions, as a p o s t g r a d u a t e student u n d e r his supervision at the University of Leicester, thus opening up my lifelong interest in the subject and h y d r o t h e r m a l processes in general.

Composition = Na - K - Ca - CI - S O 4

References

Salinity = 0 to >50 weight % salts Ore metals = generally at ppm levels

Origin = various (see below)

/,/'#/f///,~#,

SURFACE

';"';";;';';

METAMORPHIC

IGNEOUS INTRUSION

Fig. 8. S u m m a r y diagram illustrating the impact of recent fluid inclusion studies (see summaries by R o e d d e r 1984) on our u n d e r s t a n d i n g of the nature and origin of h y d r o t h e r m a l ore-forming fluids in the Earth's crust. T e m p e r a t u r e s span the whole range predicted by Lindgren (Table 2). H o w e v e r , the original concept that all h y d r o t h e r m a l ore deposits were linked to magmatic processes has been substantially modified in recent years mainly in the light of stable isotopic studies. It is n o w recognized that h y d r o t h e r m a l fluids, and the metals t h e y contain, m a y be derived from a variety of sources.

contents measured in parts per million (ppm) rather than percentages. The importance of basinal brines (connate waters in Fig. 8) and magmatic and meteoric waters in the genesis of the Mississippi Valley Type and granite-associated ores, as exemplified by the Pennine and Cornubian orefields, is now universally recognized. It is equally apparent, from comparable studies on other deposit types elswhere in the world, that seawater plays a dominant role in the formation of volcanic-hosted massive sulphide (VMS) deposits and that carbon-dioxide-rich metamorphic fluids developed during dehydration and decarbonation processes accompanying metamorphism are, at least in part (but see Boyle 1991 for discussion) responsible for the formation and modification of gold-quartz veins in Archaean greenstone belts. I am grateful to all the former staff and students at Imperial College, L o n d o n for much stimulating discussion on the role of fluid inclusion studies in relation to metallogeny. I also thank my colleagues at Kingston, especially my g r a d u a t e students W. Cox and

ALDERTOND.H.M. 1976. The geochemistry of mineralisation at Pendarves and other Cornish areas. PhD Thesis, University of London. 1978. Fluid inclusion data for lead-zinc ores from South-west England. Transactions of the Institution of Mining and Metallurgy, B87, B132-135. 1993. Mineralisation associated with the Cornubian granite batholith. In: PATTmCK, R.A.D & POLYA,D.A. (eds) Mineralization in the British Isles. Chapman-Hall, London, 270-354. -& RANKIN, A.H. 1983. The character and evolution of hydrothermal fluids associated with the kaolinized St. Austell granite, southwest England. Journal of the Geological Society, London, 140, 297-309. --, -& THOMPSON,M. 1992. Fluid inclusion chemistry as a guide to tin mineralization in the Dartmoor granite, south-west England. Journal of Geochemical Exploration, 46, 163-185. ATKINSON, P., MOORE, J. t~ EVANS, A~M. 1982. The Pennine orefields of England with special reference to recent structural and fluid inclusion investigations. Bulletin of the Bureau de Recherche Gites Mineraux, Section II, no. 2, 149-156. BAUER, G. [Published under his latinized name of G. Agricola] 1556. De Re Metallica. (Translated into English in 1950 by: H.C. and L.H. Hoover, Dover New York). BOTT, M.H.P. 1967. Geophysical investigations in the northern Pennine basement rocks. Proceedings of the Yorkshire Geological Society, 36, 139-168. -8z MASSON-SMITH, D. 1957. The geological interpretation of a gravity survey of the Alston block and the Durham coalfield. Quarterly Journal of the Geological Society of London, 113, 93-117. BOTI'RELL, S.H. & YARDLEY, B.W.D. 1988. The composition of granite derived ore fluid from SW England determined by fluid inclusion analysis. Geochimica et Cosmochimica Acta, 52, 585-588. BOYLE, R.W. 1991. Auriferous Archean Greenstone-Sedimentary Belts. Economic Geology Monograph. 8, 164-191. BRADSHAW, P.M.D. & STOYEL, A.J. 1968. Exploration for blind orebodies in southwest England by the use of geochemistry and fluid inclusions. Transactions of the Institution of Mining and Metallurgy, 77, 437-448. BREWSTER, D. 1823. On the existence of two new fluids in the cavities of minerals, which are immiscible, and possess remarkable physical properties. Edinburgh Philosphophical Journal, 9, 268-270. BROWN, G.C., IXER, R.A., PLANT, J. • WEBB, P.C. 1987. Geochemistry of granites beneath the North Pennines and their role in ore formation. Transactions of the Institution of Mining and Metallurgy, 96, 65-76. CARPENTER, A.B., TROUT, M.L. & PICKET~, E.E. 1974. Preliminary report on the origin and chemical evolution of lead- and zinc-rich oilfield brines in central Mississippi. Economic Geology, 69, 1191-1207. CHAROY, B. 1979. Definition et importance des phenomenes deuteriques et des fluides associes dans les granites. Consequences metallogenique. Science Terre, Nancy, Memoire, CHRISTOULA, M. 1992. Fluid inclusion geochemistry of selected epigenetic, low temperature mineralization in the U.K. Ph.D. Thesis, University of London. COLMAN, T.B., FORD, T.D. & LAFFOLEY, N.D'A. 1989a. Metallogeny of Pennine orefields. In: PLANT, J.A. & JONES, D.G. (eds) Metallogenic models and exploration criteria for buried carbonate-hosted ore deposits---a multidisciplinary study in eastern England. Institution of Mining Metallurgy, London, 13-24. , JONES, D.G., PLANT, J.A. & SMITH, K. 1993. Metallogenic models for carbonate-hosted (Pennine and Irish-style) mineral deposits. In: PLANT, J.A. & JONES, D.G. (eds) Metallogenic models and exploration criteria for buried carbonate-hosted ore deposits--a multidisciplinary study in eastern England. Institution of Mining Metallurgy, London, 123-133. DAVISON, E.H. 1927. Recent evidence confirming the zonal arrangement of minerals in the Cornish lodes. Economic Geology, 22, 475-479. DAVEY, H. 1822. On the state of water and aeriform matter in cavities found in certain crystals. Philosophical Transsactions Royal Society of London, part 2, 367-376. -

-

-

-

37.

HYDROTHERMAL

OREFIELDS

DEICHA, G. 1955. Les lacunes des cristaux et leurs inclusions fluides. Masson et Cie, Paris. DINES, H.G. 1956. The metalliferous mining region of south-west England. Memoirs of the Geological Survey of Great Britain, (HMSO London). DUNHAM, K.C. 1934. Genesis of the North Pennine ore deposits. Quarterly Journal of the Geology Society of London, 90, 689-720. 1983. Ore genesis in the English Pennines : A fluoritic subtype. In: KISVARSANYI, G., GRANT, S.K., PRATT, W.P. & KOENIG, J.W. (eds) -

-

International conference on Mississippi Valley type, lead-zinc deposits. Rolla Conference Proceedings Volume, Univ. Missouri-Rolla Press, 85-112. - t~ W I L S O N , K. 1985. Geology of the Northern Pennine orefield: Volume 2. Stainmore to Craven. Economic Memoir of the British Geological Survey. , HODGE, B.C. & JOHNSON,G.A. 1965. Granite beneath visean sediments with mineralization at Rookhpe, northern Pennines. Quarterly Journal of the Geology Society of London, 121, 383-414. ELL1E DE BEAUMONT, J.D. 1847. Note sur les emanations volcaniques et metalliferes. Societe Geologique de France, Bulletin, 2e series, 4, 1249-1 FERGUSON, J. 1991. The organic geochemistry of hydrocarbon gases in fluorite from Northern England. Journal of Petroleum Geology, 14~ 221-228. GREENWOOD, D.A. & SMITH, F.W. 1977. Fluorspar mining in the northern Pennines. Transactions of the Institution of Mining and Metallurgy. 86, 181-190. GUIEBERT, J.M. & PARK, C.F. (JR.). 1986. The geology of ore deposits'. Freeman and Company, New York. HALLS, C., EXEEY, C.S. & BRUTON, E.V. 1985. A bibliography of magmatism and mineralization in southwest England. Occasional Paper No. 5, Institution of Mining and Metallurgy, London. HIGHLEY, D.E., SEATER, D. t~z CHAPMAN, G.R. 1991. Minerals Extraction-positive and negative trends. Bulletin of the Institution of Mining and Metallurgy (Minerals Industry International), No 998, 15-20. IXER, R.A. & VAUGHAN,D.J. 1993. Lead-zinc-fluorite-baryte deposits of the Pennines, North Wales and the Mendips. In: PATTRICK,R.A.D. & POLYA, D.A. (eds) Mineralization in the British Isles. Chapman & Hall, London, 355-395 JACKSON, N.J. 1977. The geology and mineralisation of the St. Just mining district, west Cornwall. Ph.D. thesis, University of London. -1979. Geology of the Cornubian tin field. Bulletin of the Geological Society of Malaysia, 11, 209-237. & RANKIN, A.H. 1976. Fluid inclusion studies at St. Michael's Mount. Proceedings of the Ussher Society, 3, 430-434. , HALLIDAY, A.N., SHEPPARD, S.M.F. & MITCHELL, J.G. 1982. Hydrothermal activity on the St. Just mining district, Cornwall, England. In: EVANS A.M. (ed.) Metallization associated with acid magmatism. Wiley & Sons, Chichester 137-179. , MOORE, J. McM. & RANKIN, A.H. 1977. Fluid inclusions and mineralisation at Cligga Head, Cornwall. Journal of the Geological Society, London, 134, 343-349. - - , WILEIS-RICHARDS,J., MANNING,D.A.C. & SAMS, M. 1989. Evolution of the Cornubian Ore Field, Southwest England : Part II. Mineral Deposits and Ore-froming Processes. Economic Geology, 84, 1101-1133. JUDD, J.W. 1908. Henry Clifton Sorby and the Birth of Microscopical Petrology. Geology Magazine, Decade V, 5, 192-204. KALYUZHNYI, V.A. 1960. Methods of study of multiphase inclusions in minerals. Kiev Izdatel. Akad. Nauk. Ukrainskoy RSR [in Russian, cited in Roeder 1948 below]. LEMMLEIN, G.G. 1956. Formation of fluid inclusions and their use in geological thermometry. Geochemistry International, 6, ]Translated from Russian]. LINDGREN, W. 1933. Mineral Deposits, 4th edition. McGraw-Hill, New York. MOSER M. R., RANKIN, A.H. & MILLEDGE, H.J. 1992. Hydrocarbon-bearing fluid inclusions in fluorite associated with the Windy Knoll bitumen deposit, UK. Geochimica et Cosmochimica Acta, 56, 155-168. PHILLIPS, J.A. 1875. The rocks of the mining districts of Cornwall and their relation to metalliferous veins. Quarterly Journal of the Geologaical Society of London, 31,319-345. PLANT, J.A. & JONES, D.G. 1989. Introduction. In: PLANT,J.A & JONES, D.G. (eds) Metallogenic models and exploration criteria for buried carbonate-

-

-

hosted ore deposits--a multidisciplinary study in eastern England.

-

Institution of Mining and Metallurgy, London, 1-4. RANKIN, A.H. & ALDERTON, D.H.M. 1983. Fluid inclusion petrography of SW England Granites and its potential in mineral exploration. Mineralium Deposita, 18, 335-347. & GRAHAM, M.J. 1988. Na, K and Li contents of mineralizing fluids in -

AND

ORE

FLUIDS

247

the Northern Pennine Orefieid, England and their genetic signifcance. Transactions of the Institution of Mining Metallurgy, 97, 99-107. --, HERRINGTON,R.J., RAMSEY, M.R., COLES, B., CHRISTOULA,M. & JONES, E. 1993. Current developments and applications of ICP-AES techniques for the gochemical analysis of fluid inclusions in minerlas. Proceedings of the Quadriennial IAGOD Symposium, 8, 185-198. ROEDOER, E. 1967. Environment of deposition of stratiform (Mississippi Valley type) ores deposits, from studies of fluid inclusions. Economic Geology Monograph, 3, 349-362. 1972. Composition of fluid inclusions. U.S. Geological Survey Professional Paper 440-JJ. 1984. Fluid Inclusions. Mineralogical Society America, Reviews in Mineralogy, 12. 1990. Fluid inclusion analysis--prologue and epilogue. Geochimica et Cosmochimica Acta, 54, 495-507. ROGERS, P. 1977. Fluid inclusion studies on fluorite from the Derbyshire orefield. Transactions of the Institution of Mining and Metallurgy, 76, 128-132. 1978. Fluid inclusion studies on fluorite from the Askrigg block. Transactions of the Institution of Mining and Metallurgy, 87, 125-131. RUSSELL, M.J. & SKAULI, H. 1991. A history of theoretical developments in carbonate-hosted base metal deposits and a new tri-level enthalpy classification. Economic Geology Monograph, 8, 96-116. SAMS, M.S. & THOMAS-BEVTS, A. 1988. Models of convective fluid flow and mineralization in southwest England. Journal of the Geological Society, London, 145, 809-817. SAWK1NS,F.J. 1966a. Ore genesis in the north Pennine orefield in the light of fluid inclusion studies. Economic Geology, 61, 385-391. 1966b. Preliminary fluid inclusion studies of the minerlization associated with the Hercynian granites of SW England. Transactions of the Institution of Mining and Metallurgy, 75, 109-112. SHEPHERD, T.J., MILLER, M.F., SCRIVENER, R.C. & DARBYSHIRE,D.P.F. 1985. Hydrothermal fluid evolution in relation to mineralization in southwest England with special reference to the Dartmoor-Bodmin area. In: HALLS, C. (ed.) High heat production (HHP) granites, hydrothermal circulation and ore genesis. Institution of Mining Metallurgy, London, Special Publication, 345-364. SMALL, A.T. 1978. Zonation of P b - Z n - C u - F - B a mineralization in part of the Yorkshire Pennines. Transactions of the Institution of Mining and Metallurgy, 87, 9-14. SMITH, F.G. 1978. Historical development of inclusion thermometry. Univ. Toronto Press, Canada. SMITH, F.W. 1973. Fluid inclusion studies on fluorite from the North Wales orefield. Transactions of the Institution of Mining and Metallurgy, 82, 174-176. -1974. Factors governing the development of fluorspar orebodies in the North Pennine orefield. Ph.D. Thesis, University of Durham. • PHILLIPS,R. 1974. Temperature gradients and ore deposition in the north Pennine orefield. Fortschrifie Mineralogie, 52, 491-494. SOLOMON, M., RAFTER, T.A. & DUNHAM, K.C. 1971. Sulphur, and oxygen isotope studies in the Northern Pennines in relation to ore genesis. Transactions of the Institution of Mining and Metallurgy, 80, 259-275. SORBY, H.C. 1858. On the microscopical structure of crystals, indicating the origin of minerals and rocks. Quarterly Journal of the Geological Society of London, 14, 453-500. SPURR, J.E. 1923. The Ore Magmas. McGraw-Hill, New York. TAMMEMAGI, H.Y. & SMITH, N.L. 1975. A radiogeological study of the granites of SW England. Journal of the Geological Society of London, 131, 415-427. WILKINSON, J.J. 1990a. The origin and evolution of Hercynian crustal fluids', South Cornwall, England. PhD thesis, University of Southampton. - - 1 9 9 0 b . The role of metamorphic fluids in the development of the Cornubian orefield: fluid inclusion evidence from south Cornwall. Mineralogical Magazine, 54, 219-230. --, RANKIN, A.H., MULSHAW, S.C., NOLAN, J. & RAMSAY, M. 1994. ICP-linked laser ablation for the determination of metals in fluid inclusions: an application to the study of magmatic ore fluids, S.W. England and New Mexico. Geochimica et Cosmochimica Acta, in press. WILL1S-RICHARDS, J. & JACKSON, N.J. 1989. Evolution of the Cornubian orefield Southwest England: Part I. Batholith modelling and ore distribution. Economic Geology, 84, 1078-1100. YERMAKOV, N. P. 1965. Research on the nature of mineral-forming solutions with special reference to data from fluid inclusions. International Monographs in Earth Sciences, 22. Pergammon Press, New York. -

-

-

-

-

-

From QJGS, | 4, 453.

On the MICROSCOPICAL STRUCTURE 0 f CRYSTALS, indicating the ORIGtN of MINERALS and ROCKS. B y H . C. SORSY, E s q . , F.R.S., F.G.S., Corresponding Member of the Lyceum of Natural H i s t o r y o f N e w Y o r k , a n d o f t h e A c a d e m y o f N a t u r a l Sciences o f P h i l a d e l p h i a , &c. [Read December 2, 1857.]

[PLAT~-S XVI.-XIX.] CONTENTS. § 2. Water contained in Crystals. § 3. Minerals contained in Secondary Rocks. a. Rock-salt,Calcite,&c. b. Quartz-~'eins. § 4. Metamorphic Rocks. § 5. Minerals and Rocks formed by cooling from a stateof igneous fusion. § 6. Minerals and Rocks formed by the combined operation of Water and Igneous Fusion. a. Minerals in the blocks ejected from Vesuvius. b. Granitic Rocks. c. Temperature and Pressure under which GraniticRocks have been formed. Description of the Plates.

History of the Subject. I. Structure of Artificial Crystals. § l. Crystals formed from Solution in Water. a, Mode of Preparation and Examination ; general and special characters. b. Number, size, form, and arrangement of Cavities. e. Expansion of Fluids by Heat. d. Effects of Pressure. e. The Elastic Force of the Vapour of Water. § 2. Crystals formed by Sublimation. § 3. Crystals formed by Fusion. § 4. General Conclusions.

lI. Structure of Natural Crystals. § 1. Methods employed in examining Minerals and Rocks.

IN this p a p e r I shall a t t e m p t to p r o v e t h a t artificial a n d n a t u r a l crystalline s u b s t a n c e s possess sufficiently c h a r a c t e r i s t i c s t r u c t u r e s to p o i n t o u t w h e t h e r t h e y were d e p o s i t e d f r o m solution in w a t e r or crystallized f r o m a m a s s in t h e state o f igneous fusion ; a n d also t h a t in some cases an a p p r o x i m a t i o n m a y h e m a d e to t h e r a t e at, a n d t h e t e m p e r a t u r e a n d p r e s s u r e u n d e r w h i c h t h e y were f o r m e d .

From QJGS,?0, 689.

THE

GENESIS

OF THE NORTH ORE DEPOSITS I

PENNINE

BY KINGSLEY CHARLES DUNHAM, P H . D . B . S C . Read

February

[PLATES X X I V

7th,

F.G.S.

1934

& XXV.]

CONTENTS

I. II. III. IV.

V. VI. VII.

VIII.

Introduction ................................................ H i s t o r y of i n v e s t i g a t i o n . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Scope of t h e present s t u d y . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Deposits o f t h e P e n n i n e t y p e . . . . . . . . . . . . . . . . . . . . . . . . . . . (a) F o r m o f t h e deposits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . (b) L a t e r a l d i s t r i b u t i o n o f minerals ... (i) F l u o r s p a r a n d b a r i u m minerals ......... (ii) Chalcopyrite, sphalerite, galena ......... (iii) P y r i t e a n d m a r c a s i t e . . . . . . . . . . . . . . . . . . . . . (iv) R a r e cobalt a n d nickel minerals ...... (v) Q u a r t z a n d c h a l c e d o n y . . . . . . . . . . . . . . . . . . (vi) A r a g o n i t e . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . (vii) Dolomite, siderite, calcite ............... (c) Vertical distributior~ o f minerals ............... (i) Sulphides . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . (ii) B a r y t e s a n d fluorspar . . . . . . . . . . . . . . . . . . . . . (d) H y p o t h e s i s o f zonal d i s t r i b u t i o n ............ (o) T e x t u r a l relations o f t h e minerals ............ (i) B a n d e d veins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . (ii) G r a n u l a r veins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . (iii) U n i t y o f t h e mineralization ............... (iv) Paragenesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Deposits of t h e Great S u l p h u r Vein t y p e ............ (a) F o r m a n d m i n e r a l o g y . . . . . . . . . . . . . . . . . . . . . . . . . . . (b) Age-relations w i t h t h e P e n n i n e t y p e ......... Ago of t h e mineralization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ore genesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . (a) L a t e r a l secretion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . (b) F r o m t h e W h i n Sill . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . (c) F r o m a s u b - P e n n i n e m a g m a .................. List of Works to w h i c h reference is m a d e ............

Page 689 690 692 693 693 694 694 695 699 699 699 699 700 700 700 702 703 706 706 708 708 708 710 710 711 712 713 714 714 715 716

From Le Bas, M. J. (ed.), 1995, Milestones in Geology, Geological Society, London, Memoir No. 16, 249-263 First published in Journal of the Geological Society, London, Vol. 150, 1993, pp. 637-651

Carbonate magmas D.K.

BAILEY

Department o f Geology, University of Bristol, Wills Memorial Building, Queens Road, Bristol, BS8 1RJ, UK Abstract: For 40 years, the case for the existence of carbonate magmas rested on field observations of carbonatite intrusions, in which the lack of thermal effects raised an apparent conflict with the high melting temperatures of pure carbonates. Since 1960, the position has changed, with the growth of experimental studies and increasing observations of effusive carbonate rocks. A nephelinite/phonolite volcano in Tanzania is currently erupting Na-Ca-K carbonate magma (around 600 °C). This is unlike all other intrusive and effusive carbonatites (350 examples worldwide) which are dominantly composed of Ca, Mg, Fe carbonates, and have negligible alkali contents. Although a number of effusive calcio-carbonatites are considered to be degraded alkali carbonatites, there are several examples (including one magnesi0-carbonatite) which are close to their erupted composition, and substantiate the existence of high T carbonate magmas lacking essential alkalis at the time of eruption. In these associations silicate magmas are absent (or minor), and in most the effusive carbonatites have been erupted directly from the mantle (with entrained peridotite debris and minerals). They provide a link with the ultramafic association (peridotite and pyroxenite), seen in some carbonatite intrusions, with the commonly associated ultramafic lamprophyres (which may also carry mantle xenoliths), and with carbonate-rich kimberlites. Many carbonatite intrusions also have little or no associated silicate magmas, putting in question a popular view that carbonatites normally form only minor parts of alkaline igneous complexes of nephelinite/phonolite type. The corollary, that the carbonatites are normally differentiates is even less sound, because in alkaline complexes the carbonatite is always last in the eruption sequence. Here the carbonatite may represent the final residua expelled from the source region. Most large carbonatite intrusions seem to have been emplaced at lower T than effusives, probably as a near-solidus mush, with the interstitial fluid metasomatizing the country rocks. A wider perspective of carbonate magma genesis is called for, to encompass various kinds of differentiation from alkaline silicate magmas, and primary carbonate magmas from various depths in the mantle (with or without silicate melts). The strongly bimodal composition distribution of calcic and dolomitic carbonatites is a further factor awaiting explanation. Half of the known carbonatites are in Africa, and their timing and distribution indicate that the activity is a response to lateral forces acting across the plate. Carbonate magmatism is waiting to be unleashed. This activity demands attention because it is now clear that carbonate magmatism is a crucial surface expression of deep mantle processes.

able to list 32; by 1966, when two books were published (Tuttle & Gittins 1966; Heinrich 1966), the total had risen to nearly 200, and at the last published count (Woolley 1989) there were around 330 (about half of which are in Africa). In choosing 11 examples from the 23 known in Africa at that time Campbell Smith (1956) could essay an all-embracing compass, summing up the salient facts, introducing the rival hypotheses, and presenting a fascinating snapshot of the state of the 'art'. A similar review now, of the same length, could do scant justice to the subject, because quite apart from the sheer increase in research, and fundamental new discoveries, the whole context has changed. Of the seven special features perceived by Campbell Smith as 'subsidiary problems' (pp. 212-213), five (magmatic associations, multiphase activity, multiple intrusion, explosive activity, and range of vent diameters) are now widely recognized as general, although not necessarily invariable, features of carbonatites. Minerals peculiar to carbonatites have received much attention since 1956 (e.g. Hogarth 1989). The concommitant high levels of mobile and silicate incompatible elements, and the extensive metasomatism

A turning point in carbonatite studies was marked in November 1956, with the simultaneous publication of reviews by two major geological societies (Campbell Smith 1956; Pecora 1956). From the time of the original use of the term 'Karbonatite' (to describe the carbonate rocks of Fen, S. Norway) a controversy had smouldered over Br6gger's (1921) proposal that these were products of carbonate magmas. By 1956 the n e e d for a review in the wider geological forum, of these hitherto rare rocks, had become evident in the rapidly increasing discoveries of carbonatites in the stable parts of the continents. Campbell Smith's Presidential Address to the Geological Society of L o n d o n was both apt and timely, and appropriately titled ' A review of some problems of African carbonatites'. Most of the new discoveries following the second world war were in Africa, and Campbell Smith had played a crucial part in the first positive recognition of carbonatites in that continent (Dixey et al. 1937). Both reviews, of course, have many parallels, and a number of interesting differences. Each highlighted the growing pace of new carbonatite discoveries, and predicted that this would not slacken. In 1956, Pecora was 249

250

D.K.

(fenitization) are intrinsic features, which must always have a bearing in any consideration of carbonatite magmatism. Fenitization has continued as an important strand in carbonatite studies, but within the larger context of carbonate magmatism can only be considered in terms of alkali mobility. A summary of fenitization, with leading references is given by McKie (1989). One of the 'two main problems' for Campbell Smith was carbonatite 'mode of emplacement', and the 'ultimate problem' (p.216) was the source of the CO2. These constituted the main questions for Pecora (1956) also, and in one shape or form they have always been, and are still the nub of the argument concerning igneous carbonates. But, as mentioned above, the context of the argument has changed radically since 1956. Not only can we now draw on extensive evidence of effusive carbonatite activity, and experiments on carbonate melting and crystallization, but entirely new concepts of melt generation colour petrological thought. When Campbell Smith (and Pecora 1956) allude to source regions for carbonatite, they refer to the deep crust, or the deeper parts of the Earth--neither uses the word 'mantle'. The dawn of our present intellectual milieu was still to break, with the recognition that the primary characteristics of most magmas are determined by partial melting in the Earth's mantle (Yoder & Tilley 1962). Regardless of whether a carbonatite is primary, or a differentiate of an associated silicate melt, considerations of carbonatite magmatism are now constrained by knowledge of mantle compositions and physical states, together with experimental evidence on mantle melting conditions. The 'two main problems' of Campbell Smith must, of course, be re-examined with this perspective, and the intention in the present paper is to do this firstly, through present knowledge of magmatic rocks and experimental melts, and secondly, through what may be termed the ultramafic connections. In his review, Campbell Smith selected his examples of African carbonatites in a series of increasing age, to illustrate different erosion levels. As a result he was able to point out (p.216) that t h e r e is a general pattern in carbonatite activity, produced repeatedly. He goes on to cite Dixey that 'carbonatites are a normal part of the magmatic history . . . of Africa over a long period of geological time'. This aspect of carbonatite activity has potentially profound implications in petrogenesis, and will be explored after the main discussion of carbonate magmas.

Carbonate magmas Underlying Campbell Smith's examination of the 'Problem of mode of emplacement' was the question of whether or not carbonatite was truly igneous or magmatic, i.e. what was the physico-chemical regime at the time of carbonatite formation. Large, central carbonatite masses have given rise to the largest range of proposals, and his perceptive discussion looks at all the then suggested mechanisms, from replacement, hydrothermal deposition, plastic flow (akin to salt domes), crystal mush, to purely magmatic intrusion. For Campbell Smith the crucial evidence lay in the dykes and cone sheets, which seemed to require a 'carbonatitic magmatic liquid' as envisaged by von Eckermann (1948).

BAILEY

Magmatic evidence from intrusive carbonatite petrography Many of the features touched on by Campbell Smith have been substantiated in later studies, especially the importance of fluidization for fragmental intrusions. Evidence of replacement is widespread, but no longer seen as necessarily an obstacle to a magmatic origin, because it mostly applies to replacement of carbonate, e.g. dolomitization, or sub-solidus recrystallization of a carbonatite protolith. Replacement has been specifically addressed by Barker (1989, 1993) who points out that many (if not most) plutonic silicate rocks have undergone sub-solidus reequilibration such that any original igneous texture may be largely a palimsest: he argues that such a process is even more likely to affect easily recrystallized carbonate rocks. Equally, the formation of some carbonatite bodies by hydrothermal/carbothermal deposition (e.g. Mountain Pass, California) does not vitiate magmatic evidence in other complexes: a continuum from magma to fluids may be expected in the natural environment. Field relations are all important, and have been fully reviewed by Barker (1989). But there must be a distinction between convincing magmatic field relations (which are likely to survive sub-solidus changes) and finding critical evidence of the nature of the original magma. The igneous characteristics of the regime may still be discernible, but the composition and original phase relations may be lost, or at best obscured. Flow structures, for instance, may be interpreted as the fabric from an intrusive magma, but replacement may have destroyed the original composition and texture. The same caveat may apply even to porphyritic texture, although here the original phenocrysts may survive to provide valuable information. More serious, however, is the fact that most intrusive carbonatites also show signs of chemical exchange with their wall rocks, as seen most dramatically in alkali metasomatism (fenitization). Even if this exchange could be accurately quantified, the facts of exchange and replacement, reveal a great residual uncertainty about how much of the material that was emplaced still remains. Still more perplexing is the unknown quantity of material that passed through, and out of, the conduit prior to and after it was filled. Any intrusion and its wall rocks are but remnants of volatile-rich systems and the cautionary reminder that igneous rocks are dead magmas must surely be imperative for intrusive carbonatite. Hence, the closest approach to original magma must be sought among the rapidly quenched effusives. These can provide better understanding of magmas and processes, and guide the interpretation of existing laboratory studies and the design of new experiments. In his gentle advocacy of carbonatite magmas, Campbell Smith differed from Pecora (1956) who leaned more towards a spectrum of dense fluids, a difference surely attributable to Campbell Smith's extensive personal contact with studies of the sub-volcanic complexes in East and Central Africa. His selected African examples are listed in order of increasing depth of erosion, and at the head is Kerimasi volcano, Tanzania, where James (1956) had just reported carbonatite on the flanks and in the crater. Effusive carbonatites could resolve many of the questions about the nature of carbonatite intrusions but neither Campbell Smith, nor Pecora (who also highlighted

CARBONATE MAGMAS the volcanic connections) felt able to speculate on what form the activity might take. In this they were undoubtedly restrained by the limited evidence on experimental carbonate-melt relations, which was then irreconcilable with geological observations (and where Pecora rightly perceived new experiments would be pivotal). By one of the ironies of science, the twin volcano of Kerimasi, Oldoinyo Lengai, had been recorded as erupting carbonate ashes several times earlier this century (Hobley 1918; Richard 1942; Guest, 1956) with the added irony that the 1917 eruption receives special mention by that great advocate of limestone syntexis, S.J. Shand in his text book, Eruptive Rocks (1927, p. 36). In 1956 the significance was missed, or had been lost from sight, and some of the essentials needed to carry forward the discussion of effusive activity (and experimental melting) would have to wait until 1960. That year saw the eruption of alkali carbonatite lavas in the crater of Oldoinyo Lengai, (Dawson 1962), the laboratory production of Ca carbonate melts at low T and P (Wyllie & Tuttle 1960), and the description of effusive dolomitic carbonatites of Cretaceous age in Zambia (Bailey 1960). There can be little question that widespread doubts about the existence (or even possibility) of carbonate melts were largely dispelled by the experimental results, closely followed by reports of the Lengai lavas, although new questions were raised by both. The experimental melts were compositionally close to calcio-carbonatite, but strongly hydrous, while the Lengai lavas were anhydrous, but quite unlike other carbonatites in bulk composition. New lines of research were opening, to be supported by growing developments in analytical methods, especially for trace elements and isotopes.

251

A l k a l i carbonate m a g m a s Oldoinyo Lengai, in northern Tanzania, is the only volcano at which flowing carbonate magma has been observed (Dawson 1962). By weight the rock is nearly 60% sodium carbonate, 30% calcium carbonate, and 10% potassium carbonate, a representative analysis being given in Table 1. Any losses, to wall rocks prior to eruption, and in gas and sublimates during eruption, are at present unknown but, regardless of what these may have been, the composition is unique. All other carbonatites are essentially Ca, Mg, Fe carbonates, with very low alkali contents (Table 1). Wall rock contamination at Lengai must be minimal because the silica content of the natrocarbonatite is extremely low. The marked difference between natrocarbonatite and all others gave rise to new concepts, including the proposal that loss of alkalis to wall rocks at depth would yield the non-alkali types (Dawson 1964), and carbonate-silicate liquid immiscibility, arising from experiments in alkali carbonate-silicate systems (Koster van Groos & Wyllie 1966). New controversies were born, most of which have at their root the question of whether natrocarbonatite is parental to virtually all other types. The question is natural and needs examination, but the debates have tended to polarize, choosing to ignore the alternative that natrocarbonatite may be simply one type among many. Fortunately, a new multi-author volume with Lengai as its central theme, is in press (Bell & Keller 1994), so that here it is appropriate to focus on the characteristics of the magma and its possible phase relations, so that it may be considered within the framework of other magmatic evidence. Petrographically, the lava is porphyritic with phenocrysts

Table 1. Compositions of effusive carbonatites, as examples of carbonate magmas SiO 2 TiO 2 AI20 3 Fe20 3 FeO MnO MgO CaO Na20 K20 PzOs H20 1. Natrocarb. (Lengai) 2. Calcio-c (Kerimasi) 3. Calcio-c (Ft. Portal) 4. Calcio-c (Polino) 5. Calcio-c (Kaiserstuhl) 6. Calcio-c (Emirates) 7. Magnesio-c (Rufunsa)

0.11

0.09

0.28

0.04

0.53

13.9

32.2

8.27

0.90

0.39

tr.

0.07

0.34

0.22

0.21

54.0

0.10

0.05

1.82

41.8

13.0

1.74

3.03

7.93

0.40

8.55

36.0

0.73

0.20

3.32

3.45 14.8

16.2

0.52

3 . 9 1 3.69

1 . 3 1 0.07

7.31

38.7

0.05

0.50

0.60

3.12 24.1

0.45

0.03

0.15

0.41

0.36

52.6

0.04

0.50

1.56

1.16 39.8

7.45

1 . 0 1 1 . 7 5 8.30

3.12

0.46

3.27

40.5

0.23

0.14

7.00

1.2

0.49

1.56

19.0

28.8

4.44

0.98

tr.

CO 2 34.7

F

CI

SO 3

SrO

2.93

4.21

2.18

1.53 1.04 0.40

0.2

0.08

0.63

0.15

0.54

0.67

0.15

24.58

References 1. Dawson 1989, table 11.3, p. 269 (Anal. 1). 2. Mariano & Roeder 1983, table 1, p. 451 (Anal. 3). 3. Barker & Nixon 1989, table 5, p. 172 (Anal. 1) F value calculated from mode and mineral analyses. 4. Lupini & Stoppa in press. 5. Keller 1989, table 4.1, p. 79 (Anal. KB2). 6. WooUey et al. 1991, table 3, p. 1160 (Anal. Type 1, Mean). 7.'Bailey 1989, table 1, p. 416 (Anal. 3: dolomite melt droplet). c, carbonatite.

BaO

1.10

252

D. K. B A I L E Y

of nyerereite (NaCa carbonate) and gregoryite (Na, K, Ca carbonate) in a matrix of the same minerals. Details of the compositions are given in Gittins & McKie (1980). Their diagram showing the natural phase compositions in the synthetic system, Na2CO3-K2CO3-CaCO3 at lkbar, is a helpful presentation, followed here in Fig. 1. Estimates of eruption temperatures suggest a small range, with the 1960 flows being reported as not incandescent at night, while incandescence was observed in 1988, and Dawson (1989) records that temperatures were consistently between 560 and 580°C, which must be close to the lower limit of incandescence. These temperatures are similar to a value around 600 °C for the gregoryite-nyerereite cotectic at 1 atmosphere pressure, estimated by Cooper et al. (1975) on the basis of experiments on the natural lava. This temperature is over 100 °C lower than those in the synthetic system (Fig. 1), presumably due to the high levels of fluxing components (especially halogens) in the natural lava. Flow morphology has varied from highly mobile, vesicular flows that reached the crater walls (200-250 m), to very viscous, short, and slow moving extrusions (Dawson 1989). All these observations are consistent with a melt near a cotectic. Reference to Fig. 1 shows that the lava compositions plot close to, and sub-parallel with, the 1 atmosphere cotectic of Niggli (1919). Especially significant is the position of the lava groundmass in the midst of the bulk compositions and close to the tie line between its phenocryst compositions. These features are precisely those to be expected from a cotectic melt, and suggest that the natural cotectic is recording a low pressure of equilibration, around 150-200 bar, for the natrocarbonatite, i.e. within 500m of the surface. It is noteworthy that Cooper et al. (1975), when reporting their experiments on a natural sample, concluded that the bulk composition is 'probably close to the cotectic'. When considering the status of the natrocarbonatite, it is useful to recall that the lava eruptions (within the crater) are a tiny fraction of the 2000 m volcano composed largely of fragmental nephelinite and phonolite (c. 60 km3: Dawson, 1989). The unusual composition has given rise to the widest diversity of hypotheses about its origin. (1) Immiscible separation of carbonate melt from a parental silicate melt in the range nephelinitephonolite. Le Bas (1989) has reviewed the petrological case, and Kjarsgaard & Hamilton (1989) the relevant experimental evidence. A corollary is that if the initial carbonate melt is more calcic, fractionation of calcite can yield more sodic residual melts.

(2) Prolonged fractionation of an original primary calciocarbonatite under anhydrous conditions (Gittins 1989). (3) Melting of troniferous sediments (akin to those around the neighbouring Lake Natron) in the volcanic substructure (Milton 1968, 1989; Peterson & Marsh 1986). With the exception of 3, all hypotheses place natrocarbonatite as the product of extended differentiation: fractionation in the silicate line culminating in phonolite, and then natrocarbonatite separation; nephelinite/ carbonatite immiscibility, followed by carbonatite fractionation; or extended fractionation in primary calciocarbonatite. The fact that the erupted melt at Oldoinyo Lengai corresponds to a low pressure cotectic is not taken into account, perhaps largely because Niggli's (1919) results have been ignored. The simplest interpretation of the phase relations would be that the lavas are minimum melts from bulk compositions close to the gregoryite-nyerereite join. All other propositions require qualifying assumptions about the P T X history of any supposed earlier melt. Fractional crystallization of calcio-carbonatite would require some yet to be defined step to effect the transposition into the natrocarbonatite system (see Fig. 2), and even then eruption of the minimum melt composition alone would still need explanation. Equally unexplained at present is how, or why, a melt with cotectic characteristics should form by immiscible separation from a high T silicate melt. Of course our present view of the natural magmatism may be blinkered; on a geological timescale we have only a single instantaneous 'freeze frame' of the activity, but this stricture applies equally to all hypotheses advocating natrocarbonatite as a general model for carbonatite petrogenesis. Purely on the basis of the phase relations, the melting hypothesis is more plausible but as many have pointed out, e.g. Cooper et al. (1975), the overall chemistry, and especially the stable isotope chemistry, is not compatible with sediment melting. A powerful consensus in favour of this view may be presumed from the absence of any mention of the hypothesis in the latest multi-author book on carbonatites (Bell 1989). There is, however, another possibility that is free from the difficulties raised by the melting of sediment, namely that the source might be older carbonates in the volcanic pile. These may form initially as high T sublimates from continued CO2-rich exhalation from silicate magmas. Subsequent ascent of magma within the volcano could mobilize these, driving the resultant low viscosity melts ahead through the axial zone. These alkali carbonate melts would not invariably be erupted as flows: in

K2CO3

;9

!

.'/

,. Na2CO3

C NY

CC

CaCO3

Fig. 1. Adapted from Gittins & McKie (1980), showing the positions of analyses for natrocarbonatite (open circles); nyerereite and gregoryite phenocrysts (triangles); and groundmass (solid square); projected onto the phase diagram for N a 2 C O 3 - K 2 C O 3 - C a C O 3 at I kbar. The broken curve C-C is the cotectic at 1 atms. calculated from Niggli (1919).

CARBONATE MAGMAS 1400

i

-

I

i

1400

i

P= 1 K b a r

/

7

/ /

1200

1200

/

L

/

oo

/

/ /

< rr

1000 NCss+L

u.l b.-

T

4¸ NY+L

1000

CC+L C

800

/

I

//

600 /

1 60

I NY

800

--

600

CC+NY

NC+NY I 80

--

I 40

Na2CO 3

I 20 CaCO 3

WT, PERCENT

Fig. 2. Schematic binary diagram to illustrate the constraints of temperatures, phase relations, and compositions on the natrocarbonatite erupted from Oldoinyo Lengai (corresponding to a melt with cotectic characteristics) adapted from Cooper et al. 1975. In the binary the natural cotectic is represented by the eutectic O. A melt near O cannot produce calcite nor a calcitic residuum, even by a hypothetical loss of alkalis, which could proceed only in the direction of NY (nyerereite) and no further. Any carbonatite with calcite phenocrysts could not have evolved from natrocarbonatite as currently erupted. The most alkaline liquid possible with calcite phenocrysts would be E, and the bulk compositions of calcite-phyric 'magmas' must lie between E and C, i.e. much less alkaline and higher in T than natrocarbonatite O. A specimen containing calcite (Dawson et al. 1987) may fall in this category, but the mantling of calcite by nyerereite suggests that it may be xenocrystic. The diagram shows that where calcite phenocrysts are abundant, e.g. bulk composition X (50% phenocrysts) the liquidus T must be very high, and the alkali content correspondingly small. Such melts would indicate markedly different conditions from those necessary to produce the natrocarbonatite magma currently being erupted from Lengai. Preliminary reports on the effects of F (Gittins 1989; Jago & Gittins 1991; Gittins & Jago 1991) show general lowering of the calcite liquidus by as much as 200 °C (at 8%F). In one of these (Jago& Gittins 1991) it is stated that CaF2 or F 'break the thermal barrier caused by nyerereite'. Critical phase relations are given only in outline, in which there are disparities (between diagrams and in the data) and informed discussion must therefore await publication of the defining experimental data (e.g. run compositions, stable equilibrium phase assemblages). In any case, whether or not the barrier is breached by equilibrium crystallization, calcite would have to be perfectly fractionated (even approaching the solidus) by which point the liquid compositions are already much richer in F than their postulated derivatives, natrocarbonatite lavas. See text and Tables 1 and 3, for F levels and temperature effects in carbonatites. a complex series of activity cycles they may accumulate in and choke the upper part of the conduit, to be intermittently erupted as ashes, which during declining activity fall back and fill the crater. The modern lava activity may represent only a quiet stage when melts are able to ooze out onto the crater floor. This mode of generation would be in keeping with the phase relations, the highly constricted and very small volume activity, the totally anhydrous composition, the U-Th disequilibrium results (Williams et al. 1986), and the unique trace element characteristics (Keller 1992).

253

Because the natrocarbonatite lavas at Lengai are the only flowing carbonate magmas observed, this composition has naturally featured strongly in attempts to account for the formation of non-alkaline carbonatites, most of which are calcitic. Links with natrocarbonatite have been sought for calcitic carbonatites in two contrasting modes: (a) for calcitic lavas and pyroclasts, by calcification of alkali carbonatites through the action of surface waters, and (b) for calcitic intrusions by loss of alkalis from alkali carbonatite, metasomatizing the wall rocks (fenitization). Calcification of original alkali carbonatite volcanics enjoys considerable support, and depends entirely on the recognition of nyerereite pseudomorphed by calcite. Gittins (1989) has provided a review of the position in terms of the phase relations, urging restraint against the over-enthusiastic adoption of the surface replacement processes. He also shows that where (as in many examples) monocrystalline calcite phenocrysts are present, the original alkali contents (if any) need not have been great. A simplified example is illustrated in Fig. 2, where it may be noted also that temperatures on the calcite liquidus are higher than those for natrocarbonatite, meaning that the observed magma is not a possible parent for any carbonatites with calcite phenocrysts. Obviously, alkali melts with a calcite liquidus may exist, but the melt temperatures will be considerably higher than natrocarbonatite, and the alkali content much lower. Actual alkali carbonate melts with calcite phenocrysts have yet to be observed in nature, and their existence still awaits confirmation. No intrusive carbonatites with pseudo-nyerereite have been reported, so that any inference about their earlier alkali content must at present be based on alkali metasomatism of the wall rocks. As noted before, the rocks are but remnants of a volatile-rich, open system, where the total through-put of material is unknown. Alkali activity is evident, but the alkali content of the original magma must be conjectural: the only certainty is that natrocarbonatite could not have been the original melt, because it is a low energy, cotectic composition from which loss of alkalis (if it were possible) could only transpose the material into sub-solidus field of nyerereite (see Fig. 2). If the assumption is that calcitic carbonatite is a residuum after alkali loss, then it may be presumed that the original melt had calcite on the liquidus. From Fig. 2 it may be seen that this would set a maximum alkali carbonate content on the bulk composition around 45%, ranging down to virtually zero. An initial melt with greater than 45% alkali carbonate would be effectively super liquidus (superheated), which entails an unconstrained presumption about the nature of the system: even this melt, however, must instantaneously lose its superheated status once it is in reaction with wall rocks. Hence the general case is represented by an intrusive magma with calcite on the liquidus. The best test of the likely alkali content lies in freshly quenched (unaltered) calcitic effusive rocks. Effusive calcio-carbonatite

Surprisingly perhaps, records of calcitic volcanic rocks slightly predate that of the landmark eruption of natrocarbonatite lava (Dawson 1962). James reported Recent calcitic carbonatite layers on the flanks of Kerimasi volcano (1956); ripple-marked calcitic tufts were reported from the Cretaceous carbonatitic volcanoes of Zambia

254

D. K. B A I L E Y

(Bailey 1960, appendix V, plate V); and by 1961 the Recent carbonatite lava flow at Fort Portal had been described (von Knorring & Du Bois 1961). Dawson (1962) also pointed out that calcitic ashes were widely distributed in East Africa, making the valid suggestion that many of these might be degraded natrocarbonatite. This possibility, however, inevitably raised a question about the original composition of any effusive calcio-carbonatite. In those cases where such rocks contain calcite phenocrysts, the erupted bulk composition cannot have been the same as the Oldoinyo Lengai lavas, as indicated above, and in Fig. 2. The whole question of effusive calcio-carbonatites has fortunately been given new perspectives in reviews by Barker (1989) and Keller (1989). Barker, concentrating on field relations combined with petrography, is able to conclude (1989, p.44) that 'srvite liquids can erupt as lavas of low viscosity'. Keller presents detailed evidence on the calcite lapilli tufts of the Kaiserstuhl, showing that a replacement origin is ruled out by the magmatic chemistry combined with 'open inter-clast frameworks and open vesicles in lapilli' (1989, p.85). Some localities where calcitic effusive rocks are closest to magmatic compositions are listed in Table 2. Many other examples of calcitic volcanics are known, especially in East Africa, but are not listed here because they have been designated as calcitized alkali carbonatites (e.g. Clarke & Roberts 1986), but in the light of the Kaiserstuhl experience (Keller 1989) it would seem prudent to keep an open verdict. In terms of texture and composition the air-fall tear drop lapilli of the Kaiserstuhl (Keller 1981) are apparently determined by surface tension at a liquid-gas interface, the droplets having the composition of almost pure calcium carbonate (Table 1). As there seems little scope for major

losses or exchange after eruption, the present composition must be close to that of the original erupted melt. In the absence of a flux (e.g. H20) such a melt would be at a high temperature and above its 1 atmosphere dissociation point, leading to the conclusion that melt fragmentation occurred under pressure in the volcanic vent, and that droplet shaping was effected while the melt was supercooled in flight. Calcite microphenocrysts indicate that the melt was on the liquidus immediately prior to disintegration; the lack of strong concentricity of phenocrysts parallel to the droplet boundaries would be consistent with very rapid undercooling and quenching. Unless substantial amounts of a component (or components) have gone from the Kaiserstuhl lapilli, leaving no trace, high melt temperatures must apply. Calcite crystals in calcitic liquid would require a minimum temperature of 1230 °C at a minimum pressure of 40 bar (Point Q, fig. 20.1, Wyllie 1989). Pressure presents no problem, the melt could rise to 150 m below the surface; indeed, melt ascent to a critically shallow depth might be a plausible fragmentation/eruption trigger. Lowering of the melt temperature by the small amounts of non-calcitic components still present would not be dramatic, suggesting values similar to mafic igneous melts (1000-1200 °C): such values would be consistent with the deduction (Keller 1989) of a genetic link with melilite nephelinite. Similar conclusions emerge from consideration of other associations. At Kerimasi there is a close association of lapilli with melilite phenocrysts and those with calcite phenocrysts, and there are melilitite lapilli tufts cemented by calcium carbonate. In the Fort Portal lava flow, olivine, diopside and phlogopite xenocrysts, are enclosed in a groundmass of calcite, periclase, perovskite, apatite and spurrite, consistent

Table 2. Effusive calcio- and magnesio-carbonatites where the observed compositions are closest to erupted magma Ref & Locality

Cognate minerals

Xenocrysts

A Calcio-carbonatites 1. Ft Portal Cc, Cs, P, Ap, Mo O1, Di, Phi, Per, Ap (Uganda) 2. Catanda Carbonates. O1, Cr-Di, Cr(Angola) Mainly Cc Sp, Phi, Kaer, Ap 3. Polino (Italy) Cc, Zr, (Mo) O1, Phi, OI + Phi 4. U.A. Emirates Cc, Cr-Sp 5. Khanneshin Cc, Ank, Ba (Afghanistan) 6. Rufunsa Cc, Cr-Sp San 7. Kaiserstuhl Cc, Mt, Ap (Germany) 8 Kerimasi Cc (Tanzania)

Silicate melts in Complex

Province

None

K u/m

None

Tinguaite (?) K u/m None

None None Lc tephrite (minor) None Melilitite*

None

Melilitite*

Inference

Primary, direct eruption from mantle Primary, direct Primary, direct Primary direct Primary direct Primary, direct Differentiate, high T Differentiate (?) High T

B. Magnesio-carbonatite

6. Rufunsa (Zambia)

Dol, Cr-Sp

Phi

None

None

Primary Direct

* Silicate melt most closely related to carbonatite Abbreviations: Cc, Calcite; Cs, spurrite; P, periclase; Ap, apatite; O1, olivine; Di, diopside; Phi, phlogopite; Per, perovskite; Sp, spinel; Zr, zirconium garnet; Mo, monticellite; San, sanidine; Mt, magnetite; Dol, dolomite; Ank, ankerite; Ba, barite; Kaer, kaersutite; Lc, leucite; K u/m, potassic ultramafic lava. References: 1. Barker & Nixon 1989; 2. Silva 1973; 3. Stoppa & Lupini 1993; 4. Woolley et al. 1991; 5. Alkhazov et al. 1978; 6. Bailey 1990; 7. Keller 1981; 8. Mariano & Roeder 1983.

CARBONATE MAGMAS with high-T eruption. Although there are no associated silicate magmas at Fort Portal, the olivine-diopsidephlogopite constitutes a heteromorph of olivine leucitite, and in the volcanic fields further south (Katwe-Kikorongo) the characteristic eruptives are olivine leucitite and melilitite lapilli tufts cemented by calcium carbonate (Lloyd 1985). Similarly the only silicate magmas recorded from the Khanneshin carbonatite volcano (Afghanistan) are minor leucite tephrites (Alkhazar et al. 1978). Hence, a close connection in time and space seems to exist for effusive calcio-carbonatites and alkaline ultramafic melts in which low silica activity is marked by the appearance of melilite, leucite, kalsilite and perovskite, and there is no evidence to indicate that carbonate and silicate melt temperatures were radically different. Furthermore, the anhydrous character of the silicate rocks and minerals (and the effusive carbonatites), when fresh, points to low activity of H20 in the larger system. Further insights are provided by the Fort Portal lava which is a small single flow ( 1 - 5 m thick, covering 0.3km z) in a much larger field of carbonatite pyroclastic rocks, and clearly constitutes a special form of eruption. Barker & Nixon (1989, p. 167) say that the flow was apparently fed by lava fountains from a fissure marked by a spatter rampart. Either the additional components in the groundmass (especially SIO2) reduced the dissociation pressure of calcite, or the melt was undercooled, or largely crystalline before it gained access to the surface. Olivine and diopside have rims of monticellite, presumably formed by reaction with CaCO3, which in any formulation, in the absence of dolomite, releases CO2: the presence of periclase in the groundmass would also be consistent with degassing prior to, and during eruption. Even though it seems necessary that this material was near its solidus prior to final eruption, the mineral assemblage in effusive calcio-carbonatite is consistent with high temperatures (in contrast with alkali carbonatite) and low activity of SiO2 and H20. At Kerimasi there is coarse grained srvite (plug?) within the crater that may be analogous, consisting of calcite, monticellite and periclase (Mariano & Roeder 1983). In this case the periclase is late as it mantles earlier magnesio-ferrite, and in view of the near surface emplacement may plausibly be attributed to final crystallization below the dissociation point for Mg carbonate. A similar mineral assemblage is reported from Polino in the Umbria-Latium province in Italy, where calcitic vent tuffisite, carrying mantle-derived olivine and phlogopite, has much of the olivine replaced by monticellite: again the volcanic association is one of leucite- and melilite-bearing silicate rocks, (Stoppa & Lupini 1993). At present there is a strong current of support for the proposition that many carbonatites form by low-P liquid unmixing from a carbonated silicate melt (Le Bas 1989; Kjaarsgard & Hamilton 1989), but the fact that the above effusive calcio-carbonatites carry mantle debris (despite very low melt viscosity) denies the general applicability of low-P unmixing, and points to their formation in the mantle. A direct source in the mantle for some calcio-carbonatites is also indicated elsewhere. In the Zambian volcanic rocks, minor amounts of calcio-carbonatite appear in a dominantly magnesio-carbonatite assemblage (Bailey 1960), where the primary nature of the activity is recorded in melt droplets containing high Cr magnesio-chromite (Bailey 1989). Extrusive calcio-carbonatite containing similar chromite has been reported from the United Arab Emirates also

255

indicating a direct origin from the mantle (Woolley et al. 1991).

Liquid immiscibility: general considerations Before discussing magnesio-carbonatites, it is appropriate to examine the case for liquid immiscibility as a means of generating carbonatite magmas, because this hypothesis is currently in favour to explain alkalic and calcic carbonatites. Although the concept applied to carbonatites is of long-standing (e.g. von Eckerman 1961) experimental demonstrations of silicate-carbonate melt immiscibility were needed for a growing general acceptance of the possibilities of the hypothesis (Koster van Groos & Wyllie 1966; Freestone & Hamilton 1980). Recent reviews of the experimental results have been given by Wyllie (1989) and Kjaarsgaard & Hamilton (1989). Le Bas (1989) has explored the possible applications to natural examples, and puts the case for immiscibility as a general explanation for a diversity of N a + Ca carbonatites. An opposite view is propounded by Gittins (1989), who advocates fractionation of primary olivine srvite (calcio-carbonatite) under hydrous and anhydrous conditions as the central petrogenetic process. With so much written on the subject, it is beyond the scope of this review to evaluate the arguments for and against immiscibility versus fractionation. Present evidence does not rule out either, so that both mechanisms may be valid, and, as indicated later, even con-jointly the two are probably only part of the whole panoply of carbonatite genetic processes. More evidence is needed on both fronts. Carbonatite/nephelinite/phonolite volcanism is typically explosive and yet good natural examples of incomplete unmixing are still scarce. The most convincing example, mixed phonolite glass/calcite ash flows (Mt Suswa, Kenya; Macdonald et al. in press) still does not provide carbonatite compositions of general applicability. Appeals may be made to the low viscosity of carbonate melt, which permits perfect melt segregation, but this requires a quiescent 'magma chamber' for which there is no great evidence in sub-volcanic sections. Indeed in sub-volcanic complexes, the carbonatite is widely reported as last in the main intrusion sequence: the carbonatite obviously cannot be derived from the silicate complex it intrudes, so that if a parent magma chamber exists it awaits discovery (or recognition) in deeply exposed sections. Differentiation of any traditional type may not even apply; the carbonatite may be simply the final residua expelled from the source region. None of this rules out immiscibility, but it does mean that the case for its wider applicability must remain open. If there were doubt about the need for such caution it must be removed by the latest information on Spitzkop, S. Africa. This was one of the complexes featured by Campbell Smith (1956) as having the classic pyroxeniteijolite-foyaite-carbonatite assemblage, and nowadays would be an obvious candidate for explanation in terms of immiscibility, but Harmer (1993) argues from the isotope chemistry that the carbonatite cannot be derived from the silicate magmas. Extrusive calcio-carbonatites, such as those at Fort Portal, are crucial in this regard. They lack any associated silicate magmas, they carry dense xenocrysts of possible mantle origin, and dense xenoliths of deep crust and mantle. An explanation by low pressure, liquid unmixing is wholly inappropriate, and such cases (together with Spitzkop) serve

256

D.K.

BAILEY

to show that for each carbonatite the origin must be judged on its merits. Only in the objective approach will the full range of carbonatite origins be uncovered.

Calcio-carbonatite intrusions: a temperature enigma Calcitic (s6vite) intrusive rocks are the most abundant carbonatites, and they pose some of the same problems identified by Campbell Smith (1956) in spite of subsequent growth of knowledge. Most examples have field relations and mineralogies indicative of emplacement temperatures lower than might be expected from the evidence in the extrusive rocks. In 1956, the problem was still in the form that the melting temperature of calcite was too high (1339 °C at 1025 atmospheres CO2: Campbell Smith 1956, p. 190), and although later experiments (Wyllie & Tuttle 1960) showed that H20 profoundly lowered the temperatures of calcitic liquids, the dilemma re-emerges in another form if account be taken of the effusive calcio-carbonatites, with low contents of H20. In alkaline silicate magmas generally, H20 plays a subordinate role to CO2, S, F and CI (Bailey & Hampton, 1990). In carbonate melts, fluorine would have a similar effect to H20, and indeed in anhydrous alkali carbonatite (Anal. 1, Table 1) halogens are very high, such that a lowering of melt temperature by as much as 200 °C may be anticipated (Gittins 1989). But effusive calciocarbonatites are not characterized by high F contents (e.g. Anals 3 & 5, Table 1). For all carbonatites the average levels, and ranges, for different elements have been most recently compiled by Woolley & Kempe (1989), and the levels for the fluxing components are abstracted here in Table 3: their chief modal expression is in early-crystallizing apatite (P,F,C1,OH), pyroxene (Na), amphibole (Na,C1,OH), phlogopite (K,F,OH) and sulphides, and in late-crystallizing barite, fluorite and fluorcarbonates. At the levels in Table 3, these constituents would not dramatically lower magmatic temperatures; some fluxing components must have been lost from intrusions but the amounts are uncertain. Fluorite in late-stage vein deposits, presumably results from build up of F in residual fluids, but in most carbonatites such fluorite would make only a tiny fraction of the total mass. Temperature estimates for intrusive carbonatites show wide variation, with many values reflecting sub-solidus re-equilibration (as might be anticipated in carbonate rocks pervaded by fluids) and it has long been recognized that

calcitic assemblages form a continuum through to hydro/carbothermal deposits, hence Pecora's (1956) leaning towards a spectrum of carbonatitic fluids. Other indicators of low temperatures of final emplacement are the general lack of thermal effects on country rocks and accidental wall rock xenoliths. Textures frequently indicate final emplacement as a fragmental or crystal suspension, and the fact that most intrusions carry dense minerals such as magnetite, in large crystals, and in aggregates, also points to higher viscosities than those of carbonate melts. One possibility that has been widely mooted is that the original intrusion was alkali carbonatite and lost its alkalis in metasomatizing (fenitizing) the country rocks. Alkali metasomatism around carbonatites (McKie 1989) varies in type and extent, and the alkalis in the wall rocks may be the integrated product of sustained passage of carbonatite (and other fluids) through the conduit. Consequently, it is not possible to quantify the amount of alkali loss from the carbonatite that remains in the intrusion. Little if any alkali remains in the intrusions now and the former presence of alkalis in some effusive calcic carbonatites is at present conjectural, based on inferred calcification of previous alkali carbonate. While such a possibility cannot be excluded, the existence of non-alkalic effusive calcio-carbonatites (with levels of other fluxing components similar to those of intrusive carbonatites) requires another explanation for some intrusions. High temperature melts unable to reach the surface may be expected to have a protracted crystallization, during which components not accommodated in the carbonates (and early crystallizing phases) become progressively concentrated in the residual liquid/fluid. Such a residuum would lubricate the crystal mush during final emplacement, and fenitize the wall rocks. In various forms, such an emplacement mechanism has long found favour (Campbell Smith 1956, p.203 [also citing Chayes 1942, p.506, for the interstitial alkalic fluid to explain the carbonate intrusions of Bancroft-Haliburton]). Once again, it would perhaps be unwise to look for one general explanation to cover all cases of carbonatite intrusion. Alkali metasomatism adjacent to carbonatites is variable in type (K or Na), reflecting in part at least the history of the carbonatite prior to emplacement, e.g. a primary mantle source versus a differentiate from carbonate or silicate parents: new observations and techniques will ultimately provide tests for distinctions, or possible inter-relations, between different kinds of intrusion.

Table 3. Carbonatites: averages and ranges of analyses for hyperfusible elements (wt% )

Na20 K20 H20 +

P205 F Ci S SO3

Ferro-carbonatite

Magnesio-carbonatite

Caicio-carbonatite Av.

No.

Range

Av.

No.

Range

Av.

No.

Range

0.29 0.26 0.76 2.10 0.29 0.08 0.41 0.88

102 105 78 119 31 8 23 15

0.0-1.73 0.0-1.47 0.0-4.49 0.0-10.41 0.0-2.66 0.0-0.45 0.02-2.29 0.02-3.87

0.29 0.28 1.20 1.90 0.31 0.07 0.35 1.08

44 44 36 51 21 1 12 13

0.0-2.23 0.0-1.89 0.08-9.61 0.0-11.30 0.03-2.10 -0.03-1.30 0.06-2.86

0.39 0.39 1.25 1.97 0.45 0.02 0.96 1.08

46 51 35 54 20 3 12 14

0.0-1.52 0.0-2.80 0.04-4.52 0.0-11.56 0.02-1.20 0.01-0.04 0.12-5.40 0.06-3.00

(From Woolley & Kempe 1989).

C A R B O N A T E MAGMAS

Effusive magnesio-carbonatite Extrusive magnesio-carbonatite has so far been reported only from the Rufunsa volcanoes in SE Zambia (Bailey 1960). These are of Cretaceous age, and the sub-aerial deposits mantle an old rift valley floor, which was cut into Karoo (Jurassic) and Precambrian basement. Based on their location in a complex intersection of major rifts and the absence of silicate magmas, it was proposed that the carbonatites had a direct origin from the underlying mantle (Bailey 1960). At that time, understanding of the relationships between igneous activity and the mantle was still at an early stage of development, which meant that the Rufunsa volcanic/mantle connection could not be pursued. The earlier deduction of a mantle origin for the Rufunsa volcanics was substantiated by the analyses of quenched melt droplets in vent tuffisite, which were composed of virtually iron-free, high Mn, high Sr, dolomite (Table 1) containing microphenocrysts of high Cr magnesio-chromite (comparable with spinels in deep mantle samples) (Bailey 1989). All the subaerially erupted material is fragmental, and most is agglomeratic, composed of debris from the vents and vent-walls, and earlier pyroclastic deposits. The matrix is largely very fine carbonate ash of dolomitic/ankeritic composition, heavily stained red-brown with finely disseminated iron oxide/hydroxide. Unequivocal melt droplets are composed of colourless dolomite, or dolomite plus calcite, although primary calcite contents are hard to quantify due to accidental incorporation of calcite from calcio-carbonatite intrusions penetrated by the volcanic vents. Some rocks contain drop-like fragments of iron-stained carbonate, adding to the sense (derived from the calcite distribution) that the erupted melts were variable (although the bulk composition of the deposits overall is dolomitic/ankeritic). Most of the melt droplets in thin section enclose accidental grains as cores, i.e. they are essentially small autoliths, showing that, after fragmentation, the still-fluid melt coated the entrained solids before it was quenched. Experimental data on dolomitic liquids is limited, and in the absence of obvious fluxing components (e.g. H20 or alkalis) or evidence of their presence at the moment of quenching, the required melt temperature and pressure would have to be high. Dolomitic melt would require quenching at temperatures >1000 °C, at c. 10 kbar, if data on the pure carbonate systems were applicable (Wyllie 1989). While some amelioration of these values may be anticipated from the small amounts of extra components (e.g. Sr, Mn, P) in the melt, the implication must be that melt fragmentation and quenching took place at high temperatures, deep in the volcanic vent: hence, melt droplets had become part of the entrained solids long before reaching the surface. Other components that may have conditioned the original melt are iron (now in the groundmass) and potassium (based on the ubiquitous phlogopite, and phlogopitized fragments in the pyroclastic rocks, and extensive K metasomatism) (Bailey 1989): but if present originally, K and Fe had been largely segregated before the melt droplets were formed, so that on present evidence high T, high P quenching seems inescapable. If so, it may be concluded that (a) similar melts could not exist at the surface, and that (b) melt fragmentation (and quenching) took place at great depths, with the tuffisite pipe extending possibly into the mantle. Although undoubted

257

mantle xenoliths are as yet unreported from Rufunsa, the possible deep tuffisite formation, the spinel compositions, and the high K activity, suggest analogies with kimberlites (or lamproites), especially as the mid-Zambezi-Luangwa rift has been a locus of kimberlite/lamproite activity (Bailey 1989). In terms of Sr and Nd isotopes, the Rufunsa volcanic rocks are also highly unusual for carbonatites generally, being transitional to Group II kimberlites (Zeigler 1992). Such a relationship gives special interest to the Rufunsa province as a whole, because the intrusive carbonatites are closely similar to the contemporaneous intrusions of the classic Chilwa carbonatites of Malawi (Bailey 1960).

Magnesio-carbonatite intrusions Although replacement of calcitic carbonatite by magnesiocarbonatite is commonplace in intrusive complexes, as pointed out by Campbell Smith (1956 p.202) and recently emphasized by Barker (1989), unequivocal intrusive relations abound (Campbell Smith 1956, p. 203). In the Rufunsa sub-volcanic sections, both intrusion and replacement are clearly displayed, and in common with many intrusive complexes the general sequence is calcio-, followed by magnesio-, grading into ferro-carbonatite. Later activity took the form of replacement by silica-iron hydroxide, and veining by calcite-quartz-barite-fluorite (Bailey 1960). In places the intrusive carbonatites show transitions into intrusive tuffisite, which led to the conclusion that all the pyroclastics bore this relationship to the intrusions: the discovery that the dolomitic melt lapillae were coming directly from a mantle source rules out such a general explanation. Some magnesian pyroclastic eruptions might have originated from shallow intrusions, but they could have carried melt only if the residual liquid in the intrusion was rich enough in fluxes to lower the temperature and the dissociation pressure. So far, no evidence has emerged to identify such a melt. Most magnesio-carbonatite intrusions show little evidence of very high temperatures, and the fact that they appear 'intermediate' between calcio- and ferro-carbonatite, which has links with late-stage mineralization in many complexes, means that they share the intrusion temperature problem, discussed above, of the calcitic intrusions. Furthermore, pure dolomitic liquid would introduce another complexity, due to its high dissociation pressure. For these reasons, the case for emplacement as a crystal mush, lubricated perhaps by fluid, is even stronger for magnesiocarbonatite in shallow intrusions.

The enigma of bimodal compositions Another question, rarely (if ever) raised, and certainly never properly aired, is posed by the compositional dichotomy between calcio-and magnesio-carbonatites. A traditional view, based on the common intrusion sequence, is that magnesio-carbonatites are differentiates from primary calcitic magmas, but a whole battery of questions is set in train by the relative scarcity of intermediate compositions. A bimodal composition distribution has been evident from the outset, in the ease with which an original two-fold classification was accepted, i.e. s6vite (calcitic) and rauhaugite (dolomite) (Br6gger 1921): it appears clearly in the statistical diagrams of Woolley & Kempe (1989, figs 1.1-1.4) with two peaks emerging in the histogram for CaO.

258

D . K . BAILEY

Few geologists who have worked on carbonatites would be likely to suggest that the 'gap' in distribution is a sampling artefact; indeed, it may be more likely that the bimodality of magma types has been partly blurred by intermediate samples that represent partial dolomitization or intrusion mixing. If dolomitic carbonatite succeeds calcitic carbonatite in intrusive sequences, and each was largely crystalline on final emplacement, it is hardly possible that they are in a crystallization sequence. There is no experimental evidence available with which to account for the contemporaneous development of two such contrasting magmas from a common source. May be it will be necessary to look to distinctive modes of melt generation in the mantle source. Providing a solution to this dichotomy represents a major challenge to carbonatite geology; it is time that it is clearly recognized.

Ultramafic and mantle connections An intimate connection between carbonatites and ultramatic, alkaline ultramafic and lamprophyric rocks, was perceived at the outset (Br6gger 1921) and was set out by Campbell Smith (1956), who was by then able to include examples where peridotite and pyroxenite were major parts of the association. A growing understanding of the relationships between igneous activity and the mantle, and the finding of low initial 87Sr/e'6Sr in carbonatites (see Powell et al. 1966 for review) effectively put the seal on the mantle source. At that time there was an unresolved question as to whether the ultramafic xenoliths in igneous rocks were cognate or accidental fragments of the mantle. Such xenoliths, reported from the classic complexes of Fen and Aln6 (Griffin & Kresten 1987) and in different forms in other examples, such as Catanda, Fort Portal, Polino, now provide independent evidence of a mantle source region for the activity, regardless of any subsequent roles assigned to differentiation in carbonatite genesis. There remains, however, a major area awaiting exploration, namely the connection between peridotite and carbonatite where revealed in deeply eroded complexes such as Shawa and Dorowa (Zimbabwe). As Campbell Smith (1956) was aware, an understanding of the peridotite connection is essential to our appreciation of carbonatite activity. Not a great deal of further progress has been made in field and petrographic investigation, although our background knowledge of the mantle now indicates this could be a fruitful area. Carbonatites a n d kimberlites

A link between carbonatite and kimberlite was also hinted at by Br6gger (1921) as noted by Campbell Smith (1956), who gave prominence to Daly's (1925) work on the carbonate dykes in the Premier kimberlite. Possible connections are manifold and have been suggested subsequently by many auihors (see especially, Barker 1989, pp. 54-56, for a review); indeed, yon Eckermann (1963) assigned a parental role for kimberlite, with carbonatite forming by immiscible liquid separation. Petrographically there are strong analogies between kimberlite and the typical melilite lamprophyres of the carbonatite sub-volcanic association (aln6ite/damptjernite) in the plutonic complexes. In one sense the link was made in the first description of kimberlite when Lewis (1897 in Yoder 1975)

suggested that 'melilite basalt' was a heteromorph. More recently the similarities were highlighted in Rock's (1991) proposed classification of lamprophyres, where kimberlite is effectively used as the basis of definition, and close links with other highly undersaturated lamprophyres, e.g. aln6ite are thereby emphasized. A genetic link between kimberlite and carbonatite has been robustly challenged, however, by Mitchell (1979, 1986) drawing attention to the differences in spinels and ilmenites between the two groups. Gaspar & WyUie (1984) questioned this in showing that the compositional ranges from the two groups overlap, but Wyllie (1989, p.539) still found 'Mitchell's arguments are persuasive'. An important additional point made by Mitchell is the occurrence of carbonatites (and their associated lamprophyric rocks) in alkalic complexes, in contrast with kimberlites: a view echoed by others, including Wyllie (1989, p.500, p.537) when he writes, 'carbonatites are normally found as small bodies associated with much larger volumes of silicate rocks'. This view also emerges in many generalized descriptions of carbonatite complexes in igneous petrology texts, but whether this represents the 'normal' arrangement is another matter. Of the 84 unquestionable carbonatite complexes listed by Gittins (in Tuttle & Gittins, 1966) for Africa, 54 are recorded with silicate magmas either absent or minor in amount. Five of the seven effusive calcio-carbonatites listed in Table 2 are in the same category (Catanda, Khanneshin, Polino, Fort Portal and Rufunsa). As noted before, the last four show evidence of connections with potassic magmatism or high K activity, and four of the five give evidence of a direct mantle source. In the Fort Portal area, diamond indicator minerals are reported from alluvial deposits (Barker & Nixon 1989). Isotope chemistry in the Rufunsa carbonatites is transitional to Group II kimberlites, spinels are the same as those in kimberlites, and there are no associated silicate magmas. It may be that many carbonatites form parts of sodic or sodi-potassic alkaline complexes, but others (notably with potassic attributes) emphasize the possible kimberlite connection. The matter is not academic: rather than rejecting this possibility, more research could be devoted to exploring it, and the differences between carbonatites in the sodic and potassic associations. As well as scientific value the results may have commercial application. High K activity is marked in the Rufunsa province by intense and extensive K feldspathization around the vents, and abundant phlogopite in the volcanics, leading to the view that the total material flux from the mantle was characterized especially by carbonates, K and Fe (abundant in the volcanic matrix). No sodic igneous rocks or minerals have so far been discovered. Melt inclusions in diamonds (Navon et al. 1988) show a number of compositional analogies to the estimated primary flux from the Rufunsa mantle (Bailey 1989), such that this would make a more favourable comparison than with the lamproite and Group II kimberlite chosen by Navon et al. (1988). This in turn indicates another reason for keeping open the links between carbonatites and kimberlites: perhaps the melt inclusions in the diamond are giving another sample of the melt/fluids involved in the generation of these rocks.

Mantle source and the primary flux In his preface to the most recent multi-author volume on carbonatites, Bell (1989) writes 'most contributors favour

CARBONATE MAGMAS the formation of carbonatites from differentiation of a carbonated silicate melt'. Only one strongly advocates, as a general case, carbonatite magma generation by direct melting of the mantle (Gittins 1989). Using other lines of argument, based on experimental studies, Eggler (1989) and Wyllie (1989) reach conclusions similar to each other, but different from Gittins, namely that typical carbonatites are derived from primary nephelinitic/melilititic magmas, with effectively only kimberlites representing carbonate-bearing magmas coming directly from the mantle source. Both Wyllie (1989) and Eggler (1989) concede that primary carbonatites are possible, but unlikely. For Wyllie (p.500), 'the high ratio of silicate:carbonatite in most alkalic complexes argues against this origin' while Eggler (p.575) having set requirements including high Mg number, essential alkalis and magmatic isolation, concludes 'Few, if any, carbonatites fulfil these criteria'. Even Gittins (1989) does not indicate an example where his proposed primary carbonatite has been erupted directly and he envisages (p.588) 'nephelinite and carbonatite magmas forming sequentially, or possibly simultaneously'. In essence, all three views try to relate carbonatite genesis to a generalization about the supposed ubiquity of the 'carbonatite/nephelinite' association (largely divorced from kimberlite generation) but as indicated in previous sections there are plenty of natural examples that show this perception to be too narrow. The notion of differentiation of carbonatites from alkaline ultramafic parent magmas has a long history (reviewed by Campbell Smith 1956, p.213) and the possibilities receive support from modern experimental studies, but there are good examples of direct eruption of carbonatite from a mantle source, and the latest isotopic data on Spitzkop (Harmer 1993) suggests that even carbonatite in the characteristic silicate association may not be derivative (possibly the reverse). Given that both primary and derivative options may be valid, what may be deduced about the mantle source? Alkaline magmas generally, whether associated with carbonatites or not, are characterized by high levels of volatiles and incompatible elements so that if generated from the mantle, some mechanism of enhancing the levels of mobile elements (compared to more common magma types) is necessary. In the case of alkaline ultramafic magmas, enhancement in the source mantle region is required, and because the activity is repetitive in continental interiors there is a case for pervasive metasomatism of the source mantle as a precursor to magmatism (Bailey 1972). Xenoliths of metasomatized mantle in alkaline eruptions permitted initial estimates of the minimum bulk additives needed as a precursor to high K magmatism, namely 'calcite (cc) and kalsilite (kp)' (to produce the typical metasomite, biotite clinopyroxenite) (Lloyd & Bailey 1975). Although subsequent studies allow refinement of the requirements in individual cases, the need to produce alkali clinopyroxenite from peridotite still leaves 'cc + kp' as a useful basic model (Bailey 1987). Although arrived at by an independent line of enquiry, there are close analogies with the deduced mantle flux through the Rufunsa carbonatite volcanoes, as well as with melt inclusions in diamonds, suggesting that fluids akin to ' c c + k p ' may be active in the mantle generation of alkaline and carbonate magmas. During the early 1970s there were parallel developments in experimental petrology, with Eggler (1974) showing that CO2 in ultramafic systems produced melts of

259

nephelinitic/melilititic affinities. Predictions from Wyllie & Huang (1976) and Eggler (1976) were that the first melts from the deep mantle (in the presence of CO2) would be dolomitic. Both fields, mantle metasomatism and melting at the vapour saturated mantle solidus, have seen huge developments in the intervening years, such that the concepts are now firmly entrenched, as evidenced in many contributions in Bell (1989). Haggerty (1989) has developed the notion that metasomatic preconditioning is a prerequisite of carbonate and alkali rich magmas; Jones (1989) too looks to mantle 'enriched in carbonate components and incompatible elements'; and Gittins (1989) refers to a source 'from carbonated mantle', for his primary melts. Eggler (1989, p. 561) proposes that primary dolomite-rich carbonatites would be the initial primary melts from 'phlogopite-carbonate peridotite' at pressures below those he proposes for kimberlite generation (55-60 kbar): but, as noted above, both he and Wyllie et al. (1989, 1990) consider most carbonatites to be differentiates from alkaline magmas, generated by asthenospheric circulation/plume activity. In these scenarios the lithosphere is an inert membrane, and metasomatism an accessory process (largely post-magmatic). In fact, Eggler, specifically rejects volatile flux as a possible part of carbonatite genesis, saying (p.575), 'No magmatic precursory events that decouple volatiles from more refractory peridotite elements are necessary'. All the source requirements are inherent in OIB (ocean island basalt) mantle, subject to partial melting and fractionation (p.561). Mantle with ocean island basalt characteristics is, of course, a model composition deduced from the chemistry of ocean island magmas, which include nephelinites and rare carbonatites, so that its choice as a source for continental carbonatites is perhaps unsurprising. It is an artefact of the need to explain island magmatism in ocean sectors dominated by MORB generation, and is widely considered to be the characteristic composition of deep mantle plumes penetrating the oceanic asthenosphere. Similar plumes are figured in sub-continental sections by Eggler (1989, fig. 22.6) and Wyllie (1989, fig. 20.12). Both proposals, volatile flux and plume generated magmas, require energy and materials released from the deep mantle and through the lithosphere. Volatile fluxing, however, will be dependent on lithosphere structure and dynamics, whereas deep mantle plumes should be independent of the plate that lies above. Unlike the oceans, the continents reveal abundant carbonatite activity, which can be examined more directly in the context of the structure and history of the lithosphere. As about half the known carbonatites are in Africa, the greater part of which has been a stable plate for more than 550 Ma, this forms a suitable example, and one that appropriately reconnects with Campbell Smith's (1956) title theme. Eruption ages, lithosphere structures and tectonic events, have been examined elsewhere (Bailey 1992) and can be summarized here.

Lithosphere structure and eruption ages Rifting and magmatism are largely controlled by ancient zones of weakness in the African lithosphere, this long acknowledged relationship being clearly realized in the most recent maps in Kampunzu & Lubala (1991). When the ages of post-Jurassic igneous activity are compiled, four major peaks emerge (Early Cretaceous, Late Cretaceous, Eocene-

260

D . K . BAILEY Oligocene, and Miocene-Recent); in most areas, activity was repeated at least once. The localization of activity by old lesions in the lithosphere, the plate-wide synchroneity, and the repetitions, rule out the possibility of a source in deep mantle plumes impinging at random on the base of the African lithosphere. Instead, the plate-wide igneous episodes are found to correlate with external events such as Africa/Europe collisions and a global change in plate motion directions (Bailey 1992). Igneous activity peaks and collision chronology are summarized in Fig. 3. Black & Liegeois (1993) also emphasize the lithosphere control over the location of alkaline magmatism in Africa, but they prefer to attribute the magma chemistry to earlier plume activity which enriched the mantle source. In this case the plume provides only precursor enrichment, not the thermal and mechanical driving forces of the magmatism. Once it is recognized that melt generation is contingent on tectonic stresses acting across the plate, it is evident that any such earlier plume enrichment of the source is needed only if all other processes of enrichment can be completely eliminated from consideration. Even then, repetition of activity at the same sites, and the mechanism of melt generation are left unexplained. Connections between alkaline magmatism and plume activity may continue to enjoy wide favour, but the necessity for a connection remains to be demonstrated.

60

(a) 30

500

400

3OO

100

200

Ma

(b)

100

0

Causes o f m a g m a t i s m

20 16

(c)

12

N

8

500

400

300

200

B

100

Ma Fig. 3. Radiometric ages of igneous activity compared with collision rates of the Africa/Europe closure. (a) Histogram of all igneous rock ages in Africa (Cahen et al. 1984). The broad peak around 190 Ma corresponds to the continental flood basalt (CFB) magmatism that marked the break-up of Gondwanaland. All subsequent ages refer to alkaline/kimberlite/carbonatite activity following break-up. Note the two Cretaceous peaks, the stepped profile in Tertiary to Recent rise in the histogram, and especially the low activity period (70-50 Ma) compared with the collision record (b), below. (b) Collision rates for Africa/Europe closure using the relative movement path depicted in Dewey et al. (1989, fig. lb). Note the zero closure interval compared with the lull in igneous activity across Africa between late Cretaceous and late Eocene. (c) Histogram of carbonatite ages in Africa (Woolley 1989) The difference in resolution compared with (a) is probably attributable to the longer time intervals used in the original compilation. The line O through the diagrams marks the initiation of Africa/Europe collision according to Olivet et al. (1987).

Given that episodic intraplate magmatism in Africa is a consequence of external forces acting laterally across the plate, it follows that small volume melts have been permitted to erupt by stress release in the lithosphere. The melts, or the volatiles, or the appropriate melt sources (probably all three) must be continually available (Bailey 1993). Lithosphere strain is ameliorated by the re-opening of old fissure systems (and some warping) allowing the release of volatiles and fluids from the deep mantle. Channelling focuses the volatiles from a large reservoir of underlying mantle into relatively narrow zones through the lithosphere. The ambient geotherm along the release path controls the nature, volume, and source-depth of the magmas (Bailey 1980) giving rise to the observed spectrum of kimberlite/carbonatite/alkaline, magmatism, with progressively steeper geothermal gradients, which in general reflect the pre-existing lithosphere thickness. Relationships to continental rifting, and with melt transport and style of eruption have been examined elsewhere (Bailey 1986). Mantle metasomatism is an integral part of the scheme, offering the means by which incompatible elements can build up in previously depleted mantle. Carbonatephlogopite peridotite has enjoyed considerable favour as a model mantle source for carbonatite magmatism, but it should be recognized that unless this source were primordial (which is not usually proposed) the carbonate-phlogopite components must have been added, at or below the solidus, subsequent to any previous melting event (Bailey 1986, p. 450). Within the magma genesis regime, carbonate magmas are of multiple origins; some may separate from kimberlite, some may differentiate from melilitite/nephelinite, some are certainly primary, some from mantle characterized by phlogopite, some from amphibole-bearing mantle. Carbonatites associated with syenites may be sourced in shallow upper mantle, within the felsic mineral stability ranges (Fig.

CARBONATE MAGMAS

800

I000

I

I

1200

T °C

~ 1

Depth km •

_

P kb I0

50 ~ . ~ i ! ) "" ~ ~ ; : ! ; . : . . "

~

I00

-

~ I

"

• ACCUMULATION ~' AND

?ENRICHMEN

..... ~ I

~~-~J~.

" ....

- 20

I,,--

-

30

Fig. 4. Effects of melt/fluid percolation through the mantle to the solidus, along an initial geothermal gradient G, appropriate to off-craton regions (taken from Bailey, 1987). G1 is the initial geotherm, G2 is the perturbation from crystallization as percolating melt/fluid approaches the solidus. Crystallizing phases may include amphibole, mica, felsic minerals and carbonates, enriching the mantle in incompatible elements (fine stipple). Carbonatite melts" could accumulate near the solidus (over the depth range indicated), their low viscosity and low density would facilitate separation with the possibility of carrying mantle debris (olivine, diopside, phlogopite) if erupted directly to the surface. Melt accumulation is possible until the geotherm is steepened to N. Melts would be calcitic in this P T range (Dalton & Wood 1992), and dolomitic at higher pressures (along less steep geothermal gradients in cooler lithosphere). Potentially syenitic protoliths (and magmas) could develop below the peridotite solidus(coarse stipple): nephelinite/melilititemagmas represent melts that segregate and ascend from points on the geothermal gradient above the solidus. Solidus OE, in the presence of H20 and CO 2 (limited) (Olafssen & Eggler 1983). 4). Formation of primary carbonatite, as in Table 2, is shown schematically in Fig. 4, which also indicates possible links with nephelinite and with phonolite magmatism. Campbell Smith (1956) brought into a wider geological perspective what had hitherto been a petrological curiosity, thus helping to stimulate interest in what may now be perceived as a key rock type in understanding deep Earth processes. Important new information has come, and will continue to come, from areas such as experimental petrology and isotope geochemistry, but it is worth recalling that Campbell Smith's insight was informed by petrographic experience, harnessed to the field studies of his collaborators. Future advances can also be expected through modern techniques in microanalysis, and in volcanological research, thence adding an appropriate tribute to all the early pioneers in the field of carbonate magmas. Conclusions

(1) Most carbonatites are intrusions (mainly calcic) and the controversy aroused by Brrgger's (1921) original proposal of carbonate magma persisted for 40 years. Evidence from effusive carbonatite is essential for understanding the wider aspects of carbonatite magmatism, and for identifying the most relevant applications of the results from experimental petrology. Three types of effusive carbonatite are available for this purpose: natro-, calcio- and magnesio-carbonatite. (2) Natrocarbonatite is the only present-day example of flowing magma, but this composition is still unique. The

261

melt corresponds to a low pressure cotectic composition, presenting difficulties in relating it to other carbonatites, and to explanations requiring that it is comagmatic with its associated nephelinite/phonolite. Geochemistry also distinguishes this type from others. Its status as a composition from which other carbonatites may be derived is therefore questionable. (3) Of seven examples of effusive calcio-carbonatite, at least four carry evidence of direct eruption from the upper mantle. These may be primary. Present compositions show low, or negligible, contents of fluxing components, and the mineralogy is consistent with high T eruption. In four cases, links with leucite and melilite-bearing silicate magmas are indicated. (4) In the only known example of effusive magnesiocarbonatite, there is a similar lack of evidence of fluxing components, which would indicate quenching in the deep parts of a tuflisite system extending into the mantle. Other possible links with kimberlite activity, are seen in Cr minerals, high activity of K, and in the Sr/Nd isotope chemistry. There are no silicate magmas, and all the characteristics are consistent with a primary origin. Smaller amounts of effusive calcio-carbonatite in this association add weight to the possibility that some calcio-carbonatites may be primary. (5) While many small carbonatite intrusions, e.g. cone sheets, may have been emplaced at high temperatures, large intrusions, especially large plugs, indicate emplacement at temperatures below those inferred for surface eruptions. In the absence of flux components in the sampled compositions, the long-standing question of mode of emplacement remains. Final emplacement as a mush still seems the most reasonable answer. (6) A compositional gap exists between calcio- and magnesio-carbonatites, which is not explicable in terms of fractionation (based on present experimental data). The compositional gap may reflect primary differences in sources, or in melt generation mechanisms. This remains a crucial area for further research. (7) An important magmatic association is that between carbonatites and nephelinite/melilitite, but this is not universal. Associations with syenites, or with ultramafic rocks may be equally important (and overlapping), and some carbonatites are erupted in isolation. Final eruption of carbonatite through alkaline complexes may represent the expulsion of residua from the source region. (8) Hypotheses of origirr strongly favour a relationship between nephelinite (s.l.) and carbonatite, stemming from a prevailing perception of the ubiquity of the association. Similar reasoning leads to the separation of kimberlite and carbonatite genesis. Many carbonatites are not minor parts of alkaline complexes, and to overlook this fact may create an unnecessary stumbling block to progress. More research on the ultramafic connections is vital. (9) Analogies between carbonatite magmatism and oceanic island volcanism (and thereby sub-lithosphere plumes) is put in question by the observation of plate-wide activity triggered by external events. This implies an origin linked to permissive release of an energy and materials flux from the deep mantle into the lithosphere. (10) Carbonatites are of multiple origins, reflecting different aspects of carbon activity in the mantle, and any attempt to explain all the phenomena in a single hypothesis may prove futile.

262

D.K.

One obvious prognostication, from looking at the aftermath of Campbell Smith's (1956) review, is that now, as then, new and unexpected data are still to come, to add to the questions that still remain. Meeting these challenges promises to unlock further secrets about the deep Earth. M y o w n e n j o y m e n t o f the c a r b o n a t i t e challenge is c o n t i n u a l l y q u i c k e n e d by the s h a r e d k n o w l e d g e o f friends a n d colleagues, w h o s e continuing c a m a r a d e r i e I h e r e salute. E a r l y e n c o u r a g e m e n t c a m e f r o m no less t h a n W. C a m p b e l l Smith, a n d it was a f u r t h e r pleasure that the invitation to c o n t r i b u t e o n ' C a r b o n a t e m a g m a s ' should c o m e t h r o u g h M. J. L e Bas. M y t h a n k s also to A . R. W o o l l e y , for calling m y a t t e n t i o n to the r e f e r e n c e s on the C a t a n d a volcanoes, in A n g o l a . Final text revisions b e n e f i t t e d f r o m the constructive c o m m e n t s o f D. S. S u t h e r l a n d a n d M. J. L e Bas.

References ALKHAZOV, V.Yu., ATAKISHIYEV,Z.M. & AZIMI, N.A. 1978. Geology and mineral resources of the early Quaternary Khanneshin carbonatite volcano (Southern Afghanistan). International Geology Review, 20, 281-285. BAILEY, D.K. 1960. Carbonatites of the Rufunsa valley, Feira District. Bulletin 5, Geological Survey of Northern Rhodesia. 1972. 'Uplift, rifting and magmatism in continental plates'. Journal of Earth Sciences (Leeds), 8, 225-239. 1980. Volatile flux, geotherms, and the generation of the kimberlite-carbonatite-alkaline magma spectrum. Mineralogical Magazine, 43, 695-699. 1986. Fluids, melts, flowage and styles of eruption in alkaline ultramafic magmatism. Alkaline and Alkaline-Ultrabasic Rocks and their Xenoliths. Transactions Geological Society South Africa, Special Issue, 88 (2), (for 1985), 449-457. 1987. Mantle metasomatism: perspective and prospect. In: FITroN, J.G. & UPTON, B.G.J. (eds) Alkaline Igneous Rocks. Geological Society, London, Special Publications, 30, 1-13. 1989. Carbonate melt from the mantle in the volcanoes of south-cast Zambia. Nature, 388, 415-418 (and 374). 1990. Mantle carbonatite eruptions: Crustal context and implications. Lithos, 26, 37-42. 1992. Episodic alkaline igneous activity across Africa: implications for the causes of continental break-up. In: STOREY, B.C. ALABASTER,T. & PANKHURST, R. J. (eds) Magmatism and the Causes of Continental Break-up. Geological Society, London, Special Publications, 68, 91-98. 1993. Petrogenetic implications of the timing of alkaline, carbonatite, and kimberlite activity in Africa. South African Journal of Geology, 96, 67-74. & HAMPTON, 1990. Volatiles in alkaline magmatism. Lithos, 26, 157-165. BARKER, D.S. 1989. Field relations of carbonatites. In: BELL, K. (ed.) op.cit., 38-69. 1993. Discriminating magmatic features in carbonatites: implications for the origins of Mg- and Fe-ricb carbonatites. South African Journal of Geology, 96, 131-138. -& NIXON, P.H. 1989. High-Ca, low-alkali carbonatite volcanism at Fort Portal, Uganda. Contributions Mineralogy Petrology, 103, 166-177. BELL, K. (ed.) 1989. Carbonatites: genesis and evolution. Unwin Hyman, London. -& KELLER,J. (eds) 1994. Carbonatite volcanism--Oldoinyo Lengai and the petrogenesis of natrocarbonatite. IA VCEI Proceedings in Volcanology. Springer, in press. BLACK, R. & LIEGEOIS,J.-P. 1993. Cratons, mobile belts, alkaline rocks and continental lithospheric mantle: the Pan-African testimony. Journal of the Geological Society, London, 150, 89-98. BROGGER, W.C. 1921. Die Eruptivgesteine des Kristianiagebietes. IV. Das Fengebiet in Telemark, Norwegen. Norsk Videnskapsselskapets Skrifter, I. Math Naturv Klasse, 9. CAHEN, L., SNEELING, N.J., DEEHAE, J. & VAIL, J.R. 1984. The geochronology and evolution of Africa. Clarendon, Oxford. CHAYES, F. 1942. Alkaline and carbonate intrusives near Bancroft, Ontario. Geological Society, America Bulletin, 53, 449-511. CLARKE, M.G.C. & ROBERTS, B. 1986. Carbonated melilitites and calcitized alkali carbonatites from Homa Mountain, western Kenya: A reinterpretation. Geological Magazine, 123, 683-92. -

-

-

-

-

-

-

-

-

-

-

-

-

-

BAILEY

COOPEg, A.F., GITrINS, J. & TUTrLE, O.F. 1 9 7 5 The System Na2CO3-K2CO3-CaCO 3 at 1 kilobar and its significance in Carbonatite Petrogenesis. American Journal of Science, 275, 534-560. DALTON, J.A. & WOOD, B.J. 1992. The effects of Fe/Mg ratio and pressure on carbonate stability and melt compositions in peridotite assemblages. Transactions of the American Geophysical Union, 73, 616. DALY, R.A. 1925. Carbonatite dikes of the Premier Diamond Mine, Transvaal. Journal of Geology, 33, 659-684. DAWSON, J.B. 1962. Sodium carbonate lavas from Oldoinyo Lengai, Tanganyika. Nature, 195, 1075-1076. 1964. Carbonatitic ashs in northern Tanganyika. Bulletin Volcanologique, 27, 81-92. 1989. Sodium carbonatite extrusions from Oldoinyo Lengai, Tanzania: implications for carbonatite complex genesis. In: BELL, K. (ed.) op.cit., 255-277. - - , GARSON, M.S. & ROBERTS,B. 1987. Altered former alkalic carbonatite lava from Oldoinyo Lengai, Tanzania: Inferences for calcite carbonatite lavas. Geology, 15, 765-8. DEWEY, J.F., HEEMAN, M.L., TURCO, E., HuTroN, D.H.W. & KNO'Iq', S.D. 1989. Kinematics of the western Mediterranean. In: COWARD, M.P., DIETRICH, D. & PARK, R.G. (eds) Alpine Tectonics. Geological Society, London, Special Publications, 45, 265-283. OlXEV, F., CAMPBELLSMITH, W. & BISSET, C.B. 1937. (revised 1955). The Chilwa series of southern Nyasaland. Nyasaland Geological Survey Bullletin 5. ECKERMANN, H. YON 1948. The alkaline district of Alni5 Island. Svertiges Geologiska Undersokning, Series Ca. No. 36. 1961. Contributions to the knowledge of the alkaline dikes of the Aln6 region. IV. Arkiv frr Minereralogi och Geologi, 3, 65-68. 1963. Contributions to the knowledge of the alkaline dikes of the Aln6 region. IX. Carbonatitic Kimberlite from Sundsvall. Arkiv fiir Mineralogi och Geologi, 3, 397-402. EGGLER, D.H. 1974. Effect of CO 2 on the melting of peridotite. Carnegie Institution Yearbook, 73, 215-24. 1976. Does CO 2 cause partial melting in the low-velocity layer of the mantle? Geology, 4, 787-788. 1989. Carbonatites, primary melts, and mantle dynamics. In: BELL, K. (ed.) op.cit., 561-579). FREESTONE, I.C. & HAMILTON, D.L. 1980. The role of liquid immiscibility in the genesis of carbonatites---an experimental study. Contributions to Mineralogy and Petrology, 73, 105-17. GASPAR, J. & WYEEIE, P.J. 1984. The alleged kimberlite-carbonatite relationship: evidence from ilmenite and spinel from Premier and Wesselton Mines and the Benfontein Sill, South Africa. Contributions to Mineralogy and Petrology, 85, 133-1 40. GITrlNS, J. 1989. The origin and evolution of carbonatite magmas. In: BELL, K. (ed.) op.cit., 580-600. & JAGO, B.C. 1991. Extrusive carbonatites: their origins reappraised in the light of new experimental data. Geological Magazine, 128, 301-305. & McKIE, D. 1980. Alkalic carbonatite magmas: Oldoinyo Lengai and its wider applicability. Lithos, 13, 213-215. GRIFFIN, W.L. & KRESTEN, P. 1987. Scandanavia--the carbonatite connection. In: NIxoN, P.H. (ed.) Mantle Xenoliths. John Wiley & Sons, New York, 101-106. GUEST, N.J. 1956. The volcanic activity of Oidoinyo Lengai, 1954. Tanganyika Geological Survey Records 1954, 4, 56-59. HAGGERTY, S.E. 1989. Mantle metasomes and the kinship between carbonatites and kimberlites. In: BELL, K. (ed.) op.cit., 546-560. HARMER, R.E. 1993. The petrogenetic association between carbonatite and alkaline magmatism isotopic constraints. Terra Abstracts, 3, 20. HEINRICH, E. WN. 1966. The Geology of Carbonatites. Rand McNally and Co. Chicago, USA. HOBLEY, C.W. 1918. A volcanic eruption in East Africa. Journal of the East Africa and Uganda Natural History Society, 4, 339-343. HOGARTH, D.D. 1989. Pyrochlore, apatite and amphibole: distinctive minerals in carbonatite. In: BELL, K. (ed.) op.cit., 105-148. JAGO, B.C. & GITrINS, J. 1991. The role of fluorine in carbonatite magma evolution. Nature, 349, 56-58. JAMES, T.C. 1956. Carbonatites and rift valleys in East Africa. Tanganyika Geological Survey. Unpublished report, TCi[34. JONES, A.P. 1989 Upper-mantle enrichment by kimberlitic or carbonatitic magmatism. In: BELL, K. (ed.) op.cit., 448-463. KAMPUNZU, A.B., & LUBALA, R.T. 1991. Magmatism in extensional structural settings. Springer-Verlag, Berlin. KELLER, J. 1981. Carbonatitic volcanism in the Kaiserstuhl alkaline complex: Evidence for highly fluid carbonatitic melts at the earth's suface. Journal of Volcanology and Geothermal Research, 9, 423-431. 1989. Extrusive carbonatites and their significance. In: BELL, K, (ed.) op.cit., 70-88. 1992. Alkalicarbonatites and Ca-carbonatites: similarities differences

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

CARBONATE MAGMAS

and petrological comparisons (extended abstract). 29th International Geological Congress, Kyoto 577. KJARSGAARD,B.A. & HAMILTON,D.L. 1989. The genesis of carbonatites by immiscibility. In: BELL,K. (ed.) op.cit., 388-404. KNORRINC, O. VON & DU BOlS, C.G.B. 1961. Carbonatitic lava from Fort Portal area in Western Uganda. Nature, 192, 1064-1065. KOSTERVANGROOS, A.F. & WYLLIE,P.J. 1966. Liquid Immiscibility in the System Na20-AIzO3-SiOz-CO 2 at Pressures to 1 Kilobar. American Journal of Science, 264, 234-255. LE BAS, M.J. 1989. Diversification of carbonatite. In: BELL,K. (ed.) op.cit., 428-447). LLOYD, F.E. 1985. Experimental melting and crystallisation of glassy olivine melilitites. Contributions to Mineralogy and Petrology, 90, 236-243. - & BAILEY,D.K. 1975. Light element metasomatism of the continental mantle. Physics and Chemistry of the Earth, 9, 389-416. MACDONALD, R. K, K.IARSGAARD, B.A., SKILLING, I.P., DAVIES, G.R., HAMILTON, D.L. & BLACK, S. 1993. Liquid immiscibility between trachyte and carbonate in ash flow tufts from Kenya. Contributions Mineralogy Petrology, in press. McKIE, D. 1989. Fenite. In: BowLs, D.R. (ed.) Encyclopedia oflgneoas & Metamorphic Petrology. Van Nostrand Reinhold, New York, 166-167. MARIANO, A.N. & ROEDER, P.L. 1983. Kerimasi: A neglected carbonatite volcano. Journal of Geology, 91, 449-55. MILTON, C. 1968. The 'Natro-Carbonatite Lava' of Oldoinyo Lengai, Tanzania (abstract). Geological Society America Annual Meeting, 202. - 1989. Oldoinyo Lengai natrocarbonatite lava: its history (extended abstract). 28th International Geological Congress, Washington. MITCItELL, R.H. 1979. The alleged kimberlite-carbonatite relationship: additional contrary mineralogical evidence. American Journal of Science, 279, 570-589. MITCHELL, R.J. 1986. Kimberlites: mineralogy, geochemistry, petrology. Plenum, New York. NAVON, O., HUTCHEON, I.D., ROSSMAN, G.R. & WASSERBURG,G.J. 1988. Mantle-derived fluids in diamond micro-inclusions. Nature, 335, 784-789. NIGGLI, P. 1919. Untersuchungen an Karbonat-un-chloridschmelzen. Zeitschrift fur anorganische und allgemeine chemie, 106, 126-142. OLAFSSON, M. • EGGLER, D.H. 1983. Phase relations of amphibole, amphibole-carbonate, and phlogopite-carbonate peridotite: petrologic constraints on the asthenosphere. Earth and Planetary Science Letters, 64, 305-315. OLIVET, J.'L., GOSLIN, J., BEUZART, P., UNTERNEHR, P., BONNIN, J. & CARRE, D. 1987. The break-up and dispersion of Pangea. Coedition Elf Aquitaine (Pau) and IFREMER (Brest) (Wall map, with text on reverse). PECORA, W.T. 1956. Carbonatites: a review. Geological Society of America Bulletin, 67, 1537-1556. PETERSON, T.D. & MARSH, B.D. 1986. Sodium metasomatism and mineral

263

stabilities in alkalinc ultramafic rocks: implications for the origin of sodic lavas of Oldoinyo Lengai. Los, 67 (Abstract V12B-03). POWELL, J.L., HURLEY, P.M. & FAIRBAIRN, H.W. 1966. The strontium isotopic composition and origin of carbonatites. In: TUaWLE& GITrINS, op. cit., 365--378, RICttARD, J.J. 1942. Voicanological observations in East Africa. I. Oldoinyo Lengai. The 1940-1941 eruption: Journal ot the East Africa and Uganda Natural History Society, 16, 89-108. ROCK, N.M.S. 1991. Lamprophyres. Blackie, Glasgow. SHAND, S.J. 1927. Eruptive Rocks. Thomas Murby and Co., London. SILVA, M.V.S., 1973. Estrutura volcanica-carbonatitica da Catanda (Angola). Boletim Servicos Geologia Minas (Angola), 24, 5-14. S~,IITU, W. CAMPBELL 1956. A review of some problems of African carbonatites. Quarterly Journal of the Geological Society of London, 112, 189-220. STOPPA, F. & LuvlYl, L. 1993. Mineralogy and petrology of the Polino monticellite calciocarbonatite (central Italy). Mineralogy and Petrology, 49, 213-231. TUrtLE, O.F. & GITrINS, J. (eds) 1966. The Carbonatites. Wiley, New York. WILLIAMS, R.W., GILL, J.B. & BRULAND, K.W. 1986 Ra-Th disequilibria systematics: Timescale of carbonatite magma formation at Oldoinyo Lengai volcano, Tanzania. Geochimica et Cosmochimica Acta, 5 0 , 1249-59. WOOLLEY,A.R. 1989. The spacial and temporal distribution of carbonatites. In: BELL,K. (ed.), op.cit., 15-37. - & KEMPE, D.R.C. 1989. Carbonatites: nomenclature, average chemical compositions, and element distribution. In: BELL, K. (ed.) op.cit., 1-14. - - , BARR, M.W.C., DIN, V.K., JONES, G.C., WALL, F. & WILLIAMS,C.T. 1991. Extrusive carbonatites from the Uyaynah area, United Arab Emirates. Journal of Petrology, 32, 1143-1167. WYLLIE, P.J. 1989. Origin of carbonatites: evidence from phase equilibrium studies. In: BELL, K. (ed.) op.cit., 500-545. & HtJANC, W.L. 1976. High CO2 solubilities in mantle magma. Geology, 4, 21-4. -& Tt;rrLE, O.F. 1960. The System CaO-CO2-H20 and the Origin of Carbonatites. Journal of Petrology, 1, 1-46. , BAKER, M.B. & WnZTE, B.S. 1990. Experimental boundaries for the origin and evolution of carbonatites. Lithos, 26, 3-19. YODER, H.S. Jr. 1975. Relationship of Melilite-bearing rocks to Kimberlite: a preliminary report on the system Akermanite-CO2. Physics and Chemistry of the Earth, 9, 883-894. - & TILLEY, C.E. 1962. Origin of Basalt Magmas: An Experimental Study of Natural and Synthetic Rock Systems. Journal of Petrology, 3, 342-532. ZIEGLER, P. 1992. Preliminary results of geochemistry, Sm-Nd and Rb-Sr studies of post-Karoo carbonatite complexes in South Africa. Schweizerische mineralogische und petrographische Mitteilungen, 7 2 , 135-142.

Received 22 December 1992; revised typescript accepted 15 February 1993

Addendum Since this contribution was in press, other papers have appeared relating to carbonate (and carbonate fluid activity) in the mantle, which bear directly on some of the central issues (see especially, Ionov et al. 1993, and references therein). Mantle carbonates must obviously be seen in the context of effusive carbonatites erupted directly from the mantle, and in the carbonatite ultramafic connections (Conclusions 3, 4 and 7 above). Mantle carbonate trace element signatures are reported as akin to those in crustal carbonatites, which is welcome news but must be greeted with some reservations. No fresh effusive carbonatites are used in the comparisons, nearly all the examples being carbonatite intrusives from a wide range of geological environments (with very wide-ranging trace element levels). Most authors accept the prevailing consensus (as in Bell 1989) that primary carbonatites are

unlikely, so discrepancies in trace element patterns are attributed to low P differentiation and contamination in carbonatites. Such processes undoubtedly contribute to carbonatite chemical variations, but there is the additional (and widely disregarded) factor that some variations may relate more directly to differences in carbonatite genesis (Conclusion 10). Relating the direct evidence of carbonate activity in mantle rocks to erupted carbonatites is clearly an imperative, and should provide a catalyst for more rigorous re-appraisal of the whole spectrum of erupted carbonate magmas.

Additional reference IONOV, D.A., DueuY, C., O'REILLY, S.Y., MAYA, G., KOPYLOVA,M.G., & GENSHA~, Y.S. 1993. Carbonated peridotite xenoliths from Spitsbergen: implications for trace element signature of mantle carbonate metasomatism. Earth and Planetary Science Letters, 119, 283-297.

Added November 1994.

From QJGS, 1 12, 189. A REVIEW

OF SOME

PROBLEMS

OF AFRICAN

BY WALTER CAMPBELL SMITH, C . B . E . M . C . T . D .

CARBONATITES

S C . D . M . A . , PRESIDENT

ANNIVERSARY ADDRESS DELIVERED AT THE ANNUAL GENERAL MEETING OF THE SOCIETY ON 25 APRIL, 1956 CONTENTS

Page I. II. III. IV. V. VI. VII. VIII. IX. X.

Introduction .................................................................. Mineral composition of carbonatites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Carbonatites in eastern and central Africa . . . . . . . . . . . . . . . . . . . . . . . . . . . Structure of the complexes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The problem of the m o d e of e m p l a c e m e n t . . . . . . . . . . . . . . . . . . . . . . . . . . . Associated igneous rocks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The fenitized rocks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Theories of the origin of carbonatites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . List of references . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

ls9 191 194 198 202 205 208 212 216 217

SUMMARY Carbonatites are known at m a n y places in eastern and central Africa, from near Mount Elgon in U g a n d a to Spitzkop and Palabora in north-eastern Transvaal. T h e y range in age from pre-Karroo, post-Waterberg in the south to Miocene-Pliocene in the north. Owing to this great age difference d e n u d a t i o n has revealed carbon. atites at various levels of erosion, the deepest corresponding perhaps with the carbonatites exposed at Aln6 in Sweden and Fen in Norway. The composition and structure of the carbonatites, the associated alkaline igneous rocks, and the altered country rocks known as fenites are all described. The carbon. aces are mainly pure calcite but some are ankeritic and dolomitic and, locally, sideritic and manganiferous. T h e y carry characterisbic accessory minerals, particularly pyrochlore. The associated igneous rocks are ijolites with, less frequently, nephelinesyenite and at some centres pyroxenite. Fenitization results in t h e formation of aegirine-felspar rocks, nearly pure felspar rocks and felspathic breccia. Evidence as to chemical changes involved in fenitization is n o t always consistent, b u t addition of K or Na, or both, and loss of SiO s are satisfactorily d e m o n s t r a t e d . Current theories are reviewed. The carbonatites are believed to owe their origin to concentration of carbon dioxide or of carbonatitic fluid of m a g m a t i c origin, derived perhaps from pyroxenite highly charged with volatiles, a m o n g which carbon dioxide played the most i m p o r t a n t part, associated with phosphoric acid, fluorine, water etc., a n d in which the elements niobium a n d cerium, a m o n g others, were also concentrated. I. INTRODUCTION T ~ E r o c k s n o w g e n e r a l l y s p o k e n o f a s c a r b o n a t i t e s m a y be b r i e f l y d e s c r i b e d as r o c k s w h i c h , t h o u g h in g e n e r a l m i n e r a l c o m p o s i t i o n s i m i l a r to limestones and marbles of known sedimentary origin, yet appear to b e h a v e as i n t r u s i v e r o c k s a n d a r e c l o s e l y a s s o c i a t e d w i t h a l k a l i n e i g n e o u s rocks. In composition they consist mainly of calcium carbonate with subordinate amounts of carbonates of magnesium and iron. They occur within complexes of alkaline rocks believed to be igneous and occasionally at volcanic centres. Their situation, habit and structures and their r e l a t i o n t o t h e i g n e o u s r o c k s a r e all s u c h as t o s u g g e s t t h a t t h e y h a v e b e e n b r o u g h t i n t o t h e i r p r e s e n t p o s i t i o n in a t l e a s t a " p l a s t i c " c o n d i t i o n a n d a r e in f a c t i n t r u s i v e .

Index Badcallian metamorphism, 26, 38-41, 42, 54 Badnaban dyke, 42 Bahama Banks, fissure fauna, 160 Bailey, E. B., 196 Ballantrae Ophiolite, 59 Baltica, 166, 167, 170 Banks, J., 5 Barberton greenstone belt, 18 barkhan dunes, 175-6 Barrande, J., 86 'barren beds', 101 barren intervals, 94 'Barren Mudstones', 99 Barrovian metamoprhism, 68, 78 Barrow, G., 74, 78, 223 'Barrow's zones of progressive regional metamorphism', 68 base metal production, 237, 238, 239 basinal brine theory, 243 Bather, F. A., 135 batholiths, 221,228, 230, 232, 233 Bauer, G. (Agricola, G.), 238-9 Beannach dyke, 42 Beartooth mountains, 32 Beaumont, E. de, 239, 240 Becker, G. F., 206, 211 bedforms and bedding aeolian, 175-6, 178 aqueous and subaqueous, 176-7, 178-80 related to wind waves, 178-9 'Belcraig Shale', 108 Belgian Stage, 108 Belingwe region, Zimbabwe, 17, 28 Bellispores nitidus-Reticulatisporites carnosus Zone, 118 Ben Nevis volcano, 196 Ben Vuirich, 58 granite, 47, 48, 49, 50, 63 Betrand, M., 58 Beyrichoceras Ammonoid ZOne, 110-11 Bighorn mountains, 31, 32 bimodal mafic magmatism, 29 Binneringie intrusions, 29 'biochron', 137 biostratigraphy Dinantian, 105-7, 110-21 Early Palaeozoic, 83-9 Moffat series, 93-101 time-resolution, Jurassic, see under Jurassic geochronology biozones, 136 Birimian orogen, 13, 14 Bisat, W. S., 107 bivalves, British Dinantian, 111 blind ore bodies, 244 'blocking temperatures', 46 Blue Holes caves, fissure fauna, 160 Bohemain Massif, 168 'boils' on river surface, 176 Bollandites-Bollandoceras Ammonoid Zone, 110 boninitic magmatism, Precambrian, 32, 33 Bonney, T. G., 189 Borrowdale arc, 60 Bosost massif, 76 Bou Azzer ophiolite, 12-13 Bowen, N. L., 205, 206, 207, 213, 214, 216, 221,222, 227, 228 Bowen trend of silica enrichment, 209, 210 Bowen's reaction series, 228 Brabant Massif, 168, 170 brachiopod faunas, Budleigh Salterton, 165-9, 171

Abitibi Belt, 17 Abukuma Plateau, 69 Acadian orogenic events, 70 accretionary orogens, origin, 11, 12, 13-14, 19 Achiltibuie ultramafic bodies, 40 ACID processes, 226 'acme' of an evolving species, 133, 135 acritarchs in biostratigraphic calibration, 87, 88 acuity, 145, 146 adhesion structures, 176 Adirondack Mountains, 74 aeolian bedforms and bedding, 175-6, 178 Africa carbonatites, 249-62 greenstone belts, 17 alkali metasomatism (fenitization), 249-50, 253, 256 alkaline magmatism, 251-3, 255, 256, 259, 260 Alpine Fault, New Zealand, 57 Alps, collision zones, 76, 77 Alston Block, 241-3, 244 Amassalik mobile belt, 30-1 Ameralik dykes, 26, 27, 28 Amitsoq gneisses, 26, 27, 28 ammonites in biostratigraphic calibration, 130, 131, 132-5, 137-43, 146, 147 ecosomatic modification, 138 ammonoids, British Dinantian, 110-11 analytical top-down subdivision, 131 Andr6e, K., 189 anisotropy of magnetic susceptibility, 200 anorogenic magmatism, Proterozoic, 15 Antarctica, 31, 32, 73 Antrim flood basalts, 195, 196, 197, 198, 200-1 Appalachian orogenic belts, 70, 74, 75 Applecross Formation, 45 aqueous bedforms and bedding, 176-7, 178-80 Arabian-Nubian Shield, 12, 16, 18 aragonite, solution and precipitation, 186-7 Archaean plate tectonics, 18-19 terranes, 16-18, 25-33, 246 Archaean-Proterozoic boundary, 29 Archaean-Proterozoic mafic suites, 27, 32 Archerbeck Borehole, 118 Ardgour gneiss, 44 Ardnamurchan intrusive centre, 196, 197 Ardnish pegmatites, 43 Arenig fauna, 166 Arenig Series, 86, 95 Arkell, W. J., 129, 133, 134 Armorica, 166, 167, 168 Armorican Massif, 224, 226 Arnsbergian Stage, 111 Arnsbergites falcatus Ammonoid Zone, 121 Arran Goatfell Granite, 196 Arundian Stage, British Isles, 108, 109, 110, 113, 114, 115, 116, 117, 119, 120, 121 Arunta Complex, 73, 74 Asbian Stage, British Isles, 108. 109, 114, 116, 118, 119, 120, 121 Asbian-Brigantian boundary, 109 Askrigg Block, 242, 244 assimilation of crustal rocks, 207, 210, 212 Australia, 17, 19, 20, 31, 32 Australian Platform cratons, 87 avalanching (grain flow), 177-8 Avalonia, 12, 167, 168, 170 Avon Gorge stratigraphy, 105, 107, 108 265

266

INDEX

brachiopod/coral zonation, 113-14, 121 Brady, H. B., 118 Brewster, D., 239 brick-pattern ripples, 178 Bridgend quarries, fissure fauna, 158 Bridport-Yeovil-Midford Sands, 146 Brigantian Stage, British Isles, 108, 109, 111,112, 114, 116, 117, 119, 120 Brinkmann, R., 142 Bristol Channel, Mesozoic fauna, 153, 156, 158 Bristow, W. H., 155 Britain, attachment to Gondwana, 166, 167, 170 British Isles, Dinantian stratigraphy, see Dinantian stratigraphy in the British Isles British Tertiary Province, 195, 196, 199, 230 see also North Atlantic Province Brittany, 69 Br6gger, W. C., 249, 258, 261 Brongniart, A., 84 Buch, L. yon, 128-9 Buckman, S. S., 131-4, 138, 142, 146 Budleigh Salterton Pebble Bed, fauna from, 165-71 Bulman, O. M. B., 95 Bushveld complex, 29, 31,208 Bute, Island of, 61 Cadomian Belt, France, 226, 227 Cadomian-Avalonian belt, 19 caicio-carbonatite, 253-5,256, 257,261 calcitization, 186, 187 Caledonides, 59, 60-2, 98, 167, 230 metamorphism, 43, 47, 49, 68, 75, 77, 229 Norway, 68, 77 Cambrian System, 86, 87, 144, 145 Cambrian-Ordovician boundary, 95-6, 100 Cambrian-Silurian boundary, 85 Cambridge Time-Scale, 143 Campbell Smith, W. C., 249-51,255, 256, 257, 258, 261 Canadian shield, 17, 29, 73, 74 gravity anomalies, 15-16 Canigou massif, 76 Caradoc beds, 86 carbonate magmas, 249-62 carbonatites, 249-50, 263 alkaline, 256, 259 effusive, 250-1, 253-5, 256, 257-8, 261,263 intrusive, 250-1, 256, 257-8 link with kimberlites, 258, 260, 261 mantle source and primary flux, 258-9, 261 Carboniferous Limestone fissure fauna, 153, 154, 155-60 stratigraphy, see Dinantian stratigraphy in the British Isles Carboniferous System, 85, 86, 110, 113, 144 Carn Chuinneag, 39, 43, 58 Carn Gorm pegmatites, 43 Carrock Fell intrusion, 205,212 Cashel-Lough Wheelaun intrusion, 48, 50 Catanda carbonatites, 254, 258 cave pearls (pisolites), 187, 188 cementation of carbonates, 187 'Cenozoic', 85 Cenozoic-style plate tectonic processes, 19 Centenary History o f the Society, H. B. Woodward, 8 Central Highland Division, 63 Central Metasedimentary Belt, Ontario, 13 central volcanoes, 197, 199, 200, 201 Chadian Stage, British Isles, 108-9, 110, 113, 114, 115, 116, 117, 118, 119, 120, 121 chalk, pelagic nature, 189 Challenger expedition, 185, 186, 189 Changbaishan volcano, 197 characteristic faunal horizons, 136 Charterhouse Carboniferous Limestone, 155 chemical modelling, 232

chemostratigraphy, 146 Chilas complex, 13, 17 Chile, 230 china clay deposits, 238, 245 chronostratigraphy, British Isles, 105, 108-9 Chugach Metamorphic Complex, 81 Churchill Province, 31 Circular, the, 8 Cleveland dyke, 200 climbing ripple cross-lamination structures, 176 clingani 'bands', 101 closed-system fractionation, 208-10, 222 Coastal Batholith, Peru, 226, 227, 228-9, 230-1 Coastal Range, British Columbia, 76 Code of Rules of Stratigraphical Nomenclature, 129, 130 collisional metamorphism, 75, 76 collisional orogens, origin, 11, 12-13, 14-15, 19 colonnade lava tiers, 199 Colonsay rocks, 63 columnar structures, formation, 180 complanatus 'bands', 101 'completeness of the geological record', 146 concurrent-range biozones, 136 Connemara schists, 46 conodonts, in biostratigraphic calibration, 87, 95-6, 97, 111-13 contact metamorphism, 68, 69, 222, 223, 233 continents, dispersal and growth, 59-60 convection in magma chambers, 209, 211-12, 231,232 convective fractionation, 206, 211,212 Conybeare, W. D., 85 cooling histories and mineral ages, 46-7, 48 Cooma Complex, 224, 225 Coral Brachiopod Zone, 118 coral]brachiopod zonation, British Isles, 113-14, 118, 119, 121 corals, composition changes with time, 186 Cordilleran granite magmatism, 222, 226, 228-30, 231 Cordilleran orogens, 11 Cornubian Batholith, 224 Cornubian orefields and orefluids, 237-8, 239, 243-5, 246 Coronation Supergroup, 14 Coronatum Zone, 147 Cotteswold Sands, 132-3 Courceyan Stage, British Isles, 108, 110, 111, 112, 113, 114, 115, 118, 119, 120, 121 Craven Basin, 107, 109, 110, 111, 119 Cretaceous System, 144, 145, 146 fauna, 161 critical melt fraction, 224 Cromhall Limestone Quarry, fissure fauna, 158 crustal accretion, Precambrian, 25-6, 32 'crustal accretion-differentiation superevent', 26, 38 crustal anatexis, 70 crustal assimilation, 207, 213, 216 crustal extension and metamorphism, 75-6 crustal fracturing, 230 crustal melting, 20, 69-70, 231 crustal temperature changes, causes, 73 crustal thickening and magmatism, 32, 69-70, 73 crystal fractionation, 206-7, 208-12, 214-16, 228, 229 crystal settling, 206, 207,208-11,213, 215-16, 228 crystallization ages, in dating, 47-51 crystallization in fluid inclusions, 239-40 Cuillin Hills, intrusions, 195, 196, 197, 200, 201 Cullis, C. G., 186 current ripples, 176 Cuvier, G., 84 Dabje Mountains, 77 Dalradian block, 60-1, 63 Dalradian Supergroup, geochronology, 46-51 Dana, J. D., 58 Darwin, C. R., 86, 189, 205,206 Davey, H., 239 Davidson, T., 165

INDEX Davies, A. M., 135 De La Beche, H. T., 84, 154 Dead Sea Rift Fault, 57 Deccan traps, fissure fauna, 161 'deep biotite granite', 224, 225 Degerloch Rhaetic bone bed, 155 Dehm, R., 156, 160 dehydration melting, 19, 70 Delhi orogen, 16 Derbyshire Dome, 242, 243 desiccation fractures, 179 destructive plate margins, movements caused by, 59-60 Devonian palaeogeography from pebble fauna, 169, 171 Devonian System, 85, 86, 144 dewatering structures, 179-80 Dewey, J. F., 59 Diabaig Formation, 45 diamond-bearing rocks, 77 diamonds, melt inclusions in, 258, 259 differentiation indices, 214-15 diffusion intercrystalline, 77 in magmas, 206 Dinant basin, 118 Dinantian stratigraphy in the British Isles, 105-6 biostratigraphy, 105-7, 110-21 chronostratigraphy, 108-9, 121 eustasy, 107-8, 121 seismic sequence stratigraphy, 109-10, 121 dinosaur bones, discovery, 156 'dirty window', 28-9 'disequilibrium', 72-3 dish structures, 180 diurnal inequality of tides, and bedding patterns, 177 Dixon, E. E. L., 189 Dob's Linn, 93, 97, 98, 99, 100 dolomitic carbonatite, 257, 258 Donegal, 232 Donegal Main Granite, 221 Dorset Inferior Oolite, 138, 141,142, 147 double (multiple)-diffusive convection, 212 Drumbeg ultramafic bodies, 40 Dundry Hill, 134, 153-4 dunes, 175-7, 178 Dunham, K. C., 237, 241,243 Dunham's limestone classification scheme, 189 Durdham Downs, Bristol, fissure fauna, 156

Early Palaeozoic stratigraphy, 83-9 East African Rift, 214 East Cornwall, biostratigraphy, 110 East Greenland lava flows, 196, 197, 208 Eastern Layered Series, Rhum, 207 ecosomatic modification of Jurassic ammonites, 138 Elles, G., 94-5 Elsevirian orogeny, 13 'emanative centres', 244 Emborough Quarry, fissure fauna, 157 Embry & Klovan's limestone classification scheme, 189 emplacement mechanism for carbonatites, 256 Enderby Land granulite terrane, 76 entablature lava tiers, 199 Eoparastaffella Zone, 118 'epeiric seas', 190 equilibrium, mineralogical, 72 Eras, statigraphical, 85 Eskola, P., 72, 221,222, 223 Etheridge, R., 156 Europe, Northwest, palaeogeography, 166-71 European Variscides granulite terrane, 76 event stratigraphy, 87 extraordinarius Zone, 99

267

Faeringhavn terrane, 28 Falkland Island fossils, 86 Fascipericyclus-Ammonellipsites Ammonoid Zone, 110 'fast exposure paths', 76 fault controlled sequences, 57-64 fauna, from fissures, 153-61 faunal horizons, 133, 135-43, 145, 146, 147 Feltar mass, ophiolitic assemblage, 63 Fen carbonatites, 249, 258 fenitization (metasomatism), 249-50, 253, 256 Fenner trend of iron enrichment, 109, 210 ferro-carbonatite, 256 filtration differentiation, 228 Finland, granulites, 76 Fiskenaesset-type layered complexes, 27, 28 fissure faunas, Southern England, 153-61 flood basalts (plateau basalts), North Atlantic Province, 195, 196, 197-9, 200-1 floral biostratigraphy, British Dinantian, 114-18 fluid inclusions, 185, 190, 239-40, 246 techniques and methodology, 241-5 fluid-absent melting, 231 fluorite, inclusions in, 242, 244 Folk's limestone classification scheme, 189 foraminiferal biostratigraphy, British Dinantian, 118, 121 Forfarshire, Northeast, map of, 67 Fort Portal carbonatites, 254-6, 258 forward modelling approach, 215, 216 fossil extraction techniques, 157-8, 160 fossils, importance in stratigraphy, 83-4, 85 fractional crystallization, 206-7, 208-12, 214-16, 228, 229 fracture patterns in bedforms, 179 Franciscan Complex, Calfornia, 77 Ftichsel, G. C., 83 fundamental fractures, 57, 58-9, 62 Gabilly, J., 138 Gahard Formation, 166, 169 Gaima Plateau, 197 Galapagos, volcanoes, 197 Galway granite, 47 Garabal Hill Complex, 227-9, 230 Gargano fissure fauna, 160 garnet, petrological studies, 77 Garwood, E. J., 107, 113 Geikie, A., 185, 195, 199, 201 geochronology of Scottish metamorphic complexes, 37-51 'Geological Inquiries', booklet of, 6 Geological Society, the, 5, 6 origins of the Journal, 5-8 Geoscientist, the, 8 Geraldton-Beardmore terrane, 29 'ghost stratigraphy', 221 Giant's Causeway lavas, 199 Giletti, B. J., 37, 38, 43, 46, 47 Gilluly, J., 58 Girvan, fault controlled sequence, 60 Girvan district, palaeogeography, 97, 98, 99 Glen Dessarry syenite, 43, 47, 48, 50 Glen Kyllachy granite, 48, 50, 51 Glencoe volcanoes, 196, 201 Glenelg inlier, 44 Glenfinnan area pegmatites, 43 gneiss terrane accretion models, Precambrian, 26-7 Goatfell granite, Arran, 196 gold-quartz veins, 246 'Golden Spikes', 130 Goldschmidt, V. M., 72, 223 Gondwana, 166, 167, 170 Gorgona Island komatiites, 18, 31 Gorran Haven, Cornwall, 168 Gower Peninsula Carboniferous Succession, 189 'gradational differentiation', 228

268 grain settling, 177-8 Grampian Group, 63 Grampian Highlands, 46, 50, 51 granite classification systems, 232 layering in, 228 magmatism, 70, 221-33 'Granite Series', the, 223-6, 227 granite-greenstone terranes, 17-18, 25, 26 granite rocks, composition change over Earth history, 19 granitization (partial melting), 223, 224, 225 granule ripples, 176 granulite metamorphism, 76, 77-8 granulite-gneiss terranes, Archaean, 18 graptolites, in biostratigraphic calibration, 86-7, 93-5, 96-7, 99 gravel dunes, 176, 177 gravel-bed rivers, 177 Graveyard dyke, 41, 42 gravitational crystal settling, 206, 207, 208-11,213, 215-16, 228 gravity anomalies, 15-16 Great Bear batholith, 15 Great Dyke, Zimbabwe, 29 Great Glen, 51 Great Glen Fault, 57, 58, 59, 62-4 Greenland, 28, 29, 31, 45, 196, 197, 208 Greenough, G. B., 6 'greenstone' belts, ancient, 27 greenstone terranes, 28-9, 246 greenstone-granite terranes, Precambrian, 17-18, 25, 26 Grenville orogen, 13 Grenville Province, 76 Grenvillian Belt, Labrador, 45 Grenvillian metamorphism, 44 Grenvillian Ocean, 13, 15 Grbs Armoricain, 165, 166 Gr~s de Goasquellou sandstone beds, 169, 171 Gr~s de petit May, 165-6 Gressly, A., 84 Grout, F. F., 212 Gruinard Bay, 40 Guettard, J. E., 83 guide-fossils, 128-9, 130, 131,132, 134, 136-8 Hall, J., 93 Harker, A., 69, 195-6, 205, 206, 207, 208, 212, 213, 215 Harker diagrams (variation diagrams), 208, 214 Harker index, 214 Hartville uplift, 31 Hastarian Stage, 108 Hawaii, volcanoes and lava flows, 197 heat production in the earth, 19, 20, 27 Hebridean basaltic plateaus, 195, 196, 201 Hebridean Province, 197 'hemarae', 133, 134 Hercynian Belt, Western Europe, 224 Hercynian orogeny, 85 Hibbard, C. W., 160 high-magnesium calcite, 187, 188 high-pressure metamorphic rocks, 77 high-temperature metamorphism, 69, 70 high-temperature-low-pressure metamorphism, 75-6, 77, 233 Highland Boundary Fault, 46, 49, 57-8, 60-2, 67 Highland granites, 228 Hill, A. J., 189 Himalayas, 11, 73, 75, 86 Hind, W., 107 Holkerian Stage, British Isles, 108, 109, 115, 117, 119, 120, 121 Holm, G., 95 Holwell quarry, fissure fauna, 153, 154-5, 156, 157, 161 homogenization temperature, 240 Hooke, R., 83 Horner, L., 240 Hottah island arc, 14, 15 hummocky cross-stratification, 178, 179, 180

INDEX Hutton, J., 11 hydraulics of bedforms, 176-80 hydrocarbon inclusions, 242 hydrocarbon maturity, 114 hydrocarbon reservoirs, 189 hydrothermal oilfields and ore fluids, 237-8, 245'6 ore-genetic theory, 238-45 Iapetus Ocean, 46, 98, 99, 166, 167, 170 Iceland, lava flows, 196, 197, 199, 200, 201 Imitoceras prorsum Ammoioid Zone, 110 immiscibility of liquids, 213-14, 215, 255-6 Inchbae facies, 43 index-fossils, 130 Inferior Oolite, Southern England, 132, 133, 134, 138-42, 146, 147 intercrystalline diffusion, 77 interface method of fossil extraction, 160 intrusions categories, 29 as cause of regional metamorphism, 69 Inverian metamorphism, 38, 40, 41 'inverted metamorphism', 75 ion-microprobe analysis, 50 Irish Caledonides, 230 Irish Dinantian stratigraphy, 110-21 Islay rocks, 63 isobaric cooling paths, 76 isoclinal folding, 99 isothermal decompression paths, 76 Ivorian Stage, British Isles, 108, 110 Jason Zones, 147 Jimberlana intrusions, 29 Johnny Hoe suture, 15 Jones, O. T., 58 Jormua ophiolite, 16 Journal, the, origins, 5-8 Journal des Mines, 6

Judd, J. W., 195, 201 Jukes-Brown, A. J., 189 Julianehaab batholith, 13 Jura, fissure fauna, 156 Jurassic geochronology, 129-31, 135, 147-8 biostratigraphic time-resolution, 127, 131-4, 135-7, 147 ammonites in, 130, 137-43, 146 estimates of, 143-6 polyhemeral chronology, 134-5 Jurassic Period, 86-7, 144 'juvenile' source theory, 243 K-Ar dating, 38, 40, 41, 43, 45, 47 Kaapvaal craton, 18, 19, 20 Kaapvaal shield, 18, 19 Kainozoic, 85 Kaiserstuhl lapilli, 254 Kangamiut dykes, 30 Kangmar dome, 75 kaolinization, 245 Kapuskasing terranes, 18 Karelian terrane, 14 Katwe-Kikorongo volcanic fields, 255 Kennedy, W. Q., 57, 58, 59 Kerimasi, Oldoinyo Lengai, carbonatites, 250, 251, 252, 253, 254, 255 Kermach, K., 157, 158 Ketilidian belts, Greenland, 31, 45 crust, Scotland, 63 orogen, Greenland, 13-14, 15 Keuper]Lias boundary, 155 Keweenawan rift, 13 Khanneshin carbonatites, 255,258 Kilavea volcano, 208 kimberlites, link with carbonatites, 258, 260, 261

INDEX Knoxisporites triradiatus-K, stephanephorus Zone, 115 Knoydart pegmatites, 43 'Knoydartian' metamorphism, 44 Kobberminebugt suture, 13 Kohistan arc, 13, 17 Kola suture zone, 14 Kola-Karelian orogen, 14 komatiitic magmatism, Precambrian, 28-9, 31, 32, 33 Koolau volcano, 200 Koslowski, R., 95 Kraeuselisporites hibernicus- Umbonatisporites distinctus Zone, 115 Krynine, P. D., 58 KUhne, W., 156, !60 Kun Lun orogen, 11 Kurunegala, granulite formation, 77 Kylesku gneisses, 38, 41

Lachlan Fold Belt, 224, 226 Lake District, 95, 97, 98-9, 107 Borrowdale arc, 60 Lambert, R. St. J., 37 lamination patterns in aqueous bedforms, 176, 177 Land6vennec Formations, 166, 169 Land's End mineral veins, 240 Lapworth, C., 86, 87, 93-4, 95, 96-7, 98, 99-100, 101,134 Laramie mountains, 31, 32 lateral displacement of faults, 57, 59 Laurentia, 12, 166 Laurentian platform limestones, 95 lava-flow structures, 199-200 Laxford Front zone, 54 Laxfordian metamorphic events, 27 radiometric dating, 38, 39, 40, 41, 54 layered mafic intrusions, 207, 208, 215,228 Lehmann, J. G., 83 Leny Limestone, 50 leucosome chemistry, 225 Lewisian Complex, 54 geochronology, 38-43 North West Scotland, 25, 26, 27, 29, 31, 32, 41, 45 Liassic fissure fauna, 155 lime-mud, origin, 187 limestones classification, 189-90 structure and origin, 185-91 Limpopo belt, 18, 20, 76 Lindgren, W., 241 liquid immiscibility, 213-14, 215, 255-6 lithosphere structure and eruption ages, 259-60 lithospheric extension and regional metamorphism, 77, 78 iithostratigraphic time-resolution, 146 Llandeilo age of Scottish shales, 97 Llandovery Series, 86, 87, 88 local range biozones, 136 Loch Torr an Lochain dyke, 42 Lochan a' Chairn facies, 43 London-Brabant massif, 170, 171 longitudinal (seif) dunes, 176 Louis, J., Count de Bournon, 5 low magnesian calcite, 187, 188 low-pressure-high-temperature metamorphic belts, 69 Lulefi-Kuopio suture zone, 13 Lycospora pusilla Zone, 115 Lyell, C., 11, 84-5 Lys-Caillaouas massif, 76, 77 Mackenzie dyke swarms, 15, 200, 201 marie magmatism, Precambrian, 25-33 MAGIC processes, 226 magma mingling, 228 magma mixing, 206, 207-8, 212, 215 magma-flow directions, 200 magmas and magmatism, 19 alkaline, 251-3, 255, 256, 259, 260

269

anorogenic, 15 carbonate, 249-62 granite, 221-33 mafic, Precambrian, 25-33 plutonic, 69 tholeiitic, 208-10 see also magmatic differentiation magmatic advection of heat, 70, 75, 78 magmatic differentiation, 205,208-11,212-16 early ideas, 205-7 mechanism, 207-12 modelling, 215 magnesio-carbonatite, 256, 257, 261 magnesium calcite, 187, 188 magnetostratigraphy, 88, 145, 146 Main Central Thrust System, 75 Malene metavolcanic rocks, 28 mammals, origin, 153 mantle metasomatism, 259, 260 mantle source and primary flux, 258-9, 261 mantle-plume-related magmatism, 27 Marathon dyke swarms, 15 marine bivalves in stratigraphy, 111 marine storm bedding, 178-9, 180 MASH processes, 226, 232 Massif Central, 224 Mberengwa aUochthon, 17 M'Coy, 86 medium-pressure regional metamorphism, 68, 69 melt fraction material, 224, 225 melt generation and tectonism, 260 Mendip Hills, fissure fauna, 155, 156 mesothemic boundary status, 107 'Mesozoic', 85 Mesozoic fissure fauna, Southern England, 153, 157 metal-bearing hydrothermal fluids, 245 metamorphism 'inverted', 75 related to extension, 75-6, 223 and tectonics, 71 see also geochronomogy of Scottish metamorphic complexes; regional metamorphism metasomatism alkali (fenitization), 249-50, 253, 256 mantle, 259, 260, 261 micro-probe analysis, 232 microstructural studies, 71 Mid-Carboniferous boundary, 110 Midford Sands, 132-3, 146 Midland Valley, Scotland, 58, 60, 61,111 migma-magma, 223 migmatites, 224, 225 mineral ages and cooling histories, 46-7 mineral isochron ages, 77 mineralization of Cornubian and Pennine orefields, 237-46 Minnesota River Valley terrane, 18 miospore zonation, British Dinantian, 114-18, 121 Mississippi Valley Type mineral deposits, 238, 243,244, 246 Mistassini dyke swarms, 15 Miyashiro facies series, 72 'mobile belts', 26 Moffat area, palaeogeography, 93, 94, 97-9, 101 Moine thrust, 43, 46, 58, 63 Moinian Supergroup, geochronology, 43-4, 45-6, 47, 49 Molson dyke swarms, 15 monogenetic volcanoes, 197 Moorbath, S., 37, 38 Moore, C., 153-6, 160 Morar Group, 44 Moray Firth,Old Red Sandstone displacements, 62 Morecambe Bay carbonate platform, 107 Mourne Mountain granites, 196 Mozambique belt, 12, 19 Mull, Island Of, intrusive complexes, 195, 196, 197, 198, 201

270 multiple-diffusive convection, 212 Murchison, R. I., 83, 85, 86, 129 Murospora margodentata- Rotaspora ergonulii Subzone, 116 Nagssugtoquidian mobile belt, 30, 31 Nahanni terrane, 15 Nain Province, 31 Namur basin, 118 Namurian boundary, 110 natrocarbonatite, 251-3, 261 Neoarchaediscus Zone, 118 Neptunian dykes, 157, 158, 160 Neptunist theory, 83 New England, metamorphism, 71, 73 New England Appalchians, 69 Newer Granites, 46 Newsletter, 8 Nicol, H., 185 Nockolds, S. R., 221,222, 227, 228, 229, 230, 231 noritic magmatism, Precambrian, 29-31, 32, 33 Normandy-Wessex Basin, 146, 148 North America cartons, 87 exotic terranes, 59 fissure fauna, 160 North Atlantic cratons, 26, 31 North Atlantic Province, 196, 197, 199 see also British Tertiary Province North Sea Chalks, 189 North West Europe, palaeogeography, 166-71 North West Scotland geochronology of Highlands, 37-51 Lewisian Complex, 25, 26, 27, 29, 31, 32, 41, 45 Scourie dyke swarm, 19, 27, 29, 31, 41-3 Northumberland Trough, 111, 113, 118 Norwegian Caledonides, 68, 77 Nfik gneisses, 26, 27 oceanic crust on the continent, 59 oceanic lithosphere, Archaean, 19 Oldoinyo Lengai volcano, Kerimasi, 250, 251,252, 253, 254-5 Onaman-Tashota terrane, 29 oolitic grain formation, 187-9 open-system magma chambers, 207-8 Oppel, A., 86, 94, 129-30, 133 Oppelian Zones, 129-30, 133, 134 Orbigny, A. d', 86, 129, 133 Ordovician series, North American, 86 Ordovician System, 86, 87, 145 Moffat Series, 93, 94, 95-6, 97, 98, 101 Ordovician to Devonain palaeogeography of Europe, 165-71 Ordovician-Silurian boundary, 100, 101 ore-genetic theory, 238-46 orogens, origin, 11-16, 19 orogeny and regional metamorphism, 68-9 P - T - t paths, 69-78 ostracodes, in biostratigraphic calibration, 87, 118, 167 Ottawan orogeny, 13 Outer Hebrides, 45 'outer limit' lines, 67 Oxford Clay, Peterborough, 131,138, 142-3, 147 oxide-oxide variation diagrams, 208, 209, 210, 214-15 oxygen fugacity, 209 oxygen isotope dating, 77

P - T - t paths, 69-78 'paired' metamorphic belts, 69, 81 palaeogeography of Northern Europe, 166-71 Palaeozoic, Early, stratigraphy, 83-9, 94, 95, 101 Pan-African belt, 12, 16, 19 partial melting (granitization), 223,224, 225 Payne River dyke swarms, 15 Payson ophiolite, 16 Pb-Pb dating, 38, 39, 40, 54

INDEX Pearce element ratio diagrams, 214 Pechenga Series, 14 Pecora, W. T., 249, 250, 256 Pennine orefields and orefuids, 237-8, 239, 241-4, 245, 246 Penokean orogen, 13, 15 Periods, statigraphical, 85, 86 Permian Reef Complex, 189 Permian System, 85, 144 Perotriletes tessellatus-Schulzospora campyloptera Zone, 115 Peterborough Member, 142 Phanerozoic, 143 tectonism, 11, 19 Phillipines, tectonic activity, 59 Phillips, J. A., 84, 85, 241 Philosophical Transactions, 5, 6, 7, 8 Pikwitonei granulites, 18, 73, 74 Pilton Shale Formation, 110 plane beds, 176 plate tectonic uniformitarian model, 11-21 plateau basalts (floor basalts), North Atlantic Province, 195, 196, 197-9, 200-1 Pleistocene, time-resolution, 146 Plieninger, W. H. T. yon, 155 plumbing systems, 226, 232 plutonism, 69, 222-3, 232, 259, 260 'place' in, 222-3 'time' in, 223-4 plutons, shape of, 221 Polino carbonatites, 254, 255, 258 Poll Eorna dyke, 42 polygenetic volcanoes, 197, 199 Polygnathus communis carina Conodont Zone, 110 Polygnathus inornatus Conodont Zone, 115 Polygnathus mehli Conodont Zone, 115 polyhemeral chronology, 134-5 Pongola Supergroup, 18 Port aux Basques Complex, 71 Portsoy beds, 48, 50 Precambrian crustal development, 25-37 plate tectonics, 19, 20 Preketilidian belts, Greenland, 45 Principle of Biostratigraphic Synchroneity, 128, 136, 137 Principles of Geology, 11 Proceedings, the, 7, 8 prograde metamorphism, 77 progressive regional metamorphism, 69 Proterozoic crustal development, 25, 29-31 plate tectonics, 11-20 protolith formation and Badcallian metamorphism, 38-41 Pseudopolygnathus multistriatus Conodont Zone, 115 punctuated orogeny, 58 Purtuniq ophiolite, 16 Pyrenees, 76, 231

Quarterly Journal the, 7-8 quenched dykes, 30 Quercy phosphorites, 156 radiogenic isotope dating, development, 77 radiometric dating of Scottish metamorphic complexes, 37-51 Raistrickia nigra-Triquitrites marginatus Zone, 116 ramps, 190 Ramsbottom, W. H. C., 107, 108, 111 Rb-Sr dating, 37-8, 40, 41, 44, 45, 47, 48, 49, 50, 77 Reaction Principle, 207 Read, H. H., 221,222, 227, 228, 230, 231,232, 233 regional metamorphism, 67, 68, 69, 78, 222, 223 orogeny and, 68-9 P - T - t paths, 69-74 recent advances, 74-7 retrograde metamorphism, 77 Rhaetic fissure fauna, 155, 156, 159

INDEX Rhegreanoch dyke, 41, 42 Rheic Ocean, 167, 171 Rhum, Island of, igneous complex, 196, 197, 206-7 Riley, H., 111,156 rimmed shelves, 190 ripples, sedimentary, 176, 178 Robinson, P., 157-8 rock-time duality, 127-8, 135 role of fault, 62 'room (space) problem', the 221-2, 232 Rossendale Millstone Grit, 107 Royal Society, 5 Ruedemann, R., 94 Rufunso carbonatites, 254, 257, 258, 259 Rule of Priority, 129, 130 Russian Platform, 87, 98 Ryoke Metamorphic Belt, 81 Sahara, collisional orogen, 13 St Austell mineral veins, 238, 240 Saint Barth616my massif, 76 St Malo Migmatite Belt, 224, 226, 227 St Michael's Mount mineral veins, 240, 244 Salter, J. W., 96, 165 San Andreas Fault, 57, 59 San-yo granitoids, 81 Sanbagawa Metamorphic Belt, 81 sand ripples, 176 sand waves, 177, 180 Sandford Lane Fossil Bed, 134 Sawkins, F. J., 241-2, 243-4 Scaliognathus anchoralis Conodont Zone, 110 Scandinavian succession, 98 Schopfites claviger-Auroraspora macra Zone, 115 Scotland, 63 metamorphic complexes, geochronology, 37-51 Southeastern Highlands, regional metamorphism, 67, 68, 74, 78 Southern Uplands, 58-64, 93-4, 97, 98-9, 101 see also North West Scotland Scourian (Badcallian) metamorphism, 27, 38-41 radiometric dating, 38-43, 54 Scourie dyke swarms, 27, 29, 31, 41-3 sea-level changes, stratigraphy related to, 87, 88-9 secular biochronological resolution, 137, 146 secular resolving power, 137, 145 Sedgwick, A., 83, 85, 86 sediment drifts, 176 sediment waves, 176 sedimentary structures, Sorby and the last decade, 175-80 sedimentation and faulting, 58 sequence stratigraphy, 101, 190 series, stratigraphical, 86, 87 Sgurr Breac pegmatites, 43 Sharyzhalgay complex, 72 sheet-like pillar structures, formation, 180 shells, in formation of limestone, 186-7 Shelveian event, 86 Sherborne Building Stone, 134 Sherborne Inferior Oolite, 133, 134 SHRIMP, 50, 51,232 Silesian Subsystem, 109 Silurian series, establishment of, 86, 87, 93 Silurian System, 85, 86, 87, 145 Moffat Series, 93, 94, 97, 98 Siphonodella crenulata Conodont Zone, 115 Siphonodella sandbergi Conodont Zone, 110 Skaergaard intrusion, 196, 207, 208, 209-11,214, 216 skeletal disintegration as source of carbonate, 187 Skye, Island of, lava flows, 195, 196, 197, 199, 200, 201 Slave Province, Canada, 14, 17, 18 Slickstones Quarry, 157 Sm-Nd dating, 39-40, 41, 42-3, 44, 47, 77 Smith, W., 83-4, 128-9, 153 soft sediment deformation, 179-80

solidification index, 214 Solomon, M., 242 Solway line, 58, 60 Sorby, H. C., 175, 178, 185-9, 191,237, 239-41,245 Soret coefficient, 212 Soret diffusion, 206, 212, 213 South America, dykes, 31, 32 South Australian noritic dyke swarms, 32 South East Greenland dyke swarms, 29, 30-1 South Harris complex, 39 South Kola belt, 14 South Tibetan detachment system, 75 South Wales Lower Limestone Shales, 110 South West Greenland, 26, 27-30, 31 Southern Brittany Migmatite Belt, 73, 74 Southern Uplands fault, Scotland, 58, 60 Sowerby, 85 Spelaeotriletes balteatus-Rugospora polyptycha Zone, 115 Spelaeotriletes pretiosus-Raistrickia clavata Zone, 115 Spitzkop carbonatites, 255,259 spring-neap cycle bedding patterns, 177 Spurr, J. E., 241 Staffa lavas, 199 Stages, stratigraphical, 86, 87, 130, 131 d'Orbigny, 130 Rule of Priority in Naming, 129 standard chronostratigraphic units, 129, 130-1,136, 145 'standard geological column', the, 128, 143 standard time-ordered succession, 127 star dunes, 176 Steno, M., 83 Steno's Principle of Superposition, 127 Stensio, E., 95 Stillwater intrusion, 29, 31,208 Stiperstones Quartzite, Shropshire, 166 Stoer Formation, 45 Stonesfield Slate, Oxfordshire, 153 storm bedding, 178-9, 180 Strathan dyke, 42 Strathmore syncline, 61 stratigraphical horizons, 136 Stratigraphical Nomenclature, Code of, 129, 130 stratigraphy, Early Palaeozoic, 83-9 Strichen granite, 48, 51 strike-slip faulting, 57, 59 Stutchbury, S., 156, 162 subaqueous dunes, 178 subduction geotherms, decrease in, 19, 21 subduction zone metamorphism, 76-7 Subzones, stratigraphical, 116, 130, 131, 145 Sudbury dyke swarms, 15 supercontinent, Proterozoic, 15 Superior Province, Canada, 17, 18, 29 suspect terranes, 59 Sutton, J., 25, 38 Svecofennian orogen, 13, 15, 16 swaley bedding, 179, 180 'syn-rift megasequence', 110 Synchroneity, Principle of, 136, 137 Systems, stratigraphical, 85-6, 100-1 Tarfside Culmination, 47 Tayvallich volcanic sequence, 50 Teall, 67 tectonic control on sedimentation, 58 tectonic processes of magmatism, 230, 232 and metamorphism, 77-8 tectonic transfer of heat, 75 tectonism and melt generation, 260 temporal scope of an analysis, 145 Tertiary System, 85, 145 textural analysis, 71 textural modelling, 232

271

272

INDEX

Theory of Earth, 11 thermal modelling of orogenic belts, 73 thermobarometric measurements, 77 thermogravitational diffusion, 212-13, 215 thermometamorphism, 67, 68 tholeiitic magmas, differentiation in, 208-10 tholeiitic magmatism, Precambrian, 30 Thompson, A. B., 72 throw of a fault, 59, 60, 62-3 Tibetan sedimentary sequence, 75 tidal bedding, 180 Tien Shan orogen, 20 Tilley, C. E., 67-8 time-correlations, 127, 128-9, 147 time-duration, 137, 143, 144-5 time-interval, 137, 145 time-markers, 128 time-planes, 128 time-resolution, biostratigraphic, see under Jurassic geochronology time-rock duality, 127-8, 135 time-scale of sedimentological events, 176-7 time-temperature trajectories, 47 Tornio-Koillismaa intrusions, 29 Tornquist Sea, 98, 166, 167 Torridonian sandstones, 43, 44-6 total range biozone-assemblage, 136 Tournaisian/Vis6an boundary, 113 Transactions, the, 6, 7 'transient', 135 transverse dunes, 175-6 Traonliors Formation, 169, 171 Tremadoc Series, 86 Triassic System, 85, 144 fissure fauna, 157, 158, 160 palaeogeography, 166-71 trilobites in biostratigraphic calibration, 118-19, 120, 121 Tripartites distinctus-Murospora parthenopia Subzone, 116 Tripartites vetustus-Rotaspora fracta Zone, 117-18 Trois Seigneurs massif, 76, 224, 225,227, 230, 231 Trueman, A. E., 135 Turner, F. J., 223 Twenhofel, W. H., 189 Tytherington Quarry fissure fauna, 158, 159 U-Pb dating, 37, 38-9, 40, 41, 42-3, 45, 48, 49-50, 51, 54, 77 Uchi-Sachigo terranes, 17 uniformitarianism, plate tectonic model, 11-21, 25 uhitary association biozone, 136 upper-stage plane beds, 176

Vallatisporites verrucosus-Retusotriletes incohatus Zone, 115 Vallis Vale, fissure fauna, 154, 156 vapour-liquid ratios, Sorby, 240 variation diagrams, 208, 209, 210, 214-5 Variscan belt, 74 Variscan massifs, 76

Vaughan, A., 105-7, 108, 113, 118 Ventersdorp rift system, 19 Verneuil, M. E., 85, 86 vertebrate fissure faunas, Southern England, 153-61 Vicary, V., 165 Vis6an Stage, 108, 113, 114, 115, 121 Vis6an/Namurian boundary, 118 volatile fluxing, 259 volcanic-hosted massive sulphide deposits, 246 volcanology, British, classic period of, 195-6 Waagen, W., 129-30 Wabigoon terrane, 17 Wales, 58, 143 wall rock assimilation models, 207 Walls Boundary Fault, 62 Ward, D. J., 160 Watson, J., 25, 38 wave-related bedforms, 178-9, 180 Wawa-Abitibi terrane, 17 Weardale granite, 243, 245 Welsh Basin, 87 Werner, A. G., 83 West African craton, 12-13 West Greenland granulite-gneiss terranes, 18 lave flows, 196, 197 Westbury-sub-Mendip fissure fauna, 160 Western Alp blueschist belts, 77 whole-rock ages, 49, 77 determination of, 39-41 'Wilson cycle', 13, 29 Wilsonian cycle of megacontinent growth, 59 wind waves, bedforms related to, 178 Windsor Hill, Shepton Mallet, fissure fauna, 156, 157 Witham, H., 185 Witwatersrand Supergroup, 18-19 Woodward, H. B., 8, 134-5 Wopmay orogenic belt, 14-15, 73 Wyoming craton, 31, 32 Yangtze Platform cratons, 87 Yeovil Sands, 132-3, 146 Yorkshire Dales, 107 Zambian volcanic carbonatites, 253-4, 255 zibar ripples, 176 Zimbabwean craton, 17, 18, 20 zircon grain analysis, 45, 49, 50, 51,232 zonal mapping, 68 Zones, stratigraphicai, 86-7, 94, 130, 131,133, 145, 147 Opellian, 129-30 Zonules, 130 zoogeographical provincialism, British Isles, 107